From burek021 at gmail.com Mon Oct 1 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Mon, 1 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20120930 Message-ID: <20121001000501.DB4C318A01DE@apolo.teamnet.rs> [00:02] is there a way to batch files up? ffmpeg -i * -vcodec libx264 *.avi ? and it will output the same filename with *.avi [00:15] for f in *.wmv; do ffmpeg -i $f -c:v libx264 ${f%.wmv}.mp4; done [00:15] or something like this [00:15] there is no builtin for what you are looking for [02:13] why is there no -probesize argument on my windows install of ffplay? [02:13] wtf [03:03] How do you find out the pxl_fmt of a certain usb camera? Is there a way to probe it? [03:07] Lns<< v4l2-ctl --all [03:20] ty creep [03:53] Hello, in order to synchronize my live stream, should i map audio to video or video to audio? [04:09] i compile ffmpeg git on debian testing linux.. anyone know how to make it compile using ccache? [06:42] hotwings: put /usr/lib/ccache/ at the front of your PATH [06:56] grepper - thanks for the help [07:09] np [08:36] Hello? [08:36] what is RTV1? [08:38] anyway, I have a video that I want to resize and then upload to youtube [08:40] it's currently 2440x540 mjpeg, but I want it to be 1920x360 before I upload [08:41] when I simply ran -s 1920x360, it recompressed it with a really low quality mpeg4 [08:41] what settings should I use for uploading to youtube? [08:43] eh? [09:26] I tried ffmpeg 1.0 with x264-snapshot-20120929-2245-stable and encoding become very slow, anyone experience this? version 0.11.1 works fine [09:32] is anyone around? [12:05] Hello there :) [12:05] Does FFmpeg support MPI? [12:05] So I can share the load between multiple machines? [12:12] no [12:13] just run multiple ffmpegs [12:14] damn [16:00] Hey folks, I am trying "ffmpeg -i GOPR%04d.JPG -b 3000 -r 12 video.mp4" but I get a "No such file or dir" & I presume because the filename pattern does not start at 1 (or 0?) is there a way to specify the sequence start? [16:04] Found in the docs the blobbing pattern support "ffmpeg -f image2 -pattern_type glob -i 'GOPR*.JPG' -b 3000 -r 12 video.mp4" [16:05] but seems my ffmpeg (that I just compiled with brew) doesn't know about it :-( ffmpeg version 0.11.2 [16:05] you need -f image2 before -i [16:05] otherwise the globbing won't work [16:05] btw, we are in ffmpeg 1.0 [16:07] @ubitux: urgh & I am shocked that the homebrew version is that behind & let me do some digging [16:08] @ubitux: is -pattern_type glob 1.0 only? [16:08] you don't need that option for your need afaict [16:10] hm & but with "ffmpeg -f image2 -i GOPR%04d.JPG -b 3000 -r 12 video.mp4" I am still getting the "no such file" while clearly the first file in the dir is GOPR0685.JPG [16:11] ah, strange [16:11] well then try to upgrade yeah [16:14] grrr & waiting for pull request https://github.com/mxcl/homebrew/pull/15187 [16:14] :) [16:19] locally hacked the formula & now I feel dirty [16:25] @ubitux: yay & 1.0 seems to work fine with the globbing [16:26] thx! [16:27] np :) [16:30] @ubitux: maybe one more question & it prints "frame= 361 fps=1.3 q=69.0 size= 1702kB time=00:00:26.00 bitrate= 536.3kbits/s dup=0 drop=388" [16:30] what is the drop? [16:30] is that OK? [16:32] shoudn't matter with still images input, but i may be wrong [16:32] you'll see if the output is correct [16:49] @ubitux: wait & -b 3000 means bitrate of 3kb/s? I wanted a little more :) [16:50] but why does it say bitrate= 597.3kbits/s [16:50] that's what is actually found? [16:50] needed? [16:52] @ubitux: crap - it's 13MB in size and all just black?! [16:59] -b 3000k [17:00] relaxed: yeah & running that right now & but it still should not have been all black [17:01] maybe the "drop" really meant that all frames where dropped & for whatever reason [17:01] pastebin your command and output for more help [17:03] hi guys, I looking to get ffmpeg to encode video using dirac (libschroedinger). I am testing with a pre-build windows build of ffmpeg that has --enable-libschroeding set. ffmpeg reports "Unsupported codec id in stream "Unsupported codec id in stream 0" [17:03] relaxed: https://gist.github.com/3248b91eab9ed33bfd69 [17:05] specify the frame rate before the input [17:05] I believe the default is 25fps and your output is 12, thus frame are dropped [17:08] i'm trying to specify a start position for an input video and it's not working. i've tried ffmpeg -i video.mov -ss 15 bla bla and it gets ignored and always starts at the beginning. i also tried -ss 00:00:15 to no avail [17:08] anyone know what im doing wrong? [17:09] i also tried putting -ss before -i [17:09] http://pastebin.com/9WbUBPpj [17:09] relaxed: nice & using "ffmpeg -r 12 -b:v 3000k -f image2 -pattern_type glob -i 'GOPR*.JPG' ../video.mp4" stdout looks much saner now [17:09] if i change "dirac" to "theora" it creates a video file ok [17:10] wait nvm i figured it out [17:10] lol [17:10] thx anyway [17:12] jafa: ffmpeg -codecs 2>&1| grep -i dirac [17:12] DEV.LS dirac Dirac (decoders: dirac libschroedinger ) (encoders: libschroedinger ) [17:13] jafa: add -pix_fmt yuv420p [17:14] If I have a mpg with mp2 audio, what is the correct way when I want to put a delay on the audio, because in the original the audio is async to the video. [17:14] relaxed: it reported yuv420p for the output stream, then same error [17:14] Stream #0:0: Video: dirac, yuv420p, 960x720, q=2-31, 448 kb/s, 10 tbn, 10 tbc [17:14] remove -r 10 [17:15] changed to 25fps, same error [17:15] (removed, reported as 25fps) [17:17] Stream #0:0: Video: dirac, yuv420p, 960x720, q=2-31, 448 kb/s, 25 tbn, 25 tbc [17:17] -f mpegts output.ts [17:18] that worked [17:18] can't pacakge dirac as ogg? [17:19] yeah, this may be a ffmpeg bug but I'm not sure. [17:20] ok, might be able to remux as a separate step [17:20] thanks [17:22] interesting, audio now works as well [17:22] source file is a opus ogg file, using -acodec copy [17:22] was getting the same error with outputting to a ogg container [19:43] hi guys. im tring to do a project using various tools and at hte moment im trying to use a camera tracker software whjich requires the clip to be in frames. say i have a h.264 avi/mov clip that is 24/25/29.blah how can i seperate that into the frames maybe PNG without any loss [19:45] sine_: ffmpeg -i input %04d.png [19:46] Would give you 0001.png, 0002.png ... [19:46] how will it know what FPS to use [19:46] or would it not matter [19:46] because it will just be the frames and the fps is something i need to tell the other softweare [19:47] That command will simply output the frames. [19:49] Using ffprobe you could save the framerate somewhere. [19:53] how can i do that to a specific point in a movie. so dump the frames of 10 seconds half way through a film etc [19:54] ffmpeg -ss 00:30:00 -i input -t 10 %04d.png [19:55] (30 minutes in, 10 seconds of frames) [20:01] thanks [20:02] however the first 10 frames are the same frame [20:03] try, ffmpeg -i input -ss 00:30:00 -t 10 %04d.png [20:27] Hey :-) [20:28] thanks relaxed [20:52] hey relaxed ! [20:52] ffmpeg -i test.ts input -acodec aac -ab 128kb -vcodec mpeg4 -b 1200kb -mbd 2 -flags +4mv+trell -aic 2 -cmp 2 -subcmp 2 -s 320x180 -title "AD Nat Geo - Chimp Diaries_ Ep 5" final_video.mp4 [20:52] give me error: [20:52] Unable to find a suitable output format for 'input' [21:05] anyone please? [21:07] cbreak: :P [21:07] DelphiWorld: dont ping random people :) [21:07] Action: DelphiWorld slaps cbsrobot_ around a bit with a large trout [21:07] cbsrobot_: was you, sory cbreak :P [21:07] DelphiWorld: get a mac [21:07] HAHA [21:08] cbreak: i do have my dude... issue is i'm a beginer on it while i am a blind person [21:08] transition from windows accessibility to mac accessibility cant be easy [21:08] then don't get an iMac. Worthless big screen :/ [21:09] :) [21:09] cbreak, i have a mbp [21:09] ffmpeg -y -i test.ts -vcodec libx264 -acodec libfaac -ab 128k -ac 2 -b 640k -threads 4 -flags +loop -cmp +chroma -partitions 0 -me epzs -subq 1 -trellis 0 -refs 1 -coder 0 -me_range 16 -g 300 -keyint_min 25 -sc_threshold 40 -i_qfactor 0.71 -maxrate 10M -bufsize 10M -rc_eq \?blurCplx^(1-qComp)\? -qcomp 0.6 -qmin 10 -qmax 51 -qdiff 4 -level 30 out.mp4 [21:09] that's my line now [21:09] but fail [21:10] DelphiWorld: it still fails after what I've told you ? [21:12] cbsrobot_: pastebin? [21:12] sure [21:14] DelphiWorld: are you encoding for a special device or just following a random old tutorial from the inet ? [21:20] cbsrobot_: old tuto [21:20] cbsrobot_: i'm encoding for my iPhone [21:20] which iphone? [21:20] 3GS or newer? [21:20] or older? [21:20] JEEB: 4S -) [21:20] ok [21:20] so newer than 3GS [21:21] yes [21:21] i saw lot of gui apps but i hate them. i prefer cmd [21:21] and I guess you will be uploading stuff via itunes? [21:21] so it has to be "itunes-compatible" [21:21] JEEB: true [21:22] ffmpeg -i derp.input -c:v libx264 -profile:v main -level 30 -x264opts ref=3 -crf 23 -maxrate 10M -bufsize 10M -c:a libfaac -ab 128k out.mp4 [21:22] test this [21:25] JEEB, -c:v didnt recognise it. should i remove -c ? [21:25] umm [21:25] sounds like you've got an old ffmpeg :) [21:25] it's a newer way of saying vcodec [21:25] shit [21:25] hold JEEB;) [21:25] JEEB: apt-get i should fuck it! [21:26] always apt-get upgrade mine to the older one, so that's a downgrade:-P [21:26] apt-get lol [21:26] cbsrobot_: otherwise yum ? [21:26] if you're on a relatively new ubuntu or debian, it's probably a libav package in reality, the newer command there is avconv. Just noting this. [21:26] of course if you've got a newer ffmpeg there as well, feel free to use that too [21:27] JEEB: i'm setting up the new deb package built by myself. [21:27] k [21:27] and JEEB thank you the big Geek for reminding me [21:30] updated [21:32] JEEB: ok, the new unmissable travel is begining;-) [21:34] JEEB: it's converting now [22:02] convertion take time right JEEB ? [22:03] yes [22:34] JEEB: working dude [22:34] vlc can play it, i'lle try iTunes [22:52] hi guys, is there a way to get ffmpeg to invoke "convert" (imagemagick) after decoding but before encoding a frame? [22:52] (or a similar custom image processing command) [22:57] oooh - just found smartblur - that might do it [23:31] night all [23:31] cbreak cbsrobot_ JEEB thank you DUDES [23:31] I did something? [23:31] well, night :) [23:32] DelphiWorld: np and cu soon [23:32] :P [23:32] cbreak: you talked and assisted. thanking is free, so thank ;-) [23:33] cbreak: maybe it's because youre from ch too [00:00] --- Mon Oct 1 2012 From burek021 at gmail.com Mon Oct 1 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Mon, 1 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20120930 Message-ID: <20121001000502.E017618A01E8@apolo.teamnet.rs> [00:01] Please read and improve! [00:02] Action: beastd is not that happy yet [00:05] I hope I got the most important stuff down in a clear enough fashion. also the whole document could be a bit more fun ;) But that is over me tonight. [00:14] [FFmpeg] michaelni pushed 2 new commits to master: http://git.io/ZwQ2fA [00:14] [FFmpeg/master] doc/fate: Move fate config example into doc subdirectory - Alexander Strasser [00:14] [FFmpeg/master] cyuv: implement raw cyuv - Michael Niedermayer [00:25] beastd: make: *** No rule to make target `doc/../tests/fate_config.sh.template', needed by `doc/fate.html'. Stop. [00:26] :( [00:26] mmh maybe an old dep. [00:26] ubitux: seems so [00:26] there is no such path in the new version [00:27] yeah my bad [00:32] beastd: i don't like the term "bitrod" and I'll be glad if you avoid it. Would you please just replace the sentence so that is says "keep their code in working condition as new versions...." etc. [00:37] iive: would "keeping their code in shape as new version..." also sound good to you? or is it too informal? [00:39] imho, working condition resonates more with the idea of maintenance [00:42] iive: OK, thanks. Done. [00:47] michaelni: http://b.pkh.me/0001-swscale-fix-To-Y-UV-extern-protoypes.patch [00:48] does this make any sense? [00:48] it's fixing all these warnings for me: http://pastie.org/4867868 [00:49] it's based on the prototypes from yuy2ToY_c and yuy2ToUV_c from sws/input.c [00:49] it is the first time i see "she" used as general term. I think it is kind of standard to use "they" even as singular). Not really that important. [00:49] and the pointer types [00:50] iive: might need further grammatical changes to that sentence. so postponing it for now [00:51] something else. It may be good idea to mentions somewhere that the position doesn't give you any additional (super) powers, only more work to do. [00:52] (typo protoypes/prototypes fixed) [00:53] btw, these warnings look weird: [00:53] libswscale/swscale.c:587:31: warning: assignment from incompatible pointer type [enabled by default] [00:53] libswscale/swscale.c:588:31: warning: assignment from incompatible pointer type [enabled by default] [00:53] libswscale/swscale.c:619:35: warning: assignment from incompatible pointer type [enabled by default] [00:54] mix of int16/int32 [00:55] iive: agreed. but please don't be upset if i won't add it soon. [00:56] beastd: no problem, I don't see good spot to put it atm. and the (super) powers thing is a joke... in case it is not obvious. [00:56] :) [00:56] and yeah you are right the document isn't as complete as i wish it to be (certain aspects that deserve mention are missing completely) [01:13] i am leaving. thank you for the help. good night... [09:52] ubitux: should I push the two remaining ffprobe patches? [10:26] saste: maybe explicit that SECTION_FLAG_HAS_VARIABLE_FIELDS requires the element_name field, but anyway should be ok [10:26] though, "variable fields" isn't really explicit but well [10:26] the other patch is ok [10:27] ubitux: variable_keys? [10:27] ubitux: variable_field_keys? [10:27] i have a problem with "variable" actually :p [10:27] variable = nonfixed [10:28] i can't tell at first what's variable [10:28] if it's the number of entries or anything [10:28] it's the number and the keys of the entries [10:28] it is describing a "tags" section [10:29] you don't know how many tags, you don't know which are the keys [10:29] you should explicit that in the doxy then :) [10:29] yes [10:29] because the name isn't that obvious imo [10:30] i'll keep the name and extend the doxy [10:30] okay [10:30] so well, both ok :) [10:31] oh Nicolas finally pushed the av_opt_set_from_string patch [10:31] \o/ [10:32] i don't remember, you were waiting for this for what purpose? [10:32] a lot of purposes :) [10:32] :D [10:32] you're gonna change the world now then [10:32] so we can drop the lame sscanf in many filters [10:33] and avoid to keep ad-hoc backward compatibility checks [10:33] ok :) [10:33] it also should simplify some stuff, for example when you add a new param in a filter which was supporting only one argument [10:33] for example the volume filter [10:34] ah right, libav changed the syntax to the key=value form [10:34] (which BTW they rewrote in libav) [10:34] would be nice to support both [10:34] i feel the dB = deciBel thing a bit silly [10:35] ubitux, no the best thing is to pull the libav filter [10:36] they did some nice improvements (like optimizations) [10:36] i meant supporting both syntax [10:36] like volume=value:... and volume=volume=value:... [10:36] Bel :D [10:36] yes [10:37] volume=+10dB:precision=... [10:37] is acceptable once you make use of av_opt_set_from_string() [10:37] same as: volume=+10dB [10:37] so we save backward compatibility [10:37] yup [10:37] and peace on earth [11:23] [~/src/ffmpeg/libavfilter]- make [11:23] rm -f [11:23] this is pretty funny. [11:44] cbsrobot_: the sound of your sample is pretty awesome :) [12:03] ubitux: how is it going with your EBU R.128 filter? [12:04] i'm trying to fix a small bias with the results from another app cbsrobot_ provided me [12:04] i've fixed locally what you pointed me out [12:04] it should be applied this week [12:04] maybe tonight if i suceed in fixing the problem [12:05] ubitux, great [12:06] did you receive any feedback relating to the CGA fonts thing? [12:06] anyway I think that's safe to apply [12:07] no, no feedback about the cga thing [13:30] [FFmpeg] michaelni pushed 4 new commits to master: http://git.io/RjLJHA [13:30] [FFmpeg/master] opt: implement av_opt_set_from_string(). - Nicolas George [13:30] [FFmpeg/master] lavu/opt: cosmetic fixes forgotten in the previous patch. - Nicolas George [13:30] [FFmpeg/master] ffprobe: generalize nesting model for the XML writer - Stefano Sabatini [14:25] 14:03:53 < twnqx> Starting with 1.7.7 MakeMKV for Windows and Mac OS comes with a (patched) copy of ffmpeg. The biggest difference between MakeMKV's mmffmpeg and a stock version, is a FLAC encoder that handles 24-bit audio. If you need this functionality, the full mmffmpeg source code is available at http://www.makemkv.com/download/ffmpeg . [14:25] i confirm the patch includes 24-bit flac encoding [14:26] it seems to make ffmpeg sws dependency optional as well [14:27] not sure if that's a good idea though [14:39] saste: would it make sense to add a timing format following hh:mm:ss.xxx instead of just ss.xxx? [14:39] +function [14:39] (or change av_ts_make_time_string) [14:40] ubitux, where do you need it? [14:40] we have code that does that in several places (ffprobe.c amongst the others) [14:41] i added the pts time in the av_log output of ebur128 filter [14:41] and i though a more expressive timing information would be nice [14:42] (where in ffprobe?) [14:42] -pretty [14:43] oh, right. [14:43] then maybe a av_ts_make_pretty_time_string would be nice [14:44] (along with av_ts2prettytimestr) [14:44] i would reduce the precision though [14:45] something like %02d:%02d:%02d.%03d [14:45] (easy to align, and re-use as cmd line) [15:10] michaelni: so if get_bits_long is used with n>32 other bits are going to be lost? [15:14] durandal_1707, yes or worse ... [15:15] get_bits_long could be changed to int64 but its not needed for 99% of the code so feels a bit odd [15:23] in 24 bit case i prefer to not do extra copy, same is already done for tta [15:26] [FFmpeg] michaelni pushed 18 new commits to master: http://git.io/frLsCA [15:26] [FFmpeg/master] indeo3: fix out of cell write. - Anton Khirnov [15:26] [FFmpeg/master] indeo5: check tile size in decode_mb_info(). - Michael Niedermayer [15:26] [FFmpeg/master] ivi_common: make ff_ivi_process_empty_tile() static. - Anton Khirnov [15:27] durandal_1707, agree but i think the allocated buffer pointer should be kept seperate [15:27] because now IIRC if its assigned to frame and then something goes wrong (return -1 whatever) [15:28] the wrong ptr would be freed [15:41] it seems the threading problem is not arch specific (the 2 threads arm instance failed) [15:42] just once: http://fate.ffmpeg.org/history.cgi?slot=armv7l-panda-gcc4.6-cortexa8-threads2 :) [15:42] yes, i saw it [15:43] its a pitty that its not readily reproduceable otherwise it would be relatively easy tp debug [15:51] michaelni: at some point, maybe some gdb scripting could help; like forcing a break if it's writing 0xB2 instead of 0xB1 at 0x000A8160 [15:51] and then run over and over again the test until it's triggered [15:52] i'm not familiar with gdb and its scripting system though, and i don't think i would be able to debug it even i triggers it :p [16:08] ubitux: yeah - it's my favorite soundtrack atm ... [16:21] (cia story: http://pastebin.com/9RBBniM1) [16:36] ubitux : shouldnt we be shaming this new hosting company due to bad hosting practises ? [16:36] and by shame i mean full out internet blacklist against this company ... [16:36] never forget never forgive. [16:37] no one should ever trust its data to a any company... [16:38] you can't blame the company for your own negligence [16:38] it wasnt him who rm rf-d [16:38] as soon as you realize every company is evil, the obvious move is to take care of its own data [16:38] whatever [16:38] cant blame the company? thats what i'm doing. [16:38] you *know* that at some point a company will disapear or just delete your data [16:39] but yeah, why wasnt cia.vc on github or some repo where people could copy it ? [16:39] was it open source ? [16:39] the code still exists afaict [16:39] it's the data that's missing (not sure what that means though) [16:39] no backups is kind of low rent, they're as much to keep the provider from fucking things up [16:39] (https://code.google.com/p/cia-vc/) [16:40] "atheme infrastructure" -> databases that weren't backed up, and possibly code for services that wasn't public, no? [19:53] j-b, you see the ML? re: DCP [19:54] Daemon404: yes, I saw that [19:54] Daemon404: I am just wondering WTH do XYZ to RGB and not XYZ to YUV ? [19:55] and also, shouldn't that be in swscale? WTH? [19:55] for display purposes, i guess [19:55] j-b, yes [19:55] it shoud be in swscale [19:55] no, swscale should not be :P [19:55] nevcairiel: so? [19:55] nevcairiel, it should be in swscale, but swscale should not be [19:56] you display rgb, rarely do you directly display yuv [19:56] what? [19:56] also, having no idea about XYZ, maybe it converts to RGB easier? :d [19:56] hes right, but thats irrelevant [19:56] im not sure how easy XYZ fits into swscale [19:56] I disagree [19:56] given how annoying even planar rgb is [20:03] anything non-YUV is really even less fun then usual in swscale [20:44] michaelni: i'm using scalarproduct_int16 (appears there is MMX only) and benefits are marginal, reference decoder is ~%33 faster [20:48] durandal_1707, theres ff_scalarproduct_int16_sse2 that should get used [20:48] pure C appears to be slightly slower than pure Delphi [20:49] my cpu have sse2 and ssse3 and sse2 is slower than ref mmx [20:49] ref have ssse3 which is even faster [20:50] (it does not have sse2 at all) [20:51] would branching cause this? [20:51] because our scalarproduct works only for order multiply of 16 [20:52] saste: ping [20:53] ubitux: pong? [20:53] [FFmpeg] michaelni pushed 4 new commits to master: http://git.io/YpWSUQ [20:53] [FFmpeg/master] Move subrip/text API change info from Changelog to doc/APIchanges. - Cl?ment BSsch [20:53] [FFmpeg/master] APIchanges: fill hashes. - Cl?ment BSsch [20:53] [FFmpeg/master] swscale: fix To{Y,UV} extern prototypes. - Cl?ment BSsch [20:53] saste: http://b.pkh.me/0001-lavfi-ashowinfo-check-plane-value-before-deferencing.patch [20:53] does this make any sense to you? [20:53] (spotted because gcc was complaining) [20:54] ubitux, ok if it fixes a problem / a warning [20:55] thx, pushed [21:06] Daemon404: maybe he stared 29.5h at swscale and then decided to do it in lavfi, which took him half an hour [21:06] lol [21:06] probably [21:40] ubitux: BTW ashowinfo should check extradata [21:44] saste: you mean extended_data? [21:44] yes [21:45] have fun doing it :) [22:11] argh f* bug, wth is wrong with my code :( [22:21] [FFmpeg] michaelni pushed 4 new commits to master: http://git.io/b-ZUJw [22:21] [FFmpeg/master] lavfi/ashowinfo: check plane value before deferencing. - Cl?ment BSsch [22:21] [FFmpeg/master] qt-faststart: speedup - Jan Ehrhardt [22:21] [FFmpeg/master] qt-faststart: dont allocate a bigger buffer than needed - Michael Niedermayer [22:48] cbsrobot_: ok i think i fixed the integrated loudness :) [23:29] ubitux: nice [23:43] [FFmpeg] michaelni pushed 2 new commits to master: http://git.io/Z53AzQ [23:43] [FFmpeg/master] swscale: move main swscale wraper to swscale.c - Michael Niedermayer [23:43] [FFmpeg/master] sws: drop unused variable - Michael Niedermayer [23:52] [FFV1] michaelni pushed 1 new commit to master: http://git.io/lQJlQg [23:52] [FFV1/master] Document minor_version - Michael Niedermayer [00:00] --- Mon Oct 1 2012 From burek021 at gmail.com Tue Oct 2 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Tue, 2 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121001 Message-ID: <20121002000501.554A718A018B@apolo.teamnet.rs> [02:07] Hi, I'm working on a program that links to libavcodec, and it crashed in avcodec_decode_video2.. I read that the usual culprits are missing calls to av_init_packet, avcodec_alloc_context and avcodec_open, but all that looks correct to me. [02:08] Is there anything about the inputs I could check before calling it? Even if that causes a crash, at least I'd know which check on the inputs caused the crash which would be a big clue. This is not yet a reproducible problem, just something from the field. [02:48] mapreduce: there was some function about registering all codecs. it should be called before anything else. [02:48] mapreduce: if this doesn't help, compile with debug support, run your program in debugger and make a backtrace. [03:45] hours later and all "ffmpeg -r 12 -b:v 3000k -f image2 -pattern_type glob -i 'GOPR*.JPG' ../video.mp4" gives me is a 10GB all black movie :-( [03:47] tcurdt: Always create a small sample to check that everything is okay before a large encode. "ffmpeg -i input -t 60 output" will create a one minute sample. [03:49] Also, your command is a little messed up. the bitrate goes after the input since it applies to the output. [03:51] To get help here you need to pastebin your command and output so we can see what's going on. [03:52] relaxed: hm& what about the -r 12? that as well? [03:52] relaxed: https://gist.github.com/25740af6ece3bbe52623 [03:56] no, that's the framerate you want to use as the input [03:57] tcurdt: 3840x2880 is huge [03:57] you didn't move the bitrate after the input [03:57] so it should be "ffmpeg -f image2 -r 12 -pattern_type glob -i 'GOPR*.JPG' -b:v 2000k -t 10 ../video.mp4" [03:58] yes, that will create a 10 second sample [03:58] do you really want the output to be 3840x2880? [03:58] relaxed: that's the input size & TBH I only need 1080p [04:02] well, 3840x2880 is 4/3 so... ffmpeg -pattern_type glob -i 'GOPR*.JPG' -filter:v "scale=-1:1080,pad=1920:1080:(ow-iw)/2:(oh-ih)/2" -crf 20 output.mp4 [04:02] add -t 10 in there [04:03] If you want better compression add -preset veryslow [04:03] (after the input) [04:05] If you don't need padding on the left/right sode then omit it from the filter chain. [04:05] relaxed: and if I want to keep the input ratio? just remove the pad? [04:06] well, the input aspect ratio is maintained but with padding on the side. [04:06] seems like you just read my mind :) [04:06] what's the -crf 20 ? [04:08] It's libx264 rate control setting. Do you really need 2000k or was it an arbitrary guess? [04:09] The of it as a quality setting that ranges from 0 (lossless) to 51 (crappy quality) [04:09] Think* [04:09] relaxed: it better be really good quality & 2MB/s was a guess & I've seen it as a default in another software [04:10] so it replaces the -b:v 2000k [04:10] I see [04:10] Yes, start with -crf 20 and work your way down until you're happy with the quality. [04:12] relaxed: just wondering - why use cry over b:v? [04:12] crf [04:12] it's the recommended 1 pass rate control method [04:13] ah & ok [04:13] yay - test worked and it wasn't all black! [04:13] In practice you should only set a specific bitrate if you're shooting for a specific size or average bitrate. [04:13] (nod) ok [04:33] relaxed: thanks you so much & finally at a point where it works as it should :) [04:33] you're welcome [10:07] Hi, we use ffmpeg for live transcoding. Is there a way, to monitor ffmpeg with nagios? [10:57] hello ! [10:58] i'm trying to replace the deprecated "m_format_context->timestamp" use in the OSG Plugin. So i'm looking for the entry "creation_time", in the context metadata. [10:58] The value type is char*. Should I simple cast it to double* ? [10:58] simply* [10:59] Bruno`, no, metadata contains textual data [11:00] saste: so i shoud do something like atof ? [11:00] Bruno`, what's the content of creation_time? [11:02] saste: attribute_deprecated int64_t timestamp; //@deprecated use 'creation_time' metadata tag instead [11:02] (avformat.h) [11:03] Bruno`, that is not what i asked for [11:03] what does the creation_time tag contains, what's the value of the stored string [11:03] saste: i don't know, that's what i am asking, kind of :p [11:04] Bruno`, if it is a metadata, you can see it with ffprobe or ffmpeg -i [11:04] right now, in my code, I use the deprecated "timestamp" attribute. So I want to listen to what ffmpeg says, and use "creation_time" instead. [11:04] and no, parsing a date with atof is not a good idea, you need a more specialized function [11:08] saste: ok, it is something like "2012-07-26 17:40:00" [11:08] what value do you want to extract from it? [11:09] strptime should be able to parse it (or even av_small_strptime) [11:19] ok. Thanks. [12:11] hi i want to stream an raw video and do an mp4 encoding onto an rtmp server i am using ffmpeg 0.11.1 with the pandaboard [12:11] on ubuntu os [12:18] another question.. My code seems to use the deprecated function "url_ferror".. how am i supposed to replace it ? [12:20] seems to be an avio_ prefixed function. [13:17] how to create thumbnail using version ffmpeg version N-44781-g299c0b3 [13:17] can somebody give me a working exmaple for that version, it seems ffmpeg is really broken into lots of different versions and some commandlines that used to work are no longer working [13:17] please provide me with the right info thank you [13:18] here is the ffmpeg -version http://www.pastebin.ca/2237915 [14:56] hi! I'm trying to play a VOB file with ffplay, but all I get is garbage on the output and lots of errors. I am pretty sure the VOBs are not corrupted as VLC plays the DVD fine (as in, if I point it to the IFO files, it works). ffplay output is at http://pastebin.com/F7hMQLX4. Is there something I'm missing / doing wrong? [15:02] blaman, could you open the ticket at the bug tracker? [15:03] it seems that either your mpeg2video is corrupted [15:03] or the mpeg2video decoder has some bugs [15:03] kendooo check out the wiki [15:04] ameeth, can you rephrase your question [15:05] wica, what exactly do you want to monitor? [15:05] yeah, I'm betting on corrupted, but then, why does VLC play it correctly? VLC uses libavcodec, so ffplay should play it OK as well...? [15:05] vlc doesn't use libav for everything [15:05] i.e. it's not just a wrapper for ffmpeg libraries [15:05] it has got its own codecs too [15:09] where to get the latest windows build? [15:10] ffmpeg.zeranoe.com? [15:10] google usually helps with questions like that [16:08] Hi all [16:08] I am trying to convert an image sequence to a video using ffmpeg [16:09] I found this on stackoverflow [16:09] ffmpeg -i frame%04d.png -vcodec ffv1 -sameq test.avi [16:09] my images are name from 0000.png to 0070.png [16:09] aren't you missing a -f image2 before -i? [16:10] -sameq is *not* doing what you expect to btw [16:10] ubitux: OK [16:10] ubitux: I remove -sameq [16:11] ubitux: But, what should I use instad of frame%04d.png? [16:11] my images are name from 0000.png to 0070.png [16:11] %04d.png? [16:12] ubitux: I also tried that, but it doesn't work [16:12] ffmpeg -r 24 -f -i %04d.png -vcodec ffv1 test.avi [16:12] ubitux: OK [16:12] what is this standalone -f? [16:13] http://www.pasteall.org/35809 [16:14] ubitux: I don't know :| [16:16] you want -f image2 [16:17] ubitux: Thanks, but the error changed: [16:18] http://www.pasteall.org/35810 [16:18] why did you remove the -i? [16:19] ubitux: yep, % ffmpeg -r 24 -f image2 %04d.png -vcodec ffv1 test.avi [16:19] ubitux: Oh, no :| [16:19] ubitux: The same error after removing -i [16:20] put it back. [16:20] -f image2 -i %04d.png [16:21] ubitux: Thanks, it worked [16:21] ubitux: Thsnk a million for helping [16:22] np [16:33] burek: I lik eto monitor if the stream is stil play. And when it is possible the number of drop frames [16:34] s/I lik eto/I like to/ [16:36] wica, redirect stderr of ffmpeg process to a log file and monitor its content [16:39] ffmpeg -i ... output.wav 2>logfile.txt [16:40] Thnx, was working on that :) [16:40] Was hoping on something betr :) [16:49] wica, better in what way [16:56] burek: beter then a script of my own :) [16:56] wica, well give an example [17:14] is it possible to simultaneously send and receive RTP (specifically H.264) packets using the same socket with ffmpeg? [17:19] relaxed: you about ? [17:20] I have a treepad lite folder where i am storing all the commands and syntax that i need and use [17:20] can someone give me the syntax for converting an VORBIS audio into mp3 192 please [17:20] its in video format so im assuming you can just ignore the -vc [17:22] ffmpeg -i lol.webm -acodec liblamemp3 [17:26] ffmpeg -i input -map 0:a -c:a libmp3lame -q:a 0 output.mp3 [17:27] It's better to map what you want. [17:27] what does the a mean [17:28] im assuming its leaving out the video codec [17:28] -map allows you tell ffmpeg exactly what you want in the output. [17:29] ok but what does the rest mean [17:29] -vn just ignores video but sometimes you want a specific stream or there may be mulitple audio streams [17:30] So -map 0:a:0 maps only the first audio stream to the output [17:30] i have copied what you have written i just want to comment it so i can understand and rmembe for next time [17:30] -c ? [17:31] -map 0:v:0 -map 0:a:1 would map the first video stream and the econd audio stream to the output. [17:31] -c:a = codec:audio [17:32] 0:a (first stream) -c:a (codec audio) CODEC -q:a 0 (highest quality?) [17:34] also i dont understand why you have done 0:a:0 [17:34] rather than 0:a [17:34] is that if you are doing a huge line [17:34] and processing the seperate streams later in the line [17:35] then it would be 1,2 ,3 [17:35] 0:a will include all audio streams in the output [17:35] ok [17:36] and yes, -q:a 0 is highest quality [17:38] thanks mate your a real help. [17:38] Action: sine_ copy and pastes [17:56] wow i haven't been here in ages [17:57] anyway .. i have a mpeg that i need to be made into a dvd .. im assuming its just ffmpeg -i file here -aspect 16:9 -target ntsc-dvd dvd.mpg [17:57] ? [17:57] also why do i have to put it -aspect 16:9 will ffmpeg tell me what it is by default ? [17:58] i don't need a menu or anything [18:00] crypticmofo: ffmpeg -i input -target ntsc-dvd -filter:v 'scale=720:720*(((16/9)*(2/3))/(iw/ih)),pad=0:480:0:(oh-ih)/2,setdar=16:9' ... [18:00] umm [18:01] how would i know that ifi iddn't come here and what does it do [18:01] ^^ should give you the sorrect aspect ratio [18:01] correct* [18:01] how would i know this thought if i didn't come here [18:01] is there like a n00b proof way of finding what i need ? [18:02] Use something clicky like handbrake? [18:02] convertxtodvd works in wine [18:03] don't have wine installed [18:04] relaxed, can ffmpeg tell me my aspect ? [18:04] well i see that its playing at 1.78.1 aspect [18:04] i know thats is around 16x9 [18:04] if im correct [19:29] tmatth why would you want that [19:30] burek: When using RTP in the context of a SIP call, PBX's expect the RTP to be sent/received on the same port [19:33] well, in theory it is, if you modify ffmpeg's source code [19:34] that's what i've been experimenting with...would a patch allowing "reuse" options for SDP/RTP be a good idea? [19:36] it would be welcome I guess [19:36] if a use-case is reasonable [19:37] burek: the case i'm looking at is peer to peer calls with freeswitch as the SIP softswitch [19:39] burek: when it detects that RTP from peer A is being sent from an unexpected (i.e. different port than the peer receives on) port, it automatically starts sending RTP from peer B to that port [20:02] hey [21:27] hye [21:27] relaxed: you here bro ? [21:27] :d [21:27] syntax king [21:31] anyone know how i can do a batch conversion of files [21:31] bash? [21:32] like all files with *.this and *.that convert into mp3 keeping their filename [21:32] Hello all, I'm trying to take two videos and combine them into one using a multiply effect. How can this be done with filters? [21:33] I dont use linux im on windows using a build [21:33] like find ./ -iname "*.wav" | while read asd; do ffmpeg -i "$asd" $(basename "$asd").mp3 ; done ? [21:33] it doesnt matter anyway im going to work now. ill check out the wiki at work [21:33] Then I have no idea. [21:33] thanks - bye. [00:00] --- Tue Oct 2 2012 From burek021 at gmail.com Tue Oct 2 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Tue, 2 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121001 Message-ID: <20121002000502.5110518A01E8@apolo.teamnet.rs> [00:42] did anyone understand what Ronald was so angry about? [00:42] (honestly) [00:43] cbsrobot_: I: -23.4 LUFS [00:43] you might want to test the latest version in my branch [00:43] ubitux: ok will do [00:44] i need to improve the code a little [00:44] and likely do some similar changes to the LRA [00:45] so i'll push this week [00:45] and will work later on the true peak [01:15] ubitux : libav side never could understand michael's english [01:15] I guess the "right way" to implement the XYZ -> RGB transform would be in libswscale [01:16] its either a pixfmt or a raw decoder , yep [01:16] conversion does not belong in filter :P [01:16] yes a raw decoder could also work [01:16] i suppose adding another colorspace in libswcale would be painful [01:18] iam planing to make it easy to add colorspaces [01:18] youve got a lot to learn about social skills^W^Wmailing list trolling guys [01:19] hey i'm not trolling :( [01:19] Daemon404, did you understand why ronald is angry ? [01:19] the email wasnt very coherent... [01:19] i think he just wanted an occasion to express his frustration about something, but took a bad thread and misunderstood it [01:20] yes ^^ [01:20] that said [01:20] i am not touching this with a 10 foot pole [01:23] Action: michaelni locks Daemon404 in a small box with "it" and burries the box 10 meters below quick drying cement [01:23] :D [01:23] this is not very social michaelni [01:24] it's a fun fact that party guests dont like being murdered [01:25] Action: michaelni pulls the box quickly back out before the cement dries and lets Daemon404 back out [01:26] /nick Houdini [01:26] :) [01:27] +2 social skills for michaelni [01:27] \o/ [01:27] michaelni: but now Daemon404 will have lifelong trauma ! [01:28] anyway, 'night ppl :) [01:28] night ubitux [01:29] Action: michaelni goes back to coding [01:39] huge letters too [02:16] michaelni : now is good time to ask ronald on the list if he has any suggestions or wanted features in swscale [02:16] extend the olive branch of discussion :) [02:16] ask him if there are any patches which should be merged [02:16] or any you missed [02:18] and it is customary culture of some english countries to apologize for any misunderstandings of statements made. of course , this feels like 2011 all over again [02:18] Action: Compn goes to do something else before brain explodes [02:32] he is frustrated by the news of the 1.0 release. [02:35] I guess he is also frustrated that ffmpeg is releasing more of the new libav code, than libav itself. [02:38] maybe hes stuck between a rock and a google. [03:02] I have no idea about that. [03:17] He was such a promising nice guy. Then the takeover happened and he turned into vicious angry master. It is heart wrenching. [06:50] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/K_18JA [06:50] [FFmpeg/master] movenc: fix edit list for the case of negative pts. - Michael Niedermayer [10:42] how can i fill gaps in an audio stream? [10:43] i could use aresample, but IIRC it only fills the initial gap [10:43] is there a way to skip n frames from demuxer? [10:46] michaelni: i kind of find a way to debug the race [10:46] found* [10:48] would it make sense to implement a timepad filter for that? [10:48] we have many filters which only work with continuous streams (e.g. amerge) [10:50] saste: you want to write some sounds generation interpolation between two points? [10:50] ubitux: a silence pad would be enough [10:51] also i see there is a asyncts, and i wonder if there is a corresponding option in ffmpeg/swr [10:52] it's done manually in ffmpeg.c [10:52] asyncts insert is disabled [10:54] ubitux, in the ffmpeg manual -async is deprecated, in favor of asyncts [10:55] according to michaelni something is wrong with it [10:55] so it's been disabled in ffmpeg [10:55] iirc [10:55] what the current behavior with regards to audio gaps? [10:56] or maybe it was changed [10:56] if i don't specify special options, and i don't insert asyncts/aresample? [11:00] http://pastie.org/4890153 got you, bitch [11:20] it seems the race is around the H264Context->ref_count[2] [11:21] (frame context) [12:03] and maybe another (?) around ff_h264_lps_range in ff_init_cabac_states (cabac.c:146) [12:13] where does cia get its input from? [12:13] (the bot) [12:15] cia is dead [12:16] btw i've seen cia bots all over freenode [12:16] is that something provided by freenode or what [12:16] no [12:16] cia.vc [12:17] but it's dead atm [12:17] there is FBI and KGB though [12:18] KGB is sloow (not in real time) [12:20] durandal_1707: write a new one and call it FSB [12:20] or MOSSAD [12:21] or... Verfassungsschutz [12:22] call it NDB - they seem to leak fast [12:41] or wikileaks :) [12:41] g2g bbl [13:14] saste, aresample can fill gaps in the middle, see the AVOption table in libswresample/swresample.c [13:14] for parameter, i dunno if its in the docs [13:19] michaelni, ok, i'll try to write a document explaining the various options [13:22] thx [15:00] i'm back [15:00] michaelni, where does cia get its input from? [15:13] burek, dunno exactly but there must be a hook on the ffmpeg @ vlc git repo [15:13] that pushed something toward CIA [15:13] I think so too [15:14] Btw, how are we supposed to set the cpu flags those days? [15:16] for ffmpeg via command line -cpuflags, for the libs via av_get_cpu_flags() and av_force_cpu_flags() [15:17] theres also av_parse_cpu_caps() [15:17] to parse a string [15:17] of things like +mmx-sse2 [15:19] get_get_cpu_flags seems to be the opposite [15:20] parse takes a string, not a flag [15:21] av_set_cpu_flags_mask is deprecated [15:22] ... and av_force_cpu_flags() [15:22] ok [15:23] great, a unique API... [15:23] thx michaelni [15:46] ubitux: the summary is nice [15:52] :) [15:52] results should be closer to what you expect [15:52] yeah looks great [15:52] i still want to improve the precision, and something still looks fishy with the LRA so i need to dig a little more [15:53] but i'll push soon a similar version [15:53] I could also crosscheck with another tool - just to make sure it's correct [15:53] feel free to :) [15:53] is the threshold ajustable ? [15:54] no, the specs impose a value [15:54] you're refering to the I and LRA gate threshold, right? [15:54] yes - so maybe it could be omited in the summary - no ? [15:55] ah this one [15:55] hmmm, it changes from file to file [15:55] this threshold is moving over the analysis [15:55] ah ok [15:55] it's based on hardcoded thresholds [15:56] (-10 for I, -20 for LRA iirc) [15:56] it's the "relative" threshold :p [15:56] it might not be useful [15:57] but i wanted to print it anyway otherwise the "sections" would have been empty ;) [15:57] ha [15:57] ok [16:03] nit: maybe the stats could also be printed depending on the loglevel [16:05] they are in visual mode [16:05] (except the summary) [16:34] [FFmpeg] michaelni pushed 9 new commits to master: http://git.io/k-VrEg [16:34] [FFmpeg/master] ac3dec: ensure get_buffer() gets a buffer for the correct number of channels - Justin Ruggles [16:34] [FFmpeg/master] smacker: read escape codes in single get_bits() call - Paul B Mahol [16:34] [FFmpeg/master] Remove some silly disabled code. - Diego Biurrun [17:47] oh saste's code went in [17:48] Action: Daemon404 updates his attached_pic logi c [17:54] mmmh [17:56] "The FFmpeg/libav developpers were begging for us to drop libpostproc" [17:56] who from the FFmpeg developpers? [17:56] ! [17:57] see my answer [17:57] where? [17:57] vlc-devel [17:58] is that where the drama moved to? [17:59] oh boy whats this [17:59] Action: Daemon404 reads [17:59] :) [18:00] i'd be interested who even uses postproc in vlc [18:00] or at all in anything [18:02] Action: Daemon404 should sync libpostproc.git soon to build with msvc like ffmpeg's [18:17] [FFmpeg] michaelni pushed 2 new commits to master: http://git.io/YRrSwQ [18:17] [FFmpeg/master] cpu: improve av_get_cpu_flags() doxy - Michael Niedermayer [18:17] [FFmpeg/master] jpeglsdec: move pict debug log under correct if() - Michael Niedermayer [18:18] j-b: bah, no drama at all in that thread [18:18] you can do better [18:19] of course [19:12] btw, what does "not required by default" means in the vlc configure context? that configure would not die refusing to continue without it, or that it won't be enabled even if present on the system? [20:38] michaelni: do you have any plan to define Opus in MOV like this? http://wiki.xiph.org/Oggless [20:43] MP4_maniac, no but maybe ask nicolas he might be interrested to work on it [20:47] michaelni: where is nicolas on irc? [20:50] i dont think he is on IRC [20:51] he is on ML only? [20:52] yes i think so [20:59] michaelni: may i ask him or post topic about opus in mov on ffmpeg-devel ml? [20:59] sure [20:59] k [21:40] where is saste? :( [21:49] speaking of saste, in my patch review he told me that it would be useful to write a test, how does it work in ffmpeg?? [21:50] we mostly use "the old way" to cover quite some cases using tests/lavfi-regression.sh [21:50] "the new way" is to add a test in tests/fate/filter.mak [21:50] but it's less effective [21:50] I see :o [21:51] note that adding a test in tests/lavfi-regression.sh will need you to run configure again and stuff like that [21:51] ok [21:52] nyuhu: check 5780f9bbd5668a372eac63dddcb4cda06edfaed5 [21:52] or 39b0d40d9219b895dbd1dd5ba5873f8594360750 [21:53] nyuhu: did you ever run fate? [21:53] no :x [21:53] do you have the samples? [21:53] yes [21:53] make fate-rsync SAMPLES=... ? [21:53] yep [21:54] then well, just make fate -j20 SAMPLES=... [21:54] oki [21:54] (without SAMPLES=... if you have --samples=... in your configure line) [22:27] KGB :( [22:27] CIA :( [22:27] ffbi :( [22:31] Bots, I am disappoint. [22:36] ok now, vobsub demuxer! [22:36] or maybe i should just watch some animes [22:36] Action: JEEB just checked the next cour's shows [22:37] dunno what to watch.. [22:37] i think i have some classics still unwatched [22:38] I have a couple of older shows around to watch atm: RahXephon, ARIA, Kidou senkan nadeshiko [22:40] the first two seasons of Aria are nice, and the last one is insanely good :) [22:40] also I need to watch Gunslinger Girl [22:40] JEEB: i couldn't watch more than 1-2 episodes of nadeshiko. but there are people who like it. [22:40] kind of nice, but way below Noir or Phantom IMO [22:41] saste \o/ [22:41] Phantom I watched until ep10 or so if you mean Phantom of the Inferno [22:41] EBU R.128 in, going to try it [22:41] aria is slice of life. rahxephon is simply classic. [22:42] yeah, also aria's manga is just pretty <3 [22:42] JEEB: Gunslinger Girl is highly recommended. [22:42] JEEB: requiem for the phantom [22:42] need to get more of it when I get to a Book-Off the next time [22:42] ubitux, same [22:43] speaking of phantom, Kurau is surprisingly nice [22:44] ah, that sounds like a nice one by Bones [22:46] and the last best anime i saw was kemonozume [22:46] anyway, if you have some kind of recommendations, i'm all ear [22:47] i like a lot of weird/peaceful/violent/mindbreaking stuff so... [22:47] ubitux: i guess you've watched madoka :) [22:48] I have this thing I hate in memory, I usually forget about most things I've watched good or not if someone asks me for recommendations [22:48] iive: yes [22:49] JEEB: don't you have a list stored somewhere? [22:49] JEEB : i'd pass on rahxephon. one of those series with little answers and lots of booooooooring [22:49] Compn, I already watched most of it and deemed it relatively enjoyable while it did have its shortcomings [22:49] iive: and pretty good surprisingly :p [22:49] :P [22:49] ubitux, I try to keep a list on anidb of what I've seen, but I often forget to add stuff there [22:52] so i'm condemned to write code tonight? :( [22:52] I liked both baccano and durarara, I liked haibane renmei, higashi no eden, bungaku shoujo (OVAs and the movie) looking at my current list [22:53] and yes, this shows how freaking short my list is :< [22:54] then I liked code geass that I watched after the summer [22:54] saw everything except bungaku shoujo [22:55] code geass? please... :/ [22:55] dunno, I enjoyed it for what it was [22:55] bungaku shoujo is kind of an odd one [22:58] anyways, I wish I kept better books of what I've watched [22:58] most of it is crap, but there are some better ones [23:00] code geass was good, they just made it twice, so the repetition was boring. [23:01] also I'm one of those weirdos that began their mecha watching with Macross F and Gundam 00 [23:03] ever seen eva ? [23:04] ok, that came before -- but I really didn't watch it as mecha back then [23:04] I think one of the renewed DVDs for Eva was the first thing I ever imported from the US [23:04] that was before I habla'd Japanese [23:13] ubitux, yeah, ebur128 is cool [23:13] i loove retro-style CGA fonts :) [23:13] :D [23:57] michaelni: ping on duration in AVFrame [23:58] saste : do you have a filter wishlist? can you put an autocropping filter on there ? [23:58] automatic cropdetect :) [23:58] i already asked you some weeks ago, you said it was somehow duplicating repeat_pict and told me to check that the patch was not breaking B-frames in mpeg or something like that [23:58] i don't know how to test that [23:58] Compn, trac [23:58] k [23:59] also I don't know what do you mean by automatic cropdetect [23:59] we have already cropdetect [00:00] --- Tue Oct 2 2012 From burek021 at gmail.com Wed Oct 3 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Wed, 3 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121002 Message-ID: <20121003000502.7023518A01ED@apolo.teamnet.rs> [00:00] Compn, ^ [00:05] automatic crop detect makes more sense for playback applications. A lot of TV programs add padding and switch different movies, with and without padding [00:06] having a way to automatically remove the black bars would be nice feature. Of course, it complicates things a lot for encoding application. [00:14] iive: something like... cropdetect write the cropdetected values to metadata, crop read and apply it [00:15] this could be done (you need to handle the logic when cropdetect can't guess senseful values) [00:15] also crop can't change the cropping area size [00:22] changing cropping size can be a wish [00:22] i just want one crop :P [00:22] one auto crop :P [00:27] well crop a fixed area, is that automatic enough? [00:28] in other words, what do you want to crop? [00:30] saste : if i have a bunch of files with encoded black bars (so 16:9 content is in 4:3) , i want a filter to automatically crop it [00:31] Action: Compn finds sample [00:32] Compn, you can still easily script it [00:33] http://samples.mplayerhq.hu/A-codecs/lossless/ALAC/bgirl_trailer.mov [00:33] the problem is that right now changing crop total size is not possible (not without reconfiguring the filters) [00:34] ok but it could just not start 'cropping' until a few seconds into the video [00:34] so it waits until $userspecified time, then starts [00:34] or until $cropvaluesstopfluctuating [00:34] yes but you need to insert the crop filter since the beginning [00:34] so you need to already specify the total area [00:34] for the crop filter [00:35] crop x:y coordinates can be changed on the fly with no problem [00:35] is w:h which can't be changed [00:35] so at the moment it is a two-pass process [00:35] you get the cropping values, and apply it [00:37] so wish : insert filters after video starts ? [00:37] :P [00:38] or cant do it ? [00:38] can't [00:38] not at the moment [00:39] patches are welcome :P [01:17] [FFmpeg] michaelni pushed 7 new commits to master: http://git.io/Oe7xTQ [01:17] [FFmpeg/master] Move xGA font data from lavc to lavu. - Cl?ment BSsch [01:17] [FFmpeg/master] lavfi: EBU R.128 scanner. - Cl?ment BSsch [01:17] [FFmpeg/master] lavfi/edgedetect: add missing minus in a comment. - Cl?ment BSsch [01:30] Action: michaelni is slightly unhappy that the slightest change to the nut muxer causes 200k diffs [01:31] wat [01:35] half the regression tests use nut, if anything changes in it theres lots of diffs to the checksums [01:36] michaelni, it might be wise to change those as a separate commit [01:36] it might obscure teh actual change in the diff if it isnt [01:37] yes but it causes troubble if one runs a bisect with make fate [01:37] ah... right [01:38] yes FATE is strictly tied to NUT [01:38] that's mostly my fault (for using it in fate-lavfi) [01:38] oh saste, nicely done on ffprobe's disposition [01:38] i switched to using it today [01:39] Daemon404, good [01:39] so you have more useful info, such as karaoke stuff [01:39] for various definitions of 'useful' [01:40] yes, that's always relative [01:54] michaelni : so will you ask ronald if he has suggestions or forgotten patches for swscale? [01:54] oh you already did [01:54] ok nevermind [01:55] Action: Compn finally catches up on mails and his bad memory [02:33] [FFmpeg] michaelni pushed 2 new commits to master: http://git.io/ml8JBg [02:33] [FFmpeg/master] ffmpeg: print muxed packet sizes in debug output too - Michael Niedermayer [02:33] [FFmpeg/master] Libspeex VAD support - Dmitry Samonenko [07:30] wo, long commit descriptions from michaelni, that's unusual :) [07:35] (thx btw) [08:10] kierank: you were interested in normalization the other day, i think this might help you: http://ffmpeg.org/pipermail/ffmpeg-user/2012-October/010090.html [09:05] ubitux: thanks [09:51] Action: ubitux didn't hear much Nicolas recently... :( [09:51] i hope he is secretly writing some awesome code [10:16] TimNich: ping [10:17] I guess you need to add SAMPLES=/path/to/fate [10:21] ubitux: something like autocorrecting to ebur128 would also be a useful feature [10:21] similar to http://www.nugenaudio.com/LMB_Loudness_batch_processor.php [10:22] cbsrobot_: you need two passes anyway [10:22] yes sure - they do it in multiple passes even [10:23] ffmpeg is stream oriented it's kind of complicated to do that in builtin in the current state [10:23] but a script in tools/ could do that yeah [10:23] shouldn't require more than 10 lines of shell scripting [10:24] I don't think is as simple as settting the value though [10:24] if you just want to reach -23LUFS it should be as simple as changing the volume [10:35] cbsrobot_: pong [10:35] TimNich: I guess you need to add SAMPLES=/path/to/fate [10:36] ^answer for your mail to the ml [10:37] nahh. just cross eyed reading my notes, its make fate-whatevertest not make fate fate-whatevertest [10:37] k [10:40] Can see what is happening but can't see why!! If you correctly set the atom, then specify a be codec it gets tagged as le according to both ffprobe and QT, hence the epic fail (psnr from ~100% ->~0%) ;( [10:41] codec copy is fine though [10:44] Tjoppen: I'll able to get back to work on the mxfdec set audio pts patch. I've been busy these last days. [10:45] and the diff filter! [10:45] :-? [10:46] sure :) [10:46] mateo`: k [11:23] michaelni: got some more info about h264 reg [11:23] interested or...? [11:23] afaict there are two races [11:24] one around H264Context->ref_count as said in the mail, and another one because simultenaous frame worker threads seems to run a ff_init_cabac_states() [11:24] worker thread -> decode_frame -> decode_slice -> init cabac states [11:26] so maybe moving ff_init_cabac_states() only once in the init or something (even if not necessary) could help [11:32] heh. [11:41] is there any reason ff_init_cabac_states() would need to be called several times? [11:42] different QPs? [11:42] afaict these tables seems to be used RO [11:42] Skyler_: it takes no parameter [11:43] Ah, it doesn't do that. [11:43] anyway, moving it to the h264 init callback fixes that specific race [11:43] and the tests still pass [11:44] now there is still the second race with the ref count [12:06] couldn't 'pattern_type glob' be auto-recognized if strpos(filename, '*') returns true? [12:06] in image2 demuxer [12:07] i mean 'glob' and 'sequence' are mutually exclusive, so there is kinda no need to specify pattern type, it is enough to analyze the file name only [13:32] oh, MIPS instances! [13:32] awesome [13:33] oh IA64 as well [13:33] seems those are old though [13:39] < burek> couldn't 'pattern_type glob' be auto-recognized if strpos(filename, '*') returns true? [13:40] and what happens to file with a '*' ? [14:00] http://www.p01.org/releases/MATRAKA/matraka.png.html ? look at the source, wtf [14:16] ubitux, normal file names can't contain character '*' [14:16] so if a filename/file pattern is provided, that contains '*' it can safely be assumed a user wants the globbing to be applied [14:17] thus, no need for an additional option ' pattern_type' [14:18] < burek> ubitux, normal file names can't contain character '*' // why? [14:18] [~]- touch \* [14:18] [~]- ls -l \* [14:18] -rw-r--r-- 1 ubitux ubitux 0 Oct 2 14:18 * [14:19] :) [14:19] if someone was to write a service using ffmpeg, and ffmpeg happens to interpret files with '*' differently, then we could broke that system by easily sending a '*' file [14:20] developers shouldn't have to anticipate all the diverse special inputs [14:21] it's really confusing that unix allows asterisk [14:22] since most of its shells recognizes that character on the cmd line automatically and expands file names [14:22] but ok, you're right [14:22] thats why you have to escape it to use it [14:23] Action: cbsrobot_ sends a :beer: file [14:23] or quote it [14:23] anyway, i think i found the cause of the third race [14:23] burek did you see how fast it vanished ? [14:23] :) [14:24] and it seems in the encoder side [14:24] this time. [14:24] burek: eighter your shell is broken or you were thisty :P [14:25] [FFmpeg] michaelni pushed 4 new commits to master: http://git.io/RTISJg [14:25] [FFmpeg/master] lavfi/transpose: add support to named options and shortands - Stefano Sabatini [14:25] [FFmpeg/master] lavfi/transpose: add passthrough option - Stefano Sabatini [14:25] [FFmpeg/master] doc/swresample.txt: fix typos - Stefano Sabatini [14:25] beer is always a top priority at qos schedulling :) [14:26] just befor voip I guess [14:27] sure :) [14:28] btw, is there somehwere some diagram like this, but for ffmpeg's git repos: http://nvie.com/posts/a-successful-git-branching-model/ [14:29] or can it be generated based on ffmpeg's git [14:30] burek: yes, https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcR55JQFT8KopQjw9hF77Bt5AOhdRaVySgthF2E2qdHxolCz2msQ0w [14:30] :) [14:31] gitk ? git log --pretty=oneline --graph ? [18:23] Hi all, I need some halp with livavformat, I want to send video frames via udp and I do not understand how to do it... I have success in write frame to a file... what is the correct steps to do it? I apologize for my bad english... [18:24] http://www.nobodygoeshere.com/wp-content/uploads/2012/09/halp-cat.jpg [18:25] simonec77, put your audio/video streams into mpegts format and use udp protocol [18:28] burek: Looks 'shopped. I can tell from the pixels, and from having seen a few 'shops in my time. [18:30] ok, I already do a mts stream video+audio and I used avio_open(&m_spAVFormatContext->pb, in_pchfilename, AVIO_FLAG_WRITE) and av_interleaved_write_frame(in_psAVFormatContext, &sAVPacket) but I got error! [18:31] simonec77, look into doc/examples [18:32] all the examples wtite only to a file....:-( [18:36] simonec77, take a look at ffmpeg.c [18:36] and follow the case of the udp output [21:27] [FFmpeg] michaelni pushed 31 new commits to master: http://git.io/7D0ORg [21:27] [FFmpeg/master] wmadec: Adjust debug printf argument length modifier - Diego Biurrun [21:27] [FFmpeg/master] avformat: const correctness for av_hex_dump / av_hex_dump_log - Diego Biurrun [21:27] [FFmpeg/master] mpegts: Drop pointless casting of hex_dump_debug arguments - Diego Biurrun [21:34] michaelni: just curious, are you updating the reference tests manually? [21:34] or you found a way to overrid the references when running fate [21:35] i tried to do something like CMP=cp the other day but wasn't able to achieve anything [21:35] :( [21:38] ubitux: you can grep for diff and sha1sum in the scripts. I was able to figure out how to trick it eventually [21:38] oh? [21:39] a command to basically go "take the current output to be correct so I can commit it" would be nice [21:39] last time i had to do it manually for all the ffprobe test was quite painful [21:40] (oh and that reminds me i never push that stuff) [21:40] pushed* [21:40] IIRC it does something like "diff ref output", which I just switched around to "cp output ref" [21:40] that's what i tried to do [21:40] with CMP=cp :D [21:40] yeah, but the order of the arguments matter for cp and not for diff [21:40] :) [21:41] so you probably copied the refs to the outputs :) [21:41] well it matters for diff as well [21:41] yeah possible [21:41] not for finding whether a test fail, technically [21:41] sure [21:41] so yeah.. I think I've done exactly that mistake twice [21:42] mmh i see the "null" cmp [21:42] i wonder if CMP=null wouldn't do the trick [21:42] ah no it doesn't look right [21:43] maybe we should just add a rule [21:52] ubitux, i use something like "make fate -k -j12 | patch -p1" [21:53] oh fun [21:53] of course! [21:53] didn't though about that [21:54] thx for the trick :) [21:54] note, this needs to be done twice if the seektests change too because they are only run if the dependancies succeeded [21:54] okay [21:59] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/mPVRBg [21:59] [FFmpeg/master] fix exit_program() prototypes - Michael Niedermayer [22:05] ubitux: Nicholas never checked in his int64 timestamp fix for srtdec? [22:06] don't remember, url? [22:10] http://lists.ffmpeg.org/pipermail/ffmpeg-devel/2012-September/130430.html [22:11] That was his alternative to my attempt at preserving precision. [22:13] i've reviewed this iirc [22:13] my comments still stand afaict [22:20] Well, he hasn't updated or checked anything in. [22:21] not much news from him recently :( [22:48] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/-Utbow [22:48] [FFmpeg/master] 8svx: remove malloc and memcpy that have become unneeded - Michael Niedermayer [23:02] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/tILNiA [23:02] [FFmpeg/master] ffserver: fix potential buffer overflow, based on wrong fscanf format indentifier. - Martin Ettl [00:00] --- Wed Oct 3 2012 From burek021 at gmail.com Wed Oct 3 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Wed, 3 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121002 Message-ID: <20121003000501.65E4E18A01EC@apolo.teamnet.rs> [00:41] hello. I'm trying to implement an streaming using ffmpeg. I've successfully implemented it using the line in of my sound card and then transmitting the sound using mp3 coding, but now I want to do it in a lossless format like wav or flac (Iit's a must for my application) [00:42] but all my efforts to do that have been failed [00:42] the command that works for me is the following [00:42] ffmpeg -f alsa -i hw:0 -acodec libmp3lame -ar 11025 -ac 1 -f rtp rtp://234.5.5.5:1234 [00:43] but when I change the acodec for pcm_s16le it gives me an error in the player [00:44] I'm using VLC and it gives me the following error "SDP required: A description in SDP format is required to receive the RTP stream. Note that rtp:// URIs cannot work with dynamic RTP payload format (97)." [00:44] how can I solve it? [00:44] how can I streaming in lossless format? [03:00] wow... go away for a few weeks, I come back and there's forums, ffmpeg 1.0 is released [03:01] lots of good stuff [05:33] just a short question, i'm running ffmpeg version 0.11.1 and was a bit surprised that "ffmpeg -i INPUT -c copy OUTPUT" is not equivalent to "ffmpeg -i INPUT -map 0 -c copy OUTPUT" is this intentional? the first version copies audio, video and subtitle, the second one does so too, includes all attachments though [05:34] in the manual it states that if you use no stream specifier it should match all streams, even if i selected attachments seperately with -codec:t copy it wouldn't copy a single attachment [05:37] just updated to ffmpeg 1.0, same results [08:42] hi [08:42] is there a way to check what subtitles an mkv file has without actually playing the file? [08:42] like check the info, etc [08:42] mkvinfo? [08:42] from the package mkvtoolnix [08:42] ty [08:43] Last message repeated 1 time(s). [08:43] np [08:43] you could also do ffmpeg -i INPUT.mkv and it would list the streams... i think ffmpeg would exit with an error code though [08:44] i see, i will try them both, thanks [08:47] yeah mkvtoolnix worked great [08:47] with ffmpeg -i i got some message like: "Invalid data found when processing input" [08:47] thanks [08:47] oh that's weird [08:48] maybe some corrupt file [09:38] is there a way to encode a video from an image sequence if they don't have filenames like image001.png image002.png etc? [09:44] rubendv, -pattern_type glob in the image2 demuxer [09:51] Action: ubitux didn't hear much Nicolas recently... :( [09:51] oups bad chan [11:17] hi to all i need help for use avconv, this is the right channel? [11:18] #libav would be that [11:19] thz [11:48] i need to create video from list of png [11:58] klaxa, that's intended behavior [11:58] somehow it makes sense to some people [11:58] to select just a/v streams from the input, by default [11:58] so, yes, you need to use -map 0 to map all streams from your first (0) input [11:59] yeah, it's a funny thing [15:07] Hi!Can somebody help me in converting a 23.98 fps .mkv to 25 fps mpeg-2 keeping the same length and without flickering? [15:08] i managed to convert it to 25 fps, but the image is flickering in every second or so.. [15:08] is there a way to eliminate this flickering? [15:09] i suspect that the couse of the flickering is the frame duplication [15:10] is it possible in any sofware way to make 25 fps from 23.98? [15:13] creespekt, can you pastebin your command and complete output [15:17] here it is: http://pastebin.com/pYVUF6Qm [15:19] sir, you are not using ffmpeg [15:19] i am trying to make a 25mbit, full hd, interlaced, 25 fps mpeg-2 [15:19] please visit libav's support channel [15:23] i tried there, but i got the information, that theres no any solution, to convert a 23.98 fps video to 25 fps [15:24] I'm really not familiar with that tool at all, so I'm not of much help here [15:24] :/ [15:24] it works the same as ffmpeg [15:24] i am asking just in general [15:24] if it works the same, why the separate project then? [15:25] and no, it doesn't work the same [15:25] is this possible to make 25 fps from 23.98 with any software? [15:25] creespekt, did you try with ffmpeg? [15:26] burek: for the same reason there's no world peace [15:26] yeah, i got the same outcome, but i cant paste you command from there [15:26] thats why iam trying in every other tool [15:27] well, if you can provide ffmpeg logs, I might be able to help [15:27] ok [15:31] i cant find any logs:/ [15:33] so in general ( no matter what program) is it possible to make 25 fps from 23.98? [15:33] it is [15:33] one way is to not preserve the speed [15:34] (just changing the fps) [15:34] the other is to maintain the speed [15:34] by duplicating/dropping frames [15:34] i need to use the second one [15:35] i need to maintain to video length [15:35] but with duplicating the frames, the video flickering in action screen with fast movement [15:36] and i try to avoid it somehow [15:36] did you try ffmpeg -i INPUT -vcodec ... -acodec ... -r 25 output [15:36] yeah, and its flickering [15:36] try some other video encoder [15:37] which handles screen changes better [15:37] it is 25 fps, but you can see, especially in the background scenes during fast movements, that ist not continous [15:37] according to your avconv command, you are doing a lot of things at once [15:37] so it is difficult to narrow down the exact source of your problem [15:37] try one thing at the time [15:37] to find what's causing your trouble [15:37] and just work around it [15:38] okok [15:39] i just wanted to know the is it theoretically possible to do this or i am just wasting my time [15:39] thanks [15:40] and sorry for the wrong channel:) [15:40] np :beer: :) [15:52] is it possible concatenate fade filter(in|out) more than 2 times? [16:16] carnau: did you try chaining mulitple fade filters together? [16:21] I'll do to be sure [16:21] *do it [20:20] Hi guys, I have problem. I convert avi file with two audio tracks. FFmpeg can't recognize their name, in output from ffmpeg - i file.avi I get this "Stream #0:1: Audio: mp3 (U[0][0][0] / 0x0055), 48000 Hz, stereo, s16, 32 kb/s" [20:22] In output file, I get two audio tracks, bat their name is "und" and "und2". [20:23] ffmpeg version? [20:24] ffmpeg version git-2012-10-01-c39916b [20:25] Distro? [20:25] Ubuntu 12.04 [20:26] Are there errors encoding it? [20:27] no errors, everything goes well [20:30] so, can I at least specify name of this stream when I use -map 0:1 [some_name] -map 0:2 [another_name] ? [20:31] can I set bitrate for encoding to wav, aiff and flac? [20:33] Element9: No but with flac you can set -compression_level [20:34] relaxed: thanks! where could I have found that myself? [20:34] Pooky5: can't you use some language metadata? [20:34] and/or title [20:35] like mmh -metadata:s:0 title='stream 0 title' -medatadata:s:1 title=... ? [20:35] ubitux: good point, I will try that and see what happend [20:35] "title" and "language" meta are special iirc [20:35] i don't know if avi is able to store that info [20:36] Element9: nowhere really... It's mentioned in `ffmpeg -h full | less` but nothing specific for flac. [20:36] mkv certainly is [20:36] you can use ffprobe to check [20:36] yea, I convert from avi to mkv [20:37] mkv is wonderful really [20:37] Element9: google or reading the source [20:37] relaxed: or asking here. :) thanks again [20:40] speaking of mkv being wonderful, they state on their website it's designed for live streaming too, i can find not much documentation on how to use it with live streaming though [20:41] i already asked in #matroska but they seem dead, is there any way i can get some documentation on libmatroska? [20:41] ffmpeg doesn't use libmatroska [20:42] there is a native muxer & demuxer [20:42] so the way i understand it i have to send a header prior to the actual stream when joining a stream mid-stream is that correct? [20:42] therefore i have to generate a header on the fly for the stream that is to be sent, right? [20:43] just sending the header of the file doesn't really work, i tried that [20:43] could i utilize the ffmpeg code to do that? [20:44] i don't know matroska [20:45] okay, thanks for reading it though [20:45] should i send a mail to the matroska mailing list? [20:45] do you think that would get me any further? [20:46] no idea, you may want to contact libav-user (https://www.ffmpeg.org/contact.html) [20:47] thanks [21:19] I have an IP camera that uses the RTSP protocol. I can view it in VLC, but ffmpeg says "Unable to find a suitable output format". I know the camera records in h264. How can I make ffmpeg play RTSP streams [21:41] relaxed: http://pastebin.com/gtdXQpD5 [21:43] No, reread it and try again. [21:47] relaxed: http://pastebin.com/Q90eL6vi [21:48] Oh, try ffmpeg -i rtsp://192.168.2.101:554/trackID=1&basic_auth=YWRtaW46YWRtaW4= [21:49] er, sorry [21:49] ffmpeg -i "rtsp://192.168.2.101:554/trackID=1&basic_auth=YWRtaW46YWRtaW4=" [21:52] relaxed: that was step in the right direction, but I'm still not getting video. http://pastebin.com/L52xxbpM [21:53] well you're specifying the input, what do you want to happen from there? [21:54] I was expecting a window to pop up and show me some video [21:54] you want ffplay [21:54] and quote the url [21:55] yeah, the & is getting confused in bash [21:55] Ok, I'll give that a shot. Thanks [21:55] that's the '[1] 24029' [21:55] So should I escape it like this? \& Or will the quotes take care of it? [21:55] quotes will do it [22:09] Hi, I'm using the latest 64 bit ffmpeg static build for windows to encode some fraps files I've created to h.264/mp3 for uploading to youtube. It seems that the archive doesn't come with presets? [22:09] Or has that changed? I read something about them being built in now [22:10] when I try to run an encode without any presets, both my CPU cores spike and after about 30 minutes, only about 350 frames were encoded [22:11] -preset veryslow [22:11] Or whatever other preset you care to use. [22:11] ok, that works [22:12] except I'm seeing the same thing where my CPUs are pegged, and the output in the terminal (I'm using mintty on cygwin) doesn't update [22:12] even killing the process takes a while [22:12] with ^c [22:13] Tried using cmd? [22:13] frame= 87 fps=9.1 q=28.0 size= 443kB time=00:00:00.13 bitrate=27190.6kbits/s <-- hasn't moved [22:13] no because I have this all scripted [22:13] and cmd batch is fail [22:16] I now wonder if writing to the terminal is causing this lag [22:19] yeah, windows shows mintty as not responding [22:21] even if I "> /dev/null 2>&1" the output file barely grows [22:22] If I copy the file over to my ubuntu box it works fine [22:22] processors on that box are much slower though [22:22] hence my desire to encode on this machine [22:22] guess I'll try cmd [22:23] OK, seems I was using the old directory's command line [22:24] so, nm :) [22:27] so which preset would be recommended for high quality YT upload? I have tons of bandwidth, so that's not a problem [22:36] CHA! I got it! [22:36] It was metadata language! [22:53] Primer [22:53] try with -preset and -crf [22:53] yes, I have been using preset slow and crf 22 [22:54] and baseline [22:54] which I confess, I don't even know what they do [22:55] dont use -profile unless you know you need it [22:55] you mean -preset? [22:55] "You should set this if you know your playback device only supports a certain profile. Most decoders support High profile, so there's no need to set this." [22:55] -profile [22:55] and baseline [22:56] baseline is a profile, not preset [22:56] I've never used -profile [22:56] is high profile the default setting [22:56] klaxa, the profile is (I think) auto-determined by the frame size [22:56] ah [22:56] but please read the docs [22:56] to be 100% sure [22:57] Primer, how did you specify "baseline" [22:58] -fpre "f:\ffmpeg-x64\presets\libx264-baseline.ffpreset" [22:58] and I don't even know if that was doing anything [22:58] can you please provide your entire command [22:58] because preset files have been deprecated for libx264 (if that's what you are using) [22:58] time $FF -i "${1}" -acodec flac -vcodec libx264 -fpre "f:\ffmpeg-x64\presets\libx264-slow.ffpreset" -fpre "f:\ffmpeg-x64\presets\libx264-baseline.ffpreset" -crf 22 -threads 0 $OUT 2>&1 | tee -a log [22:59] I see [22:59] again, in cygwin with mintty [22:59] try without -fpres [22:59] actually that's a script [22:59] well, that no longer works [22:59] hence the use of -fpre [22:59] time $FF -i "${1}" -acodec flac -vcodec libx264 -preset slow -crf 22 -threads 0 $OUT 2>&1 | tee -a log [22:59] lovely! [23:00] -preset and -crf are now options of libx264 library [23:00] so no need for external presets [23:00] cool [23:07] https://www.youtube.com/watch?v=roF1kmOWUuY [23:07] just made that with -preset slow -crf 22 [23:07] looks pretty good at the best quality [23:10] looks very fluent too [23:17] I just want to be able to fraps some game content then call a script to encode and upload, and have it look decent on YT [23:19] it's nice being able to do this with a bash script too [23:19] because using any form of windows scripting just sucks [23:21] hah [00:00] --- Wed Oct 3 2012 From burek021 at gmail.com Thu Oct 4 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Thu, 4 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121003 Message-ID: <20121004000502.1A8A918A01F4@apolo.teamnet.rs> [01:34] the mplayer dev list is quite the source of hilarity [01:34] re: mp3lib [01:50] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/x8jRyg [01:50] [FFmpeg/master] 8svx: copy start value in output samples. - Nicolas George [01:54] Daemon404 : you just wish you could fork and maintain a dead library for at least 5 years until someone comes along to maintain it again [01:54] lol [01:54] also [01:54] loling at "performance concernss" [01:54] for fucking mp3 [01:54] lolz [01:54] 10% cpu on a 150mhz is the difference between video and a slideshow [01:54] you know this [01:55] >150 mhz [01:55] i cat hear you back there in 1993 [01:55] cant* [01:56] there are lots of low powered arm devices too [01:56] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/FXGMTA [01:56] [FFmpeg/master] 8svx: fix delta_decode cliping limits - Michael Niedermayer [01:56] Compn, not that low powered [01:56] maybe a first gen ipod [01:56] from 2000 [01:56] or w/e [01:56] pretty much every device mplayer runs on , but that mplayer does not have accelerated output [01:57] e.g. 700mhz arm tablet [01:57] 700mhz is barely fast enough for most videos now [01:57] unless you have mpeg2 . [01:57] ah you know this already [01:57] im reminded of all the bikeshedding over removing dead 3dnow! code [01:57] that was only ever useful on a single cpu [01:57] diego is the one with the k3... [01:58] we cant all have i7s [01:58] :P [01:58] bro [01:59] i can go to the garbage dump [01:59] and find abetter computer [01:59] for free [02:00] Daemon404: you as a bsd lover should know that everybody lot of people have a net4501 at home &. [02:00] o u [02:00] s/everybody// [02:00] i just lol at the neckbeards who cling to ancient hw [02:01] like the slackware main servers running on a p3 until 2012 [02:01] firewall and mp3 enocer in one peace [02:01] cbsrobot_, because ffmpeg on a firewar is surely no security risk [02:01] maybe he pipes the spamd log to mp3 - who knows [02:01] firewall even [02:01] Action: Daemon404 also lols at 'pros' who run irssi on their router [02:01] also very safe! [02:02] apple has even a name for such thing &. [02:04] Daemon404 : anyway, this is curse of mplayer , trying to be the 'fastest' :P [02:05] so we support ancient hacks, and optimizations [02:05] mplayer itself is an ancient hack [02:05] <_< [02:05] am curious what player you use [02:05] depends what os im on. [02:05] ugh [02:05] well what players do you use then [02:06] i use mpc-hc with lav filters (see: nevcairiel) on windows [02:06] and vlc anywhere else [02:06] i use mplayer cross-platform [02:06] :P [02:06] i like having a gui [02:06] with a seekbar [02:06] and such. [02:06] i just bind keys to certain points in the video and never leave my keyboard [02:06] the mouse sits there alone [02:07] fine for you [02:07] not for most humans [02:07] Action: Compn cant believe anyone uses mplayer over vlc [02:07] Action: Compn keeps converting users to vlc [02:08] new strategy, send all users who complain to other software [02:08] enjoy current software with less complaints :) [02:08] oh yeah? you dont like libavcodec?! well why dont you go and use --- nvm [02:08] for a second i thought you were going to say you used quicktime on mac :P [02:09] you asked what player i used [02:09] quicktime is now a player [02:09] not [02:09] * [02:09] does quicktime still make users pay to get fullscreen ? :P [02:09] ala 'quicktime pro' [02:10] dunno [02:10] quicktime x has no pro [02:10] i havent been on a mac in a while [02:10] and lost almost all features qt 1 - 7 had [02:11] apple finally got tired of supporting its own dumb features like putting flash in mov [02:11] and i mean swf flash, not flash video [02:11] cbsrobot_, so it's like fcp x [02:11] killed all usefulness [02:11] yeah - almost as bad [02:12] you still can watch movies though [02:12] so whos ready for apple stock to start lowering when the next iphone starts to suck ? [02:12] no theyll just release a new one [02:12] that sucks 4% less [02:14] any big flamewars i miss this week ? [02:15] no [02:16] I just started one bakeshed. [02:16] im thinking of creating a com[elx versioning system for ffmpeg [02:16] based on solar cycles [02:16] and the mayan calender [02:17] calendar* [02:53] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/58-TAw [02:53] [FFmpeg/master] 8svx: avoid custom clip, avoid +128 for compressed data. - Michael Niedermayer [03:03] Daemon404: actually solar cycles is fine. I think that the sun had a big activity cycle that lasts about 10-11 years, the sun core rotates with about 1 month cycle, so we can use it for regular releases. The best thing... it seems we already follow that scheme :P [03:09] lol [03:16] dont' do one based on the mayan calendar, it's almost over [03:52] Welcome to http://www.multimedia.cx/ [03:52] This domain been suspended, please use the Registration URL link to contact your registrar [03:52] hmm [03:52] old new [03:53] where was the news posted ? [03:53] Action: Compn is out of loop [03:54] cant remember [03:54] mike was notified [03:54] and he paid [03:54] waitning now [03:58] the ole domain hostage situation [03:58] i've seen it before [03:59] i dunno how people miss paying though [04:00] i get 10 metric tons of spam saying "YOUR DOMAIN WILL EXPIRE" [04:00] if it's coming up anywhere close to soon [04:02] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/FfTxAA [04:02] [FFmpeg/master] 8svx: Fixing header size, move decoding to per call instead of the first call. - Justin Ruggles [05:48] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/Hm8MGw [05:48] [FFmpeg/master] ffv1dec: print bps for pict debug too - Michael Niedermayer [09:36] mmh is the date= thing broken in ffserver? [09:39] it seems to seek way ahead the requested pts [09:39] or well, the requested pts is outbounded (>) [09:39] so it seeks at the end [09:39] and thus has no effect whatsoever [09:46] burek: do you remember some users complaining about that? [09:47] maybe i'm using it wrong though [09:48] with this configuration: http://pastie.org/4901196, i'm trying: [09:48] ./ffserver -f ./ffserver.conf [09:48] ./ffmpeg -re -f lavfi -i testsrc 'http://localhost:8090/video.ffm' [09:49] ./ffplay 'http://localhost:8090/video.ts?date=09:47:50' [09:49] with date being a few seconds after starting the push [09:50] and it's leading to something like wanted_pts=1349250470000000 with pts_min=-40000 (pos=4096) pts_max=16720000 (pos=24854528) [09:50] wanted_pts is compared against pts_min/pts_max [09:50] it either a rescale problem or a relative/absolute ts problem [09:51] (imo) [09:51] i thought it was related to 16b9156b7 but doesn't look like to :p [14:12] is there a nice way to get decodec video ptses? [14:13] *decoded [14:13] ffprobe -show_frames perhaps :) [14:15] in the debug code somewheres [14:15] -debug_ts in ffmpeg too [14:15] showinfo filter too maybe [14:16] -show_frames/-show_packets [14:16] infinite possibilities. [14:16] I did -show_frames | grep -A5 video | grep pkt_pts= | head [14:17] I have a file which is behaving a bit strangely. trying to figure out why [14:18] -of flat might be easier to grep :p [14:19] I just want the numbers, so.. :) [15:40] ubitux, i don't remember anything related to that.. :S [15:40] okay [15:41] did you see some users having with the date= thing? [15:43] I've seen ?buffer=X and Preroll [15:43] but not date [15:44] ok [16:42] [FFmpeg] michaelni pushed 7 new commits to master: http://git.io/L3icoA [16:42] [FFmpeg/master] ARM: set Tag_ABI_align_preserved in all asm files - Mans Rullgard [16:42] [FFmpeg/master] ARM: bswap: drop armcc version of av_bswap16() - Mans Rullgard [16:42] [FFmpeg/master] segment: Pass the interrupt callback on to the chained AVFormatContext, too - Martin Storsj? [16:51] Action: Daemon404 wonders why he only found out about app verifier today [16:56] Daemon404 : you like ass subs right ? [16:57] do you like ass in srt? http://pastebin.com/mCYwKkmY [16:57] ehe [16:57] sadly, vsfilter would likely actually render that [16:58] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/90xAjw [16:58] [FFmpeg/master] framecrcenc: print flags and side data elements - Michael Niedermayer [16:58] user reports it working in windows with mpc [16:59] of course [17:00] Italians... [17:05] Compn, because vsfilter is gabest code [17:05] racissssssssst [17:06] i am gabestist [17:06] ? [17:06] ah [17:06] so you support this hybrid new thing ? [17:06] opposite [17:06] I think the only reason it works with GabestCode is because he probably was lazy to limit ass syntax to ass [17:07] and I think everyone agrees to the fact that it's an abomination no matter how you look at it [17:07] well, the biggest issue is that the first line is not escaped in {\ stuff [17:09] Compn: just remove the { } escaping in subassconvert if you want to support that crap [17:09] i explicitely added the escape because i want users to be able to print { } in srt [17:10] ubitux: but wouldn't you print the rest of the first line anyway ? [17:10] since they are outside of { } ? [17:10] ? J'ai fait {Epitech.} ? [17:11] https://encrypted-tbn3.gstatic.com/images?q=tbn:ANd9GcRYyQqRG1-lt3ZeKGnfWLqz6_tLOSQ_4MoE8bap4J3EMZ5RXwSz [17:11] et? [17:11] what happens when you want to print that? [17:11] ? comments [17:11] the more important word disappear [17:11] then, stop escaping { } [17:11] and do as we do [17:11] {\ } escape [17:12] i don't understand [17:12] but still, this is {\something} something else [17:12] um [17:13] how do you not print the "something else" part? [17:13] i would just tell the users to stop using insanely unsupported files [17:13] ubitux : only hide {\ tags with slash} [17:13] the fact that vsfilter even works is due to bugs [17:13] and it not caring what format it parses [17:13] j-b : user said it didnt work with vlc either :P [17:14] Compn: of course not [17:14] Compn: see my question [17:14] ah [17:14] <@j-b> how do you not print the "something else" part? // why would i not print it? [17:14] Compn: VLC would hide the first part {\ xxxx } [17:14] i'm sorry i'm a bit lost [17:14] you want to escape only with "{\" and not "{"? [17:15] ubitux: because the something else part is gibberish numbers ? [17:15] Action: Daemon404 cant believe you people are actually talkign about supporting this abomination [17:16] ubitux : sorry, i havent a clue what needs to be done. i just find bug reports ;P [17:16] its not a bug [17:16] it's always the same things imo [17:16] i didnt say 'valid bug reports' [17:16] lol [17:16] either you escape ass tag, or you don't [17:17] mplayer escape by default [17:17] so libass renders the subs verbatim [17:17] as it should [17:17] Daemon404: for VLC we escape all {\xxx} in srt [17:17] in this case, escape = it gets printed verbatim on screen? [17:17] no [17:17] oh [17:17] we hide those [17:18] then that is wrong. [17:18] :) [17:18] and not teh srt format [17:18] lol [17:18] there is no srt format [17:18] there must be! webvtt is based off it! [17:18] :) [17:18] :D [17:18] anyway, on this abomination, this is worse [17:18] because of {\an9\fscx73\1a&HC0&\3a&HFF&\4a&HFF&\p5} [17:18] m 49 99 s 32 131 21 161 [17:18] Action: Compn lulz [17:18] i'd personnally propose -sub-hack-no-escape [17:18] how do you decide to hide the rest ? [17:18] Action: Compn rolls d12 [17:19] so ass can interpret it directly. [17:19] it was proposed several times [17:19] j-b, you dont and you tell the person who made it to stop making hilariously wrong files [17:19] not sure what we do in ffmpeg btw... [17:19] Daemon404: well, here, sure. [17:19] Daemon404 : whip up some ass drama... keeps everyone on their toes :D [17:20] is like if i started complaining my perl interpreter wont run nodejs code [17:20] people using ass are morons anyway... [17:20] j-b, theres Nothing Better for styled subs [17:21] i dont think using it for typesetting is very nice though. [17:21] Daemon404 : you just dont know the power of RealText [17:21] Daemon404: when will the anime scene switch to photoshopped bitmap subs? [17:21] or vrhtml something -scoobydoo [17:22] ubitux, never? [17:22] :) [17:23] also, re: time text [17:23] it seems everyone and their mom is making their own 'standard' now [17:24] upon the many existing 'standards' [17:24] fun times ahead [17:25] (and come on... embedded css in subtitles? whats not to love? vlc can start linking webkit!) [17:25] Action: ubitux needs to start the new subtitles api [17:25] i need to rewrite almost every text sub codec :( [17:25] (but the encoders will be way more simpler) [17:26] Daemon404 : you ready to include xml parser ? [17:27] everyone needs an xml parser! [17:27] webkit probably does it ? [17:28] no you can't use a xml parser Compn [17:28] i tried. [17:28] it's impossible to parse sami and stuff with it [17:28] Daemon404: WebVTT! TTML 4 different specs! [17:28] \p/ [17:28] \o/ [17:28] \o/ [17:28] Daemon404: http://blog.gingertech.net/2012/09/18/what-is-interoperable-ttml/ [17:28] yeah i read it [17:30] why you guys dont just patch vsfilter to break these files [17:31] obviously its vsfilter fault [17:31] break _his_ legs! [17:31] oh. [18:17] Daemon404 : freakin firefox switched to msvc2010 so the builds wont work on win2k either [18:17] OPERA still works :D [18:17] maybe i could find someone with msvc2005 to make me win2k builds [18:17] of firefox [18:17] :P [18:18] I think they partially updated their compilers because the old ones would run out of memory during compilation or linking? :D [18:19] jerks. [18:19] at least i know what the problem is now [18:19] next time zeranoe is back in here [18:19] well i guess he still uses mingw :\ [18:19] not related [18:20] Compn: you should try linux some time. ;) [18:21] ubuntu wouldnt give me root , but wanted me to login to root whenever i made a change [18:22] kind of upset me :P [18:22] couldnt remember my root pw, had to chroot and such [18:24] you mean, sudo. try slackware, they got 14.0 few days ago. It is nothing like ubuntu, for good or bad :) [18:31] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/w6v4KQ [18:31] [FFmpeg/master] tiffenc: remove unused variable - Michael Niedermayer [18:44] Daemon404: http://www.html5rocks.com/en/tutorials/track/basics/ hmmm JavaScript inside subtitles!! [18:45] >_> [18:45] Action: Daemon404 vomits a bit [18:47] Action: j-b pushes Daemon404 [18:48] why so confrontational [18:49] bro [18:49] jsut to laugh :) [19:36] Daemon404: gix seems to have some trouble with the MSVC thing, the doc/plateform.texi might not be accurate [19:36] maybe you can help him [19:36] < gix> that's what i did. either you guys use a different version of the c99 wrapper or something is missing. it states to make sure that msvc link.exe is used, but the wrapper calls it with -o instead of -out. hardcoding LD_O in the makefile changes that, but then link.exe cannot cope with /tmp-paths [19:37] all you need to do these days is specify --toolchain=msvc and it should work [19:37] wonder whats in that doc [19:37] Action: nevcairiel reads [19:38] it says the same [19:39] did you uild your own wrapper? [19:39] or get a binary from somewhere [19:39] i compiled the current master of the github repo [19:40] (with current llvm/clang master) [19:41] humm [19:41] can you revert the latest commit to the c99 converter repo [19:41] and retyr [19:41] i have a hunch... [19:42] the cosmetics patch? [19:43] yeah [19:43] inb4 i mangled something [19:48] Daemon404: nah, doesn't seem like it. same error running configure --toolchain=msvc: http://codepad.org/8His5uDM [19:48] which link [19:48] inb4 /bin/link.exe [19:48] because msvc's link sure doesnt have --help [19:49] please note i explicitly mentioned this in the deocs [19:49] there's only one link in the path [19:49] the one from msvc11 [19:50] that is wrong [19:50] so please show me teh output of which link [19:50] $ which link [19:50] Try `link --help' for more information. [19:50] ^ wiat [19:50] is this part of our configure [19:50] /c/Program Files (x86)/Microsoft Visual Studio 11.0/VC/BIN/link.exe [19:50] -_- [19:50] yeah [19:51] this is and should be handled by c99wrap.exe [19:51] it's bizzare [19:52] ive never build compilewrap.c with msvc11 though [19:52] only 10 [19:52] must the wrapper be built with msvc? [19:53] i dont think so [19:53] ive not tested it with mingw though [19:53] i compiled it with clang [19:53] oh. [19:53] did you also compile teh converter with clang? [19:54] because that wont work [19:54] libclang and the converter should all be built with msvc [19:54] i compiled it with gcc, and it worked fine [19:55] because of some annoyances with libclang [19:55] at least c99wrap, not the converter [19:55] they hardcode compiler-specific things [19:55] but i guess thats what causing issues here [19:56] Action: Daemon404 out for foo [19:56] d [19:57] well, gonna build c99conv with msvc then. though i don't get why that should matter [20:08] is c99conv buildable with msvc? [20:08] i thought it was written in c99 [20:08] can it convert itself btw? [20:22] ubitux: just built it with msvc without problems (result is the same though) [20:23] ubitux, its not c99 [20:23] that sounds wrong [20:23] many people have been uisng and building it with msvc (msvc10) [20:23] Action: Daemon404 starts to worry about msvc11 [20:24] tho nevcairiel said it works with gcc too [20:24] hello to all [20:24] I see a lot of packets being dropped on RTP over UDP with h264 content [20:24] all the sources I have tried (my server, an IP camera, some internet sources), all have the same problem [20:25] RTP: missed 3 packets [20:27] Daemon404: but shouldn't configure call the linker with -out when the msvc toolchain is set? that's at least what it looks like with LD_O etc. [20:27] c99wrap converts it [20:27] thats the point [20:29] btw is this git master? [20:31] er, my bad [20:31] it does indeed take -out, but ffmpeg's configure DOES set it correctly [20:31] (in git master) [20:36] Daemon404: yes, 05e5a24f7 to be exact [20:37] gix, this is rather strange [20:37] can you post your whole config.log [20:38] im betting it's environmental and/or something we dotn account for [20:38] crtmpserver: you should report bug, this channel is not for reporting bugs [20:38] tx, I will [20:38] sorry for the noise [20:41] Daemon404: http://codepad.org/nfhzb09v [20:42] uh [20:42] looks like it got cut off [20:42] that doesnt contain a single test log [20:42] indeed [20:43] Daemon404: http://gix.myftp.org/config.log [20:51] gix, gonna try something [20:51] sec [20:57] gix, x86 or x64? [20:58] x86 [20:58] k [21:02] okay. [21:02] so it's not the cabac tables! [21:02] :) [21:04] gix, hrm [21:04] works here [21:06] michaelni: can we get bot that reports commits in real-time? [21:07] what happened to CIA? [21:07] microchip_: you missed all the drama.. [21:07] ? [21:08] DCMA takedown [21:08] Oo [21:09] Daemon404: with -o passed to "c99wrap link"? [21:09] microchip_: http://pastebin.com/9RBBniM1 [21:09] you mean DMCA? [21:09] gix, no [21:09] configure detects my compiler correctly [21:09] it isnt detecting yours correctly [21:09] and is using the defaults [21:09] run 'c99wrap cl' and see what it outputs [21:10] ubitux: ic [21:10] the usual output from cl.exe [21:10] drama [21:13] Daemon404: what should "c99wrap link" show? [21:13] the usual output [21:13] the ms logo [21:13] and options [21:13] who is cehoyos [21:13] is he online here [21:13] carl doesnt see the merits of irc [21:13] unfortunately [21:17] Daemon404: i see. that's where it went wrong. and you were right and wrong about link.exe. my wrappers picked up an old link.exe from coreutils in the path the wrappers are stored. [21:17] ah [21:17] isnt this so fun [21:18] the wrappers probably shouldn't do that ;) [21:20] Daemon404: anyway, want a makefile and some compat stuff (inttypes, snprintf, oldnames) to compile the wrapper with msvc? [21:25] no [21:25] i have a msvs solution [21:25] i just havent pushed it yet [21:25] also a README [21:25] for how to build a static clang [21:25] if one is so inclined [21:32] j-b, whats the status of a CIA replacement for ffmpeg git ? [21:39] michaelni: almost ready [21:50] burek : carl reads irc logs tho, and you can mail him :) [21:52] Action: Compn waves at carl [21:54] poor mike [21:54] hope he gets multimedia.cx back [22:18] [FFmpeg] michaelni pushed 2 new commits to master: http://git.io/-aUwhg [22:18] [FFmpeg/master] mux/nut: factorize ff_choose_timebase() out of nut - Michael Niedermayer [22:18] [FFmpeg/master] ff_choose_timebase: only try factors upto 14 - Michael Niedermayer [23:32] michaelni: I think this code in the h264 prober is problematic [23:32] if(p->buf[i+2]&0x0F) [23:32] return 0; [23:32] because those bits are no longer reserved; constraint_set4 and 5 exist now. [23:41] Skyler_, i think i fixed this last year: http://git.videolan.org/?p=ffmpeg.git;a=commitdiff;h=5deedf3552819e5bea3327a450155bb57643a999 [23:42] oh, it's not in the latest version, I see [23:42] Action: Skyler_ goes to slap libav people [23:44] michaelni: the bot for commit is ready. It will be tested on #videolan and pushed here when ready. [23:44] has it colors? *_* [23:45] fabulous colors needed [23:46] wow double rainbow. [23:46] ubitux: yes [23:46] \o/ [23:48] wbs: i see you're doing some segmenter work; did you see the work done in ffmpeg on it? [23:54] ubitux: I didn't, and I'm not really doing work on it for the sake of adding functionality, I'm just trying to clean up whatever code we have there already [23:55] j-b, what is it based off of? [23:55] the bot [23:55] Daemon404: you have objections against a language? [23:55] no? [23:55] im curious [23:55] i had an idea for my own setup [23:55] which may or may not suck [23:55] <-- idea stealer [23:56] Daemon404: ask thresh [23:59] wbs: well there is some kind of interesting features like HLS; also you might be able to find some fixes [23:59] anyway, you obviously do as you wish :p [23:59] ubitux: as said, that was not my current target [00:00] --- Thu Oct 4 2012 From burek021 at gmail.com Thu Oct 4 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Thu, 4 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121003 Message-ID: <20121004000501.15C9918A01F1@apolo.teamnet.rs> [00:10] hey guys, i am trying to convert a .wmv to avi [00:10] lol [00:11] is that voodoo [00:11] if you elaborate the specifics a bit mo-- wait are you trolling? [00:11] manizzle you want to convert the streams inside the containers (they are both containers) [00:12] but afaik an avi takes the same streams wmv containers do [00:12] my guess would be he wants XviD/mp3 in the end [00:13] someone said the mss2 can be converted, but the wmavoice needs to be run in windows [00:13] stripped and rejoined to the mss2->xvid [00:13] is that the only solution? [00:13] WMAPro-in-WMAVoice is fucked up yeah? [00:15] can ffmpeg decode that? [00:15] wait what [00:16] how did you get a codec into a codec [00:18] what did you try so far manizzle? [00:18] because from the command line help it looks like ffmpeg should be able to decode all the streams you named [00:19] http://sprunge.us/ceAU [00:20] hmm... try the git version, i doubt it will work though, otherwise try contacting the devs, either here or mailing list or something [00:22] any devs here? [00:22] he have 1.0 that is quite recent. [00:23] yeah therefore the git version might not be a lot different [00:23] so which ffmpeg- mailing list should i email? [00:23] manizzle: does mplayer play it? [00:23] http://ffmpeg.org/contact.html [00:23] lemme try [00:23] you could try to dump raw pcm and use that instead [00:24] manizzle: there may be another solution, while you are waiting for this to be implemented. and if it requires RE, it may take a while. [00:24] k [00:25] 32 bit mplayer and mencoder can load win32 binaries. It may be a lot trickier to build and install needed files. [00:25] um, so i play with mplayer i can hear audio [00:25] but i cant see the video [00:25] xine should also have win32 loader, not sure about vlc (and can't help with them). [00:26] mplayer says no video [00:26] is it maybe the video stream thats fucked up then? [00:26] hum.. let me check. [00:27] http://fitzcarraldoblog.wordpress.com/2011/07/13/how-to-play-mss2-codec-windows-media-video-9-screen-wmv-files-in-64-bit-linux/ [00:27] maybe the mss2 shit is the fucked up crap [00:27] mplayer uses ffmpeg mss2 decoder [00:27] you just need a recent version (1.1 ?) [00:28] so the git version would help? [00:28] i have 1.0 [00:28] ffmpeg version 1.0 Copyright (c) 2000-2012 the FFmpeg developers [00:28] built on Sep 29 2012 11:22:50 with gcc 4.7.1 (GCC) 20120721 (prerelease) [00:28] i mean, mplayer 1.1 [00:29] MPlayer SVN-r35014-4.7.1 (C) 2000-2012 MPlayer Team [00:29] manizzle you *could* try this: https://gist.github.com/ee9bdf4d7fcffa6dec29 [00:29] if it's not the mss2 thing fucking up [00:31] k mplayer2 works [00:32] 35014 is from jully 6 [00:33] MPlayer2 UNKNOWN (C) 2000-2012 MPlayer Team [00:33] yeee [00:33] it works [00:33] the mss2 support have been added few days ago. [00:33] noice [00:33] yeah vlc doesnt have mss2 support yet [00:33] oh well [00:34] mplayer for now [00:35] http://git.mplayer2.org/mplayer2/log/?qt=grep&q=mss2 [00:35] > [00:35] ? [00:37] mplayer2 is fork of mplayer [00:40] hum, you are right, mplayer should also support it with the win32 loader. can you try mplayer and put the text output in a pastebin site and give the url to me? [00:42] ill install mplayer svn [00:55] yeah mplayer-svn works too [00:55] MPlayer SVN-r35226-4.7.1 (C) 2000-2012 MPlayer Team [00:55] 198 audio & 412 video codecs [00:55] cool thanks [00:56] but i think i like mplayer2 better, so switching back to mplayer2-git [05:22] Does anyone know where to get mingw-w32api? [05:22] I need it for a cross-compile but it's not listed on the repository. [07:13] Hi [07:17] I'm having some trouble compiling ffmpeg on a x86_64 Fedora 11 machine [07:19] It fails when attempting to link ffmpeg_g with repeated undefined reference to `av_lfg_init' [07:20] I'm guessing I'm doing something stupid, but I can't figure out what [07:24] Tesseract433, did you checkout from git recently? [07:25] yeah [07:25] there were some massive compile errors on fate, and I guess they are fixed now [07:25] so try to git pull/checkout again [07:26] I'm on the latest commit [07:30] still having same issues? [07:31] can you pastebin the config.log [07:32] Yeah, still have the same issue [07:32] I just tried the 1.0 release and it linked fine [07:44] hello, can someone help me out with this ffmpeg error (http://pastebin.com/XYXj6WYq). The error is "Could not write header for output file #0 (incorrect codec parameters ?): Invalid argument". I have just pulled the trunk version of ffmpeg from git and compiled it. I need to use the m3u8 live segmenting that was added a few weeks ago. I'm not an ffmpeg expert, so not sure where to go from here. Thanks. [07:51] sorry i got disconnected. I'm goign to ask my question again. \ [07:52] i'm getting the error "Could not write header for output file #0 (incorrect codec parameters ?): Invalid argument". You can see the entire output here: http://pastebin.com/XYXj6WYq - I'm cloned from git the latest dev version, because I need to use the m3u8 live playlist generation. But I don't know what the error i'm getting is and how to go about fixing that issue. [07:52] any help is appreciated [08:06] i think i've forgotten to include codecs... maybe... [08:06] that's when i ./configure'd the src [08:11] hi [08:11] how can i watermark a video with the movie filter and keep the rest exactly the same (same bitrate...)? [08:12] /usr/local/bin/ffmpeg -i file.mp4 -sameq -vf "movie=watermark2.png [watermark]; [in][watermark] overlay=main_w-overlay_w-10:main_h-overlay_h-10 [out]" file_out.mp4 changes the bitrate from 900something to 3600something [08:12] you can't without re-encoding unless you make it as subtitles which are removable though [08:13] because of the -sameq flag [08:13] it keeps the same quality, therefore uses high bitrates [08:15] so i have to re-encode with the exactly the same birate and add the watermarking? [08:15] i think that would reduce quality maybe [08:16] but try it and see if the result is still of reasonably good quality [08:17] just removing -sameq works [08:17] just tested [08:20] thanks for helping [08:22] ok i ./configured and compiled with encoders and i used them in my ffmpeg command. Now i'm getting Could not write header for output file #0 (incorrect codec parameters ?): Invalid argument [08:24] complete log? [08:24] i think there was this... [08:30] oh sorry you already posted it... you're using a lot of functions i'm not familiar with... um... my guess would be (and this is probably wrong) to move -map 0 before the -vcodec copy -acodec copy [08:30] actually that makes no sense [08:30] disregard that [12:16] hey all [12:16] i'm trying to do a screen capture with x11grab and alsa [12:18] but I can't get alsa input working [12:19] it says the device or resource is busy, and no idea where to [12:19] start (at fixing it) [12:20] http://pastebin.com/MK1x5ZQV - here's the command and its output [12:23] http://pastebin.com/v8ihNW2U - here's my .asoundrc, if that's of any use [13:18] well i managed to get it to start by killing everything sound related [13:19] then the video just contained random background noise, and when i tried playing something, it went silent [14:59] Hi all: I'm converting an .flv recorded from FMS 4.0 w/Nellymoser to wav in order to add audio back to AviSynth and attach to video. When I convert the duration changes from 1:28 to 0:23. Any way I can grab the whole thing? Thanks [15:05] hello may i ask why when i do a screencast with video and audio i get alsa.buffer.xrun ? [15:06] I have 23.98 fps progressive material that I need to convert to 29.97 for broadcast. Does ffmpeg have a filter or process for converting through pulldown? [15:38] I did some research and it looks like it can't handle the silence. The conversion process only reads the audio when the user's microphone was active. [15:38] Does anyone know have to handle silence in Nellymoser audio conversion? It's changing the duration of my audio. [15:40] King_Rat http://ffmpeg.org/trac/ffmpeg/wiki/Capturing%20audio%20with%20FFmpeg%20and%20ALSA [15:40] also http://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20grab%20the%20desktop%20(screen)%20with%20FFmpeg [15:41] ibsk8in31 can you please use a pastebin site (like www.pastebin.com) to show your ffmpeg command and the complete console output? [15:41] tuxhat, sync issues probably [15:42] dericed, what is a "pulldown" ? [15:43] http://en.wikipedia.org/wiki/Three-two_pull_down [15:43] it's basically the act of making extra fields to make content be 30000/1001 instead of 24000/1001fps [15:43] it adds frames by repeating fields rather than frames. [15:44] yup [15:44] burek: When i use 'ffmpeg input -r ntsc output' to convert from 23.98 to 29.97 the result looks like it stutters [15:44] it's because it's doing rate conversion and not "proper" pulldown [15:44] also yes, it should look like stuttering to be honest [15:45] because you get extra pictures that shouldn't be there [15:45] I was trying to mimic pulldown by using yadif to convert frames to fields, then the fps filter (to dup or drop to convert the rate), then convert fields back to frames. is this sane? [15:45] dericed, maybe asking in #ffmpeg-devel would help? [15:45] dericed, sounds like an overcomplicated way of doing it :s and possibly incorrect [15:46] JEEB: this was occurring to me [15:46] JEEB: maybe I need to pipe to mplayer and back [15:46] I know avisynth has the capabilities to do both pulldown and pullup, but that you'd have to learn, too :s [15:47] I would rather decompress everything to deinterlaced raw video and encoded it back to normal video I need [15:47] burek: ? [15:47] burek, it's the opposite [15:47] he has progressive [15:47] 24000/1001 content [15:47] oh i see [15:47] he needs to make it 30000/1001 interlaced [15:47] well, anyway you have some information missing [15:48] so artifacts are expected [15:48] like scaling a small image to a bigger one [15:48] http://pastebin.com/MUwcX7DG [15:49] burek, you don't really lose information with the thing by itself as you're just separating to fields and adding extra fields to make it 30000/1001 [15:49] I never said you loose anything [15:49] > you have some information missing, so artifacts are expected [15:50] yes, you have missing frames [15:50] what? [15:50] I'm pretty sure that when you put those fields back together when doing inverse telecine, you will get the original chroma data unless you do lossy compression in the middle, but of course we were just talking about the processes of pullup and pulldown [15:50] burek: When i use 'ffmpeg input -r ntsc output' to convert from 23.98 to 29.97 the result looks like it stutters [15:50] and you don't lose any frames [15:50] 24 <> 30 [15:51] yes, you get extra frames, and that's why it stutters because fluid motion isn't fluid any more [15:51] exactly what I said [15:51] it's not "information missing" [15:51] :) [15:51] and why "extra frames" if nothing is missing [15:51] because you're converting from a lower frame rate to a higher? [15:52] instead of adjusting the video speed [15:52] exactly my point, again [15:52] what? [15:52] burek: because I need to achieve 29.97 fps for compatability with a broadcast system [15:52] missing frames [15:52] what? [15:52] I guess you are very badly saying that you're not having 24->48 [15:52] and thus you lack the "everything's double" to get fluid motion [15:52] it's pretty obvious that 30-24 = 6 frames missing each second [15:53] I don't see the point in this arguing at all [15:53] Nellymoser losing duration of audio issue: Do I use -async to force packets even when there is silence? http://pastebin.com/MUwcX7DG [15:53] burek, it's 24->30 [15:53] not 30->24 [15:53] thus you get 6 /extra/ frames [15:54] again.. " burek: When i use 'ffmpeg input -r ntsc output' to convert from 23.98 to 29.97 the result looks like it stutters" [15:54] yes [15:54] -r ntsc is 30000/1001 [15:54] not 24000/1001 [15:54] "from 23.98 to 29.97" [15:54] from 24000/1001 to 30000/1001 [15:54] yes [15:54] do you know how to read at all? [15:54] yes [15:54] and it looks like 30-24 not 24-30 extra frames [15:54] see the from and to [15:55] you are converting to something bigger than the input [15:55] dericed: now that everything is clarified, any suggestion to improve on -r. My next guess is to pipe to mencoder and back. [15:55] ibsk8in31, did you try to convert the input, keeping the video stream too, using wav for audio stream [15:55] just to see if it will produce the correct timings [15:55] and then just get rid of the video [15:55] dericed, mplayer /could/ do what you want but I have absolutely no experience there. I don't think ffmpeg can do pulldown (it can do ivtc nowadays tho, IIRC) :/ [15:56] geez JEEB, you didn't even read me [15:56] I said exactly the same thing [15:56] no, I did read you [15:56] if he has got 24 frames [15:56] it's pretty obvious that 30-24 = 6 frames missing each second [15:56] and he wants 30 frames [15:56] I have tried that method with DirectShowSource() in Avisynth. how can I keep the video stream in FFMPEG too? [15:56] he's got 6 frames missing [15:56] what's not clear in that statement [15:56] JEEB: no ffmpeg doesn't do ivtc. See the ticket here: https://ffmpeg.org/trac/ffmpeg/ticket/681 [15:56] oh you meant it that way [15:56] that was sure a backwards way of saying it [15:57] well, he surely doesn't have 6 MORE frames, does he? :) [15:57] you should've just said that the added six frames cause the stuttering [15:57] End goal is to stitch two .flv's recorded by Flash Media Server together side-by-side w/audio. However, I lose audio packets ;( [15:57] because when you talk about missing frames [15:57] it usually means they are getting lost [15:57] his input has less information than it's needed to create desired output, that's the correct statement? [15:57] yes, but it isn't being lost [15:57] it just never was there [15:58] the idea of losing information usually means that you had it to begin with [15:58] [15:50:23] I never said you loose anything [15:58] you misread it somehow [15:58] he's got 6 frames missing [15:58] I guess I misread this as "lost" [15:58] ok :) [15:59] but let's just say that you didn't put it in the usual way this stuff is conversed about :D [15:59] if something's missing, you've usually lost it [15:59] etc. etc. [15:59] ibsk8in31, try something like [15:59] dericed, ok -- too bad [15:59] I'll try to be more clear. I have 24000/1001 but need to convert to 30000/1001. Thus I need 6 frames ADDED. Right now ffmpeg makes new frames by duplication which causes a stuttery look. In telecine work, there is a technique called 3:2 pulldown to duplicate fields rather than frames for a less stuttery look. Can ffmpeg do something like 3:2 pulldown. [16:00] dericed, yes I know that [16:00] I'm just telling you that ffmpeg itself probably doesn't have pulldown [16:00] so you'd have to use that mplayer you mention, if that works, or avisynth [16:00] for that [16:00] I only have avisynth experience mysefl [16:00] ffmpeg -i input.flv -acodec pcm_s16le -ar 8K -ac 1 output.flv [16:00] *myself [16:00] or avi [16:00] Ok thank you. Will try now [16:00] output.avi [16:01] I have a self captured mpg File from Composite In and Audio In. Audio and Video have a little delay which I'd like to correct. I tried it with following cli (+ Log) http://pastebin.com/0mXKbwLJ But when I play this file with vlc e.g. I get no timecode at all it stays at 00:00. And ffmpeg -i shows me "0 channels" at the Audio Stream. What I'm doing wrong? Or could I even get the capture at the capturing process sync? [16:01] burek: https://ffmpeg.org/trac/ffmpeg/ticket/1782 :) [16:02] +1 :) [16:02] and it's cousin request: https://ffmpeg.org/trac/ffmpeg/ticket/681 [16:02] however, I'm not sure how nowadays video editors handle frame duplication process.. can it be specified which algorithm to use or not.. [16:03] so it might be a good idea to use a video editor for such things, if it supports such specification [16:04] Stefff, with -itsoffset it does matter what -i comes first [16:04] it defines what will get "moved" in time [16:04] which* [16:05] i.e. avoid using negative values for -itsoffset [16:05] as it can have unpredictable results [16:06] ok, I try the other way around [16:06] btw, what are these "buffer underflow" errors [16:06] I dont know. [16:06] try without -vcodec copy -acodec copy, try re-encoding, to see if it works [16:08] Stefff, one more thing, if you plan to play around with recorded material, try to capture it to something lossless [16:08] for the future usage, I mean [16:10] yeah, maybe we talked about it a while ago. I tried a little with lossless, but it would need to much space to archive it and the reencode later some other outputs from that. [16:11] well it all depends what is your goal [16:11] if the goal is the best image quality, the fastest encoding speed and the best compression ration, then you are in trouble :) [16:11] raio* [16:11] ... [16:11] ratio :) [16:11] dericed, an example of what you want to do is available on http://avisynth.org/mediawiki/Select#SelectEvery [16:11] but if you think mplayer can do it, too, great :) [16:13] burek: I tried ffmpeg -i input.flv -acodec pcm_s16le -ar 8K -ac 1 output.flv -- It produced an .flv of the correct length. [16:14] now just do: ffmpeg -i output.flv -vn output.wav [16:14] When I converted that .flv to .wav, I lost packets again! The duration of the final file was only 0:35 [16:14] or: ffmpeg -i output.flv -vn -acodec copy output.wav [16:15] I see [16:15] ok sweet. what does -vn do? [16:15] did you try playing output.flv [16:15] indeed. My goal is to archive VHS Cassettes on a 3ghz pentium4 (i think). So I need fast encoding and the (nearly) best image quality I can get with it. So my choice was mpeg2 because the hardware is capable to handle the realtime encoding without an framedrop. [16:15] try ffvhuff or something [16:15] should be rather quick :) [16:15] -vn means "no video" [16:15] it removes video stream [16:15] naturally lossless formats are also relatively bigger [16:16] which means 50-100 times bigger :) [16:16] Still 0:35 when using -vn. Trying -acodec copy [16:16] Stefff, so you need 1) the best image quality, 2) fast encoding and 3) good compression rate (since you are archiving) [16:16] those 3 never go together :) [16:17] yepp :) [16:17] 2/3 you can always manage at the cost of the 3rd :) [16:17] yepp [16:17] Burek, still 0:35 :( [16:18] so I think mpeg2 with 10Mbit Bitrate is an 80% Solution. [16:21] Stefff, you could file a bug report maybe [16:22] and provide a sample of your input? [16:22] I dont think its a ffmpeg bug. Chances are better that I need the Bugreport ;) [16:25] I use this cli http://pastebin.com/gaTg7y8x for capturing from a tvcard. Are those decent setting for mpeg2? [17:17] burek: It's also still chopping all the silence when converting to .mp3 [17:17] ibsk8in31, can you submit a bug report regarding your issue please? [17:18] I'm doing a lot of searching, do you think it's my microphone settings on the flash client side? [17:20] I don't know, but I wouldn't encode audio/video while doing a capture [17:21] I would probably do it like this: http://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20grab%20the%20desktop%20(screen)%20with%20FFmpeg [17:24] Oh I'm recording a conversation (Webcam and Microphone) through Flash Media Server, not screen capturing. [17:24] http://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20capture%20a%20webcam%20input [17:32] Thanks for the links, but I need this to be on a webpage like a flash object is. Webuser A and Webuser B -> Record the conversation & Archive [17:33] I'm getting a message that the VBV buffer isn't set. How can I set it? I'm already using -bufsize and that didn't help [17:40] JoeyJoeJo: pastebin [18:19] Hello. I created 15 thumbnails from a large video (over 4B) using this command : ffmpeg -i /tmp/media.ts -ss 00:08:33 -r 1/513 -s qvga -vframes 15 tn/%02d.png. But it took 18 minutes and a lot of memory and cpu, is there any way to optimize the process ? [18:21] try moving the seek before the input [18:24] relaxed: Here is my command and its output - http://pastebin.com/Zv07qi4D [18:28] thanks relaxed, i'll try [18:28] JoeyJoeJo, you need both bufsize and maxrate for vbv [18:31] knoob<< yes, get a better computer, and it will be faster [18:32] ( you can't just seek in a video file to any place that has keyframes ) [18:33] of course it will be faster but it's a core i5 [18:34] 18 minutes using an i5 sounds much [18:35] Hi ppl, i got a small question/pblm [18:35] yeah and it consumed like 4GB of RAM and 3GB of swap [18:35] is it a bluray disc? [18:36] no just a recorded stream on a regular HDD [18:36] how come that if I use this command line "ffmpeg -i src.avi -vcodec copy -acodec copy dest.avi" [18:36] i get a differrent bpp in src.avi and dest.avi [18:37] 12 in src.avi, 24 in dest.avi, reported by mplayer [18:38] knoob<< here are 2 modes seek and skip mode, skip mode plays the file and skips until specified time [18:38] (i use ffmpeg version 0.8.3-4:0.8.3-0ubuntu0.12.04.1, it may not help) [18:39] mh ok [18:39] ss stands for seek set right ? [18:39] does it actually skip? [18:40] i remember you have to put commands before or after -i to set skip or seek mode [18:41] Nils`: this is not ffmpeg [18:41] it's an old version of a forked project [18:41] ok, what relaxed suggested should speed up the process [18:42] Nils`: please try with ffmpeg, we can't support that fork here [18:42] ubitux: so i'll better compile the latest source and retry ? [18:42] ok [18:42] I'll give it a try [18:42] Nils`: at least we might be able to help you if it doesn't work [18:42] knoob<< in seeking mode i guess ffmpeg extracts the next keyframe [18:44] ubitux: thanks for the help. in 'real' ffmpeg, is there any way to control bpp apart from using pix_fmt ? [18:45] doesn't it do what you want? [18:46] creep: ok but what ffmpeg does when I say "-b 1/513" does it skip or seek ? [18:48] i'll give it a try with a fresh version once compiled. but in my ubuntu-packaged version, for example, according to the doc, yuv420p is meant to have 12bpp color depth, and in the re-encoded version, mplayer says it's 24bpp [18:49] the original file, when played in mplayer, has 12bpp [18:49] i'll have a look to your cmd line output [18:49] when you'll upgrade :) [18:49] sure. thx [18:53] JEEB: I tried adding bufsize and maxrate and I got a new error "MPEG1/2 does not support 15/1 fps" - http://pastebin.com/DMQ3UrB1 [19:00] well, I think that pretty much says what it means on the tin :) [19:07] I'm getting another error from libavcodec. This error is in my syslog and is generated from zoneminder - ERR [Unable to read packet from stream 0: error -541478725] [19:07] I'm trying to view the RTSP video stream from an IP camera [19:28] ubitux: same result with the latest ffmpeg [19:28] compiled from the git source [19:29] though, i cannot exclude a pebkac in my ffmpeg command line ;) [19:30] :) [19:30] has anyone tried building ffmpeg with msvc and the c99 wrapper? what magic is needed so that configure (especially the cc/ld checks) completes? [19:31] gix: check doc/plateform.texi [19:35] that's what i did. either you guys use a different version of the c99 wrapper or something is missing. it states to make sure that msvc link.exe is used, but the wrapper calls it with -o instead of -out. hardcoding LD_O in the makefile changes that, but then link.exe cannot cope with /tmp-paths [19:39] ubitux: http://pastebin.com/KC1KQxRD [19:42] Howdy - is it possible to output to a fifo? [19:42] Perhaps using stdout and redirecting to a fifo? [19:45] named pipe? on *nix I /think/ you should be able to just output to the file [19:47] Nils`: looks strange to me [19:47] are you sure it's not a bug in mplayer? [19:48] this idea came to my mind. actually i'm not sure [19:48] is there a way to check this with ffmpeg ? [19:49] you can ffprobe the file [19:49] you'll see the pixfmt matches [19:52] that information could be added [19:52] give me a few minutes [19:53] here you go : http://pastebin.com/nU9DWKx7 [19:54] not sure if it's related, but the bitrate seems almost doubled in the destination video [19:54] compared to the source [19:56] crap, it was the system's ffprobe [19:56] with latest ffprobe http://pastebin.com/4pdtXvD0 [20:03] bits_per_pixel=12 [20:03] seems to be ok to me [20:04] i have no idea why mplayer displays this [20:04] but i tried this: http://pastie.org/4903834 [20:04] and ./ffprobe -show_streams raises 12 [20:05] which is normal, since the bits_per_pixel is computed based on the pix fmt [20:07] i'll patch & test [20:09] bits_per_pixel=12 [20:10] but it's retrieved from the pix fmt [20:11] i don't think it would be even possible for ffmpeg to decode that file if that bpp was different [20:12] so it makes sense to get the same result as the ffmpeg -pix_fmts [20:12] so i'd really go for a printing problem with mplayer [20:12] ok [20:12] so the fact that the bitrate is doubled [20:13] is unrelated to this ? [20:18] hello [20:18] i have a .mkv HD quality [20:18] ubitux: i have to go, thx for your help [20:19] i need to convert it to .avi while keeping quailty [20:19] and you've got some greetings from manfred [20:19] im on ubuntu btw [20:20] t4nk730: ffmpeg -i input -c:v mpeg4 -vtag "xvid" -q:v 3 -c:a libmp3lame -ac 2 -b:a 192k output.avi [20:21] Nils`: haha ok ;) [20:21] thanks relaxed going to give it a try now [20:23] what about ffmpeg -i output.avi -vcodec mpeg4 -b 4000k -acodec mp2 -ab 320k converted.avi ? [20:36] why would you encode the avi? [20:36] im just doing it so i can play it back on my ps3 [20:42] The ps3 can decode h264 level 4.1 so you may just need to remux the streams into the mp4 container. [20:43] so that command wont work for ps3? [20:44] pastebin the output of ffmpeg -i input.mkv [20:46] http://pastebin.com/r2TZK0cj [20:47] do you require the subs? [20:48] no [20:50] ffmpeg -i input.mkv -c:v copy -c:a libfdk_aac -ac 2 -b:a 256k output.mp4 [20:51] it may need to be, ffmpeg -i input.mkv -map 0:v -map 0:a -c:v copy -c:a libfdk_aac -ac 2 -b:a 256k output.mp4 [20:51] ok gonna try that one [20:53] will this keep the HD quality? [20:54] yes, you're copying the video and encoding the audio to aac [20:55] thanks alot gonna wait untill this finish gonna take a min [20:58] think im gonna do this for now own instead of the shitty mediaserver [21:00] is there a way i can make subtitles only on non english parts? [21:00] or ill have to do it for the whole movie? [21:04] With the 1.0 release of ffmpeg, a new file named "time.h" had been added to the libavutil folder. This is conflicting with the ANSI time.h when trying to build ffmpeg. Has anybody found a work-around for this other than changing the ffmpeg code? [21:06] bitpurity: this is known and has only one comment thus far https://ffmpeg.org/trac/ffmpeg/ticket/1783#comment:1 [21:09] microchip_: thanks. I was using the wrong search option on the ffmpeg site [23:32] In Avisynth, Is there any way to be selective about what gets displayed for .Info()? [00:00] --- Thu Oct 4 2012 From burek021 at gmail.com Fri Oct 5 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Fri, 5 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121004 Message-ID: <20121005000502.268E218A02C9@apolo.teamnet.rs> [00:02] michaelni, wbs: if it is too complicate to merge segment patches from martin, i'll try to merge them later [00:03] my only problem is that i don't have much time to spend on ffmpeg in this period [01:34] [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/QIEtPQ [01:34] [FFmpeg/master] movenc: force video timebase to be 0.1ms precisse at least. - Michael Niedermayer [03:10] [FFmpeg] michaelni pushed 2 new commits to master: http://git.io/YNs3jA [03:10] [FFmpeg/master] afq: update remaining samples variable. - Michael Niedermayer [03:10] [FFmpeg/master] afq: sanity assert on remaining_samples - Michael Niedermayer [06:55] >.> [06:55] i can't believe people are -supporting- a file named time.h [09:11] wow it's crazy [09:11] the ffmpeg h264 fate tests are suddenly waaaay faster in [09:11] helgrind [09:13] or maybe it's a placebo [09:17] derp, mistake. [10:06] thresh: so you're here to bring an awesome git rainbow bot? :) [10:06] exactly [10:06] thresh: why are you not in siberia? [10:07] http://www.wired.co.uk/news/archive/2012-09/18/russian-diamond-smorgasbord [10:07] av500: those are technical diamonds, not fancy enough for us Moscovites [10:08] ubitux: got anything to push? :) [10:08] it should work now [10:08] mmh i don't think so, but lemme check [10:09] Action: ubitux is looking for a typo [10:14] well i'll wait instead :) [13:30] thresh: michael pushed some stuff [13:30] doesn't seem to appear [13:31] yeah I've just seen. [13:31] michaelni: can you show me the output of git push? [13:43] thresh, last git push http://pastebin.com/y00wR2w0 [13:44] weird, but thanks [13:45] maybe it's a problem with the dns redirect? [13:45] no, I've got it [13:47] I think I fixed it [14:34] http://www.flickr.com/photos/flameeyes/7906103846/lightbox/ VDD12 [14:34] Daemon404: where are you btw? :) [14:35] since i "missed" you then :p [14:39] [FFmpeg] michaelni pushed 25 new commits to master: http://git.io/3-62Aw [14:39] [FFmpeg/master] blowfish: Factorize testing into a separate function - Martin Storsj? [14:39] [FFmpeg/master] blowfish: Fix CBC decryption with dst==src - Martin Storsj? [14:39] [FFmpeg/master] blowfish: Add more tests - Martin Storsj? [15:00] thresh: cone? [15:00] j-b: it will join once there is a commit to show [15:01] you provide a channel to spam to in a hook, not in a daemon instance [15:01] oh, nice [15:32] j-b, thresh http://pastebin.com/gimSnw4T [15:32] oups.. [15:32] michaelni: well, well well [15:33] michaelni: re-push, please [15:33] ffmpeg.git 3Michael Niedermayer 7libavcodec/mpeg12enc.c libavcodec/mpegvideo.h: mpeg2enc: support and use frame_rate_ext when needed [15:36] done, thx [15:36] ive also disabled the github KGB [16:01] michaelni: frame.data[X] for planar format seems to not come one after another [16:03] so instead of using single malloc function i need to use each one per channel [16:04] issue doesn't seem to happen with 16bit but does for 24(32) [16:04] why would you malloc the channels manually, there is get_buffer to do that [16:05] i need that for < 24 bits per sample [16:12] huh [16:12] why does that make a difference [16:13] in > 16 bit case i do not need extra buffer, i just use frame [16:47] michaelni: where can I find a file with repeat_pict != 0? [16:48] saste, some telecine stuff should use it [16:49] ubitux, you can only see my head [16:49] middle-ish... [16:49] michaelni, so the question turns to: where can I find some telecine stuff? ;-) [16:51] samples archive probably, i dont know a specific file ATM either :( [16:51] uhm... well I'll have to do a brute force scan [16:51] something in fate samples? [16:52] sawe can narrow it down a little [16:52] it's mostly used on DVDs [16:52] and very unlikely in TS streams [16:53] as these tend to be real interlaced norally ;) [16:54] saste, maybe iive or rich know a sample [16:54] uhm so in some DVD vobs? [16:54] I'd expect it there, yes [16:54] but even there it is unlikely [16:54] almost all anime dvds ever [16:54] for one [16:54] you are looking for a NTSC DVD with both 24p and 30i content [16:54] repeat_pict, that was used for software telecine? [16:55] yes [16:55] Action: saste realizes how useful -t option would be in ffprobe [16:55] protip: anything that isnt TIVTC is crap for ivtc [16:55] we need a "ffprobe -find-the-sample-i-want" option :) [16:55] hehe [16:56] we have Compn for that. [16:56] maybe we need a ffprobe index of all fate samples :P [16:56] michaelni, i'm going to try something link ffprobe -t 2 -show_frames | grep pict_type | grep -v pict_type=0 [16:56] but i need to implement -t it will take forever [16:56] *or it will ... [16:57] saste, you could just av_assert0() on the value you want and then run ffmpeg -f null - over all and search for a Abort in the output [16:57] ( ffprobe -show_frames | grep pict_type | grep -v pict_type=0 ) & sleep 1; killall ffprobe ;) [16:58] also a stream selector would be useful (and maybe i should replace -t 2 with -n 20) [16:59] saste: i can upload you some telecine dvb samples if you want [16:59] nevcairiel, yes that would help [16:59] soft telecine is what you want, right? so repeat flags in the stream [17:02] saste: http://files.1f0.de/samples/Dext.mpg and http://files.1f0.de/samples/eviltrees.ts [17:03] both contain repeat flags [17:04] nevcairiel, thanks [17:10] Daemon404: can you give some more hints, or show me somehow? :) [17:11] maybe is mspaint-circle it in a bit [17:11] ;p [17:11] s/is/ill/ [17:18] thx :) [17:18] Daemon404: version it, so we can contribute by adding some nicks [17:27] isnt it easier to use something with a tagging ability [17:27] does flickr have that? [17:28] no idea [17:28] flickr has [17:29] youd have to contact diego to do that [17:29] hes not usually on irc, i think [17:29] Daemon404 : avtime.h ? :P [17:31] Daemon404 : dondiego is on irc right now (if you mean diego biurrun ) [17:32] no i mean flameeyes [17:36] ffprobe -select_streams or -specify_streams? [17:37] or both through an alias... [17:38] oh [17:38] hehe [17:41] ffmpeg.git 3Michael Niedermayer 7ffplay.c: ffplay: add a 3rd state to infbuf for autodetection [17:41] ffmpeg.git 3Michael Niedermayer 7ffplay.c: ffplay: autodetect realtime streams and enable infbuf [17:43] michaelni: what's with this indent? :DD [17:53] michaelni, who owns the nut trademark? [17:55] apple? [17:59] :)) [18:31] michaelni: cone-205 works? [18:50] j-b, so far it seems it works [18:51] michaelni: cool. Anything else? [18:55] atm it seems all fine [21:35] ffmpeg.git 3Tim Nicholson 7libavformat/movenc.c tests/ref/fate/acodec-pcm-s24be tests/ref/fate/acodec-pcm-s32be: movenc.c: Add support for >16bit BE flavours [21:53] michaelni: http://guru.multimedia.cx/pp-vs-spp-filters/ does not work.. [21:54] michaelni: i guess its related to http://multimedia.cx [22:06] oh fuck.. that mail was from first oct.. ;S [22:06] bah. fail [22:08] lul [22:08] thine virginal earholes [22:49] burek: how do you envision ffmpeg would supply an official deb package? PPA? [23:08] llogan, you mean this: https://help.launchpad.net/Packaging/PPA ? [23:12] I was thinking more like doing it with our own infrastructure, but if launchpad is willing to build for those architectures itself (x86, AMD64 and LPIA), sure, why not [23:12] even less trouble [23:14] the goal is to provide the alternative to debian/ubuntu users, who still favor ffmpeg over avconv [23:18] burek: i don't know the capabilities of launchpad. i'm just wondering what would be easiest for users. [23:21] well, imho, it is more convenient to have an apt repository (like launchpad), but even providing only a deb package (to be used with dpkg) is also good [23:22] i've noticed many of my friends are moving from windows to linux world, choosing ubuntu mostly [23:22] would you be willing to set this up or maintain it? [23:22] so, I guess it is very popular, hence the idea of deb package for ffmpeg [23:23] well, of course, I just need some help because I never did something like that before [23:25] i think it is a good idea, but i too have no experience with this (i've only "maintained" PKGBUILDS) [23:25] hi all - new here - trying to use av_read_frame as input to apple's hw decoder. (yes, i know about vda.c, but for what i'm doing, i need to run vda externally). av_read_frame produces raw nals when reading .mp4, and start codes when reading HLS. Since the apple stuff wants a nal format, is there a filter that would convert start codes to nals? [23:28] burek: users might expect it to work as a libav replacement for any dependencies. not sure how you would address that. [23:33] well we can just copy/paste the logic that was present before, while ffmpeg was in debian repo [23:33] to be honest, im not sure how did they resolve dependancy problems with avconv [23:33] i mean libavtools [23:33] did they keep the same names and paths and stuff or what [23:34] or all the projects had to manually change the deps to libav [23:34] i assume the first. plus keeping faux-ffmpeg for winff and other cli users. [23:35] i'm not sure how they will do it once they adopt upsteam that no longer contains ffake-ffmpeg. [00:00] --- Fri Oct 5 2012 From burek021 at gmail.com Fri Oct 5 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Fri, 5 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121004 Message-ID: <20121005000501.1E35E18A02C7@apolo.teamnet.rs> [01:13] anyone works with red5 (i know that is not the right channel to ask it.. sorry for that) [08:04] michaelni: ping [09:21] Hi. FFMPEG 1.0 was released a few days ago [09:21] there is a change "SubRip encoder and decoder without embedded timing" [09:21] what does this mean? [09:23] praveenmarkandu: it's an api change, this changelog entry was moved [09:24] (or maybe not this one in particular) [09:24] praveenmarkandu: do you want the technical details? [09:24] @ubitux, not sure if I would understand the technical details [09:25] but sure, why not [09:25] ok [09:25] is that documented anywhere? [09:25] git log [09:25] basically, when you have a .SRT file, the demuxer split that file into chunk including the timing information [09:25] (the hh:mm:ss.xxx --> hh:... thing) [09:26] yup [09:26] along with the payload (the markup text, like blablabla ...) [09:26] so the decoder receives these two information [09:26] in a "text" payload [09:26] the problem is, formats like matroska allow some SubRip markup [09:26] (the blablabla... ) [09:27] so the matroska demuxer needs to send this to the subrip decoder [09:27] but it can't send chunks like the SRT demuxer [09:27] ok. i actually understand so far [09:27] since the timing information is stored differently [09:28] previously it was constructing and adding the "hh:mm:ss.xxx --> hh:..." string into the packet before sending it to the decoder [09:28] which was quite ugly [09:28] now the matroska demuxer can just say it's subrip markup [09:29] (the subrip decoder has now two branches: one for "SRT" packets, and one for subrip - without timing in payload so -) [09:29] is that clear enough? [09:30] yes [09:30] thanks. lol. from the description, i thought it would solve my mpegts subtitles streaming problem [09:30] hahahaa. hopeful & naive [09:30] :) [09:33] i'm really curious to know if you can put some actual *text* subtitles in mpeg-ts [10:05] @ubitux, yes i am very curious myself :D [10:05] i'm afraid you might be able to put one of the N TTML flavors [10:05] where do you guys want commits notifications come to? [10:05] this channel or ffmpeg-devel? [10:05] devel channel obviously [10:35] Hi, with vlc I can use video encoder of mp2v, but with ffmpeg I can't even though vlc says to use ffmpeg - is the name different with ffmpeg? [10:36] parse error [13:14] hi [13:14] http://ffmpeg.org/pipermail/ffmpeg-devel/2012-September/131785.html [13:14] Apparently the TAK demuxer only works on TAK files that don't have APE tags [13:15] Also, I can't get the ffmpeg plugin in deadbeef to load TAK files [13:16] I recommend you either comment on the mailing list, or on #ffmpeg-devel [13:16] the guy's nickname begins with durandal [16:57] hi there, I'm trying to convert a H264 video to DVCPRO with ffmpeg 7:0.11.2-dmo1. I get a strange color problem : see http://pastebin.com/Zvjx3whP, there is a screenshot here : http://dev.diroots.info/H2642DVC.jpg any idea? would that happen because of the pix_fmt conversion from yuv420p to yuv422p? [17:31] hi there [17:32] I have a ffmpeg conversion problem and am looking for advice [17:32] I have a bunch of avi files with 1 xvid video stream and 2 mp3 audio streams, and I need to remove the second audio stream from them [17:33] I am using: ffmpeg -i file.avi -codec:v copy -codec:a copy new.avi [17:33] khali: you need to use -map parameter to set which streams get copied [17:33] but the resulting file can't be played by ffplay or vlc... either the video is very slow, of the sound is very choppy [17:34] and please put your ffmpeg output to a pastebin so we can actually help you. [17:34] Mavrik_: the streams I get in the output files are the ones I want, so I doubt it'll help, but I will try [17:39] Mavrik_: http://pastebin.com/fkb0r6Wp [17:40] can someone help me with this : I'm trying to convert a H264 video to DVCPRO with ffmpeg 7:0.11.2-dmo1. I get a strange color problem : see http://pastebin.com/Zvjx3whP, there is a screenshot here : http://dev.diroots.info/H2642DVC.jpg any idea? would that happen because of the pix_fmt conversion from yuv420p to yuv422p? [17:40] ffmpeg plays sound OK but video is too slow, vlc plays video OK but sound is completely choppy [17:41] hmm [17:41] I found that I could get a playable output file by using mp4 instead of avi as the container, or by reencoding the audio stream, but I would like to avoid doing that if possible [17:42] khali: it's probably an index problem [17:42] khali: try using mencoder to reindex the file after conversion [17:42] Mavrik_: I tried adding -map 0:0 -map 0:1 but the output file is exactly the same, to the bit - so that's not the solution [17:42] mencoder output.avi -o output_fixed.avi -forceidx -ovc copy -oac copy [17:44] Mavrik_: thanks for the suggestions, I tried that,but it doesn't solve the problem... actually it is worse, as video looks corrupted now [17:45] corrupted video by copying the stream? [17:45] Mavrik: yes I know it sounds odd [17:49] Mavrik: there was a lot of "Skipping frame!" and the resulting file is ~8% smaller than the original [17:50] Mavrik: but if I use mencoder directly on my source file, the output looks reasonable [17:51] Mavrik: one thing looks wrong in my input file: audiocodec: framecopy (format=55 chans=2 rate=48000 bits=0 B/s=16000 sample-0) [17:51] Mavrik: it should be bits=16 [17:51] I tried forcing it with -bits_per_coded_sample:a 16 but it did not seem to help [17:52] Mavrik: another thing I noticed when using ffmpeg : if I do .avi -> .mp4 and then .mp4 to .avi, I get a playable avi file [17:53] this looks like an avi muxer issue [17:57] Mavrik: I was wondering if there was any option I could add to the command line to get the output file right directly [17:58] Mavrik: I tried -vbsf dump_extra, -vbsf remove_extra and -bits_per_coded_sample:a 16, neither helped [18:02] now tried -bits_per_raw_sample:a 16 too, but no luck [18:02] Mavrik: are there options that affect the avi muxing process? [18:30] khali: ffmpeg -i input -map 0:v -map 0:a:0 -c copy output.avi [18:31] would remux video and the first audio stream [18:33] relaxed: thanks for the suggestion, but the result is the same :( [18:34] try with an updated version of ffmpeg [18:34] except that now ffplay says: [avi @ 0x7f443c0008c0] non-interleaved AVI [18:34] relaxed: I'm using ffmpeg version 0.11.1, is is considered old? [18:35] When a problem occurs anything but git is old. You can try my static build http://goo.gl/DPrRY [18:35] I'll be happy to rebuild my pwn ffmpeg from source as I was doing in the good old times [18:35] That will work too [18:57] hello, anybody online? [19:02] there are billions online [19:02] BILLIONS [19:03] :) [19:04] and at least one in billion who accept error reports for ffmpeg-1.0 (git pull)? [19:06] http://pastie.org/4909806 [19:09] for some strange reason, configure fails to find /usr/local/ version I've compiled; the one used is ubuntu-12.04 flite-dev package [19:10] but I don't think that breaks the test, anyway (correct me please, if I'm getting it wrong) [19:11] you're running fate? [19:12] if you're asking for "make fate", yes (make check) [19:14] I don't have any experience with fate but it looks all green here. http://fate.ffmpeg.org/ [19:15] What are you trying to do? [19:16] just to compile ffmpeg with as many possible conversions as possible (making it as useful as possible) [19:19] readelf gave me "6: 0000000000000000 0 NOTYPE GLOBAL DEFAULT UND cmu_lex_init" on /usr/lib/libflite_cmu_us_slt.so.1, that's why I wanted to use my own compiled library [19:45] update: I succeeded (executing all fate tests without errors) when I _omitted_ --enable-libflite; it seems that package flite1-dev on Ubuntu is broken; I'll now try to see in configure what blocks the recognition of libflite manually compiled (if anybody knows, please interrupt me), this could be helpful in resolution of this error on Ubuntu 12.04 [20:42] relaxed: wow, it worked! [20:43] relaxed: using ffmpeg built manually from the git repository, your command line works [20:47] relaxed: the output has only 11 bytes differing from the non-working variant, (at least) 5 of which are from the different version numbers of Lavf [20:47] probably 6 [20:47] do that's just 5 different bytes fixing it :) [21:03] relaxed: now I feel a bit stupid as obviously I should have tried this first :/ [21:04] Mavrik_: thanks to you too, BTW :) [21:05] relaxed: thanks a lot for your help, I should be able to convert all my files and save some disk space now :) [21:10] I made it, do you accept patches? :) [21:14] hello, maintainers? [21:17] vlaad_, patches are welcome :) [21:17] where to put them? (one patch, one pkgconfig file) [21:17] anyway, what did you make? [21:18] I've made pkgconfig for libflite, then I had to change configure to use require_pkg_config instead of require2 [21:18] vlaad_, file a bug on trac, but patches get more attention on the ffmpeg-devel mailing list [21:18] in order to bypass ubuntu's dynamic build [21:19] uhm, can you pastebin the patch? [21:19] yes [21:19] i'm the author of flite so maybe i can help [21:20] http://pastie.org/4910416 [21:21] vlaad_, what's exactly the issue? [21:22] I was not able to compile with ubuntu's dev package, so I wanted to use the /usr/local one (mine), for some strange reason (later, I deduced that static build is the only solution, test returned errors for undefined symbols) [21:22] I'll be back in 15 minutes [21:35] I'm back, did I miss something important? I'm going to bugzilla to report [21:39] vlaad_, https://ffmpeg.org/trac/ffmpeg [21:40] also, can you explain the issue (on trac is even better)? [21:40] sure, basically this affects ubuntu (perhaps debian too) people only [21:42] you create a new account, then you click on new ticket and file the bug [21:42] if you have a patch you can add the patch file for it [21:42] I did, just reading the "tracker manual" [21:42] I want to be nice :) [21:52] vlaad_, don't be too nice, save your time ;-), I won't be picky with you bug report [21:52] what I need is just an explanation of the issue [21:52] i'm on ubuntu 11.10, and it works fine here, so I need to understand why is failing on other systems [21:53] ar you on multiarch or i686 system? [21:53] s/ar/are/ [21:55] basically, the only change to flite-1.4 is that I've made libflite.pc and put it in /usr/local/lib/pkgconfig (thus, modified ffmpeg's configure can find it, instead of always using /usr/lib/ shared version provided by ubuntu) [21:57] vlaad_, that is you *created* a .pc file? [21:57] yes [21:57] that's beyond of what is fixable in ffmpeg [21:57] should be rather be addressed by debuntu/flite upstream [21:58] uhh... OK, but issue is also that without patching, ffmpeg's configure will not search (or at least is not finding on my machine) the /usr/local version [21:58] I can reproduce it [21:59] vlaad_, did you try with --extra-ldflags=/usr/local/lib ? [21:59] also you may want to fix globally you environment, and add /usr/local to LD_FLAGS [21:59] no, I was relying on environment LDFLAGS [21:59] then it shouldn't fail [22:00] what's your LDFLAGS? [22:00] hmm, I'll try with --extra-ld-flags [22:03] configure correctly reads my LDFLAGS='-rdynamic -lpthread -lgcc -Wl,-Bsymbolic-functions -Wl,-z,relro -Wl,-z,now -lgomp -L/usr/local/lib -L/usr/lib -L/usr/lib/gcc/x86_64-linux-gnu/4.6 -L/usr/lib/gcc/x86_64-linux-gnu/4.6/32' [22:03] so local lib is there, and I'm getting "ERROR: libflite not found" [22:04] I'll put now my config.log [22:04] on pastie [22:08] it is here, I've copied just the part when configure tries to compile one libflite dependent test file: http://pastie.org/4910604 [22:09] therefore, I added the following to libflite.pc in order to resolve this: "Libs: -L${libdir} -lasound -lflite_cmu_time_awb -lflite_cmu_us_awb -lflite_cmu_us_kal -lflite_cmu_us_kal16 -lflite_cmu_us_rms -lflite_cmu_us_slt -lflite_usenglish -lflite_cmulex -lflite" [22:14] vlaad_, silly flite... [22:14] so it is linking against libasound (which is never used in the current app) [22:14] and failing [22:15] the best solution would be to use a pkgconfig file, but since it is *not* provided by flite, the best we can do is to add a dependency to libasound [22:16] OK, should I report this as a bug? or not? [22:16] vlaad_, wait, give me some minutes and i'll give you a patch to test [22:17] OK, cool :) [22:27] vlaad_, http://pastebin.com/jqssR1w2 [22:31] saste, this was not working; however, when I added unconditionally -lasound to LDFLAGS, it passed the test; so definitely, it requires libasound as prerequisite [22:32] vlaad_, why is not working? [22:32] now when you do --enable-libflite, lasound is automatically added to ldflags, so it should work if my understanding is correct [22:32] do you want me to put the config.log on pastie? [22:33] wherever you want [22:40] here it is: http://pastebin.com/Y4dkppRw [22:40] I made it with shorter options, in order to make it more readable [22:48] vlaad_, are you sure you applied the patch? [22:49] yes, that crossed my mind too :) I pulled the git again and compared, it differs by 7 lines you posted [22:49] vlaad_, also, can you try with configure --disable-indevs --enable-libflite [22:49] in your case even configure --enable-libflite should do [22:50] vlaad_, you can patch configure by hand [22:50] just replace: enabled_any alsa_indev alsa_outdev && ... [22:51] with enabled_any alsa_indev alsa_outdev libflite && ... [22:51] this "configure --disable-indevs --enable-libflite" didn't work, now I'll do that change manually [22:52] vlaad_, and make sure you didn't mess up configure in other parts [22:52] do git stash to stash away all the local changes [22:52] then patch by hand [22:53] did that, but again it fails :( [22:54] lasound is not added to the LDFLAGS [22:58] I've found what is going on :) [22:58] I'll post you different patch for configure now [22:59] I've just moved that line above the actual check :) http://pastie.org/4910828 [23:00] indeed, to speak the truth, this makes it uglier and messier [23:01] vlaad_, yes [23:01] my patch was suboptimal [23:01] i see why it was failing [23:02] i'll try to think of a better fix, or i'll just push the quick'n dirty solution if it has no side effect [23:02] OK, but your patch was correct [23:02] just tell me your name so i can mention it in the commit [23:02] Vladimir Kraljevic [23:02] or i'll just call you "vlaad_", as you prefer [23:02] thanks :) [23:03] you're welcome [00:00] --- Fri Oct 5 2012 From burek021 at gmail.com Sat Oct 6 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Sat, 6 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121005 Message-ID: <20121006000502.A3D1F18A018E@apolo.teamnet.rs> [00:44] Daemon404: care to write a news entry that ffmpeg can be compiled now with MSVC? or if you give me the text I'll update the site. [00:47] (or anyone else who isn't MSVC ignorant like me) [01:00] 20:33 <@ubitux> http://www.flickr.com/photos/flameeyes/7906103846/lightbox/ VDD12 [01:00] 20:34 <@ubitux> Daemon404: where are you btw? :) [01:00] 20:35 <@ubitux> since i "missed" you then :p [01:00] 20:39 < KGB> [FFmpeg] michaelni pushed 25 new commits to master: http://git.io/3-62Aw [01:00] 20:39 < KGB> [FFmpeg/master] blowfish: Factorize testing into a separate function - Martin Storsj? [01:00] 20:39 < KGB> [FFmpeg/master] blowfish: Fix CBC decryption with dst==src - Martin Storsj? [01:00] 20:39 < KGB> [FFmpeg/master] blowfish: Add more tests - Martin Storsj? [01:00] 21:00 <@j-b> thresh: cone? [01:00] 21:00 <+thresh> j-b: it will join once there is a commit to show [01:00] 21:01 <+thresh> you provide a channel to spam to in a hook, not in a daemon instance [01:00] 21:01 <@j-b> oh, nice [01:00] 21:12 < fflogger> [newticket] eric@&: Ticket #1787 (Losing Silent Audio When Converting From Nellymoser Produced By Flash ...) created https://ffmpeg.org/trac/ffmpeg/ticket/1787 [01:00] RT|AO ? [01:01] 21:14 < fflogger> [attachment] eric@&: 1_1349110964561.flv attached to Ticket #1787 https://ffmpeg.org/trac/ffmpeg/attachment/ticket/1787/1_1349110964561.flv [01:01] 21:32 <@michaelni> j-b, thresh http://pastebin.com/gimSnw4T [01:01] 21:32 <+thresh> oups.. [01:01] 21:32 <@j-b> michaelni: well, well well [01:01] 21:33 <+thresh> michaelni: re-push, please [01:01] 21:33 < cone-205> ffmpeg.git Michael Niedermayer libavcodec/mpeg12enc.c libavcodec/mpegvideo.h: mpeg2enc: support and use frame_rate_ext when needed [01:01] 21:35 <@michaelni> done, thx [01:01] 21:36 <@michaelni> ive also disabled the github KGB [01:01] 21:59 < fflogger> [editedticket] saste: Ticket #1783 (libavutil contains file with the same name as a system header) updated https://ffmpeg.org/trac/ffmpeg/ticket/1783#comment:10 [01:01] 22:00 -!- mode/#ffmpeg-devel [+o durandal_1707] by ChanServ [01:01] 22:01 <@durandal_1707> michaelni: frame.data[X] for planar format seems to not come one after another [01:01] 22:03 <@durandal_1707> so instead of using single malloc function i need to use each one per channel [01:01] 22:03 <@durandal_1707> issue doesn't seem to happen with 16bit but does for 24(32) [01:01] 22:04 < nevcairiel> why would you malloc the channels manually, there is get_buffer to do that [01:01] 22:05 <@durandal_1707> i need that for < 24 bits per sample [01:01] 22:12 <+kierank> huh [01:01] 22:12 <+kierank> why does that make a difference [01:01] 22:13 <@durandal_1707> in > 16 bit case i do not need extra buffer, i just use frame [01:01] llogan could you +mute RT|AO for a while [01:01] 22:47 <@saste> michaelni: where can I find a file with repeat_pict != 0? [01:01] 22:48 <@michaelni> saste, some telecine stuff should use it [01:01] 22:49 <@Daemon404> ubitux, you can only see my head [01:01] until it empties its buffer [01:02] 22:49 <@Daemon404> middle-ish... [01:02] 22:49 <@saste> michaelni, so the question turns to: where can I find some telecine stuff? ;-) [01:02] 22:51 <@michaelni> samples archive probably, i dont know a specific file ATM either :( [01:02] 22:51 <@saste> uhm... well I'll have to do a brute force scan [01:02] 22:51 <@saste> something in fate samples? [01:02] 22:52 < divVerent> sawe can narrow it down a little [01:02] 22:52 < divVerent> it's mostly used on DVDs [01:02] 22:52 < divVerent> and very unlikely in TS streams [01:02] & [01:02] 22:52 < divVerent> as these tend to be real interlaced norally ;) [01:02] burek: you can mute the channel, not a single person [01:02] it woul require to voice everyone else [01:03] usually +b does mute also [01:03] ah right :) [01:03] but yes, +m is for channel :) [01:04] oops, wrong mouse button is pressed. ;( [01:05] *it happens :)) [01:08] "Irssi: Pasting 5 lines to #ffmpeg-devel. Press Ctrl-K if you wish to do this or Ctrl-C to cancel." # Irssi pasted and showing this. WTF [01:10] that's to stop you from accidentally flooding channels with large pastes. [01:11] quite useful. [01:11] :) [01:13] ffmpeg.git 3Michael Niedermayer 7libavcodec/aacenc.c: aacenc: fix out of array writes [01:13] ffmpeg.git 3Dmitry Samonenko 7libavcodec/libspeexenc.c: libspeexenc: Add an option for enabling DTX [01:13] ffmpeg.git 3Dmitry Samonenko 7libavcodec/libspeexenc.c: libspeexenc: Updated commentary to reflect recent changes [01:28] llogan, im currently working on shared lib support [01:28] Action: Daemon404 wrote a proper script to generate a def file [01:28] rather than hack it into the converter [01:38] yeah, scripts make more sense for stuff like that [01:38] my intention is to treat it like gas-preproc [01:43] burek: heh, sorry. i was afk. [03:17] ffmpeg.git 3Michael Niedermayer 7libavcodec/mpeg12data.c libavcodec/mpeg12data.h libavcodec/mpeg12enc.c: mpeg2videodec: fix list of supported frame rates to include sane ext rates. [03:28] "< Skyler_> that's to stop you from accidentally flooding channels with large pastes." # but it actually pasted to channel, then showing it afterwards, which make it useless. :( [04:03] Kim Dotcom?s Gaming Lag Hints at New Spying Controversy [04:03] hah! [04:03] thats awesome [04:04] lag, the government is stealing my packets [04:16] Compn, every comcast connection ever must be bugged then [05:57] ffmpeg.git 3Duncan Salerno 7libavformat/http.c: http: add option to prevent probing for HTTP seekability [05:57] ffmpeg.git 3Duncan Salerno 7libavformat/http.c: http: prevent the Range header being sent when seekability probing isnt used [05:57] ffmpeg.git 3Duncan Salerno 7libavformat/hls.c: hls: Disable http seekability probing [07:37] ffmpeg.git 3Ronald S. Bultje 7libavcodec/h264_cabac.c libavcodec/h264_cavlc.c: lavc/h264: don't touch H264Context->ref_count[] during MB decoding. [09:36] michaelni: i want to write scalarproduct_in16 for SSSE3 [10:08] Kim Dotcom?s Gaming Lag Hints at New Spying Controversy <- Compn, that's actually a common thing :) [10:09] I've found a numerous cases where people are being "proxied" through the systems intended to log/analyze all the traffic of the given isp [10:09] the easiest case to test for the most widely used "tool" is to try to connect to a bogus ip, using telnet, and if the connection succeeds, well, you know what it is :) [10:14] burek: hehe [10:14] in my company WLAN here, all FTP is intercepted [10:14] I can ftp to 1.1.1.1 [10:14] SUPPOSEDLY they do this as a crappy implementation of active-FTP-through-NAT [10:15] well, ftp is rarely monitored, mostly http and irc is of interest [10:15] that might be just some kind of (reverse) proxy [10:15] or something [10:15] right... but why would they monotr in that way [10:15] and not on IP packet level [10:15] that wouldn't be so easily detectable [10:16] well, mikrotik routers have that functionality to proxy at tcp level [10:16] and also, I once did monitor FTP ;) [10:16] when at the Chaos Communication Congress a few years ago [10:16] they had many FTPs with interesting... stuff on them [10:16] but it's so tedious to ask everyone... [10:16] so I sniffed for USER, PASS and CWD requests [10:16] problem: apparently, even CCC members are stupid enough to do POP3 from an open WLAN [10:17] which also uses USER and PASS... [10:17] mikrotik routers, for example, they fake syn/syn-ack/ack with you and buffer some amount of bytes to figure out what are you trying to do and then redirect all the traffic for that connection to a configured route or so [10:17] so that you can configure one route just for p2p, which might be useful :) [10:18] :) [10:19] also, just to mention, since I work for one isp, the internal affairs bureau, or how is it called world-wide - police? :) they officially asked all the isps in the country to provide a way for them to be able to remotely sniff all the traffic without being noticed [10:19] yes... but I hope no ISP dares to do such things [10:19] i mean, wtf.. [10:19] faking a SYN-ACK is highly problematic for many software [10:19] yes, let's say I am working externally for an ISP too [10:20] and our company actually did a part of the "lawful interception" (or was it lawless?) solution [10:20] but that one sure doesn't fake SYNs [10:20] it is connected via a "monitor port" (receive only) to the customer [10:20] so it can't even "accidentally" forge packets/replies [10:21] that's the cisco way I guess [10:21] yes, we actually don't do this on cisco [10:21] cisco's name is monitor port though [10:21] but this interfaces to some weird ADSL hardware [10:21] main reason is for doing it this way is BTW [10:22] that in Germany, the ISP only does part of the interception [10:22] the interception device is a blackbox that the government places in the datacenters, roughly spoken [10:22] and we have to provide for ways to route the correct customer's traffic to the boxes [10:22] so obviously we don't trust these boxes, and that's why they are connected receive-only [10:22] yeah [10:22] you are lucky they provide you with such hardware [10:22] they'r even on separate electrical circuits with lesser SLAs ;) [10:23] so in case they short out, nothing else breaks [10:23] here, they accepted a law that legally binds you to provide them with the hardware too [10:23] and also you are obligated to log all traffic for 6 months.. [10:23] and we say china is doing something bad.. :) [10:23] they just block what they dont like and dont bother investigating :))) [10:24] 6 months... not sure, may be 3 here [10:24] but same thing [10:24] but well [10:24] why is Germany using it? because USA pressured them to do it [10:24] long live ssl :) [10:24] it's all kissing ass of the RIAA/MPAA [10:25] hi, i'd like to understand the difference between strict gop vs closed gop. For both mode, the gop size is fixed and if i understand well in strict gop mode, a picture from another gop could be referenced in the current gop ? [10:25] and MAYBE a bit of anti terrorism [10:25] well, people that govern the EU and federal reserves are the same people, so it's no wonder [10:25] mateo` http://forum.doom9.org/archive/index.php/t-105129.html [10:26] brb [10:28] burek: thx :), so strict gop = gop size fixed + open gop [11:38] michaelni: I wonder if you can tune cone-205's output format. The git rev is missing, and missing separator between fields which is hard to view in colorless situations. [11:42] can one of our mpegtsenc expert give a look at this trac issue ? http://ffmpeg.org/trac/ffmpeg/ticket/1673 [11:51] j-b: what i can find DTS-HD-MA encoder? [11:51] what? [11:57] or wav file of any DTS-MA sample so i can get bitexact decoder? [11:58] search the usual shady places for DTS-HD Master Audio Suite 2.0 :P [12:06] burek : interpol ? [12:13] burek / divVerent : do your govt catch any 'bad guys' or is it all political / riaa-mpaa 'bad guys' ? [12:13] that they sniff on [12:16] good question... they didn't admit to catch anyone, IIRC [12:16] because doing that would make people learn to encrypt their data, or the like [13:00] Compn, well we did have one raid on child porn guys [13:00] that were imposing on social networks as teens [13:00] they were all like 40-50 yrs old and stuff [13:00] really sick.. [13:01] but, divVerent has made a good point, which reminds me why I didn't read anything about it in news [13:02] usually its in the news to prove the system is working [13:02] at least for childprn [13:03] what is not in the news is that they caught them via their facebook updates and not extensive data tapping.... [13:07] 13:02:31 +av500 | usually its in the news to prove the system is working [13:07] right, which is why I suppose that the tapping hasn't found anyone yet [13:07] but, looking through facebook did [13:08] probably quite simple to explain: those dumb enough to not even try to encrypt such data are dumb enough to talk about it everywhere [13:10] what's really ironic is that we've been running a lot of dc++ hubs (file sharing) on our servers and I regularly scanned through the shares of all the people connected, reporting people who shared child porn to the interpol just to find out [13:11] that the interpol contacted our upstream isp, asking about my private data, like name, address and stuff.. [13:11] instead of replying to at least one of my emails, directed to them, they decided that I was the cause of all those issues.. geez.. [13:11] who is "our"? [13:12] so I really doubt interpol has any interest in chasing those kind of people, unless someone influential tells them to do so [13:12] our server's upstream isp [13:12] no, you said "our servers" [13:12] what "entity" is that? [13:13] finding out if your credible is just due diligence [13:13] and the most likely narc is kin ... [13:13] we have a lot of rented servers throughout the world and on one of those servers I received the notification of our hosters to provide them with my full details and stuff [13:14] so I asked them just to give me their phone number so I can call them and let them know whatever they need to know, but they refused that [13:14] they were just interested in my identity, don't know why and honestly don't care [13:14] but I stopped sending any reports from that moment on [13:14] I realized they don't really care that much [13:32] burek : interpol is for stuff like catching dangerous criminal anakata or julian assange [13:32] not your petty crime nonsense! [13:32] right [13:32] Action: Compn curious why burek is hosting dc++ hubs tho [13:32] why are you honeypotting ? [13:32] like William Wallace would like to say: freedom :) [13:39] the thing is [13:39] the dc++ clients you reported [13:39] are probably ran by interpol [13:39] catching the downloaders :P [13:39] siretart: "Version 4:0.5.9-1 uploaded on 2012-09-29" // you're still maintaining an old 0.5 release? (https://launchpad.net/ffmpeg) [13:39] well, thats how it would work [13:39] or that's libav? [13:39] or something else? [13:41] ffmpeg in buntu is libav [13:41] Compn, probably you're right [13:42] nevcairiel: yes i know but i'm not sure in this case since the latest release is 0.6 [13:42] it looked like to me like an old ffmpeg or something [13:43] libav didnt exist in .5 /me ducks [13:43] hehe [13:53] saste: i'll review your -select_streams patch tonight :) [13:53] ubitux, thanks [13:53] it's very simple, the only thing i'm not really sure is the option name [13:53] otherwise i'd have already pushed it ;-) [13:54] hehe well i like select [13:54] i thought about "filter" as well [13:54] or "pick" [13:54] but select is file [13:54] fine* [13:54] specify sounds wrong: you're not asking ffprobe to specify anything [13:55] you *are* specifying things :P [13:55] is there a video filter/source that can create a test video, containing timestamps in the video, so I can use it for some testing? [13:55] -f lavfi -i testsrc [13:55] that simple? :) [13:55] yep [13:55] cool [13:55] thx :) [13:55] that's a pretty great source btw [13:59] ubitux: yes, "select" sounds better than "specify", i did the same reasoning [13:59] i'll wait you to review anyway, if you want to do some comment [13:59] or i'll push tomorrow [13:59] yes please wait a bit [14:00] 6 hours of waiting maximum ;) [14:03] is there any paper that describe dtshd format (.dtshd extension) ? [14:03] ubitux, ok, I don't want to bug you right now, I know you're busy, just to say that the examples that I put on wiki do work, but for some reasons ffmpeg always fails at first 3 frames, the rest of frames are exactly as expected [14:04] burek: fill a bug then; and btw, if you're filling a bug, you can use -f lavfi -i testsrc source, so it's easy to reproduce :) [14:04] ok :) [14:06] burek: hey btw, about the thumbnail wiki page, i'd quote various other things like : [14:06] - the tile filter to make a galery [14:06] - the scene detection [14:06] - some vf select + scale in the same filtergraph [14:06] durandal_1707, can this help http://www.dts.com/professionals/sound-technologies/codecs/dts-hd-high-resolution-audio.aspx [14:07] and even better, some images with testsrc so output is shown etc [14:08] I see, well tile filter has got its own page I think.. yes http://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20take%20multiple%20screenshots%20to%20an%20image%20(tile%2C%20mosaic) [14:10] multimedia.cx is back :) [14:10] i found some draft [14:10] are you sure it's right to split so much? :) [14:10] durandal_1707, wrong link before, did you try http://wiki.multimedia.cx/index.php?title=DTS-HD [14:11] ubitux, if you are talking to me, well I've based the wiki articles on people's search queries and questions on the forums and such [14:11] burek: that is for codec...... [14:12] adapting it for them to easily find what they are looking for, didn't look too much to organize things into bigger pages :) [14:12] but you can rearrange if you like :) [14:12] durandal_1707 you need it for the format only? [14:13] /mute burek [14:14] ok :) [14:14] burek: well at some point it might be good to not split too much and try to give some hints about more things [14:15] burek: ah also, there is a vf thumbnail filter you might want to quote [14:15] ubitux, I've used links to docs wherever I use some filters, specific options and such [14:15] anyway, one page talking about tile, scene detection, select and thumbnail filter is IMO a good one to have [14:15] ok [14:15] but you're the master there [14:15] master of disaster :) [14:16] just my opinion; but as you just saw on #ffmpeg, the user wasn't aware at all about all these other solutions [14:16] and he was kind of interested by them [14:16] having them pointed out in the wiki would have prevent this i believe [14:16] i agree [14:16] anyway, just my remark of the day :) [14:17] :beer: :) [14:17] (ah also, these info are already split into the official documentation, the wiki is here to link all of these to various user needs) [14:17] Action: ubitux mute himself for now [14:36] durandal_1707: the spec [14:37] kierank: yes? [14:38] i have linked you to it before no? [14:38] i found that ffmpeg chokes on it just recently, so i found it by accidend [14:51] arh i think i understand why ffserver is broken [14:52] looks like it's one of the consequence of yet another purification :( [14:54] what should be do with ffserver, anyway? [14:54] nobody is maintaining it, and "avserv" looks a bit like vaporware [14:54] well the regression i have seems related to the fact that avformatctx->timestamp was removed [14:54] (and it seems to break the date= seek system) [14:55] saste: about avserv i dunno, and avserver is broken out of the box in libav [14:55] given that nobody does active development on it, i'm surprised it even works [14:55] it doesn't [14:55] i actively tried not to do development over it, due to the socis task [14:55] ffserver kind of works though [14:56] yes, it might be worth trying to get something out of it now [14:56] i believe it's a pretty nice tool :) [14:56] and needs way more love [14:57] all we need is more love [14:58] money is a good surrogate (just put devs at fixing bugs for money) and we lack that as well ;-) [15:01] :) [15:06] ffmpeg.git 3Martin Storsj? 7libavformat/segment.c: segment: Add a missing space [15:06] ffmpeg.git 3Martin Storsj? 7libavformat/segment.c: segment: Properly create new AVStreams for the chained muxer [15:06] ffmpeg.git 3Michael Niedermayer 7: Merge commit '0edae4e6286096023cdd6adea74722fa06029867' [15:07] j-b, btw, cone could also display the short git hash of each commit [15:08] and RT|Chatzilla asked about a seperator between the fields [15:17] ffmpeg.git 3Martin Storsj? 7libavformat/segment.c: segment: Use the public av_write_header/av_write_trailer functions [15:17] ffmpeg.git 3Michael Niedermayer 7: Merge commit '73871dc96ff78053b9dcd0eb259b7f5a5308ec87' [15:31] its separated by color =P [15:51] ffmpeg.git 3Martin Storsj? 7libavformat/segment.c: segment: Free and reinit the muxer before calling avformat_write_header [15:51] ffmpeg.git 3Michael Niedermayer 7: Merge commit 'eb447d515956b3ce182d9750083131735f00324c' [16:02] michaelni: for the bot? [16:03] j-b : can you remove the 'merge commit' notifications ? [16:03] theres going to be lots of those [16:04] Compn: why?, they are regulat commits [16:04] *regular [16:05] yeah but its just michael : 'merge commit' : big useless hash number [16:14] why don't rebase on merge commits? [16:15] j-b, yes, short hashes for the bot would allow one to simply copy and paste the hash to the terminal and look at the diff [16:17] in a perfect world, it would spit out a url to the diff :D [16:18] shortened url, that is [16:19] ffmp.eg [16:21] michaelni: very well [16:23] yepee i fixed ffserver [16:24] \o/ [16:24] ffmpeg.git 3Martin Storsj? 7libavformat/segment.c: segment: Add an option for disabling writing of a header/trailer to each segment [16:24] ffmpeg.git 3Michael Niedermayer 7: Merge commit '378a6315b7c48195ffd94e6aa9aa6d663d42b35e' [16:34] wonder if the bot can ignore merge commits [16:37] ffmpeg.git 3Martin Storsj? 7libavformat/segment.c: segment: Set the resend_headers flag for each segment [16:37] ffmpeg.git 3Michael Niedermayer 7: Merge commit 'f7b240434c015056bc6319ddbdb8483757cc13e2' [16:41] nevcairiel: it looks worse in colorless situation (says logs) [16:42] ffmpeg.git 3Martin Storsj? 7libavformat/segment.c: segment: Flush buffered data before finishing a segment [16:42] ffmpeg.git 3Michael Niedermayer 7: Merge commit 'a854362b40f0e458db5a1fb0d2612a5702ee0ace' [16:43] I'd say the old CIA, or libav's ICIA format is better. [16:46] and now fixed properly \o/ [16:47] commit! [16:49] no, i'll first show the world my awesome patch [16:49] and now the world is aware. [16:50] beware ! [16:55] :) [16:57] ffmpeg.git 3Martin Storsj? 7libavformat/segment.c: segment: Add an option for omitting the first header and final trailer [16:57] ffmpeg.git 3Martin Storsj? 7libavformat/segment.c: segment: Add comments about calls that only are relevant for some muxers [16:57] ffmpeg.git 3Diego Biurrun 7configure libavcodec/Makefile: build: Factor out mpegaudio dependencies to CONFIG_MPEGAUDIO [16:57] ffmpeg.git 3Diego Biurrun 7libavutil/x86/cpu.c: x86: ff_get_cpu_flags_x86(): Avoid a pointless variable indirection [16:57] ffmpeg.git 3Diego Biurrun 7libavutil/x86/cpu.c: x86: cpu: Break out test for cpuid capabilities into separate function [16:57] ffmpeg.git 3Mans Rullgard 7configure: configure: add --enable-lto option [16:57] ffmpeg.git 3Michael Niedermayer 7: Merge commit '65d12900432ac880d764edbbd36818431484a76e' [17:08] ffmpeg.git 3Diego Biurrun 7libavutil/x86/Makefile libavutil/x86/cpu.c libavutil/x86/cpu.h libavutil/x86/cpuid.asm: x86: Add YASM implementations of cpuid and xgetbv from x264 [17:08] ffmpeg.git 3Diego Biurrun 7configure libavutil/x86/cpu.c: x86: Drop CPU detection intrinsics [17:08] ffmpeg.git 3Diego Biurrun 7libavutil/x86/cpu.c: x86: get_cpu_flags: add necessary ifdefs around function body [17:08] ffmpeg.git 3Ronald S. Bultje 7libavcodec/h264_cabac.c libavcodec/h264_cavlc.c: h264: don't touch H264Context->ref_count[] during MB decoding [17:08] ffmpeg.git 3Michael Niedermayer 7: Merge remote-tracking branch 'qatar/master' [17:13] michaelni, whats with all the merge commit lately [17:13] instead of a branch [17:14] maybe they are left over commits [17:14] which are hard to merge, or delayed for various reasons [17:14] Daemon404: it shows repo, not branch ;) [17:14] but theyre not [17:14] technically, i dont think git cares which it is [17:15] functionally [17:31] RT|Chatzilla: it really shows both, repo/branch :P [17:32] nevcairiel: but I didn't see "master" string in its lines ...? [17:33] ffmpeg.git Michael Niedermayer : Merge remote-tracking branch 'qatar/master' [17:33] i see it! [17:34] oops ;) [17:54] Daemon404: c99conv doesn't pull variables up to the beginning of the enclosing block, does it? [17:55] no it doesnt [17:55] because libav doesnt used mix code / var decls [17:55] nor does ffmpeg [17:56] it's a Future (TM) item [17:56] (and i personally think it is bad software dev practice) [17:56] well, rmdec.c doesn't compile here because of that (lines 372+) [17:56] is this a new change? [17:56] fate should have caught it if so [17:57] last msvc build was 94 min ago [17:59] let me check [18:03] gix, [18:03] if ((ret = rm_read_extradata(pb, st->codec, codec_data_size - (avio_tell(pb) - codec_pos))) < 0) [18:04] ^ line 372 [18:04] i dont see mixed code/vardecl [18:06] Daemon404: yeah, false alarm. was in another branch. [18:10] ah [18:11] ffmpeg.git 3Giorgio Vazzana 7libavformat/oggparsetheora.c: oggparsetheora: fix comment header parsing [18:21] Action: TimNich off for the weekend ............... [18:22] s/ \.+$/./ [18:31] saste: btw, forgot to ask in the mail, what happens if you try to filter non-existing things? (like filtering only video in an audio file) [18:33] ubitux: nothing [18:33] will the output always make sense? [18:34] you could have empty output [18:34] for example -select_streams v in only audio, will result in no output [18:34] which is expected [18:35] as long as it's not broken.. :) [18:35] such as stuff like {"streams": {, }} or stuff like that [18:51] lemme try [18:51] in that case it would be an unrelated bug [19:10] ubitux: looks correct [19:10] you have an empty array, as expected [19:11] ok ok great :) [19:11] sorry for the bother then ;) [19:12] http://paste.org/55070 [19:14] looks fine; it's funny how the output are mixed [19:22] ubitux: how? [19:23] opening '{', then the av_dump_info thing, then the rest :p [19:23] mixed stderr/stdout i guess [19:24] ah yes, nothing i can do about it [19:24] sure :) [19:24] that's because we open the container before to open the file [19:24] in case we need to print versioning info [19:24] yup ok :) [19:25] ffmpeg.git 3Paul B Mahol 7configure: configure: dts demuxer needs dca_parser [19:26] saste: btw, when will we have meta info communication between filters? [19:27] well i'm slacking off [19:27] i'd like to work on a beat and pitch audio detection filter, but i think it will make sense only with this [19:27] did i already point you to the thomas repo? [19:27] it's kind of frustrating to have to print stuff in detect-like filters :p [19:27] he already implemented that (and it is pretty simple) [19:27] mmh possibly, but i don't remember [19:47] hey nyuhu, how is eq filter going on? :) [20:13] ffmpeg.git 3Carl Eugen Hoyos 7configure: Fix showspectrum dependencies: Add rdft. [21:04] ffmpeg.git 3Carl Eugen Hoyos 7libavfilter/libmpcodecs/vf_pullup.c: Do not print debug output for the (MPlayer) pullup filter. [21:45] ffmpeg.git 3Carl Eugen Hoyos 7configure: Fix libcdio detection. [22:35] arg [22:35] ffplay -> plays in sync [22:35] ffmpeg -i input.mp4 out.avi [22:35] ffplay out.avi [22:35] 20 seconds off [22:35] no warnings, no errors [22:38] Daemon404: any info with increased verbosity? [22:40] no. [22:40] [mov,mp4,m4a,3gp,3g2,mj2 @ 0x1e88340] multiple edit list entries, a/v desync might occur, patch welcome [22:40] fuck yeah [23:12] yay vobsub demuxer starts to work. [00:00] --- Sat Oct 6 2012 From burek021 at gmail.com Sat Oct 6 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Sat, 6 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121005 Message-ID: <20121006000501.7A61318A018B@apolo.teamnet.rs> [06:25] Hi All, Has some issue when converting mov to flv. Below is the error [06:25] Impossible to convert between the formats supported by the filter 'src' and the filter 'auto-inserted scaler 0' [06:25] i am able to convert almost all other format like mp4 to flv [06:26] any ideas? [06:28] the command i used is [06:28] ffmpeg -i /home/rome/videos/abc.mov -ar 44100 -ab 96 -f flv /home/rome/converted/abc.flv [06:29] sure [06:33] http://pastebin.com/CfiqN3sJ [06:33] here is the complete command [06:34] hmm... i never came across that, maybe wait for the devs do answer [06:35] is the file realls just 1.92 seconds? [06:36] *really [06:37] its a small file [07:11] how can i make a video fit a smaller video size without distortions. [07:11] ? [07:13] guys the issue was actually related to the video. I tried another mov file and it worked just fine. [07:13] Thanks Klaxa [07:13] right now i em trying this... but the video is distorted [07:13] for i in *.flv; do avconv -i $i -b 250k -r 15 -s qvga vid/new/`echo $i| sed 's/.flv//g'`.mp4;done [07:14] manitpaulose: actually .mov is just the contaner and not the codec [07:14] the issue was i was not able to convert .mov to .mp4 or flv [07:14] the issue was actually wit the file [07:14] and not exactly the most important part. the .mov you were using just now seems to be a raw video [07:15] k [07:15] tried another mov file and it worked file [07:15] djapo: you mean keeping the correct aspec ratio? [07:17] klaxa: no, can i scale a widescreen video to fit inside of a qvga screen without cliping or stretching, i em ok with adding more negative space [07:17] you mean adding letterboxes so you have black bars on the top and the bottom? [07:18] yes that's what i mean [07:21] djapo: http://superuser.com/questions/26416/how-to-convert-a-169-movie-to-a-43-letterbox-version [07:22] klaxa: thanks :D [07:22] :) [07:44] klaxa: can it also be done with avconv [07:45] ? [07:45] wasn't ffmpeg just renamed to avconv? [07:45] klaxa: i think so, but i tryed with ffmpeg and it didn't work [07:46] klaxa: no, avconv does not recognize those options. [07:48] avconv is part of a fork of ffmpeg [07:48] and that fork is distributing an old broken binary of ffmpeg [07:48] sounds awful [07:48] i think one could use the drawbox video filter to achieve the same result? [07:48] and ubuntu/debian are distributing that fork [07:48] under the name of libav. [07:49] erhm [07:49] under the name of ffmpeg sorry [07:50] ah djapo from the manpage: [07:50] >All the pad options have been removed. Use -vf [07:50] >pad=width:height:x:y:color instead. [07:50] https://www.ffmpeg.org/ffmpeg.html#pad [07:50] there are some examples a bit below here [07:51] klaxa: ubitux: thanks [07:51] djapo: you should also upgrade to ffmpeg [07:51] because that's not what you are using right now :p [07:53] ubitux: i em using ffmpeg but it could not set the options and it tells me that ffmpeg is deprecated use avconv [09:31] Hello! [09:48] hey, I'm trying to resize some PAL content to square pixels. Unfortunately, ffmpeg adds small black borders to the sides of the output, instead of just scaling the input from 720x576 to 768x576. dump: http://pastebin.com/DhLvgzUx [09:49] I tried various settings with -sar and -aspect, but somehow ffmpeg seems to read the DAR from the source, which has non-square pixels, and so thinks it's necessary to calculate its own size instead of just friggin do what it's told ;) [10:24] nobody? :( I'm currently trying to muck around with the setdar and setsar filters, but whatever I do, ffmpeg always keeps doing it's own "thinking". I just want it to resize the darn video and let me specify the sar and dar values [10:27] from the codumentation: The scale filter forces the output display aspect ratio to be the same of the input, by changing the output sample aspect ratio. [10:27] how is one supposed to ever change the DAR then? when i specify a DAR, ffmpeg does a silent rescaling of the output [11:12] Hey, quick question if anyone is alive, I'm trying to copy a mp3 to a new mp3 and the source mp3 has some weird mjpeg stream as stream '0' I'd like to drop [11:12] What do I specify in the options to do so? [11:13] Specifically it's "Stream #0:1: Video: mjpeg, yuvj420p, 450x600 [SAR 96:96 DAR 3:4], 90k tbr, 90k tbn, 90k tbc" [11:17] Action: microchip_ didn't know mp3's can contain video tracks [11:18] microchip_: Me either, I find it confusing [11:18] But I think I'm closer to my solution except.. Invalid encoder type 'libmp3lame' [11:20] microchip_: How do I drop a stream? [11:20] i've no idea [11:21] Success! [11:21] -vn [11:22] yes [11:23] I'm having difficulties to create thumbnails from a video [11:23] i.e select only 1 frame every 10 minutes [11:23] the doc says "-r 1/600" but it doesn't work [11:24] What does it do? [11:26] what's your cmd line knoch? [11:27] ffmpeg -ss 00:12:00 -i /home/file.ts -r 1/600 -vframes 15 -s qvga tn/%02d.png [11:28] there are only a few seconds between the first three thumbnails [11:31] the fourth seems to be good, 10 minutes after the third [11:35] ok just a min [11:35] trying something [11:36] ok thank you ubitux :) [11:40] knoch: can you try: ffmpeg -i /home/file.ts -ss 00:12:00 -vf 'select=isnan(prev_selected_t)+gte(t-prev_selected_t\,60*10)' -frames:v 15 -vsync vfr -s qvga th/out%02d.png ? [11:41] btw, what's your final goal of this? [11:42] knoch: oh and where did you see "-r 1/600"? [11:42] (where in the doc) [11:43] wait [11:43] here : http://ffmpeg.org/trac/ffmpeg/wiki/Create%20a%20thumbnail%20image%20every%20X%20seconds%20of%20the%20video [11:43] arh [11:43] Action: ubitux hits burek [11:43] haha [11:44] i guess i'll have to complete that page [11:44] and fix it [11:44] my final goal is to create navigation in a video with thumbnails [11:44] you basically click on the thumbnail and it seeks [11:45] knoch: did you try the scene detection? :) [11:45] to the corresponding timestamp [11:45] not at all [11:45] I have to keep the timestamp [11:46] try replacing 'select=isnan(prev_selected_t)+gte(t-prev_selected_t\,60*10)' with 'gt(scene\,0.4)' [11:46] of and btw, you should replace the -s qvga with a scale filter at the end of the filtergraph [11:46] anyway, tell me if it works first [11:46] you have a point, the scene detection would be better to avoid having bad thumbnails [11:46] the first cmd line ? [11:47] a scale filter ? I am a newbie at ffmpeg as you may have noticed I'm sorry [11:48] ffmpeg -i /home/file.ts -ss 00:12:00 -vf 'select=isnan(prev_selected_t)+gte(t-prev_selected_t\,60*10),scale=320:240' -frames:v 15 -vsync vfr th/out%02d.png [11:48] ffmpeg -i /home/file.ts -ss 00:12:00 -vf 'select=gt(scene\,0.4),scale=320:240' -frames:v 15 -vsync vfr th/out%02d.png [11:48] yes ok, I'll try these two [11:49] you move the -ss option after the input [11:49] why ? [11:49] moved* [11:49] more efficient [11:50] more accurate* [11:50] (might be slower though) [11:51] ok :) [11:52] you might be interested in the tile filter as well [11:52] yes but I don't want to bother you too much [11:52] like: ffmpeg -i /home/file.ts -vf 'select=gt(scene\,0.4),scale=320:240,tile' preview.png [11:53] bother me? mmh? [11:53] you're on #ffmpeg, that's the purpose of the channel to help you achieving what you want with ffmpeg :P [11:54] yes but.. it's great what you people do [11:54] so thank you ! [11:54] np :) [11:54] and I have tested the first command lin [11:54] e [11:56] the first thumbnail is incorrect [11:57] it's not what the video shows at 00:12:00 but it is at 20:00 [11:59] mmh wait, try to move back the -ss before -i, just to be sure [11:59] ok [12:00] so I enter this : ffmpeg -ss 00:12:00 -i /mnt/backup/media/dance_nation.ts -vf 'select=isnan(prev_selected_t)+gte(t-prev_selected_t\,60*10),scale=320:240' -frames:v 15 -vsync vfr tn/%02d.png just to be sure [12:01] so you can follow exactly what I do [12:03] well it works like a charm [12:03] perfect [12:04] can you explain the filtergraph ? [12:04] please [12:39] ubitux: I tried the tile filter and.. it's amazing [12:40] same: an explanation would be great :D [13:05] ubitux? :) [13:05] knoch, what exactly didn't work with 1/600? :) [13:06] please don't get mad :( [13:06] [11:29:00] < knoch> there are only a few seconds between the first three thumbnails ackjewt [13:06] [11:31:46] < knoch> the fourth seems to be good, 10 minutes after the third [13:06] knoch can you please use a pastebin site (like www.pastebin.com) to show your ffmpeg command and the complete console output? [13:07] sure [13:11] burek: you're one of the ffmpeg dev ? [13:12] no [13:12] im just the noise :) [13:16] oh ok [13:16] I lost the previous console output [13:16] running it again [13:18] burek: http://pastebin.com/hCHv49rT [13:18] bash colors [13:23] hm, I'm not sure what the issue might be, knoch, but those "skipped MB" and "invalid mb" tell me something is wrong :) [13:23] did you try moving -ss after -i option [13:24] like ffmpeg -i /mnt/backup/media/dance_nation.ts -ss 00:12:00 -r 1/600 -vframes 15 -s qvga tn/%02d.png [13:24] it will run slower but more precise [13:24] although I'm not sure that will actually help.. [13:25] where did you get that dance_nation.ts from? [13:27] it's a live capture from a TV stream [13:28] moving -ss after -i causes my computer to freeze [13:29] it takes a lot of memory and CPU [13:29] yeah, I was afraid of that [13:30] it is now not skipping through your input fast, but rather decodes all frames and drops them until it reaches the specified -ss time [13:32] < knoch> can you explain the filtergraph ? // select is filtering some frames, then the output is scaled with the scale filter; for more information look at http://ffmpeg.org/ffmpeg.html#select [13:33] about the tile filter: http://ffmpeg.org/ffmpeg.html#tile [13:35] many thanks to both of you [13:37] so [13:40] the scene detection is better if we don't have a limited number of thumbnails I guess [13:40] hi. i'm trying to stream with ffserver, but i get: [tcp @ 0x1e63240] TCP connection to localhost:58180 failed: Connection refused [13:40] otoh, ffserver -d shows a POST, and i can access the port with nc [13:41] and the streaming quits after a second or so with 85.8kbits/sav_interleaved_write_frame(): Connection reset by peer [13:41] ok [13:41] http://sprunge.us/YZUX ffserver.conf [13:42] http://sprunge.us/cZTL commandline + output [13:42] ubitux: is it possible to select only 15 thumbnails all over the video while taking advantage of the scene detection ? [13:43] chris2, you don't need -acodec/-vcodec in the ffmpeg line, when using ffserver [13:43] use something like this: ffmpeg -re -f video4linux2 -i /dev/video0 -isync -f alsa -i hw:0,0 http://localhost:58180/feed1.ffm [13:43] ok, that keeps running [13:44] and define your vcodecs and acodecs in ffserver.conf [13:44] knoch: nope [13:44] knoch: you can somehow do it like this: use the scene detection to output all the "scene" pictures [13:44] ok [13:44] knoch: and then filter out yourself with a script or anything what pic you don't want [13:45] maybe something like: ffmpeg ... -vf ... -f nut - | ffmpeg -f nut -i - -r 1/10 .. output [13:46] ^ knoch [13:46] ? [13:46] what are you answering to burek? [13:46] to filter out every 10th frame [13:46] ah no [13:46] you can do that with -vf select that's not the problem [13:46] or he would like some fancy selecting ? [13:46] the problem is that you have N output pictures [13:47] and you always want X of them [13:47] oh [13:47] so you need to pic using X/N and so you need to know in advance the number of output pic [13:47] btw, why doesn't -r 1/600 work? [13:47] it used to work for me at least [13:47] burek: is that related to this error? Fri Oct 5 13:47:18 2012 Codec for stream 0 does not use global headers but container format requires global headers [13:47] burek: git bisect then, i don't know how that's supposed to work [13:48] chris2, take a look at ffserver sample configuration [13:48] burek: but vf select is appropriate for this, you should update the examples [13:48] chris2 and type ctrl+f and then "global_header" [13:48] thanks [13:48] thank you ubitux [13:49] np [13:49] ubitux, the idea was to select 1 frame each 600 seconds [13:49] hence -r 1/600 [13:49] dropping all the others [13:49] i'm not sure how that could have work but well [13:49] if you say so.. [13:49] at the moment it doesn't anyway [13:49] anyway, that's exactly the purpose of vf select [13:50] and my cmd line is based on the vf select example from the documentation [13:50] (http://ffmpeg.org/ffmpeg.html#select) [13:50] what do you mean "i'm not sure how that could have work" ? [13:51] because -r is to set the frame rate of the output, and it would have a special meaning for the image2 muxer [13:51] specifying an output rate would make ffmpeg drop/dup frames to achieve it, no? [13:51] I should really learn how vf works, I have coded a basic streaming application which sends MPEG2-TS over RTP [13:51] hrm, this breaks down after a few seconds of streaming [13:51] and now I want to implement fast motion feature [13:51] burek: the image muxer is always tricky, because they are still/standalone images [13:51] so no timing stuff [13:51] and -r is for setting the framerate of the container, not really filtering images [13:52] knoch: look at vf setpts [13:52] knoch: http://ffmpeg.org/ffmpeg.html#asetpts_002c-setpts [13:52] (http://ffmpeg.org/ffmpeg.html#Examples-14) [13:52] yeah I have seen setpts [13:53] but to code it in C is a hard work when you don't have a clue of how it works [13:53] http://git.videolan.org/?p=ffmpeg.git;a=tree;f=doc/examples;hb=HEAD [13:53] AVFilterGraph etc [13:53] look at the filtering_video.c example [13:53] and just change the scale filter into a setpts one :) [13:54] it *might* work :) [13:54] great thanks I love you :D [13:54] but I would like to understand how it basically works [13:56] it's a graph of connected filters, but how are they connected ? do they always expect decoded frames ? [13:57] knoch: http://ffmpeg.org/filters.html [13:57] knoch: some filters are "source" filter [13:57] that don't expect any input [13:57] example: try ffplay -f lavfi -i testsrc [13:58] and you can then add a "normal" filter after: ./ffplay -f lavfi -i 'testsrc,hue=H=2*PI*t:s=sin(2*PI*t)+1' [13:59] you can name some outputs, reuse them between filters taking more than one input, etc [14:00] waow [14:00] you are a developer ? :D [14:00] somehow [14:00] :) [14:00] anyway you master ffmpeg [14:02] i'm just grepping examples :) [14:02] but you understand them [14:05] what is the hue filter ? [14:05] http://ffmpeg.org/ffmpeg.html#hue [14:05] "Modify the hue and/or the saturation of the input."| [14:05] argh.. sorry [14:05] no worry :) [14:06] I should have search the documentation before asking [14:06] +ed [14:08] and another question, I would like to implement (fast) rewind but I believe it is covered anywhere, so we know that fast motion is based upon modifying PTS and DTS, is it possible to do it for rewind ? [14:08] I was thinking [14:09] buffering from a Iframe to another (so a GOP), change the PTS but not the DTS [14:09] and send the frames from the buffer to the decoder [14:11] ths PTS would be changed by keeping the gap, then reversed to display the last frame in first [14:14] Action: ubitux is lost [14:29] ubitux: tell me if you think this is completely absurd [14:47] knoch: i just don't understand, but i'm a bit busy anyway now :p [14:52] when was -padleft etc changed to pad? [14:53] yes, pad filter [14:55] but when was this changed_ [14:55] ? [14:55] I have MediaMosa installed, and that uses the old parameter [14:55] Action: RoyK wonders why the old format wouldn't be allowed - it wouldn't hurt... [14:56] and wouldn't break things... [14:56] it would [14:57] because the processing is taking out of ffmpeg and moved to libavfilter [14:57] it complexifies a lot the code to keep these -pad options [14:57] it was removed a long time ago [14:57] http://ffmpeg.org/ffmpeg.html#pad [14:57] well, any idea when it was changed? [14:57] ok ubitux, tell me when you can [14:59] RoyK: going to look at the log [15:00] RoyK: Date: Fri May 7 12:16:23 2010 +0000 [15:00] Remove messy pading hack in ffmpeg.c. [15:01] thanks [15:54] Huh, is there a way I can extract program id from mpegts? I saw [15:55] http://ffmpeg.org/trac/ffmpeg/ticket/995 buw I get very strange errors [15:55] Failed to compensate for timestamp delta of -33698.589720 [15:56] To be more interesting I have a live stream via network and when I try to use -map I see always different ids for the same stream [16:13] i'm using this to stream: ffmpeg -v 2 -r 25 -s 640x480 -f video4linux2 -vcodec mjpeg -i /dev/video0 -f alsa -acodec copy -vcodec copy -f mpegts 'udp://10.153.59.22:1234?pkt_size=188&buffer_size=65535' [16:14] how do i tell these flags ffplay? it fails to detect the audio [16:14] "Could not find codec parameters for stream 0 (Audio: aac_latm ([6][0][0][0] / 0x0006), 0 channels, s16): unspecified sample rate" [16:14] maybe you should encode the audio [16:15] i tried -acodec libvorbis too [16:15] but i always get above message [16:16] wait... [16:16] you have -f alsa, but no input [16:16] or is the input from the video device? [16:16] also i think the -vcodec copy is redundant [16:17] if you have a seperate audiodevice in alsa, try adding -i hw0,0 or something [16:17] i tried that too [16:17] [aac_latm @ 0x7f066802af60] multiple layers are not supported [16:17] no idea why it thinks aac? [16:18] without -acopdec copy it recodes to mpeg2... [16:19] can i even have mjpeg in a mpegts? what mux should i use over udp? [16:21] you can cram almost any codec into any container; the question really is, is the device i intend to playback this mpegts upon going to understand mjpeg [16:21] well, i want to use ffplay [16:22] then TIAS! [16:23] tias? [16:23] try it and see [16:23] ffplay thinks its aac_latm :( [16:24] Stream #0:0: Video: mjpeg, yuvj422p, 640x480, q=2-31, -4 kb/s, 90k tbn, 24 tbc [16:24] Stream #0:1: Audio: mp2, 32000 Hz, stereo, s16, 128 kb/s [16:24] that i send [16:24] Stream #0:0[0x100]: Audio: aac_latm ([6][0][0][0] / 0x0006), 0 channels, s16 [16:24] Stream #0:1[0x101]: Audio: mp2 ([3][0][0][0] / 0x0003), 32000 Hz, stereo, s16, 128 kb/s [16:24] that i receive [16:25] maybe there's a bug in ffplay [16:26] hrm, my mplayer2 uses ffmpeg too of course [16:29] an explicit -vcodec mjpeg results in Stream #0:0[0x100]: Unknown: none ([6][0][0][0] / 0x0006) [16:42] I'm trying to create a "multiply" effect between two videos. How is this accomplished with the use of filters? [17:26] aleksm: you want two videos side by side? [17:28] relaxed: no, more along the lines of the "multiply" blend mode (http://en.wikipedia.org/wiki/Blend_modes) [17:28] so just one video [17:33] Did you look at the fade filter and the overlay filter in the man page? [17:36] You should be able to overlay images to achieve alpha blending [17:51] relaxed: thanks for the response, but would this achieve the same effect as multiplying? I'm not looking to simply overlay one video over the other. [18:00] any suggestions for broken ffmpeg default settings detected use an encoding preset (e.g. -vpre medium) preset usage: -vpre -vpre speed presets are listed in x264 --help profile is optional; x264 defaults to high [18:10] any suggestions guys [18:11] Guest31191: Yes, use a recent version of ffmpeg. [18:11] It now uses libx264's internal presets. example: -preset veryslow [18:14] Guest31191: you can get a list of valid presets from `x264 --fullhelp | less` [18:14] are you saying i should use this value at command prompt [18:14] I'm saying after you upgrade to a recent version of ffmpeg you should use that value at the command prompt. [18:15] Which version are you using? [18:16] ffmpeg 2>&1| sed q [18:16] Ok Thanks [18:17] aleksm: I'm really not sure. [20:26] im trying to convert mkv to mp4 with ffmpeg -i input.mkv -vcodec libx264 -sameq -b 2089k -acodec libfaac -ab 192k video.mp4 but the audio comes out distorted heavyly [20:27] whats the best way to keep the audio intact [20:28] didn't i help you yesterday? [20:29] yup [20:29] 1) loose -sameq and never use it again [20:29] lose* [20:29] but the command u gave me it wouldnt play in my ps3 [20:29] so i used ffmpeg -i input.mkv -vcodec libx264 -sameq -b 2089k -acodec libfaac -ab 192k video.mp4 and it played but the audio was messed up [20:33] How was it messed up? [20:34] was sound was distorted [20:34] kinda low [20:34] but you could hear it [20:36] remove -sameq, add "-ac 2 -level 41 -t 60" for a minute long sample and see if that works. [20:37] ffmpeg -i input.mkv -vcodec libx264 -ac 2 -level 41 -t 60 -b 2089k -acodec libfaac -ab 192k video.mp4 ? [20:37] yes [20:37] if that plays okay remove "-t 60" to encode the whole video. [20:37] gonna give it a try [20:55] that wrked fine [20:55] gonna save this command [20:55] its a good one thanks relaxed [21:31] t4nk651: you're welcome [21:44] I'm trying to create a "multiply" blend between two videos. How is this accomplished with the use of filters? [21:48] aleksm: ffmpeg does not have its own multiply filter, but it does support frie0r which probably has one [21:52] llogan: thank you :) [21:57] aleksm: ...or maybe not. i can't find one [21:58] oh, it does appear to have one [21:58] llogan: yes, I can't either -- do you happen to know of any other way I can multiply blend two videos using CL tools? [21:58] oh? [21:59] Action: llogan builds with frei0r support [22:03] Action: aleksm should clearly not rely on binary packages of ffmpeg [22:23] i don't know if ffmpeg can support frei0r multiply and i've never used frei0r with more than one input [22:35] aleksm: might be easier to use something like kdenlive. it also supports frei0r [22:41] llogan: I really need to do this on the CL -- it's supposed to be a server process :) [22:54] aleksm: i can't figure it out but i'm also preoccupied at the moment. it would be nice to know if ffmpeg can even support the frei0r mixer2 "filters" [22:57] llogan: I'm going to be playing around with it over the weekend, but I'll pop in here every now and then and share what I find with you [22:57] Action: aleksm is clocking out [23:05] Hi, i am using arch linux ffmpeg 1.0 [23:05] I was wondering if pthreads is enabled for this? [23:06] shadylog: ldd `which ffmpeg`| grep thre [23:07] [root at ion ~]# ldd `which ffmpeg`| grep thre [23:07] libpthread.so.0 => /usr/lib/libpthread.so.0 (0x00007fbbfc466000) [23:07] So I guess it's in there? :) [23:26] Hi, I would like to filter incoming stream (live stream over network in mpegts) into more separate streams [23:26] I use -map:p:program_name [23:27] but it simply ignores command and if I use -map 0:0 - map 0:1 is runs a circular buffer overrun error [23:27] any thoughts? [23:27] I use latest ffmpeg (as of today) [00:00] --- Sat Oct 6 2012 From burek021 at gmail.com Sun Oct 7 02:05:03 2012 From: burek021 at gmail.com (burek) Date: Sun, 7 Oct 2012 02:05:03 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121006 Message-ID: <20121007000503.3531618A01DE@apolo.teamnet.rs> [00:01] ffmpeg.git 3Stefano Sabatini 7Changelog doc/ffprobe.texi ffprobe.c: ffprobe: add -select_streams option [00:01] ffmpeg.git 3Stefano Sabatini 7ffprobe.c: ffprobe: reindent after previous commit [00:06] ffmpeg.git 3Stefano Sabatini 7Changelog: Changelog: add empty line after version lines [00:20] durandal_1707, pelase refrain from giving such useless reviews [00:21] Daemon404: not useless [00:21] "I do not like this." [00:21] useless [00:21] as [00:21] fuck [00:21] _not_ useless [00:22] how [00:22] on earth is that useful [00:22] do explain [00:22] im all ears [00:22] maybe he should use telepathy [00:22] to figure out whats wrong [00:22] and why you dont like it [00:23] "I do not like this." <- always try to give at least some kind of a reasoning [00:23] otherwise it's as useless as a spit on a face [00:23] and leads to "OK, so he doesn't like it. How can I fix it?" [00:24] instead of "OK, X could be done better." [00:24] people are just ignorant [00:24] it soudns mroe liek youre jus ta ragain asshole [00:24] let me restate that without typos: [00:24] youre an asshole. [00:24] it's never boring here :) [00:24] durandal_1707, it might be clear to you why something's bad [00:25] but it might not be for the other person [00:25] tl;dr explain it [00:25] Daemon404: thank you for so kind words [00:25] im doing my paul b mahol impression [00:25] i hope you like it [00:26] (but if you dont, feel free to say so. but not why.) [00:28] Daemon404, moderate terms [00:28] you made your point, no need to give names [00:28] blame the attitude, not the person [00:28] saste, youre right [00:34] "<@Daemon404> it soudns mroe liek youre jus ta ragain asshole" <-- I like how you only spelled one word right, and it was the most important word. It kinds adds emphasis in its own, "he must be drunk or something" kinda way ;D. [00:34] "it" is also spelled correctly! [00:34] Action: Daemon404 looks at the beer on his desk [00:34] nevcairiel: Yeah, but that's too short to count :-D. [00:35] i spell 'it' wrong all the time [00:35] hey he screwed up shorter words in that sentence [00:35] as 'i' [00:35] Daemon404: I had a friend from Canada who only typed correctly when he /was/ drunk. [00:35] we call them newfies [00:35] XD [00:38] michaelni, I need something like 576 words for xface bigint [00:38] lavu/integer supports 8 words [00:39] should I extend the number, make it configurable or use a custom biginter implementation? [00:39] second option will break compatibility, but not that this would be a problem, since possibly nobody is (still) using it [00:40] ah and the author of the code (~1990) agreed to relicense the code [00:41] this reminds me [00:41] we can ditch x11grab [00:41] vlc has xcbgrab [00:41] whic his lgpl [00:41] and also uses xck [00:41] xdc* [00:41] Daemon404, patches welcome :) [00:41] saste, that requires me to touch stuff related to X [00:41] never. again. [00:42] i somewhat envy mplayer for x11 output [00:42] from time to time, people tries to reimplement X from scratch [00:42] i recall a nice project, it was called "Y" [00:43] wayland sounds like it may become something [00:43] its got people like keithp on it [00:46] It definitely has a lot of people and groups behind it. [00:46] It'll be interesting to see what happens. [00:46] I probably won't start using it until Fedora switches to it as the default (if that ever happens, which I would think it will). [00:47] I'm too lazy to do otherwise. [00:56] saste, what exactly is "bigint" needed for ? [00:56] it's used to store the image as a bit integer [00:56] image is a 48x48 bitmap [00:57] uhm no, wait when i'll be awake [00:57] but anyway, i may need to extend the number [00:58] but seems i can't count now [03:21] ffmpeg.git 3Michael Niedermayer 7libavcodec/ffv1.c: ffv1: remove commented asserts [03:21] ffmpeg.git 3Michael Niedermayer 7libavcodec/ffv1.c: ffv1: change w/h asserts to check as the condition can likely happen [03:33] ffmpeg.git 3Michael Niedermayer 7libavcodec/ffv1.c: ffv1enc: fix assert in put_vlc_symbol() and update to av_assert2() [05:08] ffmpeg.git 3Michael Niedermayer 7libavcodec/arm/mpegvideo_armv5te.c: mpegvideo_armv5te: change asserts to av_asserts [06:21] ffmpeg.git 3Carl Eugen Hoyos 7Changelog doc/general.texi libavcodec/Makefile libavcodec/allcodecs.c libavcodec/avcodec.h libavcodec/codec_desc.c libavcodec/targa_y216dec.c libavcodec/version.h libavformat/isom.c: Pinnacle TARGA CineWave YUV16 decoder (fourcc Y216). [12:15] ffmpeg.git 3Michael Niedermayer 7libavcodec/libvorbisenc.c: libvorbisenc: fix afq delay setting [12:15] ffmpeg.git 3Michael Niedermayer 7libavformat/segment.c: segment: reverse pass avoid_negative_ts from the muxer [12:53] ffmpeg.git 3Stefano Sabatini 7doc/filters.texi: doc/filters: fix typo [13:57] ffmpeg.git 3Anton Khirnov 7libavcodec/avcodec.h libavcodec/resample.c libavcodec/resample2.c libavcodec/version.h: lavc: deprecate the audio resampling API. [13:57] ffmpeg.git 3Anton Khirnov 7avplay.c libavfilter/af_asyncts.c libavfilter/af_resample.c libavresample/audio_data.c libavresample/audio_data.h libavresample/avresample-test.c libavresample/avresample.h libavresample/utils.c: lavr: change the type of the data buffers to uint8_t**. [13:57] ffmpeg.git 3Anton Khirnov 7doc/APIchanges libavresample/version.h: lavr: bump major to 1 and declare it stable. [13:57] ffmpeg.git 3Anton Khirnov 7doc/APIchanges: doc/APIchanges: fill in missing dates and hashes. [13:57] ffmpeg.git 3Dmitry Samonenko 7libavcodec/libspeexenc.c: libspeexenc: Add an option for enabling DTX [13:57] ffmpeg.git 3Dmitry Samonenko 7libavcodec/libspeexenc.c: libspeexenc: Updated commentary to reflect recent changes [13:57] ffmpeg.git 3Mans Rullgard 7libavcodec/x86/dsputil_mmx.c: x86: dsputil: kill VLA in gmc_mmx() [13:57] ffmpeg.git 3Mans Rullgard 7libavcodec/ppc/fmtconvert_altivec.c: ppc: fmtconvert: kill VLA in float_to_int16_interleave_altivec() [13:57] ffmpeg.git 3Mans Rullgard 7libswscale/ppc/swscale_altivec.c: ppc: swscale: rework yuv2planeX_altivec() [13:57] ffmpeg.git 3Mans Rullgard 7configure: build: error on variable-length arrays [13:57] ffmpeg.git 3Anton Khirnov 7libavfilter/vf_pad.c libavfilter/vf_scale.c: vf_pad/scale: use double precision for aspect ratios. [13:57] ffmpeg.git 3Diego Biurrun 7libavutil/parseutils.c: parseutils-test: Drop random colors from parsing test [13:57] ffmpeg.git 3Diego Biurrun 7tests/fate/libavutil.mak tests/ref/fate/parseutils: fate: Add parseutils test [13:57] ffmpeg.git 3Diego Biurrun 7: Give all anonymously typedeffed structs in headers a name [13:57] ffmpeg.git 3Anton Khirnov 7libavformat/yuv4mpeg.c: yuv4mpeg: return proper error codes. [13:57] ffmpeg.git 3Michael Niedermayer 7: Merge remote-tracking branch 'qatar/master' [13:57] ffmpeg.git 3Michael Niedermayer 7libswresample/swresample-test.c: swr-test: avoid VLA [14:07] ffmpeg.git 3Michael Niedermayer 7libavcodec/snowenc.c: snowenc: get rid of VLA (well it wasnt really variable anyway) [14:07] ffmpeg.git 3Michael Niedermayer 7libavutil/pca.c: pca: get rid of VLA [14:07] ffmpeg.git 3Michael Niedermayer 7configure: configure: enable -Werror=vla [15:02] ubitux: yes, debian/stable still ships 0.5 [18:38] ffmpeg.git 3Bobby Bingham 7libavfilter/libmpcodecs/vf_tile.c: vf_tile: fix typos/grammar in comments [18:42] do we still need this filter? [20:23] ffmpeg.git 3Michael Niedermayer 7libavformat/movenc.c: movenc: fix regression with yuyv caused by c5f23d [21:32] what's so awesome about sed?? [21:33] tetris.sed? [21:33] arkanoid.sed? [21:33] nothing [21:33] perl > sed [21:34] i mean what's so awesome about s/PIX_FMT/AV_PIX_FMT/ [21:35] apart from annoying michael at the next merge [21:35] ironically, when we added AV_ to the codec ids [21:35] they added a #define CodecID [21:36] which i had to udnefine in my downstream project [21:36] yes, but what's so awesome about cleanup? [21:36] ... because they pollutedt he namespace [21:36] it's necessary, but is boring [21:36] [15:36] <@saste> yes, but what's so awesome about cleanup? <-- as a downstream user, clean up is very welcome and i appreciate it [21:36] since libav* libs have always been catastrophic messes [21:37] Daemon404, libav* mess is overrated [21:37] i disagree [21:37] cleaner than most libs i know of [21:37] if i didnt work on the libs themselves, using it work be even more horrible [21:37] some stuff is still impossible with public api only [21:37] finding a lib which is really clean is hard, i challenge you [21:37] its really fun [21:37] [15:37] <@saste> finding a lib which is really clean is hard, i challenge you <-- ffms2 [21:37] which i in fact use in favour of libav* many times [21:38] will we sed s/AV/AVS/ when the project has an honorable subtitles support? [21:38] Daemon404, but it's another layer of complexity [21:38] the classical carpet over the dirt [21:38] libav* api is too generic [21:38] and low level [21:38] it doesnt even have "get me a frame" [21:38] most people do NOT want that [21:39] also it has fraem accuarte indexed seeking [21:39] i dare you to try and implement that in libav* [21:39] yes i wonder if there is a reason for *not* to implement that cleanly in libav* [21:39] sometimes people try, and disappear [21:40] no i never tried that [21:40] i often hear "vfr" when talking about that [21:41] ffms2 handles vfr properly [21:41] so whats your excuse? [21:42] "less than straightforward and less than perfectly documented libav API" [21:42] Daemon404: what do you think is preventing that in libav*? [21:42] except "lavf is a mess" [21:42] saste, thats entirely accurate [21:42] especially when it was written, eyars ago [21:42] ;) [21:42] this is the usual argument: ouch this is a mess, i'll build another layer of complexity rather than fix it properly at the right level [21:42] then the next guy comes and see the wrapper [21:43] saste, libav* will never ever have that simple of an api [21:43] ever [21:43] ouch this is a mess, i'll build *another* layer to cover this bloody mss [21:43] and again and again [21:43] because certain people have moral objections [21:43] to simplicity [21:43] and non-genericism [21:43] this is the story of audio in linux [21:43] "it should only decode packets!" [21:43] but people is free to do whatever they seem fit them [21:44] i would really like a lot of the logic from ffmpeg.c [21:44] moved to api [21:44] it would be -really- useful [21:44] that's what freedom is about after all [21:44] freedom doesnt mean its practical [21:44] Daemon404, my Great Evil Plan is to move all ffmpeg functionality to libavfilter [21:44] eh [21:44] tahts just not feasible [21:44] then you basically need a movie source, plus some filters [21:44] for a lot of libavcodec specific things [21:45] like? [21:46] things that would make useful higher level api for lavc [21:46] without mixing in lavfi [21:46] which [21:46] shock [21:46] adds a layer of compelxity [21:46] yes that's true [21:47] i realized it soon after [21:47] but at least it is in the same low level project [21:47] that is irrelevant [21:47] (fwiw i am very soon disabling libavfilter entirely for us) [21:47] and moving to vapoursynth [21:48] ah windows specific stuff? [21:48] no [21:49] its cross platform [21:49] avisynth replacement [21:49] i plan on writign lua bindings and makign an input module for lavf [21:50] supports video atm, and after that gets more or less finished there probably will be some kind of audio support [21:50] official bindings are for python [22:16] (did i miss something?) [22:50] ubitux, no, just (fwiw i am very soon disabling libavfilter entirely for us) [22:50] ok :( [22:50] thx :) [22:50] keep in mind i am keepign essential ones for ffmpeg.c [22:51] audio mostly. [22:51] Daemon404, why ? [22:51] you should remove them too, really :) [22:51] ? [22:51] im not sure i follow [22:52] well, i think if there are problems in libavfilter, they should be fixed [22:52] if people want a higher level interface that is simpler [22:52] one should be added [22:52] the problems cannot be fixed [22:52] its the very design that is wrong [22:52] i.e. a monolithic library [22:53] Daemon404: we're almost to the point were we could have a 150-line tools feeding libavfilter with a filtergraph supporting the whole chain; like: ./ffstreamer "movie=in.avi,...,outmovie=out.avi" [22:53] it's "scripting" rapes my eyes [22:53] more or less [22:53] id rather have real langauge bindings, and many newer plugins [22:54] not to mention easy frame accurate seeking for input [22:54] well libav closed the api recently so it doesn't help supporting additionnal users filters or stuff like that :( [22:54] its still wrong [22:54] Daemon404: did you see the [a]sendcmd from saste? [22:54] it would still require a recompile [22:54] (of the user app) [22:54] Daemon404, this is not hard to fix [22:54] ubitux, tl;dr it's no repalcement for avs or vs and its use cases [22:54] ive said this so many times [22:57] Daemon404, if it fails for some use case, please open a ticket on trac and explain the use case as a feature request [22:58] i think libavfilters low level API is quite well designed, whats missing is as you say / i understand you dynamic loading a nice easy API for such plugins [22:58] i dont feel liek opening several hundred tickets [22:58] and a real way to bind languages [22:59] i.e. lua or python [22:59] Daemon404, you seem to feel like complaining alot though (not just about libavfilter) [22:59] and no offense intended, this is IMHO not helping anyone [22:59] we should try to improve what doesnt work as it should [22:59] you are indeed correct [23:00] you must understand though, many times when i mention things, "why arent you usign libav* for this?" and similar rhetoric shoved down my neck [23:00] thats in fact what suprred this (in a different network/channel with a different dev) [23:01] apologies for carrying it over. [23:01] Daemon404: we would really appreciate a well written mail/blogpost/article/whatever about the needs and expectations you and some other people have about a filtering library, and how libavfilter fail to achieve that, as well as what kind of things would be nice [23:01] because we can't do much except listening to your incomplete complains here [23:02] i summed it up once before for libav [23:02] general response i got was "we're nto gstreamer" [23:02] or [23:02] "why would you want to do that?" [23:03] and at least one lecture on why dlopen is bad [23:03] i will write one now though and send [23:03] thank you [23:04] warning: i know ill forget some [23:04] ffmpeg.git 3Michael Niedermayer 7libavcodec/get_bits.h: get_bits: get_bits_long() support n=0 as the docs allow it [23:04] ffmpeg.git 3Michael Niedermayer 7libavcodec/x86/dsputil_mmx.c: dsputil_mmx: put optimized gmc code back and avoid a VLA without loosing features. [23:05] Daemon404: take your time :) [23:06] its the public api for lavfi documented? [23:06] is* [23:06] iirc there were complains -- i dont know if it's been rectified [23:07] there are filtering examples [23:07] http://git.videolan.org/?p=ffmpeg.git;a=tree;f=doc/examples;hb=HEAD [23:07] exampels are not documentation [23:07] i personally still find the public api a bit complex [23:08] aren't they? [23:08] they are not a substitute. [23:08] the functions are likely doxified [23:08] oh [23:09] i should also sent an email about why print_options is not a good solution [23:09] (separate issue) [23:09] also a tricky one. [23:20] "Greek harmony" = golden ratio? [23:22] saste: didn't you have a patch to remove the '-' in the ffmpeg help for the filters? [23:23] and pretty-printing related patches [23:23] kierank, yes, join the phi lovers club [23:24] ubitux, yes, if i remember where i put it [23:27] i wonder if we shouldn't remove the "AVOptions" thing too [23:27] anyway, adding some more padding, remove the '-' and put some little indent would really help [23:27] iirc your patch(es) was/were improving this [23:28] btw, kind of related, but the "Getting help" message should appear with no input [23:29] michaelni : i have a complaint about ffmpeg. the whole thing is monolithic, and gigantic! and i dont have any idea how to make it smaller, but its all rubbish ... :) [23:29] :P [23:30] ffmpeg's libraries have deps on the otehr libraries in ffmpeg in a bad way [23:30] i.e. non oublic symbols [23:30] nonpublic* [23:30] so yes, youre sort of right [23:30] :) [23:30] we should make every symbol public and that way mplayer wont have trouble.... ;) [23:31] lol [23:31] or we could just rename every symbol every three months so it looks like we are developing code... [23:31] breaks api? who cares! [23:32] eh [23:32] already renamed the main programs, that was important... [23:32] api breaks more than every 3 months [23:33] ffmpeg will be depreciated so use avconv! [23:33] wonder when it will finally be depreciated... [23:33] in the year 2022 [23:34] wait [23:34] did you just complain about ffmpeg's libavfilter options being like a script [23:35] and then immediately say it should be like avisynth's .. options ... being like a script that you have to make a script to use? [23:35] Daemon404 : do you want feature request to be able to use ffmpeg via script like avisynth ? [23:36] Compn, either that or people stop pushing lavfi on me as an avs replacemeny so zealot-like [23:36] Compn: give him a chance to express him properly, let's not ask him to much for trolling here :) [23:36] ubitux : lol ok , we troll later :) [23:37] we'll troll after the post explaining deeply the issue [23:37] k [23:37] which will be far more interesting and constructive :) [23:37] (hopefully) [23:37] Daemon404 : who said to replace avisynth with libavfilter ? thats dumb. libavfilter doesnt have half of the filters that avisynth has [23:37] its not ready to be a replacement [23:37] Compn, response: [23:37] i said stop it Compn! :( [23:37] PORT IT ALL [23:37] PATCHES WELCOME [23:38] see... :() [23:38] :( [23:46] Action: Compn realizes that avisynth is based on vfw, remembers getting scolded about using ffdshow because it was too 'ancient' and lav filters wat hte future [23:46] Action: Compn brain explodes [23:47] ah avisynth 3 was going to use a real language for scripts, instead of crazy-reinventing windows batch files [23:47] ubitux : so talk to me about something so i stop trolling [23:47] multimedia.cx is back up [23:47] do you love pok?mons? [23:47] yay [23:48] no not really [23:48] :( [23:48] like power rangers, their novelty has died on me [23:48] and pogs [23:48] pogs... i almost forgot about that :D [23:49] 22:23:31 kierank, yes, join the phi lovers club --> can you perhaps call it the golden ratio instead [23:50] i don't think i've ever heard it called greek harmony in english [23:51] kierank, it was subtly humorous [23:51] oh [23:51] whoosh [23:51] but i don't mind if you want to change it (PAWE) [23:59] ffmpeg.git 3Michael Niedermayer 7libavcodec/x86/dsputil_mmx.c: gmc_mmx: enable also for large pictures when emu edge isnt needed. [00:00] --- Sun Oct 7 2012 From burek021 at gmail.com Sun Oct 7 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Sun, 7 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121006 Message-ID: <20121007000502.2CD0E18A018E@apolo.teamnet.rs> [00:09] rainmaker1 can you please use a pastebin site (like www.pastebin.com) to show your ffmpeg command and the complete console output? [00:25] burek: sure if you are still awake :) [00:25] :) [00:25] cool :) [00:29] Here it is http://pastebin.com/UB3PzGpC [00:38] burek: ant idea how to discard some streams? ie subtitles [00:41] im here [00:41] just a sec [00:43] ok ok [00:45] rainmaker1, you did select correct program id [00:45] so, now, you want what? [00:45] discard everything except audio/video? [00:47] hmmmm [00:47] if you look at pastebin [00:47] I have a mux of multiple live channels [00:47] every channel I want to transcode [00:48] audio/video [00:48] and nothing else [00:48] is this possible? [00:48] well, you selected only one program (channel) 7002 [00:49] meaning, you selected all of its streams [00:49] yes and I want to use audio/video only and I don;t want subtitles as I receive above error [00:49] and there are 4 of them, exactly how ffmpeg selected them [00:49] ok [00:49] ok, so you want to specify only audio/video from program 7002 ? [00:49] bingo :D [00:50] take a look at the format of the -map option: http://ffmpeg.org/ffmpeg.html#Advanced-options [00:50] -map [-]input_file_id[:stream_specifier] [00:50] now, what is ? take a look here: http://ffmpeg.org/ffmpeg.html#Stream-specifiers-1 [00:50] p:program_id[:stream_index] [00:50] now, you might try: -map 0:p:7002:a -map 0:p:7002:v [00:50] instead of: -map 0:p:7002 [00:51] Aha! [00:51] Now tha main error [00:52] Failed to compensate for timestamp delta of -21867.146792 [00:52] I always get this error [00:53] Now I have read something like I can specify a synchronization stream or something like that but it is not clear to me [00:53] how I know what is synchronization stream? [00:53] pastebin? [00:54] sure [00:56] http://pastebin.com/amaZH6Ly [00:59] rainmaker1, for some reason that didn't work [00:59] try this [00:59] -map 0:p:7002:0 -map 0:p:7002:1 [01:00] same error :* [01:00] :( [01:00] not * but ( [01:00] :D [01:00] pastebin? :) [01:01] sec [01:03] here it is [01:03] http://pastebin.com/HEJymGdt [01:05] pastebin sure got slow in last couple of days [01:06] became* [01:06] I dont use it on a daily basis [01:06] :) [01:06] that's good so far (your pastebin) [01:07] now [01:07] can you start with something like this [01:07] ffmpeg -f mpegts -y -i udp://237.0.0.230:5000?overrun_nonfatal -map 0:p:7002:0 -map 0:p:7002:1 output.ts [01:07] and slowly add more and more options [01:08] until you get your current command line [01:08] just to be able to see where does the error occur [01:08] well I am not sure [01:08] frame= 0 fps=0.0 q=0.0 Lsize= 246kB time=00:00:00.00 bitrate= 0.0kbits/s dup=0 drop=344 [01:09] is this ok to have a huge number of drops? [01:09] well, if fps stays at 0 thats not good [01:09] well yes [01:10] it's a strange [01:10] either your input is invalid, or the ffmpeg's decoders are buggy [01:10] I belive in second :( [01:10] it would also be good to capture a raw input and save it as a dump file [01:10] because input is used on some pro equipment [01:10] like 10-20 mb [01:11] in case you need to provide a sample, so that developers can see what is wrong [01:11] and fix the bug (if any) [01:11] I will try to open a bug report just need to read a rules how to post a bug report [01:12] Thank you for your time burek, and have a good night :) [01:12] you might try also: ffmpeg -y -f mpegts -i udp://237.0.0.230:5000?overrun_nonfatal=1 -map 0 -c copy -t 120 output.ts [01:13] np :) [01:13] nope [01:13] [mpegts @ 0x2eeb200] sample rate not set [01:14] Could not write header for output file #0 (incorrect codec parameters ?): Invalid argument [01:14] it would help to capture that udp input to some raw file.. [01:14] I'm just not sure how [01:14] yes, via tcpdump or tshark [01:15] I guess [01:15] will try [01:15] thanks again :) [01:15] :beer: :) [01:15] and good night, take a look at a clock :P [01:15] :) [01:16] the night is young :) [01:16] gn :) [01:16] :D [06:07] hey have you guys heard of http://www.theaudiodb.com? [10:54] how do you do I-frame only mpeg2? [12:35] can someone explain how I can map two different audio streams to the same output file? [12:36] im currently using "ffmpeg -y -f x11grab -s "1366x768" -r 10 -i :0.0 -f alsa -i pulse -f pulse -i default -vcodec libx264 -s "640x360" -acodec libmp3lame -ab 128k -map 0:0 -map 1:0 -map 2:0 asdasd.mp4" [12:36] King_Rat, -map, check examples in the doc [12:36] i did, and got confused [12:38] i'd have thought it would work but it only seems to take a first audio stream [12:38] pb King_Rat [12:39] http://pastie.org/pastes/4921439/text [12:40] King_Rat, looks correct [12:40] what's wrong with it [12:41] (FUCK I need to scroll xchat to see the posted messages) [12:41] King_Rat, if you want to *merge* the audio streams you need amerge/amix [12:41] oh right [12:45] so I don't need to use maps at all? [12:46] King_Rat, I don't think so, you should find some examples in the docs (amerge/amix, with somewhat different features) [12:46] oh cool, i got it working [12:46] sorry, i naively thought ffmpeg would merge the streams by default [12:48] no, fortunately ffmpeg doesn't try to be "clever" and preserve the input structure [12:48] yeah, i suppose that's much more sensible [12:49] thanks for the help, and sorry about my incompetence [12:49] you're welcome [12:49] ::donate button:: :) [12:52] im afraid i dont have much money to spare [12:52] i can draw you a picture of you like though [12:54] King_Rat, if you're an artist you may consider to send us a logo [12:54] we have a sort of contest for a new winter logo [12:55] http://ffmpeg.org/trac/ffmpeg/wiki/SubmitALogo [13:01] i'm a terrible artist [13:02] and im generally against software having logos [17:10] I recently read news that ffmpeg 1.0 was out. [17:10] So why does mine say "ffmpeg version N-45080-gd9dfe9a"? [17:10] (Just downloaded it.) [17:11] What absurd version number is that? [17:12] hi all [17:12] cheeseduck<< its the trend, you don't like that cool name ? [17:13] hi, i'm using Serviio which uses ffmpeg to transcode video and serve it over dlna/upnp or flash ( http ), I used this guide to build it myself: http://wiki.serviio.org/doku.php?id=build_ffmpeg_linux and my version info looks like this: http://pastebin.com/BDbQKYZa I'm on a ubuntu server 11.04 with a amd 760g chipset and the kind of output I always get is this: http://www.zimagez.com/zimage/schermafdruk-05-10-12-142254.php , and the sound is ver [17:13] y screechy and unuseable. does anyone have any idea what could go wrong? I searched for a solution since a week without result. [17:13] No. [17:13] Can you explain what the hell is going on? [17:13] i don't even read those numbers [17:14] doesn't really matter [17:14] i have one problem with ffmpeg currently, it fails to handle the datastream of my cheap webcam [17:29] how to compile ffmpeg using vs on windows? [17:49] bigmeow<< no idea [17:50] Hi, I'm on debian screeze and I'm trying to record a screencast for youtube. I'm trying to capture lossless video first, using this command: ffmpeg -f x11grab -r 30 -s 1280x800 -sameq -i :0.0 -f alsa -i pulse meow.avi [17:51] and it doesn't work? [17:51] But it seems kind of laggy when I play back the video. [17:51] the sound? [17:51] No, the video, the sound is fine. [17:51] try mjpg [17:51] Also, I have no idea what I should transcode it to, if anything, for upload to youtube. [17:51] or something faster [17:52] h264 mkv is the smallest [17:52] Okay, thanks, I'll try that... [17:53] you might want to try out h264 fastest setting, if your computer can handle that [17:53] as a compromise [17:55] Is there a way to list available codecs? [17:56] I tried -vcodec libx264 and it says it cannot find that codec. [17:56] But I've installed all the x264 packages in squeeze [17:58] Specifically, Unknown encoder 'libx264' [18:22] hi, I need some help with encoding [18:23] I'm using the comman ffmpeg -i \[Coalgirls\]_Mobile_Suit_Gundam_Unicorn_01_\(1920x1080_Blu-ray_FLAC\)_\[CF793FEE\].mkv -c:v copy -aq 200 Gundam_Unicorn_01.mp4 [18:23] command* [18:23] and I'm getting the error: Error while opening encoder for output stream #0:1 - maybe incorrect parameters such as bit_rate, rate, width or height [18:26] paste the terminal output onto a pastebin and link there [18:26] *here [18:26] I'm pretty sure the actual error is higher up :P [18:29] it was [18:30] hold on [18:31] http://pastie.org/4922687 [18:31] "Codec is experimental but experimental codecs are not enabled, see -strict -2" [18:32] yup [18:32] -strict experimental or -strict -2 [18:32] is needed [18:32] to enable the internal aac encoder [18:32] also I'm pretty sure aq 200 doesn't do what you think it does [18:33] variable rate encoding? [18:33] I thought that was supposed to be variable rate encoding for aac with 255 the highest quality, and 0 the lowest [18:34] I think it's the other way around, and I'm not sure if it even goes 0-255 [18:35] oh ok [18:35] at least looking from the side of H.264 where QPs go from 0-51, but not like I know how to the -aq setting works [18:38] any idea how to extract the subtitles from an mkv? [18:39] -c:s copy [18:40] ok [18:41] mooglenorph<< try ldconfig [18:47] -c:s copy isn't working [18:48] -scodec copy does try to copy the english subs, but it errors out [18:48] mainly because the codec is not supported in mp4 [18:50] try -c:s mov_text [18:51] you won't retain the advanced ass markup though [18:51] ubitux: ok [18:51] your version might be too old btw. [18:52] you are two major versions behind [18:52] at least upgrade to 1.0 [18:53] ok [18:57] I downloaded a static build and I'm getting this now: [libvo_aacenc @ 0x2d62b60] Unable to set encoding parameters [19:02] ok, maybe I now have it working, switched to mp3 for audio codec [19:38] Hi All, trying to extract subtitles with: ffmpeg -i /data/Masters\ of\ Money_121001.rec -an -vn -sbsf mov2textsub -c:s copy -f rawvideo sub.txt and I'm getting this error: "Unrecognized option 'sbsf' Failed to set value 'mov2textsub' for option 'sbsf'" Am I doing something wrong? [19:38] The video is from a Topfield PVR [19:59] I recently read news that ffmpeg 1.0 was out. So why does mine say "ffmpeg version N-45080-gd9dfe9a"? [19:59] What absurd version number is that? [20:17] a git identifier [20:18] [/tmp/ffmpeg-1.0]- ./version.sh [20:18] 1.0 [20:18] it's likely your 1.0 is build from a different version [20:19] likely a more recent one [20:21] with git/master you would have: N-45112-g51211d3 [20:21] N-45080-gd9dfe9a is 32 commits before the latest bleeding edge "version" [20:22] Hrm... [20:22] ubitux: I got mine from here: http://ffmpeg.zeranoe.com/builds/ [20:22] "FFmpeg git-d9dfe9a 64-bit Static (Latest)" [20:22] Which looks like the last stable? [20:23] a bit more recent [20:23] Then I have no idea what the actual latest stable is on that page. [20:23] Why a billion download links? [20:23] why do you care about the "stable" one? [20:23] unless you are a distributor, the latest git version is recommended [20:23] Because I don't like running unstable stuff? [20:24] git head isn't less stable [20:24] Let's say I *am* a distributor. Where is that? [20:24] you have to build it yourself [20:24] I don't understand this weird ambiguity. [20:24] And why they don't use proper version numbers. [20:24] do you understand what git is? [20:25] Yes. [20:25] zeranoe is just distributed (automated?) build based on the latest version [20:25] you have multiple build for multiple configuration preset [20:25] we recommend our users to use the latest git because we don't release "stable" version often [20:26] maybe you could ask in the zeranoe forum for 1.0 releases [20:27] so well, the version you see is just a commit identifier [20:28] if you want to do yourself some "stable" ffmpeg windows builds, you can go grab the 1.0 sources on http://ffmpeg.org/download.html#releases [20:28] and make it yourself [20:28] Well... [20:29] most (all?) the developers are *nix only people, so we don't have windows-manpower to do that ourselves [20:29] I don't see why I'd want to build it locally. [20:29] i don't know your needs [20:30] The latest is probably gonna cut it for me, but I still find it odd. [20:30] you can discuss this with this particular distributor [20:30] as ffmpeg developers, we don't really control that [20:53] guys, question - there is a video stream, made available to us as .mpg file. the file is huge though 400 mb [20:54] can ffmpeg convert this to an acceptable quality, to .mp4 or some other smaller, more efficient container / format? [20:54] I dont wanna backup 10 x 400 mb :( [20:54] 5-30% quality loss is absolutely fine [21:11] hello, i have some problems with an MTS file. i want to turn into a .mov but i get this error [21:12] [mov @ 0x8312ae0] Application provided invalid, non monotonically increasing dts to muxer in stream 0: 1 >= 1 [21:12] av_interleaved_write_frame(): Invalid argument [21:12] what does that mean? [22:49] Hellooo [22:50] Is it possible to get [22:50] frame= 23 fps= 5 q=23.9 Lsize= 813kB time=6.60 bitrate=1008.9kbits/s [22:50] More frequent? [23:58] Hello? [00:00] --- Sun Oct 7 2012 From burek021 at gmail.com Mon Oct 8 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Mon, 8 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121007 Message-ID: <20121008000501.80BDA18A018B@apolo.teamnet.rs> [04:26] hey guys [04:27] got a question here, not exactly related to ffmpeg but i thought you guys might be able to help [04:27] i'm trying to write a vfw codec that utilizes intel's quick sync feature [04:27] and i'm having trouble building the encoder example [04:28] this is the sdk: http://software.intel.com/en-us/vcsource/tools/media-sdk [04:28] according to the documentation, it should be built using vs2012 on windows 8 [04:54] :/ I've got an mp4 video and I'm trying to cut a short section out. The video has aac audio and h264 video, and apparently I want to split it off of an I frame (whatever that is), so setting -vcodec h264 works fine for that (and the video starts where I want it to), but the audio seems to be playing from that I Frame instead. [04:55] I'm doing -acodec aac and setting the other settings to match the old one [04:55] trying -async 1 now [04:56] audio matches up at about 4seconds in, but is missing before that :/ [05:28] hi, im using ffmpeg to encode some image files into a video, ffmpeg -f image2 -r 1 -i pix/%03d.png -target ntsc-dvd -aspect 4:3 out.mpg the command im using [05:28] for some reason some of the png files is opened but skipped and not encoded [05:28] (from what i see from strace) is there a reason how that can be? [05:29] i looked at the images being skipped, the dimensions to them are bigger than the ones that are encoded 3264 x 2448 [05:29] other than the dimension it the png encoding is the same [05:46] hello [05:46] i have a problem with a video h264 [05:46] can somebody help me? [05:48] can anybody help me? [05:48] http://pastebin.com/HtWy1CSi [05:48] there it is the line i use and the reslt [05:48] result [05:58] please anyone? [06:03] can anybody help me? [06:03] i have a problem with a video h264 [06:03] http://pastebin.com/HtWy1CSi [06:03] there it is the line i use and the result [06:05] i found some online answers saying it has to be the same size, so i guess i need to resize them [06:33] Im getting a weird result when i add -timecode my footage end up 5 times longer than it should.. [11:11] Is it possible to get [11:11] frame= 23 fps= 5 q=23.9 Lsize= 813kB time=6.60 bitrate=1008.9kbits/s [11:11] Is it possible to get [11:11] More frequent?* [13:43] Marlinc, what exactly are you trying to accomplish? [13:44] can you provide a pastebin of your command line + output? [13:45] thevdude can you please use a pastebin site (like www.pastebin.com) to show your ffmpeg command and the complete console output? [13:52] ffmpeg -y -f x11grab -i 1300x744|:0.0+65,24 -r 60 -s 2806x900 -b 50k -minrate 50k -maxrate 50k -bufsize 1835k -vcodec libx264 -f flv hoi.flv [13:52] brb [13:53] Back [13:53] burek, ^ [13:54] are you sure this is valid "-f x11grab -i 1300x744|:0.0+65,24" [13:54] http://ffmpeg.org/ffmpeg.html#x11grab [13:55] maybe you wanted to say something like this: -f x11grab -s 1300x744 -i :0.0+65,24 [13:56] I thought the same [13:56] unless your 'hostname' is actually named '1300x744|' [13:56] which is not likely [13:56] But okay [13:57] My question was about the " frame= 23 fps= 5 q=23.9 Lsize= 813kB time=6.60 bitrate=1008.9kbits/s" messages [13:57] Marlinc can you please use a pastebin site (like www.pastebin.com) to show your ffmpeg command and the complete console output? [13:58] https://gist.github.com/08b41b27125186e90d05 [13:59] you are not using ffmpeg [13:59] apparently avconv is the same as ffmpeg [13:59] According to the ffmpeg command [14:00] if it was, you would be using ffmpeg [14:00] go to libav support channels and ask there for the support [14:00] it's just a fork project of ffmpeg [14:00] The ffmpeg command saids that ffmpeg is deprecated [14:00] ubitux :) [14:00] And that I should use avconv [14:00] we should really create that wiki page [14:00] *** THIS PROGRAM IS DEPRECATED *** [14:00] This program is only provided for compatibility and will be removed in a future release. Please use avconv instead. [14:00] that explains all the mess that debian/ubuntu maintainers have created.. [14:01] Marlinc, long story short [14:01] I see... [14:01] that's a lie.. [14:01] sadly.. [14:01] Okay.. [14:01] they forked ffmpeg project and wanted all people to just forget about ffmpeg [14:01] just like it never existed [14:01] can't help much about that [14:01] but if you install ffmpeg I might help [14:01] Okay that's pretty sad [14:01] until then, you'll have to ask there [14:02] i know.. [14:02] As far as I know I have ffmpeg installed [14:02] we will soon create a page where people will be able to get ffmpeg deb package from [14:02] Okay [14:02] but until then, you'll just have to compile it from source.. [14:02] So [14:02] I'm creating a software package [14:02] That uses ffmpeg (or avconv) [14:03] And I would like to use the apt software manager for Ubuntu [14:03] So using debs [14:03] As a way to distribute the software [14:04] Does the matter if I specify ffmpeg as dependency? [14:04] Because it looks like it installs avconv if that's what you're saying [14:05] you might join #ffmpeg-devel and ask there, because it's kinda mess right now with debian/ubuntu [14:05] so to get the precise answer [14:05] you would need to ask developers for advice [14:06] Damn [14:06] Are there bug differences? [14:06] Big* [14:06] obviously :) [14:06] Can I use ffmpeg and avconv with the same arguments? [14:06] I wouldn't know, since I don't use avconv, so, I dunno really.. [14:07] Okay [14:07] But okay is it possible in ffmpeg [14:07] To get the " frame= 23 fps= 5 q=23.9 Lsize= 813kB time=6.60 bitrate=1008.9kbits/s" messages more frequently [14:07] I'm assuming ffmpeg has it too? [14:08] why do you need those log messages more frequent? [14:08] Because I'm using the fps and frame parts in my application [14:09] I'm trying to make a indicator that activates if the fps drops or the amount of frames get slowed down [14:09] maybe -log_level 99 ? [14:09] And I can't really create a accurate indicator if I don't get the frame amount that often [14:09] did you check the docs? [14:09] Yes but I didn't know that to look for [14:10] I did use the man page [14:10] brb [14:12] could anyone give me some help with using x11grab for fullscreen wine windows? [14:12] back [14:12] i dont really have a clue how fullscreen programs like that work [14:13] Unrecognized option 'log_level' [14:13] Failed to set value '99' for option 'log_level' [14:13] :p [14:13] burek, ^ [14:13] it's not ffmpeg.. [14:14] do any of x's numerous little tools allow you to list all displays? [14:14] maybe -loglevel [14:14] instead of log_level [14:14] or something [14:14] O yes.. [14:14] Marlinc: http://ffmpeg.org/ffmpeg.html [14:14] ctrl+f loglevel [14:14] I thought I changed the command but yes.. Ubuntu uses avconv [14:15] What mess [14:15] King_Rat: https://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20grab%20the%20desktop%20(screen)%20with%20FFmpeg [14:15] Is it possible to provide a binary with my software? [14:15] Or would it cause library problems [14:16] Marlinc, if you're asking for licensing stuff or so, read here: http://ffmpeg.org/legal.html [14:16] for libs dependencies and stuff, please ask in #ffmpeg-devel, since I can't help a lot with it [14:16] Okay [14:16] when I try and capture in fullscreen everything goes black [14:17] Really realizing this software is getting harder and harder [14:17] when creating a video from from images, with an explicit video size (i.e "-s WxH"), is there a way to have the image preserve its aspect while fitting within the boundaries of the video (the images are resizing to WxH, but don't necessarily have a W:H aspect ratio) [14:18] m3thyl you might use pad filter [14:18] http://ffmpeg.org/ffmpeg.html#pad [14:18] King_Rat can you please use a pastebin site (like www.pastebin.com) to show your ffmpeg command and the complete console output? [14:19] ok; i was looking at that but was wondering if ffmpeg could do the calculations for me; thanks [14:19] Marlinc, at least it's never boring :) [14:19] wait, i'll brb in 10 minutes or so [14:19] m3thyl, it can, just take a look at examples [14:19] alright [14:36] burek, https://gist.github.com/a830ddb9d1038367bacb [14:36] Whats the difference between the two frame amounts [14:37] One says 281 and the other says 323 [14:38] shevy, did you try: ffmpeg -i input.mpg -map 0 -c copy -c:v libx264 -crf 25 -preset slow output.mp4 [14:39] Marlinc one is reported by ffmpeg (that is calling the libx264 lib) and another is the output of the libx264 lib itself [14:39] the timing of those 2 is what makes it different [14:39] Okay.. [14:40] (I guess) [14:40] btw, if you need that accuracy, why don't you code your own libav* wrapper [14:40] instead of using cmd line [14:41] or modify ffmpeg to display that log line more frequently and recompile it again [14:41] I'm using Java because it makes it easier to make it cross-platform [14:41] burek not yet, will try, thanks [14:41] and java is calling ffmpeg? [14:41] shevy :beer: :) [14:41] hehe [14:41] Yes at the moment I'm using the ProcessBuilder class from Java [14:42] that means you'll expect from the end-users to have to install ffmpeg on their computer prior to using your app? [14:43] I'll let the application download it I think [14:43] Or provide it in the package [14:43] For Ubuntu I can simply add a dependency to the deb file [14:43] if you provide it in the package, you'll have to know the target platform before you pack it [14:43] I know [14:43] to know which ffmpeg binary to pack (exe, elf, etc) [14:44] But I already need to that because of SWT (the window library I'm using) [14:44] also, I think, legally, you'll have some issues if you "just pack ffmpeg binary" into your app package [14:45] Then I think I should do the same as ffsplit [14:45] did you read this already: http://ffmpeg.org/legal.html [14:45] it might help [14:45] Okay [14:45] just so you don't waste time on something that will end up like not recommended approach anyway [14:46] Yes I would need some help with that [14:48] But okay [14:49] I'm running the ffmpeg library quite literally [14:49] The application just provides the right arguments [14:50] you mean binary? [14:51] burek: still need help? [14:51] always :) [14:51] binary yes [14:51] Sorry [14:52] i guess i'll have to backlog? [14:52] Marlinc is one of those debian/ubuntu users [14:52] who were affected by avconv and stuff [14:52] so, in order not to advise him badly, I've asked you (and others) to suggest Marlinc what should he do [14:52] Jup.. [14:53] does the static build solve his problem? [14:53] You're asking burek right? [14:54] yeah but talking about you :p [14:54] :p [14:54] I didn't advise him to use static build, as I didn't know what exactly is his use case [14:54] Marlinc: did you try the static build? :) [14:54] Nope [14:54] ok ok well, i just woke up, give me a few minutes i'll backlog [14:54] ok :) [14:54] I just heard about the difference between avconv and ffmpeg [14:55] I though it was the same so I could just use the Ubuntu package and have the same results on other platforms using ffmpeg itself [14:58] ok well, first i don't even understand what the problem is [14:58] Marlinc: do you mind explaining again what's not working? [14:59] There wasn't really a problem [14:59] More of a question [14:59] I'm using the "frame= 323 fps= 10 q=69.0 size= 192kB time=27.65 bitrate= 56.8kbits/s" messages to provide some useful information in my application [14:59] But because it only shows op like one's a second? [15:00] I can't really provider accurate data [15:00] Provide* [15:00] So my original question if it is possible to get those status messages more often [15:01] So my original question was if it is possible to get those status messages more often [15:01] check if the -progress option has a higer rate [15:01] higher* [15:02] Okay thanks [15:02] I didn't know what to look for [15:02] So thats why I came here [15:02] it might be the same, but it looks easier to exploit for apps than the cmd output you're pasting [15:03] i never used it so well.. [15:03] anyway, i'm not sure it's available in the fork [15:03] so if not [15:03] Okay thanks [15:03] you can try this ^ [15:03] Is it allowed to provide the binary package with my application? [15:03] and check if it helps with your problem [15:03] it depends [15:04] It isn't commercial [15:04] is the source code available? [15:04] Fully open-source [15:04] it also depends if you build ffmpeg with --enable-gpl or no [15:05] I read about some license problems with libx264? [15:05] On the legal page [15:05] as well as if you build it with --enable-nonfree and stuff like that [15:05] Because I do need libx264 [15:05] i'm not sure, but afaict, the safe choice is to have your app under GPL [15:06] but i'm not a lawyer so.. [15:06] But one moment going to take a look at the progress option [15:07] mmh afaict it seems the progress rate is also 1 second [15:09] do you really need a higher rate? [15:09] I'm trying to make a indicator so the user of the application can listen to the amount of FPS he has [15:09] While in game [15:10] So he can hear what happens to the FPS when he is in game [15:10] Without looking at the application [15:10] huh? [15:11] How do I explain it [15:11] When in game you can't see the FPS of ffmpeg right? [15:11] Because your games window is in front of it [15:12] You can't see ffmpeg and you can't see my application that runs ffmpeg [15:12] Okay? [15:12] yeah right ok [15:12] and then why ~1 second is not fast enough? [15:12] My application makes a sound every 50 frames or so [15:13] So the use can sort of listen to the speed of the frames [15:13] So the user can sort of listen to the speed of the frames [15:13] Sometimes it doesn't work correctly because there is to much time between progress update [15:14] well, increasing the rate will likely slowdown the processssssssss [15:14] oups [15:14] Could have various reasons one being the FPS goes very high. Higher then 50 fps and thus missing a beep [15:14] progress* [15:14] Really that much? [15:15] i'm not sure how much but well... [15:16] anyway, your approach looks weird [15:16] I don't know what is does on the ffmpeg backend but isn't it just a message? [15:16] what would happen if you have a sound each 50 frames and ffmpeg was producing 50 frames each 0.0001 sec? [15:17] That's not likely [15:17] why? [15:17] It is for live streaming [15:17] ok [15:17] anyway, the fps is printed every second for obvious reason [15:18] Pretty obvious indeed [15:18] since it counts the number of frame per second :p [15:18] another solution would be to use the api [15:18] and do something every 50 frames [15:18] Okay [15:19] Are there Java bindings for the API? [15:19] no [15:19] ah maybe you could trick with -vf showinfo [15:20] What would that do? [15:20] and parse the n value [15:20] Moment I'll try it [15:20] See what it does [15:23] How does it work ubitux ? [15:23] try ffmpeg -f lavfi -i testsrc -t 10 -vf showinfo -f null /dev/null [15:24] ffmpeg -f x11grab -i ... -vf showinfo -f null /dev/null [15:24] *"-f null -" is even simpler!) [15:24] ("-f null -" is even simpler!) [15:24] oh right [15:24] Ah I see [15:25] And n provides the frame? [15:25] it's the frame count [15:25] Ah nice [15:25] So I get info about every frame [15:26] http://ffmpeg.org/ffmpeg.html#showinfo [15:26] yes [15:26] Thats I'll take a look [15:26] Thanks* [15:28] What did the -t option do? [15:28] Time in seconds? [15:28] http://ffmpeg.org/ffmpeg.html#Main-options [15:28] Ah [15:28] Okay [15:29] Thanks I'll see that I can do with this [17:12] It's been a while. My issue is that audio isn't starting from where I'm telling the video to start from when cutting a clip out, it's starting from the I frame before the cut. [17:14] command is ffmpeg -ss 00:05:13.0 -i ~/videos/to_trim.mp4 -vcodec h264 -acodec aac -ac 2 -ar 48000 -ab 225k -t 00:00:14.0 -strict -2 -y ~/videos/trimmed.mp4 and output is http://sprunge.us/ZjcS [17:18] thevdude: move the seek after the input, and -vcodec libx264 [17:18] :/ my video is fine, but I'll try it [17:18] the video is starting in the right place, the audio isn't. [17:18] add -async 1 [17:19] that makes the audio sync up after about 4 seconds [17:19] and makes grey frames show up at the start, let me try it with all the changes [17:19] :D [17:22] holy poop it works [18:29] So I just noticed that ffmpeg can run using pipes [18:29] Is possible to have one output [18:29] And multiple input processes [18:30] And switch between multiple inputs so that something else appears on the output stream? [18:30] If so can somebody push me in the right direction? [18:31] burek, ubitux ? [18:33] i don't understand [18:34] You can pipe video output from one process to the other right? [18:36] yes? [18:36] -f rawvideo -pix_fmt ... [18:37] Okay [18:37] So is it possible to switch between inputs? [18:39] i don't know what that means :p [18:40] So I would have multiple ffmpeg [18:40] ffmpeg -i ... -i ... -i ... ? [18:40] So I would have multiple ffmpeg's that read a file and put the output to stdout [18:40] Okay? [18:40] And one ffmpeg that reads the video from stdin [18:41] really i have a hard time figuring out what you want to achieve. [18:41] I don't know how to explain it.. [18:42] describe what are your different inputs, and what's your output [18:42] For example [18:42] One ffmpeg captures my screen [18:42] One captures a file [18:42] And then I have a ffmpeg that reads from stdin and puts it on a livestream [18:42] See where I'm going? [18:43] how do you want to merge those 3 inputs? [18:43] you want one livestream out of these 3? [18:43] Yes but I don't want to merge [18:43] I want to be able to select which one gets send to the livestream [18:43] And switch between them without stopping the streaming [18:44] i'd say you would push to 3 ffm streams on a ffserver and let the user decide which one to lurk [18:44] otherwise i dunno [18:45] with the api most likely [18:45] with the tool dunno [18:45] Okay okay.. [18:45] I think I'm just going try some things and see what it does [18:47] quick question, how do i apply the fade video filter correctly? as of now it isn't giving me any results here's the ffmpeg output and input: http://pastebin.com/9cG4txEH [18:49] works for me, and it should start fading around frame 960 [18:49] with your cmd line [18:50] ah fuck... that was my bad the video has 900 frames :P [18:51] it should do something like what you get with ./ffplay -f lavfi -i testsrc -vf fade=out:50:40 [18:51] (at 2 seconds) [18:51] (25 fps) [18:54] yes works perfectly, was totally me being stupid [18:54] klaxa: btw, sameq doesn't mean sameq [18:55] same quality* [18:55] :) [18:55] also, why do you select mpeg4 codec? [18:58] ubitux, could 'select' video filter maybe be upgraded so that it supports selecting an input stream (just like -map does) based on some criteria [18:58] that could solve the switching input's problem :) [19:01] someone could write a streamselect right :) [19:01] could be fun [19:01] shouldn't be that hard [19:01] maybe there are some problem between framerate or resolution mismatch [19:02] yes but it can always spit out an error and refuse to play along if prerequisites are not met :) [19:03] it would be like watching the tv :) [19:03] you tv spit out errors? [19:03] your* [19:03] i'm not willing to write such filter anyway [19:03] but yeah, that could be done [19:04] I was just asking if it would be too much of work, so that I know if it makes sense to write a feature request :) [19:05] could make sense yes [19:07] <@ubitux> also, why do you select mpeg4 codec? <-- because the video will be played on a machine i don't know, i don't want it to lag, quality is good enough, i didn't notice any artifacts [19:09] ok [19:42] http://www.youtube.com/watch?v=9JsDU3CuWNs this is what you guys helped me do. <3 thanks. [19:44] is it me or most of the sound is on the left channeL? [20:50] Codec is experimental but experimental codecs are not enabled, try -strict -2 [20:50] ^^^ Anyone knows how to enable experimental codecs? so I can omit typing -strict -2 [20:50] is from converting a .flv to .mp4 [21:15] shevy, why do you need to omit -strict -2 [21:15] or -strict experimental (same thing) [21:24] burek I dunno, it seems to be not useful for me to type it [21:26] I suppose there is no configure option to allow this [21:26] (I compile ffmpeg from source usually) [21:35] it's not a configure option [21:35] but cmd line option [21:37] why not just alias ffmpeg in your .bashrc (or whatever shell you are using) [21:58] hello, can anybody help encoding mp4 to flv? i am getting an error [21:58] i can past it in paste bin [00:00] --- Mon Oct 8 2012 From burek021 at gmail.com Mon Oct 8 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Mon, 8 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121007 Message-ID: <20121008000502.AD05118A01DE@apolo.teamnet.rs> [01:42] reading Daemon404 mail now [01:42] u mad? [01:42] :P [01:42] ;p [01:43] somewhat hilarious that no one has a problem with it but you [01:43] :P [01:43] Action: Compn trolls harder [01:44] noone has a problem [01:44] because noone uses it [01:44] you know mplayer can use it , right ? [01:44] -vf lavfi [01:44] mplayer devs == ffmpeg devs [01:44] theyre the same people [01:44] and this not relevant [01:44] well doesnt it answer your question about what its supposed to be ? [01:45] a filter library for ffmpeg and 3rd party encoders/players ? [01:45] i'm only asking dumb questions because i am dumb [01:45] so dont be offended by them [01:47] to answer your question on dynamic plugins , yeah i think its a great idea [01:47] it would just take one dynamic loader to be compiled, then any filter could be dropped in [01:58] ok finished reading your mail [01:58] you did quote someone who has a problem with lavfi :) [01:58] dont mind me [01:59] Daemon404 : i am curious since you bring up NIH, do you want ffmpeg to just add avisynth support or do you want them to port filters ? [02:00] ffmpeg has avisynth support [02:00] and im working on vapoursynth support for my own purposes [02:00] then i'm curious what you want lavfi to be [02:00] i dont htink lavfi should exist. [02:00] do you have a better solution on linux? [02:01] you think its not good for users and wastes time for devels ? [02:01] vapoursynth runs on linux [02:01] and mac [02:01] and windows [02:01] and even bsd. [02:01] ok [02:01] ubitux : there is also avxsynth , bbut i'm not sure what it is :P [02:01] Compn, its an abomination created by netflix [02:01] i just reading wikipedia, i dont know these things [02:02] to answer your question about why lavfi started [02:03] was it my idea? so ffmpeg could take the filters, and mplayer wouldnt have to maintain them ? [02:03] so vlc could use them too [02:03] and then make it a wrapper for all other opensource filters [02:04] combine forces [02:04] since ffmpeg had all the decoders and muxers [02:05] [20:03] < Compn> and then make it a wrapper for all other opensource filters [02:05] problem right here [02:05] dat monolthism [02:06] its like arguing the kernel should stop supporting systems ... because its too big [02:06] or something [02:06] a strange position to be , i dont understand [02:07] actually [02:07] the linux kernel [02:07] design wise [02:07] is the shittiest of all modern kernels [02:07] of course , i come from mplayer , which bundled so much stuff ... [02:07] it just has buttloads of devs [02:07] it wasnt to long ago when you had to make powerpc drivers pretend to be i386 [02:07] to make the kernel happy [02:08] i just want to be clear that you are arguing against what basically open source was founded on (supporting everything) [02:09] no [02:09] im arguing it's a crap design [02:09] and that all teh filters exist in better places [02:09] opensource wasnt foudned om shitty design [02:09] it was founded on technical merit, and political ideology [02:10] not recreating the wheel 9000 times [02:10] thats true, it wasnt my idea to rewrite the filters [02:11] im not saying it shouldnt exist [02:11] i'm remembering someones quote [02:11] im saying it shouldnt exist in its current form [02:11] as its utterly useless [02:11] a call to arms [02:11] so to speak [02:11] 'theres no reason to reinvent the wheel, but if the wheel is square...' [02:11] except [02:11] our wheel is also square [02:11] were porting logn ancient filters [02:11] quite possible :) [02:12] those filters are old [02:12] what about a sphere wheel? [02:12] i like sphere wheel idea, more stability [02:12] me and a prominent avs plugin dev were giggling in teh VDD room for lavfi [02:12] when they were like "hwdn3d is so cuttign edge!" [02:12] "and gradfun! [02:12] (hes teh AUTHOR of gradfun) [02:14] Action: Daemon404 goes to grab a slice of pizza quickly [02:14] ugh [02:14] forgot to eat dinner because of trolling... [02:14] google for vaporsynth not get results for vapoursynth [02:14] Action: Compn upset [03:21] ffmpeg.git 3jamal 7tests/Makefile: tests/Makefile: fix ffprobe-test.nut with target-exec [03:21] ffmpeg.git 3Michael Niedermayer 7libavcodec/h264.c: h264: fix integer avoption types [03:21] ffmpeg.git 3Michael Niedermayer 7libavcodec/libvpxenc.c: libvpcenc: fix flags voption types [03:21] ffmpeg.git 3Michael Niedermayer 7libavcodec/mpeg4videodec.c: mpeg4videodec: fix integer avoption types [03:21] ffmpeg.git 3Michael Niedermayer 7libavformat/mov.c: mov: fix integer avoption types [06:06] ffmpeg.git 3Michael Niedermayer 7libavformat/movenc.c libavformat/movenc.h: movenc: support an alternative to edit lists to handle the first DTS != 0 case. [06:09] oh wow [06:09] michaelni, whats the functionality of use_editlist [06:10] it might be exactly what i was going to implement [06:15] btw michaelni, do we support edit lists properly with timestamp offsets when demuxing? [06:15] i.e. skip the first N samples based off the edit list [06:15] thats the other thing i was going to implement [06:34] Im getting a weird result when i add -timecode my footage end up 5 times longer than it should.. [11:25] " michaelni, whats the functionality of use_editlist" <--- its to prevent the muxer from writing edit lists as some software doesnt support it [11:25] michaelni: hey, where do you want the hash id in the string? [11:25] Daemon404, i belive the full commit message explains this [11:26] thresh, dunno, no real preferrance, where it looks good or maybe where CIA had it [11:43] ffmpeg.git 3Martin Storsj? 7libavformat/smoothstreamingenc.c: smoothstreamingenc: Check the output UrlContext before accessing it [11:43] ffmpeg.git 3Martin Storsj? 7libavformat/smoothstreamingenc.c: smoothstreamingenc: Properly return errors from ism_flush to the caller [11:43] ffmpeg.git 3Martin Storsj? 7libavformat/smoothstreamingenc.c: smoothstreamingenc: Move the output_chunk_list and write_manifest functions up [11:43] ffmpeg.git 3Martin Storsj? 7libavformat/smoothstreamingenc.c: smoothstreamingenc: Try writing a manifest when opening the muxer [11:43] ffmpeg.git 3Martin Storsj? 7libavformat/smoothstreamingenc.c: smoothstreamingenc: Ignore the return value from mkdir [11:43] ffmpeg.git 3Martin Storsj? 7libavformat/smoothstreamingenc.c: smoothstreamingenc: Add a more verbose error message [11:43] ffmpeg.git 3Anton Khirnov 7doc/RELEASE_NOTES: doc/RELEASE_NOTES: update for the 9 release. [11:43] ffmpeg.git 3Mans Rullgard 7libavutil/parseutils.c: parseutils: fix parsing of invalid alpha values [11:43] ffmpeg.git 3Justin Ruggles 7libavformat/ffmdec.c libavformat/ffmenc.c tests/ref/lavf/ffm: ffm: do not write or read the audio sample format [11:43] ffmpeg.git 3Justin Ruggles 7libavcodec/vorbisenc.c: vorbisenc: use float planar sample format [11:43] ffmpeg.git 3Justin Ruggles 7libavcodec/libmp3lame.c: libmp3lame: use planar sample formats [11:43] ffmpeg.git 3Justin Ruggles 7libavcodec/libvorbis.c: libvorbis: use planar sample format [11:43] ffmpeg.git 3Justin Ruggles 7libavcodec/adpcmenc.c: adpcm_ima_wav: simplify encoding [11:43] ffmpeg.git 3Justin Ruggles 7libavcodec/adpcmenc.c: adpcmenc: fix 3 instances of variable shadowing [11:43] ffmpeg.git 3Justin Ruggles 7libavcodec/adpcmenc.c: adpcmenc: move 'ch' variable to higher scope [11:43] ffmpeg.git 3Justin Ruggles 7libavcodec/adpcmenc.c: adpcmenc: use planar sample format for adpcm_ima_wav and adpcm_ima_qt [11:43] ffmpeg.git 3Justin Ruggles 7libavcodec/aacenc.c: aacenc: use planar sample format [11:43] ffmpeg.git 3Justin Ruggles 7libavcodec/ac3enc_fixed.c libavcodec/ac3enc_float.c libavcodec/ac3enc_template.c libavcodec/eac3enc.c: (e)ac3enc: use planar sample format [12:02] ffmpeg.git 3Michael Niedermayer s537ef8bebf8a 7libavformat/movenc.c libavformat/movenc.h: movenc: support an alternative to edit lists to handle the first DTS != 0 case. [12:03] michaelni: like that? [12:04] thresh, hmm s537ef8bebf8a doesnt work 537ef8bebf8a does [12:04] oups, indeed [12:04] otherwise it looks fine [12:05] sweet [12:12] s is obviously not a char valid in a hex hash :p [12:53] ffmpeg.git 3Michael Niedermayer 87244c8f2015 7libavformat/matroskaenc.c: matroskaenc: remove MATROSKA_ID_VIDEODISPLAYUNIT 3 [12:56] nevcairiel: yeah, it was a leftover I had when copy-pasting message template into ffmpeg.git update hook. [13:47] one unusual use case question :) would using: "ffmpeg -i input -map 0 -c copy -vcodec libx264 ..." cause all the streams to be copied except the videos, which would be re-encoded? [13:47] the point is in using "-c copy -c:v " [13:48] was it designed this way at all? [14:04] Daemon404, ubitux, JEEBsv, etc.. can you please check #ffmpeg and advise a guy accordingly [15:31] Daemon404: how does the metadata work in *synth? [15:31] should we consider some communication channel such as dbus or something? or just associate some AVDictionary with each bufferref? [15:34] saste: the logger you're talking would be some kind of logging callback? [15:34] talking about* [15:34] yes [15:34] ubitx: dbus?? [15:34] no please don't [15:34] :) [15:35] how would you configure the logging callback? [15:35] metadata injection was implemented with a simple AVDictionary in the buffer [15:35] opaque [15:35] and how would you use it to make our filters communicate between then? [15:35] them* [15:35] through the evil opaque field [15:35] mmh [15:36] opaque could be used to define a callback [15:36] communication: several flavours [15:36] metadata or commands [15:36] and a combination of them [15:36] for example your filter processes metadata, and send a command to another filter [15:36] the filter receive a command, and send metadata [15:37] the exact interaction depends on the tackled problem [15:38] i see two problems to solve: making our own filters re-use information from the previous one (like -af ebur128,volume), and make some metadata available for lavfi user (like the silencedetect filter mentioned) [15:39] mpf, afk for a while, i'll be back in a few minutes [15:45] ubitux, i already replied [15:45] re-use information: you write metadata in the buffer, the next filter reads it, process it [15:46] export metadata: you may have a metadata sink, which prints the metadata (or logs it, or sends it, or speaks it, etc.) [15:53] yep right ok [16:00] speak sink filter, yay :D [18:17] i get segv in avfilter when checking decoder for robustness [18:17] is runtime switching between channel number supported? [18:35] is it allowed for decoder to change number or channels in decode_frame? [18:39] yes, a decoder can do that [18:40] yeah quite some decoders do that, doesnt mean lavfi can deal with that, though :) [18:40] but without using get_side_data stuff? [18:40] no need for side data [18:41] than why i get segfault, codecs get 10 channels, and i get segfault when copying stuff to frame.data[] [18:42] hmm you arent changing channels in the demuxer ? the decoder can the demuxer cannot directly (would be a race) [18:42] for more then 8, you need to use frame.extended_data [18:42] yes that too [18:42] aww!! [18:43] and extended data is poorly tested so it may have bugs [19:07] just to say, I'm not sure if the guy has created a new ticket or just uploaded a sample media file (Love Thy Brother.m4v) to ffmpeg's ftp server, but if he didn't create an appropriate ticket, then this is the post related to that file: http://ffmpeg.gusari.org/viewtopic.php?f=11&t=695&p=1030 [19:08] I'm not sure should I create the ticket or ask him to create the ticket [19:43] ffmpeg.git 3Michael Niedermayer 1822aee7e6d9 7libavcodec/dsputil_template.c: dsputil_template: replace assert() by av_assert2() [19:43] ffmpeg.git 3Michael Niedermayer 979b9b1f470a 7libavcodec/h264.c: h264: switch some asserts to av_assert1/2() [20:00] ffmpeg.git 3Paul B Mahol 27a341518e91 7libavformat/avformat.h: avformat: fix typo in comment [20:58] hm, how does one flush any remaining frames out of a lavfi filter graph? right now my yadif is "losing" the last frame, because i don't flush it out, but i couldnt directly find a way how to [21:00] burek : just create the ticket for him [21:00] if he makes another we can merge/close whichever [21:00] rather have dup tix than no-tix [21:07] Daemon404: hey we have some nice filters! [21:08] :( [21:10] where? [21:10] audio? yes. [21:10] video? no. [21:10] Action: michaelni like his mandelbrot filter ;) [21:10] overlay, select, tile, life :D [21:10] ebur128 \o/ [21:11] (ass!) [21:11] [15:10] <@ubitux> overlay, select, tile, life :D <-- all in avs and vs in betetr capacities [21:11] and ass STARTED as an avs plugin [21:11] life in avs? [21:11] ebur128 might be useful [21:11] what is it? [21:11] ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16 [21:11] now what! [21:11] also mandelbrot [21:11] isnt really, like [21:11] useful. [21:12] pff :( [21:12] deshake should be improved [21:12] Daemon404: we have showspectrum/showwaves too which are kind of nice [21:12] deshake as a better version of itself [21:12] which is foss [21:12] but not gpl compatible [21:13] :/ [21:13] showinfo filters are nice too [21:13] also existed in every otehr framework [21:13] + as you said, audio. [21:13] everything you have mentioned as existed in avs (and now vs) for almost 10 years [21:14] (ah and testsrc e) [21:14] yes, audio i have no alternative to [21:14] i should have made it clear i was talkign about video [21:14] lavfi is kind of nice to deal with multiple and different inputs & outputs types [21:15] not really better than anything else [21:16] ./ffplay -f lavfi -i 'testsrc,hue=H=2*PI*t:s=sin(2*PI*t)+1' e [21:17] 21:11:21 <@Daemon404> ebur128 might be useful [21:17] 21:11:24 <@Daemon404> what is it? [21:17] Compn ok [21:17] loudness thing [21:17] right [21:17] audio [21:17] audio 2 video to be exact [21:17] ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]" [21:18] IMO the biggest problem right now is the metadata [21:18] but according to saste it should be fixable without much effort [21:18] Action: michaelni wonders if theres another filterframwork than libavfilter thst can be used allocate buffers that can be used as internal buffers for a decoders like h264 [21:19] that is direct rendering / avoiding memcpies [21:21] i dont think anythign sane would do that [21:21] thats needlessly coupled with a decoder [21:21] and is messy [21:21] also it's nto the bottleneck [21:21] filtering is [21:21] so it really is irrelevant [21:22] come to think of it, ffms2 might decode directly into avs's buffer [21:23] so you are saying its doing something messy and irrelevant ? [21:23] it's micro-optimizationt [21:23] athat im willign to be, when filtering, has very little benefit [21:23] with teh cost of coupling libarries to one another [21:25] theres no coupling though [21:26] its just using the get/release buffer callbacks of the decoder [21:26] i need to check [21:26] i think avs and vs can do this [21:26] with input plugins [21:26] it should be entirely possible [21:26] i was mistaken abotu what you meant before [21:26] apologies. [21:36] michaelni, i think i semi-agree with paul about the boatloads of packed yuv stuff being put in decoders [21:37] at the very least i think all the vXXX decoders should be in one file [21:37] its geting out of hand.. [21:37] Action: Daemon404 can volunteer to do it, if he wont get flamed [21:37] which decoders would you merge together ? [21:38] v.c [21:38] ah [21:39] Action: michaelni has no plans to flame [21:40] might want to see what carl thinks :P [21:40] i dont think hell mind [21:40] but i can ask [21:40] always good to ask authors / maintainers [21:40] were adding a huge number of packed formats to libavcodec [21:40] its pretty ugly [21:56] looks liek avs would handle direct render perfectly fine [22:30] ffmpeg.git 3Justin Ruggles d58b25aaa261 7libavcodec/adpcmenc.c: adpcmenc: ensure calls to adpcm_ima_compress_sample() are in the right order [22:30] \o/ [22:45] Daemon404, well compressed (0.1k) ---> 1007 22:07 Buitenhuis, Der (0.1K) Public Key for FATE [22:46] compressed? [22:46] there is no attachment [22:47] uh [22:47] yeah im awesome [22:47] let me resend that [22:47] is it just me, or is that something everyone screws up at least once every other week? :d [22:48] sent [22:49] Daemon404, i confirm reception, you need to wait for baptiste though, his admin powers are needed to add keys [22:50] k [23:34] ffmpeg.git 3Michael Niedermayer f9b0694cc8ff 7libavcodec/motion-test.c: motion-test: fix height parameter [23:34] ffmpeg.git 3Michael Niedermayer 2714e841bcfd 7libavcodec/x86/motion_est.c: x86/motion_est: assert->av_assert2() [23:34] ffmpeg.git 3Michael Niedermayer f2a7e1a62b6d 7libavformat/mux.c: mux: change 1 assert->av_assert1() [00:00] --- Mon Oct 8 2012 From burek021 at gmail.com Tue Oct 9 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Tue, 9 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121008 Message-ID: <20121009000502.0298E18A018E@apolo.teamnet.rs> [03:54] hi everyone, I am trying to compile ffmpeg on arm arch and getting error http://fpaste.org/ROCy/ , any idea? [04:56] Hey, how do I drop a stream during transcoding? [05:05] I've figured out it needs to be -map [05:05] But what option do yous pecify to go 'delete this' [05:06] Oh, -vn [05:06] That appears workable. [06:30] If I want my resultant file to be x264 encasulated with mkv, and I've got a video file and an audio file that aren't neccessarily the same length, is ffmpeg capable of taking the video and the audio and truncating the longer (audio or video) as soon as the other part has finished [06:30] basically audio = 30 seconds, video = 29 seconds, ffmpeg would encode audio and video together for 29 seconds but last second of audio would be lost [12:09] knoch: -shortest [12:56] relaxed: -i_dont_understand [13:01] buu or -map -0.1 [13:01] minus in front of a stream index makes what you want [15:01] knoch: sorry, that was intended for kn1000 [15:01] but he parted [15:17] what is the meaning of MPEG-TS interleaved streams? [15:17] and differen between interleaved stream and non-interleaved stream. [15:17] and difference between interleaved stream and non-interleaved stream. [15:18] hey guys [15:18] i get this error from a script when i try to make a .mov out of a .MTS [15:18] [mov @ 0x94e1ae0] Application provided invalid, non monotonically increasing dts to muxer in stream 0: 1 >= 1 [15:18] av_interleaved_write_frame(): Invalid argument [15:19] what does that mean and how can i fix it? [15:19] it worked on my old system :/ [15:30] EricAhn: actually it is quite hard to create mpeg-ts that doesn't have interleaved steams. [15:31] iive : hm I understood [15:32] non-interleaved stream is when you have audio as one big chunk and video as another big chunk [16:12] like this? http://pastie.org/5014898 ,... sorry [16:20] KING_LEE: well, they're probably going to want to see the actual args you're passing FFMPEG inside your script [16:26] thats the script [16:26] http://pastie.org/5018997 [16:27] i try to make my MTS files from a canon xa10 work in cinelerra [16:28] i had this other script but the files came out huge (176mb .mov from a 11mb MTS) and i still cant work with it in cinelerra prperly [16:44] KING_LEE: That script is awful. [16:47] I'm using FFmpeg for windows and once and a while it freezes. This time, I try to run again and it claims ffmpeg is not a valid win32 application. Has anyone else experienced this? Is it reboot time? [16:48] relaxed: i have no clue about scripts and ffmpeg, i only want to get these videos to work in cinelerra. this other script a have from http://www.g-raffa.eu/Cinelerra/HOWTO/get_media_ready.html#_how_to_convert_your_hd_original_files_to_dnxhd produces these huge files and i want to have something lighter [16:48] KING_LEE: for a in *.MTS; do ffmpeg -i "$a" -vcodec copy "${a%.*}".mov; done [16:49] KING_LEE: The problem in this case is not the script but you should change it. [16:51] KING_LEE: cinelerra can't take mpegts as input? [16:53] it does not take the MTS. i have worked with other files about a year ago on my old system, then i had these MTS files and i got stuck with it and left it alone. now on this machine i want to get it going again [16:54] hmm, rename one of the files to filename.ts and see if it will import it. [16:56] no, that makes cinelerra think its a raw PCM [16:58] the MTS will import but there is no video and the audio is just loud noise [17:03] rename it to filename.m2t [17:07] KING_LEE and plus, you are not using ffmpeg [17:07] you are using a fork of ffmpeg [17:08] burek: what do you mean? [17:08] avconv is not the same thing as ffmpeg [17:09] take a look at your output: "Copyright (c) 2000-2012 the Libav developers" [17:09] libav is a fork of ffmpeg [17:09] He is using ffmpeg but it's from libav. All you're doing is confusing him further. [17:09] they just kept the ffmpeg tool, falsely added a message that ffmpeg is deprecated and let you believe you are still using the ffmpeg [17:09] which you are not [17:10] ffmpeg is ffmpeg, no need to confuse people with "ffmpeg from libav" [17:10] such thing shouldn't even exist [17:10] relaxed: assumes raw PCM if i rename it to .m2t [17:10] and if it was the same thing, they wouldn't create avconv (as a synonim for ffmpeg tool) [17:12] KING_LEE: ok run the followin the dir container the mts files -> for a in *.MTS; do ffmpeg -i "$a" -vcodec copy -acodec pcm_s16le "${a%.*}".mov; done [17:14] burek: Users could care less about ffmpeg/libav politics. [17:15] couldn't [17:15] oh, right :) [17:16] also, containing* [17:17] never mind, I'll just stop wasting time helping people using something else but ffmpeg [17:19] relaxed: it produces the same error as my first script [17:25] KING_LEE: ugh, you can try rawvideo if you have the space. ffmpeg -i input.mts -vcodec rawvideo -acodec -acodec pcm_s16le output.mov [17:32] relaxed: that gives me another error message i cannot make sense of :/ [17:32] http://pastie.org/5019297 [17:36] maybe its my ffmpeg? [17:36] hold on [17:36] or avconv or whatever i am using ^^ [17:38] KING_LEE: try this [17:38] and check if it works better [17:43] KING_LEE: yes, a newer version my help and you can also try ffmpeg -i input.mts -vcodec ffvhuff -acodec pcm_s16le output.avi [17:44] Can somebody help me to convert MPEG Layer I files to MPEG Layer III? [17:57] relaxed: that is still quite large and has no video [18:03] this is a room I can ask questions about the API of libavformat? or is there another room for that? [18:04] duvnell2, yes you can ask here [18:04] or in the libav-user mailing list [18:06] thx.. does libavformat (or other parts of the library suite) support getting 'cue' or 'labl' tags from .wav files and the equivalent from other formats.. basically meta data within the files that goes with the audio tracks? [18:06] e.g. 'labl' chunks in .wav are points within the audio with an attached label text.. [18:09] I'm guessing those types of things would be just additional streams in the file of some type other than audio or video [18:11] Running files through mp3gain. Some complain saying that they are Layer I and not Layer III. Can somebody advise me how I convert the mp3's to layer III? [18:13] ziggyzero, re-encode with -acodec libmp3lame [18:14] if you don't have libmp3lame enabled, ffmpeg will create MPEG Layer II files (IIRC) [18:18] KING_LEE: maybe someone in #cinelerra can guide you in the right direction. [18:21] relaxed: thanks for your help, i will try my luck there again. i think i had this script from #cinelerra and they forwarded me to #ffmepeg for further details [18:22] ask them which lossless formats it supports [18:24] ok, but i need some food first [18:24] laters [20:13] google shows some unpromising result.. is there a good C++ library that wraps encoding/decoding of files? [20:13] (using ffmpeg) [20:23] duvnell2, why do you want a wrapper? [20:24] presumably to make things easier [20:24] wrapper => more complicated [20:50] hi guys. i would like to have ffmpeg decode a stream and store it as a series of jpg images. i'm using format image2 with codec mjpeg and output file pattern like out%03d.jpg. would it be possible to set up ffmpeg to loop over that specified range instead of continuously increasing the file "number"? [20:51] i need to keep the latest x images at any given time [20:51] anyone know what the AC3 (A/52) fourcc is? [20:57] ac-3 would do the trick. avi uses twocc for audio and 0x2000 may be the one. [21:00] well, I have an aiff that is referenced by an mov that is flagged as PCM (twos) when it's really AC3 [21:02] just seeing if I can fix it in file space without forcing a command-line option [21:03] hmm. It appears VLC will open the adjusted .mov but won't actually decode the audio... I wonder how AIFF stores such information (it's in an SSND chunk) [21:04] I've forced ffmpeg to decode it before, but I'd rather just make it usable as-is [21:07] can anyone tell me which guide is best to follow for installing ffmpeg on mac 10.7.5 ? [21:08] **FIXED** All thanks for the suggestions with my MP3 issue. It was that the MP3 headers were corrupt. Running them through mp3val restored :-) I think was able to analyse them with mp3gain and apply the gain adjustments. Thanks. [21:08] I see a few links out there, but all seem varied in what they do.. [21:23] @^(*@*^@( chunked filetypes [21:32] burek: oh cool. [21:37] Sashmo, !wiki [21:37] => compilation guide [21:38] i now have access to a OSX machine, so I'll eventually re-write the OSX guide [21:38] or at least i tell myself i will [21:39] thanks everyone [21:51] llogan, there are a few ancient tickets related to osx [21:51] could you check that they still stand? [21:57] saste: ill take a look once i learn how to use it. (also it's the girlfriend's so i won't get to hog it). [21:58] llogan, they're mostly building bugs, i suppose they are already fixed but we need someone to check [21:58] ah and there is the openal thing [21:59] if someone has an old cheap mac to spare, let me know :) [21:59] powerpc? [21:59] yeah that would be interesting [21:59] i might be able to acquire some really old stuff. any model in particular? [21:59] one of the few bigendian CPUs [22:00] i know a guy who has a personal Apple "museum" [22:00] i'll see waht he may have and i'll get back to you [22:00] no no model in particular, that would be for testing purposes only [22:14] saste: would shell access suffice? [22:15] llogan, that's better than nothing, but only useful for building problem [23:11] All, I have film videos @ 640x480. Should I try to scale them up to 720x480 for widescreen? Would that affect the quality? [23:11] also, I have this script i use. Does anyone notice anything glaringly wrong about this? [23:11] http://pastie.org/5020914 [23:11] i feel like the vbitrate may be a little to high [23:11] but i am a real noob and i don't even know how i got this far. : [23:11] :p* [23:27] lake: you're using mencoder. this is ffmpeg. [23:33] llogan: llogan but mencoder is an interface to ffmpeg, right? [23:36] i don't know much about it actually. i only use ffmpeg. [00:00] --- Tue Oct 9 2012 From burek021 at gmail.com Tue Oct 9 02:05:03 2012 From: burek021 at gmail.com (burek) Date: Tue, 9 Oct 2012 02:05:03 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121008 Message-ID: <20121009000503.38CAF18A01DE@apolo.teamnet.rs> [00:16] >I don't think lavfi should be remove. [00:16] liaaaarrrrrrrrr [00:17] :P [00:34] haha Compn :) [02:16] ffmpeg.git 3Michael Niedermayer adcfc0535db8 7libavformat/mxfenc.c: mxfenc: fix av_log data type for dts paramater [02:16] ffmpeg.git 3Michael Niedermayer fc6860a3ebae 7libavcodec/8svx.c: 8svx: remove unused variable [02:16] ffmpeg.git 3Michael Niedermayer 106790a4e92f 7libavcodec/ffv1.c: ffv1: fix array data types [03:18] ffmpeg.git 3Michael Niedermayer bd2613a32205 7libavcodec/rangecoder.c: rangecoder: fix "incompatible pointer type" warning [03:18] ffmpeg.git 3Michael Niedermayer c5fdd0696ab5 7libavcodec/tiff.c: tiff: fix "assignment discards qualifiers from pointer target type" warning [03:18] ffmpeg.git 3Michael Niedermayer b9a77198280f 7libavcodec/tscc.c: tscc: fix "assignment discards qualifiers from pointer target type" warning [03:18] ffmpeg.git 3Michael Niedermayer 89074e9066d1 7libavcodec/wmalosslessdec.c: wmalosslessdec: remove unused variable [05:32] ffmpeg.git 3Michael Niedermayer 43bbc3f477a2 7libavutil/xtea.c: xtea: give constants the correct type [05:32] ffmpeg.git 3Michael Niedermayer f464b02d2214 7libavformat/mpegts.c: mpegts: fuzzy crc check for not so spec compliant files [05:45] ffmpeg.git 3Pavel Koshevoy 9425dc3dba0b 7libavcodec/ppc/fmtconvert_altivec.c: Fix build failure on osx 10.5.8 ppc [09:07] ffmpeg.git 3Cl?ment BSsch f7c46d251c9a 7ffmpeg_opt.c libavformat/ffm.h libavformat/ffmenc.c: ffserver: fix seeking with ?date=... [11:04] ffmpeg.git 3Cl?ment BSsch 208a5d132282 7tests/Makefile tests/ref/fate/ffprobe_compact tests/ref/fate/ffprobe_csv tests/ref/fate/ffprobe_default tests/ref/fate/ffprobe_flat tests/ref/fate/ffprobe_ini tests/ref/fate/ffprobe_json tests/ref/fate/ffprobe_xml tests/test.ffmeta: fate/ffprobe: add some stream metadata. [11:05] isn't the bot supposed to trim the affected files? [11:06] would be nice to have an url to the git with a shortened hash too [11:13] ffmpeg.git 3Paul B Mahol d7a473926504 7Changelog configure doc/general.texi libavcodec/Makefile libavcodec/allcodecs.c libavcodec/avcodec.h libavcodec/codec_desc.c libavcodec/tak.c libavcodec/tak.h libavcodec/tak_parser.c libavcodec/takdec.c libavcodec/version.h libavformat/Makefile libavformat/allformats.c libavformat/takdec.c libavformat/version.h tests/fate/lossless-audio.mak tests/ref/fate/lossless-tak: TAK demuxer, decoder and parser [11:20] :) [11:20] j-b: libav is switching decoders/encoders to planar audio sample formats so some sunny day VLC will fail to play anything remotely useful [11:21] durandal_1707: is there any actual use of using planar? [11:22] simpler decoders, easier filtering sometimes? [11:22] j-b: each decoder do not need to interleave stuff on its own.... [11:23] ok. Still playback needs interleaving [11:23] durandal_1707: I guess we'll just reinterleave it [11:25] durandal_1707: all APIs decode_ do that [11:25] durandal_1707: all APIs decode_ do that? [11:27] j-b: ??? [11:28] durandal_1707: avcodec_decode_audio3, avcodec_decode_audio4, avcodec_decode_audio5, avcodec_decode_audio6 or whatever the new name is? [11:29] j-b: no, it is decoder specific, it it sets SAMPLE_FORMAT_U8P/S16P/S32P ... [11:30] durandal_1707: ok. So I need to block to a version before that, right? [11:32] version of what? [11:34] of libavcodec [11:35] j-b: such sample formats are introduced long before coders started to use them.... [12:08] alac decoder already does it today [13:21] ffmpeg.git 3Martin Storsj? e67b0f99520e 7libavformat/gxf.c: gxf: Include the right header for the avpriv_frame_rate_tab declaration [13:21] ffmpeg.git 3Diego Biurrun 62ae37decde7 7libavdevice/timefilter.c: timefilter: De-doxygenize normal code comments and drop silly ones [13:21] ffmpeg.git 3Justin Ruggles 5364327186fb 7libavcodec/adpcmenc.c: adpcmenc: ensure calls to adpcm_ima_compress_sample() are in the right order [13:21] ffmpeg.git 3Justin Ruggles 37f701f1c396 7libavcodec/utils.c: avcodec: allow either planar or interleaved sample format when encoding mono [13:21] ffmpeg.git 3Justin Ruggles 7b556be67353 7libavfilter/af_resample.c: af_resample: avoid conversion of identical sample formats for 1 channel [13:21] ffmpeg.git 3Michael Niedermayer 43c157f4a46c 7: Merge remote-tracking branch 'qatar/master' [13:23] s/deoxygenize/deoxydize/; s/deoxydize/reduce/ ;) [13:37] hello [13:40] ffmpeg.git 3Paul B Mahol 7d7a473926504: TAK demuxer, decoder and parser * 3http://tinyurl.com/9fqd8ex3 [13:40] ubitux: ^^ [13:42] thresh: why tinyrul? [13:42] you could make a source.ffmpeg.org url [13:43] with a trimed hash [13:43] wouldnt that be longer? [13:43] does it really matter? [13:43] well, I used what's already available in irker without modifying it [13:43] i'm not very fond of the tinyurl thing :( [13:43] isn't it configurable? [13:47] you can make your own tinyurl if you want [13:49] ubitux: looks like it isnt [13:50] :/ [13:51] let me try updating just in case [14:05] suggestions on colorscheme are also welcomed, btw [14:05] colors are fine imo [14:07] ...and last test [14:08] ffmpeg.git 3Paul B Mahol 7d7a473926504: TAK demuxer, decoder and parser * 3http://tinyurl.com/9fqd8ex3 [14:10] looks like it won't be the last [14:11] personally i prefer bit.ly for shortening, its just shorter then tinyurl :P [14:12] please send patches to irker upstream :) [14:13] i don't think it's worth shortening these urls [14:14] also, the standard short-hash form is 7 chars. ;) [16:04] michaelni: one of the main other races raised by helgrind is ff_h264_decode_mb_cabac writing a lot of stuff on the AVFrame s->current_picture.f [16:04] do you see any simple fix for this? [16:06] same in lavc/h264_pred.h:decode_mb_skip [16:06] it seems there are a lot of write around this [16:07] and since it's touching the private data, it leads to races [16:48] the 12 vs. 7 thing will be sorted soon-ish, irker upstream says [16:48] also, they'll fix the short hash generation in the commit links [16:49] so once they do that, I'll update videolan instance, and it'd be set to do just what you wanted [17:10] thresh: thank you :) [17:53] what am I doing wrong with libswscale [17:53] when after converting yuv444p to yuv420, the chroma plane changes like this: [17:53] http://rm.sudo.rm-f.org/img/uploaded/081e22f10498dbb69456b2a41eaf667a.png [17:53] saste: i grep'ed all over my irc logs and couldn't find where you linked to the metadata in bufferref :( [17:53] I fwrote() out the u plane right before and after calling into swscale... [17:54] it's the fast bilinear resampler, BTW [17:54] does the resampler even matter when not changing the size? [17:56] also, what resampler SHOULD I be using [17:56] when I convert yv12 to yuv444p, then want to screw with non subsampled chroma, then convert back to yv12? [17:56] I want the yv12 -> yuv444p -> yv12 conversion to not change chroma at all when I do no changes in between [17:57] nearest is no option, because that handles the pixels wrong that I did change [17:57] ubitux: https://github.com/kuehnelth/libav/commits/geotiff-encoder [17:57] oh i saw that link, but not the metadata thing [17:58] ubitux: https://github.com/kuehnelth/libav/commit/b8e9baf7f93c1b8e6c9e897dbb23393e25e51a93 [17:58] yeah i saw it now :) [17:58] i just looked at the url first :p [17:59] saste: btw do you mind to dig a bit about the -help full formating patch? :x [18:00] ubitux: i can already remove the "-" prefix [18:00] other formatting can go into separate commits [18:00] "lavfi: Keep frame-based metadata." "a year ago" ? has it been submitted? [18:00] saste: sure [18:01] ubitux, never [18:01] mmh. [18:01] so it will need some rebasing, but seems ok overall [18:01] no activity since a long while [18:02] i'll try to pick it and make something out of it, eventually. [18:02] i contacted the author, and he wrote that he has no time [18:02] ok [18:56] divVerent, are you downscaling something ? fast bilinear is just for upscaling [19:15] saste: any idea what kind of convention we should follow in filter's metadata for the keys? [19:15] like "." [19:15] "filter.." [19:15] "filter." [19:15] ? [19:15] ubitux: that's a good question [19:15] also how can we avoid to meddle with external metadata? [19:16] i suppose there is no way [19:16] maybe we could write a function taking a AVFilter and key [19:16] we could even think about keeping two different metadata dictionaries [19:16] but i'm not sure i like it [19:17] well maybe the AVDictionary can have a signature or something [19:17] a meta-meta tag [19:17] :D [19:17] or just keep two different dictionary for the moment [19:17] well we can just add a const char *id in the AVDictionary [19:17] *dictionaries [19:18] ah but i'm stupid it will require multiple dict [19:18] then in the entry. [19:18] multiple dictionaries could be an idea [19:18] well it's a pain [19:18] so you have metadata_dicts [19:19] but yes it could be a pain [19:19] looks overkill [19:20] i'll go for "lavfi." [19:20] X-Lavfi/foo=bar [19:20] key="X-Lavfi/foo" ? [19:20] yes, keep it simple for the moment [19:20] doesn't solve the (theoretical) namespace conflict [19:21] since you don't know what the user may write in the metadata [19:21] i'll kill the first user who will want to put a key named "lavfi.silence_start" in the metadata of his file [19:21] and anyway, it's a frame-level meta [19:21] shouldn't even reach the file metadata [19:21] don't forget filters could be inserted multiple times [19:22] you need something to identify them uniquely [19:22] and the user may want to export the created tags (in the encoded file) [19:22] so a simple solution may consist into having - for the moment - two dictionaries [19:23] iive: it's a pain to extract in the common cases if you do that [19:24] ubitux: hum? [19:24] in other words, at some point you have to decide if to put the metadata back to the file, or to drop it [19:24] also, i think it would be useful to be able to *specify* a filter instance name [19:24] iive: if you have a meta key like "filter..silence_start" if the user want to extract the value, it's harder than just av_get_dict("lavfi.silence_start") [19:25] so you don't have to rely on the somehow unpredictable auto-assigned name [19:25] mmh [19:25] something like silencedetect/first_one=..., silencedetect/second_one=... [19:26] well i think we need a simple injection first [19:26] we'll try to overthink it after :p [19:26] at that point you can tell (globally?) to the internal filters to always use a custom prefix for metadata entries, like X-Lavfi/... [19:26] which is then discarded by the buffersink [19:26] ubitux, yes get it working, get it correct, get it fast [19:27] in this order ;-) [19:27] are here any lavfi experts? [19:28] eeeh we don't have a lavfi/utils.c :( [19:29] durandal_1707, what's the problem? [19:32] saste: i get segv in lavfi when samplerate changes midstream [19:33] durandal_1707, yes you filed a ticket, right? [19:33] yes [19:33] i believed it was supported, but I may be wrong [19:34] at least at some point we had "normalization" in the buffer source [19:34] now i don't know how it is handled [19:43] ffmpeg.git 3Anton Khirnov 778071a1420b4: pixfmt: add AV_ prefixes to PIX_FMT_* * 3http://tinyurl.com/98goofl3 [19:43] ffmpeg.git 3Michael Niedermayer 7ae77266fce26: Merge commit '78071a1420b425dfb787ac739048f523007b8139' * 3http://tinyurl.com/9g7dyrc3 [20:05] ! [20:05] oh it's the typo in old_pix_fmt.h [20:05] michaelni: fate is gonna get red if you don't merge the typo fix :p [20:07] red is for rage [20:07] FATE SMASH [20:07] :D [20:17] ffmpeg.git 3Anton Khirnov 789715a3cf187: lavu: fix typo in Makefile * 3http://tinyurl.com/97fta693 [20:18] :) [20:21] Action: beastd starts fate client again [20:21] i am still waiting for baptiste to approve my key [20:21] :V [20:22] saste: any idea why avfilter_copy_buf_props() is never called with a classic ffprobe -f lavfi -i amovie=... -show_frames? [20:23] ubitux, uhm... no [20:23] ffmpeg -f lavfi -i amovie=... -f null - calls it correctly [20:23] but ffprobe never [20:23] i wonder how that's supposed to happen [21:19] ffmpeg.git 3Anton Khirnov 7716d413c1398: Replace PIX_FMT_* -> AV_PIX_FMT_*, PixelFormat -> AVPixelFormat * 3http://tinyurl.com/95kjza43 [21:19] ffmpeg.git 3Michael Niedermayer 7ac627b3d38d3: Merge commit '716d413c13981da15323c7a3821860536eefdbbb' * 3http://tinyurl.com/8mbx7en3 [21:20] why are some hashes pink? :d [21:21] i think that these that start with number are pink [21:21] wut [21:22] no [21:22] i think it's if they are part of merges [21:22] indeed, i'm wrong. [21:22] and grey? [21:22] :D [21:32] color change looks like a bug to me [22:03] ffmpeg.git 3Anton Khirnov 78728b958ff1b: lavu: fix typo in Makefile * 3http://tinyurl.com/9htqjhw3 [22:03] ffmpeg.git 3Luca Barbato 70826d8513d14: segment: drop global headers setting * 3http://tinyurl.com/99pxe3p3 [22:03] ffmpeg.git 3Luca Barbato 7175d0d94da11: doc: initial nut documentation * 3http://tinyurl.com/8duknz53 [22:03] ffmpeg.git 3Luca Barbato 791f5f8756168: doc: remove a warning from filters.texi * 3http://tinyurl.com/9jgs9aq3 [22:03] ffmpeg.git 3Luca Barbato 7d19d01bf6281: doc: support the new website layout * 3http://tinyurl.com/8vgljly3 [22:03] ffmpeg.git 3Janne Grunau 7f101eab1be1a: x86: call most of the x86 dsp init functions under if (ARCH_X86) * 3http://tinyurl.com/8f728pk3 [22:03] ffmpeg.git 3Janne Grunau 7cb36febcbc59: x86: cavs: call ff_cavsdsp_init_x86() under if (ARCH_X86) * 3http://tinyurl.com/9kj7dhx3 [22:03] ffmpeg.git 3Janne Grunau 77e522859fc46: x86: vc1: call ff_vc1dsp_init_x86() under if (ARCH_X86) * 3http://tinyurl.com/8rx3lrv3 [22:03] ffmpeg.git 3Michael Niedermayer 752dc18d414f4: Merge remote-tracking branch 'qatar/master' * 3http://tinyurl.com/9hj6ewk3 [22:06] saste: is there something obvious i missed when testing flite in ubuntu 12.04? [22:06] llogan, no i realized what the problem was [22:07] the guy told that he compiled a local static library [22:07] if you have a dynamic lib, it is not required to specify the -lasound flag [22:07] so this explains the failure [22:07] the failure should be reproducible, with a static build of libflite [22:08] i'm gonna to try that later if i find the time, and push it [22:10] google doesnt do a good job and telling me where to find libflite [22:10] the 3rd result is ffmpeg... [22:11] did you find it? http://www.speech.cs.cmu.edu/flite/download.html [22:11] yes [22:11] libflite -> flite [22:12] flite reminds me of a prank call to a friend telling him to report to work at Taco Bell at 7 am. [22:21] maybe i shouldn't have been sitting on my ass so long on that new site design... [22:24] llogan, customizable skin? [23:23] saste: i think i get it [23:24] ffmpeg is always using avfilter to get buffers; those buffer have props, which are copied to a local AVFrame etc [23:24] ffprobe is just getting a AVFrame directly from the lavfi device [23:24] i think the lavfi device should set the meta [23:26] mmh i'm missing a step. [23:57] saste: it doesn't look simple to do the bufref->metadata to avpacket to avframe :p [23:57] (in case of lavfi input device) [23:57] (which looks like the only way to make ffprobe communicated with lavfi) [23:57] :( [23:58] ubitux, what's the problem? [23:58] lavd/lavfi fills AVPackets [23:58] so at that point [23:58] how do you get the metadata from the bufferref [23:58] up to the avframe? [23:59] you need to copy that [23:59] where? [23:59] in the buffersink i suppose [23:59] i'm not sure how that will help [00:00] --- Tue Oct 9 2012 From burek021 at gmail.com Wed Oct 10 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Wed, 10 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121009 Message-ID: <20121010000501.CF4D918A0185@apolo.teamnet.rs> [01:10] can anyone tell me why when creating a mp4 from a set of jpg's everything works fine until I specify a frame rate of 5 (-r 5) at which point is only encodes a small percentage of the available frames... [01:52] hi, is it possible to get metadata from a byte range? [01:57] to get the metadata of an online og resource i meant [02:07] ? [05:54] can anyone help me out? trying to compile on mac http://pastebin.com/z7myyHCs [06:10] sashmo: perhaps wrong version of texi2html ? [06:12] Im trying again [06:16] sashmo: guess you could always --disable-doc as a last resort [06:16] good idea [06:17] what is texi2html? [06:18] documentation system [06:19] "info ffmpeg" for example [06:20] or man ffmpeg? [06:23] well, I guess "2html" is a clue. I was speaking in general about Texinfo [07:00] can anyone point me to some tutorial for using ffmpeg and a webcam on mac? [07:25] dunno, on linux you would use /dev/video0 or similar [08:31] how to i create multi-channel wav from separate tracks? [08:41] newenc: i'm not sure if ffmpeg supports it, but sox does, maybe try that out [08:44] I've been playing with sox for the past hour but I'm not getting very far [08:46] you might progress to shoes one day.. [08:50] if nothing works and you have a gui, you could try out audacity [09:24] @zap0 touche [09:25] well heres the deal [09:26] i have a bunch of prores files with 7 audio tracks. 1-6 are mono 5.1 and 7-8 are a stereo pair. [09:27] i want to take all 8ch and mix them into one 8ch track [09:28] and so far this kind of seems to work http://pastebin.com/HE5Jega7 [09:30] so far so good. [09:34] the audio is all there in only one track. [10:05] is there a deinterlacer that works directly with 4:2:2 material in ffmpeg? [10:06] yadif is craping out [10:06] aliasing [10:21] Hi... I need some small help using ffprobe [10:21] how to get XML output? [10:21] https://gist.github.com/3857325 - I tried ffprobe -print_format xml and failed horribly [10:22] divVerent: -show_format, -show_streams, etc [10:22] -v 0 is your friend too [10:22] ah, thanks [10:23] manpage a bit non-obvious, but good help here ;) [10:24] now wondering which of the formats is best to use with shell or perl scripts... ;) [10:24] flat [10:24] with sep=_ + eval [10:24] well, eval probably is bad due to evil characters possibly somewhere [10:25] it's escaped [10:25] hopefully correctly [10:25] codec_long_name=H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 [10:25] that is sure not correctly escaped ;) [10:25] ffprobe -of flat=s=_ ? [10:25] [rpolzer at nb-04 ~]$ codec_long_name=H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 [10:25] bash: /: Is a directory [10:25] oh, it doesn't have the flat option [10:25] I thought you meant default [10:26] okay, looks like I need to compile a current build then ;) [10:26] 1.0 should have it [10:26] okay, know what to do then [10:27] system upgrade, and if it's still a 0.11 one, click the nice trolly "Mark outdated" button ;) [10:27] :) [10:28] yes, they have newer ffmpeg [10:28] ah, flat format looks indeed somewhat shell compatible [10:28] in case I find issues, is the escaping "supposed" to be shell compatible? [10:28] or is that just a coincidence [10:28] it's supposed to be shell compatible [10:28] but it still uses dots by default as sep ;) [10:28] but nice, that's the smallest issue one can have [10:29] haha, I actually see a small bug already ;) [10:29] streams.stream.8.index=8 [10:29] ... [10:29] streams.stream.tags.filename="mriam.ttf" [10:29] streams.stream.tags.mimetype="application/x-truetype-font" [10:29] the .8. is missing [10:29] 10:25:46 <@ubitux> ffprobe -of flat=s=_ ? [10:29] use flat=s=_ [10:30] streams_stream_8_nb_read_packets="N/A" [10:30] streams_stream_tags_filename="mriam.ttf" [10:30] same issue [10:30] i don't have that problem with git/head [10:30] maybe it was fixed by saste [10:30] ah, ok then [10:30] right :) [10:30] streams_stream_2_disposition_attached_pic=0 [10:30] streams_stream_2_tags_creation_time="2008-05-27 18:40:35" [10:30] streams_stream_2_tags_language="eng" [10:30] streams_stream_2_tags_handler_name="Apple Alias Data Handler" [10:30] I of course could work around this from my code... but not sure if worth it [10:30] if I can just compile from git [10:32] while I am waiting for the compile... is there a way to find the number of streams from the output? [10:32] divVerent, -show_format tells the number of streams [10:32] ah, ok [10:32] just asking because there MAY be already a variable streams_stream_N_whatever in the shell ;) [10:33] yes, it's me and my weird corner cases that never happen anyway again ;) [10:33] divVerent: if you have issues with the escaping please raise them [10:33] it might even depends on the shell [10:33] here are the two functions FYI: http://git.videolan.org/?p=ffmpeg.git;a=blob;f=ffprobe.c;h=2187ce2652a9cae5d7bc33d0d546aaac7d5abd30;hb=HEAD#l870 [10:35] looks MOSTLY good to me... except for some smaller suggestions [10:35] it is probably better to use single quotes, as then only \ and ' need escaping at all [10:36] \ to '\\' and ' to '\'' [10:36] also, this \n and \r escape doesn't work, but you probably know that [10:36] shells don't interpret these in double quoted strings [10:36] but none of these cause actual breakage, and changing newlines and carriage returns to \n and \r anyway may actually be for the better [10:37] e.g. so one can grep the output for a prefix easily before using it [10:37] mmh [10:38] iirc ' escaping is a bit special with some shells :( [10:38] divVerent: if you have a failing string in mind... [10:38] for the \n i wanted to avoid multilines [10:40] right [10:41] no, I see nothing failing [10:41] I don't care for csh, though ;) [10:41] but for posix shells, this is all right except for the \n/\r, and these two are harmless [10:46] how to use -dump_attachment? [10:47] I tried -dump_attachment:0:3 filename.ttf -i filename.mkv [10:47] this does extract it, but complains at the end with "At least one output file must be specified" [10:48] i.e. the extraction works fine, but I want exit status 0 ;) [10:49] mmh [10:49] looks like a bug [10:49] hm... from the source of 0.11.2 I have lying around here, nb_output_files really has to be > 0 [10:51] -f null -frames 0 /dev/null kind of works around ;) [10:52] haha :) [10:52] divVerent: care to open an issue? [10:52] sure, if this is a bug, that is [10:53] i believe so, doesn't look correct [10:53] I see why it does this, and I see no good way to fix it... but e.g. sox has a workaround for this kind of thing [10:53] it has a -n option which means "no output" [10:53] if you e.g. just want to run analysis on a file [10:54] my goal BTW is to try to implement automatic language based stream selection, extracting fonts, coercing fontconfig, then trying to use -vf ass [10:54] from a frontend script [10:54] I am unsure about the coercing fontconfig part though ;) [10:55] I now have the stream ID -> language mapping working, and the font extraction too [10:55] hehe, srt2ass already has been written too :P [10:56] ffmpeg -i in.srt out.ass ? [10:56] haha, really? [10:56] sure. [10:56] that's even neater... [10:56] especially given I need to extract the subtitle script anyway [10:56] so I can just select the subtitle stream, and let it write to the ass file [10:56] ffmpeg -i in.mkv out.ass ? :p [10:56] exactly [10:56] apart from selecting the stream still ;) [10:57] :) [10:57] which is easy, as I already have the data in a variable [10:57] what will be harder is making the selection a bit more clever [10:57] especially, cleverer than mplayer's [10:57] which e.g. tends to lead to viewing english with english subs ;) [10:57] mplayer's selection is kind of nice afaict [10:58] an annoyance not really fixable without a scripting facility [10:58] it is [10:58] and fixing this is not really easy there [10:58] what I want is something like "try jpn/eng; if it failed, try eng/-" [10:59] but doing such things just doesn't fit into mplayer's option syntax, or in fact, any option syntax I have ever seen+ [10:59] this is more like a job for some kind of control file [10:59] you should look at honoring disposition/default as well btw :) [10:59] how is that set exactly? [11:00] libavformat/matroskadec.c: st->disposition |= AV_DISPOSITION_DEFAULT; [11:00] no, I mean [11:00] what is it used for [11:00] is this for the english subs that are only meant to replace e.g. signs [11:00] but not dialogue? [11:00] if the stream is the default or no so... [11:00] I wanna decide the default, not the author of the file ;) [11:01] if you have two audio streams and the second has default bit, then it should be selected by default :p [11:01] no ;) [11:01] the one which I understand should [11:01] but I see, if multiple matches remain, these should be preferred [11:01] so this is useful e.g. for different audio encodings of the same language [11:01] to pick the "better" one [11:01] i was talking about the case where you have no preferences obviously [11:01] (like when doing the default picking stream) [11:02] sure [11:02] I will keep that in mind when writing my script [11:02] and yeah it could influence the selection with preferences [11:02] in case of equality, like you said [11:03] there is more disposition flags though... [11:03] :) [11:03] is there a spec? ;) [11:03] avformat.h? :) [11:04] nope.avi [11:04] that just tells me their bitmask values [11:04] for some of them, that is [11:04] I am mainly interested about what "disposition/dub" is [11:05] and "original" [11:06] i can't tell :) [11:56] hello again [11:57] i had a problem with MTS files yesterday [11:57] i solved that now but i now noticed that the recoded clips i get are all shortened by a second [12:01] http://pastie.org/5023383 [12:01] this i my script [12:01] http://pastie.org/5023386 [12:03] i gave cinelerra and ffmpeg a fresh install and it worked then, i have this script from the cinelerra mailing list and added the deinterlaceing [12:11] KING_LEE, http://stackoverflow.com/questions/965053/extract-filename-and-extension-in-bash [12:13] also "-deinterlace is deprecated, use -filter:v yadif instead" [12:13] http://ffmpeg.org/ffmpeg.html#yadif [12:16] i testes i few clips and i dont really see a difference but the duration is always shortened by 1 sec [12:17] you might run your ffmpeg with -debug_ts [12:17] to see if all the frames are being processed [12:17] and that shortens the clips? [12:19] oh, its even more in that longer clip [12:20] burek: my battery runs out, can i get back to you tonight? [12:20] ping me, and if im online ill answer [12:20] thanks man [12:21] can ffmpeg generate mkv format files? [12:22] xxthink it should [12:22] ffmpeg -formats [12:22] E matroska Matroska file format [12:26] ok [12:26] I use -f mkv, [12:26] it should be -f matroska [12:26] -f matroska [12:26] yes [12:47] Hi all [12:47] bcoudurier: did you receive my email? [12:53] which of the aac encoding libraries is recommended? [12:54] if I see it right, possibly good chances have libfdk-aacenc and libvo-aacenc, but which is better? ;) [12:55] fdk [12:55] ok, thanks [12:55] vo-aacenc is barely ffaac level IIRC :< [12:55] ah, good to know [12:55] and fdk is the "fast" fraunhofer encoder [12:55] <3 [12:55] and does fdk-aac have to be current git or is 0.1.0-1 good too? [12:56] it should be in that release too, methinks (if that's the "1.0" release) [12:58] and how does libfdk-aac compare to neroAacEnc? [12:58] (asking while compiling it) [12:58] should be similar or better [12:59] divVerent recommend for what exactly [12:59] burek: listening to sound of videos with crappy headphones ;) [12:59] ed* [12:59] the only negative point of fdk is that it's not GPL-compliant atm [12:59] :< [12:59] for that you can use wav too [12:59] previously I was using the aac or libfaac codecs for that, but these are both crap [12:59] but not like nero is [12:59] and not like faac is :D [12:59] sure [12:59] yeah, faac is derp, too by now [12:59] I know I am creating an illegal ffmpeg binary [12:59] sue me ;) [12:59] nah, only illegal to distro [13:00] IIRC [13:00] sure [13:00] and only if it's GPL [13:00] is it then BTW also illegal to e.g. back up my notebook to an external HDD? ;) [13:00] lol [13:00] sure I am linking in some GPL stuff [13:00] dunno, don't care [13:00] but I consider it equally evil as the nvidia driver [13:00] I wish fdk would've not written their own license [13:00] except that it actually has SOME source [13:01] I actually wonder how the fdk's license works within the Android framework [13:01] also, anything is better than "aac" ;) [13:01] vo-aacenc is close to that [13:01] afaik [13:01] better in what? [13:01] some japanese guy tested [13:01] burek: "aac" codec in ffmpeg is REALLY broken [13:01] even does clipping [13:01] recommended/better is usually used with some criteria [13:01] it also is way below LAME for mp3... [13:01] at same bitrate [13:01] yes, LAME is actually pretty good [13:01] encoder-wise [13:01] what are your goals [13:01] and being at least as good as LAME/mp3 is something I'd really expect of an aac codec [13:02] while ffaac is well... something that semi-works [13:02] burek: no audiophile stuff though [13:02] just getting videos to work at all on the iCrap [13:02] for that it has to be aac [13:02] my point is that each of those codecs have something they are good at [13:02] yes, faac seems to be not buggy [13:02] but also below LAME level [13:02] yeah [13:02] so, it depends what exactly do you need [13:02] faac is one of the LC-AAC reference implementations [13:02] (with a small wrapper IIRC) [13:03] why is it so bad then? ;) [13:03] it should beat LAME then [13:03] if by the "level" you refer to a perceived audio quality, then try libaacplus [13:03] burek: fdk has HE-AAC as well [13:03] at the same rate you would usually use aac [13:03] so derp [13:03] JEEBsv, we already covered that topic [13:03] burek: sure, I always used 128k though [13:03] did we? [13:03] as that makes "aac" bearable and "libfaac" somewhat good [13:03] it doesn't correctly produce HE-AAC v2 [13:03] aka aac+ [13:03] is it incorrect or just uses more features? [13:04] I don't think anyone got an opinion from a person that has read the AAC spec [13:04] well, I don't know, but it's not comparable to libaacplus [13:04] if it doesn't produce aac+ [13:04] well, it does [13:04] it does have problems with QuickTime tho it seems [13:04] V: [13:04] whatever [13:04] divVerent if you like, try libaacplus at -ab 48k [13:04] and see the results [13:05] Action: JEEBsv sighs [13:05] also, compare to whatever mp3s you have [13:05] divVerent: vo-aacenc and faac are the two reference implementations lol [13:05] they're both relatively bad [13:05] to terribad [13:05] 13:04:35 JEEBsv | it does have problems with QuickTime tho it seems [13:05] ah, that means it's out ;) [13:06] because one doesn't simply install another player on the iPhone [13:06] (or rather, they all are either slow or based on Quicktime...) [13:06] dunno if the hardware decoder fails [13:06] just quicktime fails with its he-aacv2 output. he-aac seems to work [13:06] thing is, an app can only access the HW decoder through quicktime [13:06] sandbox madness... [13:07] well, I'm not sure if anyone has tested it on the actual portable stuff [13:07] and I've never seen an app that decodes video on HW and audio separately (with other formats supported) [13:07] it may be quite hard to do that on iCrapOS [13:07] the quicktime I meant was the OS X app IIRC [13:07] not even sure which version [13:07] anyways, that should be looked into [13:07] yes, but I assume Apple uses quite similar code on iOS [13:07] anything else would be madness [13:07] well, if it uses a hw decoder on the portable things [13:07] it still has to demux and verify some stuff [13:08] but yes, the actual decoder may even be by some other company [13:08] I don't think there's a HW decoder ASIC on the PC versions [13:08] yes [13:08] the iThings have been able to decode shit QT hasn't been for ages [13:08] ever since 3GS [13:08] or so [13:08] but compat problems TYPICALLY are at container level, because someone wrote moov with too many os or stuff like that ;) [13:08] or at "let me just annoy you" level, e.g. try feeding VFR video to an iPhone [13:09] lol [13:09] it displays the first frame then stays there [13:09] the PSP handled VFR well, surprisingly [13:09] and the best part: how does it even know it's VFR [13:09] at the first frame already [13:09] IIRC in mov/mp4, the time base is just the denominator [13:10] so for NTSC stuff, the numerator goes up by 1001 each frame [13:10] which looks quite close to VFR already ;) [13:10] it does have the timestamps in the index methinks, no? [13:10] yeah [13:10] the parsers have to be crazy limited to fail at VFR and so [13:10] well, if it actively wants to find out, it sure can check if all frame timestamps are divisible by 1001 in the index [13:10] or stuff like that ;) [13:10] but why should anyone do that [13:11] I think I've mostly just found integer overflows with timestamp parsing with regards to MP4/MOV [13:11] mp4 is basically an always-VFR format due to the encoding of timestamps [13:11] nah, it can be exactly CFR, unlike MKV [13:11] true :P [13:11] but it naturally /can/ have VFR in there too [13:11] MKV did lately get proper X/Y stuff too I think tho [13:11] mp4 can be "totally sane CFR with timestamps going up by one each frame" only if fps are integer [13:11] i.e. in Europe only ;) [13:12] nah, I didn't mean like that [13:12] but yes, I know that mkv issue [13:12] in mkv before this you couldn't even get 24000/1001 there exactly [13:12] which can't even represent many frame rates [13:12] at least PAL works [13:12] yeah [13:12] anyways, they added stuff to make proper X/Y rates possible [13:12] for mkv4 I think? [13:12] not finished [13:12] it's on the spec now tho [13:13] does that mean... that they didn't even fix this when rebranding a subset of mkv as WebM? [13:13] yeah [13:13] morons at google... [13:13] just to troll everyone, we should invent a container format that stores timestamps as float32 [13:14] or float16 ("half" like GPU vendors tend to call it), which is accurate for just long enough for a single rickroll [13:14] not even that, probably :P [13:15] but at least they finally fixed that now, and now they might want to do something about preroll [13:15] for intra refresh in H.264 and opus [13:16] BTW, I noticed ffmpeg got a lot better in the last year... I just managed to actually write a script to pick streams by language and render subtitles... even using attached fonts [13:16] the latter was awful XDG_ vars hacks to tell fontconfig where the fonts are, though ;) [13:17] just some TheDailyWTF worthy error messages... [13:17] fontconfig: Selected font is not the requested one: 'CronosPro-BoldDisp' != 'Cronos Pro Display' [13:17] lol [13:17] wonder how that can even happen... I am amazed that it found the font, but is that fuzzy matching? [13:17] why do we do fuzzy matching in a font system? ;) [13:18] is that an old Microsoft hack so they could remap TmsRmn into Times New Roman in Windows 3.1? ;) [13:18] and fontconfig had to perfect it [14:14] I wrote an application using libav* libraries, the goal is to stream MPEG2-TS over RTP. To accomplish that, I create an AVFormatContext to read an input file, which is in MPEG2-TS format, then I create another AVFormatContext to remux all streams from the file, and finally, I use an AVFormatContext RTP muxer. It works fine with the VLC client but not with another RTSP client. What can I do? [14:14] I basically followed this : http://libav-users.943685.n4.nabble.com/Output-mpeg-ts-to-rtp-td2234066.html [15:44] Hello, still having trouble with avconv/ffmpeg dropping frames on me if I try to specify the framerate when encoding a stream of jpg's to a mp4 [15:44] Please see the following pastie: [15:44] http://pastie.org/private/ojuply3xjyxuxxwcyiana [15:45] I have no idea why this is happening, the only change I make is adding "-r 5" to the command line [16:06] zule: if ffmpeg doesn't drop frames. the video would loose sync with the audio. [16:07] zule: are you trying to change the speed of the video? [16:07] there is no audio [16:07] can I prevent the drop somehow? [16:07] this is pretty much just a fast paced slide show [16:09] i'm not that familiar with ffmpeg... [16:09] wait... you encode jpg's? not another video? [16:10] jpg to video [16:10] no audio is present or mixed in however [16:11] try to move the -framerate before the -i [16:13] I'll be damned [16:13] :D [16:13] the position of the options matter, because it may apply to different system. for example, if you put it in the decoder section, then the demuxer (reading files) would use a default value and the codec section would try to change from one fps to another, while preserving same speed. [16:13] why would it matter, I thought I was specifying output framerate to 5, why would a jpg stream need a framerate? [16:14] regardless that does it [16:14] because it is abstracted as virtual container, where each jpg is a frame. [16:14] thank you! [16:16] i'm glad i was able to help you :) [16:25] pwd [16:57] what am I doing wrong when I get [16:57] Stream #0:1[0xa0], 0, 1/90000: Audio: pcm_s16be, 48000 Hz, stereo, 1536 kb/s [16:57] from the Input section [16:57] but later [16:57] [abuffer @ 0x1bf9400] Setting entry with key 'sample_fmt' to value '(null)' [16:57] [abuffer @ 0x1bf9400] Setting entry with key 'channel_layout' to value '0x3' [16:57] [graph 0 input from stream 0:1 @ 0x1bf9880] Invalid sample format '(null)' [16:57] paste everything at pastebin [16:57] don't let as guess the rest [16:57] how do I force audio codec on input? is it ffmpeg -acodec foo -i input -acodec bar output , with input in format foo and output in format bar? [16:58] tonsofpcs yes [16:58] thanks burek :) [16:58] I'm doing crazy now :D [16:58] https://gist.github.com/3859359 [16:58] burek: not much more info, though [16:58] ( ffmpeg -acodec ac3 -i input.aiff -acodec copy output.ac3 ) [16:59] divVerent as I can see ffmpeg has a hard time recognizing your input [16:59] ffplay plays it fine though [17:00] including audio [17:00] tonsofpcs, sorry, not -acodec foo [17:01] but -f format [17:01] the file is a container [17:01] that contains multiple acodec/vcodec/scodec streams [17:01] burek: do I maybe need special options to read directly from a DVD ISO? [17:01] container/format is what you want to set [17:02] burek: mounting the iso and playing from there is no option, because these stupid VOBs are split at every GB [17:02] divVerent, not sure, never used it :) [17:02] burek: it seems to be working... [17:02] burek: haha [17:02] "by not using ffmpeg" [17:02] is the answer there [17:02] ffmpeg -acodec ac3 -i input.aiff -acodec copy output.wav makes an output that vlc can handle... [17:03] okay, assume I have the iso mounted... how would I do it then? [17:03] I'll try ffmpeg -f ac3 -i input.aiff -acodec copy output.wav and compare them [17:03] < divVerent> burek: do I maybe need special options to read directly from a DVD ISO? // there is still no dvd input device :( [17:03] the problem is that a title consists of multiple VObs [17:03] but I want o "seams" [17:03] so mount -o loop :( [17:03] that alone doesn't help [17:03] I want the second vob follow right after the first one without any clicks, stutter or the like [17:04] just like on a real DVD player [17:04] wonder if catting the vobs and encoding from that does the right thing? [17:04] mc [17:05] it's mpeg2 so it should work [17:05] you have a concat protocol for this [17:05] damn... that means large changes to the encoding script... :P [17:05] as I rely on e.g. ffprobe on it [17:06] divVerent: working on the dvd input device is welcome :) [17:06] maybe it'll rather be dvdunauthor then... [17:06] which eats over 9000 MB temp space [17:06] oh wait, concat: is an URL scheme [17:06] that works for me ;) [17:07] burek: interesting. If I use -f ac3 , the output is a WAV header followed by an AIFF header (FORM, COMM chunk, ANNO chunk, etc.). If I use acodec, it outputs just a WAV header and data. [17:11] hello ubitux, have you read my problem ? [17:11] knoch: i have no idea about your problem [17:12] can one hardsub a DVD subtitle somehow? [17:12] yes [17:12] do you know someone who can help me ? [17:12] okay, how? [17:12] nicolas added something recently [17:12] just a sec [17:12] also, is it possible with -filter or only with -complex_filter? [17:13] is there a way to properly parse byte-reversed AC3? If I try decoding it, I get endless frame sync errors. [17:13] tonsofpcs, why are you using -f before -i at all? [17:13] why dont you let ffmpeg auto-recognize the content [17:14] burek: because the content is byte-reversed AC3 in a container that flags it as PCM. [17:14] thank you again ubitux [17:14] knoch: i can't help you :p [17:14] (well, technically the container flags it as none and the pointer to the container flags it as PCM... long story :) [17:14] divVerent: mmh i can't find it again :( [17:14] but if you know someone who can, then you helped me :D [17:15] (read: AIFF is a horrible RIFF implementation) [17:15] knoch: i don't know any [17:15] it's sad [17:15] divVerent: ah, got it [17:16] divVerent: http://git.videolan.org/?p=ffmpeg.git;a=commitdiff;h=88fc1438c693ffb7793aeb111d89775440491840 [17:16] divVerent http://blog.rot13.org/2009/10/encoding_dvd_into_hq_flash_video_using_ffmpeg.html [17:17] ubitux: okay, so this really needs a complex filter.... will do this tomorrow then :P [17:19] ubitux: I take it, assuming I use -filter_complex with implicit output [17:19] -filter_complex FOO gets a filter appended by -filter_complex "FOO [XYZZY] ; [XYZZY] BAR" [17:19] where XYZZY is an unused label [17:20] with that, it'll just be a job of string appending again ;) [20:15] hello [20:15] burek: you here? [20:20] i changed the -deinterlace to -v yadif and testet my script with a longer clip, it still shortens the duration, seems like it plays it slightly faster [20:45] http://pastie.org/5025636 [20:46] the original MTS had a duration of 8:08 and it came out 7 mins and 35 sec [20:47] http://pastie.org/5025641 [20:47] thats the script [20:47] KING_LEE, what is the ffmpeg's cmd line? [20:47] try running: bash -x [20:47] or sh .. something [20:47] for debug [20:47] ok [20:49] http://pastie.org/5025652 [20:49] like this? [20:53] just a sec [20:54] hm [20:54] Input #0, mpegts, from '00009.MTS': Duration: 00:07:35.04, start: 0.481967, bitrate: 23837 kb/s [20:54] your input is 7:35 long [20:55] but when i play it in my videoplayer it says 8:08 [20:56] which file exactly? [20:56] 00009.MTS ? [20:56] yes [20:57] try: mediainfo 00009.MTS [20:59] gotta install that first [21:01] what that mediainfo? it says that it cannot be trusted [21:01] WARNUNG: Die folgenden Pakete k?nnen nicht authentifiziert werden! [21:01] libtinyxml2.6.2 libzen0 libmediainfo0 mediainfo [21:01] ? [21:01] do you have vlc [21:01] no [21:02] eer yes [21:02] do you have any other tool [21:02] you can check that file with [21:06] ok, when i play it in vlc it says 7:35 [21:10] seems like totem is not very reliable [21:10] the mts plays wierd in vlc [21:10] and i cant close it [21:10] hmm,... vlc... [21:10] well, where did you see 8:08? [21:11] in totem [21:11] in the info box and on the timeline [21:16] how wierd, when i open it in gnome mplayer it says 11:46 on the timeline [21:18] KING_LEE, your input might be damaged [21:19] try extracting the video with various tools [21:19] to get uncompressed video [21:19] and check if all those are of the same length [21:19] you'll surely figure out that something is wrong, using one of the tools [21:20] i am not very experienced in taht [21:21] my main goal is getting that stuff to work in cinelerra [21:22] what do you mean to work in cinerella? [21:22] you cant open .MTS files in that video editor? [21:22] no [21:24] it assumes a raw pcm and there is no video and the audio is plain noise [21:25] try ffmpeg -i 00009.MTS -map 0 -c copy output.mkv [21:25] just remux it to mkv and try opening in cin. [21:25] getting MTS into cinelerra would be even better but i have no info about that. [21:28] http://pastie.org/5025843 [21:37] hm [21:37] KING_LEE, it should work if it's ffmpeg, but try this instead [21:37] cant really read ffmpeg errors [21:37] ffmpeg -i 00009.MTS -map O.0 -map 0.1 -c copy output.mkv [21:40] Hi. I'm having problems when capturing data from an rtsp stream. The output I get is http://pastebin.com/BrAB4zcx . The problem is that it takes forever in order to get that output. The command I use is "ffmpeg -i rtsp://192.168.1.90/11 output.mp4" [21:40] why don't you paste entire cmd output, not just fragments burbas [21:41] http://pastie.org/5025918 [21:41] same thing [21:42] burek: Sorry, thought it was that part that was interesting [21:42] I'll repost the entire thing [21:45] KING_LEE try just: ffmpeg -i 00009.MTS -c copy output.mkv [21:46] jep [21:47] but it does not load into cinelerra [21:47] why not [21:47] same errors? [21:48] yes [21:48] assuming raw pcm [21:49] then your audio/video stream is either broken or unsupported by cinerella [21:49] so, find another video editor [21:51] hmm,... i guess the camera bring the same output in every file but i will check a newer one [21:51] it's not mine and we have 4 of them [21:53] hmm, i got kinda used to cinelerra and i want something good. not just a simple editor only for putting clips together [21:53] btw KING_LEE, it's -map 0 (zero) [21:54] not -map O (letter 'o') [21:54] like 0-indexed input [21:54] 0.0 = input 0, stream 0 [21:54] so -map 0.1 means input zero, stream one [21:54] and -map 0 means all streams from input zero [21:54] ok? [21:56] ffmpeg -i 00009.MTS -map 0.0 -map 0.1 -c copy output.mkv [21:56] ? [21:59] or just -map 0 [21:59] if you want all streams [21:59] it's shorter to type [22:00] you mean ffmpeg -i 00009.MTS -map 0 -map 0 -c copy output.mkv [22:00] or [22:00] ffmpeg -i 00009.MTS -map 0 -c copy output.mkv [22:00] ffmpeg -i 00009.MTS -map 0 -c copy output.mkv [22:01] -map 0 = -map 0.0 -map 0.1 -map 0.2 -map 0.3 ... [22:04] still no cinelerra input [22:05] try kdenlive [22:13] i'll give it a try [22:13] but i would really like to stick with cinelerra [22:14] a reasonable script to turn the MTS to something that fits into it all i need [22:15] it's not MTS that cinellera has problem with [22:15] it's either audio or video [22:16] it doesn't support ac3 I guess [22:16] since h264 is widely supported [22:16] by most popular video editors [22:25] well, the script i have now is ok [22:26] the size of the output is a bit large sometimes but i guess i have to live with that [22:38] the 0009.MTS is 1.4 gb and the 0009.mov gets up to 9.7 gb [22:44] kdenlive looks ok but cinelerra seems to have more functions to me [22:49] is there a way to cut down some size? [00:00] --- Wed Oct 10 2012 From burek021 at gmail.com Wed Oct 10 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Wed, 10 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121009 Message-ID: <20121010000502.D4A0E18A018B@apolo.teamnet.rs> [00:00] given a ./ffprobe -f lavfi movie=... -show_frames [00:01] filter buffer -> avframe [00:01] what am i missing? [00:02] the lavfi device is a classic format filling AVPackets [00:02] so you av_read_frame() and just get a AVPacket [00:02] how can the tool (in this case ffprobe, but it's the same problem with the other) can get extra info? [00:03] i may be missing something as well :p [00:05] saste: you don't see the problem? [00:06] in the case of a lavfi device, the buffersink is only in the scope of that device/format [00:08] well anyway, the other way would be to make a sink printer like you proposed, with the probe writers :p [00:09] also, there is still side_data :-? [00:13] well yes lavfi is like a rawvideo demuxer, so can't store metadata info [00:13] right now only tiff supports that [00:13] yes side data is a possibility [00:14] what should i put into the side_data? [00:14] metadata? [00:14] badly worded sorry [00:14] Action: saste don't have a clue about side_data [00:14] how am i supposed to store it? :) [00:14] i guess i can't store a pointer [00:15] you can copy the textual data [00:15] klv? :p [00:15] ffmpeg.git 3Tim Nicholson 7a02762995bc5: movenc.c: Force correct value for "Samples per packet" for pcm audio * 3http://tinyurl.com/9h85j6l3 [00:15] but again, that's beyond my current knowledge [00:15] [len][key][len][value] ? :p [00:16] or well, just key\0value\0 should do the trick. [00:16] ok, let's try this. [00:28] Action: beastd is going to sleep [00:28] bye... [00:29] i think i'll do the same.. :) [00:29] Action: ubitux & [00:29] fg [00:33] no :( [01:35] why are some hashes pink? :d [01:35] you noticed it well :) [01:35] ffmpeg.git 3Anton Khirnov 7716d413c1398: Replac [01:35] this should have been [01:36] ffmpeg.git 3Anton Khirnov 7 716d413c1398: Replac [01:36] (if you look at the url) [01:36] the problem is the bot uses mirc color codes [01:36] ctrl+K+7 [01:36] and when another digit is added, it becomes ctrl+K+77 [01:39] just using double digits, like 07, should work [01:40] btw, who maintains cone-907 ? [02:01] ffmpeg.git 3Carl Eugen Hoyos 7b49d94e4f25e: Support decoding of targa files with 32bit palette. * 3http://tinyurl.com/8qd3zzv3 [02:03] burek: it's an irker bot I think. maybe j-b or thresh? [02:05] oh, ok, they are both here, so they can read the logs [02:59] so where are the separators in bot's commit log? separating with color doesn't work in plain text logs. [03:14] hah, got firefox 15 working on win2k [03:18] and most other msvc2010 compiled apps i hope [03:18] Action: Compn flaunts to Daemon404 [03:18] :P [03:21] only took a bunch of hacked beta winxp dlls :P [03:21] a completely unofficial service pack :D [03:36] Action: llogan sees no use for win2k [03:39] it's an operating system :] [03:44] it kinda was an os [03:44] 12 years ago [03:45] and xp 11 years ago, heh [03:47] kinda makes me feel better about win2k :P [03:48] that sounds retarded [03:49] differently abled? [03:49] xp is just as obsolete as win2k, i'm still using xp [03:50] there's not much practical consequence [03:50] software that doesn't work on win2k or xp you just wouldn't use, hasn't really happened yet [03:52] I was under the impression a lot of stuff doesn't work for 2k [03:52] basically everything should work with xp (sp3) though [03:53] about the only thing that may not work is hardware [03:53] and things that use XP, Vista, windows 7 features that define the software (surprisingly almost none of this exists) [03:54] it's always been common practice to dynamically load or detect new features, too; as part of being able to run on different versions of windows, you may get degraded functionality but stuff just failing to start is very unusual [03:54] even the way microsoft distributes runtimes comports with this usage [04:24] ffmpeg.git 3Michael Niedermayer 78da7907a4a89: adpcmenc: switch to av_assert() * 3http://tinyurl.com/8fr4kbx3 [04:24] ffmpeg.git 3Michael Niedermayer 772c2d8a34692: dsputil: convert asserts to av_asserts * 3http://tinyurl.com/8gmeafw3 [05:27] ffmpeg.git 3Michael Niedermayer 7d07940b76d03: motion_est: switch asserts to av_asserts * 3http://tinyurl.com/8kuq7su3 [05:54] ffmpeg.git 3Bobby Bingham 73d9cdfdce76a: targa: use named constants for flag values * 3http://tinyurl.com/8ms9c5y3 [05:54] ffmpeg.git 3Bobby Bingham 7c2eec3df8996: targa: support 2-way and 4-way interleaved files * 3http://tinyurl.com/952gzr83 [05:54] ffmpeg.git 3Bobby Bingham 750787fe350c2: targa: remove unused context members * 3http://tinyurl.com/9qgs3jg3 [05:54] ffmpeg.git 3Bobby Bingham 7b56f94cc363c: targa: cosmetics - add some whitespace * 3http://tinyurl.com/9777r6m3 [06:19] ffmpeg.git 3Michael Niedermayer 7e73bac484f71: configure: add support to nicely enable ftrapv * 3http://tinyurl.com/9xvwzfc3 [06:19] ffmpeg.git 3Michael Niedermayer 703f5043f5d13: eval: Fix eval test with ftrapv * 3http://tinyurl.com/8j5u6sl3 [06:19] ffmpeg.git 3Michael Niedermayer 71e83e6ad7a8b: ra144: fix code with ftrapv. * 3http://tinyurl.com/8nfcnkg3 [06:34] Who do I talk to about being able to close tickets I've fixed? [07:17] burek: it is a bug in irker, was fixed this night [09:53] thresh great :) [10:07] huh, is AVFrame->metadata unused at the moment? [10:09] it should be used by GeoTIFF [10:11] ah right, indeed [10:11] thx [10:12] also png could make use of it... [10:15] ffmpeg.git 3Carl Eugen Hoyos 761a9f099b7da: Write 32bit palette to Targa files. * 3http://tinyurl.com/8crp2et3 [10:32] hmm, why fate reports memory leak for tak that have single key in ape2 tags? [10:35] durandal_1707: -f s16le -f crc - [10:35] -f re-dup [10:35] ? [10:35] in the fate test [10:35] remove -f s16le [10:36] it will fix the leak [10:36] but other tests had it too [10:36] i don't think so [10:38] fate-lossless-monkeysaudio: CMD = md5 -i $(SAMPLES)/lossless-audio/luckynight-partial.ape -f s16le [10:39] md5 [10:39] ` crc [10:39] md5 doesn't add a -f md5 or something [10:40] crc will add a -f crc [10:40] and then, format=av_strdup("s16le"); [...] format=av_strdup("crc") [10:41] leak should not happen [10:41] see the av_freep() commented in the appropriate place [10:41] michael removed it because it was causing problems [10:42] durandal_1707: see a1bcc76e6036e78f25cbb7323c145056cfca9d93 [10:42] but it was reverted [10:43] durandal_1707: the leak should not happen, but the -f should not appear two times in the fate test anyway [10:43] you can remove the -f s16le, it doesn't affect your test [10:44] related: 6a0dfe3b9dc6b9420b25bfd089c655583cf045d6 [10:51] ffmpeg.git 3Paul B Mahol 73a2d3df0e02d: fate-lossless-tak: remove unneeded -f s16le * 3http://tinyurl.com/9rzldn63 [10:53] I've updated the irker instance on videolan git, the colors should be fixed now :) [11:02] how to profile ffmpeg? [11:02] there are no configure options now [11:02] ffmpeg.git 03Paul B Mahol 07238e904df398: DTS-HD demuxer * 03http://tinyurl.com/9xfmf4k03 [11:02] it works good now thresh :) [11:03] Action: ubitux still doesn't like tinyurl :( [11:04] this SUXX [11:04] what [11:10] What is a DTSHD demuxer ? a demuxer for .dtshd files? How is that different from .dts? samples? [11:11] j-b: you do not have samples for .dtshd ? [11:12] durandal_1707: unmuxed? no. [11:13] what? [11:13] you have files with dtshd extension or not? [11:13] xxthink: mmh, valgrind/cachegrind, START_TIMER, ... ? [11:14] callgrind* [11:14] ok [11:14] durandal_1707: no. [11:14] ubitux: I use oprofile [11:14] add the -pg options to the CFLAGS [11:15] if it makes sense, maybe add a --enable-oprofile :p [11:19] j-b: http://dotwhat.net/dtshd/10632/ [11:24] j-b: I read on internet your vlc is borken [11:27] av500: how is it? [11:27] ubitux, there could be a simple redirect, created within .htaccess on videolan that would translate all these: http://git.videolan.org/?p=ffmpeg.git;a=commit;h=61a9f099b7da38722f758db38d9c89d1b39f3a87 into something shorter like this: http://url.videolan.org/ffmpeg/61a9f099b7da [11:28] that way you'd avoid tinyurl or any other external dependency [11:28] j-b: that avengers dvd or sp [11:28] so [11:28] av500: indeed. ARCCOS DRM [11:28] burek: if you trim the hash to 7 char it's short enough [11:29] I just followed cone's output :) [11:29] but ok, even better with 7 chars only [11:37] durandal_1707: that does not answer the question. [11:37] ubitux: burek: I tried to register v.lc domain... I failed [11:38] :( [11:38] what about g.it? [11:45] j-b, v.lc is already taken, but the owner is: Nic LC Admin [11:45] so you might ask for just a DNS A record maybe [11:45] burek: nic is the registar of .lc [11:45] yes i know [11:46] I asked them :) [11:46] :) [11:46] they told me: "do a correct business proposal" [11:47] j-b: http://www.atsc.org/cms/index.php/standards/standards/296-a1032012-non-real-time-content-delivery [11:50] j-b how about videol.an [11:51] .an exists? [11:51] batman ananananaan [11:51] http://www.controloye.com/whois/an.php [11:51] j-b: yes [11:51] anyway domain whois shows it's not been taken [11:51] ok, that's because our name registar does not [11:52] video.xxx [11:52] +1 [11:52] hotvlc.xxx [11:53] probably already taken [11:53] bringing you the best of variable length coding... [11:53] Le domaine videol.an existe d?j?, et ne peut pas ?tre cr??. [11:53] according to ovh [11:53] getv.lc [11:54] j.bk [11:55] unlikely [11:55] if I cannot get v.lc with a popular software, I doubt I can get a single-letter domain for me [11:56] what is correct business proposal? [11:57] that is a good question [11:57] but > 10k probably [11:57] per year? [11:58] per acquisition, I guess [12:31] btw, ffmp.eg is free if there is an interest for such thing.. http://lookup.egregistry.eg/english.aspx [12:35] people dont remember domain names anyway [12:35] they google "yahoo" [12:35] I stopped typing domains long ago [12:35] I just mash a few matching letters into el goog [12:35] fmfge : [12:35] :) [12:44] ffmpeg's -vf scale is great [12:44] expression evaluator... [12:57] what extensions libdca have? [13:10] ffmpeg.git 03Janne Grunau 07ea14a655f7b8: avutil: skip old_pix_fmts.h since it is just a list * 03http://tinyurl.com/8qgjj4p03 [13:10] ffmpeg.git 03Mans Rullgard 0727a310e3813c: doc: allow building with old texi2html versions * 03http://tinyurl.com/8uq22tk03 [13:10] ffmpeg.git 03Martin Storsj? 0766d652cbf38e: rtpdec_vp8: Make the depacketizer implement the latest spec draft * 03http://tinyurl.com/95rahqn03 [13:10] ffmpeg.git 03Martin Storsj? 07c9b10cc4dbb6: rtpenc_vp8: Update the packetizer to the latest spec version * 03http://tinyurl.com/9fmn3hs03 [13:10] ffmpeg.git 03Mans Rullgard 078f23907f3e4c: build: add -Mdse to PGI optimisation flags * 03http://tinyurl.com/8m3jbsq03 [13:10] ffmpeg.git 03Mans Rullgard 07643933f51d1c: build: add LTO support for PGI compiler * 03http://tinyurl.com/8cf975t03 [13:10] ffmpeg.git 03Mans Rullgard 07f79364b2c30a: ppc: fix Altivec build with old compilers * 03http://tinyurl.com/9javb5w03 [13:10] ffmpeg.git 03Yusuke Nakamura 07e04826c34e9b: file: Set the return value type for lseek to int64_t. * 03http://tinyurl.com/93ycxaz03 [13:10] ffmpeg.git 03Martin Storsj? 07c136a813d77e: rtp: Support packetization/depacketization of opus * 03http://tinyurl.com/9zza5m303 [13:10] ffmpeg.git 03Diego Biurrun 07f75f4194d1ea: Restructure av_log_missing_feature message * 03http://tinyurl.com/8n8kxn303 [13:10] ffmpeg.git 03Diego Biurrun 07a75b9a180476: mingw/cygwin: Stop adding -fno-common to gcc CFLAGS * 03http://tinyurl.com/8zhwlsp03 [13:11] ffmpeg.git 03Michael Niedermayer 07ef9fe5bedd19: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/8casapo03 [14:09] FYI: Firefox 15 signals https://ffmpeg.org as insecure [16:29] FYI: Firefox 15 signals https://ffmpeg.org as insecure [16:29] https://support.google.com/chrome/bin/answer.py?hl=en&answer=95617&p=ui_security_indicator [16:29] it shows me: The site uses SSL, but Google Chrome has detected insecure content on the page. [16:30] which means there are some src="http://..." content inside html [16:30] i.e. not all content is from https://ffmpeg.org [16:31] for example this image http://americancensorship.org/images/stop-censorship-small.png is included above the text "FFmpeg supports the fight against American Internet censorship." [16:31] and that's considered an insecure content [16:40] burek: ah uh [16:55] ffmpeg.git 03Anuj Mittal 07ce19aec15b42: Check resync marker only when enabled. * 03http://tinyurl.com/8p5bgkx03 [16:55] ffmpeg.git 03Anuj Mittal 078d2e0e2c7058: Check for resync marker based on vop coding type- vop_fcode_forward and vop_fcode_backward L * 03http://tinyurl.com/9j5k3bz03 [17:18] saste: any reason you're not storing the the g_XX as bitfields? [17:19] ubitux: i just ported the original code [17:19] i didn't feel like i wanted to process the table with an hand-written program [17:20] ah you say to change uint8_t with uint8_t:1? [17:20] yes that's possible [17:20] what I'm mostly concerned is the integer implementation, which is rather tricky [17:20] *concerned with [17:21] i don't think you can do a uint8_t:1 [17:21] you'll need to change the code :p [17:21] whatever syntax it is [17:21] i don't think the syntax will allow it in this context [17:22] anyway reply by email or i'll discard it in the next iteration (which may well be after a few months) [17:22] i was just asking a random question :) [17:22] ffmpeg.git 03KO Myung-Hun 073a45688abc5c: lavc: include os2threads.h in vp8.h if HAVE_OS2THREADS is enabled * 03http://tinyurl.com/9yv3bxu03 [17:22] ffmpeg.git 03Andrey Utkin 07d2b18c8f5b08: Introduce ff_network_wait_fd_timeout() * 03http://tinyurl.com/9ncesq503 [17:22] ffmpeg.git 03Andrey Utkin 07949acefc11b4: tcp: use timeout option consistently * 03http://tinyurl.com/9ma2ojv03 [17:23] saste: anyway, please at least commit the '-' thing for the help output :( [17:40] ubitux, i wonder if it is a good idea after all... [17:40] why? [17:40] i'm talking about removing it only for the filters, right? [17:40] the output looks less structured [17:41] no you can't do that only for filters [17:41] all or nothing [17:41] mmh. [17:42] some reindentation may help [17:42] but that's kinda subjective, also that would need some changes in opt.c code [17:42] that was one of the other patch [17:42] i would personally add 2 spaces index [17:42] indent* [17:43] (and more padding for the other options btw) [17:44] what about XML output? [17:45] uhm no i never reindented [18:26] ffmpeg.git 03Andrey Utkin 072e009c6042bd: tcp: accept params through avio_open2() options * 03http://tinyurl.com/95zg2bo03 [18:26] ffmpeg.git 03Andrey Utkin 07b6f435fbc87c: http: add 'timeout' AVOption * 03http://tinyurl.com/9k4fwu703 [20:12] if someone wants to setup a ffmpeg IOC fate box, it seems there are now also binaries on the IOC page (http://embed.cs.utah.edu/ioc/) [20:33] michaelni : guess its good that ffmpeg supports clang :) [20:33] ehe [20:33] since ioc is based on clang [20:46] libav has had an ioc fate instance for ages [20:46] should be trivial to set up [20:47] on another note [20:47] Action: Daemon404 still waiting for baptiste [20:48] he answered the same day for me when i setup mine [20:48] you must've hit him at a bad time =p [20:58] one question, what does 'start' represent in input's info: Duration: 00:07:35.04, start: 0.481967, bitrate: 23837 kb/s [20:58] or this, if more convenient: http://pastie.org/5025652 [21:45] it seems that -map option is broken [21:46] http://pastie.org/5025843 [21:46] http://pastie.org/5025918 [21:51] burek, is that a -map O or -map 0 ? [21:53] btw, if the 2 l00k the same t0 y0u then y0u maybe sh0uld try an0ther f0nt [21:53] geez .. he typed letter 'O' [21:53] thanks :D [22:20] - [X264_LOG_INFO] = AV_LOG_INFO, [22:20] + [X264_LOG_INFO] = AV_LOG_VERBOSE, [22:20] hehe [22:20] ffmpeg.git 03Duncan Salerno 07bd2f8e8f79e9: Fix second use of AVOptions in HLS * 03http://tinyurl.com/9oy5s9903 [22:20] reminds me something :)) [22:27] INFO = INFO, i fail to see why everyone wants to change it =p [22:27] :) [22:31] ffmpeg.git 03Cl?ment BSsch 076ac5e3fe9d92: lavfi/silencedetect: add av_opt_free() call. * 03http://tinyurl.com/8bfsj4w03 [22:41] nevcairiel, inorite [22:41] because it offends them or something [22:41] how DARE a lib have its own output [22:42] TAG:lavfi.silence_start=3.40735 [22:42] yay. [22:43] >_> [22:43] outputting via tags... [22:43] really? [22:44] ? [22:45] lavfi will use the metadata for communication [22:45] i see [22:45] at the end of the chain ffprobe is able to dump them :) [22:45] thats seems majorly bad for anything with lots [22:45] like, i was thinking, scene change info [22:45] so you get the xml/json/etc benefits [22:45] this will be really ugly... [22:45] Daemon404: yes that's what i was planing to do [22:45] (see my mail) [22:45] very easy to do now that i did the hard part [22:57] ffmpeg.git 03Duncan Salerno 07f3f35f7430f1: crypto should allow passing of options to the underlying protocol via the url_open2 interface * 03http://tinyurl.com/9m3hngr03 [00:00] --- Wed Oct 10 2012 From burek021 at gmail.com Thu Oct 11 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Thu, 11 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121010 Message-ID: <20121011000502.5C27718A01E8@apolo.teamnet.rs> [02:53] Action: Daemon404 builds clang ioc [03:35] make: *** No rule to make target `doc/../tests/fate_config.sh.template', needed by `doc/fate.html'. Stop. [03:35] make: *** Waiting for unfinished jobs.... [03:35] ... [03:35] i know ive seen this mentioned somewhere before [04:08] man this thing needs a ton of ram [04:33] are we supposed to handle these kind of cases where some project X is based on ffmpeg and when an issue pops up in that project, they redirect users to our support: http://ffmpeg.gusari.org/viewtopic.php?f=12&t=697&p=1043 [04:33] what is the usual practice in such cases anyway? [04:34] if its our bug [04:34] they should report a sample to us and follow proper bug report guidelines [04:34] "I posted on the Serviio forum, and it was suggested that I post on the ffmpeg forum, hence why I did!" [04:34] now, his ffmpeg log is clear [04:35] oh [04:35] pebcak :P [04:35] he needs to use ps3mediaserver or something [04:35] :P [04:35] the known issue is that we simply do not support dts-hd [04:36] oh yes, durandal11707 was speaking about that recently.. [04:36] hes trying to reinvent ps3mediaserver i think? [04:36] just point him in right direction and hope he steers that way :) [04:46] oh [04:46] is servillo the thing like ps3mediaplayer ? [04:46] er ps3mediaserver [04:46] maybe i misunderstood his report [04:47] looks like he has a mkv which stops after 3 minutes unless he uses -ac 2 [04:47] maybe we need his sample to reproduce [04:50] well, yes, he said it partially works [04:50] I redirected him to the bug tracker [04:56] ffmpeg.git 03Michael Niedermayer 0725f44b041c42: x86/dsputil_mmx: change assert() to av_assert() * 03http://tinyurl.com/9xbujvh03 [07:25] <@Daemon404> make: *** No rule to make target `doc/../tests/fate_config.sh.template', needed by `doc/fate.html'. Stop. // the file was moved, 'must be a .d lying [07:25] distclean should solve the problem :p [07:42] ubitux, right. [09:29] meh [09:29] i don't remember the trick to specify protocols options... [09:36] oh my bad i think i get it [09:54] ffmpeg.git 03Paul B Mahol 07b4e516e30e70: pcx: use meaningful return values * 03http://tinyurl.com/9m2rtaz03 [10:10] j-b: does libdca handle xxch? [10:10] of course not. [10:10] libdca handles core and core [10:11] and core [10:11] did I mention core-only? [10:11] looks like libav one have more and ffmpeg one even more [10:13] yes [10:13] this is not really news [10:13] so why vlc still use libdca? [10:14] I thought FFMpeg has X96 and both have XCh, or so [10:14] durandal_1707: SPDIF [10:18] libdca fails to decode the lfe channel also [10:18] it is only silence [11:09] ffmpeg also has XBR and XXCH now [11:10] full support for HD HRA :) [11:11] nevcairiel: does it? [11:12] nevcairiel: never new that [11:12] knew [11:12] only thing missing is XLL for lossless mode (also known as HD MA) [11:13] and lbr and x96 [11:13] lbr is not really important, IMVHO [11:13] oh right it doesnt do X96 yet [11:13] lbr is only used on secondary audio on blu-rays, nice to have, but not crucial, imho [11:26] nevcairiel: weren't there 2 different ways to store some extensions in the stream, btw? [11:27] yes [11:27] you can have a core+extensions, usually called dts-es [11:27] or is DTS HD, which has a HD header, and extensions after that [11:27] but core can only have x96 or xch, really [11:28] nevcairiel: but XCH and X96 can be in dts-es, no? [11:29] yeah [11:29] so you can have 6.1 DTS-ES [11:29] or even 96Khz [11:32] nevcairiel: and FFmpeg supports XCH in both core and ext? [11:32] yes [11:33] and lately, even XXCH, which usually only appears in HD extensions [11:33] have a 7.1 DTS HD HRA file which plays fine now :) [11:34] nevcairiel: cool [12:22] this is absurd [12:23] cehoyos first flames me in one email about me not motivating people to submit tickets+samples [12:23] so that developers can investigate more on users' issues [12:23] instead of just showing them ways to workaround stuff so that it just works for them [12:23] (losing the opportunity to send user to bug trac to upload samples of media) [12:23] and when I did start to redirect them [12:24] then his attitute (again and again) drives people crazy [12:24] somebody should really tell him not to close any bug reports on the first reply [12:24] because it is really rude to cut off people like that [12:25] who have spent their time to register and create a new ticket, which takes time most of people don't have [12:25] yeah he can be quite annoying on the bug tracker [12:25] so at least that time and will should be respected and not just ignored by simply closing tickets [12:27] I mean, after all, ffmpeg team needs those reports, people dont care.. they will either fix things, pay someone to that for them or wait for sime time until the bug is fixed [12:27] so no big deal for them if they dont report bugs [12:27] but its a huge deal for ffmpeg if we loose those reports [12:40] Apparently, MPEG streams can be enclosed in MOV files [12:41] yes [12:41] codec tag " m1s".. [12:51] .long_name = NULL_IF_CONFIG_SMALL("FFM (FFserver live feed)"), [12:51] oops, sorry [13:05] ffmpeg.git 03Duncan Salerno 07a6363e3d89e7: url: Don't treat slashes in query parameters as directory separators * 03http://tinyurl.com/8k5wh7s03 [13:05] ffmpeg.git 03Duncan Salerno 07eea003814cc5: url: Handle relative urls being just a new query string * 03http://tinyurl.com/9e683wd03 [13:05] ffmpeg.git 03Duncan Salerno 0733893e6abcdc: url: Handle relative urls starting with two slashes * 03http://tinyurl.com/8oya5u703 [13:05] ffmpeg.git 03Martin Storsj? 077bc433b36dbb: fate: Add tests of the ff_make_absolute_url function * 03http://tinyurl.com/95mvl3q03 [13:05] ffmpeg.git 03Janne Grunau 073fbda309e553: avcodec: free extended_data instead address of it * 03http://tinyurl.com/9rwjtyx03 [13:05] ffmpeg.git 03Janne Grunau 078dd0650fe673: rtpdec_mpeg4: au_headers is a single array, simple av_free is enough * 03http://tinyurl.com/9h8uk4t03 [13:05] ffmpeg.git 03Diego Biurrun 07ac56ff9cc9d4: build: non-x86: Only compile mpegvideo optimizations when necessary * 03http://tinyurl.com/94dh2tb03 [13:05] ffmpeg.git 03Justin Ruggles 0723d53c547391: atrac1: use planar sample format * 03http://tinyurl.com/8wymev903 [13:05] ffmpeg.git 03Justin Ruggles 07c1a9cfd1feb2: mace: use planar sample format * 03http://tinyurl.com/99pbnrd03 [13:05] ffmpeg.git 03Justin Ruggles 077e5f04503901: pcmdec: use planar sample format for pcm_lxf * 03http://tinyurl.com/9av8zho03 [13:05] ffmpeg.git 03Justin Ruggles 07176db0b8928a: adpcmdec: use planar sample format for adpcm_ima_qt * 03http://tinyurl.com/92qfd3h03 [13:05] ffmpeg.git 03Justin Ruggles 071b9ac7290868: adpcmdec: use planar sample format for adpcm_ima_wav * 03http://tinyurl.com/9xxgskc03 [13:05] ffmpeg.git 03Justin Ruggles 074356d66d784c: adpcmdec: use planar sample format for adpcm_4xm * 03http://tinyurl.com/8osnn6503 [13:05] ffmpeg.git 03Justin Ruggles 078b854283c346: adpcmdec: use planar sample format for adpcm_ima_ws for vqa version 3 * 03http://tinyurl.com/8cxa5uk03 [13:05] ffmpeg.git 03Justin Ruggles 07d2b6ae02aa4d: adpcmdec: use planar sample format for adpcm_xa * 03http://tinyurl.com/9pa62yw03 [13:05] ffmpeg.git 03Justin Ruggles 07da9620e8e593: adpcmdec: use planar sample format for adpcm_ea_r1/r2/r3 * 03http://tinyurl.com/9ck2k2k03 [13:05] ffmpeg.git 03Justin Ruggles 07327cdb04e354: adpcmdec: use planar sample format for adpcm_ea_xas * 03http://tinyurl.com/9ptkwdl03 [13:05] ffmpeg.git 03Justin Ruggles 074ebd74cec75c: adpcmdec: use planar sample format for adpcm_thp * 03http://tinyurl.com/8wah7oj03 [13:05] ffmpeg.git 03Justin Ruggles 07cbcd497f384f: adxdec: use planar sample format * 03http://tinyurl.com/9vtmtfe03 [13:05] ffmpeg.git 03Michael Niedermayer 07eadba3e94daa: Merge commit 'cbcd497f384f0f8ef3f76f85b29b644b900d6b9f' * 03http://tinyurl.com/8d7wcdx03 [13:06] planar! [13:06] vlc borken [13:07] yep [13:12] i've said that for years. [13:13] cptspiff: unlike your software... [13:13] xbmc supports planar sample formats? [13:15] lol. take the lame joke for what it is [13:16] i still use rather crude small C snippets to interleave the audio again, i should add native support to my own audio buffers and let avresample do it :d [13:16] vlc is the best rat oriented video player there is. i just don't like the rat, so if anything, i'm broken. [13:24] ffmpeg.git 03Mans Rullgard 070fb3b24adac1: build: link test programs only against static libs * 03http://tinyurl.com/8gvj3mz03 [13:24] ffmpeg.git 03Janne Grunau 0753e122dd4afb: swfenc: error out for more than 1 audio or video stream * 03http://tinyurl.com/93f8r3k03 [13:24] ffmpeg.git 03Janne Grunau 074ffbe3f3a5d9: matroskaenc: check cue point validity before reallocation * 03http://tinyurl.com/8ejpe8t03 [13:24] ffmpeg.git 03Janne Grunau 0718ff4d20201a: avconv: simplify memory allocation in copy_chapters * 03http://tinyurl.com/9fy7q4n03 [13:24] ffmpeg.git 03Janne Grunau 071afd7a118fd7: af_channelmap: free old extended_data on reallocation * 03http://tinyurl.com/95pa4p703 [13:24] ffmpeg.git 03Janne Grunau 07714f5ab59780: vc1dec: prevent memory leak on av_realloc error * 03http://tinyurl.com/9rkqhjv03 [13:24] ffmpeg.git 03Janne Grunau 076f8ef5320f4d: vc1dec: prevent memory leak in error path * 03http://tinyurl.com/9bwefs803 [13:24] ffmpeg.git 03Janne Grunau 078501c098687b: af_amix: prevent memory leak on error path * 03http://tinyurl.com/93dgfbz03 [13:24] ffmpeg.git 03Janne Grunau 07ac9a89562adc: af_resample: unref out_buf when avresample_convert returns 0 * 03http://tinyurl.com/948q5gh03 [13:24] ffmpeg.git 03Janne Grunau 07b94e4acb4874: cmdutils_read_file: increment *size after writing the trailing \0 * 03http://tinyurl.com/9odn5ya03 [13:24] ffmpeg.git 03Michael Niedermayer 0750b5477616c8: Merge commit 'b94e4acb4874843e914fd3cb8e089aff0756bb4a' * 03http://tinyurl.com/95xz54c03 [13:26] durandal11707, have you been working something on dts-hd lately? [13:26] why? [13:26] if so, can you please take a look at this topic http://ffmpeg.gusari.org/viewtopic.php?f=12&t=697 [13:27] i havent had a DTS-HD file yet which didnt play properly [13:27] of course, in most cases it only plays the core [13:27] but it usually works just fine [13:27] .dtshd files [13:28] ffmpeg.git 03Paul B Mahol 071470c8a9a3a7: takdec: remove redundant/wrong avio_tell() * 03http://tinyurl.com/9tlq4xs03 [13:28] ffmpeg.git 03Paul B Mahol 07eb71f027f058: dtshd: remove redundant/wrong avio_tell() * 03http://tinyurl.com/8rgobs303 [13:28] that person is playing a mkv :p [13:28] that's what he says [13:36] burek: DTS-HD have limited support, OP in this case have DTS-HD-MA file which is not supported but core should still work, in other words all 6 chans should play without problems [13:39] burek: i see nowhere that OP claim that issue is in ffmpeg, so clarification is needed... [13:40] durandal11707, ok, I was just asking if that's what you were working on, if he creates the ticket, so that you know what is it related to [13:40] are there provisions for restarting parsing etc. with a different format? [13:43] burek: DTS-HD-MA is not for streaming... [13:45] durandal11707, what do you suggest (if ffmpeg can actually decode that dts-hd stream), which format should he use? aac? [13:47] burek: he is not transcoding (which works) but he wants to stream it as is [13:48] and nothing in topic points that issue is in ffmpeg [13:49] burek: and just to be sure - ffmpeg can decode dts-hd [13:49] ok, I'll try to get some more log outputs from him, to make sure what exactly is the issue [13:49] oh, I should correct my reply then [13:50] somebody did suggest that ffmpeg doesn't [13:50] [04:35] the known issue is that we simply do not support dts-hd [13:50] :) [13:50] who is gusari.org, btw? [13:50] burek [13:51] eeeh regression in webm encode! [13:51] (libvpx or libvorbis) [13:52] Action: ubitux bisects [13:53] j-b, gusari.org is just used as a domain hoster, to avoid registering new domains just for ffmpeg related stuff [13:53] burek: that just not true, there are extensions which are part of dts-hd which are supported and there are some which are not [13:54] when we (if) decide to create an official ffmpeg forum at forum.ffmpeg.org, I'll shut that subdomain down [13:54] durandal11707, I've added a reply about that, asking him for more logs [13:54] I just wanted to be sure if I understand what the actual problem was [13:56] ffmpeg.git 03Anton Khirnov 07fb722a900fc5: avconv: remove -same_quant * 03http://tinyurl.com/8lp6ozf03 [13:56] ffmpeg.git 03Michael Niedermayer 07b7ebb49d0352: Merge commit 'fb722a900fc5cc9e003b9fef25b27ed7fc5547a2' * 03http://tinyurl.com/9v6owxf03 [14:10] ffmpeg.git 03Anton Khirnov 07233a5a807e76: lavc: split asv12 encoder/decoder * 03http://tinyurl.com/8u74zxz03 [14:10] ffmpeg.git 03Janne Grunau 0779e6e8eba203: avfilter: correct memcpy size avfilter_copy_buf_props() * 03http://tinyurl.com/8mul2oa03 [14:10] ffmpeg.git 03Janne Grunau 074a7c0c455547: http: use av_strlcpy instead of strcpy() without size checks * 03http://tinyurl.com/8npupt903 [14:10] ffmpeg.git 03Janne Grunau 07f1de23faaa61: g722enc: fix size argument in memset * 03http://tinyurl.com/9n5yxzt03 [14:10] ffmpeg.git 03Janne Grunau 07d5ef9354ce01: rtspdec: use av_strlcpy for writing into fixed size buffer * 03http://tinyurl.com/9fshz9l03 [14:10] ffmpeg.git 03Diego Biurrun 076d0beefbf6ee: swscale: Do not make ff_ symbols globally visible. * 03http://tinyurl.com/8m9dqkj03 [14:10] ffmpeg.git 03Michael Niedermayer 07cd6f5c4895c4: Merge commit '6d0beefbf6ee6dbf8efb522a9307e54c6ed5f702' * 03http://tinyurl.com/8c56f3403 [14:11] michaelni: ping [14:12] michaelni: 2c34367b4a17856584b3e8b64cefa1900342ebcd broke webm encoding (output unplayable with firefox for instance) [14:15] to reproduce: ./ffmpeg -f lavfi -i 'testsrc[out0]; aevalsrc=sin(440*2*PI*t)[out1]' -t 5 -y out.webm [14:15] ubitux, does -avoid_negative_ts 1 help ? [14:15] michaelni: yes [14:16] then add it to the write header of webm (and probably mkv) (could be under if(avoid_negative_ts<0)) [14:18] ffmpeg.git 03Diego Biurrun 07096a5d76a547: fate: Refactor setting of environment variables for groups of tests * 03http://tinyurl.com/8r93uhn03 [14:18] ffmpeg.git 03Mans Rullgard 074b895cb294d6: build: sanitize linking of tools and test programs * 03http://tinyurl.com/9o2bypd03 [14:18] ffmpeg.git 03Michael Niedermayer 07c6d39fb3c5a7: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/8ug7a3b03 [14:19] michaelni: ok will try, thx [14:28] michaelni: ffplays it fine btw [14:28] -s [14:28] wow. [14:28] ffplay plays* it fine [14:30] I wonder how avealsrc=tan(440*2*PI*t) would sound, if using floating point samples ;) [14:31] is there any comprehensive list of defined ISOM/FourCC tags, apart from the one in isom.c? [14:32] for mp4 specifically? there probably is an official list somewhere [14:32] well, I was hoping for more general .mov [14:32] well ISOM is mp4, not mov :d [14:33] true enough, my bad [14:34] divVerent: ./ffplay -f lavfi -i 'aevalsrc=tan(440*2*PI*t),aformat=sample_fmts=flt' [14:34] like this [14:34] my hardware can't do this ;) [14:34] seriously, I am just calculating its fourier coefficients [14:36] apparently, its fourier series nowhere converges ;) [14:36] pettter: http://www.mp4ra.org/codecs.html [14:37] MP4_maniac: thanks, but the tag I'm looking for isn't there.. [14:37] what tag you are looking for? [14:38] " m1s" [14:38] I'm currently checking if I can share the sample [14:38] what codec is that supposed to be? [14:39] there are many unofficial fourcc in the world :P [14:39] it looks like it's not a codec per se, but a complete MPEG stream (haven't checked if it's ES/PS/TS yet) [14:39] ew, i know of those files [14:39] they should be burned [14:39] with a passion [14:39] there is also MPEG TS in asf [14:40] TS can go anywhere :) [14:40] you can get lucky if you just open it wit ha mpeg-ts demuxer [14:40] i cant find m1s in headers in xcode and quicktime [14:40] because they are meant to re-sync when the sync bytes get lost [14:41] michaelni: the tests change, i'm not sure if it makes sense: http://b.pkh.me/0001-lavf-mkv-avoid-negative-ts-by-default.patch [14:41] Action: MP4_maniac guesses " m1s" is one of unofficial fourcc [14:41] MP4_maniac: really? I'm told the file opens in both QT and WMP [14:44] it's actually a PS stream [14:47] pettter : i made a list of isom for mp4 ... [14:48] from quicktime [14:48] but it was a few years ago [14:48] pettter : does it play with mplayer -vc +ffmpeg12 yourfile ? [14:48] Action: Compn digs up list [14:49] heh, I don't actually have mplayer installed [14:49] http://wiki.multimedia.cx/index.php?title=QuickTime_container [14:49] doh [14:50] this site lists m1s as a isom [14:50] http://wiki.awkwardtv.org/wiki/QT_Codec_Information [14:50] M1S MPEG component [14:50] google for "m1s" imdc to get good results [14:50] imdc is the name of the quicktime decoder [14:52] Compn: thanks! [14:52] pettter : is it mpeg1 or mpeg2 ? [14:53] i can just add it to ffmpeg now... [14:53] if you cant share sample [14:53] altho we love having samples :) [14:53] mpeg1 video [14:53] so mpeg-1 video in mpeg-2 ps? [14:53] format_long_name=MPEG-PS format [14:54] its not mpeg video, its supposedly a mpeg program stream [14:54] you cant justh ook that up to the mpeg decoder [14:54] is it .mov or ? [14:54] it's a .mov originally yes [14:54] m1s in .mov [14:54] heh [14:55] stripping the MOV header (990 bytes) gives me a parseable MPEG-PS stream [14:55] with mpeg1video and mp2 audio [14:55] does ffplay -f mpeg1video yourfile work ? [14:55] ffmpeg -f mpeg -i file.mov? :p [14:55] ubitux, probably makes sense ... [14:56] nevcairiel: indeed [14:57] so it's a matter of recognising the tag and restarting the parsing with the proper format? [14:57] i dont think anything like that is supported at this point [14:57] ffmpeg is setup so that you have to add .mov tags to libavformat/isom.c [14:58] yeah, that much I know, but as it's a format and not a codec.. [14:58] but i'm not sure if adding it there will fix it [14:58] yeah [14:58] sounds like dv-in-avi stuff [14:58] they kind of jammed a format in another format [14:58] myeah [15:00] michaelni: ok, submitted, i'll wait a bit for more comments [15:00] thx :) [15:11] ffmpeg.git 03Paul B Mahol 0703cc52a07d51: takdec: fix seeking * 03http://tinyurl.com/8osje4k03 [15:17] ubitux, if there are no comments then the patch should be ok (/me drops it from my que of still needing a review patches) [15:17] michaelni: i keep that in mind [15:17] i'll push tomorrow or after tomorrow then [15:18] unless you want to fix quick [15:18] no hurry [15:51] Compn, nevcairiel: http://titan.codemill.se/~peteri/scarecrowdance2.mov [15:51] that's the sample [16:00] ehe, mplayer plays it fine [16:01] huh, fancy that [16:01] it does report BUG! Invalid demuxer type in new_demuxer(), big troubles ahead. [16:01] :) [16:01] :) [16:01] well, ffplay/ffmpeg doesn't [16:01] yes, its a sample that needs some fixing [16:03] and there's no Easy(tm) fix, is there? [16:14] pettter: apparently, that sample is indeed mpeg-ps. one sample is mapped to one mpeg-ps stream. MPC MPEG splitter can split it lol [16:15] MP4_maniac: lovely [16:16] MP4_maniac: should I make a ticket? [16:19] pettter: better than nothing [16:20] right-o [16:26] that mov sample has only one track http://up-cat.net/p/0b5aed3f . so the track has two separate stream (audio and video). this is similar with dv-in-avi i think [16:29] sounds reasonable [16:30] oh dear, and that is implemented by keeping a DVDemuxContext* inside the AVIContext? [16:49] ffmpeg.git 03Matthieu Bouron 07f94edfe484e9: vf_idet: remove some unnecessary statement and unused variable * 03http://tinyurl.com/9f54ows03 [16:49] ffmpeg.git 03Michael Niedermayer 07229e33a2b6de: avidec: override sample size of 1024 for VBR AAC * 03http://tinyurl.com/8dx2wr303 [17:02] michaelni: av_log(s, AV_LOG_DEBUG, "overriding sample_size\n", ast->dshow_block_align); looks wrong [17:03] how am i going to use ld for using the development version of ffmpeg? I have a foo.c file using libavsample, and after setting the correct PKG_CONFIG_PATH I get this error: /usr/bin/ld: cannot find -lavresample [17:03] I tried to remap LD_PATH, but probably I should do also $ ldconfig. Is it going to break something? [17:04] yay animated gif patch \o/ [17:04] LD_LIBRARY_PATH maybe? [17:04] is avresample built? (--enable-avresample) [17:05] not and should not [17:05] same error, note that I have not done anything with $ ldconfig. [17:06] ubitux: yay [17:06] or you could just use libswresample. [17:07] ubitux: probably not, I was looking at the trac for some issues, and found one notifying a memoryleak there. [17:09] is Vitaliy Sugrobov on IRC? [17:09] ffmpeg.git 03Michael Niedermayer 0773ad355d23ab: avidec: fix 10l typo * 03http://tinyurl.com/8n2m2ws03 [17:53] ffmpeg.git 03Paul B Mahol 07ee109c6bc2cf: dtshd: fix seeking * 03http://tinyurl.com/8msoeun03 [18:27] durandal_1707, do not push things with useless commit messages while blatantly ignoring replies. [18:36] michaelni, just fyi, ffmpeg passes clang-ioc [18:36] ill set up an ioc vm once my ram arrives this week [18:36] and once baptiste gets my key approved... [18:42] Daemon404, great, thx [18:50] ffmpeg.git 03Paul B Mahol 072c5b1a2a6abb: txd: remove duplicated header inclusion * 03http://tinyurl.com/9c4rwu303 [18:51] ffmpeg.git 03Michael Niedermayer 0792c3173c84a9: qt-faststart: fix printf argument type * 03http://tinyurl.com/8nkrugp03 [18:51] ffmpeg.git 03Michael Niedermayer 079586db6ae504: lavc: docs: the field_order is set by the demuxer and not libavcodec for decoding * 03http://tinyurl.com/8wkuq2v03 [19:14] can someone help me with some h264 raw video from logitech c920 camera [19:14] a guy wrote a tool to capture the h264 stream from the web cam [19:14] https://github.com/csete/bonecam/blob/master/capture/capture.c#L498 [19:14] im just not sure how to pipe it to ffmpeg [19:15] something like ./capture -o | ffmpeg -f ? -r ? -i - ... [19:17] burek: -o - 2>/dev/null| ffmpeg -f h264 -vcodec h264 ? [19:17] let me try [19:18] you may also need framerate, framesize and pix_fmt [19:18] http://pastebin.com/R2BnEbYC [19:19] ill update ffmpeg just to be sure [19:19] add -r and -s [19:20] i dont know where to see -r [19:20] -s would be 640x480 i guess [19:22] output a sample from ./capture (without piping) and analyze the stream [19:22] I have a sample video.raw if it will help [19:22] :) [19:26] this is with the latest ffmpeg: http://pastebin.com/QGSAZC99 [19:41] -s $framesize -pix_fmt yuv420p [19:42] you should not need -s for an annexb strea, [19:43] Option video_size not found. [19:43] it does not allow -s before -i [19:43] i do not think its feasible to pipe h264 like that [19:43] ./ffmpeg -f h264 -s 640x480 -pix_fmt yuv420p -vcodec h264 -i video.raw -c copy -f null - [19:43] it uses v4l2 [19:43] well c920 has its own internal h264 encoder (in the webcam) [19:44] the proper solution is to add h264 support to our v42l code [19:44] if it is not there yet [19:44] so im just trying to pipe/mux that with ffmpeg and thats it [19:44] yes, there was some talk about v4l2 not yet supporting h264 pix format [19:44] something like that [19:44] but i think its implemented in their latest head [19:45] I can upload a sample [19:45] just a sec [19:45] support needs to be added to out code too [19:46] http://ffmpeg.gusari.org/uploads/sample-raw-h264-video-from-logitech-c920-webcam.raw [19:46] first 10 MB of data [20:03] burek, the raw can be played with ./ffplay -f rawvideo -pixel_format yuyv422 -s 640x480 -i ... [20:04] oh great :) [20:04] thanks :) [20:12] michaelni, is that sample a rawvideo or h264 video or both :D [20:13] i would expect less size with h264 for 200 frames [20:14] it's raw.. [20:14] i forgot -f [20:14] lol [20:19] is it normal to have an aweful quality with libvpx and -quality best? [20:20] (and it's slow to encode) [20:23] it looks like something fishy: http://lucy.pkh.me/libvpx/orig.mkv ? http://lucy.pkh.me/libvpx/encoded.webm [20:24] i64? [20:24] (encoded with ffmpeg -i orig.mkv -avoid_negative_ts 1 -quality best -q:a 7 encoded.webm) [20:25] -quality best being a libvpxenc specific option [20:25] durandal_1707 ? [20:25] yes it's in a i64? [20:27] ubitux: bisect [20:27] yeah i guess so :( [20:27] ubitux: you do not set bitrate? [20:28] i don't, isn't it facultative with a "quality" setting? [20:29] dunno, maybe it is merge bug [20:32] michaelni, this was supposed to be the raw h264 output from the webcam: http://ffmpeg.gusari.org/uploads/sample-h264-video-from-logitech-c920-webcam.raw [20:32] sorry :) [20:33] when I run it like this: # ./capture -f -o | ffmpeg -f h264 -i - -f null - [20:34] it doesn't start at all, but after I press ctrl+c, it says this http://pastebin.com/JKbp1bMH [20:34] same in 1.0... same with -crf... :( [20:35] ubitux: bisect from last time it worked, do you know when it was? [20:35] i don't know if it ever worked [20:36] i'm wondering if it's not a bad usage or something [20:38] burek, -f h264 [20:39] i don't get it, the encode really is a lot slower with quality best [20:39] but the video is still awful [20:39] ubitux, -crf works fine [20:39] michaelni, it works like: ffmpeg -f h264 -i sample.raw ... [20:39] vimeo uses it [20:39] Daemon404: what setting did you use? [20:39] but it doesn't work like this: ./capture -f -o | ffmpeg -f h264 -i - -f null - [20:40] ubitux, -crf [20:40] as seen by x264 [20:40] generally 18-20 [20:40] Daemon404: yes right, but what value? [20:40] -o just redirects output to stdout instead of file [20:40] i've tried 15-18 [20:40] it's still awful [20:40] from what input [20:40] and preset [20:40] i didn't set any preset [20:40] medium then [20:40] 20:23:59 <@ubitux> it looks like something fishy: http://lucy.pkh.me/libvpx/orig.mkv ? http://lucy.pkh.me/libvpx/encoded.webm [20:40] this input [20:40] >webm [20:40] why are you using h264 with webm [20:41] ffmpeg -i orig.mkv -crf 18 -q:a 7 encoded.webm ? [20:41] thats vp8 [20:41] nto x264 [20:41] -_- [20:41] crf is x264. [20:41] where did i said it was h264? [20:41] >crf [20:41] is an x264 option [20:41] (libvpx also has a crf private option) [20:41] ubitux, well its abusing it [20:41] because avaik libvpx has no concept ofr 'crf' [20:41] of* [20:41] {"crf", "Select the quality for constant quality mode", offsetof(VP8Context, crf), AV_OPT_TYPE_INT, {.dbl = 0}, 0, 63, VE}, [20:42] whatever, i tried to play with quality and crf [20:42] and it doesn't help [20:42] thats not what crf is in x264 [20:42] constant rate factor != constant quality [20:42] anyway, i cant help with vp8 [20:43] because... lolvp8 who uses that [20:47] wholly cannoli it works :))) [20:47] thanks a lot michaelni :))) [20:51] :beer: to relaxed and Daemon404 too, of course :))) [20:51] this is awesome... cpu on my machine is 0% :) [20:51] i can now use even raspbery :) [21:47] using the bitrate seems to work [21:47] but well.. [21:48] ffmpeg.git 03Michael Niedermayer 0794f5470a20d4: lavf: add a AVPROBE_SCORE_RETRY instead of using hardcoded values. * 03http://tinyurl.com/8gbjnua03 [21:48] ffmpeg.git 03Michael Niedermayer 07b47396b6143c: img2dec: detect .raw files only with a low score as img2 * 03http://tinyurl.com/9lur9mv03 [21:48] ffmpeg.git 03Michael Niedermayer 07317505b56691: lavf: do not prematurely accept a format with low score in init_input() * 03http://tinyurl.com/8uvudqb03 [22:53] fuck. people are STILL thinking FFmpeg is dead due to that damned "This program is not developed anymore" message in ubuntu. [22:54] Action: llogan leaves to pick up ppc machine for saste to monkey with. [23:08] llogan: i proposed burek to write a wiki page about this [23:08] but he's not confortable with it [23:08] comfortable* [23:08] so feel free to write one [23:08] if they didnt care to do anything but read that line that it output [23:09] what makes you think theyd find said wiki page [23:09] also why do you care what some lazy person thinks [23:09] what are they supposed to do? [23:09] "oh ok ffmpeg is renaming the tool okay" [23:10] now you're not on #ffmpeg, and regularly ppl are coming there "hey i have a problem with avconv" etc [23:10] it would be easier to paste them the link to the wiki page [23:10] with a brief summary [23:10] instead of wasting time derping about ffmpeg, libav, debian etc [23:11] like it happens half of the time [23:12] Daemon404: i don't think you realize how much users are abused by that message :p [23:13] itll go away with 0.9 [23:13] at least [23:13] (libav) [23:13] lts and stuff will likely ship it for a while [23:14] and it's just a fucking patch to apply to drop/reword that message [23:14] they just refuse to change it... for obvious reasons. [23:14] btw, there is debian bugzilla entry, and somebody (you) should post a patch to change this message with a better one. [23:14] it's already done [23:14] i'm not a debian user [23:14] i agree the message is overly vague [23:14] siretart is really just dragging his feet on the formality that there is no patch. [23:15] is it just proposal, or real patch? [23:15] no it's not vague Daemon404, it's a pretty obvious lie :) [23:15] no it isnt [23:15] it's true [23:15] for libav. [23:15] which is what they package. [23:15] it's just misleadingly vague [23:15] yes but it's packaged under what name? "ffmpeg" [23:16] it is misleading at best and outright lie. [23:16] and the sentence doesn't say it's only true in the small libav mind scope [23:16] thats for backwards compat with debian [23:16] and it is being changed [23:16] quite soon i hear [23:16] ubitux: this too. you should ask Debian packagers to stop abusing your trademark. [23:16] ffmpeg package is beign dropped entirely. [23:16] but in that situation, it's a lie Daemon404. [23:16] also [23:16] lol @ you all [23:16] so butthurt [23:16] it's not the problem about "it's gonna change" [23:16] Daemon404: it have been 2 years. [23:16] it's been 2yr like this [23:16] and it will stay for at least 2-3 yrs [23:16] talking to siretart, he's wanted to change it for a long time [23:16] (lts & stuff) [23:17] but been held up by otehr package maintainers [23:17] yeah [23:17] lol [23:17] please. [23:17] i didn't get that impression [23:17] Daemon404: ffmpeg have been dropped from debian 2 years ago. without any discussion on the wim of a single developers. [23:17] i met and talked with him irl [23:17] so im pretty sure i got teh right impression [23:17] Daemon404: he is the sole maintainer... [23:17] ubitux, OTHER package maintainers [23:17] who use the fucking package [23:17] asa dep [23:18] why do they care about the message? [23:18] i was talkign about teh "ffmpeg" package name [23:18] which offends you so [23:18] Daemon404: this is why you don't drop the original project when a fork appears. [23:18] Daemon404: in that case, if it has to be kept as "ffmpeg" the message must be changed [23:18] i agree the message should be less vague and misleading [23:18] because it's willingly misleading [23:19] i think you should stop peppering everythign with conspiratorial words [23:19] it's been 2yr man [23:19] and you're not doing user support [23:19] .. you are nto familiar with debian i see [23:19] Daemon404: are you trying to say that we are over reacting? [23:19] yes [23:20] yes i am [23:20] this is kind of insulting. [23:20] stop being butthurt [23:20] youre reminded me of mans and diego [23:20] you don't understand Daemon404, this is a real problem [23:20] with your ffmepg tribalism [23:20] we have to do avconv support on #ffmpeg most of the time [23:20] and explain users [23:20] it's a fucking problem for 2 years [23:20] Action: llogan wouldn't say "most". [23:20] Daemon404: then tell me, why is there no ffmpeg package in debian. [23:21] it's not about butthurting [23:21] it really is [23:21] it's a real pain in practice [23:21] every time libav makes a commit oyu dont like [23:21] there's this bitchign session [23:21] that is another story [23:21] its the same thing. [23:21] all one all one all one, all one or none! [23:21] and the answer is really simple [23:21] "debian switched to libav. go to #libav." [23:22] end. of. story. [23:22] it is the wrong approach altogether. [23:22] this is not helping the users, and doesn't help us either [23:22] it is the correct approach [23:22] Daemon404: did you have a bad morning or something? [23:22] send them to teh channel of the people who maintain the fucking package [23:22] (especially when only ffmpeg is solving their problem) [23:22] its simple logic [23:22] a 2 year old coudl do it [23:22] yes [23:23] we care about our abused users :) [23:23] no, it is not users fault that the most spread distribution have switched versions. [23:23] remember what i said about tribalism [23:23] and butthurt [23:23] if they want ffmpeg, send them to a distro that has it available :P [23:23] ubitux, youre doign them a disservice. [23:23] how? [23:23] what am i doing wrong? [23:23] [17:22] <@Daemon404> send them to teh channel of the people who maintain the fucking package [23:23] i believe they are happy about the services i provide [23:23] either you do this [23:24] or you stop bitching [23:24] because its your own damn fault [23:24] Daemon404: actually it is the opposite of tribalism. We are helping their users too, because we are nice guys. And for this they may decide that they do want ffmpeg. [23:24] now it's my fault... [23:24] Daemon404: users installed "ffmpeg" [23:24] "i need help with this package" [23:24] Action: llogan now blames ubitux for the dumb libav message [23:24] i believe they wanted ffmpeg. [23:24] llogan :DD [23:24] "we dont maintain this package but let em entertain you with our fork backstory before we try and help you with a different packages" [23:25] i don't do that. [23:25] i make them try the static build first [23:25] thats altogether wrong [23:25] and.or help them with the avconv cmd line [23:25] if youre having them install or build ffmpeg [23:25] youre doing it wrong [23:26] i'm not [23:26] 23:25:29 <@ubitux> i make them try the static build first [23:26] Daemon404: if you were been sold iPhone6 and go into apple store, and instead of kicking you, people there give you a real iphone5, would you be happy? [23:26] ^ tribalism [23:26] "ffmpeg is better so its ok" [23:26] isn't it? [23:26] :) [23:27] if we're working on ffmpeg it's because we believe the project is better, same for libav... [23:27] Daemon404: the whole point is, when you buy iPhone, you expect that it is Apple product. This is why you go to apple Store, not Samsung or Nokia [23:27] yes but users dont give a fuck [23:27] exactly [23:27] so dotn waste their time tryign to convert them [23:27] send them to the proper place for THEIR package [23:27] so i'm helping the users instead of making them go in the wrong direction [23:27] aka: not to the project they didn't selected [23:27] ... you people [23:27] serious [23:27] for real? [23:28] and yes i often point out the libav support [23:28] sending them to teh support channel for the teh fuckign package theyre using is wrong? [23:28] because the distro chose for them? [23:28] if they insist on using the non-working avconv [23:28] what sort of fuckign convluted logic is this? [23:28] ppl come to #ffmpeg to get help on the ffmpeg package they installed [23:28] they ask why it's deprecated often [23:28] but its not a ffmpeg package :) [23:28] ^ [23:28] or just use avconv [23:28] nevcairiel: yes, because it's a lie [23:29] ugh [23:29] make them go complain at debian or ubuntu why their ffmpeg package installs libav [23:29] sanity has no place in libav or ffmpeg [23:29] Action: Daemon404 goes off for a whi;e [23:29] nevcairiel: that's generally in the panel of options i provide [23:30] "try this cmd line" -> "try static build" -> if it works: build/install it (!wiki), and/or complain to your distro + long explanations [23:30] this is why a wiki page with a summary of these steps would be welcome. [23:31] if it works with the avconv, well then be it, i don't want to enter in long explanations, unless they ask [23:31] Action: michaelni upgraded debian and read a message that debian is deprecated and not maintained anymore and that i should use gentoo instead, after asking on the debian lists iam beint told that my package is now maintained by gentoo and i should ask for support there aha [23:31] :D [23:31] this may be a good 1'st april joke [23:32] oh again a deprecated message user who doesn't understand on the ml [23:32] Daemon404: i don't think you understand how much pain it is to deal with this [23:32] use Windows :) [23:33] ubitux: i think he understands, and his position is that we shouldn't deal with it at all. [23:34] back to the message...the "original" said: "This program is not developed anymore and is only provided for compatibility. Use avconv instead (see Changelog for the list of incompatible changes)" [23:34] "5B G5;>25:, =5B ?@>1;5<0" as Stalin likes to say. [23:34] the new says: "*** THIS PROGRAM IS DEPRECATED ***\n "This program is only provided for compatibility and will be removed in a future release. Please use avconv instead." [23:34] haha [23:34] much better!... [23:34] caps! [23:34] anyone know which ubuntu versions, if any, still use the old one? [23:35] indeed. much better. at least now it means that this program may be still developed. [23:35] the message should be "This program is deprecated in the Libav project, an alternative to the FFmpeg project" [23:36] i tried to get some wording like that, but it was not accepted. i'll try again. [23:37] honestly, we should first force debian to stop package libav as ffmpeg [23:37] i can understand they have some technical issues with the name [23:37] they are quite sensitive to IP violations and we can make a huge fuss about it. [23:37] this is not our problem. [23:37] iive: i don't know if that will cause dependent programs to hate us. [23:37] but refusing to change the message is a blatant hostile move [23:38] the easiest solution would be to package ffmpeg. [23:38] i mean, the real ffmpeg. [23:38] ffmpeg is deprecated iive ! [23:38] trademarks exists to prevent THIS exact situation. [23:39] that is, somebody smearing your good name. [23:39] you should stop wasting your time on libav.. just find a decent lawyer and sue them for copyright infringement and take some money from that [23:39] i'm too poor to need money. [23:39] they don't care about changing that message, since they came up with the same, for a good reason [23:40] what copyright infringement? [23:40] it's easier in the debian policy to just add a patch changing the message [23:40] but since they are a fork, they are not allowed (legally) to use your brandings, names, copyrighted stuff and trade marks (like FFmpeg or ffmpeg) [23:40] "This program is deprecated in the Libav project, an alternative to the FFmpeg project" really sounds perfect for that... [23:40] we should just ask debian to stop packaging ffmpeg [23:40] especially not in a context "Copyright by Libav developers team" [23:40] libav doesnt even have a ffmpeg anymore, debian is either shipping a rather old version or they re-added the app for compat reasons :p [23:40] :P [23:41] nevcairiel : they cant just axe ffmpeg binary, too many programs _require_ it [23:41] nevcairiel: can you say that for stable/unstable/testing and the various ubuntu releases? [23:41] the point is libav shouldn't (and mustn't) use the name ffmpeg anywhere in their fork [23:41] they can always make a symlink named ffmpeg. [23:41] ok. my next step is to see what ubuntu versions use the old really crappy message, if any, and re-visit the bug report to get it re-worded. [23:41] no matter if it breaks the whole damn debian [23:42] burek : mail fabrice [23:42] looks like it was released after the last release, libavs last major release is quite old already [23:42] the problem doesn't lie in libav anymore burek [23:42] er [23:42] removed after the last release [23:42] burek: they don't have ffmpeg for a long time now [23:42] it's just in the debian packaging [23:42] for retro compat issues and stuff [23:42] that is EXACTLY why we are pushing to fix that message [23:42] Compn, you guys are way longer in this whole business to suggest me to send an e-mail to Fabrice :) [23:43] i believe Fabrice is aware and don't want to mess with that [23:43] well, then why are we even discussing it anymore :) [23:43] if he doesn't care, why should we? [23:43] because we do user support [23:43] no one wants to mess with it [23:43] :P [23:44] arh i'm tired :( [23:44] I would mess with it a lot, if it was my project [23:44] so email fabrice then [23:44] dont matter if you are new or old [23:44] those guys showed who they are by the way they were acting [23:44] fabrice is nice guy [23:44] burek: remember, libav can't do much now to change that [23:44] i dont think i've ever actually talked to him... [23:44] well, it's a kinda insult that a newbee cares more about the project, then its developers who are far longer in all this.. [23:44] burek: the problem is now totally on debian-like distro side [23:45] so, I'm gonna skip sending that email [23:45] burek: and that mean the packager(s?) are just refusing to fix the message [23:45] the debian package maintainer for that package is a libav dev, so there is that [23:45] Fabrice is aware [23:45] is he really a "dev"? [23:45] well some sort of dev anyway [23:45] you should've sue debian community for changing the official ffmpeg source code with their lie [23:45] but anyway, there is nothing to change on libav side [23:45] except the maintainer(s) [23:46] i'll revisit the "message" issue, so hopefully i can resolve that, and then i'll try to get ffmpeg package to actually be FFmpeg in Ubuntu+. or at least see what they have to say. [23:46] fabrice holds the ffmpeg.org domain too, so the restoration of the project under this name happened with his blessing. [23:46] llogan: would you mind proposing my message so they just have to do a copy/paste? :) [23:46] yes. a copy/paste lazyproof is in mind. [23:46] He is not active ffmpeg developer at the moment, so there is no reason for him to engage into legal conflicts on his own. [23:48] but there is that thing with the trademarks that if you don't protect them, you can loose them. [23:48] So, writing a page with explanation, then publishing a news on the homepage would be a good start. [23:49] burek : also is good faith to give fork team some time to migrate [23:49] Compn, they gave you a fair fight for you to give them anything? :) [23:49] burek : thats the point, we are nice people [23:49] I would just ddos them for a month to return the favor of taking over the servers like a bunch of kids [23:50] 1 year should be enough for such elite coders like them. [23:51] burek: ddos-ing service provider is quite lame and may not be entirely legal. We should not drop to their level. [23:51] anyway, I gave up on that topic, because I'm too fresh and newbee to even discuss those things.. you guys do whatever you feel like it's right [23:51] ok, i should actually do some meatspace work today... [23:51] iive I agree, I'm just comparing things [23:52] we have given them enough time to sort out the problems. [23:52] I want to know if everybody agree that we cannot tolerate the situation in its current state. [23:53] Action: ubitux just wants to fix the message [23:54] there is not a single technical problem to do that [23:54] (in comparison to renaming the package, or trying to distribute ffmpeg in addition to ffmpeg or other solutions) [23:54] there is one, it should be done by the person who wrote the original message. [23:55] mmh that message comes from libav come, and in the libav scope it's "true" [23:55] and it got droppped from their repo so who care [23:55] again, the problem is in the packaging policy [23:56] anyway, unrelated: no one to comment/review the metadata patchset? [00:00] --- Thu Oct 11 2012 From burek021 at gmail.com Thu Oct 11 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Thu, 11 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121010 Message-ID: <20121011000501.5381618A01DE@apolo.teamnet.rs> [00:49] I'm having a big problem with ffmpeg/libavformat not saving (or maybe corrupting?) options I'm setting in the muxer.... the problem is described and shown here: http://dpaste.com/811746/ [00:50] like the metadata I set, as well as the bitrate, time base etc. are all removed/corrupted on the final output file, but dumping the context itself all shows the correct infop [00:50] info* [00:50] and I have no idea why this is happening :( [00:50] I'm using libavformat directly in my own application [00:50] and latest version (1.0) of ffmpeg [00:51] on x64 arch linux [00:53] it's missing your cmd line [00:53] huh? [00:53] for the encoded_by, -map_metadata is your friend [00:53] I'm not using the command line program [00:53] I'm using the libavformat library itself [00:53] oh, ok sorry [00:53] in my own application [00:53] then it's better with the code :) [00:54] well yes... but I was hoping there was something I obviously overlooked or that someone might know to check something, before having to create an entire (smaller) test program that reproduces this, since I can't post the full code [00:55] at least paste the ffmpeg related code :p [00:55] it's hard to guess what you are doing [00:56] ok [00:56] it's a bit late for me to start reading some other code though so... [00:56] gl ;) [00:56] Action: ubitux & [00:56] (hint: you have a dedicated ml for api usagee) [00:56] (might get more help there) [01:03] hi guys, I am using silencedetect filter and it works well thus far. [01:03] It generates outout like this: https://gist.github.com/3862017 [01:04] I am trying to figure out the best way to remove those silent parts from my input. [01:04] I want to automate it since I have about a hundred mpeg files I need to use it on. [01:05] before i go writing a ruby or bash script, i figured i would ask if anyone had advice [01:12] lake: i just sent a patch to simplify this [01:12] but for your needs right now, just split the string :p [01:13] lake: at some point, you will be able to exploit results like in the ffprobe example at the end: http://ffmpeg.org/pipermail/ffmpeg-devel/2012-October/132180.html [01:13] ubitux: can you please link me to your patch. i would be more than happy to try it out and comment if it helps [01:13] :) [01:13] ubitux: http://pastebin.com/F073SBxv [01:13] there's the code [01:14] ubitux: btw, thanks so much for your email months ago. you were quick to respond to me! thanks! [01:14] what mail? :p [01:15] i send quite a bunch of mails everyday i'm sorry i don't remember :p [01:15] bparker: did you base your code on the files in doc/examples? [01:16] yea [01:16] ubitux: i don't expect you to. i'm just excited to see you here. Re: [ffmpeg] Silencedetect filter [01:16] from July [01:16] lol [01:17] i don't remember :( [01:18] no worries, really, i was just asking for more information and you helped. thanks. [01:19] okay :) [01:19] bparker: ok so one issue at a time... [01:19] are you able to keep the encoded_by with the ffmpeg cmd line? [01:20] if so, check how its done in ffmpeg.c/ffmpeg_opt.c [01:21] I tried this: ffmpeg -i ~/121009-1AAA.mp4 -metadata encoded_by=test -vcodec copy test.mp4 [01:21] about the encoding settings, same thing, compare with ffmpeg tool, and try to play with libx264 AVOptions maybe (x264opts for instance) [01:21] but test.mp4 does not have encoded_by tag [01:21] I've been playing with them for days on end now >< [01:22] really out of ideas as to what to try [01:22] there might be a bug with the metadata then [01:22] or just unsupported feature [01:22] like the mp4 muxer just ignoring them. [01:23] anyway, it's 1:22 o clock here, and i have to wake up in 5?h [01:23] so i'm leaving for real now :) [01:23] mov ignores it also, but then I tried flv and it worked. [01:23] sigh [01:23] ok [01:24] well thanks [04:43] http://www.reddit.com/r/AskReddit/comments/117q87/c/c6k1tou [08:05] Hi, what can be wrong if ./configure for ffmpeg says libv4l2 not found ? [08:14] lfiebach: because you added --enable-libv4l2 switch without having the libv4l2 library available? [08:14] note: this library is a wrapper, you don't actually need it most of the time [08:15] it just helps with some particular devices [08:19] ok, what do i need for webcam support ? Only indevs ? [08:20] --enable-libv4l is enabled but sysroot is different so i think ffmpef did not find it [08:28] ubitux: [08:47] lfiebach: yes indevs should be enough [08:47] grep V4L2 config.h to confirm it is enabled [08:47] you should have: #define CONFIG_V4L2_INDEV 1 [08:48] hello [08:49] I'm trying to loop a video for a precise amount of time [08:49] (I mean, not x times, just something like precisly 1h) [08:49] I'd like to cut out a part out of the end of the original video as well. [08:49] I've found -loop but it doesn't exists [08:49] I'm using ffmpeg version 0.8.3-6:0.8.3-7 [08:50] you sure it's ffmpeg? [08:50] we are in ffmpeg 1.0 [08:50] that's the latest version available in my repos [08:50] it's sid [08:51] that's not ffmpeg then, it's a fork, but well. [08:51] I should comp ffmpeg1.0 and come back then ? [08:51] let's try to solve your problem instead :) [08:51] so anyway, what is this loop thing, i don't understand [08:52] you want to store loop info in your output file? [08:52] I have a video which is a musical thing [08:52] it's very short, something like 30 sec [08:52] and I'd like to loop it for a certain amount of time, like 1 hour [08:52] oh ok. [08:53] mmh let me think.. [08:54] i'm not sure that's supported but i'm gonna test something [08:54] give me a while [08:54] sure, thanks for your help [08:59] mmh i'm kind of able to do it with a loop count, looking for doing it with a duration now [09:00] ubitux: don't give yourself a headache huh [09:01] the original has a fixed time [09:01] I can fussy the loop count to get the right duratio [09:01] duration* [09:01] s/fussy/guess/ [09:02] mmh it doesn't work :( [09:03] a cmd line like ffmpeg -f lavfi -i movie=loopme.mkv:loop=0 -t 60 -y out.mkv was supposed to work, but it doesn't unfortunately [09:04] i'm opening a bug [09:10] well at least I made you discover a bug [09:11] the ticket if you want to watch it: https://ffmpeg.org/trac/ffmpeg/ticket/1799 [09:11] now i don't see much solution [09:11] do you have the line to loop it x times ? [09:11] except concatening manually several times with the concat filter or something [09:12] ashka: i thought it was working, but it isn't :p [09:12] even for a non-given duration [09:12] ?* [09:12] yes [09:12] it's the same issue [09:12] oh well. [09:12] thanks for your help [09:12] sorry :) [09:13] i'm trying to find another trick. [09:14] maybe with a pipe.. [09:14] ashka: what's the input format? [09:15] WebM [09:15] but I could convert it [09:16] ashka, how about playback side? mplayer can loop [09:16] I'm open to any solution [09:16] yet, the machine has no X server [09:17] uhmm, explain where the playback will happen, elaborate, ill be back in 10, smoke time .... [09:18] I want to get the new video into a file [09:18] ashka: i'm going to try to add the feature somehow, do you have some time? :) [09:18] looped video* actually [09:18] ubitux: if you've got motivation, well yes [09:18] I've got pleny [09:18] plenty* [09:19] perfect, then give me something like 1 hour to see if that's possible easily [09:19] I have videos recorded at 640x480. When I use dvdauthor it complains about "unknown mpeg2 aspect ratio" [09:20] my capture script looks like this: https://gist.github.com/3863702 [09:21] ashka: while i'm trying to patch it, you should start cloning the repository and build ffmpeg [09:21] so you can apply the patch later [09:21] I have just cloned it [09:21] thanks for your time [09:32] again one of those "why doesn't this work" questions... https://gist.github.com/3863736 [09:32] reordering the filter_complex commands doesn't fix it [09:32] i don't think -filter_complex options can be stacked [09:33] they can't? damn [09:33] oddly, I get the same error when I reverse the order of the options [09:33] so it apparently does store them both [09:33] -filter_complex '[0:0]null[VIDEO_IN]; [VIDEO_IN]null[VIDEO_OUT]' [09:33] ubitux: yes, that works [09:33] but that's hard to edit ;) [09:33] I prefer the command line to stay orderly... so if possible, I'd like to avoid that [09:33] calling -filter_complex again will overrid the string [09:33] (i believe) [09:33] ubitux: nope ;) [09:33] when I reverse them [09:33] it still complains about [VIDEO_IN]null[VIDEO_OUT] [09:34] there can be more than one graph, just how to connect them [09:34] it might be parsed, then overrided [09:34] why would anyone code that ;) [09:34] cause you're not supposed to need multiple -filter_complex? [09:34] what is the nb_filtergraphs variable good for then? [09:34] ok i think i've the loop patch but i'm unable to test it :( [09:35] divVerent: ah dunno, then i might be wrong [09:35] maybe for multiple outputs? [09:35] (output files) [09:35] apparently, the feature to have multiple graphs is intended [09:35] and I am trying to figure out how to use it [09:36] yes, possibly the graphs have to be independent [09:36] that'd be an annoying limitation though [09:40] oh btw ubitux [09:40] what is it with avconv ? [09:40] ffmpeg says that avconv should be used instead [09:44] it's the fork i was talking about [09:44] http://blog.pkh.me/p/13-the-ffmpeg-libav-situation.html [09:44] oh okay [09:45] tl;dr: ^F packaging on that page [09:45] btw, are you using a fork of the git repo for your patch ? [09:45] I could push your changes into my local repo to build ffmpeg [09:45] i'll likely send you a patch to git am, if i succeed [09:46] that's not yet guaranted :( [09:51] oh i get it working. [09:51] but it will only work with mpeg files (so you have to re-encode or remux it) [09:54] ashka: do you have a working ffmpeg git/master? [09:57] ashka: anyway, with git/master: wget 'http://b.pkh.me/0001-lavf-file-WIP-loop.patch'; git am 0001-lavf-file-WIP-loop.patch [09:57] and then something like ./ffmpeg -fileloop 1 -i loopme.mpg -t 3600 -y out.mkv [09:58] (assuming you ffmpeg -i loopme.webm loopme.mpg) [09:58] i'll submit this patch later [10:54] ashka: so? :) [10:55] ubitux: it just finished compiling [10:55] currently trying it out [10:56] [matroska,webm @ 0x1da3340] Unknown entry 0x18538067\n[matroska,webm @ 0x1da3340] Unknown entry 0x1A45DFA3 [10:56] got a whole lot of these all of the sudden [10:56] it's spamming [10:56] not sure if it's still writing output [10:56] what's your cmd line? [10:56] no, output is stuck [10:56] ffmpeg -fileloop 1 -i in.webm -t 3600 out.mkv [10:57] oh nvm [10:57] I didn't see the line assuming blah [10:57] my bad [10:59] explaination: mpeg streams are concatanable, so you i can just restart sending packets from the beginning [10:59] it's not possible with mkv [11:00] ideally we should fix the movie=...:loop=0 thing, but hopefully the patch should be a temporary workaround for your needs :p [11:00] btw, you might even be able to ffmpeg -fileloop 1 -i loopme.mpg -t 3600 -c copy -y out.mpg [11:01] oh really ? [11:01] should be way faster [11:01] I'll try [11:01] hmm nope [11:01] av_interleaved_write_frame(): Invalid argument [11:01] ok :( [11:01] ([matroska @ 0x222dca0] Can't write packet with unknown timestamp) [11:01] note the mpg ? mpg [11:01] oh [11:02] [mpeg @ 0x2e38ca0] packet too large, ignoring buffer limits to mux it\n[mpeg @ 0x2e38ca0] buffer underflow i=0 bufi=44059 size=44461 [11:02] still, is that okay ? [11:02] dunno i didn't try [11:02] yet it looks like it's copying anyway [11:02] I'll wait for it to be done [11:20] hmm [11:20] ubitux: it worked [11:20] I have a little additional question [11:21] can I cut out a few frames every time at the beginning of the original ? [11:21] since it's a music thing I need to make it so it syncs [11:21] in multiple steps :p [11:22] ./ffmpeg -i in.webm -ss 12 loopme.mpg [11:22] to skip 12 seconds [11:22] oh [11:22] how can I skip 15 frames ? [11:22] mmh a bit more painful [11:22] hmm [11:22] maybe a precise amount of ms [11:23] is it only video? [11:23] hm... I suppose I am compiling ffmpeg wrong... with virtually identical options, my mplayer based libavcodec encoding code outperforms ffmpeg's own conversion by far... just looking for ideas where to look [11:23] video + audio [11:23] then use a precise ts [11:23] -ss 12.345 [11:23] okay, I'll try [11:23] ah, okay... the x264 parameters differ, just why... ;) [11:23] haha, yeah -- ffmpeg has problems if you just want to do trim(0,100) or something like that, because you have to set stuff by times [11:24] divVerent, ffmpeg should nowadays use stuff pretty close to x264's defaults by default [11:24] exactly [11:24] that's why I wonder [11:24] so does my mplayer based encoding code (NOT mencoder) [11:24] I found one differing x264 option... took it out now [11:24] I think one of the only things it does differently from command line x264 is that libx264 doesn't limit refs by level [11:25] but that's comparing x264cli and libx264 [11:25] tune=animation... okay, that brings ffmpeg up to 54fps, still with mplayer I get like up [11:25] yes, and I am comparing two different libavcodec/libavformat frontends [11:25] which are compiled against the same library [11:26] CPU usage is 200% (I have two cores) in both cases [11:26] AH... I see ONE difference. The aac codec... [11:26] wonder if libfdk_aac is slow ;) [11:26] nope, that's not it [11:27] is there any way to "Profile" ffmpeg in a simple way? [11:27] like, to get simple output like "25% time in decoding, 30% in filtering, 50% in encoding, -5% in lying"? ;) [11:27] someone was asking on the devel channel yesterday i think :p [11:28] HA! found the cause [11:28] ffmpeg not at fault ;) [11:28] stupid notebook... suddenly reduced clock speed [11:28] you have a -benchmark(_all?) option btw [11:28] still, such profiling would be nice [11:29] BTW, style question regarding ffmpeg options: be lazy or add stream suffix always? ;) [11:29] for options that only the v codec but not the a codec knows [11:29] e.g. -tune:v animation [11:30] no idea :) [11:30] and thanks, -benchmark_all is nice but weirdly inaccurate [11:30] probably needs processing the output (summing up by the various step names) [11:31] divVerent: you'll have to use a tool like gprof on ffmpeg build with debug information in [11:31] for profiling [11:32] right, I didn't want the large cannons though ;) [11:32] just the little info useful for tuning options, like, in which of the major steps most time is spent [11:32] so I know if I e.g. have to change the swscale parameters or the x264 codec ones [11:32] gprof slows down the run a lot, which makes it somewhat unattractive [11:33] (okay, actually, compiling with -pg does ) [11:34] yeah well, that's because it's collecting data about function calls :) [11:34] sure, I know what gprof does... it's just not the tool of choice in many cases [11:34] in fact, I'd be highly surprised if it even works right when linking against a non-profiling libx264 [11:38] ubitux: works fine :) thanks a lot for the workaround [11:38] divVerent: it works, just the x264 data is missing :) [11:38] ashka: great, i'll submit a patch tonight, it might get upstream later [11:38] anyway, as you noticed, if you want fast benchmarking you won't get accurate results [11:39] sure [11:41] why is the hall of shame page down "until it is updated"... C&D? [12:11] oh btw ubitux [12:11] this is minor, but you might be able to fix it [12:12] a video of 10h will last 10:00:00.01 [12:12] totally minor [12:14] ubitux: my current (mostly autogenerated) filter chain: [12:14] -filter_complex '[SUBVIDEO_IN]scale=max(480\,floor(320*dar/2+.5)*2):max(floor(480/dar/2+.5)*2\,320)[scaled]; [scaled] setsar=1:1 [VIDEO_OUT];[0:0][0:1]overlay[SUBVIDEO_IN]' [12:51] .long_name = NULL_IF_CONFIG_SMALL("FFM (FFserver live feed)"), [12:51] is this format considered stable in ffmpeg? [12:51] i.e. can I "safely" use this as interchange format from a program to ffmpeg? [12:51] it is probably the simplest somewhat feature complete format we have [12:52] the one catch probably is that it depends on some enums in ffmpeg headers, especially AVCodecID and that pixel format enum [12:53] (in my application, I'd only want to send rawvideo and PCM audio, and let ffmpeg do the encoding) [12:53] divVerent, did you check format 'nut' [12:54] ffm is ffserver-specific format [12:54] nut is quite complex [12:54] I want something I can generate from like 100 lines of code [12:54] or is there a spec of nut and what ffmpeg's muxer does is way over the top? [12:55] I mean, probably nut is far from that complex, when you restrict it to the particular use case (rawvideo/pcm) [12:55] check the source code :) [12:55] the source code is exactly what doesn't help me here :) [12:55] nutenc.c is quite complex still [12:56] a spec would help here, obviously [12:56] did you try [12:56] :) [12:57] I actually did some months ago [12:57] but found nothing useful [12:57] ah, now I see [12:57] the spec is hidden in mplayer sources [12:57] wtf [12:58] http://code.google.com/p/mplayer-mirror/source/browse/trunk/DOCS/tech/mpcf.txt?r=11131 looks like what I had wanted [12:58] fflogger doesn't like nut format apparently :) [12:58] hehe [13:00] one thing about nut format I don't get though [13:00] is it allowed if the timestamps are "messy"? ;) [13:00] like, can one happily encode half a second of video, then half a second of audio, etc. [13:01] or do timestamps have to monotonous across all streams (like e.g. ogg requires IIRC) [13:02] http://wiki.xiph.org/Nut_Container [13:57] is it a bug or a feature that I can't extract subtitles from an ogm file to ass directly [13:57] but can when going via mkv? [13:57] ogm uses CODEC_ID_TEXT subtitles [13:57] trying to convert this to ass says that there is no decoder for the codec, which is true [13:57] but using the "copy" codec to plug into mkv, then going from mvk to ass works [13:59] i.e. "works as intended", or "to the tracker"? [13:59] https://gist.github.com/3865174 - the shell script part in question [14:01] $t here is the codec_type [14:05] the mkv file claims to have the subtitles in "subrip" format, which is what I would also expect here [14:07] i don't understand the question/problem [14:07] what are you trying to do? [14:07] I want to export ogm subtitles as .ass [14:07] so I can use them with the vf_ass filter [14:10] haha, I now see why my hack works [14:10] {"S_TEXT/UTF8" , AV_CODEC_ID_SUBRIP}, [14:10] {"S_TEXT/UTF8" , AV_CODEC_ID_TEXT}, [14:10] {"S_TEXT/UTF8" , AV_CODEC_ID_SRT}, [14:10] I start with AV_CODEC_ID_TEXT, which I plug into mkv via -codec copy [14:10] it becomes S_TEXT/UTF8 in the mkv [14:10] now, when READING this mkv file again, it becomes AV_CODEC_ID_SUBRIP [14:10] which can be converted to .ass fine [14:12] so... doesn't that mean that the ogm would have chances to work, if the demuxer decided on AV_CODEC_ID_SUBRIP instead of AV_CODEC_ID_TEXT? [14:13] if yes, this sounds like trac material [14:27] divVerent: and ffmpeg -i in.ogm out.ass doesn't work? [14:27] exactly [14:28] to do it with two short commands: ffmpeg -i in.ogm -codec copy -map 0 temp.mkv && ffmpeg -i temp.mkv out.ass [14:28] works fine [14:28] oops, actually the latter may not work [14:28] needs -vn -an probably ;) [14:28] but you get the idea [14:37] can i have a sample? [14:37] don't have one at a place from where I can upload... but my guess is that any ogm with subs will work [14:38] hm... maybe I can make one quickly somehow [14:41] I just encoded real crap quality... don't care ;) [14:42] divVerent: -map 0:s would copy just the subs [14:42] damn... ffmpeg refuses to write the ogm file I want... need ogmmerge then ;) [14:43] it also refuses to plug srt INTO ogm [14:43] :) [14:46] http://ompldr.org/vZnRxbw/out2.ogm [14:46] test file for this [14:46] thx [14:47] nice video test source [14:47] hehe [14:47] it's a test image generator I am working on [14:47] maybe we could improve our -f lavfi -i testsrc :) [14:48] probably not. Different applications need different test images. [14:48] haha, that one also uses a LCD hack [14:48] to get simple digit rendering code ;) [14:48] :) [14:50] my filter BTW is a dynamically loadable filter for some mplayer fork... it PROBABLY should be easy to port to other code bases [14:50] given it basically works on raw yuv444p output [14:50] in planes [14:50] the background is BTW a nice test for telecine/detelecine filters ;) [14:50] Dialogue: 0,0:00:00.50,0:00:02.50,Default,Hello, world! [14:50] mplayer's -vf filmdint horribly fails it, -vf pullup works [14:50] ok got it. [14:51] but only via mkv, right? [14:51] these are the exact times I set [14:51] just a quick hack [14:51] - st->codec->codec_id = AV_CODEC_ID_TEXT; [14:51] + st->codec->codec_id = AV_CODEC_ID_SUBRIP; [14:51] hehe, I see [14:51] in libavformat/oggparseogm.c [14:51] i'm looking at making a text decoder [14:51] i though we had one.. [14:52] divVerent: do you know the markup of subtitles in ogg? [14:52] no markup at all? that's really plain text? [14:52] don't know [14:52] never seen them have any markup [14:52] they always look plain [14:52] not even and crap like that? [14:52] haha, now I see the difference between CODEC_ID_SRT and CODEC_ID_SUBRIP [14:52] I am not aware of any [14:52] ok [14:53] yeah the SUBRIP is to workaround a problem [14:53] originally the packets included the timestamps [14:53] right [14:53] like in SRT files [14:53] and it was a pain for mkv demuxer for examples [14:53] (it had to write the ts in the payload) [14:53] and SUBRIP is the timestamp-less version [14:53] which uses pts [14:53] yes, that's actually the "codec" [14:53] the srt demuxer should be fixed [14:53] to output subrip packets [14:54] so is CODEC_ID_SRT still in use? [14:54] ah, THERE it still is used ;) [14:54] and we could get rid of CODEC_ID_SRT [14:54] :) [14:54] right [14:54] i need to do a lot of work on the subtitles [14:54] personally, I think the right way to handle ogm is to use the SUBRIP format [14:54] a long work in progress :) [14:54] because you almost always embed srt files into ogm [14:54] really? [14:54] this is just how these are made with ogmmerge [14:54] it wants srt input [14:54] and just sticks them in with no conversion of markup if any [14:54] what happens if you merge a file with markup? [14:54] it doesn't care [14:54] it just plugs it in [14:54] great.. [14:55] well then i guess the patch i propose could be pushed [14:55] if you can wait until tonight i'll submit it [14:55] ogm isn't very common any more [14:55] (or you can submit the patch right now) [14:55] and I doubt thsi has ever been specified [14:55] given ogm was created as a hack to make an "avi replacement that can embed subtitles" [14:55] based on ogg [14:55] what about ogg? [14:56] at least ffmpeg's ogg demuxer has no subtitle support [14:56] not sure if the container supports them [14:56] of course, one can always cause the ogm specific code to be invoked ;) [14:56] as ogm basically is a superset [14:56] with FOURCCs and such crap [14:56] xiph.org probably should know if this ever was intended... [14:56] i mean, does ogg defines the way to store text subtitles? [14:57] anyway, i'll submit tonight for comments [14:57] we'll see [14:57] right [14:58] if anyone has a complaint, they will say so ;) [14:58] rhaa i need to find some time for all the subtitles thing :( [14:58] this MAY break players that use lavf to decode subtitles (hint: mplayer) if for some reason they handle ogm/text but not ogm/srt [14:58] can't imagine why though [14:58] why would it break? [14:58] the demuxer is changing the codec [14:58] exactly [14:58] so mplayer will be aware of it [14:59] if a player for some reason only supports CODEC_ID_TEXT but not CODEC_ID_SUBRIP [14:59] then it will break [14:59] in fact, it looks like it WILL break in mplayer2 at least [14:59] but wonder how it plays srt-in-mkv then [15:00] https://gist.github.com/3865504 - this code section makes me think that [15:00] oh that sucks. [15:01] just, IF that is the case, shouldn't mkv playback with such subs already be broken [15:01] i remember seeing some patches in mplayer indeed [15:01] but not mplayer2 [15:01] since it's still mostly based on libav by default [15:02] but mplayer2 has its own demuxer by default so.. [15:02] it IS broken [15:02] mkv* demuxer [15:02] in mplayer2 [15:02] but only with -demuxer lavf [15:02] because it has its own mkv demuxer [15:02] and thus by default doesn't hit this issue [15:02] the mkv demuxer is pretty nice in mplayer2 :p [15:02] okay, go ahead then [15:02] so no reason to fallback on lavf :D [15:02] it's easy to fix in mplayer2 [15:02] no it's not [15:02] sure it is [15:02] because libav has no SUBRIP codec id [15:02] they just have to support CODEC_ID_SUBRIP too [15:02] have fun. [15:02] haha [15:03] even then, should this really stop ffmpeg? [15:03] okay, mplayer-svn then can add CODEC_ID_SUBRIP to that list ;) [15:03] it's already done i believe [15:03] and mplayer2 uses libav anyway so they won't ever see the patch [15:03] (in mplayer) [15:03] mplayer2 is supposed to be buildable against the two [15:03] so you need some conditional crap [15:03] I just wonder one thing [15:03] you might want to discuss this with uau :) [15:03] the alternative would be better CODEC_ID_TEXT support, BUT... [15:04] in case of ogm, the proper type is actually SUBRIP I am pretty sure [15:04] we could introduce a codec id text decoder [15:04] but it's different [15:04] I think I once saw tags on the screen with mplayer years ago [15:04] in ogm files [15:04] it would mean raw text [15:04] so apparently someone did it [15:04] right, in case of ogm, the actual source is typically srt though [15:04] and ogm encodes no "more exact" info [15:04] maybe we could assume text is subrip in all/most of the case [15:05] because most muxers will end up muxing crap at some point [15:05] under the "text" name [15:05] is the markup even a subrip feature [15:05] or is that just an extension by many players and then used by srt scripts? [15:05] it's supposed to be [15:05] i don't know much the history [15:05] but it's associated with it at least [15:06] basically, in my opinion two things are needed to fulyl resolve all of this ;) [15:06] 1. the ogm demuxer change (it is correct for typical ogmmerge usage, and ogmmerge IS the one reference ogm muxer) [15:06] 2. adding proper CODEC_ID_TEXT support wouldn't be bad either ;) [15:06] ok so far [15:07] as CODEC_ID_TEXT can still come out of other sources, even mkv [15:07] it's pretty easy to write actually [15:07] sure, probably copypaste the srt file and remove all parsing ;) [15:08] yes [15:09] ok i'm going to write it asap [15:09] hopefully submitted tonight [15:09] so much pending patches today.. [15:10] lavfi meta inject, loop in file protocol, webm regression, ogg/text/subrip, and now text decoder... [15:10] quite a productive day [15:13] and I have replaced my mplayer encoding use by ffmpeg for a change... just wondering whether it'd be a good or bad idea to release these horrible shell scripts ;) [15:13] which do language based stream selection, hardsubbing (both of DVD and ASS subs) and still support custom filter options by the caller [15:13] :D [15:14] is there BTW an easier way to do this: [15:14] scale=\ [15:14] $mode($w\\,floor($h*dar/$div+.5)*$div):\ [15:14] Action: ubitux thinks he's going to support all the mpl2 vplayer and crap in one row.. [15:14] $mode(floor($w/dar/$div+.5)*$div\\,$h)\ [15:14] I basically want to scale with 1:1 pixel aspect so that it in both dimensions is >= 480x320 (iPhone half res) [15:15] $mode is max here :P [15:15] so the general idea is, width = larger of (original width, target height * DAR) [15:15] and height = larger of (original height, target width / DAR) [15:15] did you look at the different variables in libavfilter/vf_scale, and the function in eval? [15:15] yes [15:15] i don't see any avg() [15:15] but I saw no easier way [15:15] :( [15:16] the $div crap is also needed... [15:16] mainly because x264 refuses odd dimensions [15:16] rint() is missing BTW [15:17] it's pretty easy to add functions in eval [15:17] oh, BTW, the reason why I do so weird scaling... the iPhone basically has a zoomed out view (default, image is letterboxed) and zoomed in view (i.e. cropped to screen aspect, scaled as large as possible) [15:17] and I optimize for the zoomed in view here [15:18] I just don't like havikng to use an expression evaluator for this... there should really rather be a way to say "fit into 480x320" or "crop and center to 480x320" [15:19] i agree with this [15:19] like the rescale in imagemagick? :) [15:19] yes [15:19] I was just looking up imagemagick's syntax for that ;) [15:19] i often wonder about this [15:19] 480x320^ [15:19] -resize WxH [15:19] would be imagemagick's name for what I want [15:19] and 480x420 would be the letterboxing version [15:20] 480x320! is the aspect-breaking version [15:20] maybe it would be possible to have a keep-aspect-ratio thing [15:20] anyway, i won't do it, but feel free to send a patch :) [15:20] yes, basically I was thinking of adding flags for this [15:20] two modes for that, obviously... just like imagemagick [15:21] of course [15:21] if the expression evaluator could work with complex numbers... [15:21] then we could do -filter:v "scale=aspect_letterboxed(w,h,480,320)" [15:21] and scale would work with a single expression returning w*i+h ;) [15:22] you can define custom functions [15:22] sure [15:22] with eval [15:22] but with the current way, you'd need two expressions still [15:22] like, width_letterboxed(w,h,480/320):height_letterboxed(w,h,480/320) [15:22] s!/!,! [15:22] g [15:22] oh, right. [15:22] that' [15:22] that's because of the nasty format :) [15:22] s how I got to complex numbers [15:23] args[strcpsn(s,":")]='x' [15:23] here you go \o/ [15:23] divVerent: there's no good spec for what markup subrip "should" support AFAIK [15:23] so CODEC_ID_SUBRIP would still be ambiguous [15:23] ubitux: the x is really the smallest issue ;) [15:24] mmh i'm stupid yeah. [15:24] also, width_letterboxed() even contains one ;) [15:25] also there are files with markup that could not conform to any sane spec [15:26] divVerent: i don't think that's really a problem to define two local functions in vf scale named lboxw() and lboxh() :p [15:26] like relying on the behavior of some players where libass tags are interpreted too (because srt support is implemented with an ASS renderer, and the implementation fails to properly quote things that can be interpreted as ASS tags) [15:29] hi [15:37] ubitux: it is not [15:37] but -vf scale=lboxw(DAR,320,240):lboxh(DAR,320,240) [15:37] is still a lot more verbose than [15:37] -vf scale=320:240:lbox [15:37] yup better syntax :) [15:38] also, lboxw is not sufficient alone ;) also need to round to codec specific mutliples [15:38] so... [15:38] -vf scale=round(lboxw(DAR,320,240),8):round(lboxh(DAR,320,240),8) [15:38] vs [15:39] -vf scale=320:240:lbox:round=8 [15:39] also, this is stupid anyway [15:39] real men would use [15:39] -vf scale=DAR 320 240 lboxw 8 round DAR 320 240 lboxh 8 round [15:39] and then wonder... why not... [15:39] you can add another filter [15:39] using the same internals as scale [15:40] with a different syntax [15:40] i don't remember how scale syntax was extended [15:40] -vf scale=8 DAR 320 240 4 dupn lboxw exch round lboxh exch round [15:40] ;) [15:40] it's longer, bit shows you know your RPN ;) [15:40] the current vf scale syntax parsing is quite hacky atm [15:40] yes, especially the comma abuse [15:41] using comma as separators both inside and outside is stupid [15:41] but... I know no better idea [15:41] comma? what comma? [15:41] semicolon is also already a separator in filter chains [15:41] in function args [15:41] oh in the eval [15:41] you actually can't do -vf scale=func(x,y):func(x,y) [15:41] yeah but i wasn't talking about that [15:41] but need -vf "scale=func(x\\,y):func(x\\,y)" [15:41] look at the sws flags parsing [15:41] mplayer has the same issue :P [15:41] or interl=1 thing [15:42] ah, I see [15:42] maybe you can just add another hack like strstr(args,"ratiorules=") [15:42] hehe [15:42] hack-on-hack-on-hack... ;) [15:42] :) [15:43] my favorite solution for -vf scale is still using complex numbers as an alternate interface [15:43] if only one expression is given, real part is w and imaginary part is h ;) [15:43] my favorite solution would be to have it in swscale if the api allows it [15:43] (through sws flags) [15:43] hehe [15:43] swscale that huge mess ;) [15:43] i'm not sure if you can change the specified sizes with sescale [15:43] I recently found lots of nasty bugs/features in swscale [15:43] maybe michaelni can tell [15:44] my favorite one: it loves writing between the image row end and the next row [15:44] i.e. it writes into the stride spacing [15:44] I know why it does that, it makes for faster SIMD code [15:44] and I also know the workaround - make sure your width is 16 bytes aligned [15:44] the main issue is that this fact is nowhere documented [15:45] and at any time, someone could write a SIMD scaler that works in 32 bytes blocks [15:45] ffmpeg.c is MOSTLY unaffected by this issue [15:46] except that when the block size of swscale is raised, there may happen reads beyond allocation which needs slightly larger av_malloc where images are allocated [15:47] did you notice some valgrind issues? [15:47] no [15:48] I had abused libswscale to scale part of an image to part of another image [15:48] then there is no problem ;) [15:48] (convert, actually) [15:48] there is, it does read beyond the stride [15:48] and write [15:48] but NORMALLY this is no issue [15:51] I can produce a valgrind log with a "tightly allocated" image, but the thing is that this is no bug, it's just missing documentation ;) [15:52] it's sensible that libswscale behaves the way it does [15:53] ideally, there should be a macro in swscale.h that defines the alignment libswscale wants, and a comment explaining that writes may happen in blocks of that size [16:40] hmm [16:40] ^ nevermind that [16:47] hey [16:48] can some give me a good idea on codecs for screencasting with ffmpeg [16:48] someone^ [16:50] how do they relate? [17:55] Hi. Is there a way - from ffplay - to see when ffmpeg is streaming an rtp channel? Thank you [18:58] Spideru, can you rephrase your question please? [18:59] burek: ok, thank you. If i start ffplay (with SDP file) and then connect an rtp stream with ffmpeg, all things works well. Then, if I stop ffmpeg, I can't recognize it from ffplay. How can I do that? [19:00] I need to know from ffmpeg and ffplay If the stream is not working [19:00] what does this mean: and then connect an rtp stream with ffmpeg [19:01] can you show some sample command lines? [19:01] (please use pastebin if commands are too big) [19:01] yes thank you [19:03] burek: http://pastebin.com/eTg9MW1e [19:03] the stream is perfect and finally the patch for SDP generation is up :) [19:04] but I would to know if something stop working [19:04] Spideru, if I get this right, you are using ffmpeg to feed some unknown rtp server (?) and then you use ffplay to connect to the server? [19:04] I use ffmpeg to feed ffplay server [19:05] what is ffplay server? [19:05] :| just realized now that I'm using it in wrong way [19:05] ffplay is just a player [19:05] your streaming server is located at 192.168.1.95 [19:06] and ffmpeg is just a stream source [19:06] Well, what can I use to wait and get the ffmpeg stream source? [19:06] I don't understand :/ [19:07] Ok, I try to explain it better [19:08] I would to transfer audio live stream from a client to a server. What can I use to do that? [19:08] Now I'm using ffmpeg as client, and ffplay as server. But ffplay is not a server [19:09] So, what I should use instead of ffplay? [19:09] and, ffmpeg is the right choice as client side? [19:10] ffmpeg - as a source [19:10] ffserver - as a server/broadcaster [19:10] ffplay/vlc/winamp/... - as a player [19:10] ahh so -> ffmpeg ----> ffserver <---- ffplay ? [19:10] yes [19:11] \o/ thank you! [19:11] :) :beer: [19:11] of course [19:11] where are you from? [19:11] still earth :) [19:11] serbia :) [19:11] I can offer you a pizza if you'll come to Italy :) [19:12] I've tried your pizzas and they are good :) [19:12] I know, and my fat too [19:14] :) [19:17] burek: What kind of protocol should I use? RTP or RTMP? [19:18] I'll prepare two pizzas instead one [19:19] :) [19:19] i use ffm between ffmpeg-ffserver and flv between ffserver-media players [19:19] so its more compatible for streaming [19:23] ffm? Never heard about it before [19:23] of course excluding youporn [19:24] read a little bit about it [19:25] Yes, I'll read all the doc, of course. [19:26] and thank you [19:26] your help saved me a lot of time [19:28] you're welcome :) [19:36] 2.7.2 The audio and video lose sync after a while. [19:36] Yes, they do. [19:36] Wonderful [20:31] does the latest ffmpeg release support apple prores 422 decoding? it seems to give me a black video output [20:59] built from git head and am seeing the video now, so that's good [21:34] hi [21:35] We need to uncompress an MPEG-2 file (Derived from a DVD) to an uncompressed AVI. We've got usable video (using -vcodec rawvideo), but what would the equivalent be for audio? I assume we just need to specify 16 bit PCM somehow? [21:36] -acodec pcm_s16le [21:36] ffmpeg -codecs | less [21:37] I'm aware of the pcm codecs but I wasn't aware which one was most suitable. [21:37] The There's no documentation other than that they exist. [21:38] signed 16bit little endian is the most common [21:38] hmm [21:39] sound works [21:39] picture less well [21:39] I guess "Rawvideo" just gives us raw 8 bit rgb [21:39] is there something else we might try that a nonlinear editor would deal with better? [21:40] which one are you using? [21:40] premiere [21:40] but it hardly matters [21:40] most of 'em would read an uncompressed AVI... usually. [21:42] Action: Jan- tries v210 [21:43] bingo :) [21:44] it always starts with "error decoding stream #0.1" [21:44] is that bad? [21:44] It can't be good. [21:45] is the exit status 0? [21:46] it didn't exit [21:46] it just said error decoding stream and continued. [21:46] To be fair weirdness is sort of expected, this is a VOB off a dvd. [21:46] the issue is that we need to extract a chunk of it that crosses a VOB boundary, and trying to get a nice clean decode of every single frame each side of the join is tough. [21:47] you could use mplayer's -dumpstream to read the dvd [21:47] wanted to avoid having to do that if possible [21:48] om nom nom all ur disk space r belong 2 mplayer etc [21:48] but I guess we're already having to convert the entire preceding vob. [22:09] what does -vcoded copy mean? [22:10] the command in question is this: ffmpeg -i input-file.m2ts -ab 256k -vcodec copy -acodec aac output-file.mp4 [22:10] lake: "vcodec copy" means "use the same video codec in the output as the input" [22:11] copy the video stream [22:11] tmatth: does that result in loss of quality? [22:11] lake: simply put, it means to "copy and paste" the video stream from input to output. it does not re-encode. [22:11] no loss o' quality [22:12] cp llogan pasteeater [22:12] would this result in any loss of quality: ffmpeg -i input-file.m2ts -ab 256k -vcodec x264 -acodec aac output-file.mp4 [22:12] i assume it would result in a reduced file size [22:13] you probably mean "libx264", not "x264" [22:13] and yes, it would reduce quality with the default settings. [22:13] llogan: yes, sorry, just trying to wrap my mind around it [22:14] since you are re-encoding to a lossy format..but you may not notice a difference with high enough quality settings/bitrate [22:14] so, we are talking about transcoding vs compressing then? [22:14] of course x264 can encode lossless as well, but it doesn't mean that there will be no loss (such as going from rgb to yuv). [22:15] sorry, i'm a noob, but finding that ffmpeg is amazing [22:18] transcoding refers to re-encoding while using information from the input such as motion vectors. re-encoding means to decode the stream into individual pixels and encode with no additional information from the input [22:18] at least that's how i interpret it [22:19] most of the time those terms are used as if they are the same process [22:20] is there a way to specify the library path for libx264 when building from source? [22:20] seems to be not finding it or using the wrong one [22:24] wallerdev: did you try --extra-cflags and --extra-ldflags? [22:25] no i didnt [22:25] some examples here: http://ffmpeg.org/pipermail/ffmpeg-user/2012-August/008552.html [22:26] would it just be something like --extra-ldflags "-L/usr/local/lib" or somethin [22:26] i don't know how "good" those examples are [22:26] ill check that out thanks [22:26] --extra-cflags="-I/path/to/prefix/include" --extra-ldflags="-L/path/to/prefix/lib" [22:27] that's better [22:31] well hopefully it works, otherwise something is just messed up with my system haha :) [22:32] burek, relaxed: interested in making a mini guide on trac wiki on how to make static builds? [22:53] ashka: i don't think i'll submit the patch now, it still has some issues [22:53] oh okay [22:53] well the workaround is okay for me so it's great [22:54] :) [22:54] hi, i have a nub question for you all: how can I use ffmpeg to combine (multiplex) audio files into one [22:55] when did noob become nub? nubs seem to outnumber noobs now. [22:56] haha [22:56] i like the way it sounds, nuhb [22:56] nublet [22:59] it's&cuter [23:00] burek: how can I tell to ffm (of ffserver) that I don't want video inside feed? [23:00] mykul: you mean to simply add several audio streams into one container, or to concat them all into one continuous stream? [23:01] mykul: look at concat, amerge, pan and amix filters [23:01] depending on your needs [23:01] llogan, not concat, but have them combined so that they play at once. is that what adding to one container means? [23:02] thanks ubitux, i'm drowning in options :D [23:03] Found it thank you [23:05] mykul: not exactly what i had in mind. i guess you want "option 3" that i didn't think to list. as ubitux mentioned, see amerge and amix. [23:06] cool, thanks llogan, i am armed with some language now. [23:06] https://www.ffmpeg.org/ffmpeg.html#amerge [23:14] amix worked great [00:00] --- Thu Oct 11 2012 From burek021 at gmail.com Fri Oct 12 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Fri, 12 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121011 Message-ID: <20121012000502.E84C218A01E8@apolo.teamnet.rs> [00:00] ubitux, by just "changing the message", you are actually accepting that it's ok for libav to distribute your own tool.. wouldn't it rather be ok to ask debian to either entirely remove ffmpeg or to entirely remove the problematic (3rd party) message [00:00] and if they really want to inform their users about avconv, they can do it in the description for the deb package itself [00:01] and not crippling the binary with falsified messages that confuse a lot of people [00:01] by changing the message, i'm accepting the debian has technical and political issues with packaging ffmpeg [00:01] ffmpeg.git 03Carl Eugen Hoyos 071a104bf641b1: Fix broken timestamps for some mp3 in avi samples. * 03http://tinyurl.com/9crqlqs03 [00:01] burek: well, if debian decides to package libav, which is their choice, they have to inform the user that a given program will disappear [00:02] when people see that ffmpeg binary outputs that "message" it is normal to expect that they understand that FFmpeg developers wrote that message.. that's what is wrong in all this [00:02] (and that syntax will change etc) [00:02] yes [00:02] that's why i just want to fix the message :) [00:02] you don't need to make effort and ask them politely to "change" the message [00:02] really it's not much. [00:02] you should either sue them for copyright infringement or demand they remove it immediately [00:03] i'm not familiar with the legal system, and i don't like that stuff [00:03] well, if someone takes your work and changes the output of your program, claiming that you did it, then what? [00:04] any country will take you to the jail for that [00:04] but, anyway, as one smart person here told me once :) I have much important things to do then to fight other people's battles :) [00:05] burek: really, it's quite complex to switch package names in debian [00:05] since they want to keep retro compat [00:05] it's a "stable" distro [00:05] it's their problem, not yours [00:05] yes but i understand it's not solvable now [00:05] imho, switching the project is not the most compatible thing to do. [00:05] and if they aren't even willing to fix a fucking message [00:06] they won't take the pain to fix that correctly [00:06] lol at going to jail for modifying an open source program's output [00:06] gnafu, you are not modifying it, you are distributing it [00:06] not just * [00:06] Still I have never once heard of someone going to jail for that. [00:06] enough derping for tonight [00:06] Action: ubitux & [00:06] 'night [00:07] gnafu: over-reaction, maybe [00:07] burek: the LGPL explicitly allows you to do that [00:07] Trademark law is what's relevant here, not copyright. [00:07] I don't know why I even spend my time on this topic any more [00:07] do what ever you want guys [00:08] its you who will profit/suffer from all the results [00:08] You're probably better off not unless you understand the details involved [00:08] (@devels) [00:08] Just yelling about suing people doesn't actually solve problems [00:08] I agree [00:08] & :) [00:10] ubitux: hi is not changing the message because the issue is obscure and he can get away with it. But if it is a public issue... then pointing that he had rejected a perfectly acceptable proposal would be quite damning. [00:10] hi/he [00:10] (Basically, using legal solutions to social problems isn't generally a good idea) [00:11] absolutely. court should be the last instances, and the first thing they would ask is "what you did in order to try fixing this problem?" [00:11] [22:33:23] burek, relaxed: interested in making a mini guide on trac wiki on how to make static builds? <- yes, but i _really_ need some help with that, since i'm in no way the best person to write about such thing [00:12] that cron for static builds i've made was.. so to say.. a lucky shot :) [01:16] burek: ok. it was just an idea, and i'm willing to help [01:18] no problem, we'll see in next couple of days to get it done :) [01:18] days? i usually take months, but you're uncommonly fast (ie fflogger, etc) [01:21] i hope it will stay that way :) [01:43] who wants an account at coverity for ffmpeg ? [01:46] michaelni : just make up some new accounts and you can pass them out whenever someone wants one :P [01:47] I think you should ax this question on the mailing list [01:47] llogan, yes [01:47] Compn, i need to enter full name and email address of each person [01:47] but compn has no last name [01:48] sorry, you're excluded [01:48] john doe - random at ffmpeg.org etc [01:48] Action: Compn has never truthfully filled out a form on the internet [01:48] ehe [01:48] lots of people live at 123 fake street [02:00] true :D [02:23] ffmpeg.git 03Michael Niedermayer 07cac749a551b2: vf_idet: fix free after use * 03http://tinyurl.com/8goykuz03 [02:23] ffmpeg.git 03Michael Niedermayer 074d4f431ab7b1: vf_idet: zero pointers after freeing references * 03http://tinyurl.com/9q5jscz03 [02:23] ffmpeg.git 03Matthieu Bouron 07e782d8728f1e: fate: add vf_idet filter to lavfi regression tests * 03http://tinyurl.com/9k5hvap03 [02:23] ffmpeg.git 03Michael Niedermayer 07fac1ccbda1bb: swscale-test: fix freeing of uninitialized variable * 03http://tinyurl.com/94tzctg03 [03:05] ffmpeg.git 03Michael Niedermayer 074e4ae2f82caa: sha: change loop condition to be tighter. * 03http://tinyurl.com/9g2tzya03 [03:05] ffmpeg.git 03Michael Niedermayer 07989c91b5042c: asrc_aevalsrc: Fix use of uninitialized pointer inside av_strtok() * 03http://tinyurl.com/95c795603 [03:07] j-b, can you or thresh make cone print more than the first line ? (for example one cant see if a commit fixes some/which ticket coverity id now) [03:08] was this coverity scan just published or something? [03:11] oh [03:11] i had a naughty idea [03:11] argh no that wont work, nevermind [03:38] ffmpeg.git 03Michael Niedermayer 07bdcff5af7f08: af_volumedetect: fix use of uninitilaized variable in case of planar audio. * 03http://tinyurl.com/8lxcxdv03 [03:38] ffmpeg.git 03Michael Niedermayer 074334ba043e96: ffprobe: fix use of uninitialized pointer in av_strtok() * 03http://tinyurl.com/8zeftwy03 [04:03] ffmpeg.git 03Michael Niedermayer 07e47ab0b2c93e: ffmpeg_opt: dont fail for sameq/same_quant. * 03http://tinyurl.com/8aczbce03 [04:38] Program received signal SIGSEGV, Segmentation fault. :) [04:47] ffmpeg.git 03Michael Niedermayer 077df9f595c912: swri_resample_init: unsupported sample formats are an internal error. * 03http://tinyurl.com/8oym4k603 [04:47] ffmpeg.git 03Michael Niedermayer 07492b8ec4c5f5: av_opt_set_from_string: fix memleak * 03http://tinyurl.com/8wn8mow03 [07:26] what's this CID thing? [07:30] oh, a coverity thing. [10:11] ffmpeg.git 03Paul B Mahol 078cd1c0febe88: pcx: convert to bytestream2 API * 03http://tinyurl.com/8sza24f03 [10:54] saste: a few seconds to encode a xface? oO [10:54] fear :D [10:55] ubitux, ~1000 bigint divisions, and apparently our division algo can be improved [10:55] Action: durandal_1707 wtf, facebook ask me to log in [11:10] ffmpeg.git 03Stefano Sabatini 07396648cc6a3e: configure: link flite against libasound * 03http://tinyurl.com/9fa28bu03 [11:18] damn my last commit is wrong [11:18] disallow compilation wherever asound is not available [11:24] ffmpeg.git 03Stefano Sabatini 07c4aaff8c02ce: Revert "configure: link flite against libasound" * 03http://tinyurl.com/8lbb8k603 [12:45] ubitux: any plan to support metadata printing through drawtext? [13:08] saste: nope but that would be nice [13:09] saste: btw, about moving scene detection out of select can be done but... [13:09] how do you add metadata support in select? [13:09] yes that's no easy [13:09] maybe we could have a separate selectmeta [13:09] and same for drawtext actually [13:10] we need another formatting systeme [13:10] -e [13:10] yes, desigining it in a generic way may be tricky [13:10] (so we could add meta replace, or timestamp, or things like that) [13:11] the string interpolation thing i wrote a while ago might help for the drawtext thing [13:11] but for select i don't know [13:11] we need to extend the eval api [13:11] ideally: drawtext="$$pts:$pts $$pts_time:$pts_time ${meta/comment}" [13:12] to support an AVDictionary or something [13:12] well you could imagine in eval something like "@scene" [13:12] so it will pick in the passed dict [13:49] saste: unrelated: https://www.youtube.com/watch?v=KJe9H6qS82I :) [14:08] ffmpeg.git 03Diego Biurrun 07ada12f836657: svq1: K&R formatting cosmetics * 03http://tinyurl.com/95akuqh03 [14:09] ffmpeg.git 03Mans Rullgard 0741e46a5fbacc: parseutils-test: do not print numerical error codes * 03http://tinyurl.com/965ukvg03 [14:09] ffmpeg.git 03Diego Biurrun 0763a46c6101ff: svq1: Drop a bunch of useless parentheses * 03http://tinyurl.com/953zgp903 [14:09] ffmpeg.git 03Jean-Baptiste Kempf 07507dce2536fe: arm: call arm-specific rv34dsp init functions under if (ARCH_ARM) * 03http://tinyurl.com/9a8anm303 [14:09] ffmpeg.git 03Janne Grunau 07706a559b3008: a64multienc: change mc_frame_counter to unsigned * 03http://tinyurl.com/8gm6wxy03 [14:09] ffmpeg.git 03Mashiat Sarker Shakkhar 077fb35ee931c8: vc1dec: Use correct spelling of "opposite" * 03http://tinyurl.com/8cqtaam03 [14:09] ffmpeg.git 03Luca Barbato 0782569b01a1cb: mpegtsenc: set muxing type notification to verbose * 03http://tinyurl.com/9ud7n6t03 [14:09] ffmpeg.git 03Luca Barbato 07b522000e9b2c: avio: introduce avio_closep * 03http://tinyurl.com/8rngceh03 [14:09] ffmpeg.git 03Michael Niedermayer 07de31814ab07f: Merge commit 'b522000e9b2ca36fe5b2751096b9a5f5ed8f87e6' * 03http://tinyurl.com/9xnz9ga03 [14:12] \o/ [14:32] ffmpeg.git 03Luca Barbato 0726db9100b2fa: segment: support applehttp style list * 03http://tinyurl.com/9dn55ux03 [14:32] ffmpeg.git 03Mashiat Sarker Shakkhar 0788058d9a994f: vc1dec: Set chroma reference field from REFFIELD for 1REF field pictures * 03http://tinyurl.com/8d8mry603 [14:32] ffmpeg.git 03Michael Niedermayer 07b6c3487e7f2f: Merge commit '88058d9a994f42e4e9ed4e67baf696bbfe53128c' * 03http://tinyurl.com/8crhvvn03 [14:36] ffmpeg.git 03Mashiat Sarker Shakkhar 077cc3c4e1d417: vc1dec: Invoke edge emulation regardless of MV precision for 1-MV chroma * 03http://tinyurl.com/8v98af403 [14:36] ffmpeg.git 03Michael Niedermayer 07ce27c9eb2598: Merge commit '7cc3c4e1d4179aeabcd891090e31ee5e5bfd9692' * 03http://tinyurl.com/8lsxjqf03 [14:39] ffmpeg.git 03Mashiat Sarker Shakkhar 07eb657ecefdeb: vc1dec: Set opposite to the correct value for 1REF field pictures * 03http://tinyurl.com/94hcfq203 [14:39] ffmpeg.git 03Michael Niedermayer 07a75dd13b1bc2: Merge commit 'eb657ecefdeb8b2ed9bfb55d3c2c9e0f568486bf' * 03http://tinyurl.com/9aabvqk03 [15:02] if input is pipe there is no way to find its size? [15:02] no, pipes dont tell you their size [15:02] what would the size represent? the number of bytes written? [15:10] how parser than work for last packets when using pipes? [15:18] what's about this coverity account? [15:51] ffmpeg.git 03Mashiat Sarker Shakkhar 0735a35d49d23c: Double motion vector range for HPEL interlaced picture in proper place * 03http://tinyurl.com/8qnj74u03 [15:51] ffmpeg.git 03Janne Grunau 07bd141f5ec97f: mxfdec: return error if no segments are available in mxf_get_sorted_table_segments * 03http://tinyurl.com/8q9ec2e03 [15:51] ffmpeg.git 03Janne Grunau 076d556e8327f6: indeo4/5: remove constant parameter num_bands from wavelet recomposition * 03http://tinyurl.com/9pxaala03 [15:51] ffmpeg.git 03Janne Grunau 07c466eb174699: flashsv: propagate inflateReset() errors * 03http://tinyurl.com/9fq5yxt03 [15:51] ffmpeg.git 03Janne Grunau 0725227c3a78fe: averror: explicitly define AVERROR_* values * 03http://tinyurl.com/8dvnddg03 [15:51] ffmpeg.git 03Janne Grunau 078d09d39a4b87: avconv: remove bogus warning when using avconv -h without parameter * 03http://tinyurl.com/9l8hcg203 [15:51] ffmpeg.git 03Janne Grunau 07b404c6605627: fate: add h263 obmc vsynth tests * 03http://tinyurl.com/9krchdj03 [15:51] ffmpeg.git 03Mans Rullgard 07568c70e79ee2: lavfi: convert input/ouput list compound literals to named objects * 03http://tinyurl.com/8rtgys803 [15:51] ffmpeg.git 03Mans Rullgard 074436f25a1682: build: remove references to unused EXTRAOBJS variable * 03http://tinyurl.com/8lte4qf03 [15:51] ffmpeg.git 03Michael Niedermayer 07526cb36e4b23: Merge commit '4436f25a1682ada3f7226cb6fadf429946933161' * 03http://tinyurl.com/9xfd3rz03 [16:00] ffmpeg.git 03Mans Rullgard 07e5c6e9a6f2d0: build: remove single-use variable THIS_LIB * 03http://tinyurl.com/8wpfwne03 [16:00] ffmpeg.git 03Mans Rullgard 071c7428e6554a: build: whitespace cosmetics * 03http://tinyurl.com/9puetax03 [16:00] ffmpeg.git 03Mans Rullgard 07effe443877c5: build: do not use LIB as variable name * 03http://tinyurl.com/8f6f3kz03 [16:00] ffmpeg.git 03Janne Grunau 071a2c7880aa1c: averror: make error values proper negative values * 03http://tinyurl.com/8hz8qy703 [16:00] ffmpeg.git 03Mans Rullgard 0725dc79bc1433: sh4: add required #include, fix build * 03http://tinyurl.com/8d8q99c03 [16:00] ffmpeg.git 03Luca Barbato 072d6caade2233: dsputil: split out mlp dsp function * 03http://tinyurl.com/9an7mkq03 [16:00] ffmpeg.git 03Luca Barbato 071ec629308652: mlpdsp: adding missing file * 03http://tinyurl.com/94n5pjx03 [16:00] ffmpeg.git 03Michael Niedermayer 07bb3586475919: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/97ebva303 [16:07] durandal11707, if you want an account i can create one, you can then see all the bugs coverity found in ffmpeg once the account is approved by them, i dont know how long that takes [16:09] michaelni: yes i want to kill bugs [16:12] durandal11707, account requested [16:15] Tjoppen, mateo`, ctx leaks in mxf_read_local_tags() in the error cases (CID733800) [16:16] do you 2 want accounts ? [16:18] michaelni: thx for adding me :) [16:19] ubitux, did you already get the account or are you still waiting ? [16:19] Action: michaelni us curious how long it takes [16:20] Is [16:21] michaelni: uh.. where? I don't see anything on the ml [16:22] Tjoppen, what do you want to see on the ml ? [16:23] michaelni: nope i'm still waiting [16:25] what error case are you talking about? [16:26] Tjoppen, mxf_read_local_tags allocates MXFMetadataSet *ctx [16:26] its not freed on returns in case of some errors [16:27] and not stored anywhere where it could be freed later [16:28] oh yeah, you're right [16:28] it also breaks strict aliasing [16:30] oh, goody: [16:30] if (ctx_size && tag == 0x3C0A) [16:30] avio_read(pb, ctx->uid, 16); [16:30] so if I put a UID tag in the primer package it'll overwrite.. MXFPartition *partitions in MXFContext [16:32] mxf_metadata_read_table[] ought to be split up into two parts - one for reading directly into MXFContext and one for reading into an MXFMetadataSet [16:34] anyway, ctx should only be av_free()'d if ctx_size > 0 [16:35] heh, it does check ctx_size before reading UID. doi! [16:39] has anyone built static ffserver, ffplay and ffprobe? [16:39] for linux [16:48] many, why? [16:49] are there any docs about it? [16:49] what kind of problem are you having? [16:51] dlopen and dns-related libs wont compile statically [16:51] so I'm looking for some docs of the people that managed to build it [16:51] to see what I was doing wrong [16:53] yeah, Daemon404 probably understood as a static bin that doesn't have /everything/ from the OS side in it. Those system libs are always "fun" to hack to be "static" [16:53] aka "there be dragons" [16:54] Tjoppen, do you want a coverity account so you can see bugs like that mxfdec leak ? [16:55] sure [16:55] WHERE ARE MY DRAGONS!!! [16:55] ffmpeg.git 03Michael Niedermayer 077457da3698c6: drawtext: fix leak with timecodes * 03http://tinyurl.com/9ad2xor03 [16:55] ffmpeg.git 03Michael Niedermayer 079ba2484ece53: af_aresample: fix leak on alloc failure * 03http://tinyurl.com/95y7cqp03 [16:55] ffmpeg.git 03Michael Niedermayer 074b20b21b8dab: tiff: fix leak on error return in doubles2str() * 03http://tinyurl.com/93akwgr03 [16:56] burek, dlopen and statis are mutually exclusive [16:56] static* [16:56] and yes i have build an entirely static ffmpeg [16:57] built* [16:57] but not with networking crap [16:57] with kernel in it? [16:57] obviously not relevant [16:57] I have a new PGP key btw [16:57] but iwht libc and pals [16:57] with* [16:57] Daemon404, well I didn't ask about ffmpeg, read again :) [16:57] I should send in a patch to update MAINTAINERS I suppose [16:58] burek, statically linking network libs is a Bad Idea [16:58] end of story [16:58] dont do it. [16:58] Tjoppen, coverity account creation requested [16:59] k [17:07] framesha256 [17:16] ffmpeg.git 03Michael Niedermayer 07229ccce6cca7: libxvid_rc: fix leaks in ff_xvid_rate_control_init() * 03http://tinyurl.com/9p7hsvx03 [17:16] ffmpeg.git 03Michael Niedermayer 07c9454cb643f5: av_tempfile: fix leak in error case * 03http://tinyurl.com/9nrzk6e03 [17:19] this analyzer looks quite useful :) [17:44] how do I search for filenames on coverity? [17:45] you already have your account? [17:45] yes. I'm in there looking around. the interface is very JS heavy [17:46] maybe i should check my spam inbox... [17:48] ubitux: my is in spam!!! [17:48] there we go. this is rather neat [17:48] durandal_1707: ah indeed mine too [17:50] heh fun. [17:59] ah i'll fix the ebur128 derp. [17:59] Action: ubitux & [18:01] the warnings in the h264 templates are rather.. interesting [18:02] h264 encoder in c920 logitech webcam is ok, but is far from libx264.. libx264 could stream 640x360 @ 30fps at 512k without any artifacts, I don't know how :) [18:07] ffmpeg.git 03Michael Niedermayer 07104b1d9e103f: libvpxenc: fix memleak on error path * 03http://tinyurl.com/8skfd6r03 [18:07] ffmpeg.git 03Michael Niedermayer 0734bbab432ca0: jpeglsnec: fix memleak of state in error case * 03http://tinyurl.com/8vkgbo703 [18:07] ffmpeg.git 03Michael Niedermayer 07b96d1859d523: jpeglsenc: favor av_freep() for saftey over av_free() when a variable is still accessible afterwards * 03http://tinyurl.com/8nyzb7h03 [18:08] michaelni: truemotion ones looks like false positive [18:14] durandal_1707, if they are false positives, please mark them as such [18:19] hi philipl, do you want a coverity account ? [18:32] ffmpeg.git 03Michael Niedermayer 07adcbb3fd8b23: yuv2rgb: fix declared array sizes, so they match actuals. * 03http://tinyurl.com/8uym7vz03 [18:32] ffmpeg.git 03Michael Niedermayer 077fe554853152: random_seed: fix out of array read * 03http://tinyurl.com/9ydq9bg03 [18:33] ffmpeg.git 03Michael Niedermayer 0726474d1098eb: random_seed: fix digest size * 03http://tinyurl.com/8eny4t603 [18:39] ffmpeg.git 03Cl?ment BSsch 079ad1ea13e08f: lavfi/ebur128: fix typo in condition. * 03http://tinyurl.com/9meuw3703 [18:40] does "Fix submitted" means still pending for review? [18:40] or i'm supposed to select the fixed/tested/documented option? [18:42] also, what's the "Apply + Next"? [18:46] probably 'apply changes and move to next bug [18:47] saste: there are a lot of easy bugfixes all over lavfi and stuff spotted by coverity :p [19:22] saste: i got the ppc, but i can't figure out how to separate it from the existing lan. duh. [19:24] pull the cable [19:24] oops. wrong one. [20:07] ffmpeg.git 03Paul B Mahol 07313b40efbd63: bmp: unbreak non BMP_RGB compression for v4 and v5 * 03http://tinyurl.com/9lhq4y703 [20:29] michaelni: how to handle out of read in tgq_decode_block? it may cost too much [20:42] durandal_1707, the index from the scantable is 8bit thus if block is oversized a bit no overread can happen [20:43] michaelni: the coverity login worked fine [20:44] michaelni: the i++ happens more than once and it can be bigger than 64 inside loop [20:46] i is used as index in perm[] that is a array in a struct, after the array is another array of 64 elements [20:46] so i can be up to 127 [20:47] perm[] again is used as index into block[] but this index is just 8bit [20:47] hah [20:47] so if block has some spare elements at the end it looks safe from a quick look [20:54] the mpc8 one is false positive but code looks fishy [21:01] hi ffmpeg developers [21:01] anyone arround ? [21:08] opening rtmp stream sucsessfully but hanging on99,99 [21:09] llogan: HERE!:-P [21:10] llogan: sadly the bug gui of ffmpeg is not spoken using my screen reader [21:11] then describe your issue, including steps to duplicate if possible [21:12] llogan: issue: opening RTMP stream using ffmpeg or ffserver hang. [21:12] describing how to duplicate. [21:12] FYI, i used --enable-librtmp [21:27] llogan: log here: http://dpaste.com/812586/ [21:34] DelphiWorld: i don't have any ffserver experience, but ffplay plays the stream just fine if that means anything [21:35] llogan: is my config correct ? [21:35] llogan: do you have output using ffplay ? [21:35] no, but it tells you that the issue isn't with decoding [21:35] llogan: but ffplay have sound output ? [21:36] yes, it is a simple player [21:36] llogan: i dont have ffplay [21:37] maybe someone can help you on the ffserver-user mailing list. i can't offer much help with ffserver and rtmp junk. [21:37] i compiled ffmpeg from git. where's ffplay ? [21:38] you need sdl libs to get ffplay built [21:38] #define HAVE_SDL 1 [21:38] (config.h) [21:38] will be set to 1 automatically if you have the necessary libs [21:39] DelphiWorld: this isn't the user channel btw [21:39] ubitux: i know, i thought this is a ffmpeg bug [21:40] ubitux: can you see that junky bug? [21:40] sorry i didn't backlog enough [21:40] dunno for your bug [21:43] ffmpeg.git 03Paul B Mahol 073632f35c8e16: bethsoftvid: check return value of av_packet_new_side_data() * 03http://tinyurl.com/8r3tt5g03 [21:43] opening rtmp stream sucsessfully but hanging on99,99 [21:43] ubitux: log here: http://dpaste.com/812586/ [21:44] i saw, but i don't have time for this, sorry :p [21:44] ubitux: i give you time:-P [21:45] Action: ubitux smallows it all [21:46] ubitux: ;) [21:55] ubitux: work using a feed. but not using a file. so i am using feeds. [22:06] how would I be able to add a custom protocol? ffurl_register_protocol is considered "private"... [22:08] bryno: contribute it? [22:09] err.. no? [22:11] no? heh [22:12] if that's the answer, then why expose any of the register functions? :P [22:13] i'm just wondering what happened to that protocol layer [22:13] didn't you say it was private? [22:13] the other register functions like for registering input/output formats or codecs [22:20] bryno: it got removed long ago [22:20] michaelni: any comment on the metadata in AVPacket->priv? [22:21] if you want it back fill bug report (and it was removed by fork) [22:25] ubitux, i dont like using pkt.priv for metadata [22:25] it feels wrong [22:25] do you have any other suggestion? [22:26] what needs to pass metadata to what through avpackets ? [22:27] the lavfi device use its internal filtergraph [22:27] and it needs to get the metadata out of the buffer ref from its sink [22:27] and transmit them up to the AVFrame [22:27] since it produces AVPacket, we need a way of communicating through them [22:28] maybe it would make sense to use this flag *only* for this lavfi device [22:28] AV_PKT_FLAG_HACKLAVFI [22:28] or even FF_PKT_FLAG_HACKLAVFI [22:29] Action: durandal_1707 it ruins my day when i read HACK in code [22:30] i can call it FF_PKT_FLAG_ILY_DURANDAL if you prefer [22:30] but that's less accurate [22:30] heheh [22:30] echo "hello //HACK world" :P [22:30] if its just for lavfi i dont care so much but [22:30] michaelni: the other solution i tested was to store the pointer to the metadata in the side data, but that's even crappy IMO [22:31] (and it also needs to introduce a weird flag) [22:31] please no hacks [22:31] ptr to metadata im sidedata now sounds like a really really bad idea [22:31] durandal_1707: i don't see any proper solution for this problem [22:31] what about storing the metadata itself in sidedata ? [22:31] michaelni: i meant an allocated data with the pointer written in it, not abusing the pointer [22:32] consider someone (un)intentionally might stream copy lavfi into some container and then read it again [22:32] a pointer in side data would not make sense after that [22:33] priv would be lost (that may be ok or not) [22:33] i don't follow the stream copy thing [22:34] if you store AVPackets with a muxer side data will be stored too [22:34] .priv will not [22:34] new funny flags also likely will not [22:35] so a question is what we want to happen in this case [22:35] note that the priv is just pointing to a locally allocated metadata (in the lavd/lavfi device) and will be automatically deteletd from there [22:35] its existence is limited to one AVPacket scope [22:35] so av_dup_packet() will crash later ? [22:35] when you demux another packet, the old metadata is destroyed [22:36] let me check [22:36] this sounds like not only a hack not actually not working at all :) [22:37] av_dup_packet doesn't seem to care about the priv pointer [22:37] well either it looses it (=NULL afterwards) or it preserves the pointer [22:37] looks like it will loose it [22:38] which doesn't really matter (from a leakage or crash point of view) [22:38] afaict [22:38] i dont see point in abusing metadata and using tags for this, just add another struct [22:38] so lavfi will not work if an application callas dup_packet [22:38] it's not that it won't work, it's just that the metadata would be lost if you don't copy it [22:39] yes [22:39] what does lavfi have to do with packets, doesnt is use decoded data? :d [22:39] michaelni: do you see a better solution? [22:39] durandal_1707: can you be more specific? [22:40] ubitux: write new API to handle such stuff [22:40] ubitux, what about putting the metadata in side data itself ? [22:40] michaelni: right, storing key\0value\0 per metadata? [22:40] yes [22:40] iirc it was causing problems for mem leak [22:41] side data will also be copied correctly in dup_packet [22:41] like you needed to add av_free in the frame destruct callback but i think it was a problem if the frame->metadata was a pointer to a metadata somewhere else [22:41] (and not a dup like in this case) [22:42] also such side data could be used by a demuxer returning per frame metadata [22:42] durandal_1707: that's vague, please be more specific [22:43] michaelni: don't you think it's problematic to arbitrary dict free in the frame destruct callback the AVFrame->metadata ? [22:43] since AVFrame->metadata can be pointing to a shared dictionary [22:43] iirc that was the main issue i had with that solution [22:44] i dont think i fully understand the problem but shared dictionaries sound like a fragile thing [22:45] ubitux: do not use AVFrame.metadata [22:46] you're repeating yourself now [22:46] i don't think that's what he meant [22:46] ohsix: i didn't understood it as is :p [22:46] durandal_1707: right ok but well.. [22:46] what are you proposing instead? [22:47] i cant tell you what is right thing to do, but i know what is not right thing [22:47] from a user PoV, using the AVFrame->metadata is pretty nice [22:48] michaelni: i was thinking of lavc/tiff.c, where the internal context has an AVFrame->metadata stored & destroyed regularly [22:49] and i was wondering if it wouldn't cause problem to have the output AVFrame->metadata destroyed outside this codec [22:49] which will be required if we construct a new AVDictionary in the AVFrame in case of meta in side data [22:49] ubitux: i you like AVDictionary so much (I really see no point in that) add another AVDictionary to AVFrame [22:50] sure, we can do that [22:50] along with key/value in side data [22:50] would you guys prefer that? [22:51] ubitux, why do you want to destroy it outside the codec ? [22:51] but how that will work in future? would it be better/faster than vapoursynth? [22:51] michaelni: the side_data to avframe->metadata will need to be constructed outside any codec [22:51] so it will be needed to destroy it in the destruct avframe function [22:52] now i dont understand you [22:52] anyway, we can indeed add a AVFrame->filters_metadata and use the AVPacket->side_data with key\0value\0 [22:52] michaelni: ok just a sec, will quote code [22:53] durandal_1707: i guess optimizing AVDictionnary would make sense :) [22:53] lavfi device would generate avpackets with side data but this would not go to tiff [22:53] because lavfi device returns no tiff i guess but raw [22:53] ubitux: i doubt that have any future [22:54] michaelni: that's right, but when you decode the (raw) video/audio, you'll need to construct an AVDictionary from the side data [22:54] when would you destroy it? [22:54] same as other AVFrame fields [22:55] it's about the lifecycle of something that isn't strictly frame data, but needs to be accessed with a frame, right? [22:55] avcodec_free_frame()? or otherwise the dict would need to be the avctx, right? [22:55] and the problem is in case we would want to destroy it in avcodec_free_frame() [22:56] (because the dict would be destroyed there and not in the tiff for example) [22:56] ohsix: kind of yes [22:57] durandal_1707: you doubt about what exactly? [22:57] AVFrame fields are never destroyed in avcodec_free_frame() [22:57] aren't extended data destroyed there [22:57] ? [22:57] yes, thats the single exception [22:58] and its a bad idea [22:58] i agree [22:58] that's why i wanted to avoid putting the metadata free there [22:58] but where would i put it then? [22:58] since it can't be stored in any specific decoder [22:59] where the decoder frees/reuses other fields [22:59] we are dealing with raw audio & video [22:59] (and it could be another decoder?) [23:00] it is likely the dictionnary construction will remain in the common a/v decode functions [23:00] i guess other decoders could contain metadata in their bitstream [23:00] then the metadata will need to be stored somewhere in the avctx or something [23:00] could be in the private context of the raw decoder [23:01] is there a raw audio decoder? [23:02] Action: michaelni thinks: libavcodec/pcm.c [23:02] will need to introduce a new context there mmh okay.. [23:04] you could use AVCodecInternal too if that simpler [23:05] ok so we add a local avdict metadata in raw a/v decoders, and we use them exactly like the tiff decoder except that we will put a new constructed dict each time from the metadata side data [23:05] idk the AVCodecInternal, 'will look [23:05] won't that be a bit heavy processing btw? [23:05] not that's a real issue yet but.. [23:06] the build of AVDict ? [23:07] per frame [23:07] yes [23:07] re-allocating each entry and stuff [23:07] as well as constructing the side data from the buffer ref av dict [23:08] its ugly but iam not sure what alternative there would be [23:08] no copy would mean reference counting and mutexes probably [23:09] thats except that AVDict itself is poorly implemented [23:09] i'm ok with that, i'm just wondering how it would be possible to improve it later [23:10] btw, another related question [23:10] should we introduce a new dedicated metadata dict in AVFrame? [23:10] like filters_metadata or something [23:10] so we don't need the key prefix thing [23:11] this is a good question [23:12] in what use cases would we want to treat them differently, in what the same ? [23:13] i might want to run a advertisement detect filter and store the found positions with the frames in a file for example [23:13] like other metadata [23:14] the only benefits are avoiding naming conflicts with other metadata, as well as an arbritrary "hacky" prefix to workaround that first problem [23:14] i don't see much more [23:16] btw that reminds me that the nut spec says something about X- prefixes for not in the spec things [23:16] we forgot following thatg ... [23:16] noone noticed and iam not sure it would really be better with half the fields having X- prefixes [23:18] metadata mime types :> [23:19] full address and phone number of the person adding the tag, so you can ask what he meant ;) [23:24] j-b, do you know maybe what is the analogy in vlc for ffmpeg's -analyzeduration ? (i have a h264 stream which has a gop of 20-30 seconds so it never gets recognized correctly) [00:00] --- Fri Oct 12 2012 From burek021 at gmail.com Fri Oct 12 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Fri, 12 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121011 Message-ID: <20121012000501.E1B3C18A018E@apolo.teamnet.rs> [00:13] I am using ffmpeg and dvdauthor to make some video disks. Im having a problem with the final dvd's aspect ratio. The first video, a short 5 minute cartoon is 4:3, however the main feature is 16:9. When I play the dvd, the first clip plays fine, but the main feature is stretched to fill the entire 4:3 space. [00:21] Action: pfifo better pick up a dvd-rw a wal-mart [00:30] i have video source @ 640x480, with ffmpeg cropping borders, the size is 624x464. I want to transfer it to a dvd. should these dimensions cause issues? [00:31] will i need to scale it up to 704x480? [00:34] i'm trying to figure out how to write a vfw codec [00:34] do you ffmpeg devs think that the vfw portion of the ffmpeg source is a good example? [00:35] lake, yes it will create issues [00:36] pfifo: can i scale it to 640x480 and it be okay? [00:36] lake, im not sure, theres a list of resloutions that are allowed, you have to crop/scale/letterbox to what you need [00:39] i think 640x480 is svcd [00:46] pfifo: thanks. i'll try it at 704x480 [00:49] Hi... [00:49] how are you? [00:49] I have a problem with ffmpeg [00:50] I want to add head and tail to a mov file [00:50] how do I compile ffmpeg into a dll file? [00:50] I use "-ss -00:01:0" and this add a Head (repeat first frame) [00:51] How I do this to append frames after the last frame? [01:14] How can I append frames before video?... These frame are a copy of last frame.. [01:14] Thanks... [01:17] efs071: see the "concat" protocol in ffmpeg and http://ffmpeg.org/faq.html#How-can-I-join-video-files_003f [01:18] whiplash, why? [01:18] what's the point? [01:22] Ok... Thanks... but is imposible without concat with other video. I can add copy of first frame with "ffmpeg -i movie.mov -ss -00:01:0 output.mov" [01:24] efs071 you're better off using some video editor or something [01:25] I need a ffmpeg command [01:26] burek: so it can be used as a dshow codec [01:26] and just as an exercise [01:27] whiplash, did you check zeranoe's builds? [01:27] yes, that's where i got the source from [01:28] If I have a video movie1.mov and I create a movie2.mov with last frame of movie1. How can I create movie2.mov with exact codec and params to concat it?. And Concat only work with mpeg videos? [01:28] i'm using the october 9th revision [01:28] f3f35f7 and i'm building under cygwin for i686 on a windows 7 x64 box [01:29] efs071, what's the purpose for such thing? [01:29] this is what i invoke .configure with: ./configure --target-os=mingw32 --extra-cflags=-mno-cygwin --extra-libs=-mno-cygwin --enable-cross-compile --cross-prefix=i686-pc-mingw32- --arch=i686 --disable-doc [01:29] as far as i can tell, this page: http://ffmpeg.org/platform.html#toc-Crosscompilation-for-Windows-under-Cygwin is wrong [01:30] --extra-cflags=-mno-cygwin isn't part of cygwin's gcc anymore [01:31] I need expand video with head and tail. The can create hear (repeat first frame) with: "ffmpeg -i input.mov -ss 00:01:0 output.mov". But I known how append tail to video (repeat last frame for example 1 minute) [01:31] hmm, i never noticed these external libraries on zeranoe's page before, maybe there's some that cygwin ports doesn't cover [01:33] efs071: if you want to use ffmpeg you will have to create a cat-able movie of the frame with the duration you want it to appear. [01:33] there is no "frame hold" option as far as i know [01:34] as in: ffmpeg -loop 1 -r ntsc -i input.png -t 5 -c:v mpeg2video -q:v 2 output.mpg [01:34] or similar [01:34] then continue as shown in the faq [01:37] whiplash: if you believe it is wrong then consider submitting a bug report (although I don't know shit about cygwin). [01:39] Ok [01:40] But concat only work with mpeg? [01:40] i don't know. [01:41] And how can I create a video from frames but with other video format? [01:41] Becouse I need two videos with equal format to concat [01:42] ffmpeg -i input.mp4 -vframes 1 -ss output.foo [01:42] efs071, also take a look at http://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20concatenate%20(join%2C%20merge)%20media%20files [01:42] Thanks a lot! [01:43] i still haven't figured out what the deal is with -q:v 1 with mpeg*. carl always says to avoid it, but doesn't 'splain. [01:43] well ok, but how about building ffmpeg as a dll? [01:44] do i need to install mingw? [01:44] whiplash: maybe Zeranoe will know. [02:17] whiplash: What is your question:? [02:18] compiling ffmpeg as a dll [02:19] the whole thing as one .dll? [02:19] whiplash: Why not just use each dll like avcodec-54.dll for example [02:20] well i gotta admit i only kinda know what i'm doing here [02:20] but i want to use ffmpeg as a dshow codec [02:20] so don't i need an ffmpeg.dll to use regsvr32 on? [02:20] Try downloading one of my builds, a shared one, and use the .dll files in the /bin dir [02:22] hmm... which one should i be registering? all of them? [02:22] wait, let me rephrase that [02:22] which should i point fdshow-tryouts to? [02:23] whiplash: IM sent. [05:25] greets. I'm having a little difficulty here. I have a file containing raw speex packets, and I need to transcode them to pcm_s16le. I'm doing `ffmpeg -acodec libspeex -ar 16000 -ac 1 -i raw_input_file -acodec pcm_s16le -ac 1 -ar 16000 -fmt wav outputfile.wav` and it complains "rawinputfile: Invalid data found when processing input" [05:26] so I think it's looking for a speex container on the stream, which doesn't exist. [07:42] i have a bunch of mpeg4 videos i want to combine, at the bottom of the man page it says: ffmpeg -i test1.avi -i test2.avi -vcodec copy -acodec copy -vcodec copy -acodec copy test12.avi -newvideo -newaudio [07:42] first question - do i need ot have the -vcodec and -acodec in there twice? [07:43] second question - i want to input a lot of files, but my syntax doesn't seem to be working (i believe it's saying no input) http://pastebin.com/qM29TAkh [07:47] ffmpeg 0.6? oO you are 6 release far away you know? :) [07:47] we are in ffmpeg 1.0 currently (0.7, 0.8, 0.9, 0.10, 0.11) [07:48] i don't remember how -new* behave, and they are removed now [07:48] "combining" videos is quite vague btw [07:49] it can mean a lot of different things [07:55] it's what centos/rpmforge have in their repos... [07:55] i mean concatinate, one video after the other (they're in dates so i want to show 20121001.mp4 then 20121002.mp4 etc) [08:01] you can use the -vf concat added in the latest version [08:01] or you could use an intermediate concatenable container like mpeg [08:01] ok, i'll just get the new version then [08:01] thanks =) [10:53] hi all, I've a simple question: ffmpeg library is under LGPL v 2.1 or LGPL v 3? [10:53] Hello, I have tried to insall a ffmpeg on an ubuntu server and follewed the instruction from https://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide. [10:54] danisan: read LICENSE [10:54] but 2.1 pretty much [10:58] checking the compilation I see that the ffmpeg still uses the "old" libvpx 0.9.5 - instead of the new installed libvpx 1.1.0. This is because the libvpx-dev is not created and the compiler uses the old version (configure failes) [11:03] danisan: it depends on what pieces you compile in [11:04] danisan: most of it is LGPL, some libraries and outside encoders are GPL, some are even non-free [11:04] danisan: you get library license displayed at the end of "./configure" call [11:04] default compilation is LGPL though [11:16] Marwik: the ./configure reports: "LGPL version 2.1 or later", So the question is what means the "or later"? [11:26] hey, what would the easiest way to generate a video with a timer on it be, i am using Subtitle() with avisynth tehen rendering it with ffmpeg at the moment [11:27] but i just wrote a python script to generate a billion Subtitle() lines in an avs [11:27] seems kinda ghetto [11:36] ubitux: https://github.com/divVerent/ffstuff you may find this interesting... obviously lacks documentation, though [11:36] basically, shell based filter graph editing... nasty, but works [11:36] heh sorry yesterday i didn't have time to look at the subtitles thing, certainly for this week end [11:36] sure [11:36] i'll have a look :) [11:37] retardant: maybe you can use the timecode feature in drawtext [11:37] but that's a smpte timecode, so you have some frame rate constraints and such [11:38] retardant: you can also use the current time in drawtext (aka not pts) [11:38] ok [11:38] and you can also generated a subtitle stream yourself and burn it somehow [11:39] a pts2time could be nice in the filter [11:39] inb4: making a pts.srt file that contains 120 minutes of srt elements, one per millisecond ;) [11:39] with timecodes as text [11:40] actually... it wouldn't be THAT insane to have such a thing... would probably be a fun way to see bugs in players [11:47] mplayer has some... ISSUES dealing with such an insane srt file [11:50] i've been using this movie i made to find bugs in my rendering code [11:50] since there is some time bug [11:58] i should dig up that [11:58] MOST AWESOME ASS FILE OF ALL [11:58] i used to have it lying around atleast [11:58] a .ass for rendering ultra fancy karaoke effects and other stuff onto a video [11:59] would actually render with libass, just not anywhere near realtime. [12:04] theholyduck: I think I have that file (or such a file) somewhere [12:04] it does a dissolve effect by rendering every character 9001 times with different clip rectangles, and moving the parts around [12:09] divVerent, something like that yeah [12:17] how do i tell if a bit of h264 is constant or variable [12:17] Duration: 00:00:15.04, start: 0.000000, bitrate: 5899 kb/s [12:17] Stream #0.0(eng): Video: h264 (Baseline) (avc1 / 0x31637661), yuv420p, 1280x720 [SAR 1:1 DAR 16: [12:17] 9], 5721 kb/s, 30 fps, 30 tbr, 30k tbn, 60 tbc [12:17] i think my ffprobe is probably 900 years old [12:17] there is not really such a thing as CBR H.264... or rather [12:18] there is, but nobody uses it, what people tend to use where CBR is required is some buffering algorithm (keyword: VBV) [12:18] which is locally VBR, but has a "globally constrained" bitrate [12:18] and not sure if there are tools to detect if a file fulfills that [12:19] I think you can get complete CBR, if the frame size is set to something static. then there's the VBV-constrained VBR with random padding that makes it CBR. [12:19] sure you can :P [12:19] trying to hunt down a bug in my app where sometimes a user sets an in and out point to trim the video and it's miscalculating it by up to 45 seconds sometimes [12:19] heh which is a ridiculous margin of error [12:19] but nobody really does, because it's a quality loss and most H.264-using playback devices (and standards) have published buffering requirements [12:20] yeah [12:20] except for some dumb broadcast places IIRC [12:20] where you have to have nal-hrd + CBR padding [12:20] sounds stupid [12:20] this is just for kids at a museum [12:20] then you have some very limited encoder/decoder components used for web video etc. that use locked frame sizes [12:20] and yes, both of these use cases are mostly dumb by now [12:21] retardant: you are cutting by file sizes? [12:21] retardant: why is that an issue in your app? do you estimate the position of a frame by just calculating file length * pos / total time length? [12:21] that can go wrong quite a lot, even with the typical VBV-using H.264 encoding mode [12:22] simply because having a temporarily lower bitrate is always allowed in the VBV model [12:22] am using avisynth to put 2 prerendered bits of h264 onto the start and the end of some h264 coming out of a camera and then overlaying some png images over their video to give it a kind of fun effect (so to speak) [12:22] but they get to select what they want to keep from waht they recorded [12:22] note how x264 help screen doesn't even show in the examples how you would actually do CBR ;) [12:23] and rendering it all with ffmpeg [12:23] wondering if there is a better plan othre than avisynth but in my research it looked the best idea [12:24] where does bitrate come into the equation there? [12:24] sorry i didn't originally clarify i was interested in VFR [12:25] BTW, even CBR MP3 is typically not exactly CBR... silent frames are typically encoded a lot shorter even in CBR mode [12:25] but you guys were saying something interesting so [12:25] oh, frame, not bit rate... [12:25] now I see why it irritates avisynth - it has no timecode support ;) [12:25] with many formats, e.g. mp4, it's absolutely not easy to verify that a file is really VFR [12:26] because it only stores a denominator for the time base in the header [12:26] so if you record 23.98fps NTSC, you actually have a 24000 as denominator, and the timecodes go up by 1001 each frame [12:26] so to properly check if a given MP4 file is really CFR, you have to actually read every single frame's timecode [12:26] to see if it is a constant amount after the previous one [12:27] but, fear not, there is a solution [12:27] you can extract timecodes to a file, then work with avisynth, and when calling e.g. x264 to encode again, you can make it import the saved timecodes back [12:27] and possibly you will have to process the timecodes file by a script to e.g. put your stuff you added to the beginning [12:28] --tcfile-in and --tcfile-out are the x264 options that can help you there [12:28] ok cool [12:28] and also IIRC there is some matching avisynth support for this [12:29] the file is a text file, so you should be able to do all necessary processing from your program [12:29] yeah i saw a plugin that generates a .ffindex file that i think does that [12:29] is there a notation to get ffmpeg to work like ffprobe [12:29] but the thing is, you really have to assume that your input is VFR, you can't really limit it to CFR - VFR input is way too common already [12:30] e.g. when using inverse telecine filters, but scenecuts happen very often or some parts aren't even telecined [12:30] yeah the camera spits out vfr [12:30] avisynth has a cry [12:30] hehe [12:30] why would a CAMERA output vfr... ;) [12:30] i think it might be the source of my cropping problem [12:30] well avisynth occasionally falls over saying the framerates don't match [12:30] but nothing in terms of the software is changing [12:30] weird [12:31] BTW, I wonder if you can do what you want with ffmpeg directly [12:31] there is quite a bunch of filters, you can also of course overlay the image with a png image too [12:31] i need to overlay graphics at certian time intervals and bookened the video with 2 bits of prerendered video and of course crop the video, which is straight forward [12:31] and ffmpeg is properly VFR aware (tends to need -vsync vfr, though) [12:31] ok [12:32] not sure if you can do the overlay easily, but I'd bet you can somehow [12:32] i'd rather reduce the moving parts, also then i could move away from windows for the host [12:32] there is a filter that can - based on timecode - accept and reject frames [12:32] this MAY help, but not sure [12:32] seems to be a pretty sparse programming field... doing video stuff [12:34] BTW, I wonder if what you are overlaying is something like possible with DVD subtitles [12:34] then you could just use that format, and use "spumux" to adjust the start/end times of the overlays with it easily [12:35] but even if not, I wonder if a short RGBA rawvideo you can quickly encode from a few empty frames and your overlay frames, with custom time codes, would do [12:36] i was looking at direct show stuff [12:36] because i already use it to talk to the camera [12:36] since ffmpeg couldn't get the right pin for the 264 stream [12:36] but of course, the other way to do it would be avisynth + timecode files [12:42] http://i.imgur.com/o1upg.png <- This is a close-up of a black-and-white video clip which appears to have weird greenish spots. It's extremely annoying and I think it must have something to do with how it was digitzed. [12:42] Because the original is completely black and white. [12:46] gah all my tests are to the frame accurate [12:47] i need to log more data i guess [12:57] hi folks, is it possible to create 60fps video in h264 [12:57] it somehow creates the video in 25 fps even though i enter -r 60 [12:58] ffmpeg -i %04d.png -i ../foo/bar.mp3 -c:v libx264 -preset veryslow -crf 18 -r 60 -c:a copy baz.mp4 [12:59] this is the commnad i used, pardon me for pasting here if that was wrong, i pasted because it was nearly a one-liner [12:59] so this somehow outputs 25fps video, any ideas? [13:04] is it possible that 1600x1200 res and 60fps was a bad idea? :) [13:15] anybody there [13:23] kenanb: i think you need to specify rate for input and not for output [13:24] img2 demuxer default framerate is 25 [13:25] durandal11707: wups, i didn't even know such difference exists, so how should i do that [13:25] put -framerate 60 as first arguments and see if it helps [13:26] will do so, thank you so much durandal11707 [13:27] http://i.imgur.com/o1upg.png <- This is a close-up of a black-and-white video clip which appears to have weird greenish spots. It's extremely annoying and I think it must have something to do with how it was digitzed. Because the original is completely black and white. [13:27] Any idea what's causing this? [13:29] cheeseduck: you have cmd line that produces this? [13:30] durandal11707: No... [13:30] I didn't even make the video. [13:30] Just thought that people in here have video knowledge. [13:45] cheeseduck, 16 bit RGB 565 is sometimes misinterpreted [13:46] and results in higher than average green-ness [13:47] Higher than average? [13:47] There is supposed to be ZERO greenness. [13:50] average, as in... a 5bit-level of blue, does that correspond with the same VALUE of 5-bit green (which is not full green) in a rgb565 encoding.. or is it supposed to correspond to the same meaning of full, thereby being 6bits of green. [13:50] then there is the issue is colour temperature being part of the weighting. [13:53] Ugh. [13:54] I wish there were a quick button to switch between black-and-white and colour. [13:54] cheeseduck, you've used a colour sensor to take this image? what colour standard does it adhere too? [13:54] It is very distracting when it's got "some" colours. [13:54] zap0: I took a screenshot with my media player. [13:54] what colour standard does adhere too? [13:55] No idea. [13:55] Media Player Classic Home Cinema. [13:55] what format is the recording in? is it a MP4? [13:55] h264? [13:55] what codec? [13:56] Video: Xvid 720x560 25fps 886kbps [Video 0] [13:56] Audio: MP3 48000Hz stereo 128kbps [Audio 1] [13:57] ok, so the underlying colour format is likely: yuv 4:2:0 which means it is rather trivial for a filter to generate a black & white image from [13:57] you should be able to find a media player with a "black & white" filter/fx. [14:01] i have 250 pngs, i want to use the images 2,4,6 (so with framestep 2)how should i define a framestep in ffmpeg for the input? [14:10] ah, i guess it is -vf "framestep 2" [14:19] hmm, the output is bigger than 60fps version in size [14:19] that is odd [14:22] kenanb: so you only want a video with 125 frames? [14:23] relaxed: exactly [14:23] ffmpeg by default reads at 25fps and outputs 25fps [14:24] so that would give you 5 seconds of video- what does your output look like? [14:25] relaxed: i rendered images to be 60 fps, and there are 250 images in total, but 60 fps seems to high for playback of this resolution, so i want to make it 30 fps and use 1 frame for every 2 images rendered [14:26] relaxed: the result seems as i expected (including the delay caused by my cpu's lack of rendering 1600x1200 at 60 fps :D) but the file size is somehow bigger than 60fps version [14:27] do you really need the output framesize that large? [14:28] stupid client expectations :) [14:29] okay, just checking :) [14:29] some people see pie in the sky and have to chase it [14:29] they really need to see the delay on their own machines to realize they can't use that framesize indeed [14:29] is there any way to use ffmpeg filter chains with ffplay? [14:30] e.g. to render DVD subtitles [14:30] divVerent: I think so, ubitux would know. [14:30] relaxed: so is the -vf "framestep 2" option right? [14:30] I noticed that -vf "[0:0] [0:3] overlay" does not work [14:31] kenanb: yes, I believe so. [14:32] kenanb: though you may have to script this. [14:32] because it seems right when i check video properties they seem ok, then i look at file sizes, two files are encoded with the same "-i %04d.png -i bar.mp3 -c:v libx264 -preset veryslow -crf 18 -c:a copy baz.mp4", but 60fps one is 4 mb while 30 fps one is 5mb [14:33] which seems strange to me in encoding using CRF [14:33] but i am an ueber-newbie in encoding, my logic is probably faulty [14:35] is there a lot of motion in the video? [14:35] yes [14:36] it then may simply detect the motion badly at the lower fps [14:36] ah, do you think because of the constant motion the 60 fps one is encoded better? [14:36] i see [14:36] this may happen, yes [14:36] I know -crf behaves differently based on the framerate [14:37] that too [14:37] I never knew encoding is such a deep field [14:37] respect! [14:37] the other thing is, the higher fps are, the less "important" a single frame's content is [14:37] which crf may also take into account (not sure if it does, but it'd make very much sense) [14:38] this basically tells you: you can't say the videos have same quality just because you used the same crf value [14:38] compare them visually [14:38] and compare THAT to the file size ;) [14:38] anyway, it doesn't really matter because 1600x1200 doesn't playback smoothly even in 30 fps, so I'll just let them see the 60 fps version and hate it [14:39] :D [14:39] the thing is, crf is basically black magic [14:40] it is an attempt to define the "requested" visual quality to the encoder, and it does "its best" to get that done the best way possible [14:40] it is not a constant quantizer, but does all sorts of stuff to improve efficiency [14:41] quality still may vary between different videos, but the general intention is to make it roughly the same for all, whatever "perceived quality" is ;) [14:41] < divVerent> is there any way to use ffmpeg filter chains with ffplay? // ffplay -f lavfi -i [14:41] i see, i guess professionals go the two-pass route to achieve what they want in the optimum file sizes and quality, right? [14:41] ubitux: okay... how do I get a given stream from an input file [14:41] kenanb: depends ;) [14:41] divVerent: you can't use the filter complex thing, so movie and amovie source filters [14:41] crf is quite comparable to the two-pass route regarding bitrate/quality ratio [14:41] but 2pass allows you to actually set the target size in advance [14:42] ah, I see now, movie takes a si= stream index [14:43] or not, as stream_index is deprecated [14:43] it's a bit tricky sometimes but it's quite useful [14:43] so what is one supposed to use now? [14:43] mmh? i don't think si= is deprecated [14:44] "Specifies the index of the video stream to read. If the value is -1, the best suited video stream will be automatically selected. Default value is "-1". Deprecated. If the filter is called "amovie", it will select audio instead of video." [14:44] says the manpage [14:44] ah it's because of libav i believe [14:44] they gave no alternative [14:44] don't worry about that. [14:44] haha, so libav can just deprecate a ffmpeg feature without replacement? [14:45] what kind of nonsense is that [14:45] and I mean, deprecate even in ffmpeg [14:45] it's "deprecated", or more correctly useless, in the conversion tool [14:45] avconv for them [14:45] AH, now I see, wait [14:45] there IS an alternative [14:45] you can use streams= too [14:45] which uses stream specifiers [14:45] huh? [14:45] so stream_index indeed is deprecated [14:46] oh ok [14:46] the streams specifiers [14:46] movie=foo.avi:streams=1+2 [14:46] would read streams 1 and 2 [14:46] right ok, i was wrongly bad mouthing then [14:46] my bad :) [14:46] hehe, you'll find many other good reasons ;) [14:47] the part I like least about libav are the borderline copyright violations of ffmpeg [14:48] http://blog.pkh.me/p/13-the-ffmpeg-libav-situation.html - basically, the taking over of commits (token-wise unchanged, only reformatting) while not attributing the original author [14:49] hi all, I have a QT (prores 422 HQ) with 1 'stream' or track of audio that has 14 discrete channels within it. Is there a one-line ffmpeg solution that will take the 14 channels, split them out and turn them into 14 discrete 'streams' or tracks, keeping the video stream as is? [14:55] crashd: look at channelsplit in the man page [14:55] divVerent: yeah, i wrote that article, i'm quite aware of the problem ;) [14:56] ah, that's you :P [14:57] Action: relaxed forks ubitux [14:57] I am waiting for them to do that on a somewhat bigger commit [14:57] which "likely" constitutes copyright [14:57] and then libav being mentioned on the ffmpeg hall of shame ;) [14:57] our HoS page is unmaintained :( [14:57] and down [14:58] we need more such pages... [14:59] piratebay legal threats got no updates either for a while, was quite entertaining while it lasted [14:59] relaxed: thanks, im not using 1.0 release so it wasn't coming up in any man pages. i'll get it upgrade. thanks again! [15:19] relaxed: does channel_layout support none-standard audio channels, eg: not 5.1,7.1, stereo etc..? [15:25] crashd: I'm not sure. Give it a whirl. [15:37] ah, looks like we can only use pre-defined 'channel_layouts' which only seem to go upto 8 total chans, would need to create a patched version to support some arbitrary 14 track layout. thanks anyway relaxed! [15:42] crashd: sox may be able to do it [16:40] burek: I have one more question about ffmpeg ---> ffserver <--- ffplay and is about ffm. Is there a way to use - for example - RTMPS over ffm? [16:40] just a sec [16:40] from ffmpeg to ffserver [16:40] sure [16:41] (now is working like a charm with pcm :) ) [18:20] any idea why android refuses to accept my video? i made it with: ffmpeg -r 12 -i %06d.png -sameq -vcodec libx264 -preset fast -crf 25 simulator.mp4 [18:21] they're about 1700 pngs made at 1024x768 [18:21] it works fine in vlc and dragonplayer(kde default media thingy) [18:30] amstan, why -sameq [18:31] btw, try with global headers [18:31] and qt-faststart tool [18:31] also read about -profile [18:31] it might help [18:44] I have interlaced video tapes from a Sony Digital8 camcorder. Scenes with panning show fringing. How cal I best deinterlace the videos and put them on DVD? [18:47] burek: -sameq was removed in git, so one day it will trouble us no more :) [18:47] thats the great news :) [18:48] of course we're talking years, haha [18:48] oh.. [18:48] :D [18:50] The interweb is full of examples using it, so keep your trigger finger ready. [18:51] Hmm, maybe it could be backported to the earlier branches. [18:52] that would be great if possible [18:52] but I believe that, as soon as we provide official deb packages somehow, people will just update [18:52] and that's it [18:53] which reminds me :) [18:53] ping ubitux :) [18:53] priv? :) [18:54] mmh? [18:54] can I ask you something in PM [18:54] just ask in pm... [18:54] :p [19:02] i'm trying to concatinate mpeg4 files but i'm not having much luck - http://pastebin.com/qJPvLSs5 [19:02] i also tried it with -f mpg4 which didn't change the outcome [19:03] hi [19:03] anyone works with mediaroom IPTV plataform? [19:03] this is possible to stream consume live from ffserver? [19:03] iam8up: use MP4Box's -cat [19:05] iam8up: MP4Box -cat 1.mp4 -cat 2.mp4 -new combined.mp4 [19:06] is there any way to make the input easier? i've got dozens of files [19:06] say -cat *.mp4 or accept from - with ls|sort -n [19:07] iam8up, you can't just cat *.mp4 [19:07] it will concatenate file headers too [19:07] and render it unusable [19:09] i mean, some players might recognize it and play it without problem [19:09] but others can fail, so.. [19:09] the correct way is to re-encode all your videos and produce one big output [19:09] ok i gotcha, just looking for a solution other than typing -cat 1.mp4 -cat 2.mp4 all the way up to 30 or 50 or whatever every month [19:10] man MP4Box :) [19:10] see if there is an option to provide a file list from a file [19:10] just doing svn now [19:10] need to get the packages, no binary for centos [19:28] there is not, and you can only have 20 -cat per command [19:33] relaxed, thanks a bunch for the tip [19:35] very odd result, it opens up 5 windows in VLC? [19:42] try ffplay [19:47] i don't have any machine with linux and a monitor [19:47] i used the win32 build of mp4box and it made the video just fine, something is wrong in my command [19:48] yeah, I've never heard of that happening. [20:45] hey everyone [20:45] Action: DelphiWorld yel at burek [20:45] please, how to build a Shoutcast compatible output using FFserver ?stream [20:45] what format should i chouse [21:00] hi Sast [21:00] can you look at some bug in ffm? [21:01] or michaelni [21:08] DelphiWorld: if you have a bug, search the bug reports. if it is not listed, then submit a bug report. [21:09] llogan: #ffmpeg-devel ;-) [21:17] hey, i just compiled rtmpdump v2.4 from git, but it still gives me "WARNING: HandShake: Type mismatch: client sent 6, server answered 9" [21:26] saschagehlich: without context it is hard for people to get an idea of what you're doing [21:27] llogan: okay [21:27] https://gist.github.com/3874907 this should be enough context? [21:27] meaning you need to show your commands or whatever so people can easily duplicate the issue [21:28] hm, i assumed you were using librtmp via ffmpeg. [21:28] this isn't really a support channel for rtmpdump, but someone here may know more [21:29] llogan: me or saschagehlich ? [21:29] is there an rtmpdump support channel? I thought it's developed by the ffmpeg team since it's hosted at git.ffmpeg.org [21:29] but okay, I'll try to compile the latest ffmpeg on my machine, give me a sec [21:30] what would be the respective ffmpeg command? [21:32] you can try the mailing list: https://lists.mplayerhq.hu/mailman/listinfo/rtmpdump [21:33] or this forum: http://stream-recorder.com/forum/rtmpdump-f54.html [21:45] okay, ffmpeg compilation seems to be going well until a certain point when it totally breaks https://raw.github.com/gist/4bd047b2bb68afce408c/028013307aba221c378e501a7c39e80e4e51e938/gistfile1.txt (scroll down) [21:46] saschagehlich: what was your ./configure? [21:47] llogan: https://gist.github.com/a250be7a3bb2866452c5 [21:48] this is ffmpeg from git? [21:49] yup [21:49] master branch [21:49] commit 313b40e [21:51] works for me. [21:52] any tips? [21:53] i mean it looks like gcc is totally broken here, it sees syntax errors where there are none [21:55] ppc? [21:56] nah intel [21:57] checked out v0.6.1 - works [21:57] duh. i forgot about your configure...it shows arch [21:58] sounds like a regression to me and should be reported [21:58] I'm checking out the 1.0 release& if that one works, I'm gonna report it immediately [21:59] is this os x? [22:00] yup [22:00] mountain lion [22:00] 10.8 [22:00] i'll have access to one of those later today. maybe i'll try to duplicate the issue. [22:02] alright [22:02] 1.0 release: same issue& trying 0.9 [22:04] DUUUUUUUUDES [22:04] i am happy now llogan [22:04] llogan: let me pb my new file [22:05] DelphiWorld: glad you figured it out. [22:05] llogan: lovely, let me share [22:05] also looks like burek made a ffserver guide: https://ffmpeg.org/trac/ffmpeg/wiki/Streaming%20media%20with%20ffserver [22:06] saschagehlich: a git-bisect would do better to pinpoint the commit then introduced the issue [22:06] *that [22:06] llogan: never tried git-bisect, but will do now [22:07] llogan: http://dpaste.de/yB0Gd/ [22:07] saschagehlich: tel me what you're trying to do [22:07] DelphiWorld: I'm trying to find a bug in ffmpeg that leads to compilation issues on os x mountain lion [22:08] saschagehlich: ah, out of my knoledges [22:08] :D been using git for 3 years now, but that's something I never touched [22:08] ubitux: http://dpaste.de/yB0Gd/ [22:11] llogan: thought ? [22:11] i know nothing of ffserver [22:16] llogan: just read and compare what did i change :-P [22:16] llogan: my issue soleved [22:24] omfg git bisect is awesome :O [22:31] and that's the point where commiting in small steps becomes awful... [00:00] --- Fri Oct 12 2012 From burek021 at gmail.com Sat Oct 13 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Sat, 13 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121012 Message-ID: <20121013000502.501A918A01DE@apolo.teamnet.rs> [00:26] ffmpeg.git 03Michael Niedermayer 0759cbedfc3df3: update_initial_timestamps: increase pts_buffer size * 03http://tinyurl.com/9tgnybs03 [00:26] ffmpeg.git 03Michael Niedermayer 071b5069aa41ce: sanm: fix off by 1 error in draw_glyph() * 03http://tinyurl.com/9v79elq03 [00:26] ffmpeg.git 03Michael Niedermayer 070c6b9b9fe5ed: rtspdec:read_line: fix use of uninitialized byte * 03http://tinyurl.com/9bgha6g03 [00:26] ffmpeg.git 03Michael Niedermayer 07e0909ff1584d: lavc: Fix use of uninitialized field. * 03http://tinyurl.com/9m3lszg03 [00:38] i'm trying to understand something in lavf [00:38] in the swf demuxer, when reading packets, an audio stream can be found [00:38] need_parsing is set to AVSTREAM_PARSE_FULL [00:39] then i guess the decoder is supposed to set the sample format on its own [00:39] it's for example the case for lavc/pcm [00:39] it appears that it doesn't seem to work as expected [00:39] ? "[swf @ 0x2b0a240] Could not find codec parameters for stream 0 (Audio: pcm_s16le, 5512 Hz, 1 channels, 88 kb/s): unspecified sample format" [00:39] any idea what could cause this? [00:48] does resample have support for 5512hz ? [00:49] or the swf demuxer i guess [00:49] is what that message is from [00:50] ah you think that's the issue? [00:53] ubitux, you might want to look at "[PATCH] Parse DEFINESOUND tags in swf" its maybe related [00:55] oh that's pretty recent. [00:57] oh awesome [00:57] that's exactly my problem [01:06] the patch works fine [01:32] what's the AV_PIX_FMT for rgb 15-bit? [01:32] (if any) [01:53] ubitux, wouldnt it be rgb555 [01:56] comment say 16 bits so dunno [02:18] rgb555 is stored in 16bits, 1 bit unused [02:21] ffmpeg.git 03Michael Niedermayer 075e689b65ce9c: dv: zero dsp before init, this fixes use of uninitialized dct_bits * 03http://tinyurl.com/8rdgfw603 [02:21] ffmpeg.git 03Michael Niedermayer 07e576105d841c: motion-test: zero dsp context * 03http://tinyurl.com/8o4kolg03 [02:21] ffmpeg.git 03Michael Niedermayer 075a75924dfd43: rtmpproto: fix out of array write * 03http://tinyurl.com/8zwshom03 [02:27] Action: Daemon404 pokes bcoudurier to OK his fate key [04:26] ffmpeg.git 03Michael Niedermayer 07927d866a9961: tscc2: fix out of array access * 03http://tinyurl.com/8ql5okd03 [04:26] ffmpeg.git 03Michael Niedermayer 0762722ae2d4b2: nellymoserenc: fix array element ordering * 03http://tinyurl.com/96gj8sv03 [04:29] Daemon404 : email him [04:30] uh [04:30] thats kinda what i did [04:30] >fate-admin email [04:30] as seen on fate.html [04:30] maybe it went to spam :P [04:30] no [04:30] michaelni confirmed reception [04:30] ok [04:30] but only bcoudurier can add the key. [06:19] sowy [06:36] arch name matches *arch* -> test failure :> [08:22] burek: nope [09:26] omg, i just found source of old bug [09:32] Action: ubitux is wondering why zlib's uncompress never works while a manual zstream with only inflate Z_FINISH with the same buffer sizes just works [09:32] Action: ubitux really hates this lib [09:33] ubitux: what is using zlib? [09:34] ffmpeg [09:34] do you have an alternative? :) [09:34] i mean where it is failing? [09:35] i'm writing some code to support zlib compressed raw images in swf [09:35] basically i've in and out len, i allocate the buffers, copy the data [09:35] uncompress() fails (-5) [09:35] one manual inflate just works without any problem [09:36] but manual inflat is 50 lines instead of 1 [09:39] -5 beans buffer is too small [09:39] s/beans/means [09:41] yes but i put a huge one and it didn't help [09:42] why you copy data? [09:42] how is image data stored in container? [09:43] i can avoid the copy, it's just temporary [09:44] the image is just a raw image zlib compressed [09:44] continous? [09:45] http://pastie.org/5039931 [09:46] durandal_1707: well there is a palette, and then the data, in my case [09:46] but it's not really the issue here :p [09:52] ubitux: you set out_len ? [09:53] yes [09:53] and using cast for out_len is bad idea [09:53] make out_len unsigned long [09:54] that was just copied from another code in ffmpeg, i should change it yes [09:54] 09:35:56 @ubitux | one manual inflate just works without any problem [09:54] 09:36:17 @ubitux | but manual inflat is 50 lines instead of 1 [09:54] ubitux: really? [09:55] it's shorter than that in the DarkPlaces engine ;) [09:55] ubitux: is see no cast in ffmpeg code whice use uncompress [10:03] libavformat/mov.c: if (uncompress (moov_data, (uLongf *) &moov_len, (const Bytef *)cmov_data, cmov_len) != Z_OK) [10:05] moov_len is long so it is less wrong but still wrong [10:06] but i do not think anything bad will happen if moov_len is negative [10:07] oh [10:07] that was the problem [10:07] thx durandal_1707 :) [10:08] see, casts are dangerous! [10:08] yes :( [10:09] they're more fun on big endian platforms ;) [10:49] [~/src/ff/ffmpeg]% grep -m 1 pal libavformat/avidec.c [10:49] uint32_t pal[256]; [10:49] [~/src/ff/ffmpeg]% grep 'memcpy.*pal' libavformat/avidec.c [10:49] memcpy(pal, ast->pal, AVPALETTE_SIZE); [10:49] [~/src/ff/ffmpeg]% git grep 'define.*AVPALETTE_SIZE' [10:49] libavutil/pixfmt.h:#define AVPALETTE_SIZE 1024 [10:49] is this really ok? [10:49] oh, forget it. [10:49] i'm stupid. [10:52] Action: ubitux hates himself [11:28] ffmpeg.git 03Paul B Mahol 0756519d7d14ac: takdec: s/bits_per_coded_sample/bits_per_raw_sample * 03http://tinyurl.com/8zjch4703 [11:28] ffmpeg.git 03Paul B Mahol 074dcf71aacca6: takdec: stop decoding in case of unknown bps * 03http://tinyurl.com/8k9vue503 [11:44] yepee, added locally support for 500+ swf samples. [11:46] ubitux: can you explain why av_x_if_null is a good idea to have especially as a function, not as a macro? [11:46] it silently loses type info... [11:46] calling function [11:46] so you can have a foo*f, and do bar *b = av_x_if_null(f, "argh"); [11:46] av_x(char *)av_x_if_null(av_get_sample_fmt_name(outlink->format), "?"), [11:47] and it'll compile without warning [11:47] yes, what I dislike here is the lack of any warning on an error [11:47] libavfilter/avfilter.c: src->name, srcpad, (char *)av_x_if_null(av_get_media_type_string(src->output_pads[srcpad].type), "?"), [11:47] a macro would solve this better, but unfortunately evaluate its arg twice [11:47] etc. [11:47] yes, that was the point [11:47] we could have both [11:47] wonder if it'd be wiser to have the function next to a gcc specific macro [11:47] (foo ?: bar) actually does the same as the function :P [11:48] so in GNUC it could use the ?: operator in a macro evaluating everything only once [11:48] just add the macro version :) [11:48] (generic one, not gnuc) [11:48] but that one evaluates twice, and is slow ;) [11:48] which is kind of the point of the function to avoid it [11:48] yes, but when it's not a function you can use the macro :p [11:49] I suppose what you REALLY wanted was template inline T av_x_if_null(T x, T y) { return x ? x : y; } [11:49] but C unfortunately lacks templates :P [11:49] "unfortunately" [11:49] :) [11:49] *fortunately [11:49] actually... didn't C11 solve this even? [11:50] no, not really, damn [11:50] or Perl's interpretation of the || and && operators would help here [11:50] i.e. not a "0 or 1" result, but "the last one evaluated is the result" [11:51] at least a C11 type-generic expression could be used to special case the function for void * and const void * [12:04] and a C89 hack actually allows comparing sizeof() of the types :P [12:08] bcoudurier ? [12:21] ubitux: some fun hackery ;) https://gist.github.com/3878566 [12:21] but, it doesn't verify the return type, and has the side effect of doing something REALLY bad if the warning occurred [12:21] also, lacks parentheses, but this hack isn't supposed to be used anyway [12:21] and is slow [12:25] ?&&&&&&&?&&&&&&&(void) foo[sizeof(x == y)], \ [12:25] this is actually better, somewhat... ;) (add * 0 after the sizeof) [12:25] https://gist.github.com/3878585 - better hack [12:25] sizeof serves to prevent the compare from actually being executed no matter how dumb the compiler is [13:33] divVerent: erk :) [13:35] divVerent: is there a relevant thread on the ml? [13:35] Tjoppen: no [13:35] I was just wondering about the possibly dangerous av_x_if_null function [13:35] dangerous, in the sense that type mistakes aren't noticed by the compiler [13:36] not as in really broken [13:36] but I can't find a good constructive solution, it looks like standard C allows nothing beter [13:36] but GNU C and C++ do [13:36] oh yeah.. why isn't that a macro? [13:36] because the first arg is usually a function call [13:37] so evaluating it twice would be bad [13:37] (bad as in unnecessarily slow, I've not seen it with a function with side effects) [13:37] stick it in a struct or something perhaps? [13:38] the probably sanest way to do this in C would be to extend the functions that currently may return NULL by a "return this if not found" arg [13:38] but that'd be a hell of a lot of API changes, as some of these functions are public [13:38] I was just about to suggest a macro that does a check before calling, which is what you pasted [13:39] yes, but it can't test the return value [13:39] unless said macro would also do the assigning [13:39] but that gets ugly then [13:39] ./libavfilter/graphdump.c: format = av_x_if_null(av_get_sample_fmt_name(link->format), "?"); [13:39] this is the typical use [13:40] sure, in this specific case evaluating twice is harmless (just ugly) [13:41] actually... it looks to me like currently ALL uses of this operate on const char * ;) [13:41] but if it is to become type specific, it'd get a longer name (like av_x_if_null_const_char) which is ugly too [13:42] can't there be a macro that defines a local function perhaps? [13:42] *static function [13:42] that also takes the type [13:42] but even then one would have to specify the type explicitly when calling [13:42] probably pointless, and still error prone [13:42] DECLARE_AV_X_IF_NULL(char) [13:42] sure, but then you'd need av_x_if_null_char() :P [13:42] not if you only use one version of it in each file :) [13:42] hehe [13:43] now that's fun [13:43] but... probably it should just stay as is [13:43] it was interesting to think about it, this thing IS harmful, but there isn't really a better solution in C [13:43] plus, this is a public API already [13:43] C already contains its own such harmful functions [13:43] e.g. strchr takes a const char * input string, but suddenly returns a non-const char * [13:44] durr [13:44] lack of function overloading led to this [13:45] I never asked for this [13:57] ffmpeg.git 03Paul B Mahol 07caa7e24eb1d4: truemotion2: remove unreachable code * 03http://tinyurl.com/9vbf4ey03 [14:19] ffmpeg.git 03Luca Barbato 0726b3fde6a78f: doc: update the faq entry about custom I/O * 03http://tinyurl.com/8cvmfn603 [14:19] ffmpeg.git 03Martin Storsj? 071093383d6cf7: random_seed: Support using CryptGenRandom on windows * 03http://tinyurl.com/9kml9s803 [14:19] ffmpeg.git 03Diego Biurrun 070a75d1da23b8: options_table: refs option is not snow-only * 03http://tinyurl.com/9cvwwgh03 [14:19] ffmpeg.git 03Michael Niedermayer 0743cce41267a7: Merge commit '0a75d1da23b8659ec49391469bb592da12760077' * 03http://tinyurl.com/99aoosd03 [14:33] ffmpeg.git 03Diego Biurrun 079734b8ba56d0: Move avutil tables only used in libavcodec to libavcodec. * 03http://tinyurl.com/9cafmd403 [14:33] ffmpeg.git 03Michael Niedermayer 07e33565837084: Merge commit '9734b8ba56d05e970c353dfd5baafa43fdb08024' * 03http://tinyurl.com/8djjnr803 [14:41] ffmpeg.git 03Janne Grunau 07e578f8f4680f: prepare 9_beta1 release * 03http://tinyurl.com/95qnzv503 [14:41] ffmpeg.git 03Mans Rullgard 0715ba7f6525c0: parseutils: fix const removal warning * 03http://tinyurl.com/9kqsbj403 [14:41] ffmpeg.git 03Michael Niedermayer 07d6135a886d85: Merge commit '15ba7f6525c0f56f0c8e3e3e0c0c5129de054f41' * 03http://tinyurl.com/9aa8go803 [14:43] michaelni_: is CID 700222 irrelvant? [14:54] how can I ask for a change in http://ffmpeg.org/download.html [14:54] durandal_1707, i dont know [14:55] burek, patch or pull request [14:55] i would like to add the 'id' to all those

sections, so that they can be easily accessed by http://ffmpeg.org/download.html#some_id [14:55] hm, let me see if I can create my first patch :D [14:56] (lots of swf improvements incoming tonight!) [14:57] burek, git patch against git://ffmpeg.org/ffmpeg-web is absolutely needed [14:57] dont send a patch of the .html [14:57] its not stored that way so we couldnzt apply [14:57] is there any url where I can see the source online? [14:57] is it on the ffmpeg gitweb already? [14:58] localhost after git clone :) [14:58] :) [14:58] ok :) [14:58] let me try [14:59] ffmpeg.git 03Mans Rullgard 0768e360a83c5b: parseutils-test: various cleanups * 03http://tinyurl.com/8uz3bzt03 [14:59] ffmpeg.git 03Mans Rullgard 07366b72f149a7: svq3: replace unsafe pointer casting with intreadwrite macros * 03http://tinyurl.com/9dkk8d203 [14:59] ffmpeg.git 03Mans Rullgard 07fb32f31af76c: svq3: fix pointer type warning * 03http://tinyurl.com/9lb8zoq03 [14:59] ffmpeg.git 03Martin Storsj? 079a92aea27bad: avutil: Add functions for allocating opaque contexts for algorithms * 03Error03 [14:59] ffmpeg.git 03Martin Storsj? 07e002e3291e6d: Use the new aes/md5/sha/tree allocation functions * 03http://tinyurl.com/8g65syl03 [14:59] ffmpeg.git 03Michael Niedermayer 07f391e405dfde: Merge commit 'e002e3291e6dc7953f843abf56fc14f08f238b21' * 03http://tinyurl.com/8ago2z703 [15:14] michaelni_ is this correct http://pastebin.com/fAX2Jcx2 :) [15:27] burek, its missing commit message and author [15:28] git commit -a -s and git format-patch -1 should help [15:30] do i really need to do al that :S [15:30] can you put yourself as author and random commit msg? :) [15:31] i dont really care about credits and stuff, i just need those ids :) [15:36] do i really need to do al that :S [15:36] burek: how that is useful? [15:37] ^patch [15:39] ffmpeg.git 03Martin Storsj? 07da18e918a4ec: md5: Allocate a normal private context for the opaque md5 context pointer * 03http://tinyurl.com/9fl4fyk03 [15:39] ffmpeg.git 03Mans Rullgard 07fdd666094d42: build: add support for Tru64 (OSF/1) * 03http://tinyurl.com/9ynbqjs03 [15:39] ffmpeg.git 03Anton Khirnov 07d2fcb356caf3: pixdesc: add functions for accessing pixel format descriptors. * 03http://tinyurl.com/8vnas2f03 [15:39] ffmpeg.git 03Anton Khirnov 07b7f1010c8fce: tools: do not use av_pix_fmt_descriptors directly. * 03http://tinyurl.com/8rj5djq03 [15:39] ffmpeg.git 03Michael Niedermayer 07a33ed6bc74b6: Merge commit 'b7f1010c8fce09096057528f7cd29589ea1ae7df' * 03http://tinyurl.com/8n9mtrv03 [15:43] ok, you're right [15:54] i get "FF_API_PIX_FMT is not defined, evaluates to 0" spam [16:09] ffmpeg.git 03Anton Khirnov 070a7068fa5dda: sws: do not use av_pix_fmt_descriptors directly. * 03http://tinyurl.com/9ay69hy03 [16:09] ffmpeg.git 03Anton Khirnov 079953ff3cd844: mpegvideo: fix indentation * 03http://tinyurl.com/97fa62403 [16:09] ffmpeg.git 03Michael Niedermayer 07a9bd51b1e647: Merge commit '9953ff3cd844eb5f6d8dfce98cad94b78a0fc7dc' * 03http://tinyurl.com/965skrm03 [16:27] ffmpeg.git 03Paul B Mahol 073d179edf6d2a: yop: check return value of avformat_new_stream() * 03http://tinyurl.com/8p4czqo03 [16:52] ffmpeg.git 03Anton Khirnov 0750ba57e0ce63: lavc: do not use av_pix_fmt_descriptors directly. * 03http://tinyurl.com/8lkgsea03 [16:52] ffmpeg.git 03Michael Niedermayer 07af7dd79a3230: Merge commit '50ba57e0ce63d9904269ea0728936a0c79f8bfb5' * 03http://tinyurl.com/9o6nfmz03 [17:04] michaelni: that av_pix_fmt in lavc thing is incomplete, is there other such stuff left to merge? [17:07] durandal_1707, you mean the FF_API_PIX_FMT warnings ? [17:07] no, av_pix_... which should now be static.... [17:08] please elaborate what & where [17:09] i think shared build is broken _now_ [17:12] durandal_1707, ok ill look into it once fate tells me what is missing, thx for the vague hints [17:12] michaelni: i can fix this right now [17:13] ffmpeg.git 03Anton Khirnov 0759ee9f78b0cc: lavfi: do not use av_pix_fmt_descriptors directly. * 03http://tinyurl.com/8rycd7h03 [17:13] ffmpeg.git 03Michael Niedermayer 0713afee951a49: Merge commit '59ee9f78b0cc4fb84ae606fa317d8102ad32a627' * 03http://tinyurl.com/8ztj7qa03 [17:14] durandal_1707, please do, and thanks [17:17] anyone an app to generate swf? (except ffmpeg) [17:17] i'd like to ask for a custom sample... :p [17:18] ubitux: Adobe Flash? [17:18] for example :p [17:18] probably that's not what you wanted to hear [17:18] well i don't want the app :p [17:18] I don't have it, I just know it exists ;) [17:18] i'd like someone to generate a sample for me ;) [17:18] ah, you just accidentally the verb [17:18] ah indeed [17:19] community/mtasc 20091229-4 [17:19] An open source flash (swf) compiler [17:19] wonder if that would help? wouldn't know what to do with it though [17:21] i'm looking for a swf with a "lossless bitmap" in rgb 15 [17:21] :p [17:21] i have 500+ samples with pal8, 30+ samples with rgb24 [17:21] but none with rgb15 :( [17:22] hm... I have a directory with a bunch of swf files, maybe one of them has it... let me see [17:24] ffmpeg.git 03Anton Khirnov 0722c8cbc0da8b: lavu,lavd: do not use av_pix_fmt_descriptors directly. * 03http://tinyurl.com/94s5n9703 [17:24] ffmpeg.git 03Michael Niedermayer 0727ccc82e1bb1: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/98ngbvj03 [17:25] damn, swfextract doesn't really show the type [17:25] it just says "so many PNGs" [17:26] oh wait... actually [17:26] [014] 17792 DEFINEBITSLOSSLESS defines id 0020 image 236x180 (8 bpp) [17:26] this looks greppable [17:27] no, none match, but I have a way to test [17:33] ffmpeg.git 03Paul B Mahol 07a51540d81167: lavc: do not use av_pix_fmt_descriptors directly * 03http://tinyurl.com/8kbzk2603 [17:35] swfc only has 16bit support, not 15 [17:40] ffmpeg.git 03Michael Bradshaw 07a41c824c539c: Parse DEFINESOUND tags in swf (fix ticket 1638) * 03http://tinyurl.com/9zbd9be03 [17:40] what is swfc? [17:40] divVerent: it's likely the same [17:41] there are 3 cases [17:41] rgb24 (32 bits actually), pal 8, and the rgb15 (16 bits) [17:42] ffmpeg.git 03Paul B Mahol 0777e6b085a338: imgutils: do not use av_pix_fmt_descriptors directly * 03http://tinyurl.com/9xa6qs603 [17:44] Action: durandal_1707 this is too much typing [17:47] ok [17:48] I also noticed that swfc unfortunately can't generate 16bits, because it only supports png input which has no 16bpp support [17:48] at least imagemagick can't output such png [17:56] http://www.mindless-labs.com/samhaxe/doc/1.0/files/src/NsFastXml-hx.html this is what I would try next, but have to go [17:57] ffmpeg.git 03Heesuk Jung 07e3301459f691: avidec: Use sample size in case incorrect timestamps for aac in AVI (Ticket #1755) * 03http://tinyurl.com/8ob83eh03 [17:58] ffmpeg.git 03Michael Niedermayer 070d002de4a452: exr: fix twin ; * 03http://tinyurl.com/9g8agrv03 [18:01] michaelni: where is syntax to mark ;; red in vim? [18:01] i need to write one... [18:04] ffmpeg.git 03Paul B Mahol 0782eba2266739: lavd: do not use av_pix_fmt_descriptors directly * 03http://tinyurl.com/9ur9haf03 [18:12] durandal_1707: ok :) [18:13] ubitux: np :) [18:20] michaelni: ok to push "[PATCH 4/4] lavf/swf: rename some tag defines to match the specs." ? [18:21] philipl: i'm going to make the mov_text share some code with a "raw" text subtitle [18:21] do you have any objection? [18:22] michaelni: exporting ff_* in lavu is nonsense [18:24] ubitux, iam not a big fan of renaming things unless theres a maintainer on the ffmpeg side who maintains the specific code [18:25] it can make merging harder [18:25] matching the spec looks important in these cases :( [18:25] (there are not much, and having the "sound" info for instance is useful) [18:26] but well i won't insist [18:26] durandal_1707: btw, if you have some swf with nellymoser and/or speex, i'm also interested! :) [18:26] ubitux, just volunteer as maintainer for swf :) [18:27] durandal_1707, yes it is [18:27] michaelni: okay! :) [18:30] michaelni: what's the procedure? :) [18:30] ubitux: you do merges from libav [18:31] right, looks safe so far [18:31] the biggest threat is random file renames [18:31] michaelni: https://github.com/ubitux/FFmpeg/compare/master...swf-misc [18:31] i'll soon push up to the N-1 commit [18:35] ubitux: you do _all_ merges [18:35] sure right.. [18:35] not just swf file .... [18:36] huh? [18:36] no :( [18:39] michaelni: so is it forbiden making ff_* used outside like in lavfi to be renamed to av_* ? (or this is just cosmetics ?) [18:51] ubitux, if you want me to merge something, just tell me what [18:52] ok, give me some time to finish the latest commit then [18:52] durandal_1707, ff* should not be used from outside, code violating that should be changed [18:52] existing ff symbols probably should be kept for compatibility until the next major bump [18:56] and what about adding new av_* stuff (which is just ff_* with different name)? [18:59] ffmpeg.git 03Paul B Mahol 07a291345b1e82: sws: do not use av_pix_fmt_descriptors directly * 03http://tinyurl.com/935tzc903 [19:00] Hmm! 1 second quicktime in 8min 32 sec quicktime out:( [19:03] durandal_1707, sure can be done, or avpriv_ depending on intended visibility/use [19:04] 5e6439a12508 [19:07] errr. 047dcfabc bad, bo2943 probably, what has happened to cut n paste... [19:11] TimNich, can you elaborate on bad ? [19:11] how can i reproduce? [19:12] ffmpeg.git 03Michael Niedermayer 076f557a2e125b: caca: fix 10l typo * 03http://tinyurl.com/8nwzjxh03 [19:13] that is damn copy/paste [19:14] 1 second uncompressed yup mov -> proreHQ, after that comitt comes out with a reported duration in QT of 8:32, but plays for a second, then the counter keeps on counting with a freeze frame. Import into Avid takes 10 mins instead of 5seconds and is rubbish.. [19:16] b02493e476 [19:18] you can see the wrong duration in ffplay too [19:21] TimNich, i see no odd duration with tests/data/fate/vsynth1-prores*mov [19:21] that is with ffplay [19:22] I will try a simpler command string [19:25] OK with no vf it is OK [19:26] michaelni: sample aspect ration 0....100... / 0 fails in lavfi , so ffplay cant display image, i think any sample ration with den 0 is invalid [19:28] sorry wrong file, still get it with a simple command line [19:30] ffmpeg_bisect -i in.mov -y -f mov -map 0:v -flags +ildct -top 1 -c:v prores -profile:v 3 -an in-gb02493e.mov [19:32] nitrate is also reported as 119kb/s instead of 64694kb/s [19:32] s /n/b/ [19:40] durandal_1707, 0/0 == nan should not cause problems [19:41] it causes problems with lavfi... [19:41] when using with ffplay [19:41] so instead of adding broken workarounds in decoder lavfi should be fixed [19:42] TimNich, a 3:07.63 file is still 3:07.60 with that with ffpllay [19:42] one of such workarounds is in dpx [19:44] dpx checks also for aspect ratio <0 which surely is invalid [19:44] michaelni: that commit definitely breaks with my source material. my 1 sec is 22M would you like it somewhere? [19:44] but 0/0 should work like 0/1 [19:45] TimNich, yes, ill look once i have time [19:47] michaelni: not sure what other info I can provide, my material has a timecode track and is interlaced, that and the bigger frame size and the hq coding are the differences from the fate sample. [19:52] TimNich, you can provide the / a file that is needed to reproduce it [19:55] michaelni: 3 files in uploads.ffmpeg source, script and txt. [20:00] script is a bit convoluted as its a general purpose test script, but you get the command line from it. [20:02] Action: TimNich of home now [20:13] TimNich, fixed locally, ill push after some tests [20:13] michaelni, is there any interest in registering ffmp.eg domain, for future user? [20:13] use* [20:15] would be useful for fancy short ulrs if needed, dunno [20:15] burek, feel free to register it if you want ... [20:26] ffmpeg.git 03Michael Niedermayer 073a48e38ad0e3: motion_est: more complete SAB diamond size check * 03http://tinyurl.com/9ha6grx03 [20:26] ffmpeg.git 03rogerdpack 078684e396563a: add dshow maintainer * 03http://tinyurl.com/9eoadc303 [20:26] ffmpeg.git 03Michael Niedermayer 07b8d64559d5dd: movenc: fix timescale for timecode track * 03http://tinyurl.com/96qjlw403 [20:50] here we go, even more samples supported. [20:56] michaelni, ok :) btw, is this better: http://pastebin.com/8eaFv5z9 [21:10] wait [21:10] i did something wrong :/ [21:17] i hope this one is correct http://pastebin.com/D0sj643f [21:38] o my key is approved [21:39] ill enable mingw native fate ubuilds, ioc, and maybe inspxe-cl [21:53] ffmpeg.git 03Cl?ment BSsch 070c40220b9dbc: lavf/swfdec: fix flushing with compressed swf. * 03http://tinyurl.com/8wotkx403 [22:16] burek, patch applied [22:25] michaelni great :) thanks :) [22:53] ffmpeg.git 03Michael Niedermayer 075e885275f1ed: mpeg4videodec: disable frame multithreading for GMC, its not implemented at all * 03http://tinyurl.com/9bhr46b03 [23:11] ffmpeg.git 03Cl?ment BSsch 07cd78192d09d1: lavf/swfdec: return more meaningful error codes. * 03http://tinyurl.com/9regwa903 [23:11] ffmpeg.git 03Cl?ment BSsch 0791fad50f631f: lavf/swf: define more tags. * 03http://tinyurl.com/9vnwg2403 [23:11] ffmpeg.git 03Cl?ment BSsch 07a1d2210debdd: lavf/swf: re-align after previous commit. * 03http://tinyurl.com/9d9ltr903 [23:11] ffmpeg.git 03Cl?ment BSsch 07cdfa92641531: lavf/swf: transform the swf tags define in an enum. * 03http://tinyurl.com/9boczqh03 From burek021 at gmail.com Sat Oct 13 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Sat, 13 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121012 Message-ID: <20121013000501.4A70818A018E@apolo.teamnet.rs> [01:32] Can anyone tell me why I keep getting double the frame rate with my encoding? If my source is 29.97, I get 59.94, can someone show me what I should change in my command line http://pastebin.com/gCeT5837 [01:35] Sashmo_at_work: why are you not using the presets? [01:35] found that this worked most reliably with my setup [01:35] declaring most individual x264 option is not recommended [01:35] also, you forgot to include the complete console output [01:35] so how would you change that? [01:37] a good start would be http://pastebin.com/xpi2XY8S [01:38] ok Ill try that, but if the bitrate dosnt change? [01:38] what next? [01:39] what do you mean by, "if the bitrate doesn't change"? [01:40] SORRY! I mean frame rate! [01:40] you should pastebin the console output with your command [01:40] always include your command and the complete console output when you are asking for help [01:41] ok will do [01:45] http://pastebin.com/PhsvjbJP your command, no audio, bu fps is same [01:46] does it make sense that I have to deinterlace the video before I pass it to the encoder? [01:50] is the input interlaced? [01:50] I suspect [01:50] yes [01:55] is there a work around there? [02:00] Sashmo_at_work: does the output look fine? [02:01] well, on computer it looks ok, but on a television I can see a strobe effect, like there are duplicate frames [02:22] Hey guys I am running CentOS (yes, I know, I even pick on myself for it) and I was using an outdated version of FFMPEG (as is the life for someone using CentOS) so it wasn't converting to webm, but it was converting to MP4 (I need a copy in both). So I googled and found http://ffmpeg.org/trac/ffmpeg/wiki/CentosCompilationGuide which I followed and now I can convert to webm but cannot convert to mp4 [02:23] http://pastie.org/5038547 [02:25] teamcoltra|mba: you didn't include your actual command. [02:25] Action: llogan wishes for psychic powers [02:26] http://pastie.org/5038560 [02:27] sorry llogan :) [02:27] [libvo_aacenc @ 0x28d0a00] Unable to set encoding parameters [02:27] not very useful... [02:28] probably has issues with the 6 channel audio [02:31] teamcoltra|mba: so either try adding "-ac 2" or copy the audio from input to output since i think mp4 can handle ac3 (but i'm not sure) [02:31] i don't know how good/bad ffmpeg is going from 5.1 to 2 channels. [02:32] or what the "recommended" procedure is [02:34] I just googled it and saw this http://forum.videohelp.com/threads/345323-Convert-5-1AAc-in-2channel-stereo-in-mp4?s=102de7dcf92c0e77e4b30bd183a13122&p=2154880&viewfull=1#post2154880 llogan [02:35] I was going to copy his example but it looks like it also messes with bitrates [02:37] anything that is going to be re-encoded is going to be messing with the bitrate [02:39] Do you mind walking me through the basics of what his command does? My problem is that my script needs to be a kind of one size fits all so I am hoping it's not going to degrade (horribly) the audio quality of every video it messes with [02:40] teamcoltra|mba: https://ffmpeg.org/trac/ffmpeg/wiki/x264EncodingGuide [02:40] see crf section [02:40] as for audio, since your input channels may vary, you will probably end upp just using -ac 2 [02:41] i suppose i should add fdk-aac to the CentOS guide [02:42] llogan I am getting the same error anyway [02:45] llogan: so, any ideas how to fix my issue with the frame rate? [02:46] Sashmo: don't deinterlace for tv? [02:46] http://pastie.org/5038655 llogan [02:47] option placement matters [02:47] llogan: I cant change my source&. or can I? [02:47] ffmpeg [input options] -i input [output options] output [02:47] Sashmo: sorry, i don't understand. [02:48] http://pastie.org/5038664 [02:48] llogan: My source is interlaced video, after encoding, I get double FPS, is there a way to avoid this? What should I add to the command to avoid the double FPS? [02:50] teamcoltra|mba: now you're using it as an input option telling ffmpeg to decode the input as two channels. [02:50] but it probably ignores it [02:50] lol [02:50] lets try again [02:51] Sashmo: yadif=0 (default) should provided 1 frame per frame, so i don't think it is doubling your output frame rate, AFAIK. what makes you think it is double? [02:51] Okay where does it go? because that time it told me "[NULL @ 0x2298240] Unable to find a suitable output format for '2' [02:51] 2: Invalid argument" when I put it right before the output videofile [02:52] ffmpeg -i input -ac 2 output [02:52] llogan: well, vlc and ffmpeg read the output as 59.94 when the source is 29.97, what else could it be? [02:53] http://pastie.org/5038683 [02:54] teamcoltra|mba: you have "ffmpeg 2" [02:54] lol I promise I am not screwing with you [02:54] Action: teamcoltra|mba fixes [02:55] well hot damn [02:55] Sashmo: ffmpeg shows "29.97 fps" for your output [03:18] if I do ffmpeg -i myfile it shows it at 59.94 [03:19] there is definitely something wrong with the encoding [10:27] ? [11:03] great! I successfully made a small preview version of a video by cropped, rescaling to smaller resolution, dropping frames to make fps smaller, and encoding in lower bitrate. [11:03] ffmpeg is great. [11:12] I want to compile ffmpeg in static , but it needed ld-linux.so.3, how to fix this problem? [12:13] Hi. I'm using ffmpeg - ffserver - ffplay to stream pcm16be @16000Hz . The system structure and the commands are here http://pastebin.com/t2zM3GGA . The server.conf is here http://pastebin.com/bmgK7YN6. I would to stream over rtmps between ffmpeg and ffserver. Is possible? Thank you [12:18] Spideru: Did you build ffmpeg with librtmp support? Did you look at the ffmpeg wiki? (there might be something on it) [12:19] relaxed: I've looked at ffserver documentation and I didn't find something about ffm plus rtps or only rtps. I'll check ffmpeg doc, thank you [12:20] Is a package from deb, and apparently was built with librtmp support [12:20] (thank you again) [12:21] Spideru: google's first result for "ffmpeg rtmp" http://sonnati.wordpress.com/2011/08/30/ffmpeg-%E2%80%93-the-swiss-army-knife-of-internet-streaming-%E2%80%93-part-iv/ [12:22] thank you relaxed, I'll read it [12:28] relaxed: the link is wonderful, but I was looking for a way to tell to ffserver: "hey, you must handle rtmp/rtmps flow". I found this and I'm reading it now http://stream-recorder.com/forum/use-ffserver-ffmpeg-rtmpdump-vlc-convert-live-t9529.html?s=6f8b7b0635fe9dc538fe5c96d8046e0a& [12:29] also: thank you for your support :) [15:14] hi FFmpeg guys [15:14] how to build FFmpeg with fPIC flaghs ? [15:15] you did something wrong [15:15] i bet you got a message "or recompile with -fPIC" [15:16] :) [15:17] burek: yeah recompile with fpic... DUDE. how to do it ? [15:18] judging from my experience, whenever I got that kind of err message, it meant I ran 'make', got an error, changed some things, and ran 'make' again (creating a mess) :) [15:18] CFLAGS="-fPIC" ./configure --enable-pic [15:18] so, whenever I did make distclean [15:18] ./configure again [15:18] and make again, it worked [15:18] DelphiWorld: but you need ALL libraries that ffmpeg depends on compiled with PIC [15:18] without that fPIC msg [15:18] mmmmmmmm [15:19] Mavrik: where i put it? in ./configure ? [15:20] DelphiWorld, please try 'make distclean; ./configure ; make' again, before enabling pic [15:20] just to make sure :) [15:20] burek: okay;-P [15:21] burek: building a FFmpegified VLC [15:22] +1 [15:22] burek: make -j 6 ;) [15:29] burek: same issue [15:31] then read above ^ [15:32] burek: where i put that FLAGH [15:33] well Mavrik gave you an obvious example [15:34] buuuuuuuuuuuuuuuurek: :P [15:34] i didnt understand it, magn3ts and burek explain:-P [15:34] Mavrik ... sory magn3ts [15:37] hi michaelni [15:37] how do they work anyway ;) [15:39] michaelni: how to build FF with fPIC? [15:47] Action: DelphiWorld beat burek ass [15:47] burek: --host-cflags="-fPIC" [15:49] DelphiWorld, a lot of times [15:49] when you ask a generic question [15:49] i take a look at the google search [15:49] burek: ;-) [15:49] anyway [15:49] make burek clean && halt [15:49] and if the answers can be found among first 3 results [15:49] i just ignore the question [15:49] so that you know :P [15:50] make burek clean && halt [15:50] try: make distclean && ./configure && make [15:50] if it breaks, use pastebin to show the results (complete output log or config.log) [15:50] Action: DelphiWorld make distclean burek && echo 'shutting down burek, Good bye FFmpeg:-P' | wall && halt [15:52] burek: issue is not in FFmpeg. issue is that VLC need FFmpeg to be fPIC [15:59] hi amstan [15:59] amstan: Ucraine? [15:59] DelphiWorld: no, Romania [15:59] amstan: ah so close ;) [16:00] amstan: never heard the name stan only in Ucraine;) [16:15] DelphiWorld, I've build 100s of VLCs with ffmpeg so far [16:15] not once I needed -fPIC [16:15] burek: i dont know why but i'm fighting it. about to get it dude [16:15] good luck :) [16:16] Action: DelphiWorld kick burek out of his media center and coppy all all all flac encoded Musics;-) [16:16] I don't use flac :) [16:17] (random tests) [16:17] ok it works [16:19] burek: hey, hey. why there's a branch called Qatar ? [16:20] ? [16:20] burek: there's a ffmpeg branch called "Qatar" [16:20] what the name indic? [16:21] where [16:24] burek: vlc done [16:30] burek: branch of FFM [17:14] now is there any option to auto reload FFserver config ? [17:14] without restarting it [17:32] burek: hello, do you have two minutes for a question? [17:45] Spideru: better to ask ;) [17:48] burek: http://www.youtube.com/watch?v=jVfUCS860UU [17:48] Ok, now I'm using ffmpeg - ffserver - ffplay to stream pcm16be @16KHz. These are the commands http://pastebin.com/GZaBb1jy and this is the server.conf http://pastebin.com/c9igTxZd [17:49] i have many home videos that i am trying to archive. I am importing them as uncompressed avi files. And wow, they are just massive! Is there a compressed format yielding similar quality with less disk space footprint? I want to 1) convert to mpeg2 for dvd and 2) use the import as a archive on amazon glacier. [17:49] I have problems recognizing on ffplay when ffmpeg stop to stream (for example) [17:50] if I stop ffmpeg, then restart it, ffplay still playing "old" stream, without sound of course [17:50] And: is there a way to stream with rtmps to ffserver? [17:51] I seen there is a conf for rtsp but not for rtmps [17:51] (i would to stream encrypted data from ffmpeg to ffserver [17:51] ) [18:04] Can anyone help me with this dbouel framte rate issue with a interlaced source for input? how can I get it back to normal? http://pastebin.com/PhsvjbJP [18:15] in other words, is there a good compressed format viable for archival purposes? [18:15] good meaning produces high quality with relatively low size [18:23] mutt [18:25] whoops, that was supposed to be for zsh [18:29] lake: yes, I would use ffmpeg -i input -c:v libx264 -preset veryslow -c:a flac output.mkv [18:29] er, add -crf 18 after the input [18:30] that will give you highly compressed (lossy) video and lossless audio [18:30] relaxed: darn, i get "Unable to parse option value "-1" as pixel format [18:30] pastie.org your command and all output [18:33] relaxed: http://pastie.org/5045110 [18:34] is it interlaced? [18:34] i believe it is [18:34] i didn't not explicitely deinterlace it [18:34] you should, hold on a second [18:38] ffmpeg -i input -filter:v yadif:1,format=yuv420p -c:v libx264 -crf 18 -c:a flac output.mkv [18:42] lake: what is the output fps from the command? [18:42] 29.97 [18:42] i get this tho: No such filter: 'yadif:1' [18:43] change it to yadif=1 [18:44] that should double the framerate of the input. [18:46] relaxed: getting somewhere now. must figure out how to set ntsc for v4l2 options [18:46] ah, -standard [18:47] relaxed: damn, that quality is epic. [18:49] So the output is ~60fps, correct? [18:50] 59.94 tbc [18:50] is that it? [18:51] good, yes [18:52] now to figure out cropping and getting my alsa device to work [18:52] :) [18:52] thanks relaxed [18:53] You're welcome. Read about the crop filter in the man page. Be sure and add it after yadif in the filter chain. [18:53] Can anyone help me with this dbouel framte rate issue with a interlaced source for input? how can I get it back to normal? http://pastebin.com/PhsvjbJP [18:55] Sashmo_at_work1: You want one frame per field? yadif=1 [18:55] let me try that [18:56] yadif=0 for one frame per 2 fields (=frame) [18:57] =1 will look smoother [18:58] thanks a bunch I will try this now [19:11] hey, did anything change in 1.0 for setting rtsp transport? - before I had AVDictionary *d = NULL; av_dict_set(&d, "rtsp_transport", "tcp", 0); in my code to force tcp - but since upgrading to 1.0, it seems UDP is being used again for rtsp streams, any ideas? [19:20] relaxed: that didnt seem to work for me, Im still getting more frames [19:33] Sashmo_at_work1: yadif=0 will match the input's framerate. [19:34] relaxed: looks like its being encoded at the right bitrate, i can see that in the outpuf of ffmpeg, but every time I try to play the file, its reported as double the rate exactly [19:35] relaxed: any idea why i am getting alsa buffer xruns? http://pastie.org/5045534 [19:35] no audio is playing back, playing through mplayer [19:36] relaxed, is there a way you could create a bug report for lake's issue with yadif [19:36] I mean, shouldn't ffmpeg auto-recognize it and apply the yadif filter [19:37] judging by this message: [avi @ 0x1944240] Could not find codec parameters for stream 0 (Video: none (422P / 0x50323234), 704x464): unknown codec [19:37] it seems like ffmpeg doesn't support that kind of input [19:39] burek: he has no issue with yadif [19:39] he doesn't but ffmpeg does :) [19:40] Did i miss something? [19:40] in one case ffmpeg says it cant find codec parameters [19:40] and when he added yadif, then it worked [19:40] if I understood correctly the above conversation [19:40] can anyone tell me where the x264 profiles are hidden away? I want to adjust one [19:40] Sashmo_at_work1, type x264 --help [19:41] burek: it was a colorspace issue [19:41] never mind, i'll create it [19:42] the bug has nothing to do with yadif. [19:44] and everything to do with 422P -> 420p conversion not being done by default. [19:44] relaxed ok :) [19:44] I'm not saying it should but ffmpeg loves vague error messages. [19:46] relaxed: im confused, when ffmpeg uses the x264 profiles, where are those profiles? I cant see to find them, fyi using ubuntu [19:46] burek: From your bug report, can ALSA_BUFFER_SIZE_MAX be set on the cli? [19:47] Sashmo_at_work1 ffmpeg uses x264's built in profiles now [19:47] no more file presets and stuff [19:47] relaxed, er.. what? [19:47] `x264 --fullhelp|less` will give you a list of presets, tunes and profiles [19:47] burek: so if I want to change one of the settings, i need to just overide it? [19:48] yes [19:48] burek: https://ffmpeg.org/trac/ffmpeg/ticket/615 [19:49] hey guys! i am facing a strange problem with my script to turn .MTS files to .mov. it worked 2 days ago when i testet it with the yadif option and now i wanted to make use of it and it does not work it tells me about a invalid loglevel. it worked 2 days ago and I did not change anything [19:50] KING_LEE: pastie.org your command and all output. If you paste a script I will hurt you. [19:51] http://pastie.org/5025641 thats the script and thats the the output http://pastie.org/5045671 [19:51] heh [19:52] oh,... I learned pastie.org now ^^ [19:52] just a sec relaxed, let me take a look [19:52] KING_LEE: change "-v yadif" to "-vf yadif" [19:53] ok, i will [19:53] relaxed :) "Opened 11 months ago" :) [19:53] also, that script is wrong! [19:53] if only i could now remember what was it all about... [19:53] burek: answer my question! [19:53] :) [19:53] i am just veryvery confused that it worked 2 days ago [19:53] KING_LEE: for a in `ls *.MTS` is poor bash. Don't parse 'ls' output. Use for a in *.MTS [19:53] I'm confused it ever worked. [19:54] relaxed, in the source code [19:54] btw, it gets even worse.. [19:54] when ambient becomes too dark for webcam [19:54] then it really starts to spit a lot of those [19:54] overruns or such [19:55] relaxed: what do you mean by wrong? well, it did, i tested a few clips and now i wanted to use it with a whole folder [19:56] KING_LEE: remove -sameq and replace it with "-q:v 1" [19:58] relaxed: this ouputs audio that i can hear: " ffmpeg -f alsa -i hw:6,0 out.ac3" [20:00] lake, https://ffmpeg.org/trac/ffmpeg/wiki/Capturing%20audio%20with%20FFmpeg%20and%20ALSA [20:00] http://pastie.org/5045758 [20:02] burek: i can successfully capture audio by itself. it breaks down here: http://pastie.org/5045534 [20:03] lake, welcome to the v4l2+alsa hell :) [20:03] KING_LEE: Here's your new script: http://pastie.org/pastes/5045786/text [20:03] burek: it works fine with mencoder [20:03] I couldn't solve the problem since 11 month ago, so, I can't suggest anything worth of use to you :/ [20:03] but that's on an uncompressed avi [20:04] KING_LEE: Here's your new binary: http://goo.gl/DPrRY [20:04] I just switched to vlc [20:04] lake, if your device supports mjpeg, try that kind of input [20:04] if alsa errors are gone, most probably it is some sort of usb bandwidth issue or something like that [20:05] KING_LEE: well, change the script "1920:1080", I made a typo [20:06] relaxed: thank you very much, I check that in a minute [20:07] KING_LEE: new script, take 2: http://pastie.org/pastes/5045786/text [20:16] hmm,... not working, but why? [20:16] http://pastie.org/5045959 [20:17] burek: okay, well, maybe i can get away without flac [20:19] i'm having problem decoding AAC with avcodec_decode_audio4. i'm using it the same way i was doing MP3 decoding, but for AAC i get the following error: [aac @ 0xcb4bf0] channel element 1.0 is not allocated [20:19] is there something extra i should be setting up for aac? [20:20] hi [20:21] I'm writing a flv muxer while getting into encoding / muxing etc... so just a learning experience :). I'm writing video frames (x264) which is working great. [20:21] now I want to add audio and I was looking into speex. The FLV specification only states that the audio body in the FLV bitstream, "varies per format" [20:22] now I'm wondering if someone knows if this simply means that the raw bytes can just be stored, or if I need to add some sort of speex-related header? [20:27] Is there any way for me to add libvorbis to my codecs without having to recompile? I want to convert aac to ogg. [20:33] relaxed: i don' run a 64bit system [20:33] knicholes, no [20:33] burek: K, thx :D [20:33] knicholes, wait [20:33] you can try static builds :) [20:35] sweet [20:36] do i need to set up ffmpeg again? i thought i made a good decision with this http://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide [20:55] when building ffmpeg with libx264 support, any idea howto satisfy ./configure when libx264 is in non-standard location? [20:56] so far i added x264.pc to /usr/lib64/pkgconfig/ without success [20:58] did you run ldconfing [20:59] also, did you check ./configure --help [20:59] maybe --extra-libs [21:07] burek: give KING_LEE a link to your 32bit builds [21:07] KING_LEE I give you a link to my 32bit builds :) [21:08] oh, sorry [21:08] well, just [21:08] or priv msg fflogger with !help [21:14] eeer, ... ok, sorry to ask this question but how do i install that and why was it wrong to follow that compilation guide. i am still confused that it worked 2 days ago [21:15] following the compilation guide is fine [21:17] I was just trying to get you a quick fix. The static builds include a ffmpeg binary that you can run from the current working directory like this, ./ffmpeg [21:17] KING_LEE, no need to install anything [21:17] download-extract-run [21:18] just like zipped exe file [21:18] ok, just asking because there was no readme file [21:18] there is one, as soon as you open that given link :D [21:21] but this one is empty http://ffmpeg.gusari.org/static/32bit/README.html [21:21] http://ffmpeg.gusari.org/static/ [21:24] cd [21:24] oops, wrong window [21:26] burek: ldconfig -v shows fine: libx264.so.128 -> libx264.so.128, prob is configure wants x264.h and --extra-libs doesnt help (expecting library) [21:29] extra-cflags extra-ldflags? [21:31] and how do i get this static build to run my script? please excuse my stupidity i have a bad headache today but i want to get this ........ going [22:02] can anyone help me out here, I keep getting double the frame rate after encoding. I use yadif=0 and the ffmpeg output shows 30fps, but everytime I check the output I still get double frame rate, any other ideas? [22:03] what's your input framerate? interlaced? [22:03] 29.97 interlaced [22:06] yadif=0 should net you 29.97p [22:06] let me pastebin what I am getting, its werid [22:09] http://pastebin.com/D7NTwiXs [22:11] output is 29.97p [22:11] vlc is telling me that its 59.95 [22:11] 59.94 [22:11] when I try to pull the url [22:12] Do you need to set all those libx264 options? Output to a file instead of streaming and check the results. [22:12] Action: relaxed & [22:14] sorry to be anoying, i just cant get the static build running to work through the folder, whats the syntax to combine ./ffmpeg and the script? [22:16] relaxed: i told him yesterday that the output seemed to be ntsc [22:17] KING_LEE: what script? [22:18] ...and i mentioned the deal with the options [22:20] that one http://pastie.org/5047234 burek and relaxed were so kind to hgelp me with that and told me to run http://ffmpeg.gusari.org/static/ but i cant combine them. [22:21] it works great on one file but i want to use the script to do loads of files [22:25] [19:54:32] KING_LEE: for a in `ls *.MTS` is poor bash. Don't parse 'ls' output. Use for a in *.MTS [22:38] burek: we changed that, this is a new script [22:40] but when I try to run it with ./ffmpeg it does not work anymore [22:55] KING_LEE: then show the output so we can see any errors/messages. "not work" is a very common "issue". [22:57] I guess it is probably just a matter of systax combining "./ffmpeg" and "sh my_ffmpeg-script.sh" [22:58] the script and the ffmpeg build work superb on their own [22:59] http://pastie.org/5047656 [23:01] KING_LEE: without seeing the actual contents of the scripts i don't know what we can offer. [23:01] http://pastie.org/5047234 [23:02] posted it earlier [23:03] burek: --extra-*flags works, thank you [23:11] foonix :beer: :) [23:12] KING_LEE, obviously, if you need any kind of help [23:12] a more extensive logs are needed [23:12] what exactly doesn't work [23:31] hey burek Mavrik https://dl.dropbox.com/u/97608608/MEFTAH%20TAYEB%20ENNAHAR%20TV.mp4 [23:32] what is that? [23:33] i didn't say nice to you [23:33] but to fflogger dying :D [23:33] I clicked a wrong button :) [23:36] LOL burek [23:42] How do I debug what passlog FFmpeg is looking for in 2 output scenario? Tried `./ffmpeg1 -i ...-passlogfile ffmpeg2pass-0 output0.mp4 ... -passlogfile ffmpeg2pass-1 putput1.mp4" doesnt work. [23:54] Mista_D http://ffmpeg.org/ffmpeg.html#Options-3 [23:58] burek: you suggest to use stats of the x264 options? Under section 5.7 it still lists passlogfile as a valid option in the same manual. Thank you for a hint. [00:00] --- Sat Oct 13 2012 From burek021 at gmail.com Sun Oct 14 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Sun, 14 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121013 Message-ID: <20121014000501.8A51318A018C@apolo.teamnet.rs> [00:00] how do you print video stats/info using ffmepg? [00:00] Mista_D, when you did -pass 1 you should have specified your -passlogfile [00:00] which you will use in -pass 2 [00:01] creep ffmpeg already does that [00:01] by default i.e. [00:02] burek<< i mean i have mkv file, and print its information only [00:02] --analyze [00:02] burek: I didn't specify a name, but I saw ffmpeg2pass-0.log and -0.log.mbtree, so I copied them to *-1.log [00:03] burek: any way to see what passlog file FFmpeg is looking for for second output? Tried -debug 2-9 already. [00:03] Mista_D [00:04] ffmpeg ... -pass 1 -passlogfile f1 out1 ... -pass 1 -passlogfile f2 out2 ... [00:04] for 2nd pass [00:04] ffmpeg ... -pass 2 -passlogfile f2 ... -pass 2 -passlogfile f2 ... [00:04] ffmpeg ... -pass 2 -passlogfile f1 ... -pass 2 -passlogfile f2 ... [00:04] typo [00:04] creep try ffprobe [00:05] I want to use one passlog file for pass 2 for 2 files (they only have slight bitrate differnece). [00:06] thanks trying [00:08] Mista_D, seriously... [00:08] 1 passlogfile per 1 output [00:08] why would you want all that in 1 file? [00:09] btw, I think in ffmpeg-devel you would get a reply like "patch welcome" or something :) [00:09] just to test something [00:09] burek: I know (: [00:09] :) [00:24] Mista_D, you want to use [00:24] 1 passlog file (to analyze just first file) [00:24] and use that passlog for both outputs in pass 2? [00:27] lake, are you still here [00:28] if not, when you get back, can you please take a look at this ticket https://ffmpeg.org/trac/ffmpeg/ticket/1808 [00:34] creep: also see ffprobe [11:36] hi. how can i encode a video to an aspect ratio of 1.85:1 with ffmpeg? [11:38] iive: you know? [11:40] what? [11:41] iive: how can i encode a video to an aspect ratio of 1.85:1 with ffmpeg? [11:41] Can someone help me with a little problem in a command? [11:41] https://gist.github.com/fe200d5bd0d8c1cb39e8 [11:42] The first command fails but the second doesn't.. [11:42] The only that that has changed (as far as I know) is the file [11:42] iive: when i try 1.85:1, i get "Invalid aspect ratio: 1.85:1" [11:43] Invalid chars ':1' at the end of expression '1.85:1' [11:43] funyun: try withuout the : then :) [11:44] iive, do you know? [11:45] Simex: Looks like genuine bug. Unfortunately you are using the LibAV fork, instead of real ffmpeg. [11:45] Yes I know the problem between ffmpeg and avconv... [11:46] Action: iive checks if they have newer version than 0.8.3 [11:46] iive: it produces the same result. here's my syntax "ffmpeg -i INPUT.mkv -map 0:0 -map 0:1 -aspect 1.85 -target ntsc-dvd -crf 19 -ab 224k -threads 0 -ss 500 -t 60 test.mpg" [11:47] iive: it's forcing the video to be 480x720 [11:47] likely the ntsc-dvd is the one doing the forcing. [11:47] iive: the source is 8408x1920 [11:48] funyun: why do you place the width last? [11:48] opps 808x1920 [11:48] iive: not sure what you mean? [11:48] The problem is I can't find a good way to provide the right ffmpeh [11:49] ffmpeg* [11:49] funyun: maybe you can try moving the aspect to the right. in case it takes it for input aspect override. [11:49] I was going to use the Ubuntu repositories with my application [11:52] iive: i still get the same result [11:53] iive: shouldn't it add black bars if it's going to force me to stay at 480x720? or is there a way i can add black bars? [11:53] funyun: there is something else... mpg have a limited set of aspect ratios. mpeg2 dvd set have just 3 afair [11:53] 4/3 16/9 2.21 [11:54] Simex: yeh, i'm looking at the ffmpeg download page to figure out the repositories that contain ffmpeg packages and to check what versions are available... [11:55] iive: so there's no way around this? [11:55] Okay [11:55] funyun: why do you think so many dvds have black bars? [11:56] iive: i thought because of the same problem i have? [11:56] iive: that's what i mean. is there a way for me to add black bars? [11:57] sure there is. ffmpeg have full fledged image filter system. [11:57] and it had expand even before that. [11:57] I just don't know the exact syntax. [11:58] iive: i just tried 2.21 and now it looks a lot better. but still stretched a little [12:02] seems like -vf pad is the one that can put black bars [12:03] from the documentation, try [12:03] -vf pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2" [12:06] iive: i get "Negative values are not acceptable." and "Failed to configure input pad on Parsed_pad_0" [12:06] ohh.. [12:08] Simex: btw, i see that the error happens right after shared memory access. you do have /dev/shm mounted as tmpfs and having enough free space? [12:09] Mmm [12:09] Moment [12:09] Yes it is mounted [12:10] iive: it appears i made a mistake also. 2.39 is the AR i need [12:10] not sure if that matters [12:11] Enough free space yes I'm sure [12:11] well, i'm still too sleepy, otherwise i should have noticed that 800/1920 is not 1.85 :O [12:12] haha [12:14] iive: any idea on how to correct the syntax you gave me? [12:16] funyun: actually this [12:16] funyun: actually the above command would pad the width... not height. [12:17] this is why it gives negative size... [12:18] iive: so -vf pad="ih*16/9/sar:ih:(oh-ih)/2:(ow-iw)/2" ? [12:19] sorry, you just got me at a bad moment. [12:19] i'll be more helpful later. [12:21] iive: i really needed to encode this so it's ready by the time i get home from work. i'm about to hop in the shower. if you get a chance within that time, could you possibly message me the syntax? i really appreciate your help [12:21] funyun: you can probably do the math on your own and put it in the parameters. [12:22] iive: trust me, i can't lol. it took me about 2 hours to figure out my current syntax of "ffmpeg -i INPUT.mkv -map 0:0 -map 0:1 -aspect 1.85 -target ntsc-dvd -crf 19 -ab 224k -threads 0 -ss 500 -t 60 test.mpg" [12:24] iive: if you need to go, it's cool. still thanks for your help :) [12:25] ii'll try to make a video of similar size and see what options works. [12:25] i assume you use at least 1.0 ? [12:25] i can make you a sample if you need [12:26] 1.0? [12:26] ffmpeg 1.0 [12:26] iive: oh. yes [12:27] iive: i'm not totally sure how to tell the exact version but it was built on Sep 18 2012 [12:28] "ffmpeg version git-2012-09-18-91af760" [12:37] iive: http://www.sendspace.com/file/iamuh4 [12:37] brb in 20 [12:45] for some reason width and height are reversed in the sample. no surprise it doesn't work. [12:47] funyun: "ffmpeg -i INPUT.mkv -map 0:0 -map 0:1 -vf pad="1920:1080:(ow-iw)/2:(oh-ih)/2" -target ntsc-dvd -crf 19 -ab 224k -threads 0 -ss 500 -t 60 test.mpg" [12:47] this one seems to work in my setup. [13:11] iive: worked for me too. thanks so much! :) [13:14] Action: iive pheu, that was close. [16:28] Hello. I would to connect to ffserver using rtsp. now I'm using this command to stream from ffmpeg to server http://pastebin.com/Bj6hsucS . This is ffserver.conf file http://pastebin.com/Ya7CEY5v . Any suggestion? Thank you [16:46] I get this error from ffmpeg with -f rtps and rtps://yaddayadda: "Could not write header for output file #0 (incorrect codec parameters ?): Invalid data found when processing input " [16:47] on server side i can see [ANNOUNCE] "rtsp://localhost:8091/feed1.ffm RTSP/1.0" 200 166. What am I missing? [16:47] hi Spideru [16:47] :) [16:47] Hi burek, as usual I'm crying behind ffmpeg-ffserver [16:48] :) [16:48] (It's powerful but for a noob is hard to use it at best) [16:48] I've taken your suggest to use ffm between ffmpeg and ffserver [16:48] and the sound is great :) [16:48] Spideru, did you read this maybe http://ffmpeg.org/trac/ffmpeg/wiki/Streaming%20media%20with%20ffserver [16:49] and no, you can't stream test.wav [16:49] wav format is not streamable [16:49] uhm wait [16:49] but you can put uncompressed audio into aac maybe [16:49] there is something that I am missing [16:49] wav is a file storage format [16:49] not a streaming format [16:49] yes [16:49] ok [16:50] aac might help better [16:50] but I'm streaming from ffmpeg to ffserver without wav [16:50] but it's lossy encoding [16:50] you are not streaming from ffmpeg to ffserver [16:50] you're just feeding ffserver with ffmpeg [16:50] wav is from ffserver to ffplay [16:50] yep [16:50] ah [16:50] try ogg/flac [16:50] I'll try it [16:51] I must use lossless [16:51] for ffmpeg-ffserver, just use: ffmpeg -f alsa -ac 1 -ar 16000 -i default -f ffm http://localhost:8090/feed1.ffm [16:51] thank you again for support and for that doc. I've searched a lot on ffmpeg doc but I've never seen it before [16:52] burek: thank you [16:52] there is the way to know when ffmpeg is no more feeding ffserver? [16:52] I can see on server GET / POST [16:52] POST is printed when i close ffmpeg [16:53] Is there another way? [16:53] I would to know on ffplay side when ffmpeg is no more feeeding ffserver [16:54] also try this http://pastebin.com/hEArynzm [16:56] Spideru, when ffmpeg stops feeding ffserver, you'll know :) [16:56] i.e. ffplay won't play anymore :) [16:56] ffplay remains open [16:57] without sound. I need to get an alarm [16:57] an event [16:57] well, I noticed that also [16:57] ffplay continues running even when its input is dead [16:57] yep [16:57] so you might check man ffplay [16:57] to see is there an option [16:58] to quit if the input is dead [16:58] or you can use vlc/winamp/something [16:58] ok, I'll check it [16:58] instead of ffplay [16:58] must check if vlc/something get an alarm [16:59] in extreme case i can parse output from ffplay. When there is no more data, a flag inside output fall to 0 [17:00] well, i think vlc will quit if you tell it to [17:00] you put vlc://quit or vlc://close [17:00] in the playlist, after the URL [17:00] my curiosity: Is there a protocol that can take commands from player (pause, stop, vol+, vol-) and send to ffmpeg/ffserver? [17:00] cool, thank you [17:00] so when it can't play that url anymore, it advances to the next playlist item and quits [17:01] nice tip, thank you [17:01] rtsp can do that [17:01] nice, so I need to use trsp [17:02] *rtsp [17:02] but between ffserver and ffplay I can imagine [17:02] and between ffserver - ffmpeg? [17:02] you don't put anything between ffserver - ffplay [17:03] I am missing something on high level [17:03] must study better [17:03] ffmpeg feed ffserver [17:03] with ffm [17:04] then ffplay ask to server the stream [17:04] ffserver take the data it's feeding from ffmpeg and stream to ffplat [17:04] *ffplay [17:05] I would to send a vol+ from ffplay to ffmpeg (that is feeding ffserver), in order to (for example) raise volume on ffmpeg device alsa [17:06] I expect a VOL+/yadda/dothat command, and will be my work to handle that command [17:09] Spideru, I'm not sure I understand why do you need all that [17:09] but if it works, then great :) [17:09] I need: 1) stream lossless 2)send messages like vol+ to ffmpeg from ffplay [17:10] first point it's ok (with your precious help) [17:14] I can take a player, put some custom button on it and send custom messages from player to audio soruce [17:14] *source [17:14] I would do that [17:17] Spideru, what is your logical goal? [17:17] what are you trying to implement? [17:18] I have a custom device that stream audio (no need to have bidirectional channel) [17:19] and I would to receive it on a PC, record it, set volume on custom device from PC, and other custom commands [17:20] at the moment I'm using ffmpeg for the stream, and a socket TCP for commands [17:20] Could be wonderful to have all funcions inside the same "data channel" [17:22] Can see from server and from PC when source is no more online [17:22] maybe stream on autenticated channel [17:23] *authenticate [17:28] you might want to check vlc/vlm [17:28] they even have an admin web page [17:29] where you can control all that [17:31] ok thank you. Last thing: with your ffserver.conf, when I play with ffplay http://localhost:8090/stream.ogg I get: Format ogg detected only with low score of 1, misdetection possible! http://localhost:8090/stream.ogg: End of file [17:32] stream.flv and Format flv [17:37] Sat Oct 13 17:31:39 2012 Audio codec 'flac' not compatible with FLV [17:40] But I'll work on that. Thank you burek. It's late, need to go. See you later or tomorrow, bye :) [17:41] how do i use mp3 for an audio codec [17:42] ffmpeg -formats shows mp3 being installed for D and A [17:45] ihsw mp3 is a decoder [17:46] libmp3lame is an encoder [17:46] in any event i'm at a loss of how to install libmp3lame [17:47] is there any alternative to compiling ffmpeg [17:48] i get the distinct impression you're trying to avoid telling me the package name that apt-get would need [17:49] i don't understand [17:51] indeed [17:51] well, you are the one who needs help, not me [17:51] if you don't want to explain, fine [20:28] when ffmpeg decodes TrueHD, DTS-HD, etc&does it output 24-bit PCM? [20:35] i have a sync problem - i'm trying to convert a .flv into a .mp4. if i extract the video and audio separaetly they work fine, but if i combine them together, the video is jerky - running slow for .75 seconds, then too fast for .25 seconds. the audio is fine [20:36] the video and audio are never out of sync, the video is just running at a strange frame rate - if the frames were just displayed with an equal delay, it'd be perfect. is there a sync setting i'm missing here? [20:37] just attempting ffmpeg -i in.flv -vcodec copy -acodec copy out.mp4 results in a synced but jerky output video, where the input is fine and the separately extracted streams are fine [22:02] Hi, I have a 2.35:1 aspect video I am trying to covert to DVD, which i would like to do while preserving and not squishing the video. Does anyone have a command handy that does this effectively? [22:03] running debian squeeze, linux [22:36] burek: hey, i'm looking at this now https://www.hackerschool.com/ [22:36] shit, not that [22:36] https://ffmpeg.org/trac/ffmpeg/ticket/1808 [22:45] burek: i will upload sample of the video and provide a link to you. will you be able to add it to the ticket? [23:02] does anyone have a script or a command handy that'll let me convert a 2.35:1 (letterbox) to a NTSC-DVD without cramming or stretching the video to 16:9? Like a command with a pre-determined padfilter, perhaps? [23:03] I'm probably not asking my question correctly [00:00] --- Sun Oct 14 2012 From burek021 at gmail.com Sun Oct 14 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Sun, 14 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121013 Message-ID: <20121014000502.B69B018A01E8@apolo.teamnet.rs> [00:00] --- Sat Oct 13 2012 [00:02] hey saste! :) [00:02] hey [00:02] saste: did you see my OT link the other day? :) [00:03] ah yes the life thing [00:03] yeah [00:03] :D [00:03] there are plenty of things of that kind... [00:03] okay :) [00:03] btw my attempt with libvisual failed some time ago [00:03] the API is not meant to the low level usage that we require [00:04] that is there is no way to pass audio data and get video data [00:04] weren't you willing to look into milkdrop instead? [00:04] milkdrop is even worse [00:04] since it is tightly integrated with opengl iirc [00:04] or maybe the other one i never remember the name [00:04] something like projectx [00:05] you could make a sink of it, but not that it's very useful... [00:05] btw an opengl sink or device would be very welcome [00:05] projectx? [00:06] i don't remember the name :p [00:06] there are plenty projects named like that [00:06] proof that people has no inventive [00:06] ah projectM! [00:07] yes projectM is integrated with opengl [00:07] that is there is no way to use it programmatically in a transmedia filter [00:07] integrated? :p [00:07] (http://projectm.sourceforge.net/) [00:07] ok [00:08] that's too bad then [00:08] what about frei0r [00:08] again, you could create an output device [00:08] frei0r? doesn't deal with audio [00:08] saste: btw, any idea about my question on the buffer ref metadata copy thing? [00:08] (the =NULL & dict copy thing) [00:09] no i still have to read that [00:09] what's "deal" anyways [00:09] you want some visualization where audio goes in, and you get a bitmap back? [00:09] ohsix, yes [00:09] as simple as it seems, yet i couldn't find any project delivering that feature [00:09] not that anyone has done anything interesting with this on linux, but doing it without acceleration is pretty much a dead end [00:10] you could use pbo's to stream the output back after it's been rendered with one of the opengl visualizers, but that way lies doom [00:10] pbo? [00:10] pixel buffer objects, just a sane way to access the contents [00:11] yes the problem is that they are all focused on visualization [00:11] while that's not our problem, that's only the final step [00:11] saste: and we can't grab the opengl content or something? [00:11] so we need to handle with raw data buffers [00:11] i must have missed the premise then, what are you trying to do? [00:12] ubitux, no experience with opengl, so maybe i'm missing something completely obvious [00:12] ohsix, an audio visualization wrapper [00:13] and given that, what does "being focused on visualization" mean? [00:14] not audiobuffer2videobuffer oriented [00:14] ohsix, that we don't want to immediately visualize it, we need to process it like other data [00:14] but draw-the-audio-buffer-on-opengl-surface, i guess. [00:14] still, i'm pretty sure we could grab the opengl content :p [00:15] process the output or the input? i thought it was implied that you wanted audio in and a surface out [00:15] a video buffer on surface out [00:15] so it can be filtered in the filterchain, encoded, etc [00:15] from what i remember [00:16] the API i checked (libvisual) works like this [00:16] you pass an audio buffer [00:16] right, so by focused on visualization you mean it creates its own window or whatever, even if it ends up in a buffer it's still focused on visualization :p [00:16] then there is a thread processing the audio and sending the video data to opengl, which does the rendering [00:17] from our point of view such API is not useful at all, since we need to know when the video data is ready [00:17] but again, i may be wrong [00:17] saste: btw, did you look at coverity? [00:18] but it was enough to realize that i couldn't easily get along with what i wanted, and i gave up [00:18] ubitux, no [00:19] shrug, from what i understood of frei0r it's data in data out, you could do this with it but you'd need to write your own plugins that behave that way [00:20] saste: that's pretty neat: http://ubitux.fr/pub/pics/_coverity.png [00:20] ohsix, are you aware that we already have a frei0r filter? [00:21] (no idea if that particular issue is right or wrong though) [00:21] yes [00:21] but frei0r is video only [00:21] we need audio in -> video out [00:22] we could do it natively, but since geeks worked on that since years, i thought we could leverage their work [00:22] i was apparently wrong [00:23] there's nothing stopping you from putting audio data in the source images :p but like i said, you'd need to write your own visualization plugins for it to do anything useful [00:24] we already have two visu in lavfi [00:24] we were willing to support advanced visualizrs [00:25] what practical use is it when you probably have something in mind when you do such a thing, and can do it better in a video editor [00:25] uh audio data in video buffers, that's an hack [00:25] no it isn't, heh [00:25] not talking about the audio settings :p [00:26] let's encode into wav and put them into frames and them to frei0r [00:26] and they aren't "video buffers" anyways, that's just their typical usage [00:26] i can already imagine some people screaming and with their eyes bleeding [00:26] :) [00:26] frei0r basically does nothing but find plugins in standard places and provide a process() routine [00:27] ohsix, but i don't wont to write plugins, i want to leverage existing code [00:28] *wont -> want [00:28] right, that's why i said it was useless [00:29] you will definitely have to adapt something though [00:29] cuz you don't even really get much with the linux visualization projects either [00:30] <@saste> i can already imagine some people screaming and with their eyes bleeding // http://z0r.de/3564 [00:30] ubitux, ahahah [00:33] it's a shame because projectM plugins are f***ing cool [00:33] well we could still implement an output device in the worst case [00:33] i have some contacts with mylkmist developers, i'll see if they can help :) [00:33] we have already caca after all [00:34] abusing frei0r for just general data moving would probably be a bad idea, but as far as visualization (and most mathematical transformations) they are equivalent :p [00:34] i say, if you're going to be doing work, port AVS; it's great [00:35] ohsix, they're already doing it [00:35] or better, ffmpeg supports avisynth since ages [00:35] not that i'm fascinated with this junk anymore, but once there were 3d cards the 2d effects were kind of lame [00:35] saste: advanced visualization studio, the winamp plugin [00:36] ffmpeg.git 03Georg Lippitsch 0705b7315412c3: Fix DPX decoder * 03http://tinyurl.com/9546knz03 [00:36] ffmpeg.git 03Georg Lippitsch 0724778c32d80e: Fix writing 12 bit DPX * 03http://tinyurl.com/9n4cbsu03 [00:36] ffmpeg.git 03Michael Niedermayer 07c2340831b8e9: aacsbr: change order of operation to prevent out of array read * 03http://tinyurl.com/9mn7gqe03 [00:36] ffmpeg.git 03Michael Niedermayer 070f46825d9833: ffserver: prevent nb_streams from becoming too large * 03http://tinyurl.com/8fddhns03 [00:36] it supports a bunch of 2d effects that you compose [00:36] ohsix, i wouldn't whence to start [00:36] ohsix, i wouldn't know... [00:37] download winamp, run it in wine and play with some of the presets :p [00:37] bah, too much work [00:37] the source is available, but not under a great license, but i know the guy to be flexible [00:37] and write a script to send a screenshot command, grab the picture and inject it in libavfilter with a filebuffersource filter [00:38] careful, you have to run screenshot commands fast enough to grab all the pic [00:38] 25 commands per second [00:38] while you could get the opengl output, you probably wouldn't want to, it's a whole can of worms (like for one, needing an accelerated display, of the X variety to even work) [00:40] oh wait, AVS is bsd licensed [00:40] if you want the best outcome for the least amount of work, i would port AVS, it is amazing [00:41] not to mention being able to use public presets, of which there are a million, and some very popular/legendary ones even; something linux really doesn't have [00:43] http://www.1014.org/code/nullsoft/avs/ [00:44] google video "advanced visualization studio" to see some of the presets people have made, if you really can't be bothered to find a windows machine :p [00:45] ohsix, that's not a problem, i have dual boot and a win7 VM [00:46] it works with wine too [00:46] it looks like libvisual already does AVS [00:49] ah it's something else entirely [00:49] it's a reimplimentation of one component of it, without all the goodies [00:49] how should i add an xface test to FATE? [00:50] i want to test both encoder and decoder [00:50] most image tests are in image.mak, but they only test the decoder iiuc [00:50] will you add a saste.xface? :) [00:51] more something like lena.xface [00:51] hello. I just checked out ffmpeg from git, and I can't build the ffmpeg command for some reason [00:54] what error Gnosis- ? [00:56] anyways, AVS is freakin' great, if you want "meat", there it is, you'd have to port it or reimpliment it but the presets & scripting language is das infinito [00:56] Compn, no error. The make was successful [00:57] it's just that it only compiled ffplay, ffserver, and ffprobe [00:59] nothing uses cpu like an old school 2d effect! :D (actually wasted memory bandwidth making pretty colors, bring it) [01:00] Compn: http://bpaste.net/show/8RmaJrWkYE1UUj6dR4vQ/ [01:01] I added --enable-ffmpeg at the end, but that didn't fix this problem [01:08] ubitux: i wonder when/if ioni/wonder will implement your request for the arch ffmpeg package. he used to be fairly quick, IIRC. [01:08] heh you're stalking me! [01:09] nah, i just trawl the bugs on occasion [01:09] but i was the "damned dirty ape" who commented! [01:12] Gnosis-, --disable-avfilter ----> no ffmpeg [01:14] ah, okay, thanks! [01:14] why does it have that effect in 1.0 but not 0.10 or earlier? [01:14] ffmpeg depends on some filters now [01:15] --enable-ffmpeg should do more than what it does though (thats a bug) [01:16] yeah, I added it because it seemed logical based on what was in ./configure --help; I didn't see it listed there, though [01:17] ubitux: +k)Yd; K(|`g]9BSqmrYkq#^P(y-P'r-c4u#?8|@de:G{]Q)Z>pf/az3sLsNl&+;(wRBInu4 M)}~Lap^q.:4wL&By6f8c6h; .QSIY%'\thr0/K);0KZW#Dx9g4Ko!3m>V+2y*Zs5?'rCn=[~?Qf3-*9Ovp:aK".KM at M9F*>X6sl/ god bless you. [01:18] i was going to say it saste [01:18] Action: llogan just remembered that 12.10 is a week away after messing with the guide. [01:18] ubitux: ^^ saste.xface [01:18] oh, ok [01:18] :D [01:19] saste: how can i test your branch? :) [01:23] ubitux, i'm going to send an updated patch in a few minutes [01:24] i'll check this out tomorrow then :) [01:24] 'night :) [01:34] if i want to change the default 'analyzeduration' of '5000000' is this the right place to change it: ./libavformat/options_table.h:51:{"analyzeduration", "how many microseconds are analyzed to estimate duration", OFFSET(max_analyze_duration), AV_OPT_TYPE_INT, {.i64 = 5*AV_TIME_BASE }, 0, INT_MAX, D}, [01:34] more precisely -> 5*AV_TIME_BASE [01:35] ? [01:50] burek, yes [01:53] thanks :) [01:55] ffmpeg.git 03Michael Niedermayer 07f374e9989be2: vf_fade: fix memleaks of args * 03http://tinyurl.com/9k8hg3c03 [01:55] ffmpeg.git 03Michael Niedermayer 07d2a618ab2213: af_pan: fix memleak of arg * 03http://tinyurl.com/98mzh2903 [01:55] ffmpeg.git 03Michael Niedermayer 0754b2d317ed99: ffv1: avoid checking a double for equality * 03http://tinyurl.com/9xvu6nz03 [01:59] should we mention somewhere that avfilter is required for ffmpeg binary ? [01:59] seems like we will get a lot of reports about this . [01:59] lots of people --disable-everything then just enable a few things. probably breaks all that [02:06] so when will ffmpeg final version be released? [02:07] and ffmpeg t-shirts? [02:07] ffmpeg lollipops ? [02:10] when hungary becomes the football world champion :) [02:10] creep: tshirts are on my todo list. [02:11] Action: llogan forgot about that quote... [02:11] burek<< i know nothing about that [02:12] i just don't know how we will pay for them yet with the lack of donations [02:12] they used to sell t-shirts for money.. [02:12] a novel concept. [02:12] http://www.spreadshirt.com/i-love-boobies-C3376A8474877#/detail/8474877PC27159424 [02:14] ffmpeg.git 03Michael Niedermayer 07120b38b966b9: avio: redesign ffio_rewind_with_probe_data() * 03http://tinyurl.com/8kdfmep03 [03:37] anybody know which tools can soft crop h264 ? [03:38] ffmpeg.git 03Michael Niedermayer 07e47024d72f32: wtvdec: fix memleak on error * 03http://tinyurl.com/8q752fm03 [03:39] ffmpeg.git 03Michael Niedermayer 07f657d495b04d: rtpdec_qdm2: change one assert to av_assert0 * 03http://tinyurl.com/8zggp3v03 [03:39] ffmpeg.git 03Michael Niedermayer 073f350a482fd0: ff_celp_lp_synthesis_filterf: check that filter_length is within the supported range * 03http://tinyurl.com/9jc3w2803 [03:39] ffmpeg.git 03Michael Niedermayer 073f010421421f: ff_celp_lp_synthesis_filterf: change loop end check * 03http://tinyurl.com/8c29wwm03 [03:39] ffmpeg.git 03Michael Niedermayer 079f9307ff2a6c: rv30_decode_intra_types: make check tighter * 03http://tinyurl.com/94q4ygp03 [04:00] ffmpeg.git 03Michael Niedermayer 074acfe3d193c7: jpegls: fix off limit * 03http://tinyurl.com/9qw8zqu03 [04:00] ffmpeg.git 03Michael Niedermayer 078dc89944270a: jpegls: increase run_index to 4 * 03http://tinyurl.com/9znumsj03 [05:23] looking at libvisual, you can request an actor to explicitly render into a given buffer, and you can skip opengl plugins when enumerating [05:23] seems to me that'd work fine [05:28] so your usage _could_ be to just use an actor with a given visvideo (to which you've attached your own output buffer) and you can ignore the rest [05:33] what would be done for invariance, the visualizations use random numbers but there's no way to make stuff repeatable [05:34] presumably you'd want the same output each time [06:24] ffmpeg.git 03Michael Niedermayer 07ff814c75a3f5: ffserver: fix return value of add_codec() * 03http://tinyurl.com/8aucgho03 [09:58] ffmpeg.git 03Paul B Mahol 07efb0e4e7afc6: truemotion2: use more meaningful return codes * 03http://tinyurl.com/8f586lm03 [10:10] ffmpeg.git 03Paul B Mahol 0792b3d8bc5337: bethsoftvideo: return meaningfull error codes * 03http://tinyurl.com/8kvlcxp03 [10:25] ffmpeg.git 03Paul B Mahol 07f2f711cde277: pcx: read sample aspect ratio * 03http://tinyurl.com/8qal7wd03 [10:25] ffmpeg.git 03Paul B Mahol 0779133fd0e58f: pcxenc: store sample aspect ratio * 03http://tinyurl.com/98hselk03 [10:25] you will single handedly add many digits to tinyurl, that's not very nice [11:10] ffmpeg.git 03Paul B Mahol 07e8b50385cf90: fate: update pcx reference * 03http://tinyurl.com/8cpocnd03 [12:00] ffmpeg.git 03Carl Eugen Hoyos 076254ffe0cadc: Allow autodetection of some dnxhd files that can be decoded correctly. * 03http://tinyurl.com/9t7z9sh03 [12:50] ffmpeg.git 03Sami Pietil? 073cc025273251: vp8dec: reset loopfilter delta values at keyframes * 03http://tinyurl.com/8e6mymc03 [14:21] ffmpeg.git 03Luca Barbato 076d5600e8556a: avutil: add yuva422p and yuva444p formats * 03http://tinyurl.com/94p97he03 [14:21] ffmpeg.git 03Sami Pietila 070bf511d579c7: vp8: reset loopfilter delta values at keyframes. * 03http://tinyurl.com/8sdnygs03 [14:21] ffmpeg.git 03Michael Niedermayer 07eae35eadc0ae: rtspdec: Fix use of uninitialized byte * 03http://tinyurl.com/8l46j7k03 [14:21] ffmpeg.git 03Michael Niedermayer 072f1b2ff934e6: rtmpproto: Fix an out of array write * 03http://tinyurl.com/9zpndk903 [14:21] ffmpeg.git 03Michael Niedermayer 07c80b59f679a0: tscc2: Fix an out of array access * 03http://tinyurl.com/8p6jlv703 [14:21] ffmpeg.git 03Martin Storsj? 075a2cb7821916: rtspdec: Set the default port for listen mode, if none is specified * 03http://tinyurl.com/9mly29l03 [14:21] ffmpeg.git 03Diego Biurrun 07f6c38c5f4ed6: avfilter: call x86 init functions under if (ARCH_X86), not if (HAVE_MMX) * 03http://tinyurl.com/9bqr2td03 [14:21] ffmpeg.git 03Michael Niedermayer 073b0bb321a50c: Merge commit 'f6c38c5f4ed6683a6a61db2ed418a68bbe5f5507' * 03http://tinyurl.com/8u7r5au03 [14:29] ffmpeg.git 03Diego Biurrun 07930c9d4373e0: avutil: Duplicate ff_log2_tab instead of sharing it across libs * 03http://tinyurl.com/9adp2su03 [14:29] ffmpeg.git 03Michael Niedermayer 07d197bd4f5ee7: Merge commit '930c9d4373e0f3cb7c64fcfc129127a309f6d066' * 03http://tinyurl.com/8nmwvdk03 [14:34] this is nonsens libav folks just move stuff around and calls they are doing all the work [14:36] well, diego doesn't represent the whole libav team :) [14:37] i'd be tempted to say "fortunately for them" but that's a bit rude for him :p [14:37] and when they do not move stuff around they start to rename random stuff [14:38] btw, what's this AVCodecContext->coded_frame thing for decoders? [14:38] is it some kind of reference frame? [14:38] itn't documented somewhere? [14:39] "the picture in the bitstream" [14:39] michaelni: this *_LIBAV pixfmt is pointless, it is just there for cosmetics [14:43] ffmpeg.git 03Diego Biurrun 07d5c62122a7b2: Move av_reverse table to libavcodec * 03http://tinyurl.com/9lrxpap03 [14:43] ffmpeg.git 03Michael Niedermayer 07d6c342fdc0b4: Merge commit 'd5c62122a7b26704bf867a1262df358623bf5edf' * 03http://tinyurl.com/9unzucr03 [14:43] mmh we can't have multiple side data of the same type in a packet? :( [14:43] btw, diego is one of the root masters or at least used to be. In that meaning he literally represents the libav team, for all matters not related for coding. [14:44] ubitux: what would you do with several width/channels packet side data? [14:44] iive: sure, for the political aspect i agree, i meant technically speaking [14:44] ubitux: pick one of them with random()? [14:45] durandal_1707: easier to store the key/value thing of the metadata [14:45] "gimme the next side data for this type" [14:45] "ohai here is another key/value" [14:45] but whatever, i'll come with another way of storing :p [14:48] michaelni: somthing in eval use av_reverse [14:49] michaelni: this becomes extremly stupid, please stop merging libav, if there is anything usefull pick it with cherry-pick [14:51] moving these tables has a very real reason, you should understand thigns before you complain [14:51] making both ABIs compatible is not possible [14:52] nevcairiel: eval use av_reverse [14:52] nevcairiel: the msvc thing? [14:52] nevcairiel: once libav removes av_reverse who will fix eval in lavu? [14:52] afaict for these case a duplicate is made [14:52] so tables are not exported [14:53] iirc there was some issues with exported tables [14:53] so every table will be now duplicated [14:54] they drop snow to make some place ;) [14:54] :)))))) [14:54] msvc needs some special hackery to support tables in shared library builds, so instead of doing that, it was decided to finally clean the whole issue up, because loading data tables from other shared libs is never ideal [14:54] i don't understand how that is a cleanup :/ [14:54] table is used in one lib, so move it to that lib [14:54] seems cleaner to me [14:54] in this particular case yes [14:55] nevcairiel: not for ffmpeg [14:55] but duplicating all the tables doesn't look fine [14:55] of course if there are no other solutions.. [14:55] the problem looks stupid though :( [14:55] its not "all the tables", its like 2 math related tables which are pretty small [14:56] and some other external const as well, right? [14:56] (like the crypto size thing) [14:56] accessing cross-library even on linux is a performance overhead [14:56] hmm? why have it duplicated across source, just use single header [14:57] Action: durandal_1707 talking about ffmpeg lavu [14:57] ah right, the header looks like a good way to do it :p [14:57] duplicated in the binary, but not in the code [14:57] thats how all these duplicates work, they dont duplicate code [14:57] because thats stupid [14:58] only in shared libraries tehy are dups [14:58] i admit i didn't look closely [14:58] okay [15:03] i was planing to move it to a header, log2 is a c file ATM [15:03] being a c file sucks because when you compile against libavutil you cant count on having the c file [15:04] you only have the headers and libs to link to [15:05] arent you supposed to have the full source if building any of the libs? [15:06] oh its used from inline functions [15:06] sounds like that would always fail, because ff* symbols are private anyway [15:06] maybe for libav [15:07] ffmpeg.git 03Diego Biurrun 07c1ef30a6ba2c: De-doxygenize some top-level files * 03http://tinyurl.com/8c8t2wq03 [15:07] ffmpeg.git 03Diego Biurrun 07bc4620e5d61a: Remove libmpeg2 #define remnants * 03http://tinyurl.com/9o9adtn03 [15:07] ffmpeg.git 03Michael Niedermayer 07b4ca1b159f4b: Merge commit 'bc4620e5d61a4dd9a1f654fadd281a172aab04be' * 03http://tinyurl.com/96g73pj03 [15:10] michaelni: is there reason why avcodec.h part is not applied for this LIBMPEG thing ^ [15:11] i wasnt planing to drop that idct before theres a good reason to do so [15:11] btw, how do i test MMI ? [15:14] or diferently said what options do i need to pass to gcc for it to try to build the MMI code ? [15:18] noone ? [15:19] what is MMI code? [15:20] MIPS stuff [15:20] libav droped it vaguly pointing to "it doesnt work" [15:20] i am trying to decide if theres a point in keeping it or not [15:21] but my mips cross compiler doesnt seem to accept MMI code by default [15:22] ill probably just drop it and whoever wants it and knows how to test/fix it can then put it back [15:23] it is some PS2 code [15:28] seems there have been no non cosmetic commits to it in 9 years [15:29] did you contact the mips guy to see if it makes sense to keep it? [15:29] i will but i cant wait for his reply with the merge [15:29] its trivial to revert and put it back [15:30] michaelni: probably partially useful for old hw [15:31] if it works [15:31] but people say it doesnt [15:32] because it got broken over time? [15:32] ffmpeg.git 03Diego Biurrun 07ca411fc1d343: avcodec: Remove broken MMI optimizations * 03http://tinyurl.com/8bpqslu03 [15:32] ffmpeg.git 03Michael Niedermayer 0785fe70b64c3d: Merge commit 'ca411fc1d34329cd17b28627f697e391ae52073f' * 03http://tinyurl.com/8tzu5qy03 [15:32] durandal_1707, i suspect so [15:35] "broken beyond repair" :/ [15:35] please... :D [15:37] that's how thing works, broke some stuff and make sure fate does not detect it, when it becomes obvious that is is broken just remove whole thing [15:39] iam happy to add it back and add a fate test if someone fixes it and tells me how to test it [15:40] ffmpeg.git 03Diego Biurrun 0790558e848a29: rangecoder: K&R formatting cosmetics * 03http://tinyurl.com/8ttnfev03 [15:40] ffmpeg.git 03Michael Niedermayer 07c55bebe2cc7b: Merge commit '90558e848a29ef1e85ecb1832ad9a26eebe958e0' * 03http://tinyurl.com/8qq4qde03 [15:44] i hate some of the alignment nit "fixes" :( [15:44] it makes sense to align when the data is related [15:44] no when it's fun to draw a column of '=' :/ [15:46] Action: michaelni doesnt care much about whitespace, though yes i agree some of the fixes are of dubious value [15:46] especially some of the 80 column fixes sometimes are kinda "interresting" [15:47] i mean especially the ones where there was vertical alignment of table or function call argument columns before the improvment [15:47] how can I debug flv muxer from ffmpeg, which is used by vlc [15:48] i mean, when i set the output in vlc, like "mux=ffmpeg{mux=flv}" i get 640x368 instead 640x360 frame sizes [15:48] ffmpeg.git 03Diego Biurrun 074c66af6141a0: rangecoder-test: Set error message log level to error, instead of debug * 03http://tinyurl.com/8ck8wjz03 [15:48] ffmpeg.git 03Diego Biurrun 079e6ea3cef992: fate: add avstring test * 03http://tinyurl.com/8hhvpf403 [15:48] ffmpeg.git 03Diego Biurrun 07717addecad77: Use proper return values in case of missing features * 03http://tinyurl.com/9kwf2sb03 [15:49] ffmpeg.git 03Mans Rullgard 077e76fc528d60: mpegvideo: remove write-only variable * 03http://tinyurl.com/94566da03 [15:49] ffmpeg.git 03Mans Rullgard 07366484fff172: smjpeg: fix type of 'ret' variable in smjpeg_read_packet() * 03http://tinyurl.com/9scg8ln03 [15:49] ffmpeg.git 03Mans Rullgard 070a7005bebd23: rtpdec_xiph: fix function return type * 03http://tinyurl.com/8klmuyf03 [15:49] ffmpeg.git 03Michael Niedermayer 0718884f159b61: Merge commit '0a7005bebd23ade7bb852bce0401af1a8fdbb723' * 03http://tinyurl.com/9o367em03 [15:52] I don't know how to provide useful logs for such issue :/ [15:53] burek, maybe ask on #videolan [15:53] just did :) [15:53] and we debuged it to the point of lavf's flv muxer [15:54] for example: vlc v4l2:///dev/video0 --sout='...,mux=ffmpeg{mux=flv},dst=live.flv' [15:54] sorry, for example: vlc v4l2:///dev/video0:width=640:height=360 --sout='...,mux=ffmpeg{mux=flv},dst=live.flv' [15:54] burek, so av_logs dont show up with vlc+ffmpeg or what is the problem, ? [15:54] live.flv (as shown by mediainfo) shows 640x368 [15:55] michaelni, to be honest I don't know, I'll take a look.. btw: http://pastebin.com/iGeEz2Rf [15:55] I don't know which lines are av_logs [15:56] you could hack av_log to add a prefix [15:56] Action: michaelni leaves in the direction of the kitchen to seek something to eat [15:59] burek: 0x15fb798] main mux warning: option flags is unknown [16:00] yes, I had a line like: ...,mux=ffmpeg{mux=flv,flags=+global_header},... [16:00] and later realized vlc doesn't just support passing any option to ffmpeg [16:00] but only listed ones in the docs [16:00] so i removed it later [16:01] now it's just ,mux=ffmpeg{mux=flv}, [16:01] let me repaste logs [16:01] burek: not going to read that mess again [16:02] Action: ubitux loves av_hex_dump_log() [16:02] but i see nowhere ffmpeg logs [16:02] so ask videolan are such put in logs at all [16:02] i understand you guys, just tell me how can I extract anything useful for debug and i'll do it [16:02] im now search where av_log is defined so i can prefix it with something [16:03] burek: you need to pass -v debug or something like that to vlc .... [16:03] durandal_1707 you mean -vvv ? [16:03] i did [16:04] michaelni: the side data things seems to work fine :) [16:04] (lavfi meta inject) [16:05] burek: does that mean ffmpeg logs are debug too? [16:05] do they show in the vlc's output log too? i guess yes [16:05] let me check [16:06] our logs have nice hex numbers [16:06] and colors, i know :D [16:06] durandal_1707 does this look like that: [flv @ 0x1583860] Codec for stream 0 does not use global headers but container format requires global headers [16:07] not all the debug lines have it [16:07] (like those with a NULL context) [16:07] can you just tell in which file is av_log defined [16:08] so i can prefix it [16:08] :S [16:09] burek: libavutil/mem.c [16:09] meh [16:09] log.c sorry [16:09] thanks [16:10] look at the snprintf for av_log_default_callback [16:10] and just add some stuff there [16:10] (hopefully vlc didn't change that callback) [16:11] snprintf(line, sizeof(line), "%s%s%s", part[0], part[1], part[2]); [16:11] that? [16:12] yes [16:13] thanks a lot :) [16:13] try adding "HELLO THIS IS BUREK" before the "%s... [16:13] :) [16:13] do i need to recompile both vlc/ffmpeg now [16:13] or just ffmpeg [16:13] if it's a static build you need to relink vlc [16:14] (if vlc is linked statically against ffmpeg i mean) [16:14] what happened to hex editing? [16:14] :D [16:14] durandal_1707 :D [16:14] how can I hex edit to add something :D [16:14] the string was just "%s%s%s" [16:14] no space for prefix :) [16:15] also, to speed up things, when i make such small changes [16:15] look for free data area in the binary and dump some asm opcode in it, then add a jmp where appropriate [16:15] burek: not trivial [16:15] is it enough to delete log.obj [16:15] log.o * [16:16] or i must distclean [16:16] ubitux right :D [16:17] there should not be need to remove anything [16:17] i tried just rm libavutil/log.o && make [16:17] :D [16:17] let's see what did I scr.. up now :D [16:17] ffmpeg.git 03Paul B Mahol 076571833d1a54: smjpegdec: use url_feof() * 03http://tinyurl.com/8gnqb7q03 [16:34] do you know of any tool that can generate dependency diagram for ffmpeg source code [16:44] ffmpeg.git 03Mans Rullgard 070daac647af00: avstring-test: fix memory leaks * 03http://tinyurl.com/9tfouxa03 [16:44] ffmpeg.git 03Mans Rullgard 07a77f01c72529: configure: use utilities from /usr/xpg4/bin if it exists * 03http://tinyurl.com/9ygltjm03 [16:44] ffmpeg.git 03Justin Ruggles 0761d5313d94c9: dca: allocate a secondary buffer for extra channels when downmixing * 03Error03 [16:44] ffmpeg.git 03Justin Ruggles 07f5962229bfcb: avplay: use audio parameters from the decoded frame instead of AVCodecContext * 03http://tinyurl.com/9x77bc703 [16:44] ffmpeg.git 03Michael Niedermayer 0715ef1cfe6431: Merge commit 'f5962229bfcb14c2879e69ccdf7f1a4934168609' * 03http://tinyurl.com/98j2urm03 [16:45] is it me or AV_PERM_ALIGN doesn't have any effect? [16:45] do we have this in ffmpeg's doxygen http://www.stack.nl/~dimitri/doxygen/diagrams.html [16:47] i think it would really help people to visualize how ffmpeg is structured: http://www.stack.nl/~dimitri/doxygen/examples/diagrams/html/inherits.html [16:47] i just dont know who to bug for this feature :) [16:47] we have ascii diagrams [16:48] that's even better [16:48] do you have any link perhaps? [16:48] i clicked everything there and couldn't find :S [16:50] it was a bad joke :) [16:50] :s [16:50] ok :) [16:51] so we don't have any diagrams atm right? [16:52] except http://git.videolan.org/?p=ffmpeg.git;a=blob;f=doc/swscale.txt;hb=HEAD or http://git.videolan.org/?p=ffmpeg.git;a=blob;f=doc/ffmpeg.txt;hb=HEAD [16:52] i dunno [16:52] (i'm not even sure ffmpeg.txt is still up to date) [16:54] ok [16:54] who maintains ffmpeg's doxygen? [16:54] we also have ascii art in snow.txt and some libavfilter headers [16:54] http://ffmpeg.org/doxygen/trunk/index.html [16:55] :) [16:55] burek, do you volunteer to maintain it ? [16:55] well, I would update it to 1.8.x [16:55] and add Graphviz [16:55] :) [16:56] burek, ask arpi or reimar [16:56] they are not on irc i guess? [16:57] ffmpeg.git 03Michael Niedermayer 07d0707677fa50: ffplay: use audio parameters from the decoded frame instead of AVCodecContext * 03http://tinyurl.com/8qgw69k03 [16:58] ive seen arpi & reimar on IRC but thats months ? year ago [16:58] great :) [17:00] btw anyone has samples that the 2 avplay changes fixed ? so i can test (i failed to find a sample that didnt work before when trying a few that change parameters) [17:01] michaelni: you mean midstream change of channels/samplerate/bps and such? [17:02] i can give you generate sample which crashes lavfi [17:03] iam interrested in one that fails before these patches and works afterwards [17:03] could be on trac [17:04] ffplay does not seem to have any problem with changes to samplerate before these patches [17:06] i think avplay resample if samplerate changes [17:07] ubitux thanks, but that didn't work unfortunately :/ there is no a single prefix in vlc's output :S [17:07] i guess i'll just drop it [17:12] ffmpeg.git 03Justin Ruggles 076304f78edf6a: avplay: support mid-stream sample rate changes * 03http://tinyurl.com/9vq74hn03 [17:12] ffmpeg.git 03Mashiat Sarker Shakkhar 075d2be71b9ecf: vc1: Use codec ID from AVCodecContext while parsing frame header * 03http://tinyurl.com/8fga9sv03 [17:12] ffmpeg.git 03Michael Niedermayer 072a56e65c3b3b: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/8vr7dyu03 [17:27] there are a lot of FF_API_* ignored warnings, normal? [17:28] they should be set to 0 [17:29] s/ignored/not defined/ sorry [17:49] wtf dvaudio decoder code was just removed [17:49] and now 99 years after vlc reinvent wheel [17:51] see 7458ccbb02a98ece4b4a9018a46eb13fff05b7c2 [17:51] michaelni: should we resurect this? [17:52] the commit is from 2003, how is that "just"? :P [17:52] and in 2012 vlc added support for it in its own code [17:53] the dv demuxer seems to shuffle the dv audio around to make it proper s16le [17:53] we have bug report for this on trac to use vlc code (sic) [17:54] nevcairiel: that hack does not work for dvaudio in other containers [17:54] hmm? [17:54] jb_afk : durandal_1707 is freaking out , what happen? :) [17:54] ehe [17:55] he is permanently freaking out about something [17:56] what does one put dv audio in anyway? i bet mov :P [17:56] it has all kind of crazy codecs [17:57] imho dv demuxer idea is just wrong - it is not demuxer job to do that .... [17:57] there were similar case which have been fixed [17:58] dv audio is seen in a lot of places [17:58] its an old codec [17:58] mov , avi at least [18:02] the whole situation is just ..... [18:06] why was dvaudio code removed ? [18:06] i havent looked into this [18:06] Action: Compn distracted [18:06] removed in 2003 [18:06] actually moved from decoder to demuxer [18:07] and nobody complained.... [18:07] oh lol [18:07] decoders dont belong there! :P [18:07] but dv stuff is usually encapsulated in a demuxer [18:07] so maybe thats why [18:07] i mean, encapsulated in dv [18:08] ffmpeg.git 03Paul B Mahol 078288c2b6cb07: pngdec: read sample aspect ratio * 03http://tinyurl.com/9pd3b2m03 [18:08] ffmpeg.git 03Paul B Mahol 07f58f90238fc6: pngenc: write sample aspect ratio * 03http://tinyurl.com/8ext99g03 [18:08] ffmpeg.git 03Paul B Mahol 0793931143feb0: lavc: return s->get_buffer() error code if it errors out * 03http://tinyurl.com/9efbl5a03 [18:10] both avi and mov have fourcc/twocc for it [18:11] still could be encapsulated [18:13] except avi ones are missing in riff [18:13] for avi 2 modes exist at least [18:14] one puts a dv stream with audio and video muxed together in a single stream [18:14] into avi [18:14] so you need to chain demuxers ... [18:16] itd be nice if lavf would pass on or fake a fourcc if one is not listed in riff.c [18:17] instead of 'no streams found' it would say 'no known streams found, using 0x1337' or so [18:18] Action: michaelni has a fix for the PIX_FMT warning storm locally [18:19] or we could just add twocc to the list as we find samples. ... http://wiki.multimedia.cx/index.php?title=Twocc [18:22] similar problem with mov and isom.c of course... [18:25] or i may be remembering a problem that doesnt exist anymore [18:25] ignore me [18:25] time toeat lunchbbl [18:33] michaelni: what you mean by chain demuxer? [18:34] and how to support both cases without duplicating code? [18:43] omg, I tried using Graphviz on ffmpeg's source code :)))))) [18:44] now that's called a really big graph! :D [18:48] here we go, let's go back to some subtitles things. [18:50] ffmpeg.git 03Michael Niedermayer 07c45b829d5258: tests: fix checksums for png aspect ratio change * 03http://tinyurl.com/9kkecrf03 [18:50] ffmpeg.git 03Dmitry Samonenko 07083c7bf70131: sdp: output speex optional vbr parameter * 03http://tinyurl.com/9k69how03 [18:50] ffmpeg.git 03Michael Niedermayer 07183117fed7d0: libavutil: loose idiotic circular dependancies between version and avutil.h * 03http://tinyurl.com/9mrr5rw03 [19:12] here is ffmpeg's source code "diagram" :D http://ffmpeg.gusari.org/uploads/gv1.png [19:12] no point of zooming it up i believe :) [19:16] this is how it looks like when actually zoomed up :) http://ffmpeg.gusari.org/uploads/gvchunk.png [19:43] oh, someone already did it in doxygen :) nice :) http://fossies.org/dox/ffmpeg-1.0/dir_b4b104e2afbab17aa9ef07180757c615.html [19:44] michaelni: where samples uploaded to upload.ffmpeg.org go? [19:46] how can I upload a sample for FATE? [19:46] (xface stuff) [19:47] you upload it to upload.ffmpeg.org [19:48] in upload dir and tell michaelni about it (my experience) [19:53] any url where i can wget the file will do [19:54] michaelni: i need access to wavpack sample in upload dir if it is still there [19:55] look in http://streams.videolan.org/incoming/ and http://streams.videolan.org/ffmpeg [19:58] first one needs authorization and second one does not have it [20:03] ffmpeg.git 03Paul B Mahol 07a5e0046a730b: xbmdec: s/av_reverse/ff_reverse * 03http://tinyurl.com/8r6zn4v03 [20:05] michaelni, just sent a sample in my latest mail [20:05] the sample is just few bytes [20:07] saste: XFACE is not audio codec [20:08] nor it is subtitle codec [20:13] how does swscale do with scaling rgb->yuv420 with an odd number of lines? :d [20:14] saste: redundant bytestream usage [20:14] durandal_1707, ?? [20:14] saste: xface patch [20:15] nevcairiel: it is broken? [20:16] just wondering what it does, since an odd height is technically invalid for 420 [20:16] durandal_1707, where? [20:16] saste: also you put AVCODEC and description in wrong place [20:16] saste: xface_decode_frame [20:17] nevcairiel: try and report ;-) [20:18] durandal_1707, can you reply on ML? [20:18] where should I put AVCODEC/desc? [20:19] ah subtitles codecs... got it [20:19] saste: why? [20:21] yes i can reply to ml, but do i really need to.... [20:24] durandal_1707, no need [20:25] really we could use a one-line macro for the codec description entries [20:25] the list would be easier to read/sort then [20:25] saste, xface uploaded [20:25] michaelni, thanks [20:29] Action: durandal_1707 ....premier... [20:47] ffmpeg.git 03Michael Niedermayer 078ab0b9cabaca: trasher: check seek return value. * 03http://tinyurl.com/8wktknh03 [20:47] ffmpeg.git 03Michael Niedermayer 0780db07adfe8a: probetest: check command line arguments * 03http://tinyurl.com/8ovj2mb03 [20:47] ffmpeg.git 03Michael Niedermayer 07225d3cc1ccd8: ffeval: avoid folding EOF onto a valid char * 03http://tinyurl.com/9h8893v03 [20:52] ffmpeg.git 03Paul B Mahol 072c85727f6c17: lavc/codec_desc: add/update properties of some codecs * 03http://tinyurl.com/8ueqmuj03 [20:57] ffmpeg -codecs | grep DEVIL [20:59] dpx and sgi should not be there [21:18] ffmpeg.git 03Michael Niedermayer 07c0f0bec2f205: sws-test: check W/H * 03http://tinyurl.com/95cg26303 [21:18] ffmpeg.git 03Michael Niedermayer 073689ec3d28d7: pp: avoid overflow in w*h * 03http://tinyurl.com/9x7ns9903 [21:21] i'm dealing with a silly container format which stores all audio packets, followed by video packets [21:22] does it make sense to cache all the video packets in the encoder context, and then release then at the end? [21:22] do we already do something similar? [21:22] you mean half files is audio and another is video? [21:22] yes, silly no? [21:23] saste, what container is this [21:23] and why does it exist [21:23] you do not write such muxers at first place [21:23] http://dpg.software.informer.com/wiki/ [21:23] oh [21:24] that abomination. [21:24] Action: Daemon404 un-reads [21:24] don't ask me why it exists, people do silly things all the time [21:25] can't be worse than 8svx (left channel data, followed by right channel data) [21:25] the only way i see it is write 2 temp files and cat it them at end [21:26] durandal_1707, that would be basically a script [21:26] i wonder if it makes sense to have a native muxer/demuxer in ffmpeg [21:27] a lot of caching involved [21:27] saste: with current API not [21:27] well i could cache them, and hog the memory [21:27] for small files (the typical use case) it shouldn't be really an issue [21:28] ask them to remove older versions from internet and write new one [21:29] but why you want to write muxer? you already wrote demuxer? [21:30] no, i didn't even start [21:30] so i'm asking if it makes sense [21:30] as for demuxing, would it work, or i would need to cache data in the demuxer, when waiting for the video data? [21:31] just do seek and tell user that pipe support is fantasy [21:56] 19:12:25 < burek> here is ffmpeg's source code "diagram" :D http://ffmpeg.gusari.org/uploads/gv1.png // hehe i think i know where mpegenccontext is! [21:59] thats pretty neat [22:03] what's the problem with anonymous typedef? [22:03] i don't really see the point of writing the same name twice... [22:04] (and well, that's actually an anonymous struct afaict) [22:08] if it has a name i dont mind [22:08] it's stuff like this that annoys me: [22:08] struct derp { [22:08] int a; [22:09] ubitux, did you see the link to latest doxygen? :) [22:09] uninion {double a; uint64_t b;} [22:09] } [22:09] i.e. completely anonymous with no name [22:09] Action: ubitux likes static const struct { int a, b; } foobar[] = { ... }; [22:09] shame on you [22:09] thats horrible [22:09] why? [22:10] they did exactly what I was thinking it would be the best :) they separated that huge diagram into relevant pieces so while you browse the file structure (menu on the left) you get the relevant diagram for selected file(s) [22:10] burek: no, i didn't open, i don't care that much :p [22:10] oh or maybe i saw that one [22:10] it's ugly, harder to follow, etc [22:10] there is nothing really to follow.. [22:10] Action: Daemon404 also dislieks compound literals [22:17] ubitux : Daemon404 wants to know what the code does [22:17] by giving it a name [22:17] like moving-derp-to-array or something, it makes reading the code easier [22:18] and people can reuse code too [22:18] if they know what the heck its doing :) [22:18] Action: Compn guesses anyways. anonymous, named, makes no diff to me [22:29] saste: reviewing your xface thing atm [22:30] what is xface ? [22:30] i refuse to google it [22:31] a horrible thign from the 80s [22:32] much like stryper [23:01] anyone to review my swf patch? [23:01] and/or the lavfi meta injection through side data? [23:02] saste: btw, don't forget to bump [23:11] Action: ubitux just opened saste.xface and laughed [23:11] it reminds me some double-face mind trick picture... [23:12] without the trick [23:20] ok let's review the gif thing now. [00:00] --- Sun Oct 14 2012 From burek021 at gmail.com Mon Oct 15 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Mon, 15 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121014 Message-ID: <20121015000502.62BF418A01EC@apolo.teamnet.rs> [00:14] ffmpeg.git 03Michael Niedermayer 07f70a651b3f77: sws-test: raise limigts a bit. * 03http://tinyurl.com/9k2gjvk03 [00:14] ffmpeg.git 03Michael Niedermayer 07670b927aa221: ffv1: make sure gob_count is not 0 * 03http://tinyurl.com/9bgzjeu03 [00:14] ffmpeg.git 03Michael Niedermayer 071037e484f0f1: dnxhddata_ Fix mixup of sizeof() and array elements in ff_dnxhd_find_cid() * 03http://tinyurl.com/9trbf5l03 [00:14] ffmpeg.git 03Michael Niedermayer 076581b6cef47e: dnxhdenc: assert ff_dnxhd_get_cid_table() return value * 03http://tinyurl.com/973zc5v03 [00:14] ffmpeg.git 03Michael Niedermayer 07b3eb4f54c0d0: flashsv: check deflateInit() return value * 03http://tinyurl.com/9nd4nyc03 [00:14] ffmpeg.git 03Michael Niedermayer 078cda27b7537d: avcodec_find_best_pix_fmt_of_2: fix handling or PIX_FMT_NONE * 03http://tinyurl.com/9povcfy03 [00:14] ffmpeg.git 03Michael Niedermayer 07f4d73f0fb55e: mpegvideo: check return value of ff_MPV_common_init() * 03http://tinyurl.com/8w5rkpw03 [00:14] ffmpeg.git 03Michael Niedermayer 0720ec0d2a750a: mpegvideo: fix motion_val checks * 03http://tinyurl.com/8otb5go03 [00:27] ffmpeg.git 03Thilo Borgmann 07feaff427c0df: alsdec: fix clipping of weightings for MCC decoding * 03http://tinyurl.com/97s2kyz03 [00:35] nyuhu: ping [00:35] nyuhu: i commented on the kerndeint filter for the FATE thing in cast that's what blocking you (it seemed to be last time you asked) [00:37] oh thx, I was planning on going back with that very soon (currently workning on eq) [00:38] hell yeah eq filter! [00:38] (s/cast/case) [00:39] Ill post the histogram normalization filter tomorrow [00:39] great :) [00:39] :) [00:43] saste: i think Paul's concern was about g_XX[CONSTANT] vs just g_XX[] [00:43] g_XX[CONSTANT] makes think the array needs zero-padding [00:44] while it seems every data you use in the code are explicitely in the arrays [00:44] ah ok [00:44] i used that to document the size of the arrays [00:44] well i borrowed that from the original code [00:44] (or maybe at some point the data was not static, and thus i needed to declare the size or something) [00:45] what does 'g' means? [00:45] global? :) [01:24] ubitux, Guess [01:25] ok :) [02:21] ffmpeg.git 03Cl?ment BSsch 07dff826b4a8d3: lavfi/select/scene: move out convoluted sad variable init from loop. * 03http://tinyurl.com/8oxyvf603 [02:21] ffmpeg.git 03Cl?ment BSsch 070852648301ad: lavfi/select/scene: fix potential overread. * 03http://tinyurl.com/9jf25nc03 [02:21] ffmpeg.git 03Cl?ment BSsch 07096d96ffda8f: lavfi/select/scene: use pointer increments instead of y*linesize. * 03http://tinyurl.com/8gxzsa503 [02:25] right [02:25] michaelni, [02:26] http://fate.ffmpeg.org/report.cgi?time=20121014002519&slot=x86_64-centos-clang-ioc [02:27] thats ansic... gonna move it to c99 for future runs [02:32] ohai [02:32] Daemon404, you need that thing that is under CONFIG_FTRAPV enabled for ioc [02:33] can you be a bit more specific as to what and why? [02:34] there are overflows in the ra144 code, its checked at the end of the loop only [02:35] iam not 100% sure what is correct [02:35] ftrapv is gcc only [02:35] iirc [02:35] im using clang's ioc patches + some extras from mas [02:35] mans* [02:36] maybe rename CONFIG_FTRAPV to CONFIG_AVOID_ARITHMETIC_ANOMALIES [02:36] clang supports ftrapv too [02:36] why is that code ifdef'd out? [02:37] instead of mainlined [02:37] the ftrapv variant is alot slower [02:37] like said iam also not sure what is correct [02:37] vitor said it should be checked per iteration [02:37] let me do a run with ftrapv enabled [02:37] and c99 enabled [02:37] should take ~15 mins [02:39] to just make the non frapv code pass ioc & ftrapv you just need a few carefully placed unsigend and signed casts but according to vitor IIRC that isnt correct [02:39] i see [02:49] http://fate.ffmpeg.org/report.cgi?time=20121014004611&slot=x86_64-centos-clang-ioc [02:49] wow ok [02:49] c99 mode kicked it into gear [02:49] i wonder how much of that is valid [02:50] divVerent: ping [03:00] is there a way to set decoder avoption from the ffmpeg cmd line? [03:11] -thisoption thatparameter ? [03:12] doesn't seem to have any effect :/ [03:12] maybe i was doing it wrong [03:12] anyway, too much coding for today, i'll dig later [03:15] michaelni: btw, any comment on the lavfi/select/scene/perm patch, the swf bitslossless, or lavd/lavfi metadata inject are very welcome if you find some time ;) [03:15] (heh, and even text sub dec now) [03:18] anyway, 'night ppl [03:30] ffmpeg.git 03Michael Niedermayer 07d3d715ff1345: mpeg4videodec: Recalculate timebase in case of guessing time increment bits * 03http://tinyurl.com/8cag3em03 [03:30] ffmpeg.git 03Michael Niedermayer 07b12d92efd6c0: avoid "0xFF << 24" as it is considered a integer overflow in C99 * 03http://tinyurl.com/9latfr603 [03:33] wouldnt that only be an overflow if things are signed? [03:33] ah... right.. [03:33] Action: Daemon404 just actually looked at the patch [03:47] ffmpeg.git 03Michael Niedermayer 072fed05f53a88: avoid more "0xFF << 24" as it is considered a integer overflow in C99 * 03http://tinyurl.com/9z8gzjt03 [03:47] ffmpeg.git 03Michael Niedermayer 07014b178f84fd: g723_1: fix overflow in square_root() * 03http://tinyurl.com/8tk2jqd03 [04:08] ffmpeg.git 03Michael Niedermayer 07555f352f9931: swr/rematrix: fix C99 left shift overflow * 03http://tinyurl.com/8enzvh603 [04:08] ffmpeg.git 03Michael Niedermayer 073ee8eefbf262: sws: avoid signed C99 overflows * 03http://tinyurl.com/8n9gwxt03 [04:08] ffmpeg.git 03Michael Niedermayer 07693097c355cb: paf: avoid C99 overflows * 03http://tinyurl.com/8l4pnlf03 [04:09] Daemon404, are there any overflows left ? [04:10] michaelni, cron is going to run it in 50 mins from now [04:10] well see then [04:10] too lazy to connect to work vpn ... [04:11] Daemon404 : yeah but michael is in the mood to fix them now [04:11] gotta work with the master devel [04:11] if you want things worked on quickly ;) [04:11] this sounds dirty [04:11] well i just wanted to fix all before going to bed :) [04:12] michaelni : you work too hard [04:12] ehe [04:12] so all the dpx samples we have decode correctly ? [04:12] are we still updating the changelog ? [04:13] all the samples in that zip or what it was should [04:17] tomorrow im going to try and set up an instance with the intel race condition thingy [04:17] depending on how slow it is [04:32] uh oh [04:33] oh its just make segfaulting [04:49] Daemon404, maybe you could wrap make in something the retries when make segfaults [04:51] possible [04:51] msys make is really shitty [04:51] that's nevcairiel's fate instance, though. [05:01] ok... [05:02] so clearly i have no idea what im doign re: cron [05:02] it didnt run [05:02] i guess it helps if crond is running. [05:14] michaelni, well it looks like most are in swscale.c now [05:14] if not all [05:14] re: failures [05:14] (if youre not asleep) [05:15] Daemon404, i dont sleep, its waste of time [05:16] lul [05:18] sleep is a luxury most people can't afford [05:21] ffmpeg.git 03Michael Niedermayer 077a32ab5ed0b0: sws: get rid of C99 signed overflows try #2 * 03http://tinyurl.com/9b9ghl203 [05:34] only 2 left... by why arent all 1255 test being run? [05:35] only 1149 [05:36] ah. no seek tests due to lavf failures [05:53] ffmpeg.git 03Michael Niedermayer 072217a2249dd7: dpxenc: fix signed c99 overflows * 03http://tinyurl.com/9boo5ok03 [05:53] ffmpeg.git 03Michael Niedermayer 07d31098113fab: mp3enc: fix signed C99 overflow * 03http://tinyurl.com/995s29n03 [06:16] and ioc is all green now... [09:52] michaelni: dnxd can assert in init, because table have holes and probe does not check for this [09:52] *dnxhd [12:45] ffmpeg.git 03Paul B Mahol 07e2bf1082cc0d: svq1dec: return more meaningful error codes * 03http://tinyurl.com/9z7enxe03 [12:45] ffmpeg.git 03Paul B Mahol 078f4020d8a4b2: svq1dec: check return value of get_bits1() in more common way * 03http://tinyurl.com/9ej6fca03 [12:45] ffmpeg.git 03Paul B Mahol 077b9fc769e40a: svq1dec: use more common way to check if result is not zero * 03http://tinyurl.com/9nj9avg03 [12:45] ffmpeg.git 03Paul B Mahol 07dd5aff001e1e: svq1dec: use log level that have more common sense * 03http://tinyurl.com/8ptx4qp03 [12:45] ffmpeg.git 03Paul B Mahol 07039341eb4394: wv: do not report invalid stream duration * 03http://tinyurl.com/9fjq7qf03 [12:45] aww [12:46] sorry, svq1dec commits should not be there.... [12:51] durandal_1707, I'm used to to: git push --dry-run [12:52] then I check with git show HASH [12:52] i put all stuff in same branch [12:52] and finally i commit, that's boring but sometimes it prevented me to screw up things [12:53] anyway svq1dec changes are cosmetics [12:53] i have several branch for each "topic", then i usually cherry pick to master before pushing [12:53] yes, they don't seem to be harmful at all [12:54] before using git I was using quilt, and I had a single stack where I put all the patches [12:55] and rebasing was a huge PITA every time [14:09] ffmpeg.git 03Michael Niedermayer 0703760297b1d3: doc/nut: Clarify what/where is the official nut specification * 03http://tinyurl.com/8ngbgvc03 [14:12] michaelni did you merged libav nut changes? [14:17] durandal11707, not yet, why do you ask ? [14:18] i thought it was pushed and you skipped it [14:20] anyway the changes dont look like a terribly good idea, what are they good for anyway ? [14:21] they added tags for planar and float raw codecs [14:23] sure but whats the point of the other changes ? [14:23] like? [14:24] changing order of preferrance of tags [14:24] also i have my doubts their code works as is [14:36] ffmpeg.git 03Luca Barbato 071bd442c276e6: nut: prioritize native tags * 03http://tinyurl.com/933rmlw03 [14:36] ffmpeg.git 03Michael Niedermayer 07db51c65961bd: Merge commit '1bd442c276e6688b43777a198cad0d7e3a92123f' * 03http://tinyurl.com/8zhqqj303 [14:39] michaelni: how do you merge code? [14:41] magic [14:42] not magic at all, in many cases you undo you own change, instead you should merge by hand conflicts [14:46] now sure what you mean ? [14:46] noT [14:48] see dc945b1fa8ae65a <- that is not best solution neither is commit after that that use b2 but you got the point [14:49] b2 commit removed your code (which was not wrong in such case that is should be removed....) [14:49] ffmpeg.git 03Luca Barbato 0792281850a2d8: nut: support pcm codecs not mapped in avi * 03http://tinyurl.com/92hvu6m03 [14:49] ffmpeg.git 03Justin Ruggles 0710e645e9cb06: lavr: handle clipping in the float to s32 conversion * 03http://tinyurl.com/9rrl3l903 [14:49] ffmpeg.git 03Mans Rullgard 0774c39bc68271: eval-test: make table static const * 03http://tinyurl.com/9y8hbw703 [14:49] ffmpeg.git 03Mans Rullgard 07ac17ccf73ad0: configure: work around bug in ash shell * 03http://tinyurl.com/9txhvbk03 [14:49] ffmpeg.git 03Mans Rullgard 07741a8b724e47: configure: recognise Minix as OS * 03http://tinyurl.com/8lnnw9y03 [14:49] ffmpeg.git 03Mans Rullgard 07b5198a2637b7: configure: tms470: add mapping for -mfpu=vfpv3-d16 flag * 03http://tinyurl.com/9fnwjxj03 [14:49] ffmpeg.git 03Michael Niedermayer 0782c0055c5e71: Merge commit 'b5198a2637b7b45b0049a1d4b386a06f016f2520' * 03http://tinyurl.com/8apw9bm03 [14:53] what do you mean by b2 commit ? [14:54] 1255eed533b4 [14:59] ffmpeg.git 03Paul B Mahol 07de3b1116dab1: eatgq: fix small overread * 03http://tinyurl.com/9stwpqx03 [15:06] michaelni, anyway nut changes should be discussed on nut-devel [15:06] saste, yes [15:07] i don't think libav has the right to properly change a format which was not designed by them [15:07] did someone tried to contact them about the issue? [15:08] libav doesnt talk with me when my name appears in From: so no i didnt try [15:09] doesnt talk with me NORMALLY (there are exceptions) ... [15:09] anyone is welcome to contribute... meh [15:09] enough for today... [15:14] michaelni: there is code in mov/avi that calls dvi_produce_packet but it does not give any audio [15:14] perhaps it worked at some time and then stopped [15:14] ffmpeg.git 03Mans Rullgard 075ab432fa3500: configure: improve tms470 compiler usage with glibc * 03http://tinyurl.com/9uuxqbb03 [15:14] ffmpeg.git 03Mans Rullgard 07b6f8d635f27b: build: tms470: work around glibc math.h problems * 03http://tinyurl.com/9vpagga03 [15:14] ffmpeg.git 03Diego Biurrun 0752d113ee06ba: avutil: Rename ff_set_systematic_pal2() ---> avpriv_set_systematic_pal2() * 03http://tinyurl.com/8n6yw6903 [15:14] ffmpeg.git 03Diego Biurrun 077638f0b2fef0: avutil: Do not make ff_ symbols globally visible. * 03http://tinyurl.com/9zfp45903 [15:15] ffmpeg.git 03Michael Niedermayer 07d6f6a7557c8d: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/8mddqmd03 [15:15] durandal_1707, possible [15:37] michaelni: actually avi sample have wav tag [15:43] vlc does not support dts in mov... [15:44] vlc is one big mess [15:45] michaelni: the only way i see to support such old avi files is to let avi demuxer read wav tags [15:46] but is there any wav file with ulead audio..... [15:46] durandal_1707, which software is not a big mess? [15:48] saste: ffmpeg [15:49] durandal_1707, less than other software, but still a mess ;-) [15:50] saste: where is it? [15:50] many places, do you want specific examples? [15:50] yes [15:51] not that i want to blame anyone for that, in my philosophy being messy is a natural state for software (in rapid development, and with a long story such as ffmpeg) [15:52] my favourite: libavcodec/imgconvert.c and how things are split between libavcodec and libavutil with regards to pixfmts [15:52] color space handling in libswscale [15:52] ah the new entry, pseudo_pal in libavutil/pixdesc (that hack doesn't belong to lavu) [15:53] but overall i agree, ffmpeg is pretty clean in comparison with most software projects [16:04] ffmpeg.git 03Peter Ross 070ea55365b91c: ansi: by design nb_args may exceed the size of the args array, so guard accordingly * 03http://tinyurl.com/97yxcg803 [16:04] ffmpeg.git 03Peter Ross 076bf43441f0a7: eatgq: raise error on unsupported macroblock mode * 03http://tinyurl.com/8detkr603 [16:45] michaelni: lavfi/buffer.c: line 225 gets number of channels from channel layout [16:46] this crash if avctx->channels is smaller number [16:51] ffmpeg.git 03Hendrik Leppkes 07572781b25f98: h264: fix parsing of old lossless profile (profile_idc == 144) * 03http://tinyurl.com/9gpmwmn03 [19:30] ffmpeg.git 03Michael Niedermayer 0745bd0d15f453: lavf: fix fps detection with PAFF H.264 * 03http://tinyurl.com/9g7rt6n03 [21:07] ffmpeg.git 03Thilo Borgmann 0741bf943f709e: bgmc: fix sizeof arguments (should fix CIDs: 700724 and 608084) * 03http://tinyurl.com/94okot703 [21:07] ffmpeg.git 03Michael Chinen 07c73b1a1d8010: flac_parser.c: fix issue with looping output for small files. * 03http://tinyurl.com/8fk9j8603 [21:34] Compn: freaking about? [21:46] j-b : dvaudio decoder :P [21:47] how so? [22:01] someone said it was committed to vlc [22:01] not important [22:01] I think it was, but quite some time ago [22:20] ffmpeg.git 03Michael Niedermayer 0793ef29b6f47e: noise_bsf: fix division by 0 * 03http://tinyurl.com/928qbgd03 [22:20] ffmpeg.git 03Michael Niedermayer 07a6cac64f69bf: compute_lpc_coefs: assert that normalize and fail have a supported combination * 03http://tinyurl.com/8tmt9h303 [22:20] ffmpeg.git 03Michael Niedermayer 070de0ae5a20cf: tag_tree_decode: check node being non null * 03http://tinyurl.com/8uz5wm203 [22:20] ffmpeg.git 03Michael Niedermayer 071e6cbb01e911: j2kdec: make curtilenum integer. * 03http://tinyurl.com/9kytlkh03 [22:20] ffmpeg.git 03Michael Niedermayer 07bc077ca6b5f8: j2kdec: remove unneeded operation * 03http://tinyurl.com/997mm7903 [22:37] it would be nice to have a teletext decoder.. [22:38] rofl [22:38] ? :( [22:39] complete teletext decoder is hard [22:39] see libzvbi [22:39] is that what you use @vlc? [22:40] ubitux: yes, because a complete decoder was too much of a mess [22:40] ok [22:40] only for teletext? [22:41] decoding just teletext subitles is not that hard [22:41] complete teletext pages is ... [22:43] the subtitles is the important part for most people [22:48] Action: ubitux adds a new TODO list entry [22:48] i'll never reach peace :( [22:50] we would all be bored if we had empty lists [22:51] sure, 2-3 entries is fine [22:51] 90-100 isn't :( [00:00] --- Mon Oct 15 2012 From burek021 at gmail.com Mon Oct 15 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Mon, 15 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121014 Message-ID: <20121015000501.5C29518A01E8@apolo.teamnet.rs> [00:10] burek: regarding the ticket https://ffmpeg.org/trac/ffmpeg/ticket/1808, i am unable to convert test.avi with the "-filter:v yadif=1,format=yuv420p" filter. I still get nable to parse option value "-1" as pixel format [00:10] Error opening filters! [00:10] s/nable/unable [00:12] burek: http://pastie.org/5054514 [04:51] relaxed: http://pastie.org/5054514 would you mind taking a peak at that and double checking it for me? [05:31] does -g even exist in the current code base [05:31] wtf is a group of picture size [05:32] keyframe related? [05:32] what's the best ay to force 1 frame rate key frames [05:32] so in this case 30 seconds [05:33] need to scrub the video in adobe flash with some reasonable precision [05:35] okay so gop is analogous to i-frame [05:36] http://ffmpeg.org/pipermail/ffmpeg-user/2011-September/002349.html ? [05:48] hi, I'm trying to use ffmpeg to convert the audiostream from any given file into .wav floating point sample format [05:50] my command line is: ffmpeg -i input_file.ext -map 0:0 -af "aconvert=flt:auto" -y foo.wav [05:50] but it's always converting to s16 [05:50] any ideas? [05:54] even trying u8 for the format doesn't affect it.,. it's almost as if ffmpeg believes .wav only supports s16 [06:24] pastebin: http://pastebin.com/fQyqYgXH [06:50] hi. is anyone familiar with dvdauthor? [07:03] is it possible for ffmpeg to transcode (or post process) to a "streamable" mp4? [12:55] Does ffmpeg have a quick and easy way of detecting and adding the padfilter? Trying to convert a 2.35:1 (letterbox) movie file to DVD and I can not get the padfilter correct for the life of me [12:56] I_Died_Once, no you need a two steps process [12:56] would like to convert said movie to DVD without skewing the video in the process [12:57] "detecting and adding the padfilter" what do you mean? [12:57] saste, thanks for a responce! [12:57] i'm trying to convert this letterbox file to regular old DVD without messing up the video [12:58] saste, can you give me some sort of an example on how you would go about it ? [12:59] you basically want to preserve the aspect ratio or what? [12:59] you can find examples in the ffmpeg manual, in the pad section [12:59] yeah... i was trying to instert the pads on top & bottom [12:59] scripting may also be in order [12:59] I keep jacking up the video and stretching [13:00] I'm over here thinking "it cant be THIS difficult" [13:00] I_Died_Once, command and output, i don't even know what you're trying to achieve [13:01] I'm only wanting to achieve transferring this video to DVD format without skewering or stretching the video [13:02] which is turning out to be a massive pain [13:02] I_Died_Once, pad should be enough [13:03] but you need to master the expression evaluation in ffmpeg [13:03] that may be a pain [13:03] agreed [13:03] but there are already several related examples which should get you a start [13:03] thats why I was hoping for help with the exact command, i cant get it right [13:04] i can't read you mind (yet), don't know what you want to exactly achieve [13:05] also you should show some effort, if you want people help you [13:05] aww man that hurts, i've wasted many gigs trying to get this correct to no avail so far [13:06] people with warped looking heads [14:56] hi all [14:56] how to convert flv to mp3? [15:06] bigmeow: is the audio track in flv already in mp3? [16:08] Unknown encoder 'xvid' [16:08] anyone knows how to get a list of valid encoders for video options? [16:26] shevy: ffmpeg -codecs [16:26] external codecs are prefixed with "lib"- libxvid [16:27] ah ok cool, thanks [16:27] ubitux: how to know if the audio track is in flv? [16:27] http://superuser.com/questions/487599/why-i-cannot-convert-flv-format-to-mp4-formate-using-ffmpeg [16:28] ffprobe [17:14] heheh ffmpeg makes my cpu work Cpu(s): 98.3%us [17:49] tried to convert 320x200 Stream #0:0 -> #0:0 (zmbv -> libx264) using placebo preset Q=32 and there is fair amount of noise at moving parts, and the size is about 1/3 only [17:49] it is recodring from dosbox [17:51] full command [17:52] bigmeow: ffprobe in.flv, if you see mp3 ? ffmpeg -i in.flv -c:a copy out.mp3, otherwise you need to re-encode: ffmpeg -i in.flv -q:a 0 out.mp3 [17:52] ffmpeg -i ./game_003.avi -vcodec libx264 -preset placebo game_003.mkv [17:53] add -crf 18 [17:54] libx264 writes crf=23 [17:54] that is too high? [17:55] Seems to be, doesn't it? -tune animation may be beneficial as well. [17:58] thanks [18:19] trying out animation tuning [18:21] crf18 is better but since its redalert, only a very few parts are moving... [18:38] I have read this http://ffmpeg.org/trac/ffmpeg/wiki/Streaming%20media%20with%20ffserver . So, there is a way to feed server from ffmpeg with TLS? [18:59] alright, hopefully I can get some help this time [18:59] http://pastebin.com/7rwc6bDT [19:00] long story short - I have a letterbox movie with an aspect ratio 2.35:1 [19:01] I am wanitng to put that on a standard ntsc dvd disc, but i dont want to distort the image [19:01] dont want to distort and skewer the video when putting it to DVD [19:13] I_Died_Once<< you were wasted? ;/ [19:13] heh, yeah [19:14] has anyone ever needed to convert a 2.35:1 video to DVD before? [19:16] I_Died_Once: sure [19:16] and now you want the command line to do that ? [19:16] I'd love you for it [19:17] I keep getting distorted looking heads [19:18] ffmpeg -i '/Dimension5150/movie.mp4' -vf pad=1280:720:0:92,setdar=16:9 -target ntsc-dvd movie.mpg [19:19] crf=18 h264 mkv is 9MB vs 32MB zmbv avi [19:19] with lots of standstill area [19:19] ;/ [19:22] cbsrobot_ - thats nice, but the image is on the left hand side of the screen [19:22] let me get a screenshot to show [19:23] http://i.imgur.com/Ca73T.png [19:24] I_Died_Once, so you want to center the image? [19:24] yeah [19:26] did you try this: pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2" [19:27] trying now [19:27] Error when evaluating the expression 'ih*16/9/sar' [19:28] and the error is? [19:28] the next line after that is: Error opening filters! [19:28] then i get my command prompt back [19:29] http://pastebin.com/1HUG6xkA [19:30] no encoding takes place [19:31] can you update your ffmpeg? [19:31] "sar" variable is not supported in your build [19:31] checking [19:31] I'm running debian linux [19:33] says I have the lastest version installed.... version 6:0.8.3-7 [19:33] try pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2" [19:33] works in case you have sar=1/1, which cannot be assumed [19:34] alright, its coding [19:34] also debian ships libav fork, which is not proper ffmpeg [19:34] check !wiki [19:35] that last pad option stretches the video [19:35] allright, then you should try an updated version [19:36] sar variable was added more than one year ago [19:36] ok, ok... [19:36] also aspect ratio is doing funny things [19:41] I_Died_Once, also the version shipped by Debian is usually at least one year/six months behind the main development trunk [19:41] gotcha [19:41] I'm updating and getting a newer version from debian-multimedia [19:41] they have7:1.0-dmo1 [19:42] should be ok [19:42] from their unstable branch [19:42] (oohh unstable!) [19:42] in the wiki you find compilation instructions, it's not that hard [19:43] i apreciate the help, seriously [21:02] hey, I want to record from the rtsp source, but I want to save the video to files like file_minute1.mp4 file_minute2.mp4, etc - so every minute I want ffmpeg to record to a new file. I tried to just do a while loop with ffmpeg recording for 1 minute duraction, but that creates gaps. Any ideas? [21:03] what does this error mean? [21:03] Unable to parse option value "-1" as pixel format [21:08] I'm trying to do: ffmpeg -i rtsp://192.168.88.11/ -vcodec copy -acodec copy -f segment -segment_time 60 -y "test-%03d.mp4" - but that shows "Output file #0 does not contain any stream" -- If I replace -f segment with -f mp4, it works but the output is not in segments. Any ideas? [21:16] ah, I was missing -map 0 [21:56] ffmpeg -i ./game_003.avi -vcodec libx264 -preset placebo -crf 18 -tune animation game_003_placebo_crf18_anim.mkv [21:56] hmm, stills are good, but the grass is still "moving" a bit after the moving units [21:57] and 10MB h264 mkv vs 32MB avi... [21:59] I am trying this command [21:59] ffmpeg -f alsa -i default -f v4l2 -s 640x480 -i /dev/video0 -acodec flac -vcodec libx264 out.mkv [21:59] but no audio is played when i `mplayer out.mkv` [22:01] default may be muted or wrong [22:01] microphone and line in is muted by default... [22:05] ffmpeg -f alsa -i hw:6,0 -f v4l2 -s 640x480 -i /dev/video0 -acodec flac -vcodec libx264 out.mkv [22:05] hw:6,0 is my correct input. [22:06] still no audio [22:11] lake<< tried unmuting and selecting alsa mixer input ? [22:13] redalert short video using above command line http://www.sendspace.com/file/gzldee [22:16] http://pastebin.com/fQyqYgXH [22:16] hi, I'm trying to use ffmpeg to convert the audiostream from any given file into .wav floating point sample format [22:16] my command line is: ffmpeg -i input_file.ext -map 0:0 -af "aconvert=flt:auto" -y foo.wav 10:50:02 PM [22:16] but it's always converting to s16 [22:17] even trying u8 for the format doesn't affect it.,. it's almost as if ffmpeg believes .wav only supports s16 [22:17] any ideas? [22:23] duvnell2: -c:a pcm_f32be or pcm_f32le maybe? [22:24] unfortunately i don't know if you can put float in a wav [22:24] at least our muxer doesn't seem to know how to [22:24] but you can generate raw pcm in float [22:26] against boredness, redalert short video, with a little hack in memory http://www.sendspace.com/file/gzldee [22:26] peek&poke [22:54] is the libavcodec(-dev) that ships with Debian old as the mountains? [22:55] Honestly, it's so old it doesn't even have the AVPacket struct [22:55] anyone able to confirm? [22:56] you should complain to debian then :p [22:56] I am actually :P [22:56] and btw, debian is not packaging ffmpeg [22:56] but the fork [22:56] Hi guys. I've an mkv with an x264 stream, but my projector doesn't read the file. How can i convert to avi mpeg4 without loosing quality? [22:56] I actually have ffmpeg [22:56] on Debian [22:56] right atm [22:57] garulf: you'll likely loose some quality, and increase file size [22:57] you can try -q:v 0 [22:59] does all this mean I have to checkout ffmpeg? [22:59] ubitux, ok thanks I will. I don't mind loosing a bit of quality, but I'm tried to convert without any option and the output quality is very poor.. [22:59] barque: what do you expect us to do? [23:00] oh nothing :D [23:00] :) [23:00] I'm wondering if a checkout'll help :] [23:00] git/svn/whatever [23:01] yes, git :) [23:05] actually yeah they have more up to date headers and libs on a quote non-free repo unquote [23:06] I guess ffmpeg is not free [23:06] as in libre [23:06] cough narcissists cough [23:08] non-free repo? [23:11] yeah they have to separate what they don't consider to be completely "free" software from free software that is tainted with non-free components [23:12] s/don't/do [23:12] i don't understand your apparently troll joke :p [23:13] lol it is a joke, but I'm not trolling [23:13] you don't remember the whole firefox and iceweasel thing? [23:13] sure [23:13] just because the firefox logo was non-free they made their own branch and called it iceweasel [23:13] I think regarding ffmpeg they see the same distinction or something [23:13] right [23:14] or maybe a whole bunch of stuff they wrapped together and put it on a multimedia repository [23:14] oh right i think i get it [23:14] under the 'non-free' category [23:14] well when i was talking about the fork [23:14] i wasn't talking about debian packaging ffmpeg under a different name [23:14] i mean they are packaging a real fork (called libav) [23:14] (and under the ffmpeg name) [23:15] yeah something or other [23:16] I bet they're hosting the non-free stuff from a server in Poland [23:16] that you can only reach after a long train ride [23:16] Action: barque grumbles [23:19] I'm honestly sorry if I offended anyone with that joke [00:00] --- Mon Oct 15 2012 From burek021 at gmail.com Tue Oct 16 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Tue, 16 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121015 Message-ID: <20121016000502.7DBF218A01EC@apolo.teamnet.rs> [00:45] ffmpeg.git 03Michael Niedermayer 07a63d7213b3eb: ffplay: Prevent 0/0 aspect from being passed on to lavfi * 03http://tinyurl.com/8touvp203 [00:45] ffmpeg.git 03Michael Niedermayer 071b8f4d773c4d: ffmpeg: use framerate not fieldrate as filter input. * 03http://tinyurl.com/9lkfa4603 [01:27] /home/daemon404/build_ffmpeg/src/tests/fate-run.sh: line 68: 28651 Illegal instruction $target_exec $target_path/"$@" [01:27] wtf? [01:27] is this just me? [03:04] Guys, I'm getting the same R and G through libavcodec PIX_FMT_RGB24 [03:04] command line or your own program ? [03:04] my own [03:05] I should try cmd line ... I guess .. ? :P [03:05] makes it much easier to test for us [03:05] :) [03:06] true, I'm gonna review the code about 10x just to make sure I'm seeing this right [03:06] no one has reported something similar have they? [03:07] oh my god [03:07] this is NOT a problem of yours, but of mine [03:07] Action: barque crawls into a dark corner [03:08] with all this code I was bound to make a mistake anyway :P [03:08] first run is never perfect [03:44] poor canaidans :) [03:44] so polite [03:44] canadians* [03:57] ffmpeg.git 03Michael Niedermayer 07151469db3379: j2kenc: remove unneeded dereference * 03http://tinyurl.com/9hppmd603 [03:57] ffmpeg.git 03Michael Niedermayer 0735782bfbca0a: h264: Change asserts to av_asserts in ff_h264_fill_default_ref_list() * 03http://tinyurl.com/9o6wehf03 [07:38] nyuhu: so you like fist review? [08:11] Daemon404: ping [08:12] any objection for the webm patch according to Nicolas' comment? [08:17] ubitux, confirm with the spec then ok [08:17] im sleeping [08:44] ffmpeg.git 03Stefano Sabatini 07cd559bb49a34: lavc: add xface image decoder and encoder * 03http://tinyurl.com/brguubo03 [08:44] ffmpeg.git 03Stefano Sabatini 07304c37b216d7: tests: add fate-xface test * 03http://tinyurl.com/cf2wowx03 [08:55] Daemon404: ok :) [08:56] yay xface. [09:20] ubitux: I am back [09:20] and I got your messages [09:22] ffmpeg.git 03Cl?ment BSsch 07b08273c9ca8b: lavf/mkv: avoid negative ts by default. * 03http://tinyurl.com/cn5epy503 [12:19] michaelni: ping [13:02] ffmpeg.git 03Paul B Mahol 074f5e5a05132b: lavf/txd: cosmetics: fix identation * 03http://tinyurl.com/9ock66803 [13:14] This program is based on qt-faststart.c from the ffmpeg projects .... [13:28] TimNich, pong [13:32] michaelni: quicktime is still not right yet, b8d6455 is better but there are still quite a few places where there are wrong values in the mood atoms. See http://pastebin.com/VxHrYdx9 [13:33] s/mood/ moov/ [13:35] if you do a probe you find an error line saying "Timecode frame rate 0/1 not supported [13:36] TimNich, which values exactly are wrong ? [13:36] the diff shows which are different not which are wrong ... [13:38] The full dump is the good one, and in the diff -- good ++ bad [13:40] you mean removing code is good [13:40] and adding code is bad? [13:40] Action: Compn trying to parse TimNich's sentence [13:40] remember, english isnt everyone's first language here, so be specific [13:42] Compn: If you looked at the diff you would see good lines removed, bad lines added. Appreciate the point about non native English speakers, but in context with the diff in front of one I would have thought it made sense... [13:42] i guess he ran diff -u good bad [13:42] timescale 12800 is weird [13:43] TimNich : i dont even know what mail you are talking about. i'm just in observer mode here [13:43] TimNich, well, you want av sync then good is not good at all [13:43] Compn: [13:43] i see the pastebin now tho [13:44] good in terms of what used to work.. the latest version crahses Quicktime player and Avid... [13:44] so you're specifically talking about mov muxer ? [13:45] If there is an inherent AV sync issue then thats been around a while&. [13:45] yes [13:47] Compn: yes, usual story , two steps forward one step back. Sounds like michaelni may have unearthed a deeper issue though.. [13:48] TimNich, do you have some clue what is exactly causeing the problem ? [13:48] i mean does QT fundamentally not support such timebases or is it something else ? [13:53] From a three way diff I wonder about the timescale: 12800 etc in the mdhd atom of the timecode track as this was the same in both previous versions which at least play after a fashion in QT player [13:57] michaelni : do we have any samples that look like what mov muxer outputs ? [13:57] e.g. 12800 timebase [13:57] i'll look at who changed what in mov muxer in a minute [13:57] :P [14:00] Compn: We know exactly which commits introduced the issue. Its a case of what mechanism is at work.. [14:00] It would be interresting to have a working mov sample with fine timebase for the video and a timecode track [14:00] big buck bunny has a timecode track iirc [14:00] (set to 00:00:00:00 though) [14:01] what do you mean by "fine" I can knock up anything quicktime Pro can produce.. [14:02] TimNich, with fine i mean for example a 25fps video that has a timebase smaller than 1/25 [14:02] like a 1/1000 timebase or 1/250 or something [14:04] michaelni: Not sure what would have that, I can't see QTPro bothering to set the tiembase smaller than it thinks is needed... [14:10] michaelni: Just looked at a QTPro file and timebase is 25... [14:13] However I notice an apparent inconsistency in the bad2 headers. In the last stud atom the frame duration is listed add .000078 secs and the number_of_frames 0, both the others have .04 frame_duration [14:16] michaelni: Dave Rice has a laaarge collection of a lot of mov/mp4 samples with timecodes [14:16] you might want to contact him [14:16] (more than 1TB of samples) [14:17] he is sometimes on IRC btw [14:24] ffmpeg.git 03Justin Ruggles 078441909f49d1: lavc: update documentation for AVFrame.extended_data * 03http://tinyurl.com/924p3bv03 [14:25] ffmpeg.git 03Justin Ruggles 07977eb7d567f6: shorten: use planar sample format * 03http://tinyurl.com/9zebe7c03 [14:25] ffmpeg.git 03Justin Ruggles 077ebfe5b44a55: wmadec: use float planar sample format output * 03http://tinyurl.com/9st5ua603 [14:25] ffmpeg.git 03Justin Ruggles 072725ce7c7c34: wmalossless: output in planar sample format * 03http://tinyurl.com/9gt236k03 [14:25] ffmpeg.git 03Justin Ruggles 07f9d732c26448: wmapro: use planar sample format * 03http://tinyurl.com/9vf2xp303 [14:25] ffmpeg.git 03Mans Rullgard 07e98b02de5f92: configure: check for mprotect * 03http://tinyurl.com/9vxu5mk03 [14:25] ffmpeg.git 03Mans Rullgard 0795cd815c3663: swscale: try to use mmap only if available * 03http://tinyurl.com/9geu84403 [14:25] ffmpeg.git 03Michael Niedermayer 07ae237a117acb: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/9ylrthd03 [14:48] kierank: you're aware there is a AVDict for metadata in the AVFrame in ffmpeg right? [14:49] (see f49ec1b4) [14:51] hee is writing it for libav [14:51] ffmpeg.git 03Paul B Mahol 07916e40b5b42f: smacker: return more meaningfull error codes * 03http://tinyurl.com/8lfhlpa03 [14:52] durandal_1707: yes that's why i'm just saying it's already available and it might be interesting to keep a few things in common [14:52] again, that's a feature we have since one year now [14:54] nope, since Jul [14:54] ah my bad, commit date/author date [14:55] i still am unsure frame accuracy will work with threading [14:55] maybe it's ok with a dictionary [14:55] but with a buffer i got a lot of crashes [14:56] why would you use a buffer? [14:56] to export closed captions [14:57] and arbitary user data [14:57] any reason not to have it in packet side data? [14:57] because it needs reordering [14:57] ok [14:57] anyway, if you add some !dict metadata, please don't call it "metadata" [14:58] because dict metadata are quite useful [14:58] we can deal with it without breaking abi (we can replace the getter), but it's a pain [14:59] kierank: you're trying to support cc & teletext? [14:59] well i just want arbitrary user data [14:59] exported [14:59] the problem with a dict is that it works only with strings [15:00] yup [15:00] unless you put some base64 in it [15:00] but i suppose you're not that mad [15:10] ffmpeg.git 03Michael Niedermayer 075bac83dae84a: shorten: fix U8 to be planar too * 03http://tinyurl.com/8qu2ex503 [15:10] ffmpeg.git 03Michael Niedermayer 075c7a62aef59e: movenc: fix tmcd parameters * 03http://tinyurl.com/9j6okep03 [15:12] TimNich, can you retry if qt likes the new movs better ? [15:12] making it now.. [15:28] michaelni: We're good on QT7 and Avid [15:28] Thanks [15:30] ffmpeg.git 03Paul B Mahol 075864fe62efa4: flacdec: replace -1 with AV_SAMPLE_FMT_NONE * 03http://tinyurl.com/98x2df603 [15:30] ffmpeg.git 03Paul B Mahol 07e75357ea7c8c: xbmenc: cosmetics: reindent * 03http://tinyurl.com/9x2bm2703 [15:35] Tjoppen, TimNich, mateo` "avio_wb16(pb, mxf->tc.start + frame); // continuity count" <-- the sum does not fit in 16bit [15:35] what is the right fix for this ? [15:36] (testcase is simple make fate-lavf-mxf) [15:36] mxf->tc.start is 264363 [15:36] ffmpeg.git 03Stefano Sabatini 07e56b3a5ebb12: examples/muxing: fix case inconsistency in message * 03http://tinyurl.com/dxnbogu03 [15:37] ffmpeg.git 03Stefano Sabatini 076133149e0c70: examples/muxing: add missing error checks * 03http://tinyurl.com/d4zmd6903 [15:37] ffmpeg.git 03Stefano Sabatini 0759e31aa2cafb: examples/muxing: provide more information in case of avcodec_open2 failure * 03http://tinyurl.com/cpzzj2n03 [15:38] michaelni: not a problem, IIRC [15:43] yep, it's supposed to overflow. it was used for detecting mising content packages during transmission (S326m-2000) [15:47] Tjoppen, ok if i add a & 0xFFFF ? [15:54] yes [16:07] ffmpeg.git 03Paul B Mahol 07011f3892ff35: h264: use designated initializers for AVClass * 03http://tinyurl.com/93qaqcv03 [16:10] can someone comment on https://ffmpeg.org/trac/ffmpeg/ticket/1801? [16:10] what's the proper way to setup DTS anyway? [16:12] saste_: wait before pushed the lavc/utils patch [16:12] i think av_get_sample_fmt_name can return null [16:13] and some libc have some issues with it [16:13] (like crashing instead of "(null)" or similar) [16:13] ubitux, we have a lot of that [16:13] yeah i know :( [16:14] well whatever [16:14] anyway at that stage sample_fmt should be set to a valid value [16:14] do as you wish :) [16:14] i can do an av_x_if_null() [16:19] why is avio_wl* so stupid? [16:20] yes not using multiple avio_w8 is not making things much faster but still..... [16:22] this is sick [16:22] ffplay -f lavfi "flite=text='I am left':voice=rms[left]; flite=text='I am right'[right]; [left][right]amerge" [16:22] I can hear "left" or "right" according to what I expect to hear [16:27] now make it speak stuff from av_log [16:28] you don't really want that [16:28] gawd, this is such a hack [16:29] I have a simple patch series that adds parsing field order to the DV decoder. is this of interest? [16:31] i'm interested [16:32] it's a massive one-line patch [16:32] Tjoppen: Not sure, what issue does it solve? [16:32] plus another one before it to move a piece of code up a few lines :) [16:32] it exposes the field order (FS flag). that's it [16:33] currently bff is assumed? [16:33] dv encoder might want a similar fix, but I don't really care about that [16:33] yes [16:33] the fix for the encoder would be to put !top_field_first << 6 in the relevant byte [16:33] PC3 of the VSC pack [16:33] who can explain this: http://pastie.org/5062307 [16:34] sample_fmt is not the one shown when encoding [16:34] I'd like to know it had done it (AV_LOG_WARNING) [16:37] saste_: isn't that normal for mp3? [16:37] Actually I think I would really like field_order to be one of those AV_LOG_INFO bits like frame rate etc that utils spits out at various moments&. [16:40] I have a rather hacky patch for #241 and duplicates [16:43] TimNich: uhm, ffprobe -show_streams [16:43] or show_frames [16:44] Tjoppen: does your fix deal with the FF/FS value in the VSC pack ? [16:44] I considered FF. not sure what to do if it's zero [16:44] Tjoppen: i think the encoder should be fixed too if i remember well [16:45] prolly [16:45] anyway i should do that on the encoder part when i'll get some time [16:45] you know what? I'm going to look into that too. FATE passes with extracting FS [16:46] ugh, AV_FIELD_TB/BT [16:47] Tjoppen: -show-frames, but thats OK for a prodding an unknown stream. If it was added into the other output on a per file basis then it could be checked fro in the logs of batch jobs without the extra pass of a probe... [16:47] I don't see what this has to do with the dv decoder [16:47] that's a CLI issue [16:48] It doesn't. But my previous comment got me thinking about my more generalised wish list&. [16:49] Agree with the AV_FIELD_TB/BT but that's Apple's nomenclature, and confusing it is too... [16:51] mateo`: http://pastebin.com/k2chCe38 [16:51] TimNich: nm, I just read it from AVPicture instead [16:56] Tjoppen: there are too many places too look... [17:00] TimNich: there appears to be two places. but: the TB/BT stuff don't apply to raw frames [17:05] Well the current spec mandates the use of the fiel data in that format for uncompressed video, and michaelni said it should get it from the avctx->field_order [17:06] AVOicture doesn't actually seem relevant, did you mean AVFrame? [17:13] AVFrame picture; [17:13] right [17:14] isn't AVFrame.top_field_first set from AVCodecContext though? [17:14] ffmpeg.git 03Paul B Mahol 073f8148911c6e: img2dec: check return value of av_new_packet() * 03http://tinyurl.com/8c2dygy03 [17:17] Tjoppen: not that I could find. I gripped the whole source tree and the only time AVCodecContext->Field_order is read is either for a stream copy, or to see if the file atom should be written to a mov (and if its not set it isn't) [17:20] michaelni: suggested that it was up to the demuxer to fill it in, and the decoder to move it to the coded_frame->top_field_first, etc [17:21] no, that does not make sense [17:21] the decoder should fill top_field_first from the value actually stored in the essence, if applicable [17:22] lavc ought to deliver what the essence says, and lavf what the container says [17:22] Agreed, but in the case of the essence not containing that information, but it being available elsewhere, surely one should look there in the second instance? [17:23] correct [17:23] and vice versa [17:23] or.. I might slightly prefer if a separate library did this [17:24] this is sort of my beef with lavf - it's guessing too much. I'd rather it were dumber, and having something like libavguess instead :) [17:24] Well if you ffprobe -show_frames a mov containing uncompressed video it will give you a 0 0 [17:24] Even though it is interlaced tff which the file atom will tell you. [17:25] so the raw video decoder needs to fill in top_field_first [17:25] yes [17:25] Tjoppen: the patch should be ok :) [17:25] nice. I'll post that and the other two [17:30] Next job is to do the reverse, so the fiel atom gets written back... [17:54] ffmpeg.git 03Paul B Mahol 07295218f53152: idcin: check chunk_size value before using it * 03http://tinyurl.com/cesesba03 [18:04] ffmpeg.git 03Paul B Mahol 07d7d5b5dfc1b6: flvdec: check return value of create_stream() * 03http://tinyurl.com/cn346do03 [18:04] Action: Daemon404 tries to figure out whats making fate ioc barf [18:04] also added a static uclibc build [18:22] ffmpeg -codecs lists encoders: prores prores_anatoliy prores_kostya [18:23] your point? [18:29] oh yeah, those encoders don't share tables that they could share [18:29] last time I looked [18:29] because I'm trying to pin down a problem with it... [18:35] woo sigills [18:42] http://chromashift.org/log.txt [18:42] how is the sar denom getting set to 0? [18:43] Action: Daemon404 also stares at args [18:43] saste_, this looks like your domain (filter graphs) [18:46] michaelni: any reason to prefer the uint8_t casts over a simple & 0xff? [18:46] you're worried about some optim or something? [18:46] (avio patch) [18:50] ubitux : did you valgrind it ? [18:50] Action: Compn has no clue [18:50] valgrind? oO [18:50] Tjoppen : arent they under different licenses? [18:51] i'd better look at the assembly generated, but i suppose my gcc is smart enough for this kind of op [18:51] if so, how could they share things... [18:51] its verbotten! [18:52] ubitux, no reason its just bikeshed [18:52] ok ok [18:53] though one could probably argue uint8_t has a consistency benefit with int casts [18:53] that is where signed is needed [18:54] ffmpeg.git 03Tim Nicholson 070ee57f8b1476: rawdec.c: Extract interlace information from quicktime flag if it exisits. * 03http://tinyurl.com/bn55laa03 [18:54] ffmpeg.git 03Tomas H?rdin 072b4bbd12c2ec: dvdec: Move the VSC pack parsing to before avctx->execute() * 03http://tinyurl.com/d55kuok03 [18:54] ffmpeg.git 03Tomas H?rdin 077b383bd9d71c: dvdec: Set top_field_first from FS flag * 03http://tinyurl.com/crxqqqx03 [18:54] ffmpeg.git 03Tomas H?rdin 078cbb8f53575d: dv: Set FS flag from AVFrame.top_field_first * 03http://tinyurl.com/cbjh62z03 [18:57] wtf is LVFF file? what create this? [18:58] no idea [18:58] people starting to make up samples for excuses to add code :P [19:02] Compn: dunno [19:03] you could have the one with a more restrictive license use code from the other perhaps? but not the other way around? [19:05] ffmpeg.git 03Paul B Mahol 07bb502411ddb9: nutdec: check return value of av_new_packet() * 03http://tinyurl.com/d9gqw2j03 [19:05] ffmpeg.git 03Paul B Mahol 073ca8a2328878: lavf/audiointerleave: check return value of av_new_packet() * 03http://tinyurl.com/csbvtjj03 [20:02] ok [20:02] 1b8f4d773c4d3bfcff9cdbc26200afbbaa445bc3 is what broke it [20:02] den is getting set to 0 [20:09] Daemon404, how can i reproduce it ? [20:13] michaelni, /configure --prefix=/home/daemon404/build_ffmpeg//install --samples=/home/daemon404/fate_ffmpeg/ --enable-gpl --enable-memory-poisoning --arch=x86_64 --cc=/home/daemon404/ioc/bin/clang --extra-cflags=-fcatch-undefined-c99-behavior --extra-ldflags='-fcatch-undefined-c99-behavior -lm' --disable-asm --enable-ftrapv --disable-optimizations [20:13] is what im doing [20:13] and tetsing with: [20:13] (normal clang should work i think) [20:13] fmpeg -nostats -cpuflags all -threads 1 -thread_type frame+slice -i /home/daemon404/fate_ffmpeg//cdxl/cat.cdxl -an -frames:v 16 -f framecrc - [20:13] ^ test [20:23] Daemon404, Hmm and how is that related to "den is getting set to 0" ? [20:23] i traced the crash [20:23] [12:42] <@Daemon404> http://chromashift.org/log.txt [20:23] that commit causes that [20:27] ok, ill look into it [20:39] ffmpeg.git 03Nicolas Noirbent 07af32a1f2d1b8: doc/filters: pad examples should be a @subsection * 03http://tinyurl.com/cb9hvne03 [20:39] ffmpeg.git 03Michael Niedermayer 07547ec4d3034d: ffmpeg: check timebase validity before using it. * 03http://tinyurl.com/csedwnr03 [21:06] michaelni, fixed [21:06] ioc is now happy again [21:16] added http://fate.ffmpeg.org/report.cgi?time=20121015191457&slot=x86_64-poky-linux-uclibc-static [21:18] poky linux, kawaii~ [21:18] i use to work on yocto/poky ;) [21:20] michaelni, do you have any idea for ticket #ticket 1801? [21:20] in other words, what's the correct way to set DTS [21:21] also, does it really belong to user-space? [21:27] ffmpeg.git 03rogerdpack 07d9d547063648: docu: fix flite example * 03http://tinyurl.com/9p9jhpd03 [21:27] ffmpeg.git 03rogerdpack 077f5a78a0da8e: docu: add flite ffmpeg example * 03http://tinyurl.com/8mez4ns03 [21:27] ffmpeg.git 03rogerdpack 07a8eaa9ebc670: docu: change verb flite * 03http://tinyurl.com/97enst703 [21:27] ffmpeg.git 03rogerdpack 07dfd085aae2a1: docu: add word resize to scale filter, to make searching for the word resize simpler * 03http://tinyurl.com/9meslh603 [21:27] ffmpeg.git 03Michael Niedermayer 070cc3cd5c654d: Merge branch 'docu' of https://github.com/rdp/FFmpeg * 03http://tinyurl.com/99jf48l03 [21:30] Are there any plans to expose the protocol API again? [21:31] bryno, that's a popular feature request, please file a ticket in trac (and let people upvote it) [21:31] I know that it was public at one point. I made a lot of use of it ^_^ [21:34] bryno, evil use of it? ;-) [21:36] about tickt1801 i miss information to reproduce the problem, the timestamps should be set correctly in the correct timebase ... [21:36] ffmpeg.git 03Stefano Sabatini 07a726ac9a11b4: examples/muxing: extend usage notice * 03http://tinyurl.com/cwrd2cj03 [21:40] michaelni, cd doc/examples; make muxing; muxing out.mov [21:41] also what's the real meaning of "Encoder did not produce proper pts, making some up."? [21:42] mp=geq is crashing :'( [21:45] morning [21:49] saste_, that way one can only build muxing.c if one has updated all system libs to git master [21:49] and headers [21:49] michaelni, no you can also use the source libraries [21:49] the compiled source dir libraries [21:50] i still have to commit the README file [21:50] ill look at this once its documented how to build muxing.c [21:52] michaelni: did we do what you wanted/needed? [21:54] j-b, no iam still waiting for cone showing more complete commit messages, daily free pizza and world peace [21:54] more complete messsages? like the first line? [21:54] and no more tinyurl :( [21:55] ok [21:55] will see what we can do [21:58] ffmpeg.git 03Stefano Sabatini 0722c5cc239c25: tools/ffeval: include compat/getopt.c in case of missing system getopt() * 03http://tinyurl.com/9x2cy9k03 [22:00] roger d pack = ? [22:00] is he on irc? [22:03] usually, he is not [22:06] should we use av_log or stderr in tools? [22:45] ffmpeg.git 03Stefano Sabatini 07faa1cb50ed7d: lavfi/ass: extend syntax for ass filter * 03http://tinyurl.com/cbsx5ja03 [22:56] anyway to debug a passlog file name that ffmpeg is looking for when encoding to two different files? "ratecontrol_init: can't open stats file" is all I get. [00:00] --- Tue Oct 16 2012 From burek021 at gmail.com Tue Oct 16 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Tue, 16 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121015 Message-ID: <20121016000501.73B3418A018B@apolo.teamnet.rs> [02:31] Hey folks, my colors are coming out messed up from libavcodec [02:31] PIX_FMT_RGB24 is R8G8B8 right ? :/ [02:31] not inverse? not BGR? [02:31] etc. [02:44] i would like to ask how one would go about merging video sources that contain both 25 fps and 30 fps in a professional manner [03:02] asdsdfhbasdhbahsvdasd [03:03] libavcodec is giving me the same R and G through PIX_FMT_RGB24 decoding [03:03] HARP!!1 [03:03] asdfasdfasdf [03:03] ok now seriously... what the hell? [03:12] nevermind, my own mistake [06:44] in the interest in archiving audio, would high [06:44] quality mp3 or vorbis be better? [06:52] lake, i prefer vorbis, but depends what you care to play it on later since some things don't support ogg [06:53] memleak it something that has always been there, you could show the same thing by playing tricks with the old intelfb, i learned about it around 1995 [06:53] no, 2005 i mean [06:56] tdr: it's just for archiving. i will put it online and not touch it again for ages. and if i do, it will be to convert into some other format, dvd probably. [06:56] tdr: i would ideally use mjpeg/flac, but i get sync issues there while mjpeg/mp3 doesn't seem to [06:57] these are home videos, so quality is a concern, but i don't think lossless is needed [06:58] well i'm kind of an audiophile, i like things to sound good, i generally dont use mp3 much [06:59] tdr: i hear you, i definitely don't want it to sound crappy. i'll try a high quality vorbis and see if i like it. [07:00] doesn't have to be "high quality" really. just don't shoot super low. [07:00] plus mp3 is closed and vorbis is open [07:18] Is there sample C++ code that uses libav to transcode a audio file? (Like a generic transcode function) [11:06] hello. I cannot do 'git pull' on the ffmpeg repository. git index-pack gets stuck. [11:07] isn't it just slow? [11:07] also, is it normal that it is trying to download 50 MB, even if I clone the repository yesterday night? [11:07] are you cloning through http? [11:07] ubitux, checked with htop, it doesn't seem to do anything [11:07] ubitux, no [11:08] works fine for me [11:09] http://pastie.org/5061046 [11:10] you're doing a complete clone there, yesterday I used --depth 1 and now the problem arises with git pull. let me try to clone it completely [11:12] pffft, stuck at 99%, doing absolutely nothing [11:17] something looks wrong with your setup [11:34] ubitux, https from github works. I checked with gdb, git gets stuck in a read to stdin, but I guess this provides little info. OS is fedora 17 x64. fyi, if others with the same problem come. [11:35] ok :) [11:35] you have a git read-only on github [11:36] (which might be more appropriate than http) [12:57] Hello. I'm Trying use this; ffmpeg -vn -acodec copy -i input output; But seem to get an error "Unable to find a suitable output format for 'copy' copy: Invalid argument"; Any clues? [12:59] Erm. Wrong error pasted; Unknown decoder 'copy' [13:20] bug2000: try ffmpeg -i input -vn -acodec copy output; [13:23] cbsrobot_, It worked. Why did it work? [13:24] argument order is important [13:26] Thank you very much cbsrobot_. [13:47] bug2000, before -i you set decoder/input-side settings, while after -i you set encoder/output-side settings [13:47] JEEB, OH. thanks. [16:52] Hi everyone! [16:54] Does ffmpeg support accelerated vdpau decoding for re-encoding rather than playback? [17:01] Does ffmpeg support hardware vdpau decode for re-encoding rather than playback? [17:02] xander`: no [17:08] Well, seems like no-one's here... [17:09] ...is there? [17:11] did you have another question? [17:13] Yes. Does ffmpeg have plans for/support for a simple scripting engine (possibly in a separate program) for video filtering? [17:13] An please answer my earlier question also. Thanks! [17:13] I'm trying this command (minimal) that doesn't work: [17:14] No to your first question. [17:14] ffmpeg -vcodec h264_vdpau -i abc.mp4 test.mpg [17:15] OK. Thanks anyway. About the filtering... [17:15] ...support of a minimal scripting would be great. If not within ffmpeg, then as a separate front-end. [17:16] However, the ffmpeg docs only talk about graph2dot... [17:16] ...and not any tool to generate the filter graphs in the first place. [17:19] Can't you script it using whatever tool you know? [17:23] Yes. But that would require knowing the options each filter took in detail, to provide error checking. That defeats the whole purpose of scripting. I would like to have named options in an easily readable syntax, kinda inspired by AviSynth... [17:24] ...Implementing a script myself would be a hack job, with no value from these `scripts' to myself in the future or others. [17:29] can matroska output be forced to do not obtain metadata from source file ? [17:40] damn, i should not give up so easily, found it! -map_metadata -1 does the trick :) [17:48] hi ffmpegifiers [18:07] hey guys, i'm quite new to ffmpeg and have a simple question about encode/decode audio. Right now I have written a small program, that gets audiodata from my microphone. It saves both, not encoded RAW-Audio and encoded mp3. For encoding I use avcodec_encode_audio2(). Now I want to decode the AVPakets I get from avcodec_encode_audio2() using avcodec_decode_audio4() to understand how the... [18:07] ...function works. Therefore I created an AVFRame *decoded_frame and I try to call int ret2= avcodec_decode_audio4(c,decoded_frame, &got_frame, &pkt); [18:08] In Debugmode VS says there is an access violation at 0x0000000 at my call of avcodec_decode_audio4(). What could that be? [18:08] thanks in advance [18:41] howcome crf18 is 3x larger than default crf23 ? [18:42] creep: it uses a higher bitrate [18:42] okey [18:42] can you recommend me a conversion command for animation? [18:43] Didn't I give you one already? [18:43] tried that [18:43] 10MB h264 mkv vs 32MB zmbv ;/ [18:43] ffmpeg -i ./game_003.avi -vcodec libx264 -preset placebo -crf 18 -tune animation game_003_placebo_crf18_anim.mkv [18:44] this is my best option? [18:44] If you want a specific size use 2 passes, if you want a specific quality use -crf [18:44] stills are good, but movements have noise [18:44] most of the screen is unchanged... it is a redalert video [18:46] i would think that -anim makes the filesize 3MB [18:46] You could try -tune stillimage and lower the crf value if you still get noise. [18:47] Are you using a recent version of ffmpeg? [18:47] latest svn [18:47] you mean git [18:47] can't remember, maybe [18:47] ffmpeg 2>&1|sed q [18:47] i just copied clone command into bash script [18:48] ffmpeg version N-40640-g5edd4fc Copyright (c) 2000-2012 the FFmpeg developers [18:48] built on May 23 2012 10:28:32 with gcc 4.4.4 20100726 (Red Hat 4.4.4-13) [18:48] so -anim makes it better but not much [18:48] It's time to update. [18:50] ffmpeg version N-44818-g13f0cd6 Copyright (c) 2000-2012 the FFmpeg developers [18:50] built on Sep 28 2012 08:34:26 with gcc 4.4.4 (GCC) 20100726 (Red Hat 4.4.4-13) [18:50] sj [18:50] ah [18:50] somehow i have /usr/local/bin too and its first, i'mm deleting it right now ;/ [18:50] test it again once you know you're using the recent binary. [18:52] ah, i was using the old version [18:52] trying again then [18:52] was ffmpeg made better since may ? [18:52] ffmpeg is made better everyday [18:54] now i know what caused the configuration mismatch errors [19:53] hy [19:55] hi [19:56] hi [19:57] Last message repeated 1 time(s). [19:58] any body wanna be my friend? [19:59] try #friends [20:10] lol relaxed [20:11] mystica555 is here too... fun [20:12] i autojoin like 30 channels on freenode... half the reason im missing so often is because i excess-flood upon rejoining and then i don't want to deal with it and i close it... then sometimes it works [21:07] is it ok for ffmpeg to be dropping frames. this is my output [21:07] frame=102056 fps= 30 q=24.0 size= 2146768kB time=00:56:45.79 bitrate=5163.7kbits/s dup=0 drop=102023 [21:17] lake: pastie.org your command and full output [21:21] relaxed: http://pastie.org/5063774 [21:26] move the framerate before the video input and change it to "-r 30000/1001" [21:27] what the new command for -an and -vn ? [21:27] they should still owrk [21:27] work* [21:28] yup my bad [21:28] sorry [21:28] relaxed: now i get "[video4linux2,v4l2 @ 0xd9b4e0] ioctl set time per frame(100/2997) failed [21:29] relaxed: i mean [video4linux2,v4l2 @ 0x1dfd4e0] ioctl set time per frame(1001/30000) failed [21:30] omit it [21:32] relaxed: this is my output now [21:32] http://pastie.org/5063824 [21:32] seems better [21:35] If you want the output to be 29.97 then change yadif=1 to yadif=0 [21:36] You were doubling the framerate and then cutting it in half at the same time. [21:36] Thus, a shit ton of dropped frames. [21:36] i think i want yadif=1 [21:36] i'm going for quality [21:37] Yes, you do [21:37] thank you relaxed [21:37] you are my hero [21:38] You are welcome. If you want to archive this maybe you should think about using libx264 in lossless mode. [21:39] yes it is for an archive. i didn't realize i could use lossless x264 [21:39] Then make lossy encodes from it. [21:39] exactly what i'd like to do [21:39] would it be smaller than an uncompressed avi? [21:39] Yes but it will still be large. [21:40] change -crf 18 to -qp 0 for lossless video [21:40] add -preset veryslow for better compression [21:42] i get crazy xruns with -preset veryslow [21:44] same with -preset slow [21:45] what are you capturing? [21:46] v4l2 [21:47] relaxed: same with lossless, loads of xruns [21:47] waaaaah [21:48] my machine is super fast, so i think it's just the dazzle device i am using. [21:48] i'm not sure what the bottleneck is [22:25] relaxed: when i only capture the audio from alsa and encode into flac, i get no xruns [22:26] relaxed: so i guess the bottle neck is the usb video ccapture device [22:26] i wonder if there is a usb 3.0 capture device [22:45] Hi, I'm trying to convert PNG images to an mp4. I'm doing the following command without double quotes, "ffmpeg -f image2 -framerate 30 -start_number 1 -r 3 -i VIP-Loading-Media-%.png -s 1024x768 vip-loading.mp4" on Windows 7. [22:45] However when I execute this command it tells me it can't find me image files. [22:45] TheWarden, "VIP-Loading-Media-%.png" => invalid syntax [22:45] They are called VIP-Loading-Media-01.png, VIP-Loading-Media-02.png and VIP-Loading-Media-03.png. [22:46] you want -%02d.png [22:46] Oh okay I misunderstood. I thought the % symbol was the variable but rather its %02d. [22:48] mmm okay well that did it however the mp4 file is 13KB and there is no content. It's just black. odd. [22:49] Am I doing something wrong? Maybe ffmpeg doesn't support png images [22:50] argh [22:51] no it was: [22:52] here is the command I issued and the results it dumped out, https://gist.github.com/2b9457e6e846117c25f0. [23:04] any idea that I'm doing wrong here? must be doing something stupid. [23:05] why -framerate 30 ... -r 3? [23:05] what do you want to achieve? [23:06] I'll be honest first time doing this not overly familiar with ffmpeg. What I'm wanting to achieve is to take the three images I have and convert it to a video file with 30 secs per image will say out to a 1024x768 resolution and mpeg 4. [23:06] I might have to adjust how many seconds per image but not sure without playing with it. [23:07] you have three images, how much do you want they should last on the output video? [23:07] also: 30 secs for image with 3 images... is a waste [23:11] well I dunno a minute... even 30 seconds is good as the file will be in a loop playing over and over again. [23:12] TheWarden, how can you assume that it will loop? [23:12] also, how much time should last every loop? [23:13] saste_: I'm forcing VLC to loop the video but I suppose to be on the safe side I could just make a file that is like 2 or 3 minutes long. [23:13] TheWarden, still you didn't reply to my first question [23:13] how should the three image last in the output video? [23:13] how much... [23:14] saste_: per image your saying? ahhh not sure like 30 secs ? [23:14] so every image should last 30 secs? [23:14] yeah [23:15] repeat... this is just a video that is loaded by default until the playlist is formed and starts playing. [23:15] kiosk like setup [23:15] ffmpeg -f image2 -framerate 1/30 -start_number 1 -i VIP-Loading-Media-%02d.png -r 30 OUT [23:15] something like that [23:52] burek: NITE [00:00] --- Tue Oct 16 2012 From burek021 at gmail.com Wed Oct 17 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Wed, 17 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121016 Message-ID: <20121017000501.F03F018A01EC@apolo.teamnet.rs> [01:04] saste_: thanks so much, that did the job. [02:28] Alo [02:31] Someone here can provide me some support? [02:43] Someone alive here? [02:49] hi. is there a way to rotate a video without re-encoding it? [02:57] im here sometime already and no one until now [02:58] answered me :/ [03:10] funyun: No, I don't think so. [03:10] RunawayDevil: You've not asked a question... [04:05] Hello.. need some help with my command under windows =) [04:45] funyun: you can rotate it upon playback [04:55] hey peeps how do i install yuv420p [04:58] its not letting me convert from it to mpeg 4 [04:59] YUV450p is a pixel format, not a video codec. [04:59] *420 [04:59] ? [04:59] er ok i pb what i got on the cli [05:01] http://sprunge.us/dJaR [05:01] tghere ya go [05:02] Your ffmpeg is rather old, maybe you could update it? And if you do, something like this should work for Android non-HD: ffmpeg -i input -vcodec libx264 -preset:v slow -b:v 500k -profile baseline -acodec aac -b:a 128k output.mp4 [05:29] :( [05:29] its on my server where i encode stuff [07:01] i would like to try to pass these options to ffmpeg: x264 -q0 -m1 [07:01] -q0 works, but -m1 doesn't [07:30] what container would hold flac/mjpeg [07:31] lake: -qp 0 -subq 1, or use x264opts [07:31] mkv should hold that combination [07:31] llogan: thank you so much for the quick replies [07:31] but why mess with those settings anyway? that's what the presets are for [07:34] llogan: i could not find any preset that works like -qp 0 -subq 1 [07:35] -qp 0 is lossless [07:35] llogan: if you know what preset they fall under i would be happy to use it [07:35] i am streaming from a v4l device and i -subq 1 seems to be better for streaming live [07:35] anypreset will work with lossles, but only two are really useful: ultrafast and veryslow depending on what you are doing [07:35] no xruns [07:37] i didn't try ultrafast, actually, that might work. [07:37] you should be using lossless for streaming [07:37] *shouldn't [07:37] but i have these home videos that i want to archive. [07:37] should i just use -crf 18 [07:38] i don't understand what you are doing. first you mention live streaming, and then you mention archiving [07:40] llogan: sorry for being unclear. I am "streaming" analog (red/white/yellow cable) to a usb analog to digital converter (red/white/yellow in, usb out) to my pc. 4vl2 is my input device. [07:40] the source is a handycam from the 90s. [07:41] i see now that streaming implies over the network [07:41] that is not what i mean [07:42] ive never used anything like that before. [07:44] if you have unlimited hard drive space you can use lossless if you like. [07:44] mmm 120 minutes of film into lossless is going to massive [07:45] i have 45 films * 2 hours each [07:45] yikes [07:45] of you're limited then -crf 18 should probably be indistinguishable [07:46] what does ffmpeg show for your input streams? [07:46] and the only reason i would reconvert the output would be for dvd. [07:46] thanks llogan. i think that solves my problems! [07:47] https://ffmpeg.org/trac/ffmpeg/wiki/x264EncodingGuide [07:47] CRF section [07:48] 18 looks solid from here [11:31] I'm capturing in windows.. the audio is off about -1000ms on capture.. its there a way to compisate for this during the realtime recording to disk? [11:32] ffmpeg -f dshow -r 30 -aspect 16:9 -s %resolution% -i video="SCFH DSF":audio="Stereo Mix (Realtek High Defini" -rtbufsize 9835m -pix_fmt yuv444p -vcodec libx264 -preset ultrafast -tune zerolatency -qp 0 -acodec pcm_s16le Z:\%MM-DD-YYYY%.mkv [11:38] btw this is being called by a bat file so hence the %.......% stuff [12:41] How to make sure ffplay outputs single channels output audio for single channel input audio ? [14:18] does anyone know of any tool that can tell me what is wrong with this flv video (h264+aac): http://teststream765.chickenkiller.com:8091/live.flv [14:19] something is wrong with it, but I just can't figure out what [14:19] it doesn't play in jw player but it plays somehow in vlc and ffplay... [14:24] burek: did you try remuxing it? [14:25] er.. no, why? [14:26] try it [14:26] well it breaks the whole point kinda.. [14:26] the webcam is encoding the h264 video [14:26] i just add the audio in vlc and mux it into flv and stream it [14:27] wait, what do you mean remux it, i am muxing it :) [14:27] i thought you said re-encode it :) [14:28] webcam gives raw h264 which I'm muxing together with aac in flv (using vlc) [14:34] Hmmm, maybe flash player can't decode the h264 stream. [14:35] let me copy a part from #ffmpeg-devel [14:35] described in ffmpeg's vocabulary, vlc is doing this: ffmpeg -f h264 -i /dev/video0 -f alsa -i hw:0 -c:v copy -c:a libaacplus output.flv [14:36] when i grab the raw video (not h264) and encode both a/v in flv, then it works [14:36] but if i let the webcam encode it and just pass the video + encode the audio, again in flv, then the video doesn't work, but audio does [14:37] right, flash player doesn't seem to like the h264 stream. [14:37] how can I tell what's wrong with it [14:38] That is a good question. [14:38] Maybe nothing is wrong with it and the porblem is with flash player. [14:39] Since ffplay and vlc can play it back that would be my guess. [14:39] :/ i updated everything regarding flash.. [14:39] i would just like to have some kind of a tool [14:39] that can compare 2 flv files [14:39] so that i can tell what is different [14:39] and try to fix it [14:40] That sounds magical :) [14:41] :) [14:44] Did you try using fmpeg instead of vlc to capture? [14:44] ffmpeg* [14:44] it can't capture -f h264 [14:44] i don't know why [14:44] vlc can [14:45] http://pastie.org/private/ajwaid5gea27orjimqmmmg [14:47] Wouldn't you use -f v4l2 ? [14:47] oh yes, my bad [14:48] ffmpeg -f v4l2 -vcodec h264 -i /dev/video0 ? [14:49] you may not have to specify the vcodec [14:49] ok, what cmd should i type [15:22] What is your vlc command? [15:24] (to avoid spamming the channel) [15:30] if only I was able to grab it only using ffmpeg [15:31] I could debug it a low easier [15:31] but this way I need to bug people from videolan too [15:31] not knowing where the bug actually is [15:31] s/low/lot [15:33] doesn't work: http://pastie.org/private/7dlsd9wcflbrl5gld3z1rg [15:33] works: http://pastie.org/private/fnfbnee7wthe14d8zkexlq [15:49] and duration is certainly not 11 minutes.. it is more like a minute or less [15:49] also "368 pixels" should be 360.. man.. it is so wrong.. [15:50] burek: does this work? ffmpeg -f v4l2 -s 640x360 -r 25 -i /dev/video0 -map 0:v -c copy output.mkv [15:51] input Stream #0:0: Video: rawvideo [15:51] output Stream #0:0: Video: rawvideo [15:51] :) [15:51] ffmpeg -f v4l2 -s 640x360 -r 25 -vcodec h264 -i /dev/video0 -map 0:v -c copy output.mkv [15:51] [video4linux2,v4l2 @ 0xf79520] Cannot find a proper format for codec_id 28, pix_fmt -1. [15:52] just that end it quits [15:52] and* [15:52] ffmpeg -f v4l2 -s 640x360 -r 25 -vcodec h264-pix_fmt yuv420p -i /dev/video0 -map 0:v -c copy output.mkv [15:52] [video4linux2,v4l2 @ 0x1a23620] Cannot find a proper format for codec_id 28, pix_fmt 0. [15:53] I guess you should file a bug report. [15:54] well that is a point :) what should I report? :) [15:55] ffmpeg -f v4l2 -s 640x360 -r 25 -vcodec h264 -pix_fmt yuvj420p -i /dev/video0 -map 0:v -c copy output.mkv [15:55] [video4linux2,v4l2 @ 0x25f4620] Cannot find a proper format for codec_id 28, pix_fmt 12. [15:58] The reason the stream doesn't work with flash player is because it is yuvj420p and not yuv420p. [16:01] fmt.fmt.pix.pixelformat = V4L2_PIX_FMT_H264; //replace [16:01] does this have to do anything with that? [16:02] #define V4L2_PIX_FMT_H264 v4l2_fourcc('H', '2', '6', '4') /* H264 with start codes */ [16:02] no, i guess it's not that sorry [16:30] well I fixed my sync issues with directshow filters.. hand to flip the order, Audio=:video= and have -itsoffset 00:00:00.90 right after the video filter.. [16:31] that is case anyone else comes looking for it. [16:31] if I had -itsoffset 00:00:00.90 anywhere else it didn't work but it didn't give any errors ether. [16:35] yes, you basically put audio a little bit before the video that way [16:36] join #libav [16:36] ? [16:37] he probably misspelled it :) [16:37] :P [16:37] he's probably provocating [16:37] probably just missing / [16:37] :) [16:41] I recently want to write a program to make video monitoring, but I never used ffmpeg, where should I start? [16:42] I am also interested about the libav, what's the matter? [16:43] excuse me? [16:44] you should just check the headers related to what you'd be doing [16:44] ffmpeg consists of multiple libraries that it presents [16:45] libavcodec, libavformat, libavutil, libswscale, libswresample, libavresample [16:47] libavformat handles parsing of "containers" into streams that can be thrown into decoders or "repackaged", libavcodec handles the decoding and encoding of things, libavutil has various utlilities that other libraries might as well use, libswscale handles conversions between colorspaces, libswresample and libavresample are two libraries that handle audio resampling/mixing [16:49] Soulhunter, I should go and read some docs about libavformat & libavcodec [16:49] s/Soulhunter/so/g [16:50] yeah [16:50] there are examples available as well [16:50] otherwise the doxygen and the source code are your friend [16:51] JEEB, OK, thank you [17:14] andyhuzhill: help for command line or api usage? [17:14] ubitux:api [17:14] ubitux, I want to compiled it into my program [17:15] andyhuzhill: ok then, https://ffmpeg.org/about.html for an overview of the project, and http://git.videolan.org/?p=ffmpeg.git;a=tree;f=doc/examples;hb=HEAD is a good start for API usage [17:15] build examples etc [17:17] can anyone check what is wrong with this stream.. http://teststream765.chickenkiller.com:8091/live.flv [17:18] :S [17:18] does anything look odd/wrong or anything [17:18] in terms of video frame size, pixel format, encoding, etc [17:19] because it won't play in flash players [17:20] i can't atm, i'm behind a proxy [17:20] ok, thanks anyway [17:20] burek: do you have the issue if you wget it? [17:21] you mean, can flash player play it as a file? [17:21] i don't know what's your issue [17:21] flash players play only audio not video [17:21] so I don't know how to start debugging [17:21] you sure the h264 profile is supported by the player? [17:22] i know some apple devices are limited to a very limited level [17:22] you might need to check in that direction [17:23] https://en.wikipedia.org/wiki/H.264/MPEG-4_AVC#Levels [17:23] hmh, how to set the profile level via v4l2.. [17:23] is it even possible [17:24] the device is outputing h264? [17:24] yes c920 [17:24] huhu fun [17:24] you couldn't even imagine... [17:25] i'm not sure how you could influence the level [17:25] maybe just by requesting a different resolution [17:25] (see the -video_size input v4l2 option) [17:26] also look at -list_formats [17:26] to check if there are "different" h264 formats [17:27] yes, good idea, didn't try different video sizes [17:47] Is there any plan to support Freeframe API beside frei0r? [18:20] hello when i use http auth how should i escape the password if there is a ! [18:20] %21 does not work [18:33] arpu_: try \! [19:27] thx relaxed works fine! [20:35] can someone explain to me how video gets changed from 29 fps to 25 fps correctly. do they just skip out hte od frame here and there ? [20:51] what is the benefit of 60fps vs 30fps [20:55] sine_, you can't really do that "correctly" :< [20:55] some people merge frames, others cut frames off [20:55] some make fields from frames and encode interlaced [20:55] etc. [20:55] lake, unless you have actually a 60fps source, absolutely nothing [20:57] i have 60 fps mode on my camera and you can make smoother slow motion [20:57] you have more source frames to use [20:58] yeah, sure [20:58] you can slow down 60fps to, say, 24fps [20:58] and it'll be still relatively fluid [20:59] so if i'm capturing from v4l2, how can i know the source [20:59] source fps [20:59] uhh, the API usually tells you the options >_> [20:59] some cameras over v4l2 give for example 15fps 1920x1080 but 30fps 640x480 [20:59] etc. [21:44] if i convert a .mkv (x264/flac) => ntsc-dvd, i get a ton of buffer underflow messages from ffmpeg [21:45] join the club [21:45] i feel so welcomed [21:48] relaxed: is this a known issue [21:50] <_Adolf_Hit-ler_> lmfao [22:31] lake, you did do it with multiple passes, right? [22:31] single pass rate control is completely derp, 2pass can be somewhat better [23:01] Does anyone know a program that i can use to batch process .png or .jpg images. i want to change colour depth and contrast and invert the images for a better black and white camera tracking [23:07] sine_: you can probably use a frei0r filter in ffmpeg to change contrast, and negate filter to invert [23:09] or of course inverstigate imagemagick or graphicsmagick [23:10] ok i have saved what you have said into a text file and will investigate tomorrow after work thanks and good night. [23:12] sine_: the frei0r filter in particular is "contrast0r" [23:13] and graphicsmagick is generally faster, AFAICT. [23:13] (than imagemagick) [23:14] JEEB: thanks for the tip. i will try multiple pass [23:15] also underflows just mean that the encoder couldn't keep the stream within the maxrate/bufsize you set [23:16] JEEBsv: okay, when i play back the video it's choppy but it sounds great [23:16] i'll try multipass now and see how it goes [23:17] this has all been one crazy learning experience. i didn't realize how complex ffmpeg is. [23:17] very badass tool [23:59] what is a value between "sameq" and [00:00] --- Wed Oct 17 2012 From burek021 at gmail.com Wed Oct 17 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Wed, 17 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121016 Message-ID: <20121017000503.0323218A01ED@apolo.teamnet.rs> [00:14] michaelni: any neat trick to check for the overflow in swfdec? [00:14] given w and h being int (at most 16 bits), an >= 0 check added [00:14] and* [00:15] (and linesize with max w * 4) [00:16] mmh maybe av_check_image_size().. [00:18] av_image_check_size* [00:25] my brain doesn't work anymore [00:25] 'night ppl :) [01:16] ffmpeg.git 03Michael Niedermayer 07db3a0aae9fdc: msvideo1enc: fix initial keyframe value * 03http://tinyurl.com/8qubk8e03 [01:16] ffmpeg.git 03Michael Niedermayer 07f5581266f767: ra288: assert order to be withinn supported range in do_hybrid_window() * 03http://tinyurl.com/8vkvw2p03 [01:16] ffmpeg.git 03Michael Niedermayer 07901f4bb6fc70: utils: consider mpeg4 in mp4/mov to have a unreliable timebase * 03http://tinyurl.com/8t22sks03 [01:16] ffmpeg.git 03Michael Niedermayer 07ba31e59e83ca: ffmpeg: Override r_frame_rate by corrected timebase only for interlaced streams * 03http://tinyurl.com/8ah7zst03 [02:25] ffmpeg.git 03Michael Niedermayer 071fe45903b81f: msvideo1enc: fix interframe encoding * 03http://tinyurl.com/c4y7de203 [03:15] ffmpeg.git 03Michael Niedermayer 073896cd11a107: mxfenc: explicitly truncate continuity count. * 03http://tinyurl.com/brad5pq03 [03:15] ffmpeg.git 03Michael Niedermayer 079c669672c7fd: x86/motion_est: widen before multiply. * 03http://tinyurl.com/cy5opwz03 [04:06] ffmpeg.git 03Michael Niedermayer 074cc4ca584736: mandelbrot: fix inner=period coloring routine * 03http://tinyurl.com/bszbv6c03 [04:17] ffmpeg.git 03Andrew Euell 0736b3b2376dd9: gitignore: add *.dylib for those of us on Darwin/OS X systems. * 03http://tinyurl.com/8sgr8ux03 [10:50] michaelni: question about a very old commit :) [10:50] c45388b1 [10:50] "if(s->streams[i]->codec->codec_tag == 5){" [10:50] is this a special case for nellymoser? [10:51] ah, i missed that this chunk was already present.. [10:53] 11a8e4257 then :) [10:55] anyway, looks like a nellymoser 8khz special case to me [10:56] mmh or maybe vp6... [10:56] yes vp6a actually :p [11:18] nyuhu: there is no virtualdub repository? [11:18] only tarballs? [11:50] ubitux : theres the source code on sourceforge& [11:53] sourcecode right [11:53] but what about the history? [11:57] theres a changelog with a brief of whats new since previous version :/ [11:58] +the [12:51] nyuhu: ok :( [12:51] it's too bad because it could have given some hints about the constants [13:03] yes& [13:21] nyuhu: it looks like random to me [13:21] or at least "designed" for random [13:22] and "jran" sounds like j random [13:22] http://sprott.physics.wisc.edu/phys505/lect07.htm [13:22] Example: A = 1366, B = 150889, C = 714025 [13:22] looks similar to your constants [13:24] nyuhu: and they looks totally constant [13:24] ffmpeg.git 03Diego Biurrun 077e68c91e247b: rmdec: Move SIPR code shared with Matroska demuxer to a separate file * 03http://tinyurl.com/8hv6tec03 [13:24] ffmpeg.git 03Luca Barbato 070fbb62a8e6c6: build: support asan and tsan toolchain shortcuts * 03http://tinyurl.com/9hhcz5703 [13:24] ffmpeg.git 03Mans Rullgard 074c995fafd861: configure: simplify get_version() function * 03http://tinyurl.com/9p8xsgv03 [13:24] ffmpeg.git 03Michael Niedermayer 079270a2b374ea: Merge commit '4c995fafd861f537360b3717901cdbed6a6844e7' * 03http://tinyurl.com/9cn7fqp03 [13:24] nyuhu: so you could remove them from the context [13:30] ffmpeg.git 03Mans Rullgard 0766a1ccd7467a: configure: simplify argument handling in check_ld * 03http://tinyurl.com/crpncb703 [13:30] ffmpeg.git 03Michael Niedermayer 07d15be9108e6a: Merge commit '66a1ccd7467ab1913cd8877114c6d4c2588bb12f' * 03http://tinyurl.com/c5yu9bh03 [13:35] ffmpeg.git 03Stefano Sabatini 077ca102a7d7bd: examples: add README file with simple compilation instructions * 03http://tinyurl.com/ch82q5f03 [13:52] ffmpeg.git 03Stefano Sabatini 0725f3827e1daf: tools/ffeval: do not use UNIX-specific /dev/std{in,out} files * 03http://tinyurl.com/8w6c56g03 [13:56] Action: ubitux just founds something pretty funny with vp6 and flv [14:01] ffmpeg.git 03Mans Rullgard 0784e65c763d8f: build: simplify enabling of compat objects * 03http://tinyurl.com/9ogafhq03 [14:01] ffmpeg.git 03Diego Biurrun 07c1fcfdec7546: rangecoder-test: Return in case of an error * 03http://tinyurl.com/9nnmaxu03 [14:01] ffmpeg.git 03Michael Niedermayer 075717562c789f: Merge commit 'c1fcfdec75468009dc7de29a5d1c6adf3b2ef77d' * 03http://tinyurl.com/9jxyv6j03 [14:02] ./ffmpeg -i ~/fate-samples/ea-vp6/g36.vp6 -c copy -y out.flv ? output is rotated %) [14:04] yeah, I think VP62F or whatever was actually upside down or whatever [14:05] it works fine when muxed in mkv [14:07] ffmpeg.git 03Diego Biurrun 07f1be514540dc: rangecoder-test: Drop timer output that clutters stderr * 03http://tinyurl.com/ccl3zmt03 [14:07] ffmpeg.git 03Mans Rullgard 07099294577c8c: fate: improve md5sum utility selection * 03http://tinyurl.com/dy2f3f403 [14:07] ffmpeg.git 03Mans Rullgard 073dc06b6972cf: tiny_psnr: check for specified sample size less than 1 * 03http://tinyurl.com/cj9cosd03 [14:07] ffmpeg.git 03Michael Niedermayer 0783962004f727: Merge commit '3dc06b6972cf389269e9c36ff0a4373f80f7149b' * 03http://tinyurl.com/bt4eqq703 [14:09] and same in mov [14:13] well, there was some weird thing with muxing vp6X into flv, that's all I remember [14:13] which I remember including upside-downess [14:18] :( [14:19] ffmpeg.git 03Mans Rullgard 07bf868c4a9bc6: tiny_psnr: fix range calculation for sample size of 32 bits * 03http://tinyurl.com/bn8839n03 [14:19] ffmpeg.git 03Mans Rullgard 0707b3790d3626: build: simplify linking tools with cmdutils.o * 03http://tinyurl.com/ck8hlzl03 [14:19] ffmpeg.git 03Michael Niedermayer 078227d36bd356: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/d9z5tfs03 [14:29] I'm sorry to bother you guys again, but is there any way I could modify the flv muxer to be _really_ verbose with its err output log, so that I can see what doesn't work in vlc, when it is being used through the libav linkage [14:30] it seems it doesn't mux well the raw h264 stream (coming from the webcam, encoded already, no need to re-encode) with alsa input (encoded at the same time with vlc) [14:31] described in ffmpeg's vocabulary, it would be like: ffmpeg -f h264 -i /dev/video0 -f alsa -i hw:0 -c:v copy -c:a libaacplus output.flv [14:32] that's basically what vlc is supposed to be doing prior to http streaming [16:36] ffmpeg.git 03Anton Khirnov 071846f3b5b15f: avconv: fix -force_key_frames * 03Error03 [16:36] ffmpeg.git 03Reimar D?ffinger 078efae4cbbf51: avconv: fix parsing of -force_key_frames option. * 03http://tinyurl.com/co7nb5303 [16:36] ffmpeg.git 03Kostya Shishkov 0702b723946279: vc1dec: add flush function for WMV9 and VC-1 decoders * 03http://tinyurl.com/c67hjwo03 [16:36] ffmpeg.git 03Reimar D?ffinger 07e9ac06160f45: sipr: fall back to setting mode based on bit_rate. * 03http://tinyurl.com/bllgnts03 [16:36] ffmpeg.git 03Anton Khirnov 07bed5847563d1: bmpdec: only initialize palette for pal8. * 03http://tinyurl.com/dx7q6o903 [16:36] ffmpeg.git 03Janne Grunau 07fdb708078174: Revert "nuv: check per-frame header for validity." * 03http://tinyurl.com/bmgqeyr03 [16:36] ffmpeg.git 03Janne Grunau 076704522ca9dd: nuv: check RTjpeg header for validity * 03http://tinyurl.com/d6t2ctq03 [16:36] ffmpeg.git 03Max Lazarov 0725a1a5b1b38b: eval: fix swapping of lt() and lte() * 03http://tinyurl.com/bt52k7b03 [16:36] ffmpeg.git 03Janne Grunau 077a7229b52d19: imgconvert: avoid undefined left shift in avcodec_find_best_pix_fmt * 03http://tinyurl.com/d6xpmor03 [16:36] ffmpeg.git 03Michael Niedermayer 07da0c45766347: mpegvideo: Don't use ff_mspel_motion() for vc1 * 03http://tinyurl.com/cfjwvtb03 [16:36] ffmpeg.git 03Anton Khirnov 077124fa5d3640: lavf: don't segfault when a NULL filename is passed to avformat_open_input() * 03http://tinyurl.com/cwoswto03 [16:36] ffmpeg.git 03Justin Ruggles 07d9ffa2aca1e4: golomb: check remaining bits during unary decoding in get_ur_golomb_jpegls() * 03http://tinyurl.com/czcvgcg03 [16:36] ffmpeg.git 03Michael Niedermayer 07a1b127515bb7: alsdec: check opt_order. * 03http://tinyurl.com/d4mzppa03 [16:36] ffmpeg.git 03Martin Storsj? 076d1b91324c30: h263: Add ff_ prefix to nonstatic symbols * 03http://tinyurl.com/cne5wp603 [16:36] ffmpeg.git 03Martin Storsj? 078c0bbe51561b: vlc/rl: Add ff_ prefix to the nonstatic symbols * 03http://tinyurl.com/ctscotf03 [16:36] ffmpeg.git 03Janne Grunau 07f695bd601640: rv34: use AVERROR return values in ff_rv34_decode_frame() * 03http://tinyurl.com/d4jy28603 [16:36] ffmpeg.git 03Janne Grunau 0790575bd7dd1e: rv34: Handle only complete frames in frame-mt. * 03http://tinyurl.com/dyq2hpy03 [16:36] ffmpeg.git 03Janne Grunau 07b1ad5a21da7a: rv34: error out on size changes with frame threading * 03http://tinyurl.com/bwlxpj903 [16:36] ffmpeg.git 03Mina Nagy Zaki 07e5f4e2494228: lavfi: avfilter_merge_formats: handle case where inputs are same * 03http://tinyurl.com/dxgkwhj03 [16:36] ffmpeg.git 03Kostya Shishkov 07d4f3abca6a76: indeo3: validate new frame size before resetting decoder * 03http://tinyurl.com/clgh5ex03 [16:36] ffmpeg.git 03Justin Ruggles 07e46cf805b100: vorbisenc: check all allocations for failure * 03http://tinyurl.com/c8np5qz03 [16:36] ffmpeg.git 03Alex Converse 079aaaeba45c41: vorbis: Validate that the floor 1 X values contain no duplicates. * 03http://tinyurl.com/ccj2rfu03 [16:36] ffmpeg.git 03Martin Storsj? 0731bc3fb563b1: snow: Check mallocs at init * 03http://tinyurl.com/cf4vuur03 [16:36] ffmpeg.git 03Michael Niedermayer 07be2dd2559f54: Merge remote-tracking branch 'qatar/release/0.8' into release/0.10 * 03http://tinyurl.com/cq2nks603 [17:12] michaelni, what is the usual way of sending patches for ffmpeg-web [17:12] ml/pastebin/..? [17:22] burek: ml [17:22] classic ffmpeg-devel [17:22] ok thx :) [17:27] btw, should I attach patch or just paste it into the mail body? [17:28] ideally git send-email [17:28] oh :$ [17:28] and no, never paste it [17:28] as an attachment is fine, but it's nice if it's inlined [17:30] hm, i dont have git-send-email ? [17:30] is that normal? [17:30] git send-email, no - after git [17:30] git: 'send-email' is not a git command. See 'git --help'. [17:31] how old is your git? =P [17:31] there is an entry in man git about git-send-email(1) [17:31] 1.7.10 [17:32] solved [17:32] debian provides it in separate package it seems [17:35] how i hate debian and the practice of breaking everything up into meaningless packages =p [17:35] it is easier for them to maintain it, i guess [17:35] how so, all these git tools get updated as one package in upstream [17:35] don't know, im no debian maintainer :) [17:36] just guessing [17:36] nice :) still no git-send-email :D [17:36] how more stupid this can be [17:36] nevcairiel : you mean splitting up -dev headers from libs ? [17:37] no, apparently splitting git in multiple packages [17:37] 'why wont xxx compile?' - 'because debian' [17:37] i can understand splitting off dev headers [17:37] normal users dont need them [17:37] i had to do apt-get install git-email to be able to use git-send-email, but through the use of: git send-email :) [17:37] shortly :D [17:38] so, git send-email --to=... --subject=... 0001-blabla.patch [17:38] right? [17:38] you give it the revisions to send, not a patch [17:39] wait, I don't need git format-patch? [17:39] no [17:45] f it.. i dont have time to setup everything including smtp [17:46] i added it as attachment and sent it manually [17:49] < nevcairiel> you give it the revisions to send, not a patch // you can as well [17:58] ffmpeg.git 03Mans Rullgard 07b102d5d97dae: h264: allow cropping to AVCodecContext.width/height * 03http://tinyurl.com/cdxflrt03 [17:58] ffmpeg.git 03Mans Rullgard 070054d70f23ed: mov: set AVCodecContext.width/height for h264 * 03http://tinyurl.com/cwa2v7n03 [17:58] ffmpeg.git 03Anton Khirnov 07a60eb6ef12df: ffmpeg: fix -force_key_frames * 03http://tinyurl.com/bqgpvdr03 [17:58] ffmpeg.git 03Kostya Shishkov 070173a7966b33: vc1dec: add flush function for WMV9 and VC-1 decoders * 03http://tinyurl.com/c9c7e6p03 [17:58] ffmpeg.git 03Janne Grunau 07f31170d4e7f9: nuv: check RTjpeg header for validity * 03http://tinyurl.com/buc3k3603 [17:58] ffmpeg.git 03Janne Grunau 078812b5f16410: imgconvert: avoid undefined left shift in avcodec_find_best_pix_fmt * 03http://tinyurl.com/bomlrjp03 [17:58] ffmpeg.git 03Michael Niedermayer 07899d95efe12f: mpegvideo: Don't use ff_mspel_motion() for vc1 * 03http://tinyurl.com/crnx9ug03 [17:58] ffmpeg.git 03Anton Khirnov 0777d43bf42d76: lavf: don't segfault when a NULL filename is passed to avformat_open_input() * 03http://tinyurl.com/c68lp5803 [17:58] ffmpeg.git 03Michael Niedermayer 07b6ba39f931a8: alsdec: check opt_order. * 03http://tinyurl.com/d8rscdg03 [17:58] ffmpeg.git 03Mina Nagy Zaki 07b6c5848a1f8f: lavfi: avfilter_merge_formats: handle case where inputs are same * 03http://tinyurl.com/cr8sesx03 [17:58] ffmpeg.git 03Justin Ruggles 0761ece4137298: vorbisenc: check all allocations for failure * 03http://tinyurl.com/brmmq4503 [17:58] ffmpeg.git 03Alex Converse 07d6e250abfc36: vorbis: Validate that the floor 1 X values contain no duplicates. * 03http://tinyurl.com/d6u6z2a03 [17:58] ffmpeg.git 03Michael Niedermayer 07e28814e0e16c: Merge remote-tracking branch 'qatar/release/0.7' into release/0.8 * 03http://tinyurl.com/ce6cgz703 [18:03] ffmpeg.git 03Mans Rullgard 074eea330a2afa: h264: allow cropping to AVCodecContext.width/height * 03http://tinyurl.com/c5s7kdw03 [18:03] ffmpeg.git 03Mans Rullgard 0744e6cf3f75f0: mov: set AVCodecContext.width/height for h264 * 03http://tinyurl.com/cnsdbjb03 [18:03] ffmpeg.git 03Kostya Shishkov 07aa4121276777: vc1dec: add flush function for WMV9 and VC-1 decoders * 03http://tinyurl.com/cpzwot903 [18:03] ffmpeg.git 03Janne Grunau 07459feb7cce03: nuv: check RTjpeg header for validity * 03http://tinyurl.com/c5l56co03 [18:03] ffmpeg.git 03Janne Grunau 07fd7426ed8985: imgconvert: avoid undefined left shift in avcodec_find_best_pix_fmt * 03http://tinyurl.com/c7bojmm03 [18:03] ffmpeg.git 03Michael Niedermayer 07c82ae85a8a78: mpegvideo: Don't use ff_mspel_motion() for vc1 * 03http://tinyurl.com/c9ukn4j03 [18:03] ffmpeg.git 03Mina Nagy Zaki 07a4e277312cac: lavfi: avfilter_merge_formats: handle case where inputs are same * 03http://tinyurl.com/dxjcxub03 [18:03] ffmpeg.git 03Alex Converse 070e2f415adf5d: vorbis: Validate that the floor 1 X values contain no duplicates. * 03http://tinyurl.com/d29ve7903 [18:03] ffmpeg.git 03Michael Niedermayer 07f48d1fb16720: Merge remote-tracking branch 'qatar/release/0.6' into release/0.6 * 03http://tinyurl.com/coj95b503 [18:17] ffmpeg.git 03Kostya Shishkov 079125aa9218c3: vc1dec: add flush function for WMV9 and VC-1 decoders * 03http://tinyurl.com/dya8op403 [18:17] ffmpeg.git 03Janne Grunau 07f695be22d89a: nuv: check RTjpeg header for validity * 03http://tinyurl.com/cafb5r503 [18:17] ffmpeg.git 03Janne Grunau 077296a6b5e9c2: imgconvert: avoid undefined left shift in avcodec_find_best_pix_fmt * 03http://tinyurl.com/c95tkjn03 [18:17] ffmpeg.git 03Michael Niedermayer 076d6373dc6441: mpegvideo: Don't use ff_mspel_motion() for vc1 * 03http://tinyurl.com/csycyuy03 [18:17] ffmpeg.git 03Mina Nagy Zaki 070dfcbe5285f0: lavfi: avfilter_merge_formats: handle case where inputs are same * 03http://tinyurl.com/bwoltvy03 [18:17] ffmpeg.git 03Michael Niedermayer 07776fb2e10d4a: Merge remote-tracking branch 'qatar/release/0.5' into release/0.5 * 03http://tinyurl.com/d6665sl03 [19:04] fried potatoes or beans? [19:46] ffmpeg.git 03Michael Niedermayer 07dd3e5baa5996: swr-test: check pointers before use * 03http://tinyurl.com/8saxjkw03 [19:46] ffmpeg.git 03Michael Niedermayer 079de5b11d58ed: swr-test: fix division by 0 * 03http://tinyurl.com/8doxm2g03 [20:12] ffmpeg.git 03Michael Niedermayer 07340305646a82: riff: fix division by zero for G726 * 03http://tinyurl.com/9jh7fzw03 [20:12] ffmpeg.git 03Michael Niedermayer 07de6c15044415: nut: fix int32 overflow * 03http://tinyurl.com/8ggtx8803 [20:12] ffmpeg.git 03Michael Niedermayer 0724d6af0f7e5e: sierravmd: flip reduce arg order. * 03http://tinyurl.com/9dvadjm03 [20:31] burek: you could also use the gmail smtp server [21:47] michaelni: ping on "lavfi/scale: accept named option, make parsing more robust" [21:57] saste, does it still work with changing input resolution after the patch ? [21:57] I think so [21:58] how can i test it? [22:00] that is a good question [22:03] you can try to concatenate 2 mpegs with different size and chec per avlog that the scale filter gets differenz input sizes while the whole filter graphs is NOT rebuild [22:04] I believed we implemented normalization at some point, going to test it now [22:05] yaaargl [22:05] Action: ubitux realized something is wrong with the lavfi metadata patch :( [22:06] we dup the dict in the AVFrame in the copy props [22:06] but by default apps won't manually destroy the dict [22:06] and actually, nothing will [22:06] (except in case of the lavfi device with what we did) [22:07] any suggestion? :( [22:10] ubitux: why do you need to destroy the dict? [22:11] because there is one allocated per frame? [22:11] which will leak [22:11] ubitux, "decode_frame" free last, dup current, "codec close" free last [22:14] av_dict_free(&avctx->coded_frame->metadata); ? so we would move this from raw decoders up to the general decode frame? [22:15] ok let's try to activate my brain for once. [22:25] michaelni, same behavior as current (that is broken) [22:25] but somehow works [22:25] michaelni: codec close can only access to "coded frame", which is likely no pointing on the last decoded frame most of the time afaict [22:26] it is broken because is not doing the expected thing [22:26] saste, how can i reproduce the "broken" ? [22:26] when the size changes, it assumes the new size [22:27] ubitux, you can store a pointer to the dict in the codec context [22:27] for example if you have scale=2*iw:2*ih it will just keep the new "changed" size [22:27] michaelni: right sure, ok [22:27] i'd rather expect scale to resize the video accordingly [22:27] but => unrelated to the patch [22:30] saste, but its working if you use a fixed size like 400:300 ? [22:35] michaelni, my patch is correct in the sense that it doesn't change the current behavior [22:36] which is: when the size changes, assume the new input size as output, whatever expression the user set [22:39] michaelni, BTW can you have a look at the examples/muxing issue? [22:39] it is disturbing that we ship a broken example, and I don't know which is the best way to fix it [22:43] about the scale patch, id like to look at things before its applied, if theres a bug it should be fixed before changing the code [22:46] about the muxing issue, how can i reproduce it ? [22:47] michaelni, muxing out.mov [22:50] michaelni, i don't want to end up fixing the world before committing a simple change (which I have in my local repo since a few weeks) [22:50] so if the patches doesn't change the current behavior, i'm for applying it [23:16] puting a metadata in the codec context was actually so simple.. [23:16] life is a mess [23:17] (^^ in case you need some useful and constructive feedback) [23:18] :D [23:35] yepee! [23:35] it works, and no leak [23:36] thx a lot michaelni, it was actually the simplest thing to do.. [00:00] --- Wed Oct 17 2012 From burek021 at gmail.com Thu Oct 18 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Thu, 18 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121017 Message-ID: <20121018000502.08F4918A01F0@apolo.teamnet.rs> [00:00] "qp 0" ? [00:00] huh? [00:00] Mavrik_: I need a "near lossless" [00:01] Mista_D: for libx264? just use the highest crf value that still gives you an acceptable quality. [00:02] and sameq is gone. also it did not mean "same quality" [00:02] purge it from your memory! [00:02] llogan: ~10? [00:03] see if you can tell a difference between the source and -crf 18 [00:04] llogan: MSU VQMT will, and with enough zooming, anything is possible. [00:14] people watch videos, not "quality measurement tools". [00:17] llogan: you won't notice a change in a minute or 5, after an hour of watching a movie, you can tell crf 18 from original. [00:21] :) [00:22] Mista_D, crf assumes variable quality [00:22] over different scenes [00:22] so you won't accumulate any quality loss during time [00:23] crf is something like "quality factor" [00:23] it's not bandwidth limitation to accumulate loss if too much bitrate has been already used in past [00:24] burek: I'm up for pepsi challenge of AppleProres vs x264 cfr 18 (: [00:25] I don't use prores :) I believe you :) [00:35] Why would anyone want to use an intermediate codec that's lossy? [00:37] Someone alive? [00:38] pablomurad: I am, but burek is a zombie. And he bites people that ask useless questions. [00:39] (") [00:39] I don't believe my question is so useless, lol. I used to use ffmpegx and I couldn't edit the size of the subtitle and fonts too. [00:39] Many forums I read told me the problem was with the Ffmpeg binaries, is this true? [00:40] it might be, unless if it is not [00:40] We only support ffmpeg directly, not guis from other projects that interface with ffmpeg. [00:40] does it use ffmpeg for subtitle stuff? [00:41] pablomurad, if you can add some custom options to be forwarded to ffmpeg, add -report [00:41] or what is its name [00:42] I only knows that maybe the problem is with the last ffmpeg, I know you dont support another Apps, but it would be nice if you guys take a look on this. However, the actual binary is great [00:42] Sure, I will do it :) [00:42] pablomurad, can you somehow dig up actual ffmpeg logs, if you can, we might help [00:42] so turn on all the verbosity there is in your gui [00:42] It would be helpful if you could replicate the problem using ffmpeg. [00:42] I will get a log soon as possible!! thank you guys [02:24] practically, why would one choose to use -vf yadif=1 over -vf yadif=0 [02:24] i used yadif=1 on some home videos and it made it look so fluid that it didn't look right [02:30] the default is yadif=0 but with some content 1 might look better. [02:31] lake: This is old but you might learn something. http://www.100fps.com/ [02:36] relaxed: so it's more or less subjective? [02:36] i'll check that link out. thanks [02:42] lake: I would always use 1 [04:30] Hey I'm trying to use -i in.avs, however I keep having issues with it not loading my avs plugins. But they are sitting in the correct directorys and such. Is there something I'm missing? [05:41] relaxed: i'm interested to know if you agree with this article http://renomath.org/video/linux/interlace/ [13:06] i've made a fast motion video like that. http://paste.osuv.de/index.php/QCErl/ but how to display the first image for 5 seconds [13:33] or how to combine these two commands? http://paste.osuv.de/index.php/ru6CV/ the first image should take 5 seconds. and after that 11 images per second [14:12] markuman: you can't concatenate video with two different framerates. [14:13] relaxed: no i mean, the first image from time laps should display 5 seconds long (kind of introduction image). and then start the timelaps video [14:13] I would use rawvideo output for both and cat them into ffmpeg from stdin. [14:15] For the first command: ffmpeg -loop 1 -r 11 -i stapelzeitrafferintro.png -t 5 -pix_fmt yuv420p -f rawvideo 1.yuv [14:16] Second command: ffmpeg -r 11 -i $other_images -pix_fmt yuv420p -f rawvideo 2.yuv [14:18] Third: cat 1.yuv 2.yuv | ffmpeg -f rawvideo -pix_fmt yuv420p -r 11 -s $freamesize -i - -vcodec libvpx -b:v 5000k output.webm [14:19] where $framesize is the WxH of the images [14:21] also, there are presets for libvpx that you might want to look into, the defaults may suck. [14:21] k thx [14:23] here's the list: http://git.videolan.org/?p=ffmpeg.git;a=tree;f=presets;h=e64871fca371b05ebcd8531f66b705e5ba03236a;hb=HEAD [14:25] you would think libvpx would take a page out of libx264's book and make internal presets available [14:27] relaxed: i see you're using a bitrate for libvpx; were you able to achieve a decent quality with a quality factor or anything? [14:28] ubitux: I was just using the bitrate he chose. I honestly hope I never have to use libvpx because other than being more "open" it fails in everything compared to libx264. [14:29] Including ease of use. [14:29] i wasn't able to get a non-low bitrate quality by playing with the different quality, cfr etc [14:29] i can only get a good output with -b:v :( [14:30] Why do you need it? For html5? [14:35] relaxed: yes, playable inside the browser [14:35] quite handy for common videos [14:35] True, I guess that's where the web is heading. [14:39] if i want to show a sample to someone, i just make a sample in webm, upload it and give a direct link [14:39] that's way handy than looking for the video on youtube [14:39] get a crappy quality, bad edits, etc [14:42] You could achieve the same thing with an h264 video. [14:42] except that wouldn't be directly playable in the browser :( [14:42] And serve it from your website, dropbox, or whatever. [14:42] i wonder if html5 allow seeking through url [14:43] like a video attribute with a start [14:44] ubitux: It would be if you had a template html setup with jwplayer. Then you would only have to upload the video. [14:53] <@ubitux> except that wouldn't be directly playable in the browser :( <- why? [14:54] cause i don't want to use flash? :p [14:54] h264 video, not a container.. [14:54] ? [14:54] webm(h264+aac) [14:56] webm is vp8+vorbis [14:56] only that? [14:56] no wonder it sucks :) [14:56] well, for html5 I think [15:00] you can do (h264/aac).mp4 through html5 but most browsers won't ship with decoders. [15:01] why is it so difficult to create an open source container in the 21st century? [15:01] do we lack knowledge, time or what? [15:02] webm is pretty much matroska [15:06] the push to vp8/vorbis is due to patents [15:13] is there some way I can turn an m4v file to an mp4 file without recompressing it [15:13] I think you can simply rename it. [15:13] hmm [15:13] adobe premiere doesn't like it when we do that [15:14] I have no idea where you'd even GET an m4v file [15:14] ffmpeg -i input -map 0 -c copy output.mp4 [15:14] I mean I've come across m2v before. [15:14] Apple has to rename everything. [15:14] unrecognized option "c" [15:15] ffmpeg -i input -acodec copy -vcodec copy output.mp4 [15:15] that's what I did [15:15] and...? [15:15] hmn [15:16] premiere still only sees it as 3 frames long [15:16] but vlc plays it fine [15:16] Recompress time I guess :) [15:16] Action: relaxed blames Adobe [15:16] Action: Jan- too [15:16] ok, what would you suggest to comprehensively reencode this thing so it's completely redone [15:16] 14:56:23 relaxed | webm is vp8+vorbis [15:17] 14:56:48 burek | only that? [15:17] 14:56:55 burek | no wonder it sucks :) [15:17] sort of [15:17] if you put anything else in, it is no longer webm [15:17] but matroska [15:17] webm is basically a subformat of matroska only supporting these codecs [15:17] so a browser can say "we support WebM", which then means "matroska but only few codecs and no extra crap like attachments" [15:18] Jan- ffmpeg -i yourinput.m4v ? [15:18] what's inside? [15:18] divVerent thanks for clarification :) [15:18] it makes me think it sucks even more :D [15:18] then just use matroska :P [15:19] lavf demuxer BTW doesn't care at all if it is webm or matroska [15:20] and the muxer has a webm and a matroska mode, and all the webm mode does is skip/disable some matroska features [15:20] this actually does make sense though [15:20] WebM is a whole spec including codecs [15:20] so if WebM allowed "any codec", you could have a "WebM supporting" browser, and a "WebM video", but the browser won't play it [15:21] for Matroska you'd have exactly this situation [15:21] you could have a player with Matroska support that just can't decode the H.264 inside [15:22] burek wouldn't I need to set bitrates and suchlike [15:22] Jan-, what I meant was: "Can you type: ffmpeg -i yourinput" [15:22] and pastebin the output [15:22] so we can see what's inside [15:23] divVerent yes I understand [15:23] it's just funny how many people think html5 will auto-magically solve all their problems [15:23] and they praise it so much [15:23] hehe [15:23] until they test it [15:23] it almost did [15:23] :) [15:23] http://pastebin.com/br7LqwTF [15:23] if only it hadn't been for the codec quarrel [15:23] I wonder if that came off a blu-ray or something [15:23] and this isn't even WebM's fault [15:23] I mean, an mp4 file with chapters in it? [15:24] the issue is, half the world supports H.264, the other half supports WebM [15:24] and thus nothing was gained by html5 [15:24] uh Jan-... you really want to update your ffmpeg [15:24] because html5 did not decide on which codec/format to support [15:24] mainly Apple and Microsoft blocked WebM from being required in HTML5 [15:24] burek: no, I don't, because if i do that everything will change [15:24] i mean, if you want to encode files properly [15:24] and I won't be able to do ANYTHING. [15:24] well, ok [15:24] and many others blocked H.264, for obvious patent reasons [15:24] so in the end, HTML5 for video became worthless [15:26] webm is going nowhere the same way flac for audio is going nowhere. [15:26] Jan-, try to re-encode with: ffmpeg -i input -map 0 -acodec copy -scodec copy -vcodec libx264 -crf <0-50> -preset output.mp4 [15:26] de facto standards are HARD to move. [15:26] oh, right.. you can't even use -crf... [15:26] well I don't know what to suggest [15:27] Jan-: flac going nowhere? [15:27] use a static ffmpeg? [15:27] flac already IS somewhere [15:27] while webm didn't really establish itself, and never will [15:27] it's replacable, if Google Chrome from today to tomorrow stopped supporting WebM and would only do H.264, nobody would notice [15:27] I'm pretty sure I don't have access to -presets and if I did I wouldn't know which one to use [15:28] WebM was dead the moment Microsoft and Apple decided to block it [15:28] Jan-, too bad [15:28] that's the consequence of using ancient ffmpeg, sorry.. [15:28] you can use static ffmpeg [15:28] re-encode your files [15:28] and delete it [15:28] (ffmpeg) [15:29] while FLAC actually managed to establish itself DESPITE better formats being around (although none of the lossless formats got an absolute majority) [15:30] wav rules :) [15:30] simply because FLAC was quick enough and is now one of the only two lossless compressed audio codecs you can assume to be decodable even in 10 years [15:30] the other is probably Apple Lossless (ALAC) [15:31] and APE [15:31] the reason is mainly that when FLAC came, it actually filled a need... because it was clearly superior to other lossless formats [15:31] even about APE i wouldn't be so sure [15:31] given it has quite poor support on mobile devices for example [15:31] Unfortunately as we all know it's not really about it being better. [15:31] you find a flac player for anything [15:31] It's about a lot more than that. [15:31] but ape... [15:31] flac is little better than shorten [15:31] *compression [15:31] what is really great about flac is not the compression, but the support [15:33] but thing is, it doesn't even matter... [15:33] in the test I see here, the span between the best and worst lossless compressed format goes from 51% to 58% [15:34] nobody is gonna switch format just for 14% total space savings (and that's what you'd get by changing from the worst to the best) [15:34] now, look at WebM... it did not fill ANY need [15:34] we already HAD video playing in the browswer [15:34] can someone suggest a sensible value for -presety [15:34] sure, it was using awful codecs, and involved Flash [15:35] but that's mainly why WebM was a failure. It did NOT become the "supported everywhere" format it was trying to become... [15:35] mainly because two big players stopped it from becoming that [15:35] and if you don't need "supported everywhere", then you can just keep using H.264 which already was there for ages, including good encoders [15:36] which is "supported everywhere except where software patents are in the way" ;) [15:38] it says "the system cannot find the path specified" which is what it usually says when you try to use presets on windows [15:38] no matter where you put the presets it always just says that [15:42] pastie.org the command and all output [15:43] it's OK [15:43] I think I fixed it [15:43] what does it mean when it starts a line of output with "[libx264 @0xDEADBEEF]" or whatever [15:43] is that the memory location where x264 was loaded? [15:44] that looks like memory corruption if you actually see 0xDEADBEEF [15:45] otherwise, it's the address of the context object [15:45] con...text... object? [15:45] "some stuff holding some data" [15:45] Action: Jan- suspects this conversation is headed in the direction of load maps and function pointers, and begins to feel the Fear [15:46] Jan-: just type LOAD "*",8,1 to fix this [15:49] can "-bsf h264_mp4toannexb" be used with -c:v copy ? [15:50] and what exactly does this message say: [mpegts @ 0x995f40] H.264 bitstream malformed, no startcode found, use the h264_mp4toannexb bitstream filter (-bsf h264_mp4toannexb) [15:51] im just trying: bla1 | ffmpeg -f h264 -i - -c:v copy -f mpegts - | bla2 [15:52] Is the function location stuff output by av_log or something [15:52] I've not seen it on other programs [15:53] Jan-: it is [15:56] yay, I made an API reference about ffmpeg and it wasn't wrong :D [15:56] I knew it would happen one day :D [16:15] is there some way I can specify an actual bitrate [16:15] this was always hard with ffmpeg as it always seemed to ignore -b [17:40] why would ffmpeg create a video with no audio track [17:44] ...aargh premiere doesn't like AAC audio [17:44] hi guys [17:45] how do I just get it to do uncompressed audio!? [17:45] learn me please. 29.97 is that actually 30fps ? [17:45] im looking at iphone footage [17:45] iphone 4 [17:45] sine strictly speaking no but it's often mislabeled [17:46] how do i correctly dump the frames as png and then recombine them [17:46] dumping is not a problem i suppse its the recombining [17:47] ffmpeg -i input %04d.png? [17:47] Not sure. [17:47] I would really like to know how to make it do AAC or uncompressed audio. [18:19] Jan-: -acodec pcm_s16le [18:55] hi there channel :-) [18:55] anyone available [18:56] my ffmpeg doesn't work properly in ubuntu 12.04.1 [18:56] can anyone help me to solve it, please? [18:57] have you built ffmpeg yourself or are you using the ffmpeg binary that comes with Ubuntu? [18:59] JEEBsv: I've tried to extract the audio from .mp4 file using it.. it says to use avconv instead of ffmpeg [18:59] JEEBsv: I guess the ones that comes with ubuntu [18:59] so you are using the 'ffmpeg' binary that comes with ubuntu's default packaging. Debian/Ubuntu uses a fork project called libav and not ffmpeg. So yes, you should switch to using the 'avconv' command if you want to use that [19:01] Basically libav rewrote parts of the ffmpeg app and renamed it 'avconv', and then left the 'ffmpeg' binary to rot for a short while. The ffmpeg project merged those updates into the ffmpeg app. [19:02] if you have problems with avconv you can either hit #libav with them, or try the static builds available from the links on the topic etc. [19:02] JEEBsv: I'm going to try one and give you the error results ok? [19:02] -.- you should point that towards #libav then [19:05] JEEBsv: tell me an url the paste the results for you [19:08] JEEBsv: I've found one pastebin [19:08] JEEBsv: http://pastebin.com/RuP3bAM6 [19:09] I think that's pretty obvious and actually doesn't have anything to do with the ffmpeg binary being old [19:09] you don't have the version of libavcodec installed that has the "extra" encoders and so forth [19:09] libavcodec-extra I think? [19:17] JEEBsv: if I do sudo apt-get install libavcodec-extra-53.... it says the following extra packages will be installed libavutil-extra-51 and libopenjpeg2 Suggested packages: libfaad0 the following packages will be removed: libavcodec53, libavutil51 [19:18] yes, because it will remove the libraries without those extra things linked in :P [19:19] JEEBsv: should I confirm them? [19:19] yes [19:19] JEEBsv: all right [19:20] done it [19:21] JEEBsv: so suppose to do that ffmpeg will work fine, then? [19:21] JEEBsv: or avconv? [19:22] avconv is the newer one in the setup that debian/ubuntu has made, so if you are going to use their packages then yes, use avconv -- as that binary is more up-to-date than the ffmpeg that they give out [19:22] (ugh, my English, ended up saying something twice >_>) [19:23] JEEBsv: let me test it [19:24] JEEBsv: one sec. [19:24] also if you are going to be using that ubuntu package, I recommend you switch to #libav [19:24] because they use that project, while other distros use ffmpeg [20:20] Ok, this is confusing: [20:20] *** THIS PROGRAM IS DEPRECATED *** [20:20] This program is only provided for compatibility and will be removed in a future release. Please use avconv instead. [20:21] What does that mean? [20:21] it says to use avconv [20:22] which is the forked equivalent of ffmpeg [20:22] Oh, a fork. [20:24] And what are the differences, not counting need to modify my scripts? [20:24] nothing really, they operate the same [20:26] So just a great way to annoy users. [20:27] I think that was everyone's intended goal for the fork. [20:31] It seems so. [20:32] I'll have to change every reference of ffmpeg in the scripts. Easy but annoying. [20:34] avconv != ffmpeg [20:35] avconv: Like ffmpeg but changing some commandline options. [20:35] That would be a quite good description for whatis [20:36] it is not same, avconv use different lib [20:36] Nobody said it was the same. [20:36] you did [20:36] durandal_1707, Did not say same, said like [20:36] You misread. [20:37] >> relaxed: nothing really, they operate the same [20:38] THEY DO [20:38] putting caps will not make statement true [20:38] Then stop trolling me. [20:39] They do operate the same. I use both quite often. [20:39] durandal_1707, Then state why they do NOT operate the same. [20:40] they are different and use differen lib so can not operate same, but may operate same some time, but not all the time [20:42] wlan3: If you need support for the evil fork that does not operate 100% the same as ffmpeg, I would be happy to help you in #libav. [20:42] durandal_1707, what I see is that I have to change my scripts or manually build ffmpeg times and times for my scritps to work. [20:43] relaxed, It would be useful, but first I needed to complain a bit. [20:43] wlan3: only if you use ubuntu [20:44] or debian [20:44] durandal_1707, I use debian on most of my systems. [20:45] wlan3: your choice - i'm not forcing you to use anything, your pick [20:46] durandal_1707, but you said ONLY if you use ubuntu. [20:47] if you do not want to use non-system build or whatever it is called [20:49] durandal_1707, for unattended systems, Debian is the best working I've used so far, if you want to suggest me another option, i would investigate a bit. [20:49] what another option? [20:49] Debian is best. [20:50] durandal_1707, You should know more than me. [20:50] about linux distros? i do not know much, because i use FreeBSD [20:51] I prefer hurd [20:51] But, since it is still incomplete, I use Linux [21:02] I have a load of png from a video which was 29.97 fps [21:03] how can i encode them back again [21:03] can i lump the pngs together or do they need to be rerendered/encoded [21:03] sure you can [21:03] wanna keep max qual if you know what i mean [21:03] since everything is posible [21:04] Then you should read man ffmpeg [21:04] ffmpeg -i 0001.png -vcodec h264 /out/lolcakes.mp4 [21:05] maybe [21:05] ffmpeg [21:05] sorry [21:05] ffmpeg -f image2 -i %4d.png -r 25 -sameq myvideo.mpg [21:05] no, sameq is not what you think [21:06] ffmpeg -f image2 -r 24000/1001 -i C:\BLENDERTEST\%04d.png -vcodec mpeg2video -qscale 3 C:\BLENDERTEST\output.mpg [21:07] i want it to keep the quality though [21:07] My eyeballs hurt [21:07] i need it 29.97 as well [21:08] jeeb can you hit me with the syntax please [21:08] ffmpeg does that by default [21:08] I think (that about keeping quality [21:08] but i need a container [21:08] x264 [21:08] I think (that about keeping quality <- no [21:08] mp4 [21:09] ffmpeg or anything else has absolutely no idea of how to "keep quality" [21:09] we don't have thousands of chinese children in basements [21:09] looking at various results [21:09] i just want to watch my cgi [21:09] we don't yet have things that work like your eyes and brains :) [21:10] ffmpeg -r 30000/1001 -i %4d -c:v libx264 -crf 18 out.mp4 [21:10] for example [21:11] ugh, %4d.png I meant [21:11] and the 3000/1001 is the 29.97 [21:11] -c:v is a newer way of saying "vcodec", libx264 is the H.264 encoder [21:11] oh ok cool [21:11] codec video [21:11] and crf is the rate control mode that is closest to "constant quality" [21:12] the default is 23, and lower is more bits thrown at the source, higher is less bits thrown at the source [21:12] 18 should be more or less nice-looking esp. with HD content, SD content depends on the actual content and your eyes [21:12] JEEB, I meant, the output comes from the input, so if you do not alter options, the quality of the output is *similar* to that from the input. [21:12] tun0, nope [21:13] unless you mean that the defaults for various encoders being 200kbps mean that ffmpeg wants you to have the "similar quality" :P [21:13] only mostly used encoders like libx264 have sane defaults [21:13] like libx264 having crf 23 as the current default that looks pretty good for most people while compressing nicely [21:13] sine_, idea with crf is to find the highest crf value that still looks good for you [21:13] cool its working [21:13] im saving the code in my treepad lite. i have a section for ffmpeg [21:15] JEEB, when I want nice quality, I just use out.webm as output. [21:15] no idea what the vp8 default for libvpx is [21:15] you might want to check it out [21:15] thanks it worked. however my first cgi render is crap! [21:15] but yes, it's a newer encoder so there's a chance it's a nice default [21:15] sine_, huh [21:15] and how are you checking? [21:16] you talking to me ? [21:16] the line beginning with sine_ and the next from it, yes [21:17] im watching my cgi render ? [21:17] ive just encoded it [21:18] it was my first camera mapping experiment. i worked out mapping markers etc [21:18] I'm asking you to be somewhat less vague about various things if only possible [21:18] since it seems like the output wasn't good enough for you [21:18] no, i mean my tracking and cgi modelling skills to the track data [21:18] not the ffmpeg render [21:18] k [21:37] hey [21:37] xwma is wma ? [21:39] never hoid of it. [21:40] <- Mr. Usefull [21:45] tun0: you also forgot to weigh the fact that repository packages of "ffaux-ffmpeg"/avconv or ffmpeg age relatively quickly compared to development [21:46] hey llogan [21:49] hi [21:52] burek: I choose to help users with the tool they have handy instead of wasting both of our time trying to install another tool that does the same thing. Also, I lurk in #ffmpeg-devel :P [21:53] If someone is in #libav and they're using FFmpeg's ffmpeg I will help them too. [21:54] Why? Because I don't give a shit about the politics between the two camps. [21:54] relaxed I know you do, just it might be better to tell them what that is all about [21:54] and redirect them where they need to go [21:54] no need to waste time in here [21:54] I don't look at it as a waste. [21:55] and also, if you dont give a shit, which you dont, then no need to tell people it's the same thing [21:55] because it is not [21:56] btw, there is no avconv's ffmpeg (at least it shouldn't be, due to copyright) [21:56] libav's avconv [21:57] if it's a fork, then there can't be avconv's ffmpeg [21:57] there can only be avconv [21:57] they should have named it ffmpeg2 [21:57] they shouldn't [21:57] because it's not [21:57] it's a 3rd party fork [21:57] I know the whole story. [22:03] I've been helping people in this channel for years. Longer than any of the familiar names I see here today. I will continue to do so the way I have regardless of anyone's political agenda. If an op doesn't like it they can man up and ban me. Until please stop the whining! [22:03] s/Until/Until then/ [22:04] I'm not whining, I'm just saying you should stop telling people something that's not true [22:04] I don't have anything against you, personally [22:04] it's just wrong to tell them something like that because that doesn't help them it just confuses them even more [22:05] regardless of the "politics" [22:05] llogan, packages age? Did not understand [22:06] i simply mean that compared to git-master the packages are generally old and smell like moldy shoes. [22:06] i just heard, ffmpeg was to be deprecated? [22:06] lol [22:06] Action: llogan sighs [22:07] cbdev: here is a summary: http://stackoverflow.com/a/9477756/1109017 [22:07] I do agree that message was the ultimate troll. Keep in mind i didn't write it. [22:07] Or feel the n eed to explain it everytime someone comes in here asking about it. [22:08] ah i see :) [22:08] thanks [22:08] the link refers to an older, crappier message, but it's still relavent [22:13] Action: relaxed deprecates burek [22:13] :) :beer: [22:13] :) [22:14] Action: wlan2 deprecates ppp0 [22:15] burek: the word "deprecated" should trigger the bot [22:16] it could be like the secret word on peewee's playhouse. [22:16] relaxed, should trigger what action? [22:17] a message explaining the fork, etc... [22:18] So that we will never have to have this discussion again. [22:19] Good idea. [22:19] Also explaining why the debian package maintainer sucks? [22:20] Or trolls, if you prefer [22:20] well, there's enought trolling already. [22:20] enough* [22:21] Also, I have a friend. [22:22] She usually uses windows for recording "tv" [22:22] And I was trying to figure which command she should use in ubuntu. [22:23] burek: !fork pointing to an *external* source explaining the forking status? [22:24] I told her to try "mplayer tv://" But even that does not work. [22:25] which external source saste? [22:25] do you have a link [22:25] stackoverflow may be one [22:26] we (and with "we" I mean "you") could add a link to more resources [22:26] internal resources may be biased [22:26] :) [22:26] or a collection of links from both ff* and av* and external resources [22:27] I like all that "biased" talk :) [22:27] they intentionally put a message saying ffmpeg is dead [22:27] and you are playing "fair" with them :) [22:27] cool :) [22:27] they even try to take over the domain [22:27] but that's ok too :) [22:27] let's not be biased :) [22:28] let the users decide for them [22:28] (even if they mostly don't care at all, provided the bloody thing works for them) [22:28] exactly :) [22:28] they made some steps to lead the users "in the right direction" [22:29] so they are more than biased if you ask me [22:29] and now we bear the burden of that kind of behavior [22:29] but, as i said, let's not be biased after all :) [22:48] wlan2: "does not work" is a common description of a multitude of problems. [22:57] llogan, she is descriptive no more than that [22:59] the capturer is one of this freacky things that encode the video themselves and send it throught usb [23:03] https://pad.riseup.net/p/nGsg5q9DrLAI [23:05] That was all she gave me when I asked a big bit of times [23:17] Action: llogan kicks this router [23:40] I'm trying to do point to point streaming using rtsp but not working yet. The command im using and its output: http://pastebin.com/2TUyMRq6 Any suggestion/example that can iluminate me? Thanks! [23:43] rtsp/rtp/udp ? [23:51] burek: rtsp using udp [00:00] --- Thu Oct 18 2012 From burek021 at gmail.com Thu Oct 18 02:05:03 2012 From: burek021 at gmail.com (burek) Date: Thu, 18 Oct 2012 02:05:03 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121017 Message-ID: <20121018000503.1849C18A01F1@apolo.teamnet.rs> [00:11] ffmpeg.git 03Michael Niedermayer 076d55a40b0080: mov: print warning if ff_get_wav_header() fails * 03http://tinyurl.com/9vq6jdd03 [00:11] ffmpeg.git 03Michael Niedermayer 07cb65b32c97b0: mp3enc: remove unneeded null ptr check * 03http://tinyurl.com/8lgblqe03 [00:11] ffmpeg.git 03Michael Niedermayer 07a30972609ca3: yuvPlanartoyuy2_c: fix sign extension * 03http://tinyurl.com/8mblbvd03 [00:11] ffmpeg.git 03Michael Niedermayer 07a07e9d72a1d9: yuvPlanartouyvy_c: fix sign extension * 03http://tinyurl.com/9zhvwcu03 [00:11] ffmpeg.git 03Michael Niedermayer 073e0b29ccd075: ffmpeg: Make video filter graph reinit user selectable * 03http://tinyurl.com/9grcj2w03 [00:11] ffmpeg.git 03Michael Niedermayer 076cbb8a450f16: libavfilter/buffersrc: Do not fail hard on changes of input parameters. * 03http://tinyurl.com/8qmelxq03 [00:11] ffmpeg.git 03Michael Niedermayer 075d2b8850746b: lavfi: limit matching w/h/fmt asserts to non scale filters * 03http://tinyurl.com/98mjdlp03 [00:32] saste, your patch breaks compex filters [00:32] example: ffmpeg -i matrixbench_mpeg2.mpg -i tests/lena.pnm -filter_complex 'overlay=10:main_h-overlay_h-10' out.avi [00:32] michaelni, which patch? [00:33] "lavfi/scale: accept named option, make parsing more robus" [00:33] uhm, how's that? [00:33] auto insert maybe? [00:34] dont ask me iam just testing as i had a bad feeling [00:34] No option name near '(null)' [00:34] ok got it [00:35] well that is a pre-existing bug, but sure it must be solved [00:35] we don't have FATE test for -filter_complex? [00:36] i run make fate with my patchset and didn't catch the issue [00:37] goodnight people [00:37] n8 saste [00:37] [18:37] <@Daemon404> how do i make a build-only fate instance? [00:37] same question here [00:39] Daemon404, try makeopts="FATE=" [00:39] "try" ? [00:40] sorry wrong escaping, try 'makeopts="FATE="' [00:40] :) [00:41] still using that trick? [00:42] is that really the official way? [00:42] i mean, that's the way i workaround the problem but.. :D [00:42] 'official [00:42] well once the tilera patch is merged [00:42] ill add my tilera build to ffmpeg's fate too [00:42] Daemon404: you have some config examples on lucy.pkh.me btw [00:42] http://lucy.pkh.me/fate-configs/ [00:45] i never tried to find another way because it works ... [00:45] :D [00:46] i think i saw a build_only patch somewhere on libav [00:46] so even if there is another solution, it's likely to be hacky as well [01:25] michaelni: examples make works for me: http://pastie.org/5070391 [01:28] michaelni: maybe a shared build? [01:45] #define CONFIG_SHARED 0 [01:47] theres a "-L/usr/local/lib" in there, if i remove it it links [01:48] cc muxing.o -pthread -Lpc-uninstalled/../../../libavdevice -L/usr/local/lib -Lpc-uninstalled/../../../libavfilter ... [01:51] is it better if you replace: [01:52] CFLAGS := $(shell pkg-config --cflags $(FFMPEG_LIBS)) $(CFLAGS) [01:52] LDLIBS := $(shell pkg-config --libs $(FFMPEG_LIBS)) $(LDLIBS) [01:52] with: [01:52] CFLAGS += $(shell pkg-config --cflags $(FFMPEG_LIBS)) [01:52] LDLIBS += $(shell pkg-config --libs $(FFMPEG_LIBS)) [01:52] ? [01:53] no, same problem [01:53] ok :( [01:54] well i have to wake up ~5h so time for me to sleep.. [01:54] 'night [01:56] ubitux, good night [02:45] ffmpeg.git 03Michael Niedermayer 077b8fd298161d: examples/muxing: fix video pts * 03http://tinyurl.com/9lu7wzx03 [02:54] ffmpeg.git 03Michael Niedermayer 07adbb75dbd8ea: mov: fix time types related to mov_metadata_creation_time * 03http://tinyurl.com/8t2524303 [03:37] ffmpeg.git 03Michael Niedermayer 074e2e3d943ef0: ffv1: fix packed rgb with 1.3 * 03http://tinyurl.com/9h7xxos03 [07:46] ffmpeg.git 03Cl?ment BSsch 07711ffb84df29: lavf/swfdec: support DefineBitsLossless{,2} tag. * 03http://tinyurl.com/93kes8303 [10:05] is the aac ssr profile in use? [10:05] i mean, is there some encoders making use of it, and at least a few decoders supporting it? [10:51] i dont think i ever saw a file [10:51] and usually people give me the weirdest sample files [10:53] ok :) [11:17] how long in minutes is 2gb wav file (44100hz, 16bps, 2channels) ? [11:19] 2*1024*1024*1024/(44100*2*2) [11:19] 203 minutes or so [11:28] ubitux: I saw one ssr file [11:28] and then you woke up ? [11:28] according to a comment on top of the aacdec, there was a gsoc to support it [11:29] i guess it never reached the review process? [11:29] no diego wanted the ssr stuff from the gsoc resurrected [11:29] then I woke up [11:29] :D [11:30] no one uses it though [11:30] faad supports it iirc [11:30] i cant generate silence with 2 chan with aevalsrc [11:31] aevalsrc=0 ... -ac ? :P [11:32] i get mono [11:32] and if i specify layout i get error [11:32] aevalsrc=0:0 ? [11:33] thanks [12:02] ffmpeg.git 03Michael Niedermayer 07364c60bf64a2: sws-test: parse command line args before initing contexts * 03http://tinyurl.com/cdmbgcv03 [13:59] ffmpeg.git 03Paul B Mahol 07d6ea59b8608d: tta: datalen is unsigned integer per reference library * 03http://tinyurl.com/cvjzkyg03 [14:13] ffmpeg.git 03Paul B Mahol 071ade37ae9c2b: lavc/tta: use meaningful error codes * 03http://tinyurl.com/cwts6hp03 [14:21] ffmpeg.git 03Mans Rullgard 073b20eb25e7b9: avserver: move avserver-specific code from ffmdec.c to avserver.c * 03http://tinyurl.com/cgnopxh03 [14:21] ffmpeg.git 03Luca Barbato 0721de6ba5c12f: nut: export codec_tag provided by rawvideo * 03http://tinyurl.com/cd9zvdl03 [14:21] ffmpeg.git 03Mans Rullgard 071fbaabefc4bb: network: #include stdint.h in network.h * 03http://tinyurl.com/c6x2g4d03 [14:21] ffmpeg.git 03Diego Biurrun 07c0a6cac2920e: fate: Add rangecoder test * 03http://tinyurl.com/clec3rs03 [14:21] ffmpeg.git 03Luca Barbato 079a978b334b9b: ffv1: K&R formatting cosmetics * 03http://tinyurl.com/c54ppwe03 [14:21] ffmpeg.git 03Christian Schmidt 074a7429203a39: pcm-mpeg: correct bitrate calculation * 03http://tinyurl.com/c2y73uu03 [14:21] ffmpeg.git 03Rafa?l Carr? 07a25d912dca9c: avcodec_encode_audio(): fix invalid free * 03http://tinyurl.com/cak54ek03 [14:21] ffmpeg.git 03Michael Niedermayer 07d6e87190fd07: Merge commit 'a25d912dca9cd553440167e0476c47581359c0fc' * 03http://tinyurl.com/brazqj603 [14:34] ffmpeg.git 03Anton Khirnov 07a119c64e38bd: avconv: fix disabling auto mappings with -map_metadata * 03http://tinyurl.com/cavmbvw03 [14:34] ffmpeg.git 03Victor Vasiliev 0771e92414bfd7: lavf: move RIFF INFO tag writing from avienc to riff * 03http://tinyurl.com/c4vkeu603 [14:34] ffmpeg.git 03Michael Niedermayer 07fadfbb354b30: Merge commit '71e92414bfd79e56ea6fff174a665ff7b9b86e68' * 03http://tinyurl.com/bt5kxc803 [14:59] ffmpeg.git 03Victor Vasiliev 070bca0283ccde: riff: do not write empty INFO tags * 03http://tinyurl.com/cvholko03 [14:59] ffmpeg.git 03Michael Niedermayer 07c079da507316: Merge commit '0bca0283ccded5e32da143a462168ad1988a58fd' * 03http://tinyurl.com/bo7umz503 [15:33] ffmpeg.git 03Victor Vasiliev 0758b619c8a226: wav muxer: write metadata * 03http://tinyurl.com/cxf7dsa03 [15:33] ffmpeg.git 03Michael Niedermayer 07d8cfa9835804: Merge commit '58b619c8a226cc4564ad5af291bc99a04f89ee56' * 03http://tinyurl.com/bsok2xw03 [15:40] one stupid question.. what does -flags +global_header apply to? (streams or containers) [15:47] ffmpeg.git 03Anton Khirnov 0731c54711cc3f: lavf: split wav muxer and demuxer into separate files. * 03http://tinyurl.com/dyl6zz803 [15:47] ffmpeg.git 03Michael Niedermayer 07df5e089da95b: Merge commit '31c54711cc3f1484af101d629bbb805820d37ad1' * 03http://tinyurl.com/c3bffys03 [15:51] wee - I got an IFF file at work [15:51] it's 1985 again [15:52] why you get such files? [15:52] ffmpeg.git 03Anton Khirnov 0779922d7237ab: wav: do not fail on empty INFO tags * 03http://tinyurl.com/crto9gw03 [15:52] ffmpeg.git 03Michael Niedermayer 07940ee636301a: Merge commit '79922d7237aba2b8c6abbd2e06a0c08e4f498ad4' * 03http://tinyurl.com/cotnawj03 [15:52] I have no idea. but the client seems to want support for them [15:53] might be time to do my old "split wav into wavenc/dec, then share code with the avi demuxer" idea [15:53] in which case an iff demuxer could share code too [15:58] Action: durandal_1707 notices conflict list is becoming bigger and bigger [16:00] ffmpeg.git 03Anton Khirnov 07a43283b6f4cf: wavdec: check size before reading the data, not after. * 03http://tinyurl.com/d99cgce03 [16:00] ffmpeg.git 03Derek Buitenhuis 07c75848cd4c09: configure: Add support for Tilera processors * 03http://tinyurl.com/bn6h47h03 [16:00] ffmpeg.git 03Michael Niedermayer 07775d41b617f0: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/d4przjb03 [16:00] does anyone have any idea of what particular case Nicolas is talking about in the lavfi metadata support? [16:10] ffmpeg.git 03Paul B Mahol 079b762e2cba16: idcinvideo: remove redundant " id CIN Video: " from av_log() * 03http://tinyurl.com/cj7b9p203 [17:11] saste: http://fate.ffmpeg.org/log.cgi?time=20121017114713&log=configure&slot=x86_64-archlinux-gcc-enableshared [17:11] didn't you send some patches for this? [17:12] mmh or maybe it was carl [17:13] it was carl [17:13] and it was pushed [17:13] "just"? [17:13] no [17:13] and the patch was to make it fail when it wasnt foudn [17:14] why did i read "just".... [17:14] instead of silently contining [17:14] ok [17:14] i'll mail him, thanks :) [17:14] maybe you should check why it fails first :P [17:14] yes [17:14] though, it's a syntax error :p [17:15] lol [17:53] michaelni: that dca merge looks fishy [17:54] ffmpeg.git 03Michael Niedermayer 07a4fe661157b2: mov_probe: fix integer overflows * 03http://tinyurl.com/blobmm603 [18:00] eh all my typo are spotted [18:16] tilegx and tilepro fate builds added [19:55] what is nice way to look code changes of actual merges? [19:55] git cli or web? none [19:56] there might be a decent way in one of git's many guis [19:56] (probably mac based...) [19:56] cli only [19:56] there's only the ugly way [19:56] which is the obvious one [19:57] git is just for linux kernel maintainer [19:58] hmm, so looks like i need to setup this web gui thing [19:58] git instaweb ? [19:59] there's gitk too [19:59] but if you want to set something up, cgit [19:59] gitk isn't bad [19:59] ohsix, none of those let you see merge stuff that easily [19:59] gitweb's is horrible [20:00] wah, i look at that all the time on gitweb [20:00] you mean with extra context, not just the changes? [20:00] real changes to files no virtual crap [20:00] merge conflicts, any extra content in the merge commits themselves [20:01] and how it applies to THIS tree [20:01] not the tree you merged it from [20:01] so you mean between branches [20:01] branches is too specific [20:01] just trees [20:02] trees have branches, one or more, practically speaking it's always between branches [20:02] i suppose [20:02] a tree has all branches, comparing all branches to another set of all branches doesn't make much sense [20:03] but thats not the definition of tree i use [20:03] http://git.alsa-project.org/?p=alsa-lib.git;a=shortlog;h=HEAD commitdiff is what i use all the time, and it's good enough; interested to hear where it fails [20:03] ohsix, for example [20:03] micael merges the latest batch of libav changes into ffmpeg [20:03] via git merge [20:03] merge conflicts etc [20:04] or maybe it applies slightly differently to teh ffmpeg tree [20:04] but ig you view teh diffs for the commits themselevs [20:04] theyre always against libav's, nto ffmpeg's [20:04] and the merge commit itself is very confusing to look at [20:04] if at all even relevant [20:04] ah, can't help the confusion there, i think; because when you do a merge, they _are_ against libav's [20:05] exactly [20:05] but it's really annoying to track merged things [20:05] you would have to do all the patches individually for that to not be [20:05] there's no good way to view teh dleta as it was applied to our tree [20:05] and there really should be [20:05] since thats what it ends up as [20:05] ok, i took it to mean the web/gui interfaces were bad, not that merging was confusing [20:06] no webgui/gui displays it well [20:06] :P [20:06] afaik [20:06] you put them on a different branch and then compare them, merging only when needed [20:06] right, but the web ui isn't what's confusing you, the merge is [20:06] no i mean its way of displaying the merge [20:06] is poor [20:06] double + [20:06] like [20:06] + [20:06] ++ a [20:06] -+ a [20:06] its a shitty way to visualize it [20:07] maybe i've gotten used to it, or it never really mattered; i happen to agree with you tho i couldn't think of a better or less interesting way to do it [20:07] ;p [20:08] interesting in the "may you live in interesting times" sense, not wow this is amazing sense [20:08] i use log -m but it forks only for the first merge.... [20:10] hm, i'm not sure i wouldn't just apply the merge to another branch and compare them if i was in that situation often, isn't the patch posted before merges are done? or is this a situation where you land in the middle of a merge and you need to figure out what happened [20:11] i'm presuming you want to review a merge, not like, doing a bisect then having to untangle a merge [20:14] log -m is doing job well except i need to manually pick each merge [20:15] hm, well, i might be able to think of some other useful tool if the use-case is different [20:17] bleh [20:17] *** drop! [20:17] ^ what causes this [20:17] falling into the middle of a series of patches is usually confusing, merge or not; so i hadn't considered it as a source of extra confusion :p thus only thinking about reviews [20:18] git grep? probably a frame with a bad timestamp or something [20:18] (or an actual, genuine missed frame) [20:19] yes im looking [20:20] maybe ill send a patch for a less shitty message [20:20] by default it just drops thousands of frames for this clip\ [20:20] and the "verbose" long is "*** drop!" [20:20] s/long/log/ [20:21] it eventually drops the "your computer may be too slow!" or maybe i'm thinking of mplayer [20:21] no [20:21] nb_frames = FFMIN(nb_frames, ost->max_frames - ost->frame_number); [20:21] if this is 0 [20:21] it drops the frame [20:21] ah something else then [20:21] ok, gotta run; bbl [20:21] of course there are no comments. [20:25] and there is no indication what max_frames is at all [20:25] :| [20:26] Action: durandal_1707 yes : git lgs --min-parents=1 libavcodec/dcadec.c [20:26] looks like INT64_MAX... [20:27] Action: Daemon404 guess he asks michaelni wtf is going on [20:35] ffmpeg.git 03Michael Niedermayer 07fd9e88fe6018: libavfilter/lavfutils: remove useless NULL check on codec context * 03http://tinyurl.com/cwlzux403 [20:36] ffmpeg.git 03Michael Niedermayer 07657998b5ee46: libavfilter/lavfutils: remove useless NULL check on format context * 03http://tinyurl.com/d4xmvd203 [20:36] ffmpeg.git 03Michael Niedermayer 077fd65104f489: ffm_seek: fix division by zero * 03http://tinyurl.com/c7ez86k03 [20:36] ffmpeg.git 03Michael Niedermayer 07378a5b9c5f1a: ffm_write_write_index: check lseek() return code * 03http://tinyurl.com/c8v27fa03 [20:36] ffmpeg.git 03Michael Niedermayer 0771bc8c95d7ca: ffm_read_write_index: check lseek return code * 03http://tinyurl.com/d9chjb403 [20:50] Daemon404: is it going to be reference frames or something? cuz there'd be no need to drop them unless something was exceeded, or they got too old [20:51] its not clear whats going on [20:51] right, but there are only a handful of circumstances where frames would be dropped [20:51] small enough to investigate each one, anyways [20:51] [14:33] <@elenril> in ffmpeg it's dropping frames when 1) two frames have the same timestamps or 2) you want cfr [20:51] [14:44] <@Daemon404> elenril, that cant be accurate... [20:51] [14:44] <@Daemon404> this is neitehr vfr input [20:51] [14:44] <@Daemon404> nor are there same timestamps [20:51] [14:45] <@elenril> Daemon404: my crystal ball is out of ord [20:51] [...] [20:52] [14:35] <@Daemon404> avconv's behavior is SO much better [20:52] [14:36] <@Daemon404> Press ctrl-c to stop encoding [20:52] [14:36] <@Daemon404> ^C^C^C^C^C 1 fps= 0 q=0.0 size= -0kB time=0.01 bitrate= -17.6kbits/s [20:52] [14:36] <@Daemon404> and starts using ALL of my memory [20:52] ah, interesting [20:53] i'm not in a situation where i can read the code, so i won't distract you further [20:54] perhaps it keeps buffering stuff because timestamps are nonsense [20:54] the timestamps are fine [20:54] i checked them maually [20:54] what container? [20:54] mp4 [20:57] get sample so i can reproduce and play little with it? [20:57] *got [21:00] i dont think i can share this sample [21:00] work-related [21:05] how many streams it have? [21:06] ffmpeg.git 03Michael Niedermayer 07a0e0e1e19254: ffmdec: fix hypothetical overflows * 03http://tinyurl.com/cw3cn8r03 [21:06] ffmpeg.git 03Michael Niedermayer 07f03c0f6afcb1: ffmdec: check av_new_packet() return value * 03http://tinyurl.com/bsltcsk03 [21:06] ffmpeg.git 03Michael Niedermayer 07d185c8a79bbd: tiff: run strlen() after setting the pointer * 03http://tinyurl.com/bqv9wzm03 [21:08] Daemon404: did you check them with -debug_ts? [21:13] ubitux, with ffms2 [21:13] but same shared lib [21:14] -debug_ts output is ... dense... to say the least [21:14] :( [21:26] so riff info chunks are not 2byte aligned? [21:33] saste, michaelni (and others): i'm going to rename AV_PKT_DATA_METADATA to FF_PKT_DATA_METADATA in the next version of the lavfi metadata patch if you don't mind [21:33] ffmpeg.git 03Nicolas George 07709628aa71f2: lavfi/avf_concat: fix invalid exclusive test. * 03http://tinyurl.com/dywb5h603 [21:33] i'm actually pretty uncomfortable with what is libav going to do [21:33] name will not help [21:33] and i'm afraid they will introduce a AV_PKT_DATA_METADATA as well at some point [21:33] i thought FF_ was for internal only [21:34] Daemon404: yes, but there is no point (at the moment) to make this side data available for users [21:34] ubitux: AV_PKT_DATA_METADATA=666 [21:34] as long as FF_* isnt public [21:34] durandal_1707: that doesn't solve the issue, it's already the case [21:34] durandal_1707: the problem is if they use the same name [21:35] then put our initials in name [21:35] i was just proposing to use FF so it's private and it doesn't matter [21:36] AV_PKT_DATA_METADATA_BETTEREST [21:36] and it can be made public without much effort [21:36] llogan: AV_PKT_DATA_METADATA_FFMPEGGOTITFIRST [21:36] if you make something with a FF_ prefix public [21:36] that is wrong [21:36] and inconsistent [21:37] just FFMPEG [21:37] is enough [21:37] so you would prefer a public AV_PKT_DATA_METADATA_FFMPEG? [21:37] and thats just inconsistent with the entirety of thepublic api [21:38] AV_PKT_DATA_METADATA_LOL [21:38] Daemon404: you should worry, as a user, about a FF_PKT_DATAxxx defined somewhere imo [21:38] durandal_1707: AV_PKT_DATA_METADATA_COME_ON_LIBAV_BREAK_API_IF_YOU_CAN [21:39] anyway, what are your suggestions? [21:39] add _EX [21:39] AV_PKT_DATA_STRINGS_METADATA maybe? [21:40] so libav introducen jet another regression into ffmpeg [21:41] theyve done nothing yet [21:41] ubitux is beign preemptive [21:42] Daemon404: talking about merges - recent one being riff info thing [21:43] Daemon404: yep, i'm just cautious on this one, especially after elenril told me today :D [21:43] +what [21:45] what's the problem with anonymous typedefs? [21:45] *anonimous typdedeffed structs? [21:45] personal preference [21:46] it seemed to have caused "issues" with some public ones in header [21:46] otherwise it just seems like a new diego-ing thing as durandal_1707 says [21:46] anonymous typedeffed structs considered harmful? [21:47] what about removing gotos then? [21:47] [20:25:33] And what are the differences, not counting need to modify my scripts? [21:47] [20:25:48] nothing really, they operate the same [21:47] ? [21:47] funny project for a boring winter afternoon [21:47] why tell something like this and cofuse people even more [21:47] confuse* [21:48] burek: ask relaxed in -user. [21:48] burek: that is libav agent [21:48] yes but still [21:48] it's inappropriate to answer like that [21:48] it's all about perspective. [21:49] (not that i'm agreeing with that perspective). [21:49] well not quite [21:49] the reaction that has caused was [20:34:12] I'll have to change every reference of ffmpeg in the scripts. Easy but annoying. [21:49] and tomorrow he'll come again asking for help saying some option is not working in avconv [21:50] ffmpeg.git 03Stefano Sabatini 071f7962625cf3: examples/muxing: remove video_outbuf unused and useless code * 03http://tinyurl.com/d3598js03 [21:50] ffmpeg.git 03Stefano Sabatini 075ca298df2df8: examples/muxing: remove misleading comment about pending API change * 03http://tinyurl.com/bnnn2kf03 [21:50] ffmpeg.git 03Stefano Sabatini 07d6196d942191: examples/muxing: fix bogus setting of st->id * 03http://tinyurl.com/cq9tmsn03 [21:50] ffmpeg.git 03Stefano Sabatini 07eda0a52bf161: examples/muxing: check on frame * 03http://tinyurl.com/caxacx903 [21:50] ffmpeg.git 03Stefano Sabatini 07eebde404bc8c: examples/muxing: merge add_audio_stream() and add_video_stream() * 03http://tinyurl.com/d2o4dhc03 [21:50] if it was the same then there would be no fork at all [21:50] . [21:50] burek: yes, it's the same old bullshit. [21:53] saste: i need something faster than scene detection for my tests; i'm going to add some metadata in one of the black detection filter, which one is appropriate? [21:53] (we have two of them iirc) [21:54] blackframe & blackdetect [21:54] ubitux, blackdetect is better [21:54] blackframe is GPL, and has an inflexible syntax [21:54] ok, thanks [21:54] we should try to merge the remaining features to blackdetect and drop it [21:56] Action: ubitux is stupidely laughing at the racist debug joke he added in his local tree in the blackdetect filter [21:56] yes that filter is racist, that is colorist [21:57] why a pink frame is better than a black one? [21:57] 4chan has a game with the same name as that filter [21:57] haha [21:57] colordetect would be a better filter (and less colorist) [21:58] you should just call it what its for [21:58] e.g. commercialdetect [21:58] why such color discrimination? [21:58] but a colordetect would be slower [21:58] Compn, thats a pretty shitty commercial detect [21:58] you need a logo mask too [21:58] saste: i think it's overkill [21:59] and perhaps check channel changes and bitrate reductio [21:59] n [21:59] saste: with vf compare it won't be needed anymore [21:59] assuming we fix the code [21:59] (aka making it work with a still image) [21:59] so we can detect one frame in a stream [21:59] black, white, pink, lena, etc [22:00] i assume this has a threshold? [22:00] ask luca [22:00] right [22:00] it soudns really trivial [22:00] i wonder how our psnr would compare to that [22:00] do we have a psnr filter? [22:00] no it was discussed on ML [22:01] there were some optimization issues to be fixed [22:01] it was never committed [22:01] Action: ubitux is waiting for the filter to go in, fix the braindead N alloc & free per frame, make it work with still images, and rule the world [22:02] why dont you just tell luca [22:02] to fix it [22:02] i did [22:02] he will likely won't [22:02] grammar-sploded my head [22:02] :D [22:03] he will likely not do it [22:10] meh the metadata don't reach the end of the graph [22:10] :D [22:13] return -1 never seem to end... [22:14] -1 is very descriptive though [22:14] it tells me not to touch this part of the code [22:14] lest i endure pain [22:15] saste: how is the buffer ref transmitted to the next filter in blackdetect? [22:15] i mean, i add some meta to inlink->cur_buf, but that buffer is ignored afterward [22:15] grep -R "return -1" (find -name '*.c') | wc -l => 3412 [22:16] so any idea how we can trick this? [22:16] ubitux: can't say with no code [22:16] my memory is overrated [22:17] Daemon404, yeah permission denied to touch this code [22:17] i think i missed something :p [22:19] it seems to only work when i inject the metadata in the picref passed to start_frame() [22:31] just curious [22:31] do you guys feel like this metadata string in side data is a hack? [22:33] sort of [22:34] how so? [22:34] (soon: batch normalization possible with a few lines of code!) [22:35] the whole idea - it is not going to be used by any demuxer [22:35] nicolas mentioned some [22:35] ? [22:36] http://ffmpeg.org/pipermail/ffmpeg-devel/2012-October/132263.html [22:36] * The v4l2 device uses the priv field to manage its DMA buffers, it could need to inject metadata if some device were able to report additional information, such as autofocus info or GPS position. [22:37] also, i wonder if i couldn't use it for the timecode :) [22:37] (when at format level, like most of the cases) [22:37] what about waiting evils action? [22:38] do you think those suggestions are not reasonable? [22:46] i dunno what to think of it [23:30] ffmpeg.git 03Michael Niedermayer 0735daf3ca8173: cmdutils: remove unneeded null check * 03http://tinyurl.com/chybjur03 [23:30] ffmpeg.git 03Alexis Ballier 07916352f28285: configure: do not quote arguments passed to filter{,_out} in check_ld. * 03http://tinyurl.com/d3zyabt03 [23:39] nyuhu: seems like my lavfi fate howto will be soon "deprecated" [23:40] (the lavfi test listing at configure time is going away soon) [23:41] if you find a better solution, let me know :) [23:43] sounds legit [23:43] fine with me burek :) [00:00] --- Thu Oct 18 2012 From burek021 at gmail.com Fri Oct 19 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Fri, 19 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121018 Message-ID: <20121019000501.E41D718A01ED@apolo.teamnet.rs> [00:43] I'm trying to do point to point streaming using rtsp but not working yet. The command im using and its output: http://pastebin.com/2TUyMRq6 Any suggestion/example that can iluminate me? Thanks! [00:46] arakn0, why not just use udp output directly if you need p2p [00:47] also, your ffmpeg might be old [00:47] try 1.0 [00:49] burek: it's one of my requirements. In my scenario, I have an embeded device that acts a rtsp client. [00:50] req for what? [00:50] oh rtsp [00:50] ok [00:50] can you update your ffmpeg [00:51] sure [00:51] I built ffmpeg on the 3rd of nov [00:51] but i wonder if I'm in the right track [00:51] it doesn't matter when did you build it, but what source code did you use [00:55] ffmpeg version N-45010-g5e6439a [00:55] arakn0: you said you built ffmpeg but your pastes look like the repo version [00:55] that one should be ok [01:01] burek: weird. I just pulled master and rebuilding. [01:03] just in case, but I'm not sure that my ffmpeg version is the problem [01:04] more concern about the options [01:04] we'll see :) [01:04] that I'm using [01:05] burek: just fyi, I've tried rtp over udp and tcp and both work with no issues [01:19] hey, so I am having some issues trying to do some processing of multicast udp stream (ffprobe gives me this: mpeg2video (Main), yuv420p, 1920x1080 [PAR 1:1 DAR 16:9], 65000 kb/s, 30.22 fps, 29.97 tbr, 90k tbn, 59.94 tbc), but I am getting lots of artifacts and corruption, when I use vlc however the stream looks great [01:33] burek: thanks for the tip, here is what I am trying to do: http://pastebin.com/tn1sH7Dj [01:33] burek: same problem [01:33] addisonj, you are not using ffmpeg [01:34] join #ubuntu and ask them to explain to you what are you using actually [01:34] or read this [01:36] burek: oh sorry, just did this on a new machine, forgot about the fork, will switch and try again [01:36] addisonj, try using ffmpeg's git [01:41] the command and its output to start an rstp server: http://pastebin.com/zKuF8Kzq [01:43] you are still using old ffmpeg :) FFmpeg version 0.6-4:0.6-2ubuntu6.3 [01:43] latest is like 1.0 [02:07] burek: i don't know what happened. [02:07] here it's the right version: http://pastebin.com/b6LbnGkk [02:08] [tcp @ 0x2664c40] TCP connection to 0.0.0.0:554 failed: Connection refused [02:09] I've seen that. But I don't wan to connect to that IP. I want to bind to that IP and open that port [02:09] ffmpeg is not an rtsp server [02:09] So, I can stream point to point with rtp...but not rtsp... [02:10] you can feed the rtsp server [02:10] but if you are going to use p2p [02:10] what's the point of rtsp then [02:10] why not just send udp/rtp directly [02:11] the requirements [02:11] I know, I know... [02:11] well ok [02:11] then don't use ffmpeg [02:11] use something else, vlc for example [02:12] use something that can act as an rtsp server [02:12] i think ffserver also can [02:12] let me check [02:12] yes, I was going to mention that [02:12] ffserver [02:12] http://ffmpeg.org/sample.html [02:12] ctrl+f rtsp [02:14] yes, I see that. I was going in that direction too. It wasn't 100% if I could stream using just ffmpeg [02:14] http://ffmpeg.org/trac/ffmpeg/wiki/Streaming%20media%20with%20ffserver [02:14] that might help too [02:15] the only problem is I'm not sure how to specify -vcodec copy inside ffserver's config file [02:16] but since you want to stream a file [02:16] there is no need for that and you can safely ignore me :) [02:16] just see the sample and rtsp examples there :) [02:23] burek: thanks so much. I'm going to study it and I'll let you know! [02:25] :beer: :) [02:36] burek: our issues disappear with ffmpeg, avconv-- [02:36] thanks for the push in the right direction :) [02:36] +1 :) [03:53] Hi all, for an unknown audio source is there a way to transcode it to an uncompressed output without possible loss in bit depth? For instance with video I can use -c:v rawvideo which will produce uncompressed video in native chroma subsampling and bit depth. There doesn't seem to be a rawaudio equivalent. [05:03] hello i have a question, i need to input 3 videos crop and position into one ouput.. can this be done? [05:03] like a tile effect [05:04] jivetalkingturke: sup cracka. [05:04] i am capturing 3 webcams at 720p and would like to join them into one video output [05:05] it is possible. let me find an example. [05:05] ok thanks [05:06] the side-by-side example in the overlay filter documentation is close https://ffmpeg.org/ffmpeg.html#overlay-1 [05:06] ok.. thanks i'll have a look [05:06] overlay is the word i needed.... [05:08] thank you.. i'll have a look at overlaying and see what i can get... [05:09] that should be a good starting point. my example is ancient anyway [05:10] this binder has too many examples [05:10] no worries thanks lllogan... [05:10] and thank you dericed [05:11] ffmpeg -f lavfi -i testsrc -f lavfi -i testsrc -f lavfi -i testsrc -filter_complex "[0:0]pad=iw*2:ih*2[a];[1:0]negate[b];[2:0]vflip[c];[a][b]overlay=w[x];[x][c]overlay=w:h" -y -t 1 out.avi [05:11] aw, he's leave the room once I finish an example [05:12] Action: llogan gives dericed +10 FFmpeg Bucks anyway [05:12] sweet FFmpeg Bucks!! can i apply to a trac bounty? [05:16] dericed: consider adding your example here: https://ffmpeg.org/trac/ffmpeg/wiki/FilteringGuide [05:17] i see similar requests every once in a while [05:19] llogan: working on it [05:21] llogan: where do i upload thumbnails for wiki? [05:22] dericed: click on "Attach file" button on bottom of page, IIRC. [05:23] thx [05:23] ah...duh. i forgot there already is an example at FancyFilteringGuide [05:23] s/Guide/Examples [05:24] but a three or 4 input version would still be interesting [05:24] dericed: sorry about that [05:25] and that example still only have one input. [05:31] llogan: my image doesn't work but it is added to here: https://ffmpeg.org/trac/ffmpeg/wiki/FilteringGuide [05:32] dericed: thanks [05:39] dericed: i fixed the image. apparently you have to attach it to the specific page you are editing. i didn't know that until now, and if i did i forgot because my memory is => lesser ape. [12:00] hi there.. i want to create a timelapse of some web cam images [12:01] the images are sorted like: cam////img--.jpg [12:01] and i want to pick one image each day [12:02] any smart way ffmpeg can do that? [12:02] write a shell script to rename them to 1.jpg 2.jpg etc. [12:02] that's about it [12:03] *shrug* [12:03] ffmpeg can't take its input from a text file? [12:03] so i can just put the absolute paths in there? [12:04] ffmpeg can't take explicit names in image sequences, AFAIK [12:04] maybe look to mplayer, then [12:05] no, just write a shell script to rename them [12:05] there's LOTS of images [12:05] or, copy [12:05] so? just throw disk and time at the problem. that or use an NLE like kdenlive [12:05] and i don't want to rename any images, that's irreversible, and i don't want to copy them either, because that's a huge overhead [12:05] i use mplayer instead [12:05] ln -s then [12:06] ok, that's more interesting [12:06] ln is a more interesting approach, yeah [12:06] but the find job takes forever..... :p [12:07] 1440 images a day for 6 months :) [12:08] somethin glike I=0; for f in `find .|grep jpg` ; do ln -s $f $I.jpg && I=`expr $I + 1` ; done [12:08] then ffmpeg -i %d.jpg yadda yadda [12:08] yes :) [12:09] but they need to be sorted [12:09] I hacked support for pipe separated image sequence names at work, but it's somewhat hackish [12:09] just pipe the find output to sort then [12:10] lunch [12:18] hi, i'm fiddling around with avconv and ffserver.. [12:19] running the server works fine.. just avconv doesn't like to produce the feed: http://pastesite.com/82142 [12:20] i also tried just plain avconv -f video4linux2 -i /dev/video0 http://localhost:8090/feed1.ffm [12:20] same error [12:20] i wonder what avconv tries to tell me.. what codecs did it try? what paremeters did it try? [12:20] which ones failed? :D [12:21] works fine in vlc btw.. [just to note that the hardware does work] [12:22] and in mplayer.. [12:26] hmm, i guess it's a broken package in ubuntu.. [12:44] i tried building my time lapse now, but i'm only getting 9 seconds worth of video [12:45] my files are named 0001.jpg to 0584.jpg [12:45] and i'm running: ffmpeg -i %04d.jpg -r 30 -s 640x480 -vcodec libx264 -b 2000k cam-01.mp4 [12:46] Input #0, image2, from '%04d.jpg': [12:46] Duration: 00:00:23.36, start: 0.000000, bitrate: N/A [12:46] Stream #0.0: Video: mjpeg, yuvj420p, 2560x1920 [PAR 1:1 DAR 4:3], 25 fps, 25 tbr, 25 tbn, 25 tbc [12:46] why does it say duration 00:00:23.36? how can it know that? [12:48] from fps and number of images [12:48] but why is only a 9 second long video created? [12:49] ah, error in input [12:49] three broken images [12:49] let's try that again [12:50] argh.. the images have to be in order [12:51] and there's three jpegs that are broken [12:51] ffmpeg has no option to "skip" broken input? [12:56] just removed those images.. size 0 anyway [12:56] that was the problem [13:51] hello... was wondering if someone could help me with my terminal command for ffmpeg.. [13:52] i am trying to run multiple jobs at the same time [13:53] i don't want to use a bash script.. would rather do it all from terminal [13:53] i am using 3 webcams 1 audio interface and 1 microphone [13:54] would like to capture them at the same time [14:23] anyone know about multiple simultaneous input? [15:07] hey guys, I'm updating my project to use the latest version of ffmpeg - it's working for single channel streams but anything stereo is coming out the other side very distorted - the playback rate is fine but the sound is like it's dropped 2 octaves - example: http://hosting.ispyconnect.com/example.mp3 [15:09] t4nk124: you have not support for planar sample formats [15:09] so only mono will work [15:09] almost every decoder is switching to that [15:10] some, like flac and alac supports both [15:11] when i try and use AV_SAMPLE_FMT_S16 I get an error: Specified sample_fmt is not supported. [15:11] as i said only some decoders support selecting sample fmt [15:12] ... but it works if it's single channel [15:12] and this used to work with an old version of ffmpeg [15:12] t4nk124: as i said decoders switched to planar sample format [15:13] ah ok so basically what are my options? [15:13] and interleaving is now handled in libswresample [15:13] interleave output manually or use libswresample [15:14] you will need to take care of channel layout - that it match with output device [15:14] and libswresample AFAIK do not support that [15:14] ^custom remapping [15:15] so you are force to do it manually until this get cleaned up either in libswr or in lavc itself [15:15] *forced [15:16] do you know of any examples? I'm way out of my depth :( [15:16] but if you are interested in stereo this should not matter at all [15:16] ideally it would spit out a stereo mp3 [15:17] you meen encoding? [15:18] *mean [15:18] yes [15:19] you just need to deinterleave samples [15:20] ok sounds promising - how do i do that? [15:20] but this is overkill considering that libmp3lame support interleaved samples, so fork just removed that code.... [15:21] actually not, it was never done that way, deinterleaving was done in encoder istself :roll [15:22] yeah this was all working fine in the previous version - now i can't encode anything stereo into mp3 using AV_SAMPLE_FMT_S16 [15:23] you need to give encoder each channel in separate plane [15:24] so i should be using AV_SAMPLE_FMT_S16P ? [15:24] previously: Chan.1_sample.1|Chan.2_sample.1|Chan.1_sample.2|... [15:26] now: Chan.1_sample.1|Chan.1_sample.2|Chan.1_sample.3 | ... | Chan.1_sample.N | Chan.2_sample.1 | ... | Chan.2_sample.N [15:27] t4nk124: yes, but make sure you give deinterleaved samples [15:30] hmmm... all iv'e got coming into the method is a pointer to a byte array [19:40] hi question, I'm trying to decode my webcam's output to something compatible in multiple oses using a pipe [19:40] what type of codecs/contains would be advisable? [19:40] (it has to become a stream) [19:47] hi everyone! After 3 hours searching on internet, I haven't found out how I can preserve metadata after a conversion. I'm using version 0.7.13 under FreeBSD 9.0. Any suggestions? [20:13] hi [20:14] do somebody know how to set the headphone's left to audio track 1 right to track 2 converted to mono ? [20:14] or maybe usgin gmplayer? [20:45] creep: ffmpeg sure can do it using -filter_complex [20:45] but I don't know exactly how [20:45] look for "amerge" in the manpage, it does something like that [20:55] Good morning/evening. Can I make ffmpeg produce TIFF images in the RGB colourspace? [20:57] alezakos: several rgb colourpspaces are supported (except planar)) [21:01] I'm trying to stream an mp4 file over RTSP using just ffserver, but for some reason ffserver crashes. http://pastebin.com/pjGDzRH8 Any help/hint ??? [21:02] Good. As my experience on this field is zero, could you please provide me with help on how to split a video to tiffs readable by imagemagick? [21:05] alezakos: ffmpeg -i input out%06d.tiff [21:09] Unfortunately, that command produces tiffs in the YCbCr colourspace, and imagemagick doesn't support it (the colours are weird) [21:10] arakn0: not sure if this is still up to date, but the documentation says streaming from files is broken: http://ffmpeg.org/ffserver.html#What-can-this-do_003f [21:11] alternatively you could try using ffmpeg as a feed [21:11] i think? that kinda worked for me [21:12] alezakos: your input is than not in rgb colorspace which means you need to specify it via -pix_fmt rgba [21:14] Hi folks! I've ffmpeg compiled with 'libaacplus' on FreeBSD, and it doesn't keep into output file the metadata of input file. I've tried all -map options but nothing works. Any ideas? [21:14] durandal_1707: That works perfectly! Thank you very much! [21:14] klaxa: damn.... but it works for mpg :( [21:15] Endorgh: what container? [21:15] arakn0: like i said, no idea if that's still up to date, i have given up on setting up ffserver myself, because it lacks some functionality i want (streaming matroska with subtitles) [21:17] durandal_1707: flac to aac [21:20] durandal_1707: exactly this command: ffmpeg -i file_in.flac -ab 60k file_out.aac [21:21] durandal_1707: if I type ffmpeg -i file_in.flac -ab 60k file_out.mp4 the file_out.mp4 has the medatada, but its length is only 2 minutes, not the same length of input file... Is very strange [21:21] uhh, raw AAC doesn't have anything really to put metadata into [21:21] a container like "MP4" is needed to accomodate that [21:24] JEEB: hence, is needed to specify ".mp4" extension [21:24] ? [21:24] Endorgh: flac to mp4 with libaacplus works fine here [21:25] Endorgh, or anything else that has possible spots for the tags you need, yes [21:25] durandal_1707: do you recommend libaacplus in place of FAAC? [21:25] ok I understand... [21:25] no, you should use fdk [21:25] instead of both faac and libaacplus [21:25] handles both HE-AAC and LC-AAC [21:25] https://github.com/mstorsjo/fdk-aac [21:26] ohh, great info! [21:26] you can't give out binaries with it, just like faac and libaacplus, but it's the least bad aac encoder atm :P [21:27] JEEB: you did blind test? [21:28] not me, but in general that has been the concensus. It's the fraunhofer encoder and it definitely is better than faac and vo-aacenc with LC [21:28] JEEB: hehe, the least bad aac encoder. Then, what is the best one? [21:28] durandal_1707: really need no blidn test to compare with faac [21:29] given I DO hear difference between 128kbit/s faac and original [21:29] faac really IS that bad [21:29] and it doesn't sound bad with HE-AAC so it's at least on par with libaacplus [21:29] yeah, faac is pretty bad [21:29] and vo-aacenc is at times even worse than the ffaac [21:29] blind test may be needed for 192kbit/s faac :P [21:29] (internal aac encoder) [21:29] JEEB: sorry for the late replay, but about the libaac_fdk delay issue [21:29] and the ffmpeg builtin "aac" encoder is buggy [21:29] it does clipping ;) [21:29] or a noise similar to that [21:29] well, every encoder does that [21:29] I was not able to get it to work [21:29] oh [21:30] I thought it is a samplerate issue [21:30] I mean, apparently the "aac" encoder in ffmpeg has some overflow issue somewhere [21:30] it really is only good if "there is nothing else available" [21:30] but it does not seem so [21:30] and if you really NEED aac [21:30] divVerent, more like the ffaac encoder was left to be after some development :/ [21:30] the aac codec needs -strict -2 for a reason :P [21:30] at least it's better than the WMA encoder [21:30] lol [21:31] jeeb:didn't test THAT one [21:31] Endorgh, not sure -- Fraunhofer or Apple's (Dolby's?) [21:31] but I actually did use the aac encoder for a while [21:31] and so did I use faac [21:31] for libfdk_aac I still have to look for the artifacts ;) [21:31] well, the internal aac encoder is the best that you can still distro [21:31] even though I changed bitrate from 128 to 96 when changing from faac to libfdk_aac [21:31] vo_aacenc according to a blind test is not really better [21:32] jeeb: that I don't believe [21:32] unless vo_aacenc has similar issues :P [21:32] but well [21:32] the average case of the "aac" codec is fine [21:32] it's just the reference code >_> [21:32] it just has a bad worst case [21:32] if you want, I can switch to using it again [21:32] (and yes, I know that fraunhofer and friends base on the reference code, too) [21:32] and wait till I get a good sample to demonstrate the "clipping" issue [21:32] but at least fraunhofer generally has been optimized [21:32] when I did encounter one in the past, I verified the weird noise goes away when reducing volume before encoding [21:33] JEEB: Do you know how to prevent the .5 sec delay in libfdk_aac ? [21:33] cbsrobot, if it's the priming samples you need to signal them in the container and make sure the thing that's playing it back actually understands the way of setting priming samples [21:34] I think ffmpeg should support that (maybe even by default?), and L-SMASH's "MP4" muxer does it too [21:34] aac in mp4, played back with quicktime [21:34] basically you should check the first PTS of the first audio frame/packet, you should have the amount of priming needed there [21:34] (there's a negative pts IIRC) [21:35] how do I check that ? [21:35] hexedit ? [21:35] no idea, people usually use the API to get that figure :P [21:35] or ffprobe ? [21:36] also check if the audio length is the same in ffmpeg before and after [21:36] when you do f.ex. ffmpeg -i derp.mp4 and have it show stuff [21:36] (ffprobe probably shows too) [21:37] Ok thanks [21:38] I'll try and ping you later (maybe) [21:38] basically if there's a length difference then priming samples aren't for whatever reason pushed there :P [21:38] (output should be somewhat longer) [21:39] if that is the case, you'd have to get the amount of priming needed and re-mux it with L-SMASH's muxer or something from raw AAC and setting the amount of priming samples [21:41] basicly it a mov with prores and pcm sound convertet to h264 and aac in mp4 [21:48] ffmpeg -i input_file.flac -ab 60k output_file.mp4, generates an output file with less time length than the original input file. Why?? O_O [21:49] it results incomprehensible for me... [21:53] Endorgh: i said i can not reproduce it, you are using buggy version [21:55] nah i set software mixer on gmplayer and opened up same video 2 times, synced them, and set one language with balance left on one, and the other language on the other gmplayer on the right, i can watch CSI in both language now [21:56] durandal_1707: ok, understood! [21:58] Endorgh: what version you are using? [22:00] ffmpeg-0.7.13_6,1 [22:01] uhh, does that even have libaacplus >_> [22:01] also, that is /quite/ old [22:02] its the default version on FreeBSD port collection [22:02] you should build it yourself then, current git if possible [22:03] you are not going to get much support or help with as old version as that. [22:03] as sad as it is [22:03] Endorgh: there is ffmpeg-011 and ffmpeg-devel [22:03] and ffmpeg1 (1.0) [22:06] durandal_1707: mmm d'oh! [22:37] is there a "just work" setting for ffmpeg? [22:38] as in eat as many stream encodings as possible [22:38] without throwing errors [23:08] question, what are the required arguments for webm? [23:15] required? use extension .webm (?) [23:15] And then it automatically chooses vp8 and vorbis [23:15] hm don't know what parameters to give with webm [23:16] Encoder (codec none) not found for output stream #0:0 [23:16] and [23:16] Could not write header for output file #0 (incorrect codec parameters ?) [23:18] ffmpeg -f video4linux2 -i /dev/video0 -vcodec mpeg4 -f webm -y /dev/stdout | nc -lp 5555 [23:21] Stream #0:2(eng): Subtitle: subrip [23:21] is this the same as SRT fromat? [23:25] yes, srt is subrip [23:25] iive: how can I determine what arguments are required for a specific container/codec ? [23:26] no idea. sorry. [23:26] btw, if you try to stream it through a pipe, nut may be the container. [23:27] ok mplayer hung again, shit "MPlayer interrupted by signal 2 in module: unknown" would someone tell mplayer devs that module unknown has bugs? [23:27] hm ok thx [23:27] I seriously suck at ffmpeg-.- [23:27] flv may be a little slow, there is v3 in development, no idea how to select it. [23:27] creep: LoL! [23:28] what was signal2? bus error or ctrl+c? [23:29] iive: looks like asf is my best bet so far:) [23:29] thanks [23:30] iive<< controlc did the job now [23:31] there are really lame errors like double pressing play, or pressing controls while playlist is open resulting infinite loops, or setting window to always on top and hitting some buttons on controls [23:56] creep: interesting, i assume these are gmplayer related. could you try to fill some of them in the bugzilla, in case they are not already there. [00:00] --- Fri Oct 19 2012 From burek021 at gmail.com Fri Oct 19 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Fri, 19 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121018 Message-ID: <20121019000502.E881918A01F1@apolo.teamnet.rs> [00:27] does anyone have wav files with INFO chunks aka tags? [00:29] foobar2k can write such files, question are they spec compliant [00:29] durandal_1707: doesn't audacity allow you to add those too? [00:31] i do not have it installed [00:33] oh, I forgot you use FreeBSD [00:35] there is package but i do not want to download and install all that stuff just to test possible non-compliant software [00:35] and reading forums WMP can not read audactity wav tags [00:36] so nobody cares about compatibility [00:37] i was just being a douche [00:37] as the kids say around here [00:40] there is even 'id3 ' chunk in wav [00:41] i didn't really look into what audacity does, but here is a file: http://greasedscotsman.com/tone.wav [00:44] apparently it ignored album title, track number, and genre although those fields are available in audacity's "metadata editor" [00:46] looks like audacity picks safe side, it puts even chunks size [00:51] well if invalid INFO chunks are detected whole LIST should be just ignored, instead of failure to decode anything [01:14] ffmpeg.git 03Michael Niedermayer 07ce739e66f4f9: cmdutils: add missing check for ftell() return * 03http://tinyurl.com/clttcht03 [01:14] ffmpeg.git 03Michael Niedermayer 07eb19d89d8eb5: cache: check lseek() return * 03http://tinyurl.com/cctqgvn03 [01:17] huh, err_detect flags does not work in lavf [01:20] hah -f_err_detect does the job but it is deprecated ..... [01:21] michaelni: why is avconv mentioned in ffmpeg? [01:25] to explain that ffmpeg supports everything avconv does ? [01:26] no, err_detect (same for lavf and lavc) does not work for ffmpeg in lavf context [01:26] so there is f_err_detect [01:27] if option is named same in lavf and lavc, there is no way to set it to be for lavf(or both) [01:38] durandal_1707, hum, it should work [01:38] i mean same option name in both [01:40] if i have only one option it does not work for lavf (obviously) [03:02] did anyone here compile static ffmpeg for raspberry pi [03:02] cross-compile [03:02] hahah seems someone else got his pi today :D [03:02] :) [03:03] very cool thingy :) [03:03] i need to think of how i will stream video from my NAS to the PI :) [03:08] i was following this guide http://lumux.co.uk/2012/07/03/cross-compiling-ffmpeg-for-the-raspberry-pi/ [03:08] but all the tools from that git are 32bit [03:08] no 64bit compilers there [03:08] something is wrong i guess :) [03:14] I can't believe that many people in deed build ffmpeg on the raspberry itself.. :S [03:14] how painful is that :S [06:18] heh :) it works :) [06:18] let me see now how strong is raspberry pi :) [08:07] not very [11:05] ffmpeg.git 03Justin Ruggles 07abd8b9e7e05b: libmp3lame: resize the output buffer if needed * 03http://tinyurl.com/c7qnnvr03 [11:05] ffmpeg.git 03Mans Rullgard 0721fed588cba7: fate: add macros useful for conditionally enabling things * 03http://tinyurl.com/dycrfkt03 [11:05] ffmpeg.git 03Mans Rullgard 07b262e4559710: fate: dependencies for vsynth tests * 03http://tinyurl.com/cpqtf4603 [11:05] ffmpeg.git 03Mans Rullgard 07292d1e787438: fate: dependencies for acodec tests * 03http://tinyurl.com/cfzmql303 [11:05] ffmpeg.git 03Michael Niedermayer 0717e4b0644b5a: Merge commit '292d1e78743855404c7d07e3e7cb3f9c9ae6275b' * 03http://tinyurl.com/c3zzwhh03 [11:05] ffmpeg.git 03Michael Niedermayer 07aadaa8112a19: fate: dependencies for ffmpeg acodec tests * 03http://tinyurl.com/c4dsmmg03 [11:05] ffmpeg.git 03Michael Niedermayer 075806cd93651f: fate: dependencies for ffmpeg vsynth tests * 03http://tinyurl.com/dx4nmnw03 [11:46] ffmpeg.git 03Mans Rullgard 07fa26335003c7: fate: handle lavf test dependencies entirely in make * 03http://tinyurl.com/d2kyb6w03 [11:46] ffmpeg.git 03Mans Rullgard 0736ac9a16a19a: fate: dependencies for seek tests * 03http://tinyurl.com/ccxvz2a03 [11:46] ffmpeg.git 03Michael Niedermayer 079317b73f11c3: Merge commit '36ac9a16a19a365ce58cc871484c20cffe9b6401' * 03http://tinyurl.com/chbxr8a03 [12:18] wtf, i did not write single bit of ffv1.3 code [12:19] ffmpeg.git 03Mans Rullgard 07206a070dce5a: fate: list lavfi tests in a makefile * 03http://tinyurl.com/d85qroe03 [12:19] ffmpeg.git 03Mans Rullgard 07b93e934aeea4: mips64: mark hi/lo registers clobbered in MAC64/MLS64 macros * 03http://tinyurl.com/c8ov89p03 [12:19] ffmpeg.git 03Michael Niedermayer 073777e6b3bf65: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/d4tv3do03 [12:49] durandal_1707: well it is written on the internet, must be true [12:50] but i did not write 1.3 at all i just cleaned up some code and added more trivial code like extra colorspace stuff for ffv1 [12:57] durandal_1707: you can not argue with the internetz [16:13] ffmpeg.git 03Michael Niedermayer 071350dffdc6ad: riff: dont discard truncated metadata * 03http://tinyurl.com/99x9usw03 [16:13] ffmpeg.git 03Michael Niedermayer 07d0c27e88d2bb: riff: retry reading metadata without padding if it fails with * 03http://tinyurl.com/9d9pekr03 [20:37] ffmpeg.git 03Michael Niedermayer 072b1a2466c7f0: dv: change assert(a2 < 4) to av_assert() * 03http://tinyurl.com/9ml5hdu03 [20:45] ffmpeg.git 03Michael Niedermayer 0723b203014f5d: indeo4: prevent printing uninitialized variable * 03http://tinyurl.com/8v8v54u03 [21:13] ffmpeg.git 03Michael Niedermayer 0797d1cb5cd48c: bmv: remove unreachable default case * 03http://tinyurl.com/8lt5a7503 [22:11] ffmpeg.git 03Michael Niedermayer 072472f3facbb4: lzwenc: change assert to av_assert * 03http://tinyurl.com/9q6952s03 [22:11] ffmpeg.git 03Michael Niedermayer 075537c92f84db: mpegvideoenc: check return value of ff_MPV_frame_start() * 03http://tinyurl.com/93tku5l03 [22:11] ffmpeg.git 03Michael Niedermayer 07c753b56b4d56: ff_convert_matrix: fix integer overflow * 03http://tinyurl.com/8rfd6gw03 [22:46] divVerent: you there? [22:49] ffmpeg.git 03Cl?ment BSsch 07e807a2b64899: lavc: add raw text subtitles decoder. * 03http://tinyurl.com/8w34gf703 [22:50] divVerent: anyway, see above ^ [23:05] saste: ping [23:05] durandal_1707, pong [23:06] saste: you ok with idcinvideo explanation about fate ref chage? [23:06] *change [23:06] durandal_1707, yes, please add the explanation to the commit log [23:06] the other remarks are nits that you can safely ignore [23:07] i'm writing yet another demuxer [23:09] durandal_1707, BTW what happened of the XPM encoder/decoder? [23:09] no motivation [23:10] :( [23:10] well maybe i'm going to "adopt" it if you don't plan to work on it [23:11] saste: i have hnm4 (intra only working) if you want to work on that too [23:11] so i can use the original XPM glyphs of the effectv matrix filter [23:12] ham4? [23:12] hnm4 - game codec [23:13] ok i was confusing with IFF HAM images [00:00] --- Fri Oct 19 2012 From burek021 at gmail.com Sat Oct 20 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Sat, 20 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121019 Message-ID: <20121020000501.876C218A01DE@apolo.teamnet.rs> [00:47] hi [00:48] anyone in here [00:51] we're all waiting for a valid question. [01:21] is it possible to use a bit stream filter in ffserver? any hint? Thanks [01:56] both vlc and mplayer should be useful, but the coders surely suck in user interaction [02:41] hello! anyone know how i can convert a byte array of 2 channel 16 bit audio into a planar format that i can encode with? [02:55] t4nk157: that should be pretty trivial and there are and was numerous examples in FFmpeg source code and on the web [02:56] yeah i've spent 4 days on it so far - any ideas where to look? I've spent hours on google looking for it [02:56] have you wrote some code at all? [02:57] yes massive amounts, it's working with single channel, as soon as i add another channel it gets distorted - i think i was chatting to you yesterday about it [02:57] t4nk157: i mean code that does interleaving [02:59] i'm not sure where to start - the code i had worked with the 53 version. It doesn't work with the new version. I get a byte array of AV_SAMPLE_FMT_S16 which I can't encode with unless I use AV_SAMPLE_FMT_S16P - but I have no idea how to get that byte array into a planar format [03:01] for planar and 2 channels it is 2 planes or 2 byte arrays [03:01] how do you call encode_audio2 ? [03:02] avcodec_encode_audio2(c, &pkt, frame, &got_packet); [03:02] and where you put your pointer to array? [03:03] i get a load of byte data at a time that i have to split into frames [03:04] you put it into AVFrame->data[] ? [03:04] ret = libffmpeg::avcodec_fill_audio_frame(frame, c->channels, c->sample_fmt, pSoundBuffer, size, 1); [03:07] in that case your pSoundBufer need to be interleaved [03:07] *deinterleaved [03:07] so lets assume it have 100 samples in each frame - each sample takes 2 bytes [03:08] yeah that's what i'm after - how to do that [03:09] size & 1 must be 0 for stereo [03:09] so you divide size by 2 (because there are 2 channels in stereo) [03:10] there are many ways to do actual deinterleaving, one of it is using libswresample [03:10] do you know of any example code that does that? [03:11] ffplay.c use it [03:12] and some filters libavfilter [03:12] do you use any of them? [03:13] i've been through an enormous amount of source code trying to find an example of deinterleaving without result :( [03:13] i'm mentioning that lib because it have simd for conversion [03:15] naive approach: you take first 2 bytes and put them in X buffer. then take next 2 bytes and put them into Y buffer, you repeat this until you run out of samples [03:15] yeah i haven't gotten anywhere, my old code does deinterleaving but i can't see how it relates to the new api - everything it's calling has either been deprecated or removed [03:15] last step you concatenate X and Y buffer [03:16] ok so each packet has to be left or right channel? [03:16] packet is what encoder gives you [03:17] aka output [03:17] yeah but i need to feed into the encoder arrays of left and then right channel? [03:17] the pSoundBUffer you mentioned in avcodec_fill_audio_frame is X+Y [03:19] someone must have done this before right? [03:20] what? [03:21] written some code that deinterleaves data for avcodec_encode_audio2 [03:22] https://www.google.com/search?q=deinterleave+avcodec_encode_audio2 [03:22] no, you deinterleave pSoundBuffer you use in fill_audio_frame [03:23] it must be deinterleaved before calling fill_audio_frmae [03:26] yes, therein lies my problem - you said there are loads of examples of doing this - could you give me a link? [03:29] I thought all this was automatic - which is why you specify channel_layout and channels etc [03:30] t4nk157: i could be but it is not, you can not specify your sample format and expect fill_array to do the job [03:31] that is job for swr_convert [03:31] ok is there an example of using swr_convert? [03:31] but this could be implemented .... [03:31] t4nk157: only if you have libswresample actually instealled. do you? [03:31] yes [03:35] t4nk157: http://ffmpeg.org/doxygen/trunk/ffplay_8c-source.html [03:37] ok thanks for that [07:34] i would like to use jack as an input format with ffmpeg. i'm on archlinux but i cannot seem to get the "-f jack" flag working [07:39] looks like i have to compile it with --enable-indev=jack [11:59] hi all [12:00] I'm trying to get a raspberry pi to stream a webcam to my computer [12:00] but when i go to ip/webcan.mjpeg it starts to download a file [12:10] hi! is fdk_aac be able to convert to AAC-HE? or is libaacplus better for this task? [12:12] yes. no. [12:16] extremely precise and concise :D [12:19] and, How I can be sure that I'm achieving AAC-HE quality typing this command:?? [12:19] ffmpeg -i input_file.flac -acodec libfdk_aac -ab 60k output_file.mp4 [12:24] Yulth: doing blind ABX test with reference that have AAC-HE quality [12:27] libfdk-aacenc supports several profiles: LC, HE-AAC, HE-AACv2, LD and ELD [12:28] I understand, but in the command line I can't specify what type of AAC I want to use for encoding. I only can type "mp4" or "aac". [12:29] there is -profile [12:29] mmm, interesting, I'm going to try it [12:29] ffmpeg -h full [12:39] this command: ffmpeg -i file_in.flac -profile:aac_he_v2 -ab 60k file_out.mp4 [12:39] is giving me this error: [12:39] Unable to find a suitable output format for '60k': invalid argument [12:39] :S [12:39] I think I've followed all steps shown by man page... [12:44] -profile HE-AAC [12:51] mmm the same: [12:51] "Please use -profile:a or -profile:v, -profile is ambiguous" [12:51] and I've followed the same sintax shown by man and even given by you... [12:56] ok solved: [12:56] -profile:a aac_he_v2 [13:13] why doesn't "ffmpeg -formats" list the wmv1 and wmv2? [13:14] Element9: because they are not formats [13:14] durandal_1707: formats == containers? [13:15] something like that [13:16] durandal_1707: can I read somewhere about the difference? [13:17] it is just wording [13:18] in this specific case formats are containers [13:18] can I choose which codecs to include and which to exclude when I compile ffmpeg? [13:19] yes [13:19] durandal_1707: great. thanks! [15:16] Hi. I currently have the following workflow: {SWFs} --swfrender--> {PNGs} --ffmpeg(image2)--> avi. I'm wondering if I could remove the dependency on swfrender and do all the stages with ffmpeg. [15:17] Basically, each SWF file is an image, and I want to encode all the image into a movie. [15:18] can ffmpeg recognize swf files? [15:19] According to google, uncompressed swf could be supported. [15:20] but does it work with your files? [15:21] nope. I get this error message [swf @ 0x8e61a60]Could not find codec parameters (Video: mjpeg) [15:26] have you compiled ffmpeg with zlib support - paste full ffmpeg output [15:27] http://pastebin.com/NBerPDUH [15:28] that is 3 years old [15:30] Running a Debian stable. I need to embedded this process into an application, so user will have "outdated" versions anyway [15:31] cykl: sorry, i cant give you such support [15:32] I will try on with an up to date version in few hours and see if it works or not. [15:32] durandal_1707: thanks for your time. BTW: If somebody already did this kind of workflow I'm interested by a testimonial ;) [16:21] :( [16:45] Hey how can I figure out which codec corresponds to AVCodec with codec id: 142? [16:45] avcodec_get_context_defaults3 [16:46] ? [16:46] No that doesn't make any sense... [16:47] most likely VP8 [16:50] durandal_1707: Thanks [16:51] I just realized XCode switched to clang and now none of my stuff works, what a pain :( [16:53] meekohi: well I guess you are fucked [16:53] meekohi: thanks Apple [16:53] can you even have more than one xcode installed at once? [16:53] lol [16:53] is that permitted? [16:54] Even if it is I'm sure you wouldn't want to. [16:54] hahaha [16:55] avcodec_find_encoder keeps returning null when I look for VP8, despite linking in lvpx, and this all worked in the past. [16:55] So I'm a bit stumped what I'm missing. [16:56] you can have multiple versions of 4.x installed [16:56] there's a cli tool to choose which one is active for command line stuff [16:57] probably need a developer account to actually get anything other than 4.5 though [16:58] Any recommendations for debugging missing codec problems? [18:11] hi. i'm trying to remux a live stream with avconv: |avconv -re -i pipe:0 -acodec copy -f mp4 -movflags empty_moov [18:11] however i get: Undefined constant or missing '(' in 'empty_moov' [18:12] how do I add an empty moov atom? [19:54] #Hello everyone. I'm new to ffmpeg and would like to know if its possible to configure a webcam stream with both audio and video, and no flv? [19:55] what container do you want to use instead of flv? [19:56] I've tried mjpeg, which is almost real time, but it has no audio [19:56] @klaxa : anything but asf or flv [19:57] what is your setup so far? [19:58] @klaxa: pastebin.com/nqqE8bdq following this tutorial http://www.area536.com/projects/streaming-video/ [20:01] you could try simply replacing asf by mp4 i think [20:01] just a guess, i'm no expert [20:02] @klaxa: thanks, i'll try it right now [20:15] @klaxa: do you happen to know a good tutorial for streaming a webcam using mp4? [20:15] i'm lost here [20:15] actually no, i rarely use ffserver [20:15] i've found tutorials to stream mp4 videos only [20:15] it's ok... thanks anyway [20:16] using mp4 didn't work properly? [20:16] did it give any errors? [20:33] amendes365: i tried using the mpegts container, mpeg2video and vorbis as codecs, i get segmentation faults with the ffmpeg command though [21:29] hey guys... is it possible to use a bit stream filter in ffserver? any hint? Thanks [21:34] thank you @klaxa [21:36] did you try it and it worked? [21:42] no, but I appreciate the help [21:42] I will reinstall and start again from scratch [21:56] Hi, can we perform multipass encoding with live stream? It sounds impossible to me, but maybe... [22:11] @klaxa got the webcam working following http://bit.ly/PfJyd5 but getting 'file mystream.ffm not found' in browser [22:13] amendes365: did you copy the config? because that config isn't complete (lacks xml tags) [22:17] @klaxa here it is: pastebin.com/zTJz4KLm [22:19] swf container and flv codec? i thought you didn't want flash? [22:19] and you wanted audio [22:35] @klaxa thats the goal, but I was focusing on a working solution to stream the webcam and then change the format [22:36] try setting the contaner as mpegts, the videocodec as mpeg2video and vorbis as audio codec [22:36] then try again and see if anything crashes [22:36] ok [22:47] hey guys. i've joined 2 videos (X = A+B) with mp4box and when i open X in a player it only shows B and it says that duration of X is the same as for B. but the file size for X is s(X) ~ s(A)+s(B). ffprobe for X says [22:47] Duration: 00:00:29.44, start: -12.027937 [22:47] that 12.027937 secs is A. how can i reset that "start" to be 0 ? [23:03] @klaxa gives me an error 'Unknown AudioCodec ogg' in ffserver [23:03] replace ogg by vorbis [23:03] ok [23:03] ogg is a container [23:05] yes. ffmpeg shows 'Could not find input stream matching output stream #0.0 ioctl(VIDIOC_QBUF)' [23:06] what command did you run? [23:06] http://pastebin.com/Xws1LUzu [23:07] i ran ffmpeg -r 15 -s 640x480 -f video4linux2 -i /dev/video0 http://127.0.0.1:8090/feed1.ffm [23:09] not sure if this is the error, but ffmpeg isn't providing an audio stream [23:09] so there is nothing for the audio to be encoded [23:09] that's weird [23:09] if you run pulse you can use -f alsa -i pulse and change the recoding device with the pulseaudio volume control [23:11] ok. pulseaudio shows 'No application is currently recording audio' [23:12] it says so while running ffmpeg? [23:15] no, i checked it out in the pulseaudio volume control [23:16] and ffserver shows [ffm @ 0x9a14cf0]Could not find codec parameters (Audio: vorbis, 44100 Hz, mono, s16, 32 kb/s) [23:16] Hi folks! I don't understand why the output file (aac format, -profile aac_he_v2) generated has a length of only 2 minutes plus 20 seconds, when the input file has 5minutes length..... I'm converting to aac-he v2 into mp4, using version 1.0 under FreeBSD. Any ideas? :S [23:16] the command is: [23:17] ffmpeg1 -i input_file.flac -profile:a aac_he_v2 -ab 60k -ar 44100 -ac 2 output_file.mp4 [23:17] I'm typing wrong any option? [23:20] Endorgh: is output really that long? [23:24] durandal_1707: interesting observation!! The MP4 output file has the same SIZE as a raw aac output file, but when I play both files, the raw aac file doesn't show any metadata although it appears as correct lenght. Otherwise, the MP4 output shows full metadata, but player only can play first 2 min + 20 secs... [23:25] any incompatibility with mp4 - m4a containers? [23:28] Endorgh: does same happens with mkv container? [23:29] I don't know, let me try it! [23:35] durandal_1707: with mkv or mka containers it seems to work... [23:36] what could be the cause? or..., when I can find more information about containers, etc.... [23:37] Endorgh: you sure it works? and it is not just number set in container header? [23:38] mmm [23:39] do: ffmpeg -i input -f null - [23:39] I don't know if this could give any extra info, but ffmpeg says this when encoding into mp4: [23:39] [flac @ 0x2a4e3110] max_analyze_duration 5000000 reached at 5015510 [23:39] ok [23:39] that is just spam (completly irrelevant) [23:40] ok [23:44] this is the output: [23:44] http://pastebin.com/q2PzAqTL [23:48] Endorgh: that is expected, flac support is not broken, but do same with aac in mkv [23:49] Hey [23:49] i'm transcoding a mpeg2 stream into mpeg4 [23:49] how to output it to RAW UDP? [23:51] durandal_1707: flac support is not broken? what it means? [23:55] Endorgh: that duration of flac file is correct [23:58] durandal_1707: ok, and what is the conclusion? [00:00] --- Sat Oct 20 2012 From burek021 at gmail.com Sat Oct 20 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Sat, 20 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121019 Message-ID: <20121020000502.A294018A01F0@apolo.teamnet.rs> [00:04] michaelni: why i get stream start is in unreadable format? 2286340.981000 [00:32] what to use to convert ascii to integer? [00:34] strtol [00:35] and how to check for failure? [00:36] *endptr==ptr implies no conversion happened [01:48] ffmpeg.git 03Michael Niedermayer 07df727d408c7e: mlp_parser: print error when ff_combine_frame() fails to add the current buffer * 03http://tinyurl.com/8zok49203 [01:48] ffmpeg.git 03Michael Niedermayer 0705b0337025f6: motionpixels/mp_decode_frame_helper: assert that the first pixel doesnt reuse the last. * 03http://tinyurl.com/9etjwdy03 [02:08] ffmpeg.git 03Paul B Mahol 07aadb7b3ac4fb: lavc/c93: use meaningful error codes * 03http://tinyurl.com/9ezxumo03 [02:08] ffmpeg.git 03Paul B Mahol 07527224830aad: idcinvideo: if decoding fails return error * 03http://tinyurl.com/9q3wtnr03 [02:08] ffmpeg.git 03Paul B Mahol 074c6c6a266e9c: jvdec: use more meaningful error code * 03http://tinyurl.com/8twxyjs03 [02:08] ffmpeg.git 03Paul B Mahol 070bf40e0ef421: lavc/yop: remove redudant YOP in av_log() messages * 03http://tinyurl.com/9nz3q8u03 [02:08] ffmpeg.git 03Paul B Mahol 07be536f084a98: xxan: return more meaningful error codes * 03http://tinyurl.com/9p8n6zs03 [02:08] ffmpeg.git 03Paul B Mahol 07445f36d7c83f: kmvc: use meaningful error codes * 03http://tinyurl.com/9yj2mu203 [02:08] ffmpeg.git 03Paul B Mahol 074ebf30595106: lavc/tta: do not overwrite bits_per_coded_sample * 03http://tinyurl.com/9r9toe503 [02:20] some guy on libav-user ml says that libswscale is too slow [02:21] lets rm it and rewrite it hen [02:21] :P [03:03] ffmpeg.git 03Michael Niedermayer 07c0b17ea106b9: roqaudioenc: Fix crash with very small roq files * 03http://tinyurl.com/975s86m03 [03:51] ffmpeg.git 03Michael Niedermayer 07a06f943f9d65: roqvideodec: replace dead code by assert * 03http://tinyurl.com/9kmsbqy03 [03:51] ffmpeg.git 03Michael Niedermayer 07d50aa006fb34: tiffenc: fix integer overflow * 03http://tinyurl.com/9u432en03 [09:32] ubitux: I have seen the commit [09:32] what is with all this tinyurl waste BTW? [09:32] i don't know [09:32] i don't like it :( [09:32] do we really need to allocate a new tinyurl for each commit? [09:32] it just worsens the tinyurl service for everyone else [09:33] why not http://ffmpeg.org/gd50aa00 [09:33] would be a job for mod_rewrite ;) [09:33] even the full url with a reduced hash is fine [09:33] +1 [09:33] well, I see the point of using short urls [09:33] but compare [09:33] http://tinyurl.com/9u432en [09:33] http://ffmpeg.org/gd50aa00 [09:33] equally long [09:33] and the latter doesn't eat up tinyurl namespace [09:34] but only means ffmpeg can no longer have URLs starting with g and having no dot inside ;) [09:34] could even do http://ffmpeg.org/d50aa00 with some careful regexing [09:34] ffmpeg.org/h/ [09:34] or that [09:34] that's one more character than the tinyurl though ;) [09:34] i think we have similar redirections with sources.ffmpeg.org [09:35] source.ffmpeg.org sorry [09:35] o the tinyurl TOS even allow this? [09:35] yes, it is [09:35] well, then that's tinyurl's fault ;) [09:36] curl -I 'source.ffmpeg.org/?test' [...] Location: http://git.videolan.org/?p=ffmpeg.gittest&test [09:36] something looks fishy :D [09:36] hehe [09:36] that'd work too [09:36] http://ffmpeg.org/?hash [09:36] curl -I 'source.ffmpeg.org/?foo=bar' [...] Location: http://git.videolan.org/?p=ffmpeg.gitfoo=bar&foo=bar [09:37] the dup is weird [09:37] ubitux: it MAY still make sense [09:37] there may be a separator that isn't contained [09:37] but I can't find the one :P [09:37] curl -sI 'source.ffmpeg.org/?foo=bar' | cat -e [...] Location: http://git.videolan.org/?p=ffmpeg.gitfoo=bar&foo=bar^M$ [09:38] doesn't look like to [09:38] was just a guess, becuase such duping happens easily in shell scripts splitting strings [09:38] if s is foo/bar, then ${s%%/*} is foo and ${s#*/} is bar [09:38] but if s is just foo, both will be foo [09:39] so MAYBE source.ffmpeg.org uses a matching technique with similar issues [09:39] sure not a shell script though ;) [09:40] source.ffmpeg.org/?;a=commit;h=d50aa006 [09:41] i'm able to trick it with this [09:41] but it's ugly :p [09:41] but it still dupes, I suppose [09:41] just that this then at least works [09:41] yes [09:42] Location: http://git.videolan.org/?p=ffmpeg.git;a=commit;h=d50aa006&;a=commit;h=d50aa006 [09:42] it's a big ugly :D [09:42] source.ffmpeg.org/?;h=d50aa006 [09:42] suffices BTW [09:42] indeed [09:42] because a=commit is default if h is given [09:42] i like a=commitdiff though [09:43] i'd go for source.ffmpeg.org/?;a=commitdiff;h=d50aa006 [09:43] I don't, for a very simple reason [09:43] because the subject line then is less visible :P [09:43] how so? oO [09:44] commitdiff has the commit description [09:44] in addition [09:44] ah no sorry it's also in commit as well [09:50] yes [09:50] it is there in commitdiff [09:50] but in commit, it is right above the commit message [09:51] and in committdiff it's way more above [10:49] hey, it would be nice to have a shout protocol [13:36] ffmpeg.git 03Martin Storsj? 07b760ffdd07ea: aviobuf: Remove a senseless ifdef in avio_seek * 03http://tinyurl.com/9ukqgnp03 [13:36] ffmpeg.git 03Martin Storsj? 0753e8cd68b722: configure: Split out msvc as a separate target OS * 03http://tinyurl.com/8sttjkg03 [13:36] ffmpeg.git 03Martin Storsj? 07fc085c5b33a9: gxf: Add a local copy of the relevant parts of the frame rate table * 03http://tinyurl.com/8rrhfe903 [13:36] ffmpeg.git 03Martin Storsj? 07eaa9b2e66c04: avcodec: Rename avpriv_frame_rate_tab to ff_mpeg12_frame_rate_tab * 03http://tinyurl.com/9g8tx6h03 [13:36] ffmpeg.git 03Martin Storsj? 07d66c52c2b369: Add support for building shared libraries with MSVC * 03http://tinyurl.com/8ehw33803 [13:36] ffmpeg.git 03Mans Rullgard 07c0329748b04e: fate: add a dependency helper macro * 03http://tinyurl.com/8bw2wlb03 [13:36] ffmpeg.git 03Michael Niedermayer 07b0554fec0470: Merge commit 'c0329748b04e1f175dad8c9c2ebf22a5e2dc5b72' * 03http://tinyurl.com/9kxlvbp03 [13:54] guess now i have to figure out what needs to be fixed for shared msvc ffmpeg builds [13:54] ffmpeg.git 03Mans Rullgard 07eccd0671cc89: fate: dependencies for demux tests * 03http://tinyurl.com/9u2k9wf03 [13:54] ffmpeg.git 03Mans Rullgard 07b55dda4a7205: fate-als: add dependencies * 03http://tinyurl.com/9ogn8x503 [13:54] ffmpeg.git 03Mans Rullgard 072e05143a8df9: fate-indeo: add dependencies * 03http://tinyurl.com/9f2fhbg03 [13:54] ffmpeg.git 03Mans Rullgard 0708c6a12a562d: fate-mpc: add dependencies * 03http://tinyurl.com/8pr5yor03 [13:54] ffmpeg.git 03Mans Rullgard 0722f5149ec461: fate-twinvq: add dependencies * 03http://tinyurl.com/97d4waa03 [13:54] ffmpeg.git 03Mans Rullgard 0755351de9f43d: fate: add dependencies for misc microsoft codecs * 03http://tinyurl.com/8hmbtak03 [13:54] ffmpeg.git 03Martin Storsj? 0722310eef9f78: changelog: Mention the MSVC DLL support * 03http://tinyurl.com/9h8hzyf03 [13:54] ffmpeg.git 03Martin Storsj? 070af1fe845a9d: avformat: Fix references to the removed function av_write_header in comments * 03http://tinyurl.com/9btmu3d03 [13:54] ffmpeg.git 03Michael Niedermayer 0793f244e3abc3: Merge commit '0af1fe845a9d7112da0a58d33a4fc81fe7c47e95' * 03http://tinyurl.com/95tlxyk03 [14:10] ffmpeg.git 03Diego Biurrun 077b2121e7e2e1: riff: Move functions around to be covered by appropriate #ifdefs * 03http://tinyurl.com/8jr27k203 [14:10] ffmpeg.git 03Diego Biurrun 07e8fe208be810: fate: dependencies for screen codec tests * 03http://tinyurl.com/9x3kjv303 [14:10] ffmpeg.git 03Mans Rullgard 077a12d97eb1aa: aac: fix build with hardcoded tables * 03http://tinyurl.com/8dcsffq03 [14:10] ffmpeg.git 03Martin Storsj? 070de9380be54e: rtp: Update the check for distinguishing between RTP and RTCP * 03http://tinyurl.com/8ey2ojh03 [14:10] ffmpeg.git 03Martin Storsj? 071c37744963ac: rtsp: Vertically align a constant definition * 03http://tinyurl.com/8bf6l4k03 [14:10] ffmpeg.git 03Martin Storsj? 073f055f8f5f71: rtsp: Allow setting the reordering buffer size via an AVOption * 03http://tinyurl.com/8qtg8s203 [14:10] ffmpeg.git 03Mans Rullgard 071cd432e167b1: configure: fix libcdio check * 03http://tinyurl.com/8jjwqzr03 [14:10] ffmpeg.git 03Michael Niedermayer 0781ff0c24ef05: Merge commit '1cd432e167b1a80853760c89a33606e2b5f229c2' * 03Error03 [14:20] ffmpeg.git 03Kostya Shishkov 078774d5835806: bmv: get a new frame on every decode_frame(), so we can use direct rendering * 03http://tinyurl.com/9hmms5e03 [14:20] ffmpeg.git 03Kostya Shishkov 07169514c440a7: indeo3: do not try to output more lines than we can fit * 03http://tinyurl.com/9vrhule03 [14:20] ffmpeg.git 03Martin Storsj? 0791485e744f8c: fate: Add proper dependencies in lossless-video.mak * 03http://tinyurl.com/9ntr68f03 [14:20] ffmpeg.git 03Martin Storsj? 0761cc99748c11: fate: Add proper dependencies in qt.mak * 03http://tinyurl.com/8htepdb03 [14:20] ffmpeg.git 03Michael Niedermayer 07c5fd9d3c35cb: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/9gkanw703 [15:02] ffmpeg.git 03jamal 079434ead2f34b: fate: Handle lavf-fate tests in a makefile * 03http://tinyurl.com/9vmkneu03 [15:13] Don't we really only need tiny URLs if we plan to tweet about a commit? [15:13] Action: gnafu scuttles. [15:36] Action: cbsrobot advises gnafu to fix his tinyurl before fixing others: http://tinyurl.com/gnafu [15:39] Huh, there's a Gnafu on Twitter. [15:54] troller too? [16:11] ffmpeg.git 03Petter Ericson 077abf394814d8: mov.c: Check for stsd + m1s tag indicating MOV-wrapped MPEG-PS, and force continued probing if found. * 03http://tinyurl.com/9erz4sh03 [16:26] durandal_1707: Huh, I guess I never quite considered myself a troller. I didn't think I was good enough. I do look up to kshishkov and av500, though ;D. [16:27] And I dunno about that Gnafu, as it seems most of their tweets are in French (and I didn't bother trying to translate). [16:27] gnafu: sorry, I was just trolling [16:27] :-D [17:01] Compn: ping [17:07] http://technet.microsoft.com/en-us/security/msvr/msvr12-017 [17:10] they are still secret, so? [17:13] now that there is riff tags it is time add exif tags support [17:22] the hard part with supporting old weird subtitles format is to actually find some samples :P [17:22] how about a tool that can create them? :) [17:23] might do the trick [17:23] i have one sample for subviewer1, one for mpsub (though, i should be able to generate some more with mplayer) [17:23] hi, dumb question about the sony D10 spec, is 8 channels at 24bit allowed ? [17:23] now about vplayer, aqt or pjs.. [17:28] j-b: would vlc be interested in using lavf/lavc for subtitles at some point? [17:30] j-b: would vlc be interested in QDMC ? [17:31] ubitux: subtitles being? [17:32] durandal_1707: I think this is used in RTSP too right? [17:33] j-b: right now we have subrip, realtext, sami, subviewer2, jacosub, webvtt (basic markup for this one though), and a few others; basically the more complex ones [17:33] i plan to add the old ones supported by mplayer pretty soon [17:35] we're going to change the api anyway, but i was wondering if at some point the vlc project would be interested in moving to lavf/lavc for this [17:37] ubitux: no, I mean: do you mean decoding or rendering? [17:37] demux and decode [17:37] right now you would get an ASS string which you could directly send to libass [17:38] it will change [17:38] here is how it would look like: http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/2012-September/130474.html [17:39] basically you would be in the video player path in the diagram [17:39] (btw if you have any comment/suggestion on this) [17:39] (work is already started) [17:41] decoding, maybe, re-rendering, no. [17:42] you would be responsible for the rendering [17:42] what do you mean by "re" rendering? [17:43] the "libass to render" is in your scope; if your player has libass you can decide to encode the avsubtitle into ass markup; otherwise you can encode it in "raw text" so it can be printf'ed or displayed with a basic osd [17:44] (that's what the rendering filter in lavfi would do) [17:44] i wish someone would finish the work on libass to make it free of fontconfig on windows [17:47] zgreg did [17:49] a patch to vlc is not exactly what I would call non-fontconfig support in libass [17:50] He released 0.10.1 and updated his work on harfbuzz in order to go to this point [17:50] there is a branch in the google code git repository, but it has some open TODOs [18:11] nevcairiel: but it is on the way [18:20] ffmpeg.git 03Ash Hughes 072470851f1228: lavc: enable recursively using avcodec_open2/close. * 03http://tinyurl.com/9w5p8t903 [18:54] - h->cbp= cbp; [18:54] + h->cbp = cbp; s->current_picture.f.mb_type[mb_xy] = mb_type; [18:54] am i the only one to find this ridiculous? [18:54] (\n after ';', irssi ate it) [18:56] same for the switch cases.. :/ [18:56] Action: ubitux really doesn't get it [18:57] durandal_1707 : pong [18:57] ubitux : you saw the sub samples in the repo, right ? [18:58] Compn: does playing QDMC works for you? [18:59] in mplayer ? [18:59] i dont have latest svn mplayer, a lot of things are broken due to planar output atm [18:59] but svn mplayer have fixed that [19:00] http://samples.mplayerhq.hu/A-codecs/QDMC/rumcoke.mov works with lavf demuxer and binary quicktime codec [19:00] and it does not need to be latest, because QDMC works only via binary codec [19:00] Compn: the mpsub in a lost directory on mphq [19:01] for me , using r34835 [19:01] (not the samples directory iirc) [19:01] Compn: good [19:02] ubitux : whut? theres http://samples.ffmpeg.org/sub/ [19:02] ffmpeg.git 03Michael Niedermayer 07d86ef54476b3: avcodec_align_dimensions2: add missing breaks * 03http://tinyurl.com/999d77v03 [19:02] ffmpeg.git 03Michael Niedermayer 070fa26bd4703c: utvideoenc: fix theoretical integer overflow * 03http://tinyurl.com/8l96hz603 [19:02] ubitux : http://samples.ffmpeg.org/sub/manyfmts/ contains the mpsub samples generated by some random subtitle program [19:05] wow. :) [19:05] thx [19:06] theres also this [19:06] http://www1.mplayerhq.hu/DOCS/tech/mpsub.sub [19:12] durandal_1707 : anything else you want me to test with quicktime stuff ? [19:13] no, i just need brainz to start REing [19:16] ah [19:16] wonder if kostya did any work on it [19:17] qdm2 has lots of authors :) [19:17] decoder i mean [19:20] durandal_1707 : you've been busy RE'ing lots of stuff :) [19:20] Compn: that was old times... [19:20] ben larsson wrote this up : http://multimedia.cx/mirror/qdmc2.pdf [19:21] has some info on qdmc i think [19:22] oh , no it doesnt :D [19:22] Action: Compn says stuff before reading stuff [19:34] qdm2.c is hard to follow :) [19:56] Hey guys, oss colleague of mine is sitting down with the Visual C++ guys re: why oss devs might not be using Visual Studio/VC++ compiler. Looking to collect your feedback and pass it on as I've hacked on ffmpeg/win32 in the past. [19:56] I understand C99 is important, anything else? [19:56] C99 features are the biggest reason unless the project is C++ [19:57] ffmpeg now has MSVC support because there's a clang-based converter now [19:57] which converts /some/ C99 features into C89 [19:57] Ah yeah, saw that. That's kinda lame tho isn't it, lol [19:57] and makes ffmpeg compile'able in the first place [19:57] rrivera, well how else would you support it then o_O Because it's pretty sure that MSVC ain't going to support C99/11 any time soon [19:58] you take the most realistic path that's possible [19:58] I hear you, yup. [19:59] rrivera : patches are welcome to make ffmpeg work easier in vc++ compilers and apps [19:59] and then there's plenty of windows hate in various OSS circles, certain people not being ready to make limitations or extra changes regarding windows or MSVC specifically unless pushed a lot [19:59] as long as they dont clutter the code in huge ways... [20:00] those are the two biggest reasons I know why OSS projects don't use the MS compilers [20:00] actually we support a lot of mingw/msvc now JEEB [20:00] yes I know that [20:00] I just noted [20:00] especially because chrome uses msvc to compile ffmpeg ... [20:00] ah [20:00] ffmpeg.git 03Michael Niedermayer 0731fdf3065dac: vf_idet: reorder operations to avoid division by 0 * 03http://tinyurl.com/9f6zo4903 [20:00] ffmpeg.git 03Michael Niedermayer 07042a738b4599: vf_mp: check list in querry_format() * 03http://tinyurl.com/8jtsj5903 [20:00] ffmpeg.git 03Michael Niedermayer 07aaf78e4d14b4: vf_mp: fix null ptr deref in case of ENOMEM * 03http://tinyurl.com/9mg26dh03 [20:00] I'm just talking about overall (F)OSS traits that lead to MSVC not being used [20:00] ah [20:01] Yep gotcha. I'll put another tick in the C99 list haha [20:01] rrivera : most of the devels use linux. so they stick to linux tools. [20:01] some use msvc and its debugging tho [20:02] the first being the fact that stuff uses C99/11 features (and possibly GNU extensions), second being linux/*nix developers and possible wish to not "filth" the project with MSVC-specifics (given the thing can even be built with MSVC to begin with) [20:02] and one of ffmpeg's core goals is to be cross-os-compatable [20:03] so vc++ stuff probably doesnt work well on mac for instance... [20:03] or on ppc, arm, and other arch [20:04] they did re-add ARM again lately, but yeah [20:04] it's very specific, and if you're cross-platform then... [20:04] yeah i dont think visualstudio was cross-plat back in the year 2000/2001 when ffmpeg was started [20:05] I would think Windows is a big consumer of ffmpeg [20:05] it is [20:05] Interesting that it's a back burner OS for ffmpeg-dev [20:05] vlc distributes millions of copies every year for windows [20:06] 19:00:17 especially because chrome uses msvc to compile ffmpeg ... --> iirc not yet [20:06] not sure about that though [20:06] there was some talk of a script that converts c99 to what vc++ stuff needs [20:06] a pre-compiler script [20:07] you mustnt be very observant >.> [20:07] chrome does use msvc now (or will extremely soon-- theres stuff sitting in gerrit) [20:07] and ffmpeg has built with msvc for a while now [20:07] Interesting that it's a back burner OS for ffmpeg-dev <- people've used it with mingw toolchains for years, then linking to their MSVC-built software [20:08] yeah, people have been using ffmpeg on windows successfully [20:08] there is a ffmpeg-msvc port , why arent more vc++ people helping to develop that ? [20:09] Compn, ffmpeg git master builds with msvc [20:09] stop spreading misinfortmation [20:09] Daemon404 : tell it to kierank [20:09] i mean i don't think chrome use ffmpeg with msvc yet [20:09] iirc [20:10] rrivera : what problems are there wtih ffmpeg ? submitting some bug reports is the quickest way to get ffmpeg to fix things .... [20:10] http://trac.ffmpeg.org/trac/ffmpeg [20:10] Compn: Hmm, what? [20:11] Daemon404 : http://trac.ffmpeg.org/trac/ffmpeg/wiki/KnownCompilationIssues and http://trac.ffmpeg.org/trac/ffmpeg/wiki/CompilationGuide both mention msvc being uncompilable [20:11] i read it on teh internet [20:11] Daemon404 : if the wiki needs to be updated, yell at it for spreading misinfo [20:11] on a wiki even! [20:11] must be true! [20:11] our wiki! [20:11] lol [20:11] never mind http://ffmpeg.org/platform.html [20:11] that OFFICIAL docs [20:12] oh yeah [20:12] rrivera : so what problems are there with ffmpeg and visual studio ? [20:12] i ask because i dont know them [20:12] Compn: Not sure, I typically cross-compile. I'm just asking for feedback to pass onto Visual C++ team. :) [20:13] heh [20:13] which team / company are you talking to ? :) [20:13] Action: Compn updates wiki [20:13] I'm just passing it onto my colleague Garrett (http://twitter.com/fearthecowboy/status/259340782337671168) [20:14] well the answer is ---- ffmpeg supports vc++ :) [20:17] there. [20:18] rrivera : send him that info on the c99 to c89 script. thats probably useful for a lot of projects... [20:21] Action: durandal_1707 compiling mplayer, result: libavcodec/x86/cabac.h:170:9: error: ran out of registers during register allocation [20:22] durandal_1707 : ehe, michael vs gcc bug :) [20:22] it should have been fixed [20:22] re run the last command with -fomit-frame-pointer [20:22] i'm using clang [20:22] ah [20:22] well -fomit-frame-pointers then (i cant remember pointers / pointer) [20:22] if thats possible with clang... [20:23] why mplayer recompiles ffmpeg - how to forece it to use shared lib? [20:23] ./configure --help | grep -i ffmpeg |grep -i external [20:23] or something [20:23] or rm -rf mplayer/ffmpeg [20:23] maybe [20:23] no idea [20:24] the better question is [20:24] why does mplayer use a shitty fucked up custom build system for ffmpeg [20:24] at least mplayer2 got rid of that iirc [20:25] Daemon404: last time i checked it still use some internal stuff [20:25] have to ask dondiego [20:26] mplayer uses internal stuff, for some filters and other things [20:26] durandal_1707, i thought mplayer2 supported system ffmpeg now [20:26] mplayer2 got rid of those filters i think :D [20:26] or maybe just for mplayer2 shared ffmpeg [20:26] i bet theyre nto even useful filters [20:26] Daemon404: mplayer2 and mpv support system libav* [20:27] yeah [20:27] i havent used 'normal' mplayer in a long time. [20:28] wtf, libpostproc can not compile with latest lavu [20:28] according to mplayer output [20:29] durandal_1707: huh [20:29] Checking for libpostproc ... Package 'libpostproc' requires 'libavutil = 51.41.100' but version of libavutil is 51.65.100 [20:30] Action: Compn runs away from version hell [20:31] durandal_1707: Some pkg_config problem it seems [20:31] yea [20:31] i forgot to remove it when removing lib [20:39] Compn: looks like with latest mplayer i cant play QDMC [20:41] forget that it works [20:41] something was corrupted in binary blobs [20:53] durandal_1707 : yeah , theres some versions of the binary which dont work (qt 6.5 i think does not, but qt 6.3 does) [21:58] <@Compn> http://www1.mplayerhq.hu/DOCS/tech/mpsub.sub // i was talking about this one :) [22:55] ffmpeg.git 03Michael Niedermayer 073d48dd01fd3d: avidec: remove unneeded null check * 03http://tinyurl.com/9nyphoz03 [22:56] ffmpeg.git 03Michael Niedermayer 07d30351363f6f: url_alloc_for_protocol: fix use of uninitialized variable * 03http://tinyurl.com/9cb38se03 [22:56] ffmpeg.git 03Michael Niedermayer 07a96577df3846: avio: fix sizeof argument * 03http://tinyurl.com/9hfsfp803 [23:36] Action: durandal_1707 Yes!! [23:37] @_@ [23:40] ffmpeg.git 03Michael Niedermayer 07395caf3de84a: hls: fix integer overflow * 03http://tinyurl.com/8snqvdk03 [23:40] ffmpeg.git 03Michael Niedermayer 07ba39303050c1: gxfenc: fix null ptr dereference * 03http://tinyurl.com/9osqqhj03 [23:40] ffmpeg.git 03Michael Niedermayer 078fb8d539a4c5: hlsproto: fix integer overflow * 03http://tinyurl.com/96eg79o03 [00:00] --- Sat Oct 20 2012 From burek021 at gmail.com Sun Oct 21 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Sun, 21 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121020 Message-ID: <20121021000502.7E34E18A01ED@apolo.teamnet.rs> [00:08] ffmpeg.git 03Hendrik Leppkes 07b87ff3449662: vc1: implement vc1 field interlaced dxva2 decoding * 03http://tinyurl.com/8wlpu6r03 [00:08] ffmpeg.git 03Hendrik Leppkes 0733f2a4942380: vc1: only disable interlaced b-frames for software decoding * 03http://tinyurl.com/8lgvq8703 [00:33] saste: i have a strange issue with lavfi device and ffprobe, which is not reproducible with ffmpeg & ffplay [00:33] interested? :) [00:34] ubitux, what? [00:34] and don't forget to file a bug report, if you don't plan to work on it [00:34] i kind of try to understand it at the moment [00:34] i'm doing ./ffprobe -f lavfi -i 'amovie=input.mp3,ebur128=video=1 [out0][out1]' -show_frames [00:35] it stops immediately without error [00:35] av_read_frame() is returning an "invalid argument" err code [00:35] (coming from the request frame thing) [00:36] what if you remove ebur? [00:36] works fine [00:37] actually even a split & showwaves works [00:37] so maybe i miss something in the filter [00:37] but since it works fine with ffplay & ffmpeg.. [00:37] ubitux: gdb... [00:37] Action: ubitux prefers random av_log debugs [00:38] may be related to internal caching [00:38] i was just asking in case you may have an idea [00:48] saste: yes, the fifo looks empty [00:49] uhm... [00:52] it looks like the main difference with ffplay is that ffplay just doesn't care and continue to read frame anyway [00:55] ubitux: and ffmpeg? [00:55] ffmpeg has no issue either [00:55] and yes i remember that i saw a similar error [00:55] why not? [00:56] mmh maybe it has.. [00:56] heh it has an explicit error [00:58] well, i'll see later? [00:58] 'night [00:58] saste: tomorrow is your last chance to comment on the lavfi metadata btw ;) [00:58] Action: ubitux & [00:59] ubitux, are you waiting for reviews? [01:00] what about the issue spotted by nicolas? [01:00] anyway give some time, i'll try to review tomorrow [01:00] if not feel free to push it [01:37] seeking in he-aac-v2 have problems [02:02] ffmpeg.git 03Hendrik Leppkes 07953a3dcc4e28: Mark data symbols shared between libraries with av_export * 03http://tinyurl.com/8qq2auf03 [02:02] ffmpeg.git 03Hendrik Leppkes 0704bf2e7f0e9c: swresample: include ff_log2_tab for shared builds * 03http://tinyurl.com/9x793yo03 [02:03] hi [02:08] hi [03:06] hi ramiro [03:06] do you want a coverity account ? [11:31] ffmpeg.git 03Stefano Sabatini 076d6ccbae4cee: examples/decoding_encoding: add missing checks on avcodec_alloc_context3() * 03http://tinyurl.com/8sfr9n703 [11:31] ffmpeg.git 03Stefano Sabatini 078c4753f7f5f1: examples/decoding_encoding: remove misplaced and confusing comment * 03http://tinyurl.com/8l6ardw03 [11:31] ffmpeg.git 03Stefano Sabatini 077b116a94af99: examples/decoding_encoding: fix misc typos in the usage text * 03http://tinyurl.com/9ju6pws03 [12:12] ffmpeg.git 03Stefano Sabatini 071cd9c81ddb31: lavc/utils: extend feedback provided by avcodec_open2() * 03http://tinyurl.com/8sq24yg03 [12:12] ffmpeg.git 03Stefano Sabatini 077bc533c41b40: lavc/utils: fix a few case/punctuation inconsistencies in avcodec_open2() * 03http://tinyurl.com/8z2cs5c03 [12:12] ffmpeg.git 03Stefano Sabatini 07935ecfb00238: examples/scaling_video: remove unnecessary intermediary variable in fill_yuv_frame() * 03http://tinyurl.com/9aq6tmf03 [12:22] ffmpeg.git 03Stefano Sabatini 07cdea54b4c8cd: lavu/parseutils: rework rational reduction logic in av_parse_ratio() * 03http://tinyurl.com/9bs83ga03 [12:43] ffmpeg.git 03Michael Niedermayer 076bcdfe48d0d0: mpeg4videodec: Disable frame multithreading for GMC, its not implemented at all * 03http://tinyurl.com/9tgkg7k03 [12:43] ffmpeg.git 03Diego Biurrun 07c896aa984e88: build: Drop OBJS declaration for non-existing PCM_DVD encoder * 03http://tinyurl.com/9quocpn03 [12:43] ffmpeg.git 03Diego Biurrun 078fb1e2640505: lzo: Drop obsolete fast_memcpy reference * 03http://tinyurl.com/8nucuew03 [12:43] ffmpeg.git 03Diego Biurrun 074b587848ce03: configure: Disable Snow decoder and encoder by default * 03http://tinyurl.com/96m9mnq03 [12:43] ffmpeg.git 03Martin Storsj? 0712549db6535c: fate: Add proper dependencies for the tests in video.mak * 03http://tinyurl.com/9qdyr6303 [12:43] ffmpeg.git 03Hendrik Leppkes 07d2d08d706b8f: gitignore: ignore files created by msvc * 03http://tinyurl.com/8ncnbsc03 [12:43] ffmpeg.git 03Martin Storsj? 072f41eaa9c6b5: rtsp: Make sure the ret variable is initialized in ff_rtsp_fetch_packet * 03http://tinyurl.com/9hcvvju03 [12:43] ffmpeg.git 03Mans Rullgard 071846ddf0a787: ARM: fix overreads in neon h264 chroma mc * 03http://tinyurl.com/8ft4fht03 [12:43] ffmpeg.git 03Mans Rullgard 07c9ef43215c7d: fate-vc1: add dependencies * 03http://tinyurl.com/9zy67nl03 [12:43] ffmpeg.git 03Michael Niedermayer 0704c6ecb7da67: Merge commit 'c9ef43215c7d68c2cdcdbe02287aa114f27a32ed' * 03http://tinyurl.com/98b3o8m03 [13:30] ffmpeg.git 03Diego Biurrun 07c08536979be1: avutil/lzo: K&R formatting cosmetics * 03http://tinyurl.com/8pyva9y03 [13:30] ffmpeg.git 03Diego Biurrun 075532cf317838: avutil/mem: K&R formatting cosmetics * 03http://tinyurl.com/9audojc03 [13:30] ffmpeg.git 03Diego Biurrun 0779042ab37619: configure: Group math functions into a separate variable * 03http://tinyurl.com/966qcpy03 [13:30] ffmpeg.git 03Michael Niedermayer 076912e7a008ac: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/9cs2zuz03 [14:05] ffmpeg.git 03Michael Niedermayer 07bf52ad1e49d3: lavc: revert broken hunk from 1cd9c81ddb317ca00061d11d3562d3a34888d91b * 03http://tinyurl.com/9xw3wbu03 [14:41] <@saste> what about the issue spotted by nicolas? // i fixed it and sent a new version [17:03] oh no! bsd fail! [17:03] :P [19:39] ping for lvf [20:25] same_quant should be put back [20:28] Action: michaelni is wondering where cone went ... [20:29] j-b, is cone ok, she didnt report the last few commits ? [20:29] durandal_1707, even though it didn't work as it even was supposed to in the first place, and even if most damn users were misusing it? [20:29] it's like wtf [20:30] if it's put back [20:30] it was the most misused option ever [20:30] I think it even set quants to zero for a long time now instead of copying them [20:30] but there is guy, look above ffloger post, that is using it and it is useful to him [20:31] it's not useful for him, he just thinks it's the proper damn option [20:31] and I'm pretty sure he wasn't getting the same quants [20:31] also he could just use a low enough quant with the setting [20:35] didn't we add a dummy sameq/same_quant option displaying a warning? [20:35] yep [20:35] durandal_1707, not to mention that sameq IIRC would just give random results if the input format was different from the output one [20:36] I really don't see why sameq would have to be brought back to life [20:36] it just hack to be used with mpeg2? [20:36] ... [20:37] it was a hacked that used to be used for development of something or another to copy the quants of the same format input to the output. It has for all I know not worked for a relatively long amount of time and most of its users are thinking it's "same quality". [20:37] the first sentence already tells you why with MPEG-4 Part 2 -> MPEG-2 you shouldn't be using it [20:38] it was probably just setting a low enough quant there /by chance/ [20:38] omg [20:38] he would have better and more predictable results by just using a certain low enough quant instead [20:40] saste: well, i don't know how that buffering is supposed to work; i'll likely open an issue soon :( [20:40] (the lavfi device issue) [20:44] wtf some of my mails didn't reach ffmpeg-devel.. ? [20:45] arg again i replied in private.. [20:53] someone please copypasta what I said on the trac on that sameq issue or someone might actually end up putting it back >_> [20:53] one could fix the issues and then put it back [20:54] why? I understand it's a usable setting for development, but in this case it was different formats to begin with [20:54] JEEB: http://ffmpeg.org/trac/ffmpeg/wiki/Option%20%27-sameq%27%20does%20NOT%20mean%20%27same%20quality%27 [20:54] he was depending on (most probably broken) utility [20:54] cbsrobot, and why am I getting that linked to me? [20:55] well it's already a subject on trac [20:55] we could also just point the ticket to the page [20:56] which doesn't mention what I noted [20:56] Action: cbsrobot reads it again [20:56] "...that used to be used for development of something or another to copy the quants of the same format input to the output...." [20:56] the guy on trac wants it for MPEG-4 Part 2 -> MPEG-2 [20:56] IMHO it'd be better to just set a low'ish quant or something [21:00] This option actually means "same quantizers" and should only be used to copy the quants from the input to the output of the same format (f.ex. mpeg2 -> mpeg2). [21:02] JEEB: what option should he use instead? [21:03] he would have better and more predictable results by just using a certain low enough quant instead [21:11] but we should at least warn the user [21:11] just dropping an option like this overnight is bad [21:12] saste: option is ignored [21:12] also, aren't there some proper use of the option? [21:13] yes, but I haven't seen it used in the last few years [21:13] copy quantization coefficients may make sense when one is transcoding+filtering and re-encoding to the same codec [21:14] personally I'm just happy a setting that has caused so much herp and derp is no longer there, and I don't think re-adding it for some development work would be much work if someone becomes to have a need for it during development [21:14] (although I've heard it didn't work in the end for a relatively long time, just setting some low'ish quants) [21:15] transcoders are evil [21:17] fb722a900fc5cc9e003b9fef25b27ed7fc5547a2 [21:17] i can agree on the rationale, but breaking API should be avoided [21:17] anyway maybe it should *fail* rather than print a warning [21:18] saste: than revert michaelni commit [21:20] ahah it was even mentioned in the FAQ [21:20] no wonder people were misusing it [21:20] i misused it as well [21:21] wait.... [21:21] that's because people looks for a magic option - keep same quality [21:21] should just make ffmpeg -magic .... [21:22] having to set a quantization parameter is not that user-friendly (you have to get a clue about what the quantizer is in the first place) [21:22] Compn: ffmpeg -dwim [21:23] ffmpeg -thatfilterfromavisynth-daemonkeepstalkingabout [21:23] ffmpeg `cat /dev/brain` [21:23] Action: Daemon404 has been hilighted [21:23] michaelni, do you happen to know sort of input the mpeg2video decoder expects? [21:24] im passing it a gop offset in a mpeg2 PS [21:24] and its... -sorta- working [21:24] [15:21] <@Daemon404> http://i.imgur.com/20jJm.png [21:24] [15:21] <@Daemon404> ^ output luma [21:24] [15:21] <@Daemon404> doesnt look quite right ;) [21:24] e.g. i i know teh offset of a GOP in an mpeg2 ps file, and i read in from that offset, and pass to avcodec [21:24] but the thing is that -sameq sometimes seemed to work [21:25] at least it was better that no option at all [21:25] so the question is, why was it working? [21:25] Daemon404, the decoder expects MPEG-ES video frames [21:25] michaelni, im not -that- fimiliar with mpeg-ps [21:25] but should a gop offset point to mpeg-es frams? [21:25] shoudlnt* [21:26] what is a "gop offset" ? [21:26] sec [21:26] http://neuron2.net/dgmpgdec/DGIndexManual.html#AppendixA [21:26] the "position" marker [21:27] "The position field is a decimal number that points to the byte position in the source file at which decoding of the indexed I picture should begin." [21:27] ugh i was wrong, there was no API break (no lavc option was removed) [21:28] michaelni, what about to fail rather than printing a warning in case of -sameq? [21:28] Daemon404, a I-Picture can be split in many lets call them chunks at the PS level [21:29] or more precissely, MPEG-ES frames get concatenated and then randomly choped in a series of bytes [21:29] then headers added and interleaved with audio then more crap added and you have MPEG-PS [21:29] and direct the user towards some option which may achieve the same effect [21:30] i see [21:30] saste, theres no option that achives the same effect currently [21:32] michaelni, mention something about quality options? [21:32] there are so many of them [21:32] Action: saste feels like we need a proper manual for the lavc options [21:33] the most fitting would probably to mention that the user should select a bitrate similar to the input [21:33] but if people want this option so much we should think about fixing it and putting it back [21:33] michaelni, is there anything tangible i can do with libavcodec given these "gop offsets" [21:33] convert the quality between formats and all that [21:34] (if i understand correctly what they are) you can use them as seed for our random number generator [21:34] <.< [21:35] michaelni, I have no clue about how -sameq was working, but it was quite hackish [21:35] it wasnt working [21:35] it was *somehow* working in some cases [21:35] Action: Daemon404 thinks it was a coincidence [21:36] the problem with sameq was that it was written 10 years ago and not maintained since then [21:36] or something like that :) [21:36] yes [21:36] I hear it had been broken for a /long/ time [21:36] and just used some zero/low quant [21:39] michaelni, so given those offsets, is there a way to get it into an mpeg-es format easily? [21:39] JEEB: allright, it was copying the quality param of the input decoded frame, and putting it in the encoded frame [21:40] that info was possibly used by the encoder [21:40] Daemon404, if you have a MPEG-PS file you need to run it through a demuxer that produces ES frames [21:40] or a ES stream which you can feed through a AVParser [21:40] from avcodec.h: quality (between 1 (good) and FF_LAMBDA_MAX (bad)) [21:41] michaelni, i may only access the ps at those offsets [21:41] and start reading from there [21:41] i dont know if avparser can handle that [21:41] i.e. i am NOT using libavformat [21:41] to open the ps [21:41] if there is PS headers in the middle of the coded frames, the parser wont remove those [21:41] FF_LAMBDA_MAX (256*128-1) defined in avutil.h [21:41] you may need to write some magic to do that [21:42] nevcairiel, its certainly possible [21:42] every d2v related filter in existence has done it [21:42] in general it seems like something related to some specific codec, so I agree it doesn't make much sense to rely on that option [21:42] i suppose it was working more or less when transcoding from MPEG-family-codec to MPEG-family-codec [21:42] its not like PS parsing is that complicated [21:43] nevcairiel, yeah but i have no idea how? [21:43] i never looked at PS much. i could tell you how to deal with TS :) [21:44] ;p [21:44] im am readign dgdecode's source [21:44] pity me [21:44] Daemon404, if it is PS you could try to feed it through libavformats PS demuxer [21:45] mmm maybe [21:45] this code is lol... [21:45] // If the D2V filename has _P just before the extension, force [21:45] // progressive upsampling. [21:45] surely this cannot be a bad idea... [21:50] heh.. can lavf even open a buffer [21:50] or does it only take filenames... [21:50] you can feed it buffers, but need to give it a custom avio context [21:51] what func? [22:04] wtf is the ps demuxer called [22:04] there isnt a mpegps [22:05] its called "mpeg" [22:06] ff_mpegps_demuxer ... name= "mpeg" [22:07] i see this now... [22:11] is mkv streaming actually supported? [22:18] mpf. i need to introduce another side data thing [22:18] (codec srt -> codec subrip) [22:19] ubitux, btw make sure all the metadata stuff gets fate tests so it is all regularly tested [22:20] mmh [22:20] could easily break with noone noticing otherwise [22:20] maybe i should add a ffprobe test indeed [22:20] aha.. works! [22:21] michaelni: do you have a video fate sample in mind with 2-3 scene changes? [22:22] no but one could run a scene detect filter over all the samples to find a good candidate [22:22] small file low resolution is probably preferable [22:22] so its faster [22:22] yep [22:23] i need to find a sample with silence too [22:28] michaelni: any other comment btw? [22:29] ubitux, didnt really look at the latest patch yet [22:29] just wanted to point that fate thing out ... [22:30] yes, good point :p [22:48] meh side data doesn't get transmitted [22:51] (with subtitles) [22:57] oh it's because ffmpeg isn't reusing the decoded subtitle& [23:03] rha the subtitles api really is completely out of time :D [23:15] seems related to AVFMT_FLAG_KEEP_SIDE_DATA [23:15] what's this side data merge thing? [23:20] michaelni: av_packet_merge_side_data() in lavf/utils.c:ff_read_packet is trashing the side data set in the subtitle demuxer [23:20] any idea why? [23:22] no [23:23] :( [23:30] j [23:30] oh ok the function is moving the side data at the end of the pkt data... huh. [23:31] Action: ubitux doesn't get the point [23:32] Action: ubitux doesn't get how that's supposed to work [23:33] is there nice way to read null terminated string from avio? [23:33] ubitux, the point is that older user apps do not preserve side data but just the main data between lavf and lavc [23:34] its easy to disable the merging if its unwanted [23:34] yes with keepside option [23:34] but& [23:34] how is the side data supposed to be restored? [23:34] magic :) [23:35] oh indeed the magic thing [23:35] FF_MERGE_MARKER? :p [23:35] av_packet_split_side_data() [23:36] yeah ok [23:36] so i'm supposed to get the data back [23:37] and that function is called only for audio & video, ok, get it. [23:37] thank you :) [23:37] ahh uhm, i forgot adding that for subtitles ? ^^;; [23:37] seems so :p [23:43] i forgot name of function that reads string up to NULL or X chars whatever comes first [23:44] something like bprint + a avio_r8() loop? [23:45] ff_get_line maybe [23:45] libavformat/aviobuf.c:int ff_get_line(AVIOContext *s, char *buf, int maxlen) [23:46] michaelni: could we simply disable it for subtitle btw? [23:47] since no subtitle use side data at the moment, the "old user app" should not need it, right? [23:47] (assuming the subtitles side data are always optional) [23:47] do you plan to keep the split side data thing indefinitely btw? [23:48] ubitux: actually i need it in lavc [23:48] durandal_1707: i'd be interested in a bprint version of ff_get_line btw [23:48] too bad :p [23:48] want a bytestream_get_line? :) [23:49] avio_get_str(utf8) variant for lavc [23:49] good morning [23:50] j-b: good midnight [23:50] ubitux, we could but that would be inconsistent [23:54] i will use strnlen [00:00] --- Sun Oct 21 2012 From burek021 at gmail.com Sun Oct 21 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Sun, 21 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121020 Message-ID: <20121021000501.7907618A018E@apolo.teamnet.rs> [00:07] needs_more_info [00:07] does mkv file with aac have same duration ? [00:07] yes [00:08] mkv containing aac_he_v2 has the same duration as the original input file [00:09] what you use to play mp4? [00:11] deadbeef [00:12] does ffmpeg -i input.mp4 show correct duration? [00:17] durandal_1707: yes, it shows correct duration [00:21] so it appears to be deadbeef problem and how it calculates duration [00:22] how are deadbeef libav* plugins build? [00:22] perhaps it uses older versions of library.... [00:24] may be, I'll investigate about it... [00:25] otherwise, when I convert flac to aac_he_v2 into mp4, deadbeef, aqualung and VLC are having problems playing it [00:26] coding aac_he into mp4 works well on all of them, but not aac_he_v2 [00:26] its confusing........ :S [00:29] does aac_he_v2 decodes correctly with ffmpeg? [00:29] (it may be that such profile is just broken) [00:32] yes, perfectly. If I convert from mp4 to mp3, it works perfect... [00:32] in any case, How can I check if 'aac_he_v2' profile is broken? [00:35] hmm, dunno perhaps deadbeef use broken aac decoder [00:35] I understand [00:37] in last place and in order to discard any possible error made by myself, this command is 100% correct? [00:38] ffmpeg1 -i input.flac -acodec libfdk_aac -profile:a aac_he_v2 -ab 60k -ar 44100 -ac 2 output.mp4 [00:39] it should give me an mp4 file containing high quality AAC-HE v2 audio? [00:43] if you play it with quicktime and it plays correctly then yes [00:46] yo saste [00:46] hi DelphiWorld [00:46] saste: do you know Sast-NG? [00:46] saste.-ng? [00:46] durandal_1707: ok, I'll try it! [00:46] thanks!! [00:46] an improved version of saste? [00:47] saste: first of all what is sast ? [00:47] i have no idea [00:48] saste: saste-ng is a softcam ;) [00:48] saste: when i see you i remember it;-) [00:50] can't find it on the internet [00:50] saste: please kill me:( [00:50] saste: my assness [00:50] saste: it's sasc:P [00:52] nighty [12:14] hi there, when I convert my OGV videos from recordmydesktop like so: ffmpeg -loglevel quiet -i out.ogv youtube-upload.avi [12:14] the avi files become "artefacted" and not as clear as the original out.ogv [12:14] it's like converting a PNG to JPG [12:15] any tips to avoid this lossy conversion, yet still be able to upload to Youtube ? [12:15] Capture your desktop using ffmpeg. [12:15] why? just upload the original file [12:15] That's what recordmydesktop uses... [12:17] sacarasc: is there a guide to doing that somewhere? [12:18] Tjoppen: the orginal file that recordmydesktop produces: out.ogv, does not upload to Youtube. [12:19] sacarasc: for e.g. `ffmpeg -f x11grab -s wxga -r 25 -i :0.0 -sameq /tmp/out.mpg` doesn't record audio [12:20] that's strange, because youtube uses ffmpge [12:20] also: don't use avi [12:20] ffmpeg -i out.ogv -vcodec libx264 -crf 20 -acodec copy out.mov or something like that [12:23] Tjoppen: why not .avi ? [12:39] Tjoppen: btw sound is not working in the out.mov [13:23] are there any support for opencl/cuda in ffmpeg for h.264 or web encoding? [13:27] RoyK, there is a patch that's under review currently, which moves lookahead me to opencl in x264 IIRC [13:27] but [13:27] it's the usual case of "do I really want to lower the quality and get a 10% speed boost with a high-end GPU instead of just investing on a new CPU" [13:29] it?s just that a colleague of mine told me he got a 4x speed gain by using the gpu on his machine with some proprietary software [13:29] s/it?s/it's/ [13:29] first of all, did he properly test or just see the speed with random settings? [13:30] because x264 can in general be very fast f.ex. [13:31] also I'm not even sure how fast those GPU-based encoders are nowadays, but they mostly suck because GPUs need you to be able to throw a lot of stuff at the same time, aka the workload really has to be well multithreadable. Unless you strip features out of H.264 to its barebones you're not going to get that simply on the GPU [13:32] when you have something using GPU for certain tasks, that /can/ be faster, but not always without a cost and the actual CPU algorithm might've not been fast to begin with [13:33] anyone know how i get the playing length of an AVI file? [13:33] the integrated encoder chips in intel stuff f.ex. are really dedicated hardware for encoding and the results of those generally are quite a bit better than pure GPU encoders' [13:34] but those too are mostly currently riding on the side of "we're fast and we're not using the 'CPU'", and the speed might be attainable with similar/better quality from x264 with a fast preset [13:36] anyways, OpenCL/CUDA is not a silver bullet under any circuimstances. CUVID decoding can help on newer hardware stuff to take some stuff off of the CPU, but if you go fast enough on the encoder's side that might end up being your bottleneck since the hw decoder can only output as many frames per second [16:16] hi, i'm trying to analyze the audio files we have in MegaGlest (an open source RTS game, http://megaglest.org ), to get an idea whether playing them back is causing elevated CPU load (memory matters, too, but less so), and to find out ways to reduce the overhead they cause without decreasing the audible quality more than a little bit. [16:17] here's what i find, including high (?) bitrates of 1411 kbps: http://paste.ubuntu.com/1292282/ [16:17] i'm looking into this because profiing the game turns out that a rather high percentage (up to 8%) of the processing power of the test system goes into audio processing. [16:18] now any suggestions on what to look into would be welcome, i'm pretty helpless with audio and audio processing. [16:22] maybe ./ffmpeg -nostats -i input.wav -benchmark_all -f null - ? [19:57] while trying to configure ffmpeg for compiling, I got "libtheora not found". I tried with ffmpeg 1.0 package and the latest code from git. is this a known problem could it be that I am doing something wrong? [19:58] Element9: paste relevant config.log somewhere [20:02] i'm trying to grab to a mov, but [20:02] ffmpeg -f x11grab -s 1366x768 -i :0.0 -vcodec libx264 -crf 20 -acodec libfaac -f alsa -i hw:0,0 out.mov # results in Unknown decoder 'libx264' [20:02] any ideas? [20:03] before -i = settings for decoders/inputs [20:03] so in other words exactly what it told you [20:03] there is no such decoder as libx264 (and no libfaac decoder either) [20:03] probably because of the fact that you have two -is [20:04] also I would recommend fdk instead of faac if you want aac [20:04] https://github.com/mstorsjo/fdk-aac [20:06] JEEB: `ffmpeg -f x11grab -s 1366x768 -i :0.0 -vcodec libx264 -crf 20 -acodec libfaac out.mov` works, so libx264,libfaac do exist [20:06] or must exist [20:07] it's just that cmdline has no sound [20:07] so it's useless to me :) [20:08] well obviously since you don't specify an audio input [20:08] hendry, did you like ignore all of what I said? [20:09] re-read my lines [20:09] also encoders are not decoders [20:09] Action: JEEB sighs [20:09] kinda spoonfeeding here, but try: ffmpeg -f x11grab -s 1366x768 -i :0.0 -vcodec libx264 -crf 20 -f alsa -i hw:0,0 -acodec libfaac out.mov [20:10] and the fdk stuff... no idea about that though [20:12] faac is actually pretty bad, but I guess building yet another binary is not an alternative to this guy :P [20:12] klaxa: that didn't work either [20:13] you make spoonfeeding sound like a crime [20:13] opus looks good, i think it misses a lot of container implementations though... as a standalone audio-codec it has a great quality/bitrate ratio [20:13] this stuff is really rather tricky ffs [20:13] did you google? did you look at the manpage? [20:13] klaxa: yup ! [20:13] more like, put those -is before you start setting up the encoder/output side settings [20:13] what did you google? also, if you say it doesn't work, that doesn't help [20:14] otherwise ffmpeg will parse those things before an -i as input/decoder settings [20:14] as I said [20:14] complete ffmpeg log would help [20:14] this stuff is really rather tricky ffs <- yes it is if you have never done it before, but you have gladly ignored what I told you. [20:15] hendry: sorry if we sound insulting, but it's actually kinda tiring to help if people ignore you [20:15] https://gist.github.com/3924250 is the log [20:15] JEEB: IIUC you want me to group -i together [20:16] A) i tried that, I get hw:0,0: Protocol not found [20:16] are you running pulse by any chance? that would make things easier (less efficient though) [20:16] oh wait [20:16] B) this is not a problem when I use huffyuv [20:16] you are missing -f alsa [20:16] i think? [20:16] -f also before -i hw:0,0 [20:16] *alsa [20:16] and after the video -i I would guess [20:16] you didn't run my command after all [20:17] oh no you did [20:17] however... >Unknown decoder 'libx264' [20:17] hendry, I would be surprised if it wasn't a problem in case of ffvhuff or huffyuv :) Because the inputs should be the same [20:17] Action: hendry wrote https://github.com/kaihendry/recordmydesktop2.0/blob/master/r2d2.sh#L24 today, trying to make ffmpeg more useful [20:18] ugh sameq [20:18] remove that [20:18] first of all it doesn't make sense with ffvhuff or huffyuv [20:18] because they are lossless [20:18] second of all I hear sameq has been broken for years and didn't mean what you think it does [20:18] my experience of huffyuv is that they are lossy [20:19] since the picture / colours are off [20:19] they are only lossy in the sense that there's a possible RGB->YCbCr conversion going on [20:19] e.g. http://r2d2.webconverger.org/2012-10-21/ [20:19] if you need rgb, you take on a codec that can do RGB [20:19] I think ffvhuff might be able to do that, or maybe utvideo [20:19] pick your choice off the lossless codecs [20:19] JEEB: well, I'm all ears. I don't want any loss ideally. I want it to look exactly like the screen capture [20:20] and i want it to work across browsers :-) [20:20] Action: hendry tries ffvhuff [20:20] utvideo at least does support RGB [20:20] I coded that implementation ^^; [20:20] durandal_1707: sorry, got carried away, trying to fix it myself. :) here's the log: http://pastebin.com/vjKs0W0p [20:20] JEEB: where would i get a list of video codecs with their respective color space? [20:21] and i want it to work across browsers :-) <- you're not telling me that you want the in-the-middle capture to be playable with browsers, right? [20:21] klaxa, I'm not sure if it's listed anywhere easy :s [20:21] :S [20:21] also there could be defaults [20:21] utvideo is lossless? [20:21] f.ex. libx264 is defaulted to 4:2:0 [20:21] yes [20:21] it's similar to huffyuv/ffvhuff [20:21] ah [20:22] Element9: have you installed libtheora? [20:22] just has implementations on all major OS [20:22] VFW/DS for Windows [20:22] QT thingy for Macs [20:22] and now libavcodec dec/enc as cross-platform [20:22] JEEB: I don't understand what you mean by "in-the-middle capture" [20:22] hendry, you don't usually throw the lossless file you capture onto the internet as-is [20:23] because if you do capture with one, it's usually quick with enc/dec and not exactly small [20:23] durandal_1707: yes [20:23] you then usually convert those to whatever you need in the end [20:24] and in most cases you need to convert to 4:2:0 YCbCr (colloquially called "YUV") for web video [20:24] so if you don't want any change of color in the end result, that's a "tough luck" kind of thing, but the change should be minimal if you match RGB->YCbCr and YCbCr->RGB conversion matrices [20:25] durandal_1707: I have there: codec.h theoradec.h theoraenc.h theora.h [20:25] in /usr/include/theora [20:25] s/there/these [20:25] of course 4:2:0 YCbCr means it will have chroma subsampling, but not like anything else is really supported in browsers (Except for Chrome or Flash) [20:25] Flash uses Mainconcept's decoder and thus can do up to 4:2:2 YCbCr (in both 8bit and 10bit), and unless Chrome removes features from their H.264 decoder they should be able to deal up to 4:4:4, or possibly RGB [20:26] JEEB: just tried both utvideo & ffvhuff. I think ffvhuff.webm is better, but still the colours are off [20:26] I tried configuring without theora and then it says libvorbis is missing. hmmm... [20:27] > ffvhuff.webm > WhatAmIReading [20:27] anyways, quality-wise they should be the same but the utvideo encoder is not threaded so prolly slower [20:27] I should really get to that one day :P [20:27] hendry, check your log btw [20:27] or gist it [20:28] that will tell you what possible colorspace conversions were made [20:28] Element9: you will probably need to add extra flags: --extra-cflags=-I/usr/include --extra-ldflags=-L/usr/lib [20:28] at least I would think that with utvideo there should be none, as I've supported all of the usual ones (RGB24, RGBA, 4:2:2/4:2:0 YCbCr) [20:29] JEEB: https://gist.github.com/3924250 [20:30] durandal_1707: lets try that [20:31] >-t is not an input option, keeping it for the next output; consider fixing your command line. [20:31] >Stream #0:0: Video: vp8, yuv420p, 1366x768, q=-1--1, 200 kb/s, 1k tbn, 29.97 tbc [20:31] looks like it gets converted to 4:2:0 YCbCr [20:32] yes [20:32] Stream #0:0 -> #0:0 (rawvideo -> libvpx) [20:32] and instead of ffvhuff you get vp8 [20:32] probably because it's webm [20:32] but this is weird [20:32] and vorbis instead of pcm_16se [20:32] it shouldn't just override like that [20:33] http://r2d2.webconverger.org/2012-10-21/utvideo.webm.html versus http://r2d2.webconverger.org/2012-10-21/ffvhuff.webm.html [20:33] *pcm_s16le [20:33] it should instead kick you for using a container that couldn't take your selection [20:34] hendry, you could report this as a bug I guess? Because instead of herping a derp at you it auto-sets things to be compliant for webm. Instead you probably would have wanted to get an error that would tell you that those things don't go into webm in case you /explicitly/ set thme [20:34] *them [20:34] durandal_1707: now I get "Symbol mangling check failed." [20:35] what OS is that? [20:36] JEEB: where should I file a bug? [20:36] durandal_1707: xubuntu and cross compiling for windows [20:36] Element9: not my experize [20:37] durandal_1707: oh, I think I saw a flag for cross compiling. I'll try to set that one [20:37] durandal_1707: I can always try to comment out that on check. maybe it goes well :) [20:37] durandal_1707: thanks for the help [20:38] hendry, trac I guess [20:40] hendry, also in general I would really say that you shouldn't be encoding straight into something that you give out as-is on the web, because you won't be able to compress it as well when it also has to be realtime [20:40] JEEB: tbh most of this stuff if over my head [20:41] JEEB: so if you could spoon feed me what I should be doing, I'd be extremely grateful [20:41] JEEB: I'll add it to https://github.com/kaihendry/recordmydesktop2.0/blob/master/r2d2.sh#L24 [20:43] hendry: record lossless to a temporary raw, then re-encode for publishing on the web [20:43] klaxa: do you have a step-by-step example please? [20:44] hendry, I'm not going to be copying your long line, but look at this, more or less, ffmpeg -vcodec ffvhuff -acodec pcm_s16le out.avi [20:44] meaning that mumbo-jumbo on your live input devices etc. [20:44] hendry: i write something up for you [20:46] JEEB: my mumbo jumbo is pretty standard? x11grab & alsa for sound [20:46] klaxa: that would be much appreciated [20:46] hendry, I'm just not well versed in that but it looked more or less correct [20:46] but you should understand how my example would build around your inputs :P [20:47] JEEB: i'll assume utvideo is very well suited for lossless screencasts? [20:48] added -report logs to http://r2d2.webconverger.org/2012-10-21/ if you are interested [20:48] I can do this in one step with my script like so: ./r2d2.sh -d 5 -u r2d2.webconverger.org:/srv/www/r2d2.webconverger.org -c ffvhuff ffvhuff2.webm [20:48] as a format it's nice enough, supports RGB as well as various YCbCr formats, but it might not be fast enough, but I'm not sure if the native library encoder that Daemon404 wrote is any faster :) [20:49] libx264 ultrafast lossless it is? [20:49] JEEB: I'm OK with slow encoding speed as long as results look like my desktop I'm trying to capture. [20:49] http://r2d2.webconverger.org/2012-10-21/utvideo2.webm.html is failing the "look like my desktop I'm trying to capture" test [20:49] and so is http://r2d2.webconverger.org/2012-10-21/ffvhuff2.webm.html [20:50] ... [20:50] did you actually read ANYTHING I wrote? [20:50] O.o [20:50] and did you not read why I told you reporting that one thing might be a good idea AT ALL? [20:50] if you set your output file name to be something dot webm [20:50] it will auto-set stuff to be compatible with it [20:50] that is not ut video and that is not ffvhuff [20:51] JEEB: yes, I do read but i might not understand [20:51] real ffvhuff or ut video DO NOT DECODE IN YOUR BROWSERS [20:51] JEEB: I understand you want me to report a bug about something I understand very little [20:51] webm is vp8 + vorbis [20:51] there's a reason why I set the output file name as out.avi there [20:51] meh [20:52] durandal_1707, I think everyone by now knows that, but the funny thing is that instead of erroring out on incorrect video/audio codecs (that are set explicitly) it will just switch them over to compatible ones :P [20:53] gonna fix it asap [20:53] ok, I understand `ffmpeg -vcodec ffvhuff -acodec pcm_s16le out.avi`... and that's it? maybe i'll just shut up now and wait for klaxa [20:54] hendry: what ffmpeg version you are using? [20:55] JEEB: what audio codec would you suggest for lossy audio? [20:55] hendry, yes? [20:55] durandal_1707: ffmpeg version 1.0 Copyright (c) 2000-2012 the FFmpeg developers built on Sep 29 2012 11:22:50 with gcc 4.7.1 (GCC) 20120721 (prerelease) [20:55] durandal_1707: http://r2d2.webconverger.org/2012-10-21/ffmpeg-20121021-024453.log may help [20:56] durandal_1707: i'm using Archlinux [20:56] hendry: latest ffmpeg disallows muxing other codecs for webm [20:58] klaxa, depends on what your intended use case is [20:58] hendry: what is your intended use case? [20:58] durandal_1707: so I'm not using the latest IIUC [20:58] klaxa: recording my desktop [20:58] will you record your soundcards audio or your voice or both? [20:59] klaxa: ideally to container that works with HTML5 video across browsers. I don't mind 2/3 copies if it means getting it working on IOS6 safari etc [20:59] i see [20:59] klaxa: both I guess [20:59] hendry: this is strange 1.0 should not allow it either [20:59] klaxa: "-acodec pcm_s16le -f alsa -i hw:0,0" has been fine for me [21:00] well yeah that's raw pcm [21:00] you don't want that to be sent over the internet [21:00] klaxa: ok, sure, whatever is the appropriate codec for voice really [21:00] i think with webm you're pretty much bound to vorbis and vp8 [21:01] klaxa: fine, but Safari IOS6 I assume is a different beast altogether [21:02] anyone know the variables for libvx? i for one don't, i don't use libvx [21:02] i.e. profile, preset, good crf or qp values [21:03] libvpx? [21:03] unfortunately it has no such easy things like the three first ones you mentioned :P [21:03] it does have constant quant'ish thing [21:03] but nothing like crf [21:08] D: [21:08] do the default values suffice for publishing on the web? [21:08] it has some vpre files for certain stuff, but no idea how good those in reality are [21:08] well.. hendry will find out for us :) [21:08] you are most probably best off setting a constant quant or a bit rate with libvpx [21:10] https://gist.github.com/7dcccbd86fdcce3c4ced [21:10] hendry: ^ [21:11] ah fixed small thing, second step: raw.avi -> raw.mkv [21:12] klaxa: https://gist.github.com/3924250 [21:12] klaxa: first step worked, second didn't [21:12] ah fixed small thing, second step: raw.avi -> raw.mkv [21:13] reload the paste [21:13] wait what.. [21:13] -strict -2 # didn't work either as the output suggests [21:14] there is no audio in the .mkv [21:14] klaxa: when i `mplayer raw.mkv` i hear audio [21:15] wat... no there is audio in the mkv [21:15] but ffmpeg doesn't... someone... why doesn't vorbis work? [21:15] that's beyond my knowledge... try specifying bitrate with -ab 192k or something after -acodec vorbis [21:16] hendry: do not use internal experimental aac encoder (it have limited quality) [21:16] neither internal vorbis encoder (which is just not for users) [21:17] just do not use experimental things [21:17] durandal_1707: can you suggest edits to https://gist.github.com/7dcccbd86fdcce3c4ced please? [21:18] yeah i'm actually a novice myself [21:18] i feel like i should know a lot more to be able to give advice at all [21:19] i've got to go to bed since it's 3AM here, i'll be connected though and I'm desperately looking to improve https://github.com/kaihendry/recordmydesktop2.0 [21:21] hendry: use copy for last -acodec [21:23] durandal_1707: https://gist.github.com/3924250 [21:23] i guess thre is a problem with klaxa's https://gist.github.com/7dcccbd86fdcce3c4ced step 1 [21:24] hendry: what are you doing? [21:24] >[webm @ 0x1c2a6e0] Only VP8 video and Vorbis audio are supported for WebM. [21:24] i think vorbis is mandatory for webm, no? [21:25] ffmpeg -f x11grab -s 1366x768 -r 30 -i :0.0 -f alsa -i hw:0,0 -acodec pcm_s16le -vcodec ffvhuff -acodec vorbis raw.mkv # doesn't work: Codec is experimental but experimental codecs are not enabled .... [21:25] wait... don't use -acodec twice [21:25] durandal_1707: trying to record my desktop [21:25] remove the -acodec vorbis [21:25] that's for the reencode to webm [21:25] the raw is supposed to be lossless anyways [21:26] klaxa: oops, no different if i remove 2nd -acodec vorbis [21:26] so it looks like we are stuck here or is there some intermediate stage ? [21:27] klaxa: mkv can have anything for audio [21:27] i mean hendry [21:27] yeah the initial raw should have ffvhuff video and pcm_s16le audio [21:27] then re-encode with vp8 and vorbis for webm [21:28] should use -acodec libvorbis not -acodec vorbis [21:28] hendry: you cant use webm for this [21:28] pcm_s16le is overkill, it can take lot of space, use flac [21:29] good morning/afternoon/evening everyone, im dont know why ffmpeg shows 'Unknown AudioCodec: aac' even when [ffserver -formats | grep aac] recognizes it [21:30] amendes365: you want to stream aac? [21:31] ah right... libvorbis, will flac be encoded fast enough? (it's audio so i guess so?) [21:31] durandal_1707: im trying to stream H.264 video with AAC audio in MPEGTS [21:32] klaxa: if disk space is small and disk is slow but CPU is reasonaly faste use flac [21:32] following one of the tutorials in ffmpeg.org [21:32] ffmpeg -f x11grab -s 1366x768 -r 30 -i :0.0 -f alsa -i hw:0,0 -acodec libvorbis -vcodec ffvhuff raw.mkv && ffmpeg -i raw.mkv -acodec copy -vcodec libvpx encode.webm # but my machine is too slow to play this. worried the audio is out of sync.... need slep [21:32] klaxa: if disk is big and very fast it is irrelevant [21:32] hendry, reload the gist, go to sleep, come back tomorrow with a clear mind [21:32] looks like you are a bit tired [21:33] amendes365: you need aac encoder there are many of them supported by ffmpeg [21:33] klaxa: if we use flac then we get the Only VP8 video and Vorbis audio are supported for WebM. problem [21:33] web video is crazy hard :( [21:33] you are supposed to use it the other way around [21:33] flac for the raw, vorbis for the encode [21:33] because flac is lossless [21:33] and vorbis isn't [21:34] durandal_1707: i know, ffserver -formats show raw ADTS AAC, Advanced Audio Coding, libfaac AAC but when starting ffserver shows Unknown audio codec: AAC [21:34] klaxa: so we need an intermediate step to convert the flac audio to vorbis for the final webm ? [21:34] klaxa: but it is irrelevant in this case becuase it just copy it when muxing into webm [21:34] hmm true [21:34] hendry: try flac if libvorbis encoding is soo slow [21:35] hendry: basically you recored as fast as possible to keep sync and reencode and mux after recording is finished [21:36] amendes365: formats are not codecs, use ffserver -codecs [21:36] durandal_1707: understand that [21:37] durandal_1707: guess i need to figure out the intermediate step's args after some sleep [21:39] durandal_1707: thanks. ffserver -codecs showing unrecognized option '-codecs' [21:39] klaxa: thank you for the help yesterday, trying mpegts now [21:39] i can run ffserver -codecs [21:40] the only encodable line says: DEA.L. aac AAC (Advanced Audio Coding) [21:41] ill try reinstalling ffmpeg [21:41] internal encoder is experimental [21:41] use libfdk-aac [21:41] how would you install and set up an external encoder? [21:41] install encoder program and re-compile ffmpeg with the apropriate flags? [21:41] yes [21:43] so it probably won't work with pre-compiled binaries from ? [21:44] i've visited different forums where usually is recommended to use the medibuntu repositories for this type of codec [21:44] klaxa: some external encoders are not redistributable(non-free) [21:44] which means, to use best available encoder you need to compile it [21:45] is there any other audio codec i can use instead of aac for streaming H.264 video in mpegts? [21:46] mp2? [21:46] or aac [21:46] *ac3 [21:47] durandal_1707: will try mp2. this is my second day sleeping in the lab :/ [21:47] mpegts for audio acepts: mp2,mp3,ac3 and aac [21:48] mp2 quality is not perfect [21:48] amendes365: what bitrate you are targeting? [21:49] durandal_1707: i want a very lightweight setup to see if i can get realtime audio/video streaming from a robot [21:49] robot is running on debian [21:49] as long as the audio is understandable is ok [21:50] so robot is doing encoding? [21:51] yes, but im testing a scenario between two laptops first [21:51] and using a webcam [21:52] i was using mjpeg which worked perfectly, but it has no audio [21:53] and i was considering ffmpeg to have both video and audio, otherwise i would have to do them separately [21:54] amendes365: so you can only use mpegts? [21:55] i can use anything as long as it is lightweight and can get me close to real time streaming [21:56] well mkv wiht mjpeg + any audio codec should work tooo [21:56] durandal_1707: quality is not important as long as the stream is understandable [21:56] if bitrate can be very high you can be completly lossless [21:57] ok... will try that [22:01] i don't think lossless video is important if a webcam is the source... [22:02] durandal_1707: do you have experience with live streaming matroska? i had quite the trouble and didn't succeed in the end [22:06] klaxa: what trouble? [22:06] mostly that ffserver wasn't able to mux the videos correctly (h264 video, aac audio, ass subtitles) [22:07] actually i don't remember what went wrong... [22:08] i'll try again and come back... maybe i was just too stupid the last time i tried [22:20] durandal_1707: got it working with the official theora/vorbis example, will make a conf for mkv + mjpeg, which will be the container and which one is the codec? [22:20] audio will be mp2 [22:21] test [22:35] amendes365: mkv is container [22:36] excellent [22:36] durandal1707: i made a test with output from my webcam to a .mkv video, mjpg is working [22:36] im working on the audio right now : ) [00:00] --- Sun Oct 21 2012 From burek021 at gmail.com Mon Oct 22 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Mon, 22 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121021 Message-ID: <20121022000502.7093818A01F1@apolo.teamnet.rs> [00:13] srt demuxer now outputing subrip packets, check. [01:11] bot broken? [01:11] i can reproduce mp3 seeking issue (though sample rate does not change for me) [01:12] (banned by tinyurl? :p) [01:14] we are tired of the tinyurl [01:14] because we dont twitter [01:28] saste: coverity spotted some stuff in xface [01:28] ubitux, where can i find coverity reports? [01:29] http://ubitux.fr/pub/pics/_xface-coverity.png [01:29] saste: log in http://scan5.coverity.com:8080/ [01:29] don't you have an account? [01:29] ubitux, no [01:29] ask michaelni :p [01:30] there are a few spotted issues in your filters ;) [01:31] i dont know why michaelni doesnt just make generic ffmpeg shared login [01:31] :P [01:32] then anyone who just wants to look at it for 2 mins doesnt have to register [01:32] Compn: in this special case it does not make any sese [01:32] *sense [01:32] Compn: there is some kind of comment system, with login associated with who fixed the bug etc [01:33] goodnight [01:33] if saste had an account, i could have assigned him some issues ;) [01:33] 'night saste [01:33] like anyone cares who fixes what bug on a 3rd party bug tracker [01:33] :P [01:33] well, it's pretty nice :) [01:40] ubitux: jacosub support only some styles? [01:41] what do you mean? [01:42] the decoder supports only a limited subset of styles yes [01:42] i mean using mplayer -sub [01:42] i guess it will use the internal mplayer decoder? [01:43] why why why [01:43] because mplayer has its own subtitles handling [01:43] anyway, did you really find jacosub subtitles? @_@ [01:44] i thought i got all the jacosub subtitles in the world [01:44] no, i use one from fate [01:44] ah, ok [01:44] this sucks that nothing use subtitle code [01:44] because there are < 70 jacosub files in the world [01:44] yeah, a bit [01:44] but it might not be mature enough yet [01:46] it might be time to look into this new api... [01:46] but i'd like to get done with the lavfi metadata first [04:55] @ubitux> because there are < 70 jacosub files in the world <-- there's a lot more than 70 episodes of LoGH [09:23] Plorkyeran: hehe :) [09:24] http://samples.ffmpeg.org/sub/jss/ seems i don't have them :P [09:34] michaelni: btw i was wondering, isn't the split/merge side data problematic for stream copy? [09:36] (since the demuxer would add the side data to the packet, which is directly copied without the data being split at decoding) [09:36] imho the whole merged side-data thing is problematic. What if you just use avformat and send it so something else entirely afterwards? [12:39] https://trac.videolan.org/vlc/ticket/7625 [12:39] Any idea? [12:43] nevcairiel, in that case, if it doesnt work as is you can set AVFMT_FLAG_KEEP_SIDE_DATA [12:44] also if we assume the side data did has some purpose then you have to do something with it anyway [12:44] or that "purpose" will not work anymore [12:45] well isn't it problematic if for instance you remux a rgb8 avi? [12:45] or any side data thing [12:45] i don't think that's really a good idea to have that :( [12:49] do you have some testcase that fails ? if so iam happy to fix it [12:50] j-b: ratecontrol trouble [12:50] it's weird because i'd expect the rc_eq to be always set, but seems to be NULL [12:51] j-b, is that using ffmpeg or libav ? [12:51] i cant help with libav bugs [12:51] michaelni: this is a linux VLC, so your guess is as good as mine [12:51] michaelni: i don't have any file at hand now, but any media with a rgb8 videostream (and so a palette side data) will fail, right? [12:51] cvlc is only on Linux and so, of course, I don't control the packaging... [12:52] Action: j-b welcome to my life [12:52] j-b: don't you support a --report option or like that [12:53] must be an hell to have to guess which are the involved library versions [12:54] saste: yes. As if it was not hard enough with libavcodec going from 52.25 to 54.99, with versions with and without decoders, with fork from Ubuntu-universe and Debian-Multimedia, now you have at least 2 forks... [12:56] j-b, can this issue be reproduced somehow with command line ffmpeg ? [13:00] also does vlc do something with rc_eq ? the message would indicate that its NULL which is not the default [13:09] michaelni, can you create a coverity account for me? [13:10] saste, i think i created one days ago, can you check your spam folder ? [13:15] michaelni, found none [13:18] saste, created another one [13:19] thanks [13:22] michaelni: should I expect a mail or so? [13:22] saste, yes but i do not know how quick [13:22] ok [13:23] if you got nothing till tomorrow evening then ping me and provide a different email address than the gmail one maybe [13:38] in some cases the spotted issues are bogus [13:38] what should we do in that case? [13:39] for xfaceenc the fix is trivial, but not required [13:39] and adds an unnecessary check (so it slows it down, even if in an unnoticeable manner) [13:40] [NULL @ 0x2158100] [13:41] why do we print the funny hexdigit? [13:41] is it somehow useful, or should we replace it with something more useful (e.g. the called function name?) [13:45] saste, you can mark CIDs as false positive or intentional [13:45] Action: michaelni misses cone [13:47] saste, the idea behind the address for av_log is that if you have 2 abccontext to be able to distinguish them [14:41] michaelni: cone seems to be restarted [14:47] ffmpeg.git 03Luca Barbato 0771f7b22dba60: ffv1: split decoder and encoder * 03http://tinyurl.com/9qgeb3603 [14:47] ffmpeg.git 03Michael Niedermayer 0769fd0b7adbd2: Merge commit '71f7b22dba60524b2285643ae0b49d8f64977129' * 03http://tinyurl.com/922yrqz03 [14:47] cone :) [14:48] still no plan to remove that tinyurl thing? [15:02] ffmpeg.git 03Luca Barbato 074a2a4524a3f5: ffv1: propagate errors * 03http://tinyurl.com/9zlfaln03 [15:02] ffmpeg.git 03Michael Niedermayer 0799ea47fe5a55: Merge commit '4a2a4524a3f50ed302820ba971ddd48e78c7436f' * 03http://tinyurl.com/9clddju03 [16:19] ffmpeg.git 03Luca Barbato 070f13cd318719: ffv1: update to ffv1 version 3 * 03http://tinyurl.com/8rodexv03 [16:19] ffmpeg.git 03Mans Rullgard 07ebe46b8063ae: ARM: reinstate optimised intmath.h * 03http://tinyurl.com/8ber35h03 [16:19] ffmpeg.git 03Derek Buitenhuis 07f2a7236d0c73: doc/platform: Move a caveat down to the notes section * 03http://tinyurl.com/9ef5zg803 [16:19] ffmpeg.git 03Derek Buitenhuis 072d09b36c0379: doc/platform: Add info on shared builds with MSVC * 03http://tinyurl.com/8s5z93e03 [16:19] ffmpeg.git 03Michael Niedermayer 07aa760b173590: Merge commit '2d09b36c0379fcda8f984bc8ad8816c8326fd7bd' * 03http://tinyurl.com/8fa4ort03 [16:36] $(call ALLYES, FFPROBE LAVFI_INDEV SELECT_FILTER AVCODEC MOV_DEMUXER SVQ3_DECODER ZLIB) [16:37] nice dependencies. [16:37] :) [16:47] michaelni: i'll commit soon the lavfi metadata injection with fate tests for scene & silence detection, unless you want me to wait more [17:04] ffmpeg.git 03Diego Biurrun 076cfca5b6ae5a: ffv1: Add missing #includes to header file * 03http://tinyurl.com/8kn9fhp03 [17:04] ffmpeg.git 03Diego Biurrun 07886087829157: fate: Introduce ENCMUX macro for tests that require encoders and a muxer * 03http://tinyurl.com/8sduptg03 [17:04] ffmpeg.git 03Diego Biurrun 07a7d2861d3675: svq3: K&R formatting cosmetics * 03http://tinyurl.com/8arttqk03 [17:04] ffmpeg.git 03Michael Niedermayer 070e097616865e: Merge commit 'a7d2861d36756b913e85681b86ed3385274e8ced' * 03http://tinyurl.com/92ne85o03 [17:05] https://github.com/ubitux/FFmpeg/compare/master...lavfi-metadata [17:15] ffmpeg.git 03Diego Biurrun 07af1ede069cb6: svq3: cosmetics: Drop useless parentheses * 03http://tinyurl.com/9njxlrz03 [17:15] ffmpeg.git 03Diego Biurrun 072e0c410485ef: fate: adpcm: Add dependencies * 03http://tinyurl.com/8dsklj803 [17:15] ffmpeg.git 03Diego Biurrun 07620345f930d1: fate: adpcm: cosmetics: Sort test entries * 03http://tinyurl.com/995aqz803 [17:15] ffmpeg.git 03Michael Niedermayer 07d7b8a9a589d9: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/9ptywg503 [17:15] meh i need to rebase again [17:16] ubitux, make sure you dont access the new fields directly from outside avcodec [17:16] i think i saw at least one acces from lavfi [17:17] huh, really? [17:17] AVFrame.metadata i think [17:18] oh! [17:18] ok, will fix [17:18] thank you [17:21] fixed in avfilter_copy_frame_props() [17:49] wtf i just -c copy and the h264 input got transformed into rawvideo.. [17:50] oh my bad. [17:56] ffmpeg.git 03Cl?ment BSsch 076fb2fd895e85: lavc: add lavfi metadata support. * 03http://tinyurl.com/8bosxvf03 [17:56] ffmpeg.git 03Cl?ment BSsch 07de23953de2d8: lavfi/select: store scene score in buf ref metadata. * 03http://tinyurl.com/9b3v8l203 [17:56] ffmpeg.git 03Cl?ment BSsch 07fbedce6b8512: lavfi/silencedetect: export silence info to metadata. * 03http://tinyurl.com/9qwcoex03 [18:08] ffmpeg.git 03Michael Niedermayer 0738797dc31b65: compat/getopt: add {} to complex ifs * 03http://tinyurl.com/94ugnq903 [18:08] ffmpeg.git 03Michael Niedermayer 074ce9312d7617: libavcodec/cook: add {} to complex ifs * 03http://tinyurl.com/9ygau6d03 [18:08] ffmpeg.git 03Michael Niedermayer 07f72b735d41f9: libavcodec/vorbisenc: add {} to complex ifs * 03http://tinyurl.com/95pqdv803 [18:09] ffmpeg.git 03Michael Niedermayer 07217193facd3e: bavformat/mov: add {} to complex ifs * 03http://tinyurl.com/8lkucus03 [18:09] ffmpeg.git 03Michael Niedermayer 07fefe9bd78266: libavformat/movenc: add {} to complex ifs * 03http://tinyurl.com/9j2bq9903 [18:09] ffmpeg.git 03Michael Niedermayer 07b8a64d69a927: libavformat/mpegvideodec: add {} to complex ifs * 03http://tinyurl.com/8wbjlqw03 [18:09] ffmpeg.git 03Michael Niedermayer 07c2c066784f19: libavformat/mxfenc: add {} to complex ifs * 03http://tinyurl.com/94gbsnq03 [18:09] ffmpeg.git 03Michael Niedermayer 0740ceb6d49f19: libavformat/oggenc: add {} to complex ifs * 03http://tinyurl.com/8rsyoxt03 [18:12] michaelni: is there some warnings with some compiler? [18:26] ffmpeg.git 03Marton Balint 07d6e95669496a: h264: add support for AFD detection * 03http://tinyurl.com/8scfs9h03 [18:31] ubitux, i dont know but that kind of code is just fragile, the develoer sees the association based on indention but the compiler doesnt care about the indention at all [18:33] its a recipe for bugs [18:33] of course [18:33] i was just wondering how you spotted them all [18:37] i dont know if i spotted all, it was just an ugly grep command that i used [18:42] michaelni : you should put it in patcheck :P [18:42] if possible [18:42] and all other sneaky code things [18:42] does libav use patcheck ? [18:43] patcheck works on diffs, i greped *.c so its not going to work [18:44] for new code i mean [18:44] but ok [18:45] michaelni, ping on "propagate or return meaningful error codes in avcodec_open2()" [18:54] really, we should properly document all the quantization stuff (a dedicated libavcodec manual would fit) [19:02] ffmpeg.git 03Marton Balint 07da569556217f: ffplay: use framedrop by default when sync is not done to video * 03http://tinyurl.com/9ukuag203 [19:02] ffmpeg.git 03Marton Balint 07eaa91ed863a7: ffplay: fix nosync threshold check in synchronize_audio * 03http://tinyurl.com/8dyzujs03 [19:02] ffmpeg.git 03Marton Balint 0766bb5b1bc9b1: ffplay: initialize audio and video pts drift * 03http://tinyurl.com/8ebxfra03 [19:02] ffmpeg.git 03Marton Balint 0777bd595ad2ff: ffplay: fix external time sync mode * 03http://tinyurl.com/8wb66wm03 [19:02] ffmpeg.git 03Marton Balint 072a4c7e6540c4: ffplay: add serial field to PacketQueue entry and populate it * 03http://tinyurl.com/96rp3a503 [19:02] ffmpeg.git 03Marton Balint 07b2a8850969b8: ffplay: only check external clock if current frame serial matches the displayed frame serial * 03http://tinyurl.com/8smuwbz03 [19:02] ffmpeg.git 03Marton Balint 07fca16a15712b: ffplay: add get_master_sync_type function * 03Error03 [19:02] ffmpeg.git 03Marton Balint 07d30c69251f04: ffplay: use get_master_sync_type where necessary * 03Error03 [19:02] ffmpeg.git 03Marton Balint 073166a6fc3797: ffplay: if there is no audio stream, use external clock by default * 03Error03 [19:02] ffmpeg.git 03Michael Niedermayer 07747a00b688d7: Merge remote-tracking branch 'cus/stable' * 03Error03 [19:20] ubitux: http://fate.ffmpeg.org/report.cgi?time=20121021160939&slot=x86_32-msvc10-windows-native [19:20] damned windows path [19:21] but it raises an unconfortable issue [19:24] oh :( [19:24] saste: any idea how i should fix it? [19:24] automatic escaping [19:24] that's why i implemented tools/ffescape [19:25] but then we need to wrap the test in a script, which performs automatic escaping [19:25] that's why the graph syntax sucks hard, compared to avisynth [19:25] since requires up to two levels of escaping, three considering shell escaping [19:25] most users get it wrong [19:26] ubitux: also BSD is failing [19:27] http://fate.ffmpeg.org/report.cgi?time=20121021163915&slot=x86_64-freebsd8.2-clang2.8 [19:27] though I have no idea why [19:27] -video|1|14100|23.500000|14100|23.500000|1|0.001667|2006431|320|240|rgb24|1:1|I|0|0|0|0|0|0|0.850000 +video|1|14100|23.500000|14100|23.500000|1|0.001667|2006431|320|240|rgb24|1:1|I|0|0|0|0|0|0|0.840000 [19:27] looks like yet another floating thing [19:28] what is the last value? [19:28] the scene score [19:28] ah ok [19:29] saste: so how/where would you implement the automatic escaping? [19:30] ubitux, I have a patch pending [19:30] ah, great :) [19:30] i want to fix a few things before committing i [19:30] *it [19:31] but then deploying it in the test system may be complicated [19:31] ffmpeg.git 03Carl Eugen Hoyos 07fa190b3cd736: Improve MPEG-PS-in-MOV detection. * 03http://tinyurl.com/8hqxwdl03 [19:32] i wonder how i'm supposed to fix the freebsd one :( [19:32] avoiding float computations? [19:33] i want the value in the [0;1] range [19:34] ubitux: you may add an option where you specify the value in integer units [19:35] and return 0-100 integer values [19:35] maybe someone has a better idea (this may complicate the internal code) [19:47] ubitux, distclean + fate-filter-metadata-scenedetect fails with no ffprobe having being build [19:47] the 0.84/0.85 can be reproduced with --disable-asm --disable-yasm on linux [19:47] make sure swscale gets its bitexact flags! [19:48] i added -bitexact [19:48] mmh CONFIG_FFPROBE isn't enough, i will add the ffprobe dep [19:48] thanks [19:48] i'll fix all of this asap [19:48] try -sws_flags +accurate_rnd+bitexact if its not there yet [19:50] mmh ok [20:06] ffmpeg.git 03Cl?ment BSsch 07c0d56bf8a4c6: fate: fix filter metadata dependency to ffprobe. * 03http://tinyurl.com/8f39mdf03 [20:13] the sws flags don't seem to help :/ [20:17] ffmpeg.git 03Cl?ment BSsch 07e168165489fe: doc/muxers: document mov faststart option. * 03http://tinyurl.com/9okq7vw03 [20:51] ubitux, is the input to the filter the same and the output score differs or the input isnt the same in the first place ? [20:52] i'm going to check in a moment [20:53] i was just fooling around with something right now :p [20:53] saste: just added the sliding mode to showspectrum :P [20:53] nice [20:56] just sent [20:56] now serious matter, fate. [21:06] ffmpeg.git 03Michael Niedermayer 07571309181859: mpegaudiodec: Fix buffer handling on random access * 03http://tinyurl.com/8gwvuxw03 [21:12] ffmpeg.git 03Stefano Sabatini 078c2dbc380500: lavc/utils: provide more feedback in case of experimental codec * 03http://tinyurl.com/8qrq7vo03 [21:13] michaelni: the input isn't the same in the first place indeed, maybe the flag are not properly transmited to swscale [21:13] most likely a stupid mistake from me [21:14] Action: Compn waits for autocrop filter [21:14] :P [21:15] Action: Compn afk [21:18] michaelni: any simple check to see if the flags reach sws? [21:23] ubitux, print the flags from sws_scale() [21:34] ok they don't seem to be transmited [21:38] scale_args=[0:0:(null)] [21:38] mmh. [21:39] wasn't this fixed? [21:39] ubitux, is it a problem? [21:39] ok it's done in ffmpeg & ffplay. [21:39] but not ffprobe [21:39] saste: i see at least two problems :p [21:40] i mean with your fate test [21:40] ah yes it's a problem [21:40] the bitexact flags for sws are not honored [21:40] and so the input differs [21:40] (since select is converting to rgb when using scene detection) [21:41] ok i'll commit them soon [21:41] i just want to test them another time [21:41] ah you have a fix for that? i wasn't inventing thing? [21:50] well i guess i'll wait for your patches then [21:55] ubitux: 6fb2fd895e858ab93f46e656a322778ee181c307 [21:55] wow, longest commit message in FFmpeg history [21:55] :) [21:56] btw fate is failing on your last test [21:56] you mean the scene detect? [21:56] yes [21:56] yes that's what we were talking about [21:57] amovie=/amrwb/seed-12k65.awb,silencedetect=d=.1: No such file or directory [21:57] huh? [21:58] i believe it is because I'm not using SAMPLES [21:58] ah i use --samples at configure time [21:58] mmh i guess i'll need to add them to some kind of samples dep [22:00] anyway, for the scene detect, there are multiple things to do [22:00] first make sure ffprobe is transmitting the sws flags, and then add the bitexact flags to the FILTER_METADATA_COMMAND [22:00] but you told me you had some pending patches for this [22:01] about the SAMPLES issue, i don't know how i'm supposed to fix that properly.. [22:04] saste: the sws flags thing looks a bit messy and inconsistent [22:05] it's only set in ffplay and ffmpeg, but not in ffprobe; and maybe it should be set in the lavfi device? [22:12] ubitux, yes [22:12] btw if you cant fix this, the not working test should be disabled [22:13] didable broken things to make stuff appear to pass isnt a very good idea... [22:13] disabling* [22:14] michaelni: i'll fix as soon as saste push what he was talking about [22:14] it can be fixed without much trouble i believe [22:15] we just need to pass the bitexact flags properly to sws :p [22:20] ubitux: can you fix the samples thing in the meanwhile [22:20] yes, trying to figure out how to :p [22:23] Daemon404, i think we speak about different things [22:26] if code is commited that breaks something it should be fixed or reverted, ubitux is working on that. But if noone had time, the code that broke something should be reverted [22:26] isnt it new code that just doesnt work everywhere? [22:26] its not a regression, just open TODOs, basically =) [22:27] michaelni: what's my deadline for fixing fate? [22:27] its causing a regression for "make fate" [22:27] a regression is existing tests breaking, not new tests not passing, imho [22:28] anyhow fixes seem to be on their way [22:28] ubitux, no idea, it doesnt matter, just if you give up / work on something else please disable it first [22:28] ok [22:28] i can go to sleep tonight without fixing it? :) [22:29] if you continue dreaming about how to solve it :) [22:29] nevcairiel, its a regression in "make fate" [22:29] it worked before [22:30] of course ;) [22:31] "make fate" is used to test if code is ok, for example before push, if it fails developers work is hindered [22:33] thats IMHO not so much about "hiding a problem" more "not having all alarm bells ringing for days while its being worked on" [22:35] and these alarm bells could hide other new problems or regressions [22:37] michaelni, i thought you were talkign about a fate test [22:37] by bad [22:37] considering its only broken on one os/toolchain combination, i think we'll survive a day until its fixed [22:39] i think its not even the change itself, just the way the test is setup with the movie source, yes? [22:40] i think its broken on all non x86 platforms [22:41] fate looks rather normal [22:42] except the fail on native windows builds which is new [22:45] many didnt test the problematic revission yet i think [22:45] ffmpeg.git 03Stefano Sabatini 072969abd908f1: lavfi/graphparser: fix parsing error in case of NULL sws_opts addition * 03http://tinyurl.com/8tk9gkf03 [22:45] ffmpeg.git 03Stefano Sabatini 078f37a1e8dc93: lavfi/avfiltergraph: avoid to print "(null)" in the scale args * 03http://tinyurl.com/8otm4q403 [22:46] didnt it just uncover a parsing problem when a string parameter includes a :, which happens in windows path strings? [22:46] saste: thx! [22:46] are we talking about the same problem? :P [22:46] nevcairiel, you need to escape it [22:46] the problem is doing it in an automated fashion [22:47] thats rather unintuitive though :p [22:47] how do you escape in that place even? normal backslash? [22:56] ok almost fixed it [23:00] ok, fixed [23:02] i wonder how safe this is though [23:03] saste: http://b.pkh.me/0001-lavd-lavfi-honor-sws-flags.patch ? i guess that needs to be made safer, right? [23:04] (with some #if CONFIG_...) [23:10] ubitux, you are accessing global variables that reside in ff* user tools from the libs [23:11] that is definitly not safe [23:12] mmh it's in the cmdutils, missed that [23:14] i guess i'll have to do that in ffprobe [23:15] but i don't have access to the filtergraph.. [23:16] quick solution is to add a AVOption to lavfi that allows setting the sws flags [23:16] libavdevice/lavfi.c that is [23:16] what about getting a new sws context in the lavfi device? [23:17] mmh wait i'll need the link with the options [23:19] another way is to make avfilter_graph_parse() somehow set the "global" flags from the graph string [23:22] my brain is dead, i'll see that tomorrow [23:22] ok no hurry, maybe ill look at it while you sleep [23:24] you're a real guardian ;) [23:24] 'night :) [23:25] this is a test case btw: http://b.pkh.me/0002-fate-add-bitexact-flags-to-filter-metadata-test-comm.patch [23:25] and i have another patch to add the scale filter dep to scene detect as well [23:26] anyway, cya [23:26] good night ubitux [23:31] i thought boxx codec was dead, whats it doing with 10bit ? [23:41] since when "being experimental" is an error? [23:41] what about AVERROR_PINK? [00:00] --- Mon Oct 22 2012 From burek021 at gmail.com Mon Oct 22 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Mon, 22 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121021 Message-ID: <20121022000501.5BB2718A01F0@apolo.teamnet.rs> [02:39] what would be the paramaters to take png's and turn them into wmv? [02:40] i tried ffmpeg -i "C:\Users\Zelo\Desktop\earth\frames\*.png" -vcodec wmv2 -y C:\Users\Zelo\Desktop\earth.wmv [07:25] I'm trying to figure out how to include audio recording in screencasts with ffmpeg. [07:26] so far I'm at ffmpeg -f x11grab -r 25 -s 1024x768 -i :0.0 -vcodec huffyuv -sameq screencast.avi now just trying to figure out how to get it to record my voice mic and whatever audio is on the pc [07:47] Ok I almost got it figured out. I've got x11grab working and i-pulse. Might someone be able to help me figure out how to capture analog audio as well? [07:49] robertzaccour: https://github.com/kaihendry/recordmydesktop2.0/blob/master/r2d2.sh#L26 [07:50] robertzaccour: i use alsa, I don't understand how/why pulse is used by so many ppl [07:51] hendry, pulse works best for me, especially when using guvcview [07:52] hendry, [NULL @ 0x265e440] Requested output format 'x11grab' is not a suitable output format [07:52] :0.0: Invalid argument [07:58] Could someone help me figure out how to record from my speakers? I already got my screen and mic with ffmpeg -f alsa -i pulse -f x11grab -r 25 -s 1440x900 -i :0.0 -acodec pcm_s16le -vcodec huffyuv -sameq Screencast.avi just gotta get my speakers in there and I'm set. [08:03] how may I use something like Deshaker with ffmpeg? > http://www.guthspot.se/video/deshaker.htm [08:11] Is there anything in my question I'm leaving out that might be helpful in figuring out? [08:12] robertzaccour: can't you control that sort of mixing from alsamixer ? [08:12] robertzaccour: sorry, I would love to help, but I am an newbie [08:13] robertzaccour: when you figure out how to do it, I would like to incorporate it to https://github.com/kaihendry/recordmydesktop2.0 [08:45] can ffmpeg do this with deshake? > http://isenmann.wordpress.com/2011/03/22/deshaking-videos-with-linux/ [08:52] Is it even possible to record from the mic and the speakers at the same time? Could someone please help me with this? [08:53] robertzaccour: /j #alsa [08:56] fling, this is the ffmpeg channel [09:13] is there a sane way to figure out the duration of a .webm file ? [09:20] hendry, right-click, select properties, look in the audio/video tab [09:20] or double click and see how long it is [09:27] dadadew, you familiar with ffmpeg commands? [09:28] absolutely [09:30] robertzaccour: i'm on linux [09:31] dadadew, here's what I got so far so ffmpeg -f alsa -ac 2 -i plughw:1,0 -f x11grab -r 15 -s 1360x768 -i :0.0 -acodec pcm_s16le -vcodec libx264 -preset ultrafast -threads 0 output.mkv now I just need to figure out how to get my analog speakers in on the recording [09:37] fling: we have a deshake filter, but it might be that effective [09:38] fling: your post reminds me the thing from google [09:38] there is an opened issue to improve the deshake filter, if you want to contribute :) [09:39] hendry: use ffprobe [09:39] with -show_format, and you might be interested in -of as well [09:43] thank you ubitux [09:44] anyone here familiar with ffmpeg commands? [09:44] you likely need to configure it with alsa [09:44] ya think so? [09:45] ubitux: ffprobe man page is mental. what does -of do [09:45] dunno, but maybe something like making your speakers in recording mode [09:45] hendry: try -of json, or -of flat=s=_ [09:45] you'll get the point [09:48] ubitux: got it. that ffprobe man page needs a sanity check nontheless ;) [09:49] really? [09:49] http://ffmpeg.org/ffprobe.html#Writers [09:49] and the -of description looks pretty obvious [09:50] you even have an example :) [09:50] ubitux, I'm tryin to figure out how to get ffmpeg to record the speakers directly, just gotta figure out the input and where to put it [09:51] robertzaccour: i understand, but to my knowledge, everytime i had to record from the speaker, it needed some configuration on alsa side as well [09:51] and everytime, it was a pain anyway [09:51] ubitux, you mean with alsamixer? [09:51] for example [09:53] ubitux: http://r2d2.webconverger.org/2012-10-21/ubitux.html [09:54] haha [09:55] now you could use -of json, and print the info in javascript [09:56] ubitux: yes, bit too tired to that atm, but I understand [09:56] and btw, since you're using shell scripting, -of flat=s=_ makes possible to eval directly the variables [09:57] ubitux: i'll make a note of that [09:57] ubitux: i wish the output wasn't so lossy, can't clearly see my desktop [09:57] so $(eval ffprobe -v 0 -of flat=s=_ -show_format ...) and then you can access for instance $format_duration [09:59] looks readable to me [09:59] but you can play with -b:v to get a better quality [10:00] ubitux: on this line ? https://github.com/kaihendry/recordmydesktop2.0/blob/master/r2d2.sh#L34 [10:00] ubitux: what's an example value ? [10:01] i'd say on line 36 [10:01] i don't understand why you have so much ffmpeg cmd lines [10:01] why is audio & video separated? [10:01] example value? like -b:v 2000k ? [10:03] first i'm recording in flac because i'm told that's fast [10:03] ubitux: then i need to convert to vorbis as webm only accepts that as an input [10:04] ubitux: so there is 3 steps :( [10:04] you can replace the two last one with a single step afaict [10:05] like... ffmpeg -i $temp $out [10:05] codecs will be autopicked [10:05] you could explicit them with ffmpeg -i $temp -c:a libvorbis -c:v libvpx $out if you prefer [10:06] then play with the quality, like ffmpeg -i $temp -c:a libvorbis -q:a 7 -c:v libvpx -b:v 3000k $out [10:06] not sure what 2000k and 3000k mean [10:06] bitrate [10:06] and -q:a 7 ? [10:07] audio quality [10:07] variable bitrate, likely [10:07] ok, i'll play. spent many hours on this now. :} my exp is that the audio had to be in vorbis in advance [10:08] why? [10:08] ubitux: isn't there just a *lossless* codec, or is webm inheriently lossy? [10:08] not sure if vpx has a lossless mode [10:08] and afaik webm only allows vp8 [10:14] where can i figure out if vp8 has a lossless mode? [10:44] ubitux: two lines works btw! http://r2d2.webconverger.org/2012-10-21/ubitux.html [10:46] i'm on the internet @_@ [10:49] ubitux: file sizes seem a bit big. i'll have to play with -b:v [10:49] yep [10:52] ubitux: any other possibilities to improve it? I need a H264 version too, need to figure out that line from $temp [10:55] in a mp4? afaict libx264 has better defaults anyway [10:55] i'd say to just play with -crf option [11:05] ubitux: https://gist.github.com/3926442 ffmpeg -i /tmp/./r2d2.sh.8Sp5.mkv 2012-10-21/forIOSsafari.mp4 # simply doesn't work I guess I need to manually tell it to convert ? [11:06] [NULL @ 0x19efec0] Codec is experimental but experimental codecs are not enabled, try -strict -2 [11:06] so add -strict -2, or use an external aac encoder [11:06] like fdk-aac, libfaac, etc [11:08] ubitux: ffmpeg -i /tmp/./r2d2.sh.8Sp5.mkv -crf 0 -c:v libx264 -preset slow -c:a libmp3lame 2012-10-21/forIOSsafari.mp4 # works ! [11:08] yes, you can encode in mp3 as well [11:08] not sure what "-preset slow -c:a libmp3lame 2012-10-21/forIOSsafari.mp4 # works ! [11:08] not sure what "-preset slow" does [11:09] it makes it slow? :) [11:09] anyway, you'll likely want -q:a 3 or something (or reduce even more to get a better quality) [11:09] crf 0 is a bit low OTOH [11:10] -crf 21 should be more than enough [11:10] 15-25 should be the range you should play with [11:10] look at libx264 for the appropriate settings [11:11] quantizer scale is 0-51; where 0 is lossless [11:11] i want lossless don't I? [11:11] Action: hendry reads http://ffmpeg.org/trac/ffmpeg/wiki/x264EncodingGuide [11:12] ah if you want some lossless... [11:12] i'd use another codec for lossless video though [11:12] i want the desktop I'm capturing to be sharp [11:13] ubitux: surely with mp4 I'm limited to h264 ? require playing on IOS6's safari [11:13] i guess so [11:14] mp4 seems 1/3 of the size of the webm regardless! [11:14] though you can likely put another codec if you want, indeed the issue is your destination device [11:14] yes, you have to play more with the libvpx bitrate and maybe other settings [11:21] damn, doesn't work on my iphone http://r2d2.webconverger.org/2012-10-21/test.html [11:31] hendry: it's possible some apple device are limited to a few h264 levels [11:32] is crf the level? [11:33] no [11:34] hmm, doesn't work in Android either [11:34] https://en.wikipedia.org/wiki/H.264/MPEG-4_AVC#Levels [11:34] look at the levels supported by your devices [11:35] and then use -level [11:35] oh no, there is depressing http://stackoverflow.com/questions/9144574/no-matter-what-i-do-i-cannot-get-my-mp4-video-to-play-on-safari-ipad [11:35] :) [11:55] none of these presets work for me -vpre, maybe there are in a seperate ffmpeg distro? [11:56] try -preset:v [11:57] these options are a bit tricky [12:08] which filters do I need in order to be able to capture audio and video to file? [12:09] I disabled all filters when I compiled ffmpeg and I'm getting: "'aresample' filter not present, cannot convert audio formats." [12:10] anything else? ...before I just go and do the possible long compile-test-repeat cicle :) [12:17] can't get audio working on safari ubitux, video works though [12:17] this (flac -> aac) is to blame [12:18] try another aac encoder [12:18] Element9: are you using ffmpeg 1.0? [12:18] ubitux: yes [12:18] Error while opening encoder for output stream #0:1 - maybe incorrect parameters such as bit_rate, rate, width or height [12:18] ubitux: should i be using 1.0, or is it too early? [12:19] i fell like i just need the right options [12:19] why did you disable filters? [12:19] it's likely you'll need the resample and scale filters for most of the op [12:19] ubitux: what other aac encoder ? [12:19] ubitux: to get the smallest executable possible [12:19] hendry: fdk-aac, libfaac, etc [12:20] Unknown encoder 'fdk-aac' [12:21] ubitux: oh, it's not the released 1.0, it's clone from git [12:21] it's fine :p [12:21] hendry: it's an external one, you need to link it with [12:21] Element9: 1.0 or git/master is fine [12:22] -c:a libfaac # failes too :/ [12:25] hendry: same problem. [12:28] ubitux: going to try rebuild ffmpeg in desperation https://bugs.archlinux.org/task/27465 [12:29] --enable-libopencore_amrnb --enable-libopencore_amrwb [12:29] isn't this ok? [12:29] i though it was yet another aac encoder [12:34] ubitux: yes... it works with re-compiled ffmpeg! http://r2d2.webconverger.org/2012-10-21/baseline.html [12:35] :) [12:44] hello, i need help :?( [12:45] if i try to use the "configure" command to build ffmpeg gcc says that he dont know what this command means, can someone help me maybe? [12:47] CryingRiver: tryed "./configure" ? [12:47] yes [12:47] s/tryed/tried [12:47] which OS? [12:47] windows [12:47] 7 [12:48] and which compiler do you plan to use? [12:49] that the output: http://pastebin.com/N9nPgc53 [12:49] configure is a bash script [12:49] for linux [12:50] here are the guids for compiling ffpmeg: http://ffmpeg.org/trac/ffmpeg/wiki/CompilationGuide [12:51] I'm compiling it in linux for usage on windows - haven't tried to compile it on windows [12:53] mhh... my real problem is that I need to build ffmpeg with a special gcc version, and I dont know how to do it... can I use mingw with a special gcc version? [12:56] CryingRiver: did you read the compilation guide? [12:57] yes, but i got evrytime a "undefined reference to "av_register_all" error [12:58] if i build it with mingw [12:58] and use this build in my application [12:59] and everything works if you use prebuilt one? [13:00] no errors while compiling with mingw? [13:03] i dont tried a prebuilt one [13:06] Element9: no, it dont works, same error [13:06] so it's not due to the custom build then :) [13:07] yes, but whats then the source of this error... :( [13:11] you're not linking the libav dlls to your project [13:13] i link the .a files, how can i link the .dlls in the makefile? [13:44] ubitux: do you know how to get "scrubbing" or seeking working in H264? [13:46] scrubbing? [13:46] seeking should just work [13:53] ubitux: it doesn't :/ [13:53] you can't seek in the file with ffplay? [13:53] ubitux: when i move the cursor on safari's media player thingy in Youtube, it says "hi speed scrubbing" [13:54] no idea [14:06] i dotnt get it, i install ffmpeg on windows on cygwin without using mingw [14:06] i cant remember how and i need to update it for youtube-dl and other things to work [14:06] i googled like a vicous man but no go [14:18] i also didnt have yasj [14:18] yasm [14:45] i know, i must delete the older ffmpeg version [14:45] before i make install the new one [14:45] for some reason it's in use [14:45] cant find it anywhere [14:56] I'm trying to find a input method I found awhile back this year. I'm trying to convert a multi chapter DVD into one MP4. The method I saw was not a cat command and it was not http://ffmpeg.org/ffmpeg.html#join [14:56] is vorbis encoder now marked as experimental and it previously wasn't? [14:56] I cannot use a cat command because I'm already using the std input [14:57] Element9: i think you want libvorbis to use the external encoder, there is an internal one which is marked as experimental [14:57] klaxa: thanks. that's what I suspected [15:00] https://evilshit.wordpress.com/2012/09/13/how-to-combine-multiple-files-with-ffmpeg/ [15:00] that must have been it [15:04] yup:D [15:06] ubitux: ok :] [15:17] every time I stop the capature, windows thinks ffmpeg crashed and says "ffmpeg.exe has stopped working" [15:18] is that a known problem? I haven't had that problem before [15:52] ubitux: the deshaker used for this video looks very good > http://www.youtube.com/watch?v=UBeBHshMsWI [15:56] good day eveyone, can anyone help me figure out why I keep getting lip sync issues after 8+ hours of live transcoding? [16:01] oke doke [16:02] fling: great :) [16:07] fflogger: http://pastebin.com/ebjBAjwu [16:17] Sashmo: >-strict experimental [...] -acodec aac [16:17] i think that's not optimal, i was told to install fdk-aac if i wanted aac [16:18] klaxa: ? yes? [16:18] you have to recomple from source though, that might fix it [16:18] i'm no expert though... i can't guarantee anything [16:22] hmmm interesting [16:22] I'll look into it. [16:23] can anyone else recommend the same? [16:23] fdk is the best available AAC encoder atm, but just non-optimal AAC encoder shouldn't create a/v desynch [16:24] does anyeont think it has something to do with timestamps? [16:26] well, yes if the input timestamps are borked then yes, but I would check if you are always fast enough first, although the fps=26 does show that you are indeed over the seemingly 25fps input stream [16:27] also I see you get decoding errors, do those just happen at the beginning and never after that? [16:28] JEEB: yeah those are just at the begining [16:28] JEEB: how can I check if the timestamps are borked up [16:29] no idea [16:29] also, can I slap you for using manual settings with libx264? [16:29] I'd really like you to slap you [16:30] -preset was made for a reason [16:30] (and presets within x264) [16:30] slap me witha trout please [16:30] http://mewiki.project357.com/wiki/X264_Settings#preset [16:30] you set those with -preset [16:30] (and then set profile/level) [16:33] JEEB: but is that really effecting my audio issue? [16:42] Sashmo, nah -- I am just noting other things in that command line of yours [16:42] cool, thanks [16:42] also if you're streaming you probably want to use maxrate and bufsize [16:43] maxrate being the maximum rate allowed (aka the minimum speed that lets a person watch the stream), and bufsize being the buffer over which that maximum rate is calculated [16:43] JEEB: I'll play with it [16:43] JEEB: any idea why VLC reports back that the stream is 50FPs? instead of 25? [16:43] but neither might be related to the a/v desynch. Which stream? [16:46] that same one [16:46] any of my streams really [16:46] all report back double [16:52] Sashmo, I meant input or output :P [16:53] JEEB: sorry I dont udnerstand? [16:53] your mpeg-ts input or the rtmp stream you create [16:54] my source is mpegg-ts h2.64 25fps, and ffmpeg reports that its encoding at 26fps, but vlc reports it as 50fps [16:56] well, yadif shouldn't at least in your case be bobbing [16:56] in any case, not sure where the a/v desynch comes up, other than possible hick-ups in encoding [17:00] JEEB: forget about the dsync termporaily, why am I getting this dumb 50fps on vlc? [17:01] no idea :P [17:01] VLC can fail, and something somewhere else could be failing as well [17:03] all the players show it [17:07] hi is it normal that i get a 5 sec. lag between my webcam and the stream i get from ffserver [17:08] or could that be buffering? [17:15] jelly1: are you using HLS? [17:15] Sashmo: what's HLS? [17:15] Sashmo: i could paste my ffserver.conf [17:16] i want to stream only video from my webcam to a website and use the other source for opencv [17:16] jelly1: sorry cant help you there [17:16] Sashmo: ok :( [17:16] i'll continue searching [17:17] Sashmo: could it be the convertion? [17:17] from raw to in my case webm [17:17] jelly1: I think the lowest I have ever seen it is down to 2 seconds [17:17] aha [17:18] Sashmo: or does ffserver itself buffers, and then cpu speed doesnt matter for encoding? [17:26] jelly1: no idea, but I recall if you change the keyframe interval that you can get it to load faster, think of it like this, if you set a key frame interval of 60 frames, than the decoder will wait for frame 60 to output the video, therefor, 60 frames is 2 seconds, if you are encoding at 30fps, anyone else correct me here, I could be wrong, but this is what I was told before [17:27] aha [17:27] Sashmo: yes logical [17:27] jelly1: what are you setting it to now? [17:27] Sashmo: FpS> [17:27] *fps? [17:27] no key frame [17:28] and sure fps [17:28] VideoFrameRate 24 [17:28] and your keyframe? [17:28] isn't set [17:29] hmm I dont know what the defalut is, try adding it [17:29] keyint=48 [17:29] should be for 2 seconds [17:29] if your fps is 24 [17:29] why 24? [17:30] Sashmo: oh i fetched this sample from a website [17:30] try this, set your fps to 15, and see if that makes it any quicker [17:30] dont add any ket int [17:30] for a test [17:30] ok [17:30] pastebin your command [17:30] let me check it [17:30] FYI, im no expert here [17:30] just helping [17:30] ffmpeg -f video4linux2 -i /dev/video0 http://localhost:8090/webcam.ffm [17:31] hmm i should take notes [17:31] just log your irc chat, best notes ever! [17:31] Sashmo: nah [17:31] Sashmo: for the delay :p [17:31] use pastebin.com [17:31] sure [17:31] paste the full command and output [17:32] and the ffm config [17:32] Sashmo: http://sprunge.us/ACLC ffserver.conf and the command is ffmpeg -f video4linux2 -i /dev/video0 http://localhost:8090/webcam.ffm [17:33] I've never used the ffm before, so I dont know what changes work here [17:33] straight command line is what I have experiance with [17:33] sorry [17:34] aha [17:34] I guess its all the same though [17:34] there isn't much in the ffm [17:34] I dont see much of a difference [17:34] well your config is there [17:34] yeah ;) [17:34] instead of the command line [17:34] so it would be the same [17:35] hmm not sure what my stream plays now [17:35] what are you playing it with? [17:35] from the web page? [17:35] Sashmo: oh its just my webcam [17:35] your probably using swf [17:35] now i will use fplay [17:35] no I mean [17:35] Sashmo: no webm [17:35] ah ok [17:36] sorry dude, dont really know how to help you further [17:36] ask the others [17:36] ok [17:36] Im definitely not the guy to point you in the right direction, at least with this type of setup [17:36] ok [17:36] about 7 sec. with swf [17:37] Sashmo: hmm still lags about the same with VideoFrameRate 15 it seems [17:37] try the keyint command [17:37] ok [17:38] Sashmo: which value? [17:38] keyint [17:38] add that [17:38] and make it 1 [17:39] see what happens [17:39] ok [17:39] [flv @ 0x7f39a0009860] warning: first frame is no keyframe [17:39] hmm that's what ffplay also says :) [17:39] Sashmo: looks the same [17:40] Sashmo: maybe reading this should shed some light http://ffmpeg.org/pipermail/ffserver-user/2011-March/000006.html [17:42] Sashmo: googling gives me some more hits [17:42] will study them later [17:42] Sashmo: thanks btw! [17:50] jelly1: np [18:01] <_stclaws> I am republishing a live rtmp stream from a fms server. Works fine, but the incoming stream sometimes switches sources (we made it like that) which makes ffmpeg think it ends. Is there a way to make ffmpeg just keep on listening to incoming stream no matter what? [18:10] I want to edit start and stop time of some videos. It goes well if I don't use the "-t" option but when I do I get this error: http://pastebin.com/7Tq6KS7W [18:32] <_stclaws> I saw there are parameters to the rtmp protocol in ffmpeg, but how do I use them? For instance, if I want to do "ffmpeg -i rtmp//server/path/stream" and use the parameter "listen", what is the syntax? [19:42] I've enbled wmv2 encoder when calling configure (--enable-encoder=wmv2), but resulting ffmpeg don't have wmv2 encoding support. is there anything else that I should enable? something that wmv2 depends on? [20:03] Element9: can you check config.h , look for the line with CONFIG_WMV2_ENCODER [20:04] iive: I've run configure again with some other options and have a different problem now. I'll check that once I get it built again [20:04] iive: I take it that config.h is generated during ./configure? [20:05] yes, config.h config.mak and config.asm I think. there must be a config.log too, in case you need more info about failed test. [20:06] have in mind that the order of configure options may matter. [20:06] Action: iive is not sure for the last one... but. [20:07] iive: i have kind of put them in the order they appear in "./configure --help" [20:08] :) [20:08] iive: don't know if I could do better than that... except to look in the code :) [20:09] well, i'm just saying that it may be good idea to put the options that disable all codecs first, and then put the options that enable specific codecs after. [20:11] iive: oh yeah, sure. i did that :) [20:22] hello all - i've gotten awesome help from folks here in the past so I'm hoping someone can help me out with this issue today. here's the situation. I have 720p ProRes .movs that I am using ffmpeg to trim & transcode into H.264 .movs for YouTube upload. [20:22] the command I'm using is this: ffmpeg -i input.mov -ss 1201 -t 5000 -c:v libx264 -crf 18 -c:a libfaac -b:a 64k -ac 1 output.mov [20:22] (obviously the ss and t values change) [20:23] sometimes this works like a charm. other times, tho, youtube just rejects the file [20:23] often if I just rerun the same command it'll work fine the second time [20:23] any thoughts on what's going wrong here? [20:24] oh wait i miscopied the command [20:24] hang on [20:24] ffmpeg -i input.mov -pix_fmt yuv420p -c:v libx264 -crf 18 -c:a libfaac -b:a 64k -ac 1 -ss 1201 -t 5000 output.mov [20:29] ok i have to step away from my pc for a min, i will be back before too long tho. would appreciate any thoughts folks have. [20:51] compiling ffmpeg from master for windows gives an executable which crashes on exit while release 1.0 works fine. does somebody, by any chance, know when was this regression introduced? :) [23:03] Hey everyone. I am relatively new to Linux (running Kubuntu), and just installed ffmpeg, when I try to convert an ogv file from made with (using winff) desktop recorder (which plays fine) I get error messages no matter what I try, mov, mp4, avi [23:13] hi folks - asked this earlier, but didn't get a response, forgive me for the repeition: I have 720p ProRes .movs that I am using ffmpeg to trim & transcode into H.264 .movs for YouTube upload. [23:13] the command i'm using is this: ffmpeg -i input.mov -pix_fmt yuv420p -c:v libx264 -crf 18 -c:a libfaac -b:a 64k -ac 1 -ss 1201 -t 5000 output.mov [23:13] (obviously the -ss and -t values change) [23:13] sometimes this works like a charm. other times, tho, youtube just rejects the file [23:14] and often when tha thappens, I can just rerun that command in ffmpeg and the resulting file works [23:14] any thoughts on what's amiss here? [00:00] --- Mon Oct 22 2012 From burek021 at gmail.com Tue Oct 23 02:05:02 2012 From: burek021 at gmail.com (burek) Date: Tue, 23 Oct 2012 02:05:02 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20121022 Message-ID: <20121023000502.5E16F18A01EC@apolo.teamnet.rs> [00:12] ffmpeg.git 03Michael Niedermayer 07248b1ff26b7c: Fix various uses of av_log_missing_feature() * 03http://tinyurl.com/9ttzpsb03 [00:12] ffmpeg.git 03Michael Niedermayer 0739747d87d090: avfilter_graph_parse: add support for parsing sws_flags * 03http://tinyurl.com/8eevlbz03 [00:12] ffmpeg.git 03Michael Niedermayer 072d11ee4bfc7c: fate: fix fate-filter-metadata-scenedetect * 03http://tinyurl.com/9rorhfu03 [07:35] michaelni: thx for the fix! [09:08] michaelni: "fix used variable" // "unused", right? [10:07] ffmpeg.git 03Paul B Mahol 07cb0add3ce961: lavf/flacenc: disallow creation of invalid files with -c copy * 03http://tinyurl.com/9rwjvbk03 [10:42] saste: you forgot to remove gcd in cdea54b4c [10:42] btw, looking forward tools/ffescape :)) [10:43] oh and i sent a pending patch on the ml for the SAMPLES issue [10:44] hello, can I find libfdk devs here? [10:44] i don't think so [10:47] wbs^^ [10:48] ah, wbs indeed :) [10:50] btw, just curious, why working on an external encoder? [10:51] ffmpeg.git 03Stefano Sabatini 07b6e36a424447: lavu/parseutils: remove unused gcd variable in av_parse_ratio() * 03http://tinyurl.com/8cnkneh03 [10:51] ffmpeg.git 03Stefano Sabatini 07ccd6def9b3dc: lavfi/aspect: extend syntax for the setdar and setsar filters * 03http://tinyurl.com/8nserd403 [10:51] ffmpeg.git 03Stefano Sabatini 076752aac6bca7: lavfi/aspect: add max option * 03http://tinyurl.com/962my2b03 [10:52] thx saste [10:52] ubitux: were you asking me? [10:52] more wbs i believe :p [10:53] dunno what that is, but ok heh [10:53] a nickname on this channel :) [10:57] ubitux: what is $(FATE_SAMPLES_FFPROBE)? [10:59] tested the patch, and seems to work, but i'll leave review to someone more qualified than me [10:59] as for ffescape, i can hardly work on it before this night [11:00] given a filename, you need to perform this: [11:00] saste: FATE_SAMPLES_FFPROBE is a different list than FATE_FFPROBE, the same way FATE_SAMPLES_FFMPEG is a different list than FATE_FFMPEG [11:01] ubitux, ok bad naming but nothing we can do about [11:01] and these FATE_SAMPLES_ list are in the FATE_EXTERN list, which is only used when SAMPLES is set [11:01] do you want me to add this in the comment description? [11:02] <@saste> given a filename, you need to perform this: // this? [11:05] shescape(ffescape(ffescape(filename, ":") + ":$other_options", ",;")) [11:06] you need to escape ":" in the filename, then special graph chars in the resulting string, and finally shell escape [11:06] we can improve this mess in two ways [11:07] 1. by putting the filtergraph description in a file [11:07] i proposed that sometimes ago and it was rejected with the argument that sh can do it [11:07] (include a file in the command) [11:07] for the fate failure i believe we will just need to ffescape ':' in $SRC [11:08] 2. by creating an alternative syntax (more avisynth like) [11:08] filtergraph in a description file ? agree. [11:08] -f lavfi script=myfiltergraph [11:08] 2. is much work, so we can ignore it for the moment [11:09] ubitux, yes that's a possibility [11:09] -f lavfi evalscript=myfiltergraph even [11:09] ubitux, looks, maybe it is *aready* implemented [11:09] so we can add a script= later with a scripting thing like you just mentioned [11:10] ummhno [11:10] haha [11:10] but is easy to add [11:10] no way it is @_@ [11:10] ah no, it's the "str" [11:10] not a file [11:10] check for example how I did it in sendcmd [11:11] then I wouldn't be against -{a,v}f_file -filter_complex_file [11:11] but it is not required here [11:12] gottago, good luck ;-) [11:12] heh [11:13] will see later [11:13] still, generating a file for this in fate looks overkill [11:13] i'd better just ffescape the ':' [12:35] [mpeg2video @ 0177ab80] a vbv buffer size is needed, for encoding with a maximum bitrate [12:35] any idea? [12:40] -rc_max_vbv_use 1 [12:40] j-b : according to this http://www.itbroadcastanddigitalcinema.com/ffmpeg_howto.html anyhow [12:41] weird, this worked before [12:41] which commit broke it ? [12:42] no clue [12:42] like yesterday, users complaining [12:45] j-b, how can the problem be reproduced ? [12:46] according to user: "C:\Program Files\VideoLAN\VLC\vlc.exe" -I dummy --dummy-quiet --ffmpeg-hurry-up --no-loop --no-repeat --ffmpeg-skiploopfilter 4 --audio-language=any --start-time 0 --sout "#transcode{venc=ffmpeg,vcodec=mp2v,vb=7000,acodec=a52,ab=384,samplerate=48000,deinterlace}:std{access=file,mux=ts,dst="test.mpg"}" --sout-transcode-width 480 --sout-transcode-height 320 --sout-ffmpeg-keyint=3 "C:\Media\sample.avi" "vlc://quit" [12:46] let it drop [12:46] I will find a fix [13:21] ffmpeg.git 03Michael Niedermayer 076182e0a6f640: vf_aspect: unbreak avoption system * 03http://tinyurl.com/9escc9y03 [13:21] damn [13:24] j-b : vlc doesnt have an automated testing system like fate ? [13:24] no [13:24] Failed to avformat_open_input ... <= what's this ffspeak? [13:24] to detect regressions like this [13:31] j-b: you could call it "VATE" [13:31] http://www.urbandictionary.com/define.php?term=vate [13:33] in Italian/Latin it is a word (with a positive connotation, it is the attribution given to a high poet/priest) [13:38] it's hard to make test suites for video players [13:41] for jerkiness and tearing :) [13:41] and audio microcuts [13:44] ubitux: give job to several folks which will watch various clips all the time [13:45] it won't be called "VATE" then, it will be "CHINA" (i let you find an acronym for the joke) [14:07] durandal_1707: I watch clips all the time at work :) [14:07] av500: talk with j-b [14:08] av500: lucky you [14:08] ffmpeg.git 03Michael Niedermayer 07c3778df2d4c0: ffmpeg: fix negative array index * 03http://tinyurl.com/8sa9zdv03 [14:33] ubitux : whys that? it has command line interface, and i'm guessing some kind of md5output [14:33] if you mean , just for sheer amounts of tests, that would be hard to test every option on every build [14:34] Compn: you would be able to test decoders/demuxers [14:34] not what actually makes the video player [14:35] (the user experience, smooth playback, GUI interations etc) [14:35] smooth playback by timing the test [14:38] i didn't say it was impossible, but it's generally way more complex than just diffing outputs [14:44] Compn, can you add regression tests to ffplay :) ? [14:44] i wanted these for a while already ... [14:45] -vo file! [14:49] ffmpeg.git 03Nathan Caldwell 07a4aa20fbdbe8: avcodec: prefer decoders without CODEC_CAP_EXPERIMENTAL * 03http://tinyurl.com/8n42naf03 [14:49] ffmpeg.git 03Nathan Caldwell 07a893655bdaa7: avutil: Add AVERROR_EXPERIMENTAL * 03http://tinyurl.com/8se4wgf03 [14:49] ffmpeg.git 03Nathan Caldwell 07c854102da773: avcodec: handle AVERROR_EXPERIMENTAL * 03http://tinyurl.com/8tnjhnu03 [14:50] ffmpeg.git 03Martin Storsj? 07e0d5ac6ae3a8: rtsp: Update a comment to the current filename scheme * 03http://tinyurl.com/96qyhga03 [14:50] ffmpeg.git 03Martin Storsj? 07c3e15f7b39aa: rtpdec: Don't pass a non-AVClass pointer as log context * 03http://tinyurl.com/9mlgx9803 [14:50] ffmpeg.git 03Michael Niedermayer 07e3a91c51f713: Merge commit 'c3e15f7b39aac2012f09ee4ca86d2bc674ffdbd4' * 03http://tinyurl.com/98d379403 [15:01] ffmpeg.git 03Justin Ruggles 0746a86c61940e: alacdec: set bits_per_raw_sample * 03http://tinyurl.com/9zy555a03 [15:01] ffmpeg.git 03Anton Khirnov 07d6f4fe68c8e8: lavc: extend frame_size doxy. * 03http://tinyurl.com/9aolmt203 [15:01] ffmpeg.git 03Anton Khirnov 076173a8fe5d2c: riff: remove a write-only variable * 03http://tinyurl.com/8eovuy903 [15:01] ffmpeg.git 03Michael Niedermayer 07f9cf14c8da2a: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/9mfuo6b03 [15:04] someone can recommend an image formats which allows to read the values in a human readable way? [15:04] sng is a possibility, so i was contemplating the possibility to add native support to ffmpeg [15:04] XPM may be another alternative [15:05] ppm ? [15:05] ffmpeg -i tests/lena.pnm lena.ppm [15:06] vim lena.ppm... not very much [15:06] IIRC PPM has an ASCII option which we don't support [15:06] https://en.wikipedia.org/wiki/Netpbm_format [15:06] this is what i call "ppm" [15:06] yes [15:07] https://en.wikipedia.org/wiki/Netpbm_format#PPM_example [15:07] we don't support the ASCII format [15:07] saste: nit++: 's' is before 'v' [15:07] (ffescape, ffeval -- gitignore) [15:08] ubitux, that list is not sorted anyway [15:08] locally fixed [15:08] and yes xpm are indeed another one [15:09] so why esr implemented sng? [15:09] (vim even renders directly xpm :p) [15:17] replying to myself: SNG supports palette data, so it's ideal for storing/viewing/editing paletted images (like PNGs) [15:54] durandal_1707: what happend to http://ffmpeg.org/pipermail/ffmpeg-devel/2012-June/125668.html ? [15:55] rot, because nobody REing prores alpha part [15:56] it can be added, but nothing will use it, because there is no other codecs that use this format, and anyway it is not really imprtant that much because there is yuva444p16 [15:56] and switch can be done at any time once prorse alpha code comes [16:40] durandal_1707: dnxhd also has an alpha mode that isn't re'd [16:42] isn't that raw? [16:43] but that is 422 [16:43] it is vc3- based [18:12] ffmpeg.git 03Michael Niedermayer 0722793d7bb36e: ffmpeg/lavc: move experimental warnings to libavcodec. * 03http://tinyurl.com/8uw866r03 [19:44] ffmpeg.git 03Matthieu Bouron 07d8173f264f73: dv: fix a check on dv_extract_timecode return value * 03http://tinyurl.com/8jldf9x03 [19:44] ffmpeg.git 03Matthieu Bouron 07c68a8a1340e7: lavf/aiffdec: don't stop parsing after SSND chunk * 03http://tinyurl.com/8wgqwhz03 [20:46] what to use to set jpeg encoder quality/compression? [20:51] -qscale X [20:53] but from lavc user point of view [22:14] anyone to have a look to http://ffmpeg.org/pipermail/ffmpeg-devel/2012-October/132854.html ? [22:15] i added the following description: [22:15] FATE_SAMPLES_FFPROBE is a different list than FATE_FFPROBE, the same way [22:15] FATE_SAMPLES_FFMPEG is a different list than FATE_FFMPEG; these [22:15] FATE_SAMPLES_ lists are in the FATE_EXTERN list, which is only [22:15] used when SAMPLES is set. [22:21] mmh i'm wondering [22:22] what's wrong actually with the msvc fate instance? [22:22] the configuration is --samples=/d/Dev/ffmpeg/fate/samples [22:23] is the ':' also missing on fate display? [22:23] Daemon404: how is your --samples configured? [22:24] my what now? [22:24] context? [22:25] Daemon404: the msvc fate instances are yours, right? [22:25] i'd like to know you set the samples directory [22:25] theyre nev's [22:25] and if the path is indeed "/d/Dev/ffmpeg/fate/samples" [22:25] oh, ok, sorry [22:25] i have one msvc instance of a shared build, and its for libav [22:25] since ffmpeg's doesnt work yet [22:25] iirc [22:26] nevcairiel: ping :) [22:26] Daemon404: ok :( [22:26] wut [22:26] nevcairiel: what's your samples path in the msvc instances? [22:26] im working on some exotic setups atm... [22:26] nerbsd/armv5 [22:26] netbsd* [22:26] technically the path is D:\Dev\ffmpeg\... however because its running in a msys bash shell, it gets the msys-style path [22:26] nerdsb [22:26] msys should translate that to the real path before ffmpeg ever sees it [22:27] probably d:/dev/... [22:27] yes [22:27] is this that : problem again? [22:27] yes likely [22:28] right now ffmpeg sees '/d/Dev/ffmpeg/fate/samples/svq3/Vertical400kbit.sorenson3.mov' [22:28] i doubt that [22:29] well: http://fate.ffmpeg.org/report.cgi?time=20121022182926&slot=x86_64-msvc10-windows-native [22:29] expecting users to escape : in path names is dumb. [22:29] just sayin. [22:29] the script is called with that [22:29] but its translated before ffmpeg ever gets it [22:29] yup [22:29] or should be at least [22:29] msys is sometimes funny [22:30] ffmpeg.git 03Michael Niedermayer 07250fe6eeb4a1: cmdutils: apply option to codec and format contexts if possible. [22:30] this is a particular case :p [22:30] i wonder if i shouldn't just define another command in place of "run" [22:31] ah well that's stupid [22:31] it wouldn't help either [22:31] Action: Daemon404 notes a 15 gb image takes forever to dd to an sdhc card [22:32] hm how did you trick msys into not translating the path? [22:32] it did translate it when i looked at the error yesterday [22:32] i hid it in a string [22:34] i can probably specify a windows-style path in the fate config [22:35] [16:32] < nevcairiel> hm how did you trick msys into not translating the path? [22:35] this is the wrong "solution" [22:36] what solution do you see within a filtergraph Daemon404? [22:36] ubitux, fix the dang parser [22:36] so it handles VALID patsh correctly [22:36] well, it's just opening the passed path [22:36] in the recent error messages, ffmpeg is actually getting a msys path [22:36] which is '/d/Dev/ffmpeg/fate/samples/svq3/Vertical400kbit.sorenson3.mov' [22:37] because msys failed to identify the path in the middle of a argument [22:37] ah [22:37] yes [22:37] msys wont tl that [22:37] it has specific rules about what it will tl [22:37] so i just change it to windows path with wrong-way-slashes [22:37] work both for win32 and msys apps [22:38] the problem will then be the ':' [22:38] if it cant handle the : [22:38] (i guess) [22:38] thats a bug that should be fixed [22:38] period [22:38] or rather [22:38] colon. [22:38] how? [22:38] ubitux, why cant it handle : in the first place [22:38] because its the key-value pair seperator in filtergraph arguments [22:38] :) [22:39] well you choose that delim poorly. [22:39] right, but we can't change it now [22:39] i think the whole idea of the movie source is a poor one [22:39] me too [22:39] but thats just me [22:39] again, do you see another solution? [22:39] if it was sane [22:40] having sources in the filtergraph is nice IMO [22:40] hwy? [22:40] i dont see ANY benefit [22:40] from doing -i [22:40] you can describe completely your filtergraph with a string, which simplifies the API a lot [22:40] not all tools want to reproduce the filter complex thing from ffmpeg [22:41] and that's indeed not done in ffprobe, and ffplay [22:41] as the name suggest, it's "complex" [22:41] yes, one horribly ugly string [22:41] :| [22:41] i dont see a benefit to having the src specified in it [22:41] i just told you... :/ [22:41] [16:40] <@ubitux> not all tools want to reproduce the filter complex thing from ffmpeg [22:42] thats called design failure [22:42] but not making that an api [22:42] well it really is simpler [22:42] you only need to send a string to your app [22:42] there is nothing simpel about lavfi's api. [22:42] to describe a filtergraph [22:42] or even its input strings [22:43] its akin to reading lisp [22:43] for me [22:43] ;) [22:43] i wonder if we shouldn't just add a CONFIG_SANE_ENV and put SANE_ENV in the deps of these tests [22:45] the inability to be able to deal with windows path is something that should be fixed [22:45] anyway, if the path is transformed to a "d:\foo\bar" syntax, it should be possible to handle it without much trouble when ffescape is in [22:45] nevcairiel: it is able to deal with windows path [22:45] if it cant handle C:/depr.mov [22:45] no [22:45] no it cannot. [22:45] it can, you just have to escape the : afaict [22:45] thats not handling it [22:45] at all [22:45] >_> [22:45] please [22:46] im serious [22:46] that is insanely dumb [22:46] we can't just change the whole syntax of filtergraph just because you don't want to add a '\' [22:46] or add some basic checks? [22:46] or maybe that's actually a solution mmh [22:46] for e.g. /,.:/ [22:46] or something [22:46] like setting a global separator [22:46] yes [22:46] like sed :P [22:47] s#a#b# [22:47] or s/a/b/ [22:50] Action: ubitux wonders about what other character to use [22:50] mmh actually [22:51] would "movie='d:\foo\bar':xx=..." not working? [22:51] it's using av_get_token so it might work [22:52] hmm [22:52] why is this failing anyway ? [22:52] damnit, there is some bash magic in the fate test that checks if the samples path starts with a slash [22:52] unportable shit ftl [22:52] there is surely no "d" option in the filter [22:52] so d: could easily be handled [22:53] michaelni: right now because it's not passing "d:\foo\bar" but the other path [22:53] and then it will fail because of ':' escaping which IMO can be avoiding by just adding '' around $(SRC) [22:54] that should work indeed [22:55] something quite different, i think filter graph descriptions in genera need more whitespace and newlines, that would make them more readable [22:56] https://github.com/ubitux/FFmpeg/compare/master...fate [22:56] inb4 add a \ [22:56] and hit return [22:56] spaces should be skipped yes [22:56] ffmpeg.git 03Michael Niedermayer 070de41ead6f90: qt-faststart: check fseeko() return codes [22:57] < nevcairiel> damnit, there is some bash magic in the fate test that checks if the samples path starts with a slash // huh? [22:57] well there is [22:58] it uses that to try to check if its a relative path [22:58] ah, ok [22:58] and something like d:/... looks relative to it :P [22:58] haha [22:59] arent there bash-builtins to do this stuff [22:59] why the manual hackery [22:59] oh well [22:59] no fixing the path then [22:59] ok so, how are we supposed to handle that '/d/...' thing? [23:00] you're not, fix fate :p [23:00] so much trolling just to realize that there is nothing wrong with lavfi :( [23:01] it still failed on the : path before you added those sws flags [23:01] what ':'? [23:01] http://fate.ffmpeg.org/report.cgi?time=20121021161916&slot=x86_64-msvc10-windows-native [23:01] if we add the '' quoting it will work just fine [23:02] i can push my branch right now to fix that [23:02] i'd just need a ok for 01f2386 and maybe the others [23:02] i should figure out a more portable way to figure out if a path is abolute in bash [23:02] i'm sure there is a real function for that [23:03] all the quick solutions on google check for damn slashes [23:04] it's simple [23:04] we should just check for damn ':' [23:06] i could just leave out the drive from the path, its all on the same one anyway, so /Dev/... would work [23:06] but thats not a universal solution [23:06] and it probably wouldnt work with any msys tools [23:09] wait. [23:09] Error initializing filter 'movie' with args 'd:/Dev/ffmpeg/fate/samples/svq3/Vertical400kbit.sorenson3.mov' [23:09] so this is what you actually pass to the filter [23:09] if msys manages to translate the path [23:09] or the native path is passed [23:10] ah and the sws_flags thing broke that detection? [23:10] yeah, it works if its the first argument [23:10] but the sws flags appeared before it [23:11] how does this detection work? [23:11] a lot of magic, i guess [23:11] i'm not even sure it will detect it if we enclose it within '' [23:11] they might expect quoting [23:12] well we can see it with the silencedetect one [23:12] since it hasn't sws flag thing [23:13] michaelni: do you mind if i push my branch? [23:13] ubitux, if it fixes things sure push it [23:13] ok [23:14] ffmpeg.git 03Cl?ment BSsch 077be9c0c10fc5: fate: fix SAMPLES dependency for ffprobe. [23:14] ffmpeg.git 03Cl?ment BSsch 07ae69c683e58c: fate: add scale filter to the scene detect dependencies. [23:14] ffmpeg.git 03Cl?ment BSsch 072649b78384a7: fate: improve metadata filter deps readability. [23:14] ffmpeg.git 03Cl?ment BSsch 077c2d5eec29e0: fate: quote file paths in movie/amovie filtergraphs. [23:14] let's see now. [23:15] so now cone wont give an url anymore ? [23:15] haha :) [23:15] better than a tiny url! [23:16] Action: cbsrobot stares at gnafu [23:17] nevcairiel: i see you're doing some experiments with one msvc instance :D [23:18] i tried to change the path, but i found that stupid absolute-path check :p [23:19] i just pushed the '' btw, if you want to give it a try [23:23] nevcairiel: i'd like to have a look to the code doing the path translation; what part of what project is responsible for this? [23:23] in msys? i have no clue [23:23] msys from mingw? [23:24] yeah its somewhere in one of the msys runtime libs that get linked into the executables [23:24] no idea where [23:24] ok, thanks [23:25] git clone git://mingw.git.sourceforge.net/gitroot/mingw/mingw [23:25] Cloning into 'mingw'... [23:25] warning: You appear to have cloned an empty repository. [23:25] :( [23:25] mingw uses svn [23:25] why are you cloning mingw [23:25] >____> [23:25] nothing good can come of this [23:26] http://sourceforge.net/projects/mingw/develop [23:26] i see cvs & git :p [23:26] you probably want http://mingw.git.sourceforge.net/git/gitweb.cgi?p=mingw/msys-runtime;a=summary [23:26] thx [23:26] again [23:26] why [23:26] <_< [23:26] he wants to find the path translation code :p [23:26] .. [23:26] theres an entire damn page [23:26] dedicated to it [23:26] on their site [23:27] http://www.mingw.org/wiki/Posix_path_conversion [23:27] oh great [23:27] thx [23:27] http://mingw.cvs.sourceforge.net/viewvc/mingw/msys/rt/src/winsup/cygwin/path.cc?view=markup *_* [23:27] viewvc is horrible [23:27] end of story [23:28] :3 [23:29] my fate finished building, nothing changed [23:31] anyway, time for sleep [23:31] mmh this looks weird [23:31] night :) [23:33] Missing key or no key/value separator found after key '/Dev/ffmpeg/fate/samples/amrwb/seed-12k65.awb' [23:33] this is pretty weird [23:37] as if the msys escaping was just breaking everything [23:43] mmh something will need to be changed in src movie maybe [23:45] ok get it [23:59] "movie=f='c:\foo\bar'" doesn't actually receive "f='c:\foo\bar'" as arg, but "f=c:\foo\bar" [23:59] that's what causing the issue [23:59] heh sws cant handle 4x4 images [23:59] hey saste, perfect time! :) [23:59] ubitux, i want to tell that the ' should work [23:59] i commited [23:59] but it doesn't [23:59] because of what i just said [00:00] --- Tue Oct 23 2012 From burek021 at gmail.com Tue Oct 23 02:05:01 2012 From: burek021 at gmail.com (burek) Date: Tue, 23 Oct 2012 02:05:01 +0200 (CEST) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20121022 Message-ID: <20121023000501.568BB18A01DE@apolo.teamnet.rs> [00:34] <_stclaws> I am trying to compile ffmpeg with rtmp support. I first build rtmpdump. But when doing configure of ffmpeg I get: ERROR: librtmp not found. Any idea why? [01:05] _stclaws: check config.log [01:12] <_stclaws> iive: I tried everything I could think of. [01:12] _stclaws: the config.log should contain the failed check, together with the reason for the failure. [01:13] <_stclaws> It just ends with : check_pkg_config librtmp librtmp/rtmp.h RTMP_Socket [01:13] <_stclaws> ERROR: librtmp not found [01:14] looks like you installed only the library and the -dev with the headers is missing [01:15] can you find rtmp.h somewhere in your filesystem? /usr/include/librtmp/rtmp.h ? [01:15] <_stclaws> librtmp is here: /usr/local/include/librtmp [01:16] <_stclaws> I thought all would be included with rtmpdump [01:16] redhad doesn't include /usr/local in their defaults, so you may try specifying it as additional include path. [01:18] try: --extra-cflags=-I/usrl/local/include [01:21] oops, /usr/local/include .... [01:22] <_stclaws> iive: tried that path but nope (yes usr) [01:23] <_stclaws> I read somewhere that pkg-config had something to do with it [01:23] same error in the config.log? [01:23] can you pastebin a little bit more from the end? [01:25] <_stclaws> http://pastebin.com/U9tEbivS [01:26] <_stclaws> and my config line: http://pastebin.com/3L6j6Kn6 [01:49] <_stclaws> I did add PKG_CONFIG_PATH=/usr/local/lib/pkgconfig because that's where librtmp.pc is [01:51] <_stclaws> Am I supposed to install librtmp-dev also although I did install rtmpdump? [02:06] when i try to ffplay an mp3, instead of playing it, ffplay displays a new window with a spectrogram of the mp3 [02:07] why does it do that? and how can i get it to just play the file? [02:07] It should also be playing it. [02:08] ah.. i see that it was.. but just through alsa, not jack [02:09] how can i turn off the spectrogram display? [02:11] Why would you want to? [02:11] i don't want an extra window to appear, or for ffplay to waste extra processing power displaying it [02:11] it's distracting, annoying, and wasteful [02:12] Minimise it, and if it's wasteful, your computer is either really, really old or you're overly sensitive to waste... [02:13] my computer is old [02:16] <_stclaws> Ok, I am stumped, it is getting late, and I guess I soon give up... [02:22] _stclaws: sorry, I got to do other things. [02:22] installing in /usr/local usually indicates you compiled yourself, so -dev package may not work [02:23] good call with the pkg config path, but i assume it didn't work either? [02:23] try to call it directly. [02:29] e.g. pkg-config --cflags --libs librtmp [02:29] the configure actually uses pkg-config --exists librtmp [02:29] but that returns true/false to the script. [02:30] i must try rtmp myself. maybe tomorrow. [02:36] <_stclaws> when I try to call it directly it complains that libssl is not found [02:37] <_stclaws> it is in /usr/lib/pkgconfig/libssl.pc . So I add that path to PKG_CONFIG_PATH. But still saying Package 'libssl', required by 'librtmp', not found [02:55] <_stclaws> Crap, I just forgot to do export of PKG_CONFIG_PATH :) [03:10] <_stclaws> thanks for the help, iive! [03:15] <_stclaws> Does anyone know what the proper syntax is for the "listen" and "rtmp_live" parameters of the rtmp protocol? [04:39] i have another question about codec and ffmpg. if i was to output to mp4 then change to wmv, will i save some file size? [06:56] _stclaws: ffmpeg -rtmp_live 1 ... / ffmpeg -rtmp_listen 1 ... [06:57] _stclaws: I think rtmp://...?listen&timeout=... may also work [07:02] what's a good lossless video format? [10:13] <_stclaws> I am receiving a rtmp stream from a fms server, but the stream is now and then "renewed" since the fms source is switched. I want ffmpeg to just go on but it thinks the stream ended and exits. Does anyone know how to just keep it up listening and receiving indefinitely? [10:13] Last message repeated 1 time(s). [10:13] <_stclaws> sorry twice [10:17] Hello.. have a question.. is there any way to have ffmpeg streams not enter a container. I'm capturing audio/video but mp4 doesn't support pcm_s16le audio so I will get a error if I try to capture. Instead I would like to have video go into the mp4 but audio go to another .wav in realtime. [10:18] Exatrive: look at the documentation for -map [10:19] I did look at the map function.. but its not too clear if this will allow it to map the channel to bypass the first container. I will look again and see what I can do... [10:27] hello, I'm trying to build the latest ffmpeg from source following the ubuntu compilation guide and am having trouble during the make step for fdk-aac [10:27] this is the error I'm getting on make: libtool: link: more than one -exported-symbols argument is not allowed [10:27] make: *** [libfdk-aac.la] Error 1 [10:41] cjhmdm, looks like an autohell libfdk issue [10:41] saste: autohell? [10:42] cjhmdm, autohell = autotools (autoconf, automake, libtool, etc.) [10:43] saste: any way I can resolve it? [10:43] cjhmdm, ask to libfdk devs, they will be more qualified to help [10:46] cjhmdm: did you get the source from git ? [10:47] cbsrobot: yes, I'm following the guide, line for line here: https://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide [10:47] these are the exported symbols: https://github.com/mstorsjo/fdk-aac/blob/master/fdk-aac.sym [10:48] campare it to your fdk-aac.sym [10:49] https://gist.github.com/3930424 [10:49] maybe you can debug it in the makefile [10:49] after the autoreconf step [10:49] it's telling me I can't have more than 1 exported-symbol [10:50] my fdk-aac.sym file is the same as the one you linked [10:50] unless I'm blind heh [10:51] just before this line https://github.com/mstorsjo/fdk-aac/blob/master/Makefile.am#L32 [10:51] see if you can find it in the Makefile [10:52] libfdk_aac_la_LDFLAGS = -version-info 0:1:0 -no-undefined \ [10:52] -export-symbols $(top_srcdir)/fdk-aac.sym [10:53] what does your shell say about "echo $LDFLAGS" [10:53] nothing [10:53] no return [10:54] see line 384 in your Makefile [10:54] what I don't get is why the trouble all of a sudden.. I've built, rebuilt and updated ffmpeg on this machine around 6 times over the past year with no problem [10:55] maybe you can print the export-symbols jut before that [10:55] although, I do blieve the fdk-aac step is new [10:55] line 384 in my makefile is: $(MPEGTPDEC_DIR)/tpdec_latm.cpp \ [10:56] I'm still not having any luck with the map option.. the audio stream still want to be saved in the mp4 file... with a copy of the audio in the .wav [10:56] cjhmdm: search for export-symbols [10:56] I'm trying to capture video to mp4, and audio to .wav seperatly with no luck. [10:57] cbsrobot: line 297: -export-symbols $(top_srcdir)/fdk-aac.sym [10:57] the whole line: [10:57] libfdk_aac_la_LDFLAGS = -version-info 0:1:0 -no-undefined \ [10:57] -export-symbols $(top_srcdir)/fdk-aac.sym [11:01] go figure.. compiles just fine on a fresh machine with the same hardware and os as the other machine (but with only base system and necessary build tools).. but all dev tool versions are the same, so can I deduce that it's conflicting with another package on the system? [11:03] cjhmdm: search for "all:" in the makefile [11:04] and put "@echo $(libfdk_aac_la_LDFLAGS)" after that line [11:04] run make again [11:04] cbsrobot: all: all-am [11:04] see what it prints [11:04] yes [11:04] Makefile:484: *** target pattern contains no `%'. Stop. [11:04] without the quotes [11:04] @echo $(libfdk_aac_la_LDFLAGS) [11:05] I didn't put quotes :P [11:05] all: all-am [11:05] @echo $(libfdk_aac_la_LDFLAGS) [11:05] with a tab in fron [11:05] ahh [11:05] *front [11:06] https://gist.github.com/3930510 [11:07] in your shell [11:07] wait [11:07] that was the output after adding that line to the makefile and running make [11:09] add "@echo $(LINK)" [11:09] after the other line [11:09] and show me the output [11:09] hello [11:09] so the above line then @echo $(LINK) below it? tabbed also? [11:09] yes [11:09] same output [11:10] wait, I think there's a bit more on there [11:10] i have a problem with building the latest snapshot or git versions of ffmpeg where i get http://pastebin.com/raw.php?i=AGvrTcBq [11:10] echo $LD_LIBRARY_PATH says /usr/local/lib: [11:10] cbsrobot: https://gist.github.com/3930534 [11:11] mkozjak: make clean [11:11] x264 built with Command line options: "--enable-pic" "--enable-shared" "--extra-cflags=-march=native" [11:11] cbsrobot: nevermind.. it is the same output [11:11] make clean && make [11:11] cbsrobot: yeah, doesn't help [11:11] cjhmdm: swap the two lines [11:11] make clean && make distclean && make -j20 [11:11] cbsrobot: all: all-am [11:11] @echo $(LINK) [11:11] @echo $(libfdk_aac_la_LDFLAGS) [11:12] same output [11:13] <_stclaws> Could anyone point me to a good tutorial for how to do HTTP Live Streaming with ffmpeg? From an incoming live rtmp stream. [11:14] cjhmdm: hmmm [11:14] cbsrobot: [11:14] err.. yeah, hmm heh [11:14] I show you my output [11:14] the weirdest thing is, when running make on the new machine, the output is entirely different [11:14] -version-info 0:1:0 -no-undefined -export-symbols ./fdk-aac.sym [11:14] and it compiles fine [11:15] /bin/sh ./libtool --silent --tag=CC --mode=link gcc -g -O2 -o all [11:15] but the only other thing on the 'old' machine is wowza media server [11:15] firts line is @echo $(libfdk_aac_la_LDFLAGS) [11:15] and the needed libs for building ffmpeg [11:15] second $(LINK) [11:16] try to figure out where in the makefile it adds the second export-symbols [11:16] can i somehow define to configure where the system should look for libx264.so? [11:17] cjhmdm: search for libfdk_aac_la_LDFLAGS and see how it is created [11:18] hi [11:18] cbsrobot: ok [11:18] any of the ffvp8 devs around? [11:20] cbsrobot: what I don't get is this also... this is the output on the old machine: https://gist.github.com/3930568 --- this is the output on the new machine: https://gist.github.com/3930561 --- the 2 machines are identical in every way. The only difference is the old machine has other packages (unrelated to building ffmpeg) installed [11:21] ~/ffmpeg/libvpx ? [11:21] cbsrobot: libfdk_aac_la_LDFLAGS = -version-info 0:1:0 -no-undefined \ [11:21] -export-symbols $(top_srcdir)/fdk-aac.sym [11:21] cjhmdm: wrong copy paste ? [11:22] no, that's where libfdk_aac_la_LDFLAGS is created [11:22] see https://gist.github.com/3930561 [11:22] wrong paste [11:22] err yeah [11:22] lol [11:22] one sec [11:24] this is the proper output on the new machine: https://gist.github.com/3930586 [11:24] you need to make clean on this machine first [11:25] the old or new? [11:25] new [11:25] michaelni: ping [11:25] I just followed the guide heh [11:26] same output after running make clean heh [11:26] i am working on an algorithm that generates intermediate frames using the motion compensation information during decoding [11:26] i need some help with the source code for h264.c [11:27] wyuka: check #ffmpeg-devel [11:28] oh okay [11:30] cbsrobot: oh well.. I'll try not to stress it heh.. just forces me to set up the new processing server anyway. I just hate not being able to figure out why something is broken heh [11:32] cbsrobot: thanks for trying to help either way [11:32] np [11:59] <_stclaws> I need to set up ffmpeg to receive multiple consecutive rtmp live streams and re-encoding them into one single stream. Can anyone suggest the best way for doing that? [12:05] What do I have to enable while configuring for building to get WMV2 encoding? [12:05] --enable-wmv2 alone apparently isn't enought. here are all of the options I'm using if it's relevant: http://pastebin.com/u7XV393g [12:06] I figured out the mapping.. omg... my command line looks like a mess with all this mapping but at least it works.. lol [12:07] Element9: maybe --enable-encoder=h263 [12:08] cbsrobot: oh, that's why h263 got enabled by itself :) and I disabled it explicitely [12:08] cbsrobot: thanks. I'll try that [12:08] np [12:20] how do i make sure ffplay plays single channel files as single single & keep the sample rate as it is ? [13:49] hello is it possible to select a stream with the hex data like #0:1[0x45] [13:50] <_stclaws> has anyone here done HTTP Live Streaming with ffmpeg? [13:51] yes [13:51] http://ffmpeg.org/ffmpeg.html#segment_002c-stream_005fsegment_002c-ssegment [13:51] _stclaws: look for "HLS" here [13:52] arpu: isn't this a program id? [13:53] hmm could be the problem is if i restart ffmpeg the order of the streams (more audio streams) is different [13:58] <_stclaws> ubitux: I am just making my first tests but can't get the video to play on the web page. Can you see anything wrong here: http://pastebin.com/9tKdL7cJ [13:58] <_stclaws> ubitux: This is the page with the test video: http://217.70.33.15/hls/ [14:01] <_stclaws> ubitux: This was my ffmpeg command line: http://pastebin.com/HyRmtXgC [14:21] <_stclaws> Is this normal output when you do hls streaming from ffmpeg: http://pastebin.com/E55BLSrh [14:32] arpu: http://ffmpeg.org/ffmpeg.html#Stream-specifiers-1 [14:32] look for "program id" [14:33] ubitux, thx [14:33] _stclaws: ffplay 'http://217.70.33.15/hls/test1.m3u8' works for me [14:34] i will try maybe like -v:p:0x45 ? [14:34] <_stclaws> Ok, now I get video. But it only shows up after I stop streaming. [14:35] <_stclaws> While ffmpeg is working and outputting the .ts and .m3u8 nothing plays on the web [14:36] arpu: select for what? [14:36] _stclaws: you need a player on front [14:37] or you need to just output a standalone webm [14:37] Stream #0:0[0x44](deu): Audio: mp3 ([3][0][0][0] / 0x0003), 44100 Hz, stereo, s16, 128 kb/s [14:37] Stream #0:1[0x45]: Video: h264 (High) ([27][0][0][0] / 0x001B), yuv420p, 720x576 [SAR 64:45 DAR 16:9], 25 fps, 25 tbr, 90k tbn, 50 tbc [14:38] arpu: i mean, for -map? [14:38] _stclaws, did you try the +live flag? [14:42] <_stclaws> for the rtmp in ffmpeg, yes. The input is no problem. The problem is that the player on my web page cannot see the output until the streaming stops. [14:43] <_stclaws> ubitux: I am using the