Ffmpeg-devel-irc

Download

ffmpeg-devel-irc@ffmpeg.org

January 2014

1 participants
62 discussions

ffmpeg-devel.log.20140115
by burek 16 Jan '14

16 Jan '14

[01:10] <BBB> ubitux: very nice, lgtm [01:39] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:785dc146544a: avcodec/adpcm: fix sample count for stereo SBPRO3 [03:04] <BBB> ubitux: iadst16 simd on my github, working now [03:04] <BBB> ubitux: I'll do some measurements and then submit a patch [04:16] <BBB> and email sent [04:57] <cone-722> ffmpeg.git 03James Almer 07master:8bfa5f7fab3d: webp: fix capitalization of the word Exif [04:57] <cone-722> ffmpeg.git 03Michael Niedermayer 07master:ab296c7a9fe5: avcodec/hevc_ps: Override max_dec_pic_buffering when its inconsistent with num_reorder_pics [05:16] <cone-722> ffmpeg.git 03Janne Grunau 07master:8b4119187b62: cabac: remove leftovers from the cabac encoder [05:16] <cone-722> ffmpeg.git 03Michael Niedermayer 07master:2df53e0f7cea: Merge commit '8b4119187b62d6932e07aded11d33d3b24e1b42f' [05:22] <cone-722> ffmpeg.git 03Janne Grunau 07master:032d4da8af35: aarch64: configure: set arch dependent features [05:22] <cone-722> ffmpeg.git 03Michael Niedermayer 07master:9d2286761329: Merge commit '032d4da8af353e32ab5c86ee3a210f6065921a4a' [05:31] <cone-722> ffmpeg.git 03Janne Grunau 07master:d0cd2a8c461f: aarch64: bswap inline assembly [05:31] <cone-722> ffmpeg.git 03Michael Niedermayer 07master:53e6977c07a8: Merge remote-tracking branch 'qatar/master' [13:46] <ubitux> BBB: should i wait for your mc and iadst funcs before adding avx? [13:46] <BBB> whichever goes in first [13:46] <BBB> if yours goes in first, I'll simply add INIT_XMM avx to my patch [13:46] <BBB> else you add it to yours [13:46] <ubitux> well i can push my little patch right now [13:46] <BBB> go for it [13:46] <BBB> I'll amend mind [13:46] <ubitux> i didn't add anything for mc though [13:47] <BBB> mc won't be helped much [13:47] <BBB> it doesn't use any avx instructions [13:47] <BBB> for mc, we should probably just use axv2 for 16/32 wide blocks [13:47] <ubitux> ok [13:47] <BBB> and then ignore avx [13:48] <ubitux> i'll push in a moment then [13:48] <BBB> ok [13:48] <BBB> feel free to review mine after that, I'll rebase and add the relevant INIT statements [13:49] <BBB> and you said you'll do lf_8 right? [13:49] <BBB> it's very much on top of the "my performance sucks" list right now [13:53] <cone-432> ffmpeg.git 03Clément BSsch 07master:8b4190da9382: vp9/x86: add AVX for itxfm and lpf. [13:55] <ubitux> BBB: the patch seems not rebased on the recent itxfm changes [13:56] <ubitux> ah i'm looking at the wrong change.. [13:57] <ubitux> mmh no it was correct [13:59] <ubitux> lea sstride3q, [sstrideq*3] [13:59] <ubitux> oh? [14:00] <ubitux> you can do that? :o [14:06] <michaelni> with the right assembler you can, but they turn it into [reg+2*reg] [14:08] <ubitux> ooh [14:08] <ubitux> that's the trick :D [14:08] <nevcairiel> yasm is smart like that [14:11] <cone-432> ffmpeg.git 03Martin Storsjö 07master:d51f09962d5b: mov: Free intermediate arrays in the normal cleanup function [14:11] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:04cc753d0744: Merge commit 'd51f09962d5b4bc999fb70c040f330dd1873212e' [14:13] <BBB> ubitux: yeah yasm is cool like that :) [14:13] <BBB> ubitux: you can also do *5 or *9 like that [14:13] <ubitux> i didn't even though about that trick [14:13] <ubitux> i wonder if i could use that for the vertical version of lpf [14:15] <BBB> usually not [14:16] <BBB> it takes an extra lea to generate the *3, so *3 is the same as lea address, add+stride*2 and then [adress+stride] [14:16] <BBB> it works if you do it many times [14:16] <ubitux> BBB: i do it many times :) [14:16] <BBB> then it may help yes [14:16] <BBB> becasue you can read groups of 4 pixel rows instead of 2 [14:19] <cone-432> ffmpeg.git 03Martin Storsjö 07master:2620df13104d: mov: Free an earlier allocated array if allocating a new one [14:19] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:fb37d03a8708: Merge commit '2620df13104ddaa136158eb6bb1195adbf9d7692' [14:20] <ubitux> with stride, mstride, stride3 and mstride3 you can the [-8;8] range without ±7 and ±5 [14:20] <ubitux> can *do* [14:21] <ubitux> but with an extra dst you could cover everything i suppose [14:21] <ubitux> and with the 2 extra dst that's even better [14:25] <ubitux> BBB: why do you interleave 3 sub srcq, sstrideq? [14:26] <ubitux> (instead of doing a single sub srcq, sstride3q) [14:28] <ubitux> also, srq4q is just the original srcq+sstrideq [14:29] <ubitux> no idea if that makes any difference to not have the *4 [14:30] <BBB> what code where? [14:30] <ubitux> gonna send a mail [14:30] <ubitux> p [14:30] <ubitux> just a min [14:33] <ubitux> BBB: mail sent [14:33] <BBB> ty [14:33] <BBB> I'll read it at work [14:33] <BBB> gotta go now [14:33] <BBB> bye [14:33] <ubitux> cya [14:34] <BBB> oh that patch was merged a while ago wasn't it? [14:35] <BBB> anyway will check sure, if it's faster, feel free to merge it [14:35] <BBB> bye [14:36] <ubitux> ah, i guess i'm lost in the wrong mails [14:36] <ubitux> :D [14:36] <ubitux> i must have archived the wrong mails, silly me [14:41] <cone-432> ffmpeg.git 03Janne Grunau 07master:1e9265cd8f08: dct-test: test ff_simple_idct_neon only on ARM for now [14:41] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:d01f8790d93a: Merge commit '1e9265cd8f0821acbeca1db437be1361a3976b85' [14:59] <cone-432> ffmpeg.git 03Janne Grunau 07master:b7b17ed66e19: aarch64: add cpuflags support for NEON and VFP [14:59] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:d5560f1fecde: Merge commit 'b7b17ed66e199afc7246e642bf3b35c3f8eca217' [15:06] <cone-432> ffmpeg.git 03Janne Grunau 07master:71617884a2a6: aarch64: h264 chroma motion compensation NEON optimizations [15:06] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:6f001d87ff01: Merge commit '71617884a2a673908bd5c0f73d4f91fdca3da82a' [15:13] <cone-432> ffmpeg.git 03Janne Grunau 07master:8438b3f09f6b: aarch64: h264 idct NEON assembler optimizations [15:13] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:fb1c786a9dc6: Merge commit '8438b3f09f6b225d0886cc385117c38eb44ca0c1' [15:19] <cone-432> ffmpeg.git 03Janne Grunau 07master:d5dd8c7bf0f0: aarch64: h264 qpel NEON optimizations [15:19] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:19fc3c0122eb: Merge commit 'd5dd8c7bf0f0d77c581db3236e0d938f06fd5591' [15:24] <ubitux> BBB: just replied to the correct mail [15:24] <ubitux> (hopefully) [15:26] <cone-432> ffmpeg.git 03Janne Grunau 07master:c65d67ef5024: aarch64: hpeldsp NEON optimizations [15:26] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:28207bfafb90: Merge commit 'c65d67ef50241886b9749e88fd333c52090f70c7' [15:31] <ubitux> BBB: anyway, for the 8x8, should i write mmx code? [15:36] <cone-432> ffmpeg.git 03Janne Grunau 07master:36e3b1f2fd26: aarch64: h264 loop filter NEON optimizations [15:36] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:bf0470a5be1c: Merge commit '36e3b1f2fd262028834a9d7b1eb533c1218ee6c2' [15:49] <cone-432> ffmpeg.git 03Janne Grunau 07master:f896bca03fc6: aarch64: h264 (bi)weight NEON optimizations [15:49] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:92d07ea4b592: Merge commit 'f896bca03fc63b93851c1c14c9321c20b3cd44a6' [15:54] <cone-432> ffmpeg.git 03Janne Grunau 07master:fe96769bed82: aarch64: port neon clobber test from arm [15:54] <cone-432> ffmpeg.git 03Michael Niedermayer 07master:949adce12516: Merge remote-tracking branch 'qatar/master' [18:07] <cone-668> ffmpeg.git 03Michael Niedermayer 07master:fd334b9b6e89: avformat/mxfdec: fix typo in error check of mxf_edit_unit_absolute_offset() [18:35] <cone-668> ffmpeg.git 03Michael Niedermayer 07release/2.1:eda6effcabcf: Update for 2.1.3 [20:00] <cone-668> ffmpeg.git 03Michael Niedermayer 07fatal: ambiguous argument 'refs/tags/n2.1.3': unknown revision or path not in the working tree. [20:00] <cone-668> Use '--' to separate paths from revisions [20:00] <cone-668> refs/tags/n2.1.3:HEAD: avformat/mxfdec: fix typo in error check of mxf_edit_unit_absolute_offset() [20:30] <cone-668> ffmpeg.git 03Diego Biurrun 07master:dc4d726bab59: build: Add separate hidden config option for the intrax8 code [20:30] <cone-668> ffmpeg.git 03Michael Niedermayer 07master:221dc52d03e1: Merge commit 'dc4d726bab59f349d366916ba01df16289e5c81f' [21:17] <cone-668> ffmpeg.git 03Diego Biurrun 07master:aab40bbfd5a7: x86: dsputil: Simplify xvmc deprecation conditional [21:17] <cone-668> ffmpeg.git 03Michael Niedermayer 07master:6391dec82a97: Merge remote-tracking branch 'qatar/master' [21:23] <llogan> michaelni: for twitter - "FFmpeg 2.1.3 is now available! This fixes a libxavs compilation issue." does that LGTyou? [21:23] <michaelni> llogan, yes assuming ive not added a new bug ;) [21:23] <llogan> we'll fix it in 2.1.4 [21:36] <cone-668> ffmpeg.git 03James Almer 07master:b90c71cc59ed: fate: add WebP Exif metadata test [21:53] <llogan> fateserver...perl. maybe i won't add the header. [22:08] <funman> http://streams.videolan.org/issues/raw-h264/test.h264 [22:08] <funman> I can play it with ffplay or mplayer [22:08] <funman> using lavf from VLC however, av_probe_input_format fails [22:29] <michaelni> funman, vlc calls av_probe_input_format3() with 2261 bytes, ffprobe calls it with 2048 and then re-calls it with more and more until it succeeds at 1mb [22:30] <funman> love that 2048 + 213 [22:31] <funman> https://mailman.videolan.org/pipermail/vlc-devel/2009-September/066721.html [22:31] <nevcairiel> that sure is a hack [22:33] <j-b> no shit? [22:35] <ubitux> BBB: http://b.pkh.me/0001-vp9-x86-simplify-a-few-mc-inits.patch [22:36] <ubitux> i guess i have to do the benchmark? [22:36] <ubitux> i doubt it will make anything visible though [22:43] <ubitux> meh i'm too lazy to add the timings [22:58] <jedir0x> i'm having a problem with avlib where when creating an mp4 with h264 (no audio stream) the file is corrupt, but there were no errors or problems during encoding (just drawing bitmaps to the AVPicture) - mplayer (and other forgiving players) play it just fine - but it wont play on OSX or on my Android device. Any ideas on how to go about debugging this? [23:31] <cone-668> ffmpeg.git 03Michael Niedermayer 07master:1a01dc83434f: avcodec/pgssubdec: Use av_fast_padded_malloc() for rle buffer [23:31] <cone-668> ffmpeg.git 03Michael Niedermayer 07master:b948ab8132e2: avformat/jvdec: clear packet padding after downsize [00:00] --- Thu Jan 16 2014

1 0

ffmpeg.log.20140115
by burek 16 Jan '14

16 Jan '14

[03:33] <jedir0x> I'm having trouble with what i think is setting the proper PTS. My program is simply encoding and writing video frames to a file. I'm basing it on the muxing.c example. I end up with a video that is ~15k frames per second and ~1 second long. Is this the correct way to set pts on the AVFrame?: frame->pts += av_rescale_q(1, videoStream->codec->time_base, videoStream->time_base); [04:45] <albator> somebody to explain me how this command works? " -filter_complex "[0:v]setpts=0.5*PTS[v];[0:a]atempo=2.0[a]" -map "[v]" -map "[a]"" [04:45] <albator> I really don't know how to set this up [05:57] <llogan> albator: take the video stream(s) from the first input and apply setpts filter to double video speed. take the audio stream(s) from first input and increase the tempo by 2. [05:58] <llogan> *by a factor of 2 [06:00] <relaxed> is filter_comples needed with one input? [06:00] <relaxed> complex* [06:04] <relaxed> oh, it's needed for multiple input streams and not necessarily input files. [06:04] <llogan> http://ffmpeg.org/ffmpeg.html#Complex-filtergraphs [06:05] <llogan> i wonder why vf isn't just replaced with filter_complex [06:05] <llogan> but i'm not sure of any downsides of using filter_complex on a simple filtergraph [06:09] <relaxed> I guess the original -vf wasn't well thought out. [06:09] <relaxed> remember when we had to specify external files in the filter chain [06:10] <llogan> remember "vhook"? [06:10] <relaxed> like movie=some_external_files for overlays [06:11] <relaxed> vhook predated -vf [06:12] <llogan> i see users still using movie. so i regurgitate FFmpeg Tip #43 about using -filter_complex [07:29] <jedir0x> somehow all of my video frames look blue [07:29] <jedir0x> i wonder if i'm messing with sw_scale [07:29] <relaxed> sounds smurfy [07:51] <jedir0x> haha, RGBA vs. ARGB [08:08] <k0bi> i am trying to extract jpeg from fms [08:09] <k0bi> ofter running hi i run "C:\ffmpeg" -i "rtmp://ip/room_5280/90 live=1" -s 250x186 -r 6 c:\tmp\5280\frame%d.jpg and it stuck ofter saying "hasAudio FALSE"? [08:09] <k0bi> if hasAudio TRUE it works but if its false it stuck [08:09] <k0bi> am i am doing somthing wrong? [08:10] <relaxed> fms? [08:10] <k0bi> yes [08:10] <relaxed> wtf is that? [08:10] <k0bi> not sure [08:11] <k0bi> i tring to extrack pictuere from rtmp [08:11] <k0bi> its works but somtimes its stuck [08:12] <k0bi> "C:\ffmpeg" -i "rtmp://ip/room_5280/90 live=1" -s 250x186 -r 6 c:\tmp\5280\frame%d.jpg [08:12] <k0bi> thats the command line i am using [08:14] <llogan> yes, that is the command you are using. [08:15] <k0bi> http://pastebin.com/A9y8rwyS [08:15] <k0bi> and thats the result [08:16] <llogan> relaxed: what do you think about adding $HOME/ffmpeg_build/share/man/ to MANPATH in .profile for persistent ffmpeg man action? [08:16] <relaxed> "rtmp server sent error" [08:16] <k0bi> yes but if i run on other room it sed that 2 [08:16] <k0bi> qand it working [08:17] <k0bi> and [08:17] <relaxed> llogan: that would work [08:19] <k0bi> http://pastebin.com/ebjUHsL6 [08:19] <k0bi> you see its the same till the line of hasAudio TRUE [08:19] <relaxed> llogan: look at `man man`, there's ~/.manpath which might be better [08:21] <relaxed> instead of having to source .profile [08:22] <llogan> that seems mo' sane. i'll probably still source .profile for its ~/bin path junk [08:23] <k0bi> <relaxed> you know this [08:25] <relaxed> k0bi: what did you do to make it work? I don't understand the issue. [08:26] <llogan> k0bi: you should use scale filter instead of ss: "-vf scale=250:-1" which will result in 250x188 [08:26] <k0bi> ok i have 2 rooms of fms stream on one room it worknig good but in the other it stuck [08:29] <k0bi> C:\>"E:\wwwroot\base64Player\ffmpeg" -i "rtmp://ip/mygirl/room_4738 [08:29] <k0bi> the same result stuck ofter hasAudio FALSE [08:30] <k0bi> C:\>"E:\wwwroot\base64Player\ffmpeg" -i "rtmp://ip/mygirl/room_4738/23 live=1" -vf scale=250:-1 -r 6 c:\tmp\4738\frame%d.jpg [08:32] <k0bi> i think the problem is becuse of the audio [08:32] <k0bi> cuz if audio is true is worknig [08:32] <relaxed> add -map 0:v [08:33] <__raven_> hi [08:33] <k0bi> just add and its the same [08:34] <relaxed> I don't think this is an ffmpeg problem. [08:34] <k0bi> but the fms work just fine [08:35] <relaxed> WHAT IS FMS [08:36] <llogan> since it's "mygirl/room*" I assume "Flesh Media Server" [08:36] <llogan> I mean "Flash" [08:37] <__raven_> possible to detect the video content for problems such as fog or too dark picture? any thresholds for sharpness or something like that [08:38] <k0bi> as i see it on the server i got 10 rooms in all it work ok just in the rooms that has no audio [08:39] <relaxed> oh, i bet there's some flesh too :) [08:40] <k0bi> sorry? [08:40] <llogan> __raven_: not really. you might be able to figure something out with one of many available filters [08:40] <llogan> http://ffmpeg.org/ffmpeg-filters.html [08:41] <k0bi> ok so you are saying that used of a filter can fixed my problem [08:41] <k0bi> and it not with the audio [08:42] <llogan> k0bi: no. that was directed towards __raven_ [08:42] <llogan> i'm tired. it took me 3 tries to spell "towards" [08:45] <k0bi> <llogan> can you try to halp me [08:47] <llogan> k0bi: i'm streaming ignorant. does VLC play the "broken" inputs ok? [08:48] Action: llogan blames Flesh Media Player [08:48] <k0bi> yhaa it all work good the fms tream the flash play all good [08:49] <k0bi> the problem is if i connect the ffmpeg to extrck picture it stuck on roon with no audio [08:49] <coalado> I'd like to have a ffmpeg build that can be used for mux/demux tasks only. Any Ideas? [08:50] <relaxed> coalado: the download page has links to ffmpeg static builds [08:50] <coalado> relaxed: I know. but they are "too powerfull" [08:51] <llogan> --enable-weakling [08:52] <llogan> k0bi: i don't know. you're using librtmp. maybe the native rtmp protocol support will behave differently. [08:52] <coalado> I do not need any transcoding features - AND I don't want to ship them due to license questions [08:52] <llogan> k0bi: i forgot how to tell ffmpeg to use native instead of librtmp. or maybe i never knew. [08:53] <llogan> k0bi: you might be able to get help on the ffmpeg-user mailing list [08:55] <k0bi> llogan:ok i will try but mybe we can try here as well [08:56] <llogan> coalado: "--disable-everything --enable-muxers --enable-demuxers" but that's probably not what you want [08:57] <coalado> llogan: this all is pretty new to me, so I don't know if this is what I want ;-P [08:57] <coalado> I actually need to merge a dash audio and a dash video stream [08:59] <llogan> the configure options i provided will provide a "ffmpeg build that can be used for mux/demux tasks only" [08:59] <coalado> ok thanks I'll try it [09:02] <llogan> coalado: you will probably also want --enable-protocols or --enable-protocol=file and such [09:02] <llogan> see ./configure --help [09:25] <k0bi> can i put try chtch like in ffmpeg? [11:45] <calcifea_> seems that ffmpeg 2.1 does not use -threads on vp9 [11:48] <relaxed> calcifea_: ffmpeg -h encoder=vp9 2>&1| grep Thre [11:48] <calcifea_> relaxed: threading capabilities are down on every codec [11:49] <calcifea_> -threads is a encoder option [11:49] <calcifea_> it works on vp8. try the same command relaxed, same output [11:49] <relaxed> hmm, you're right [11:50] <calcifea_> i think that ffmpeg just miss that option about vp9 [11:50] <calcifea_> for some reason [11:50] <JEEB> I don't think vp9 encoding has any threading yet [11:50] <calcifea_> it does [11:50] <JEEB> in libvpx [11:52] <JEEB> calcifea_, I searched the git logs for "threading" and I can only see decoding multithreading [11:52] <JEEB> not encoding [11:52] <JEEB> can you specify a git revision it was added with? [11:56] <calcifea_> JEEB: i see. thanks [11:57] <pagioss> hi what is this error and how to resolve Could not find option 'deblockalpha' in any of the FFmpeg subsystems (codec, format, scaler, resampler contexts) [12:21] <pagioss> anyone!? [12:23] <average> pagioss: that's kind-of... [12:23] <average> pagioss: dude, when you want to solve a problem, you present people with some errors or output of FFmpeg [12:23] <average> pagioss: you don't just say "How do I solve X" [12:23] <average> like that gives any sort of useful information [12:34] <Ko_deZ> Hi. I am trying to make a simplistic h264 buffer analyzer and qp analyzer, and trying to avoid decoding the video. We have an encoder supplier that we have some problems with, and I would like to simplify our debugging/testing. I am of course hoping for an easy ride, but if anyone can point me in the right direction, that would help a lot. The data arrives in transport stream, but I can supply it in PES packets, NAL units if nessesary of course. [14:05] <calcifea> how come that if i transcode a VOB with ffmpeg i get weird visual effect in output? [14:05] <calcifea> i see the same issue while playing the VOB itself [14:05] <calcifea> but if i use some ffmpeg front-end i don't get that [14:18] <calcifea> oh [14:18] <calcifea> -pix_fmt yuv420p solved [14:33] <relaxed> calcifea: the vob wasn't in yuv420p? [14:44] <calcifea> relaxed: it didn't fixed [14:44] <calcifea> fix nothing [14:44] <calcifea> i was wrong [14:47] <calcifea> http://pastebin.com/raw.php?i=v3RFv7Ti [18:47] <albator> JEEB just to let you know -filter_complex "[0:v]setpts=1.25*PTS[v];[0:4:a]atempo=0.8[a]" -map "[v]" -map "[a]" for some reasons, audio and video are nearly in sync with these settings [18:50] <DeadSix27> anyone knows why vp9 doesnt seem to work in my build? I used this line to configure libvpx: http://pastie.org/8636309, and also enabled it in ffmpeg configure [18:50] <DeadSix27> according to the output of the libvpx build vp9 should be enabled. [18:53] <DeadSix27> but looking at "-encoders", theres only: V..... libvpx libvpx VP8 (codec vp8) [18:53] <DeadSix27> is there any configuration flag i have to give ffmpeg? [18:55] <DeadSix27> (beside enabling libvpx i mean) [19:01] <thebombzen> configure it and send us your config.log [19:01] <thebombzen> on a paste service [19:01] <thebombzen> DeadSix27: [19:03] <DeadSix27> sec. [19:04] <DeadSix27> thebombzen: http://pastie.org/private/msjngpu8jvdpvbbqwvc5ag [19:04] <DeadSix27> just noticed an error at line 272 [19:04] <DeadSix27> or i guess thats one, idk. [19:05] <DeadSix27> didnt rly touch that, just followed what i rhymed together on SO etc. [19:06] <thebombzen> don't use --prefix that's the install prefix [19:06] <thebombzen> use --cross-prefix [19:06] <thebombzen> that's the prefix where the cross-compile tools are located [19:07] <thebombzen> do ./configure --help and look under "advanced options (experts only)" [19:07] <thebombzen> that has cross-compile info [19:07] <DeadSix27> will do in a sec. [19:08] <DeadSix27> or in 4 hrs, depending how long when i finish, ill come back to you though. [19:08] <DeadSix27> until im finished* [20:28] <sledges> hello [20:28] <sledges> how can I sync a timestamp (drawtext filter) to a framerate? [20:30] <tvkid> i want to compress a raw 640x480 video to a < 100Mbps stream(any format/codec) is it possible to do with a 454MHz ARM(non-cortex) processor with 64MB RAM? [20:32] <tvkid> i mean minimum processing over the raw stream such that it becomes less than 100 Mbps stream so that i can send a single stream over ethernet [20:32] <jnvsor> tvkid: Recording my whole screen raw goes less than 100Mbps, have you run into any problems? [20:35] <tvkid> i was calculating like 640*480*24(bits per pixel)*30 frames per second = 221184000 which is around 221Mbit no? am i doing it right [20:35] <tvkid> jnvsor: ^ [20:38] <tvkid> jnvsor: i am curious how your whole screen raw data takes less than 100Mbit? [20:43] <DeadSix27> thebombzen: http://pastie.org/private/6vcitrltnjok1evgayz6lw [20:56] <jnvsor> tvkid: Ah my bad, try libx264 ultrafast - if that doesn't work I don't know what will [21:24] <thebombzen> tvkid: afaik huffyuv is optimized for fast compression, but it will be hard to read outside of libavcodec land [21:27] <thebombzen> DeadSix27: that's odd, from ./configure --help it says: --cross-prefix=PREFIX use PREFIX for compilation tools []. I'd try putting --enable-cross-compile before everything, but if ./configure doesn't recognize that then I don't know. sorry [21:27] <thebombzen> also tvkid: yuv420p uses half the space of rgb24 [21:28] <DeadSix27> i use git master [21:28] <DeadSix27> also tried git experimental [21:28] <thebombzen> so that would be 110 MBit at 30 fps, which is 88 Mbit at 24 frames per second [21:30] <thebombzen> also it appears that the next release is 2.1.3, whatever happened to 2.1.2? [21:31] <llogan> compile issue involving libxavs [21:35] <DeadSix27> thebombzen: are we sure we talk about the same library, i checked out 1.2.0 1.3.0, and master & experimental? [21:36] <DeadSix27> that option wasnt in any of them [21:36] <DeadSix27> nor did it ever say: advanced options (experts only) [21:36] <DeadSix27> only advanced options [21:36] <DeadSix27> you sure* [21:54] <DeadSix27> well gnight. [22:17] <sledges> figured out: need to use %{pts} and tweak source code to format that as hh:mm:ss:uu ; sadly there's no other easy way to sync the clock [22:19] <drwx> hey, why does ffmpeg report s32 for 24bit flac files? [22:19] <drwx> Stream #0:0: Audio: flac, 96000 Hz, stereo, s32 [22:20] <JEEB> probably because the decoder outputs 32bit for that case [22:20] <JEEB> do you get the same thing when you actually start decoding the stream? if yes, then the decoder just bit-shifts it to 32bit, no data should be lost [22:46] <drwx> JEEB: yeah [23:39] <jedir0x> i'm having a problem with avlib where when creating an mp4 with h264 (no audio stream) the file is corrupt, but there were no errors or problems during encoding (just drawing bitmaps to the AVPicture) - mplayer (and other forgiving players) play it just fine - but it wont play on OSX or on my Android device. Any ideas on how to go about debugging this? [23:47] <bgilbert> Hi, I was wondering if there was a way to transcode rtp packets in ffmpeg using only stdin/stdout? [00:00] --- Thu Jan 16 2014

1 0

ffmpeg-devel.log.20140114
by burek 15 Jan '14

15 Jan '14

[00:15] <funman> 00:15 -MemoServ(MemoServ(a)services.)- mraulet is not registered. [00:15] <funman> damned [00:23] <BBB> ubitux: ofc :) but you can fix 'em too y'know :) [00:24] <BBB> llogan: oh thanks that's a lot of attention for random stuff. the summary doesn't really mention what the rest of the blog is about (that is, it tells what frame-mt is, but not what the rest of the entry is all about) [00:25] <BBB> llogan: but this is totally fine also I guess [00:26] <ubitux> BBB: all (most?) seem related to the threading so... [00:26] <ubitux> you know my opinion on that ;) [00:26] <BBB> great opoortunity to learn all evils of threading [00:26] <BBB> \o/ [00:26] <BBB> or so [00:26] <ubitux> exactly [00:26] <BBB> lol [00:27] <llogan> BBB: i thought it would have been of interest to FFmpeg-geeks, but yeah, i'll just drop the summary [00:27] <BBB> llogan: I think you're right actually [00:27] <BBB> but whichever you think is best [00:27] <BBB> I have no strong opinion either way [00:27] <ubitux> llogan: you could also just tweet it [00:27] <llogan> ubitux: ok. i'm feeling lazy anyway. [00:28] <ubitux> not sure the "ffmpeg news" is relevant for this, it's not about new things in ffmpeg [00:28] <llogan> i'll just tweet it [00:28] <ubitux> you're always looking for things to tweet so.. [00:29] <ubitux> BBB: i've send you the samples in query btw [00:29] <ubitux> anyway, gtg [00:40] <BBB> fuzzed2 doesn't crash for me [00:40] <BBB> valgrind is clean [00:44] <cone-644> ffmpeg.git 03Clément BSsch 07release/2.1:15d7b7d7ccb1: avcodec/libxavs: attempt to fix compilation after b18c7c8d. (cherry picked from commit 71cd83e34cf7ba88d766434e3d2b4d99c14bf0f2) [00:44] <cone-644> ffmpeg.git 03Clément BSsch 07release/2.1:d9b7557732f1: avcodec/libxavs: 2nd attempt to fix compilation after b18c7c8d. (cherry picked from commit 260fc0d95b025b03b2a15116526e4c83b1ca1a31) [00:44] <cone-644> ffmpeg.git 03Michael Niedermayer 07release/2.1:c8d43c22dbb1: avformat/utils/av_probe_input_buffer2: Fix pd.buf_size [00:44] <cone-644> ffmpeg.git 03Michael Niedermayer 07release/2.1:0a055cc62e23: avformat/utils/av_probe_input_buffer2: fix offset check [00:44] <cone-644> ffmpeg.git 03Michael Niedermayer 07release/2.1:2156d9bd7db1: avformat/utils/av_probe_input_buffer2: fix buffer passed to ffio_rewind_with_probe_data() [01:25] <cone-644> ffmpeg.git 03Michael Niedermayer 07release/2.1:7ae6229b97f3: Merge commit '0673ede985a6560e7efb86dab1c58fb7f95ce587' [01:25] <cone-644> ffmpeg.git 03Michael Niedermayer 07release/2.1:bfdfeadf1154: build sys: rename STRIPFLAGS to ASMSTRIPFLAGS [01:25] <cone-644> ffmpeg.git 03Michael Niedermayer 07release/2.1:16e49d85b6bb: configure: remove code that disables striping in the absence of some flags [01:25] <cone-644> ffmpeg.git 03Michael Niedermayer 07release/2.1:45900618ae4d: library.mak: only run asm strip if ASMSTRIP flags are set [01:59] <cone-644> ffmpeg.git 03Michael Niedermayer 07master:05c78f345b62: avformat/utils: av_probe_input_buffer2 decrease difference to libav [01:59] <cone-644> ffmpeg.git 03Andrey Myznikov 07master:b79bccba8012: avformat/librtmp: Fix memory leak if RTMP_ConnectStream() fails [05:16] <michaelni> BBB, about the prores IDCT patchset, ok if i apply it ? (it makes both idcts match and if someone provide more information about the prores fdct / reference samples then we can try to make it closer to that but that should be easier when both idcts match) [05:21] <cone-596> ffmpeg.git 03James Almer 07master:0b54bc24db8b: webp: add support for EXIF metadata chunks [09:32] <ubitux> 4412 decicycles in ff_vp9_loop_filter_h_16_16_ssse3, 4193462 runs, 842 skips [09:32] <ubitux> 3010 decicycles in ff_vp9_loop_filter_v_16_16_ssse3, 4193528 runs, 776 skips [09:32] <ubitux> 3600 decicycles in ff_vp9_loop_filter_h_16_16_avx, 4193621 runs, 683 skips [09:32] <ubitux> 2678 decicycles in ff_vp9_loop_filter_v_16_16_avx, 4193742 runs, 562 skips [09:32] <ubitux> indeed BBB [09:32] <ubitux> not bad :) [09:33] <nevcairiel> how much of avx does that really use? [09:34] <ubitux> just the mov thing i suppose [09:34] <nevcairiel> still xmm i guess? [09:34] <ubitux> yes [09:34] <ubitux> it's not avx2 [09:35] <ubitux> that's pretty nice for a 15-lines diff [09:37] <nevcairiel> :) [09:41] <ubitux> i don't know if i need to check for ARCH_X86_64 or if it's implicit [11:56] <saste> anyone going to review the libvidstab patch? [11:56] <saste> ^ ubitux? [11:57] <ubitux> i didn't plan [11:57] <ubitux> since you were active on it ;) [11:57] <saste> ubitux, no the filter code [11:57] <saste> the configure patch depends on the filter changes [11:59] <ubitux> well except the shitty style it should be fine [12:01] <ubitux> i have on todo list to port that lib so it can fully benefit from the ffmpeg infrastructure [12:01] <ubitux> but it will take some little time [12:23] <ubitux> is there a delay in the seek sometimes? [12:23] <ubitux> ah, my bad [12:45] <ubitux> ok there is a delay. [12:47] <ubitux> http://pastie.org/pastes/8632251/text [12:48] <ubitux> so, is this a wanted behaviour? [12:48] <nevcairiel> sounds like something didnt get flushed [12:48] <nevcairiel> did you flush teh decoder on seek? [12:49] <ubitux> it's supposed to do it automatically [12:49] <nevcairiel> how would the decoder know if avformat changes position? :) [12:50] <ubitux> well, the function is supposed to call the flush [12:51] <ubitux> mmh maybe it only happens when read_seek2 callback is available [12:51] Action: ubitux is confused [12:51] <nevcairiel> it shouldnt happen at all [12:51] <nevcairiel> avformat doesnt know about the decoder [12:51] <nevcairiel> how would it [12:52] <nevcairiel> its not guaranteed that you use the same AVCodecContext that avformat setup, in fact it can be even be problematic to use the smae [12:52] <nevcairiel> so many setups create a copy of it for the decoder [12:53] <nevcairiel> hence, an explicit call to avcodec_flush_context or whatwasitsname is required [12:53] <ubitux> ah.. [12:54] <ubitux> i don't see ffplay doing that [12:54] <ubitux> is it normal? [12:55] <nevcairiel> avcodec_flush_buffers was the name [12:55] <nevcairiel> and i see ffplay calling it [12:55] <nevcairiel> after a seek ffplay queues a flush_pkt [12:55] <ubitux> any idea why it's done in a weird place? [12:56] <nevcairiel> which then causes it to flush later [12:56] <nevcairiel> its a bit of a odd design, but what do i know about ffplay [12:56] <ubitux> mmh [12:56] <nevcairiel> maybe its afraid of a threading thing [12:57] <nevcairiel> and the packet queue is properly locked, but the avcodeccontext is not [12:57] Action: nevcairiel is just guessing here [13:00] <ubitux> ok great [13:00] <ubitux> nevcairiel: thanks a lot :) [13:03] <BBB> ubitux: great work [13:04] <ubitux> you mean the 15-line diff to add avx? [13:04] <BBB> yes [13:04] <BBB> who cares how many lines it is [13:04] <ubitux> :D [13:04] <BBB> this is all about result [13:04] <BBB> you got it 20% and 10% faster, if I count correctly? [13:05] <BBB> that's really good [13:05] <BBB> michaelni: I'm thinking... [13:19] <BBB> michaelni: I don't really have much of an opinion either way, I do worry about moving away from the prores code, but I'm inclined to believe we just took whatever we RE'ed to fit in with the original simpel_idct (minus different scaling, hence it's its own function), so this won't make things worse, just differently unknown and at least the bitexact check goes away in the cpuflags [13:19] <BBB> michaelni: so given that I guess it's fine [13:38] <BBB> ubitux: are you going to do avx for all simd functions? [13:38] <BBB> ubitux: I'm sure idct would be helped by it also [13:39] <Daemon404> isnt avx mostly float? [13:39] <BBB> three-operand instructions works for int also [13:39] <Daemon404> ah ok [13:39] <BBB> but it's still on xmm registers, not ymm, yes [13:39] <BBB> avx2 does ymm-int [13:39] <BBB> (haswell) [13:39] <Daemon404> yeah i know avx2 does [13:39] <Daemon404> but hardly anyone has it [13:40] <BBB> buy me a new laptop and I'll write avx2 [13:40] <Daemon404> ;) [13:40] <BBB> (it has to be a macbook pro, else I won't use it) [13:40] <Daemon404> im thinking of buying a new laptop in feb, but not a mbp [13:41] <Daemon404> the current one i have is retardedly locked by the bios [13:41] <Daemon404> (it disables vt-x) [13:42] <Daemon404> cant decided between an mbp (then wiping os x) or a lenovo ultrabook [13:43] <TimNich> vt-x is disabled in a lot of BIOS's by default [13:43] <j-b> good morning [13:43] <Daemon404> TimNich, no i mean i am unable to enable it without hacking the bios [13:43] <Daemon404> because they want to sell that as a 'feature' in their 'business' laptops [13:44] <Daemon404> consumers dont need vt-x, DUH [13:45] <Daemon404> ill likely get it when im in NY in feb [13:45] <Daemon404> because screw uk kb layouts [13:45] Action: Daemon404 runs [14:10] <ubitux> BBB: feel free to :p [14:17] <BBB> oh [14:17] <BBB> :/ [14:18] <ubitux> BBB: i can do it if you want :)) [14:19] <BBB> holy crap it's $3.3k [14:19] <BBB> but then it must be haswell right? [14:19] <BBB> I hope so [14:20] <Daemon404> http://shop.lenovo.com/gb/en/laptops/thinkpad/x-series/x1-carbon/ <-- considering thi [14:20] <Daemon404> s [14:20] <Daemon404> (us version) [14:20] <BBB> http://store.apple.com/us/buy-mac/macbook-pro?product=ME294LL/A&step=config [14:21] <BBB> and then bigger cpu and bigger hd [14:21] <BBB> =3.3k :/ [14:21] <BBB> ok work time [14:21] <ubitux> BBB: ok, trying avx, will give you some bench [14:22] <BBB> \o/ [14:22] <BBB> feel free to send patches, I bet most will be trivial [14:22] <BBB> also note avx is only xmm, so it won't help for the 4x4 idct [14:22] <BBB> (just in case you want to try :-p) [14:22] <ubitux> :( [14:22] <BBB> we can make a 2x4x4 idct [14:22] <BBB> that can do xmm -> avx [14:23] <ubitux> BBB: btw, can we assume X86_64 with avx? [14:23] <BBB> no sadly [14:23] <ubitux> mmh ok [14:23] <BBB> but just put it under %if ARCH_X86_64 [14:23] <BBB> we did with most assembly so far [14:23] <BBB> I think it's fine [14:23] <ubitux> okay [14:23] <Rodeo> can't seem to go bove $8,700 on that Apple configuration page, no way to break the $10k barrier [14:23] <BBB> also iadst16 is almost done [14:24] <BBB> Rodeo: try corporate consulting [14:24] <BBB> ok work for real now, bye [14:36] <cone-768> ffmpeg.git 03Robert Krüger 07master:4a38eeec38c5: Revert "Revert "vf_yadif: move x86 init code to x86/yadif.c"" [14:36] <cone-768> ffmpeg.git 03Robert Krüger 07master:99c4abfe8862: Revert "Revert "yadif: add parens around macro parameters"" [14:36] <cone-768> ffmpeg.git 03Robert Krüger 07master:61093993668b: Revert "avfilter/yadif: Revert "lavfi: convert input/ouput list compound literals to named objects"" [14:36] <cone-768> ffmpeg.git 03Robert Krüger 07master:194ef56ba7e6: Change license of yadif from GPL to LGPL [14:44] <cone-768> ffmpeg.git 03Diego Biurrun 07master:7151c5d04aed: arm: Use full filenames as multiple inclusion guards [14:44] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:7766c7b7a009: Merge commit '7151c5d04aed3b496c21f713dcb603e2cbdb9c49' [14:51] <cone-768> ffmpeg.git 03Diego Biurrun 07master:46bacb5cc616: x86: Consistently use cpu flag detection macros in places that still miss it [14:51] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:b148a39d55ee: Merge commit '46bacb5cc6169ff5e8e982495c4925467c1d8bb7' [15:00] <cone-768> ffmpeg.git 03Robert Krüger 07master:d8e763fda772: vf_yadif: Relicense from GPL to LGPL [15:00] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:12bc33d7cd04: Merge remote-tracking branch 'qatar/master' [15:10] <ubitux> BBB: http://b.pkh.me/0001-vp9-x86-add-AVX-for-itxfm-and-lpf.patch [15:10] <nevcairiel> its nice when x86inc can do something neat like this for you, aint it [15:11] <ubitux> yeah :) [15:13] <ubitux> if someone offers me an avx2 cpu i don't mind adding more :p [15:14] <ubitux> (i'm too lazy to upload to kierank machine :p) [15:56] <j-b> who is the mxf expert? [15:56] <ubitux> maybe mateo` [15:57] <j-b> mateo`: I got MXF files unsupported in ffplay. Are you interested in getting access to them? [15:58] <Daemon404> have you tried ffmbc? [15:58] <j-b> no. [15:58] <mateo`> j-b: yes [15:58] <j-b> mateo`: it's a new RDD spec (25?) if I understood correctly [15:59] <j-b> Daemon404: no. I'm not interested in this fork. [16:00] <Daemon404> i assume because gplv3 prevents backports [16:00] <mateo`> oh well, i'm not aware anymore of the new MXF specs but i can take a look [16:00] <nevcairiel> I still dont understand how this gpl-upgrade can be a good idea. I can take a LGPL project and make it impossible for them to benefit from my development, that seems totally backwards to me [16:01] <Daemon404> it's *freedom* [16:01] <nevcairiel> it limits my freedom to take ffmbc code and backport to ffmpeg [16:02] <j-b> Daemon404: no. It's because I'm not interested in projects that do not collaborate with the community. [16:02] <michaelni> nevcairiel, you could ask the author if he is ok with backporting the code [16:02] <j-b> Daemon404: it's like the whole "this-is-open-source-but-I-object-to-any-change-to-it" and "I-m-the-maintainer. Don't-touch". [16:03] <nevcairiel> thats not the point, i'm complaining about the license BS, not an actual example =p [16:04] <nevcairiel> anyway back to writing code! [16:08] <wm4> does ffmbc have contributors? or is it a one man project? [16:09] <wm4> hard to tell, because they don't seem to have a git repo [16:10] <Daemon404> j-b, true [16:10] <Daemon404> and agreed [16:10] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:d9e556ebd04c: avcodec/proresdsp & idct: move biasing from after the IDCT into the IDCT [16:10] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:cca2772e1689: avcodec/simple_idct_template: change the idct coefficients so that they match the x86 code [16:10] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:a7ea733b721e: avcodec/simple_idct_template: fix row rounder [16:10] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:4b0cad6596bc: avcodec/simple_idct_template: fix rounding of the special DC case for 10bit [16:10] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:cb613657ee46: avcodec/x86/proresdsp_init: x86 prores IDCT is bitexact again [16:15] <ubitux> michaelni: CLIP_AND_BIAS could be renamed to CLIP :p [16:16] <nevcairiel> it could also be renamed to BANANA [16:17] <ubitux> BANANA is nice as well indeed [16:34] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:b821def9f5ee: avcodec/proresdsp: rename CLIP_AND_BIAS to BANANA [16:35] <ubitux> waaat [16:35] <nevcairiel> haha [16:35] <ubitux> :D [16:35] <ubitux> ahaha [16:35] <nevcairiel> well he didnt actually do it [16:35] <nevcairiel> but good one :D [16:35] <ubitux> :) [16:36] <mateo`> hahaha [16:37] <TimNich> aja.com [16:38] <wm4> lol [16:38] <wm4> the commit message is a lie [16:38] Action: wm4 is disappointed [16:39] <wm4> a valueable opportunity for mocking a bad (?) codec is lost [16:39] <nevcairiel> not sure prores is that bad, its just proprietary, which some might consider bad by design [16:42] <Daemon404> the tech used in it isnt really proprietary [16:42] <Daemon404> i call it fanc jpeg [16:42] <Daemon404> fancy* [17:03] <ubitux> saste: nice thing the pkg config version check btw [17:11] <kierank> j-b: can i have a look at the samples? [17:12] <j-b> kierank: sure. [17:12] <kierank> nevcairiel: the coefficient rescan is quite interesting [17:12] <kierank> allows spatial scalability [17:12] <kierank> in that domain compression isn't everything [17:13] <kierank> j-b: pure speculation but I think we don't have a tag for aac audio in mxf [17:13] <j-b> kierank: video is absent too [17:13] <kierank> oh [17:24] <kierank> j-b: you can also ask wtf this thing is to the ebu guys at fosdem but i doubt they'll know [17:30] <mateo`> j-b: did you provide the sample somewhere ? [17:32] <j-b> mateo`: I'm uploading it [17:33] <mateo`> ok [18:44] <ubitux> j-b: they lack a demuxer, not a decoder [18:46] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:b2ae92110f9e: avcodec/flashsv: check avio_read() return in mov_read_udta_string() [18:46] <j-b> ubitux: ah [19:07] <wm4> ubitux: but that means no animated gifs [19:07] <wm4> the best video codec... not supported [19:10] <ubitux> i think that's actually pretty cool to support it but well [19:26] <JEEB> lol [19:26] <JEEB> intel is distributing ffmpeg [19:26] <JEEB> with the samples of theirs [19:26] <JEEB> (media sdk) [19:27] <nevcairiel> oh right i wanted to check the new sdk [19:27] <nevcairiel> but first need to upload a new release before i have to setup new hosting with google code file uploads going down tomorrow [19:33] <{V}> JEEB, cool [19:34] <JEEB> nevcairiel, I hope you remember your intel account's uname/pw better than I do :P [19:34] <JEEB> the new SDK and the samples need login to download [19:34] <JEEB> and I got locked out when guessing mine [19:35] <nevcairiel> had to change the password on login [19:35] <nevcairiel> but thats all [19:37] <nevcairiel> i wonder whats in their HEVC sdk thing [19:38] <wm4> <nevcairiel> but first need to upload a new release before i have to setup new hosting with google code file uploads going down tomorrow <- huh, already? [19:38] <nevcairiel> well the upload option gets disabled tomorrow iirc, existing files remain accessible [19:39] <nevcairiel> "Google Code will not support creating new downloads starting January 15th, 2014." [19:39] <JEEB> I think the sdk just has a decoder [19:40] <JEEB> also if you look there you'll notice that they will be offering elecard's mp4 demuxer and vanguard's HEVC dec/encoder [19:40] <JEEB> in the future [19:40] <JEEB> (separate to their own HEVC thingy) [19:40] <nevcairiel> yeah i saw [21:20] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:2545182c294f: avcodec/mpegaudiodec_template/mp3on4: check that all channels have been decoded before returnig a frame [21:20] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:afbe8c6a8441: avcodec/mpegaudiodec_template: decode_frame_mp3on4: conceal errors in decoding instead of discarding data [21:32] <cone-768> ffmpeg.git 03Reimar Döffinger 07master:d84bd4650b1b: mxf: Set AV_FIELD_PROGRESSIVE [21:32] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:8721dda1c57f: Merge commit 'd84bd4650b1bed882e2ac1651b4a86206aa0e363' [21:43] <ubitux> mmh we don't have any avx2 code in ffmpeg? [21:43] <ubitux> that's not serious. [21:43] <nevcairiel> its new [21:43] <cone-768> ffmpeg.git 03Luca Barbato 07master:42f9132218ca: mxf: Do not use int to check the seek position [21:43] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:1171ad647ebf: Merge commit '42f9132218ca11a8e9a3c82a175b46bca092113e' [21:43] <nevcairiel> does the x86inc framework even have it yet [21:44] <nevcairiel> guess it does [21:44] <ubitux> yes there are a bunch of optims in x264 [21:44] <ubitux> we just have the flag [21:44] <ubitux> sadness. [21:45] <nevcairiel> at least cpu detection is implemented as well [21:45] <kierank> I have some patches for swscale [21:45] <nevcairiel> you can be the first, today! [21:45] <kierank> one of the bitdepths is broken though [21:45] <ubitux> nevcairiel: oh you're going to send me a cpu? @_@ [21:46] <nevcairiel> hm ...i can give you a shell on one? :D [21:48] <ubitux> that's gonna be painful [21:49] <nevcairiel> low-end cpus are rather cheap though, of course not the most fun to compile on either [21:50] <nevcairiel> hm looks like the lowest end doesnt have avx2 [21:51] <kierank> nevcairiel: there is ccache [21:52] <nevcairiel> i guess [21:53] <nevcairiel> the cheapest cpu with avx2 is the i3-4130 then [21:54] <nevcairiel> around 100¬ [21:54] <nevcairiel> ask saste for SPI funds :P [21:59] <cone-768> ffmpeg.git 03Marton Balint 07master:aa0cb16c15a5: mxf: Fix off by one error in d10 aes3 decoding [21:59] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:d938a013af76: Merge commit 'aa0cb16c15a5b30f78542f18e3fa65de005cf084' [22:09] <michaelni> nevcairiel, Intel Core i5-4570, 4x 3.20GHz is 157¬ and i wont approv funds being used for cheapest CPUs we want devels to have some decent hw [22:11] <nevcairiel> sure [22:13] <cone-768> ffmpeg.git 03Luca Barbato 07master:f5fbbbc022f7: mxf: Drop unnecessary checks [22:13] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:84ca7c28ce49: Merge commit 'f5fbbbc022f723d3ccf99afd5d658a977b51c08a' [22:22] <llogan> nice description... [22:28] <cone-768> ffmpeg.git 03Luca Barbato 07master:1a4e4ad0e0c5: mxf: Use av_malloc_array [22:28] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:e3c0fcaddabe: Merge commit '1a4e4ad0e0c5486dcab05e54b587672a498dd7cf' [22:42] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:0d6605c7ef43: mxf: Fix a possible leak of extradata [22:42] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:bd8788ea17c3: Merge commit '0d6605c7ef43f97a88950542af09078adef33b6d' [22:50] <cone-768> ffmpeg.git 03Tomas Härdin 07master:8b708f1c6b1b: mxf: Correctly support files from Pinnacle Thunder [22:50] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:633214667684: Merge commit '8b708f1c6b1baf3b97ed93226bf5dae1a9b13fb7' [23:09] <cone-768> ffmpeg.git 03Tomas Härdin 07master:cc1e3ace6307: mxf: Fix potential leak in mxf_read_local_tags() [23:09] <cone-768> ffmpeg.git 03Michael Niedermayer 07master:05fac0b45be2: Merge remote-tracking branch 'qatar/master' [23:15] <ubitux> http://pastie.org/pastes/8633851/text [23:15] <ubitux> ok, let's do 8x8. [23:34] <ubitux> mmh perf doesn't provide the most common callers? [00:00] --- Wed Jan 15 2014

1 0

ffmpeg.log.20140114
by burek 15 Jan '14

15 Jan '14

[00:34] <i_s> figured it out, my pixel conversion was wrong [01:02] <thebombzen> sorry 'bout that, testing my IRC client [02:04] <FunkyELF> I have this file... http://pastebin.com/iURM9HZd [02:04] <FunkyELF> I want to encode it with x264 or something to shrink the size. Also, I want to skip the last 3 seconds of the video.... what command line can I use for that? [02:06] <relaxed> FunkyELF: ffmpeg -i input -t 05:50:28 ... [02:07] <FunkyELF> Duration: 00:01:15.18, start: 0.000000, bitrate: 47072 kb/s [02:08] <relaxed> -t 00:01:12 [02:09] <relaxed> it's HH:MM:SS [02:09] <FunkyELF> relaxed: okay... I'm looking at the x264 encoding guide [02:10] <FunkyELF> trying to find a list of all presets [02:10] <relaxed> x264 --fullhelp [02:10] <FunkyELF> As a shortcut, you can also list all possible internal presets/tunes for FFmpeg by specifying no preset or tune option at all: [02:10] <relaxed> or http://mewiki.project357.com/wiki/X264_Settings [02:10] <FunkyELF> but then that one line doesn't work [02:11] <FunkyELF> relaxed: I don't have the x264 binary. I have a static build of ffmpeg [02:11] <relaxed> then check the url [02:11] <FunkyELF> okay... when I use handbrake there is a slider for quality [02:12] <FunkyELF> wondering what the corresponding ffmpeg swich would be [02:13] <relaxed> https://trac.ffmpeg.org/wiki/x264EncodingGuide look at -crf [02:15] <FunkyELF> relaxed: thanks [02:15] <llogan> FunkyELF: "-preset help" should list presets and tunes [02:16] <FunkyELF> llogan: ffmpeg -preset help ? [02:16] <relaxed> doesn't work here [02:16] <FunkyELF> me neither [02:17] <llogan> you replace "slow" or whatever with "help" [02:17] <relaxed> `ffmpeg -preset help` ? [02:17] <llogan> ffmpeg -y -f lavfi -i nullsrc -f mp4 -preset help /dev/null [02:18] <llogan> see. easy. [02:18] <relaxed> that's retarded [02:18] <llogan> you couldn't tell i was joking... [02:21] <relaxed> ffmpeg -h encoder=libx264 would be a good place [02:21] <llogan> it's a libx264 AVoption so "-preset help" has to be invoked via the encoder [02:24] <llogan> https://trac.ffmpeg.org/wiki/x264EncodingGuide#AdditionalInformationTips is a messy section. i'll blame roger [02:25] <llogan> and the iOS chart should include refs since level does not modify refs, IIRC [02:26] <llogan> although maybe that should be considered a bug [02:37] <bencc> are there rules on allowed resolutions? [02:38] <bencc> can I have for example 1000/750 in 4:3 aspect ratio and 992/558 in 19:6 aspect ratio? [02:44] <sacarasc> bencc: Allowed by what? [02:45] <relaxed> I allow it [02:47] <bencc> by ffmpeg and display in media players and in the browser [02:47] <bencc> releaxed: that's because you are relaxed :) [02:48] <sacarasc> Computers generally don't care about resolutions. [02:49] <sacarasc> If you were going for a DVD, then you'd have trouble. [02:49] <bencc> ok. thanks [03:37] <Fusl> uuuh... dunno if some of you guys speak german in here and if you're awake since it is 3:37am in germany but nevermind, http://www.heise.de/newsticker/meldung/Die-tausend-gestopften-Loecher-des-F… [03:38] <Fusl> ah, and here is the english one: http://j00ru.vexillium.org/?p=2211 [03:39] <mocean> Hello, it seems that ffmpeg is only using 1 cpu core. Do I have to do anything specific when converting a file to mp3 to use multicore? [03:40] <Fusl> try -threads X where X is your number of threads you want to start [03:41] <Fusl> but i'm not sure it works with mp3 [03:41] <sacarasc> I don't think you can do multi-cored MP3. [03:41] <sacarasc> You'd have to just run multiple commands. [03:42] <mocean> hmm [03:42] <mocean> so ffmpeg --threads 4 ? [03:43] <Fusl> no, single dash [03:43] <sacarasc> ffmpeg -i input.flac -threads 4 output.mp3 [03:43] <sacarasc> But it won't do anything. :D [03:43] <Fusl> it is not supposed to do anything [03:44] <Fusl> `ps auxwwe` and check if ffmpeg spawns threads there [03:44] <Fusl> ehm [03:44] <Fusl> no [03:44] <Fusl> ps fauxww [03:45] <Fusl> .. and also check if all of them eat up cpu cycles [03:52] <mocean> -threads does not seem to make a difference, only one core is used at 100% [03:52] <sacarasc> [02:41:06] <sacarasc> I don't think you can do multi-cored MP3. [03:52] <sacarasc> [02:41:18] <sacarasc> You'd have to just run multiple commands. [03:53] <mocean> I don't understand that statement [03:53] <mocean> Multi commands to convert an mp3? [03:53] <sacarasc> Yes. With one command, do one, then do another with a different command at the same time. [03:57] <mocean> but it's one file that's over a gig in size [03:59] <sacarasc> You'd probably have it done if you continued single thread when you joined. :D [04:02] <mocean> it was saying over 2 hours till completion [04:04] <Fusl> mocean: how about split them up into 4x 30 minutes parts and then concat them? [04:07] <mocean> I guess that could work [04:08] <mocean> Surprised that multi threads does not work with mp3, people have had multi core cpu's for a long time now [04:08] <Fusl> tbh, i'm surprised too :/ [04:15] <Fusl> mocean: idk if there is another automatic and better way of doing this but this is what i used last time: http://sprunge.us/QiQN [04:25] <mocean> thanks Fusl [04:25] <Fusl> youre welcome :) [06:14] <hayuto> how do i get ffmpeg compiled with all the possible options [06:15] <relaxed> hayuto: do you need all of them? [06:15] <hayuto> i want it for hls streaming my default ffmpeg installed on server gives syntax error in baseline vprofile [06:18] <hayuto> http://pastebin.com/V87TYFUT [06:19] <hayuto> if i delete baseline and acodec it works but it drops frames all the time [06:19] <hayuto> i tried -s 640x480 [06:20] <relaxed> that version is way too old. FreeBSd ports has the most recent version: http://www.freshports.org/multimedia/ffmpeg/ [06:22] <hayuto> hmm i cant install it im not admin [06:23] <hayuto> but i can compile from github [06:23] <hayuto> but theres a lot of options [06:25] <hayuto> what is the option to enable the movie:movie.avi:loop=1 stuff? [06:27] <hayuto> lavfi but theres nothing like that [06:27] <hayuto> ok i try with what i know [06:32] <relaxed> you can compile it locally but with FreeBSD that will likely be a pain. [06:33] <relaxed> which FreeBSD version are we talking about? 9.x ? [06:33] <hayuto> 9.2-STABLE FreeBSD 9.2-STABLE [06:34] <relaxed> ask the admin to upgrade ffmpeg [06:34] <hayuto> yeah i did that [06:34] <relaxed> what did he say? FUUUUUUUUU? [06:35] <hayuto> =D no they are asleep probably noone answered yet [06:35] <hayuto> im on their irc channel [06:35] <hayuto> so im waiting [06:35] <hayuto> but i tought it would be faster to compile from git.hub [06:35] <hayuto> but it isnt so easy [06:36] <hayuto> a lot of things missing on their system or i dont know how point ffmpeg to them or where they are [06:37] <relaxed> ./configure --extra-cflags="-I/usr/local/include" --extra-ldflags="-L/usr/local/lib" [06:38] <relaxed> plus whatever you're trying to --enable [06:47] <relaxed> hayuto: it might be harder if you're in a jail [06:48] <hayuto> why [06:49] <relaxed> you might be limited, I've never tried from within one [06:49] <hayuto> yeah but why i would go to jail [06:49] <hayuto> =o [06:49] <relaxed> a freebsd jail, mr. funny man [06:50] <hayuto> oh [06:51] <hayuto> i think it must be the options they give on hls site [06:51] <hayuto> baseline and this sound codec [06:51] <hayuto> exact same [06:51] <hayuto> cause now im testing without baseline, also no sound with copy [06:51] <hayuto> maybe android only can play this baseline profile [06:52] <hayuto> without it i get terrible framrate [06:52] <hayuto> if you have android its here http://37.59.202.234:13331/hls/ [06:52] <hayuto> i think only chrome beta can play it tough [06:53] <hayuto> ;/ [07:00] <hayuto> k i try with new version when i can [07:00] <hayuto> and will report back [07:39] <llogan> hayuto: you could just try a ffmpeg build from links via ffmpeg download page. should work on *bsd, right? only problem they won't have faac support [07:40] <llogan> but you can use -acodec aac -strict experimental if you need to [07:41] <JEEB> both faac and fdk-aac are nonfree so you are not supposed to be able to distribute binaries with them [08:09] <hayuto> i need to know whats best for html5 and android [08:12] <Fusl> hayuto: which container format? [08:15] <hayuto> it generates hayuto-220.ts hayuto-221.ts hayuto-222.ts hayuto-223.ts hayuto-224.ts hayuto.m3u8 [08:16] <hayuto> application/vnd.apple.mpegURL m3u8; [08:16] <hayuto> video/MP2T ts; [08:18] <hayuto> MPEG transport stream data <- when i do file *.ts [08:20] <hayuto> it kind of make me think if its releted to android not being ready to play full blown streams yet [08:20] <hayuto> or its badly encoded [08:21] <hayuto> the test files on jwplayer6 hls page are played super good and quality but its not real live stream [08:24] <relaxed> hayuto: what's your command? [08:25] <hayuto> this /usr/local/bin/ffmpeg -re -i rtmp://37.59.202.234:13335/live/hayuto -vcodec libx264 -g 10 -ar 44100 -ac 1 -f flv rtmp://37.59.202.234:13335/hls/hayuto [08:26] <hayuto> but nginx generetes the files i dont know [08:26] <hayuto> because i tried /usr/local/bin/ffmpeg -re -i rtmp://37.59.202.234:13335/live/hayuto -vcodec libx264 -vprofile baseline -g 10 -acodec libfaac -ar 44100 -ac 1 -f flv rtmp://37.59.202.234:13335/hls/hayuto [08:26] <hayuto> but it gave error, so now im waiting for new version of ffmpeg [08:28] <hayuto> yeah it actually dont stream anything [08:28] <hayuto> nginx captures it and generates this files [08:28] <hayuto> so you can use them in <video></video> [08:28] <hayuto> the actual stream is the /live/ [08:29] <relaxed> try -vpre baseline instead of -vprofile baseline and add -crf 16 [08:32] <relaxed> pastebin if you have more errors [08:37] <hayuto> ok [08:40] <relaxed> the options have changed so much over various versions that it's hard to remember what worked when. [08:42] <hayuto> it works, still hangs from time to time [08:44] <hayuto> but theres nothing in ffmpeg [08:44] <hayuto> no red text [08:45] <hayuto> it encodes ok from what i see [08:45] <hayuto> oh no theres something, [h264 @ 0x807428020] mmco: unref short failure [08:45] <hayuto> [h264 @ 0x807428020] Missing reference picture [08:46] <hayuto> drop is only 2 [09:06] <xlinkz0> i've used this in my wrapper class : double fps() { return 1.0 / av_q2d(m_codec->time_base) / FFMAX(m_codec->ticks_per_frame, 1); } [09:07] <xlinkz0> however it returns 90k for some video file [09:52] <GT1> Hi guys, I just came back to say thanks for the help and to share my findings [09:53] <GT1> if you remember I had problem with itsoffset not delaying the audio when used with overlay and all. What I ended up was just adding a little -async 1 to the end of the command which fixed my problem. The audio got to its right place and the video wasn't lagging [09:53] <ubitux> cool :) [09:53] <GT1> so If you have similar problem that your audio is not delayed by itsoffset or your overlay image gets lagging, just add this command [09:54] <GT1> :D I'm soooo happy now :D [10:00] <GT1> k, I'll be leaving then :) thanks for the help! [10:17] <Justice> Im trying to stream to twitch however i got some issues. The requirements are frame interval 2 seconds but when using -g 2 and max bitrate to 3M the quality is poor [10:18] <Justice> removing gop will make the quality better however streaming to twitch after 6-10minutes the stream will get distorted [10:19] <JEEB> -g is in pictures [10:19] <JEEB> so you get two-picture gops :P [10:19] <Justice> oh [10:19] <JEEB> also do you properly use maxrate and bufsize? [10:20] <Justice> um using -maxrate 3M and bufsize 3M [10:20] <Justice> I'm [10:20] <JEEB> if those are correct for your use case that's correct [10:20] <Justice> but how can I set Keyframe Interval: 2 seconds ? [10:20] <JEEB> well, how many pictures per second do you have? [10:21] <Justice> im setting 30 FPS [10:21] <JEEB> thus, <seconds>*<pictures per second> [10:22] <Justice> so gop should be -g 2*30 ? [10:22] <JEEB> which is? [10:22] Action: Justice let me bring my calculator up... [10:22] <Justice> haha 60 but yeah [10:23] <JEEB> yup [10:23] <JEEB> that's it [10:25] <Justice> btw if the bitrate does not achive the maxrate setting is there a reason for that? [10:25] <Justice> is that network or hardware or both? [10:27] <Justice> JEEB: http://codepad.org/MstZ1u3m [10:27] <Justice> should be right ? [10:29] <JEEB> maxrate and bufsize only limit the bandwidth (of video, methinks) [10:29] <JEEB> since I think they set the vbv settings in libx264 [10:30] <JEEB> and yes, that looks relatively OK, I would personally build ffmpeg with fdk-aac myself tho and use it to encode AAC [10:30] <JEEB> also you're using crf, nice [10:31] <Justice> hm lets see if this works stream is on 7th minute hopefully it wont corrupt [10:33] <JEEB> if you get corruption it's likely due to something else [10:33] <Justice> any ideas what? [10:33] <Justice> Playing with the settings does change how fast the corruption occur though [10:34] <Justice> before it would just show up at 3minute mark playing with rate and cfr incresed it to 6/11minutes [10:34] <JEEB> sounds like you're using too much bandwidth or something? [10:34] <Justice> that or it was just coincidence [10:34] <JEEB> I really don't know :P [10:35] <JEEB> since you're using a restreaming service it should set the limits for bandwidth [10:35] <Justice> i got 250Mbit upload so cannot say it is netowrk issues so maybe twitch thing or my Asus router crying =) [10:35] <BtbN> which platform are you using? [10:35] <BtbN> twitch is one huge network issue [10:36] <Justice> platform as OS or? [10:36] <BtbN> yeah [10:36] <Justice> Linux [10:36] <Justice> Arch to be specific [10:36] <BtbN> hm, so no OBS for you yet [10:36] <JEEB> well, look for the bandwidth limitations on your service is all I can say :P [10:37] <JEEB> also keep an eye on the encoding rate, to make sure you're not getting too slow at any point [10:38] <Justice> meh corruption [10:38] <Justice> have a look http://www.twitch.tv/commanderalchemy [10:39] <JEEB> yeah, just make sure you're actually setting your VBV limits properly, and then make sure your encoding picture rate does not drop while you stream [10:41] <Justice> encoding rate you meaning fps or? [10:41] <JEEB> yes, ffmpeg should be keeping the number up updating [10:42] <Justice> yeah both the fps and bitrate keeps at 30fps and 1100k [10:44] <Justice> hm must be twitch or something else then? [10:53] <Fusl> someone knows if i can set the thickness of showwaves dots? [11:00] <BtbN> Justice, twitch is extremely sensitive with their video. Do they show it as Excelent quality? [11:01] <Justice> BtbN: Exelent quality? [11:01] <BtbN> yeah, if they don't they also explain why not [11:01] <Justice> where does that say that? [11:02] <BtbN> on your streamer page [11:03] <Justice> BtbN: hm cannot seem to find anything like that, um channel page or ? [11:04] <BtbN> your streamer page, not the channel itself [11:06] <Justice> I must be blind cannot really see it.... [11:06] <GT1> rofl, hi I'm back again :( [11:06] <GT1> I have a strange notice from ffmpeg [11:07] <GT1> TIFF tag type (0) is not implemented. Update your FFmpeg version to the newest one from Git. If the problem still occurs, it means that your file has a feature which has not been implemented.*** [11:07] <GT1> any idea? [11:08] <Fusl> can ffmpeg also do such visualizations like the vlc spectrum one? http://imgur.com/Voe3zFo [11:08] <Justice> BtbN: you sure its not an exclusive partner thing? [11:09] <BtbN> no, iz'd noz. [11:09] <BtbN> it's not [11:09] <BtbN> it's huge label on your streamer page. It's either green, saying "Excelent" or something else, telling you what's wrong. [11:10] <GT1> Anyone have an idea why can't my ffmpeg interpret android camera made image? TIFF tag type (0) is not implemented. Update your FFmpeg version to the newest one from Git. If the problem still occurs, it means that your file has a feature which has not been implemented.*** [11:10] <Justice> BtbN: there I found it [11:10] <Justice> BtbN: its on dashboard [11:10] <BtbN> yeah, the streamer page [11:11] <Justice> odd it complains about keyframe interval and bitrate [11:11] <BtbN> twitch needs an absolutely constant bitrate, and exactly 2 seconds keyframe interval [11:11] <Justice> BtbN: it should be [11:12] <GT1> ubitux: are you here? [11:12] <Justice> BtbN: http://codepad.org/MstZ1u3m those are my settings [11:12] <ubitux> i'm hiding in the dark [11:12] <ubitux> Fusl: yes, showspectrum filter [11:12] <GT1> :)) have you ever encountered something like this? [mjpeg @ 0x404a8230] TIFF tag type (0) is not implemented. Update your FFmpeg version to the newest one from Git. If the problem still occurs, it means that your file has a feature which has not been implemented.*** [11:12] <Fusl> ubitux: the right bottom one is showspectrum but that's not nearly the same spectrum (left one) as in vlc [11:13] <BtbN> Justice, that's not constant bitrate, as far as i can see [11:13] <ubitux> Fusl: we don't have a mode like that, patch welcome i'd say [11:13] <GT1> do I need any special build configuration? [11:13] <ubitux> GT1: dunno [11:13] <ubitux> open a ticket [11:14] <Justice> BtbN: this should help no ? "-minrate 3M -maxrate 3M -bufsize 3M" [11:14] <BtbN> that's not CBR mode [11:14] <Fusl> any alternatives? i've read about some gstreamer-tools visualizations which i could use but i need some headless utils which convert via cpu (like ffmpeg) [11:14] <Justice> BtbN: all documentation I found is that there is no CBR mode [11:15] <BtbN> x264 does have one, so ffmpeg should forward that somehow [11:15] <Justice> "There is no native CBR mode, but you can "simulate" a constant bit rate setting by tuning the parameters of ABR:" [11:15] <ubitux> Fusl: you could add a mode to showspectrum :p [11:15] <GT1> ubitux: Silly question :P how to add ticket? [11:16] <ubitux> GT1: http://trac.ffmpeg.org/ [11:16] <Fusl> ubitux: yeah, combined/separate which still don't do the same as vlc's spectrum does :( [11:19] <Justice> BtbN: acording to https://trac.ffmpeg.org/wiki/x264EncodingGuide#FAQ [11:20] <Justice> there is no CBR native so min max rate is the onlyway [11:37] <Justice> BtbN: seems that the stream is exelent until some time and then it gets corruption and it tells me that the keyinterval is at 116.0 seconds even though nothing has changed on my end [11:39] <allengreen> can vlc's slider works for rtsp url? [11:42] <Fusl> can someone tell me what i've done wrong here? http://sprunge.us/GSHZ [11:42] <Fusl> i'm trying to overlay the showwave effect with the avectorscope effect [11:43] <Mavrik> Fusl, em [11:43] <Fusl> Mavrik: yes? [11:43] <Mavrik> you're saying "split audio to two channels and name them [out1] and [a]" [11:43] <Fusl> OH [11:43] <Mavrik> and then you're saying "take stuff from [b] ..." [11:43] <Mavrik> and ffmpeg doesn't like that :P [11:43] <Fusl> yeah, i forgot to change "2" to "3" [11:47] <Fusl> hmm... the avectorscope effect seems to generate a black background which completely covers the showwaves filter ... can i change the background of avectorscope to transparent or at least filter out the black color to transparent with another effect? [11:49] <ubitux> mmh i don't remember if we finally have such filter [11:49] <ubitux> saste: do we have that color to transparency filter? i don't remember :( [11:51] <saste> ubitux, color to transparency?? [11:52] <Fusl> saste: i want to overlay 2 generated video streams where the foreground one has a black background-colour which i want to remove so that the background-stream comes through [11:52] <saste> Fusl: how do you generate the background? [11:53] <saste> -> foreground [11:53] <Fusl> /usr/local/bin/ffmpeg -i mix.mp3 -ab 320k -ac 2 -acodec libmp3lame -ar 48000 -vcodec libx264 -b:v 5M -filter_complex "asplit=3[out1][a][b]; [a]avectorscope=s=1920x1080:zoom=4:rate=24:mode=lissajous_xy,format=yuv420p[scope]; [b]showwaves=s=1920x1080:mode=line[waves]; [waves][scope] overlay [out0]" -map [out0] -map [out1] -preset ultrafast -r 24 -s hd1080 -threads 16 -f avi mix2.avi [11:53] <saste> avectorscope? [11:53] <saste> ideally we should have an option to set the foreground color [11:54] <saste> ahhh!! -> the background color [11:54] <saste> so you can make it completely transparent [11:57] <Fusl> however, i also tried "blend=all_mode=overlay:all_opacity=0.7" instead of "overlay" but that doesn't work :/ [12:00] <saste> Fusl, did you read my comment? would an option setting the background color fix the issue? [12:00] <Fusl> tbh, i really don't know :S [12:01] <Fusl> how can i test if avectorscope really outputs a black and not a transparent background? [12:02] <GT1> ubintux: this is what I get [12:02] <GT1> http://pastebin.com/fkT6CFyP [12:03] <GT1> and this is the image I try to encode [12:03] <GT1> https://www.dropbox.com/s/nat0eyagy76tq06/input.jpg [12:04] <GT1> can somebody test if he can encode this image? [12:17] <Fusl> saste, ubitux: did it! i had to format the avectorscope effect with "yuva444p" and "lut"(?) it with "c3=128" (50%): /usr/local/bin/ffmpeg -i mix.mp3 -ab 320k -ac 2 -acodec libmp3lame -ar 48000 -vcodec libx264 -b:v 5M -filter_complex "asplit=3[out1][a][b]; [a]avectorscope=s=1920x1080:zoom=4:rate=24:mode=lissajous_xy,format=yuva444p,lut=c3=128[scope]; [b]showwaves=s=1920x1080:mode=line[waves]; [waves][scope]overlay[out0]" -map [out0] -map [out1] -preset [12:17] <Fusl> ultrafast -r 24 -s hd1080 -threads 64 -f avi mix2.avi ( http://dericed.com/2012/display-video-difference-with-ffmpegs-overlay-filte… ) [12:29] <Fusl> why isn't ffmpeg using all 64 cores here, althought -threads 64 is set? it only uses 1 core which is the reason it takes a few hours to convert those files O.o [12:29] <BtbN> not all codes support parallelizing [12:29] <BtbN> *codecs [12:30] <Fusl> which do? [12:30] <Fusl> (i need to upload that video later on youtube) [12:58] <theholyduck> Fusl, filtering probably isnt multithreaded (just an educated guess) [12:58] <theholyduck> Fusl, x264 is multithreaded, but, last time i checked, it only supported up to 16 threads? [13:03] <JEEB> no, it can go over that [13:03] <JEEB> it depends on the settings of course, and generally you start getting bottlenecks somewhere else in the chain [13:10] <akovia> Hi, for some reason I can't get "pad" to work. Been following the examples here http://www.ffmpeg.org/ffmpeg-filters.html#pad, but I always get the error "Unable to find a suitable output format for 'pad=1920:1080:0:12:black' [13:11] <akovia> here is my command [13:11] <akovia> ffmpeg -i Common-Poster-Errors.mkv pad="1920:1080:0:12:black" ~/Videos/Screencasts/Poster-Errors.mkv [13:11] <akovia> oh very sorry :( [13:13] <akovia> http://pastebin.com/K323GYJE [13:19] <ubitux> akovia: missing -vf btw. [13:19] <ubitux> but don't use that tool [13:20] <akovia> tyty! [14:18] <Justice> is there a reason on why max keyframe interval keeps increesing even though i have set a limit with -g ? when streaming? [14:18] <Justice> is that packet drops or on my side? [14:26] <JEEB> Justice, there shall be no GOPs longer than what you set with -g [14:32] <Justice> hm looking at docs it seems it only sets the minimum interval ? [14:32] <JEEB> no [14:32] <JEEB> maximum [14:32] <JEEB> GOPs can be longer than -g [14:32] <JEEB> uhh [14:32] <JEEB> shorter [14:33] <JEEB> but they can only go as long as -g [14:33] <Justice> okay hm found also -keyint_min [14:33] <Justice> guessing thats the min [14:33] <JEEB> yes [14:33] <JEEB> but you shouldn't touch that [14:33] <JEEB> in general only the maximum counts [14:33] <JEEB> the minimum with x264 is generally same [14:34] <JEEB> *sane [14:34] <JEEB> goddamnit [14:34] <JEEB> what's with the typos [14:34] <Justice> Thing is im streaming to twitch and after a while it gets distorted and twitch tells me its getting keyinterval att 100 instead [14:34] <JEEB> then you're pushing something wrong there, and if you've set -g then it's something else [14:35] <JEEB> because -g is the maximum, there cannot be a longer GOP [14:35] <Justice> can it change because of packetloss or something? [14:35] <Justice> here is the script I use, http://pastebin.ubuntu.com/6750469/ [14:35] <Justice> just added min to check if that helps but... [14:39] <albator> hello [14:39] <albator> I d like to play my videos slightly slower than normal speed , and I am trying this [14:40] <albator> -filter:a "atempo=0.85" -filter:v "setpts=1.25*PTS" [14:40] <albator> but I get audio "asynch" [14:40] <albator> what can I do? [14:40] <JEEB> Justice, remove the min :P x264's default is sane. And yes, if it doesn't read all of your packets then of course it can miss an IRAP [14:40] <JEEB> which is why I noted that you should check the bandwidth limits from the restreamer [14:42] <Justice> well im streaming from my desktop and my network is 250mbit/s upload so cannot see that, just checked the speed to few servers so its sane there. [14:42] <JEEB> no I mean [14:42] <JEEB> whatever the RESTREAMER wants [14:42] <JEEB> and will be giving out [14:43] <Justice> restreamer? [14:43] <JEEB> you -> restreamer -> public [14:43] <JEEB> capisci? [14:43] <Justice> yes now I do thing is that wtich tells you what is wrong with stream if something is wrong [14:43] <JEEB> no just goddamn read the docs [14:43] <JEEB> they have to say the maxrate and amount of buffer somewhere [14:43] <JEEB> and that's one of the most common problems, VBV [14:44] <albator> yes it s on the broadcasting requirement... [14:44] <albator> but they are idiots... [14:44] <Justice> 3M is their recomended for 720p [14:44] <JEEB> recommended what [14:44] <albator> https://help.justin.tv/entries/25128411-Broadcast-Requirements [14:44] <Justice> bitrate etc [14:44] <JEEB> argh [14:44] <JEEB> seems like you still don't understand [14:44] <JEEB> even if someone tells you 3M that just means that the maxrate is that [14:44] <JEEB> you have no idea of buffer size then [14:45] <Justice> they want CBR so min and max should be the same, and i read that they want a buffer thats equal or higher than bitrate [14:45] <JEEB> they don't want CBR [14:45] <JEEB> real CBR does not exist in video world except for very weird circuimstances [14:46] <Justice> "Mode: Strict CBR" [14:46] <JEEB> I WAS JUST GETTING THERE GODDAMNIT [14:46] <Justice> thats on their requirements page [14:46] <JEEB> LET ME WRITE [14:46] <JEEB> they just write CBR because many apps call VBV-limited ABR like that [14:46] <Justice> o.O [14:46] <JEEB> yes [14:47] <albator> btw , any idea on my slowing speed issue? (sorry to ask) [14:47] <Justice> JEEB: but they complain when the avg bitrate and maximum bitrate is differing so it must be CBR no ? [14:47] <JEEB> no [14:47] <JEEB> they want VBV-limited ABR that is within the VBV limitations [14:48] <JEEB> albator, gotta love it how they say NOTHING about VBV. They just expect that the goddamn tools the user uses already has a preset for the service [14:48] <JEEB> fuck that [14:48] <JEEB> with a large wrench [14:48] <albator> ofc [14:49] <albator> but this is mainstream.. when u use ffmpeg with command line, you are on your own [14:49] <albator> and u end up here :D [14:49] <JEEB> they could at least note their buffering model :P [14:49] <JEEB> fuckers [14:49] <Justice> meh its like tripple buffering on vsync i guess then [14:50] <Justice> they just call it what they want even though its not it... [14:50] <JEEB> generally you should at the very least get the maxrate and the amount of buffering in either seconds or bits [14:50] <JEEB> from seconds you can basically calculate the bufsize with <seconds>*<maxrate> [14:51] <JEEB> and that should be it, but if the fuckers don't give out the info... :V [14:51] <JEEB> anyways, do also really check that you're encoding fast enough so the encoder is not dropping behind [14:52] <Justice> well the fps is stable att 30fps [14:53] <Justice> only that is not getting up enough is the bitrate that is mostly at 1.5M when streaming [14:54] <Justice> hm i wonder what settings OBS uses for twitch since its opensource it should be somewere maybe git [14:56] <albator> btw, i broadcast there and I dont use any of those settings you have.. like buffer keyint etc [14:56] <albator> maxrate.. [14:56] <albator> just saying [14:56] <Justice> I tried without [14:56] <Justice> but after a while i get distortion [14:57] <albator> on the bottom of the picture? [14:57] <Justice> no whole [14:57] <albator> hm [14:57] <Justice> like the pixels are not updated enough [14:57] <albator> remembers me vlc [14:57] <JEEB> albator, you need to have VBV or you are DoingItWrong [14:57] <albator> and jtvlc [14:57] <JEEB> as in, maxrate and bufsize [14:58] <JEEB> otherwise the encoder does not limit itself to any bandwidth [14:58] <albator> i just setup the crf:v [14:58] <JEEB> yes, exactly [14:58] <albator> yes and I dont need to [14:58] <JEEB> eh [14:58] <JEEB> so you don't want to set any minimum bandwidth that your stream can be watched with? [14:58] <JEEB> cool story bro [14:59] <JEEB> basically, under certain circuimstances it _can_ work [14:59] <albator> well I m reasonable with the crf that's all [14:59] <albator> yes it does if your files are similar [14:59] <Justice> what cfr are you using? [14:59] <JEEB> it does not work, it just /happens/ to work [14:59] <JEEB> seriously, start using VBV [14:59] <albator> for 480p mostly 26 , 30 for 720p [14:59] <JEEB> if your rate is less than the limits, it will not do anything [14:59] <JEEB> but if you suddenly get changes, maxrate/bufsize will limit the rate [14:59] <albator> yes I understand [14:59] <JEEB> no, you do not [15:00] <JEEB> as long as you don't use it you don't understand [15:00] <albator> I understand it s to avoid peak in the video bitrate [15:00] <JEEB> no, it is to make sure that average bit rate never goes over maxrate within the buffer [15:00] <JEEB> you can have peaks [15:01] <JEEB> the whole idea of VBV is to enable safe peaks [15:01] <albator> but i am not using buffer [15:01] <Justice> JEEB: so to make it vbv just change -maxrate and minrate to vbv-maxrate and same for minrate ? [15:01] <albator> donno how they manage that on their side [15:01] <JEEB> Justice, -maxrate and -bufsize are the VBV options [15:01] <JEEB> albator, THEY DON'T [15:01] <JEEB> and you are doing it fucking wrong [15:01] <JEEB> seriously [15:02] <JEEB> add maxrate and bufsize according to what the restreamer requests [15:02] <albator> they dont request I think [15:02] <JEEB> oh they do [15:03] <albator> btw JEEB, do you know how to slow down a video in synch with audio? [15:03] <JEEB> now, seriously, please do not spread non-VBV limited streaming setups. People already have very little info on VBV even though it's a simple thing. [15:03] <albator> sorry to ask again :P [15:03] <JEEB> albator, modify the PTS? [15:03] <albator> -filter:a "atempo=0.85" -filter:v "setpts=1.25*PTS" [15:04] <albator> but I get audio out of sync [15:04] <JEEB> I have no idea about the audio part [15:04] <JEEB> sorry [15:04] <JEEB> also [15:04] <JEEB> 0.85 and 1.25? [15:04] <JEEB> shouldn't that be 0.75? [15:04] <albator> audio i think 1= normal speed [15:04] <albator> PTS = not sure [15:05] <JEEB> PTS is just a value, it will change depending on where the picture is in the video [15:05] <JEEB> it's not a tempo [15:05] <albator> makes it hard to setup [15:05] <JEEB> but yes, if you x1.25 all PTS values then you will make the clip longer [15:05] <JEEB> and pictures change slower [15:05] <Justice> JEEB: what should i change in the script to make it vbv, im looking all round google but everything is conflicting with eachother.. [15:05] <albator> yes [15:05] <albator> that's the point [15:05] <JEEB> Justice, just -maxrate and -bufsize [15:06] <JEEB> those are the VBV settings [15:06] <JEEB> as I said already [15:06] <viric> can the pts be used to change A-V offset? [15:06] <albator> I also put audio slower, but it's not sync.. [15:06] <Justice> JEEB: should i have CRF ? [15:06] <JEEB> Justice, that's the main rate control mode, VBV has nothing to do with that [15:06] <JEEB> you can use it with everything but the constant quantizer mode, which you should not use anyways [15:06] <JEEB> (only meant for development purposes) [15:07] <Justice> I see [15:07] <JEEB> basically you use some rate control X and then limit it to your VBV limits with maxrate/bufsize [15:07] <JEEB> albator, as I noted are you sure the number on the audio side is correct? [15:08] <JEEB> if you did 1.25 wouldn't it be 0.75? [15:08] <JEEB> not 0.85 [15:08] <albator> To double the speed of audio: [15:08] <albator> ffmpeg -i input.mkv -filter:a "atempo=2.0" -vn output.mkv [15:08] <albator> The atempo filter is limited to using values between 0.5 and 2.0 (so it can slow it down to no less than half the original speed, and speed up to no more than double the input). [15:08] <albator> so 1 is normal speed I think [15:09] <albator> 1.25 25% faster [15:09] <albator> i guess.. [15:10] <albator> oh I see what you mean [15:10] <albator> let's try ;) [15:12] <albator> still not good.. [15:13] <JEEB> no idea then :P [15:14] <albator> yep [15:14] <albator> maybe I can try using the -r input value [15:14] <albator> tuning a bit [15:26] <nahkunt> hi. i used ffmpeg to convert some mp4s to gifs. the problem is the gif filesizes are huge, about 16x larger [15:26] <nahkunt> anything i can do about that? [15:32] <ubitux> nahkunt: is your ffmpeg recent enough? [15:33] <ubitux> if so, try with -pixfmt rgb24 [15:33] <ubitux> -pix_fmt* [15:35] <nahkunt> ok [15:36] <nahkunt> ubitux: same size [15:36] <nahkunt> it said Incompatible pixel format 'rgb24' for codec 'gif', auto-selecting format 'pal8' [15:37] <ubitux> you're problably not up-to-date [15:38] <nahkunt> i'm using 2.1.1-3 [15:38] <nahkunt> should i switch to git? [15:42] <ubitux> should be good enough [15:42] <ubitux> can you do the paste? [15:48] <nahkunt> ubitux: http://ix.io/9Rm [15:49] <ubitux> mmh [15:49] <ubitux> rgb8 maybe? [15:51] <nahkunt> ubitux: great! now the gif file is only 6.66x the size of the mp4 [15:51] <ubitux> sounds legit [15:51] <nahkunt> reckon it would be possible to go lower? [15:51] <ubitux> you can try playing with the flags [15:52] <ubitux> try -gifflags -transdiff [15:52] <ubitux> sometimes it's bigger [15:52] <ubitux> sometimes it's smaller [15:52] <ubitux> depending on the source [15:53] <nahkunt> well thank you sir [15:54] <ubitux> it won't be as effective as other tools anyway [15:54] <ubitux> but it should be good enough for most cases [15:54] <nahkunt> ubitux: maybe i should use ffmpeg to convert the mp4s into pngs and then use imagemagick to create gifs? [15:54] <ubitux> maybe [15:54] <ubitux> or improve our encoder :p [15:55] <ubitux> but you would probably need a 2-pass mode [16:11] <k0bi> hi i run "C:\ffmpeg" -i "rtmp://ip/room_5280/90 live=1" -s 250x186 -r 6 c:\tmp\5280\frame%d.jpg and it stuck ofter saying "hasAudio FALSE"? [16:13] <k0bi> any idea? [16:49] <k0bi> any one a live here? [16:53] <jnvsor> k0bi: Nope [16:56] <k0bi> hi <jnvsor> [16:56] <k0bi> i am trying to extract jpeg from fms [16:57] <k0bi> ofter running hi i run "C:\ffmpeg" -i "rtmp://ip/room_5280/90 live=1" -s 250x186 -r 6 c:\tmp\5280\frame%d.jpg and it stuck ofter saying "hasAudio FALSE"? [16:57] <k0bi> if hasAudio TRUE it works but if its fals it stuck [16:57] <k0bi> any idea? [17:03] <jnvsor> k0bi: Sorry, you want to extract a screenshot from the video stream? [18:02] <mankyd> Howdy! I am attempting to dump the frames from a 32 second long video, but I'm only getting 55 frames. I'm wondering if someone can help. [18:03] <mankyd> The images in the frames cover the entire extent of the video (its a screen record with lots of redundant frames), but I need fine-grained timing resolution so that I can tell when in the video each frame occurs [18:03] <mankyd> the command is simply: avconv -i speedindex.mp4 -f image2 image-%5d.jpg [18:04] <brontosaurusrex> !avconv [18:07] <mankyd> Please excuse my ignorance, brontosaurausrex, are you implying that this is not the place to ask about avconv? [18:07] <mankyd> I originally used ffmpeg, it told me "This program is only provided for compatibility and will be removed in a future release. Please use avconv instead." [18:08] <mankyd> I received the same problem when I used ffmpeg, of course [18:08] <Mavrik> no, this isn't the place to ask about avconv. [18:15] <brontosaurusrex> mankyd, what os are you on? [18:16] <Mavrik> if he's getting that message he's on Ubuntu 12.04, Debian stable or equivalent [18:17] <brontosaurusrex> or debian perhaps? [18:36] <albator> is it possible to have an idea, for example, how much setpts=2.0*PTS is actually slowing down? [18:36] <albator> and what is the value that doesn't change the speed at all [19:01] <jnvsor> albator: PTS is the time value of the input, 2*PTS is twice that meaning the video will be half speed [19:02] <albator> thx [19:02] <albator> so 0.5 is 2x faster [19:02] <JEEB> yup [19:04] <albator> i m trying the filter complexe example but no success atm.. [19:04] <albator> -filter_complex "[0:v]setpts=0.5*PTS[v];[0:a]atempo=2.0[a]" -map "[v]" -map "[a]" output.mkv [19:18] <coalado> I'd like to have static ffmpeg builds for linux/mac/windows to avoid codec licensing problems (mpeg la etc.) Actually, I only need the option to mux/demux video and audio files. [19:19] <coalado> Any tipps how to do this with as less effort as possible? [19:35] <grkblood13> I'm trying to transcode a source as close to realtime as possible but it's taking ffmpeg about 5 seconds before it outputs anything. my command is `input_source | ffmpeg -f s16le -ar 8000 -ac 1 -i pipe:0 -acodec pcm_s16le -ar 22050 -ac 1 -bufsize 352k pipe:1` [19:49] <grkblood> apologies for leaving before, frrenode locked up [19:52] <grkblood> I'm trying to transcode a input source as close to real time as possible but for whatever reason ffmpeg is waiting atleast 5 seconds before it outputs anything. my cmd is `input_source | ffmpeg -f s16le -ar 8000 -ac 1 -i pipe:0 -acodec pcm_s16le -ar 22050 -ac 1 -bufsize 352k pipe:1` [19:53] <grkblood> since my bufsize is currently set to the same size as 1 second of samples, shouldn't it start outputting the transcoded data after 1 second and not 5? [19:53] <SmokedCheese> sup guys [19:53] <SmokedCheese> I have a video which is chunked and goes live piece by peace. Can I stream it with ffmpeg? [19:54] <SmokedCheese> *by piece [19:59] <jnvsor> grkblood: Are you getting a "Do you want to overwrite this file" prompt? [19:59] <jnvsor> grkblood: add a -y flag to auto overwrite it and that should stop the delay [20:00] <grkblood> jnvsor: delay is still there with a -y [20:01] <jnvsor> grkblood: got a pastebin? [20:02] <grkblood> it outputs the configuration and libs, hangs for about 5 seconds, then starts Input #0, s16le, from 'pipe:0': [20:02] <grkblood> unfortunately I don't, its on an isolated system and I can't get the pastebin across networks [20:03] <jnvsor> grkblood: Well, you can either hand type all the stuff that's in there or find a way to copy it over :P [20:09] <grkblood> jnvsor: weird, I ran the same command on another machine in which I can paste the output to you and it ran almost instantaneously [20:10] <grkblood> https://gist.github.com/grkblood13/fcf8195e373971ef2669 [20:11] <grkblood> however, on the box that has the delay the input is a live feed and not a wav file [20:12] <grkblood> same commands and same ffmpeg build, just different input sources [20:26] <grkblood> ok, I've recreated a test case, the following command takes about 6 to 7 seconds before it starts outputting the transcoded data: wget -O - http://nprdmp.ic.llnwd.net/stream/nprdmp_live01_mp3 | ffmpeg -f s16le -ar 8000 -ac 1 -i pipe:0 -acodec pcm_s16le -ar 22050 -ac 1 -bufsize 352k -f wav -y test.wav [20:30] <brontosaurusrex> grkblood, do you need to use that wget and pipe? [20:30] <grkblood> wget no, pipe yes [20:31] <grkblood> heres the output for that last cmd https://gist.github.com/grkblood13/a4d093dc8df83ad92c91 [20:32] <brontosaurusrex> uhm, so this are two inputs? [20:32] <brontosaurusrex> one is named pipe and another a pipe [20:32] <grkblood> brontosaurusrex: here's with no pipe or wget: ffmpeg -f s16le -ar 8000 -ac 1 -i http://nprdmp.ic.llnwd.net/stream/nprdmp_live01_mp3 -acodec pcm_s16le -ar 22050 -ac 1 -bufsize 352k -f wav -y test.wav [20:32] <brontosaurusrex> and same delay? [20:33] <grkblood> same delay occurs with that command [20:33] <brontosaurusrex> and without -bufsize? [20:33] <grkblood> that link is a NPR stream btw [20:33] <grkblood> same without bufsize [20:34] <brontosaurusrex> my guess is that ffmpeg is probing the input for too long [20:35] <grkblood> so how do we work around that? [20:36] <brontosaurusrex> not sure exactly, but probably by telling the ffmpeg somehow what that stream is about [20:37] <brontosaurusrex> but if do something like "ffmpeg -i intranet/file.mp4 test.mp4" the delay will vary [20:38] <brontosaurusrex> goes from 10s to almost no delay [20:38] <grkblood> hrm [20:39] <brontosaurusrex> not sure how that applies to NPR [20:39] <grkblood> arent we defining the input parameters with these commands: ffmpeg -f s16le -ar 8000 -ac 1 -i [20:39] <grkblood> which is actually wrong now btw since I changed my source [20:41] <brontosaurusrex> i meant "ffmpeg -i http://intranet/file.mp4 file.mp4" * [20:44] <brontosaurusrex> grkblood, so what exactly is the problem with the delay? It starts recording later than you want, or? [20:44] <grkblood> its an issue because I need to be realtime or as close to realtime as possible [20:44] <grkblood> a 5 second delay is hardly real time [20:44] <brontosaurusrex> yeah [20:49] <brontosaurusrex> cvlc http://nprdmp.ic.llnwd.net/stream/nprdmp_live01_mp3 does work fast btw [20:50] <brontosaurusrex> but mpv also has delay and some errors [20:50] <brontosaurusrex> so perhaps you could use cvlc to dump stream [20:53] <grkblood> that seems like a lot of extra baggage getting vlc involved [20:56] <brontosaurusrex> dunno; cvlc http://nprdmp.ic.llnwd.net/stream/nprdmp_live01_mp3 --sout=file/mp3:./file.mp3 [20:56] <brontosaurusrex> ^ this is faster than 5 secs [20:59] <grkblood> k, thanks. Ill play around with that [21:00] <brontosaurusrex> grkblood, you on debian? [21:01] <grkblood> nope [21:01] <grkblood> centos [21:01] <brontosaurusrex> mkay [21:03] <grkblood> fixed it [21:03] <grkblood> start it off with this: ffmpeg -analyzeduration 0 [21:03] <brontosaurusrex> share! :) [21:03] <brontosaurusrex> right [21:05] <grkblood> too bad thats not displayed in ffmpeg -h [21:06] <brontosaurusrex> ffmpeg -analyzeduration 0 -i http://nprdmp.ic.llnwd.net/stream/nprdmp_live01_mp3 -acodec copy out.mp3 ? [21:08] <grkblood> no, for my original line where I knew the audio source parameters. If you know the parameters for NPR I'm sure it'll work the same. [21:08] <grkblood> I'll post my cmd [21:09] <grkblood> input_source | ffmpeg -analyzeduration 0 -f s16le -ar 8000 -ac 1 -i pipe:0 -acodec pcm_s16le -ar 22050 -ac 1 pipe:1 [21:10] <grkblood> input_source is a local binary that outputs raw PCM audio from a local tuner [21:18] <grkblood> brontosaurusrex: ffmpeg -analyzeduration 0 -f s16le -ar 22050 -ac 1 -i http://nprdmp.ic.llnwd.net/stream/nprdmp_live01_mp3 -acodec pcm_s16le -ar 22050 -ac 1 -f wav -y test.wav [21:19] <brontosaurusrex> well, cool [21:20] <brontosaurusrex> mpv/mplayer also has --cache=number [21:20] <brontosaurusrex> mabye there is something like that in ffmpeg as well [21:20] <brontosaurusrex> if i do "mpv --cache=32 http://....file" its fast [21:24] <Golfgeo> Hi all [21:41] <bourbon> can you compile the ffmpeg binary against pre-compiled ffmpeg libs? [21:42] <bourbon> that is, if you compile ffmpeg without the executable binaries, can you compile just the ffmpeg executable linked against the library so's? [21:50] <Golfgeo> is there a way to have a streaming ffmpeg/avconv show what you're streaming on your local machine while it is started with the same command on the command line? [21:54] <llogan> Golfgeo: we have nothing to do with avconv [21:55] <llogan> ...but i don't quite understand your question. [21:56] <Golfgeo> Well, then you can't help... So, thanks and have a great day :) [22:31] <jedir0x> Are there going to be any inherit problems with encoding H.264 video at 1 frame per second? WIll most or all players support this properly? [22:49] <jnvsor> jedir0x: Don't see why not. Worst case scenario you could lower the fps then duplicate frames so you get a 30fps video with 30 times the same frame in a row [23:08] <thebombzen_> does anyone know how to read a DVD with FFmpeg? What I've been doing for archival is dumping the stream with "mplayer dvd://1 -dumpstream -dumpfile dump.vob" then transcoding dump.vob to H.264 to fit it in less space. Can I do this directly with FFmpeg? [23:08] <sacarasc> I don't think so. [00:00] --- Wed Jan 15 2014

1 0

ffmpeg-devel.log.20140113
by burek 14 Jan '14

14 Jan '14

[01:31] <cone-846> ffmpeg.git 03Michael Niedermayer 07master:7b41cbac7fc0: avcodec/huffman: extend ff_huff_gen_len_table() to allow >8bit [01:31] <cone-846> ffmpeg.git 03Michael Niedermayer 07master:7a8318fbb1f5: avcodec/huffman: raise the input capability of ff_huff_gen_len_table() from 8 to 14bit [04:08] <llogan> ubitux: Trivia: The movie "The Fifth Estate" contains a reference to ffmpeg at ~1h15s. [05:15] <cone-821> ffmpeg.git 03Luca Barbato 07master:48a5b155433e: hevc: Reject impossible dependent tile [05:15] <cone-821> ffmpeg.git 03Luca Barbato 07master:aead772b5814: hevc: Bound check slice_qp [05:15] <cone-821> ffmpeg.git 03Michael Niedermayer 07master:22bfb4be284c: avcodec/hevc: Check entry point arrays for malloc failure [07:29] <ubitux> llogan :) [10:58] <funman> I wonder why mraulet added the temporary (from the draft) HEVC transport stream ID to the demuxer but not the muxer [10:58] <funman> it's a bit useless if noone generates ts files with that id [10:58] <funman> https://github.com/OpenHEVC/openHEVC/blob/hevc/libavformat/mpegts.c oh well he did [10:59] <funman> michaelni: any reason against picking that up ? [11:00] <funman> woops mpegts.c is the demux [11:00] <JEEB> just make sure that's in the latest draft too [11:00] <JEEB> I don't think it's changed, but just make sure [11:00] <JEEB> unless you have access to the documents [11:00] <funman> well i don't even have the draft [11:40] <ubitux> arh [11:40] <ubitux> it seems best_effort_timestamp might be broken in case of non-flushing [11:41] <ubitux> i need to confirm this, but if you reach the end, don't flush, seek to start, decode [11:41] <ubitux> the best_effort_timestamp might be broken (with a ending ts info) [11:47] <wm4> best_effort_timestamp is in general not very reliable [11:48] <ubitux> what should i use? :P [11:49] <wm4> write your own complicated custom code [11:49] <wm4> for users, best_effort_timestamp isn't an option anyway, because libav doesn't have it [11:51] <nevcairiel> by that logic ffmpeg may as well stop developing features [11:51] <wm4> if it's something as basic as this, then yes [11:51] <wm4> and it does [11:52] <nevcairiel> by my logic, once you have enough compelling features, people will choose the right lib that works for them [11:54] <durandal_1707> huh, why cehoys replies to msgs i do not see? [11:54] <wm4> durandal_1707: spam filter? [11:56] <durandal_1707> no, i think cehoyos picked wrong pill [11:57] <durandal_1707> wm4: oops you are right, fucking retarded spam filter [12:02] <ubitux> anyway it seems to do some flushing post seek [12:02] <ubitux> that's really weird, but my code might be wrong [12:44] <wm4> there should be a ffmpeg consulting company [12:45] <BBB> ubitux: yeah we beat them by a good 15% now [12:45] <ubitux> yeaah [12:45] <ubitux> BBB: not in 32-bit and ARM though i suppose [12:45] <ubitux> BBB: with threading? [12:46] <ubitux> also we don't have any avx yet [12:46] <ubitux> i think they do [12:53] <pross-au> BBB: about? [13:00] <pross-au> disregard. think i solved it [13:19] <Rodeo> Hi! Is anyone here going to FOSDEM and still looking to share accommodations (e.g. hotel)? [13:19] <BBB> ubitux: w/o threading, w/ a lot more [13:20] <BBB> ubitux: depending on thread type basically [13:20] <BBB> tey have some avx2 in the encoder, I didn't look to see if they have it in the decoder [13:20] <ubitux> Rodeo: i don't know if some ff dev developers are going to fosdem [13:20] <ubitux> BBB: i was think of avx [13:20] <BBB> oh [13:20] <ubitux> not even avx2 [13:21] <Rodeo> neither do I, which is why I'm asking ;-) [13:21] <ubitux> thinking* [13:21] <BBB> it's trivially simple to add avx [13:21] <BBB> we already made the macros the correct way [13:21] <BBB> just add some INIT_XMM avx and we're done [13:21] <BBB> kierank has a test machine he volunteered for us to use for testing [13:21] <ubitux> well, will that save some unpacking? [13:21] <BBB> mova mostly [13:21] <BBB> avx adds three-operand instructions [13:21] <ubitux> yeah right, but that's not the interesting thing [13:21] <BBB> oh, you'd be surprised [13:22] <BBB> we use it all over the place [13:22] <kierank> I have an avx2 machine but nowhere to colocate it :( [13:22] <BBB> ubitux: example, for prores simd, it helped in the order of 10% or so, iirc [13:22] <ubitux> BBB: yeah but wouldn't the main benefit in stuff like loop filter where you don't unpack twice [13:23] <BBB> you mean we punpcklbw on the same data twice? [13:23] <BBB> or unpack = load-from-memory? [13:23] <ubitux> yes the punpck[lh]bw/repack [13:24] <BBB> where do we do it twice? [13:24] <BBB> in filter6/14? [13:24] <ubitux> yes [13:24] <BBB> hm... maybe we can merge these two [13:25] <ubitux> would that mean using half of 256b regs for the previous code? [13:26] <BBB> I hope not [13:26] <BBB> rather, for avx2, we would do 32bytes all at once [13:26] <BBB> we have 32 registers anyway [13:26] <ubitux> ah [13:27] <ubitux> also, i guess for itxfm with 16x16 and 32x32 having 256b reg would help [13:27] <ubitux> and i'm assuming you will have to write some code avx specific [13:27] <ubitux> no? [13:49] <BBB> ubitux: absolutely [13:49] <BBB> ubitux: it's not that trivial to convert from sseX to avx2, but it can be done I guess [13:49] <BBB> 32x256b regs certainly gives incentive [13:49] <ubitux> again, i'm thinking avx :p [13:50] <ubitux> avx is 256b and avx2 512 right? [13:50] <BBB> oh wait the 32 was for the generation after that [13:50] <BBB> nah I'm confusing two things, 32 is for 512b regs, the one after avx/2 [13:50] <BBB> iirc [13:50] <BBB> avx2 is integer form of avx, like sse2 is integer form of sse [13:52] <ubitux> oh so, since i have here a cpu with avx, it has absolutely no use for integers except the mova thing? [13:52] <ubitux> (and eventually float ops?) [13:52] <BBB> yes [13:53] <ubitux> mmh ok [14:02] <michaelni> funman, i cherry picked all things that mraulet asked me to pick, if theres more thats ready and should be picked [14:03] <michaelni> just say what / which hash [14:13] <cone-123> ffmpeg.git 03Luca Barbato 07master:c798a6fedc3d: vc1: Factorize out chroma MC [14:13] <cone-123> ffmpeg.git 03Michael Niedermayer 07master:1df53ff22de2: Merge commit 'c798a6fedc3dd9ed3ac1fb2d95819af58705a87e' [14:14] <michaelni> funman / and others, also if theres some hevc repo/tree that can just be git merged thats welcome as well [14:24] <cone-123> ffmpeg.git 03Michael Niedermayer 07master:6d98959c8a5a: vc1: Add avg_no_rnd_vc1_chroma_mc4_c() [14:24] <cone-123> ffmpeg.git 03Michael Niedermayer 07master:58eb9b929998: Merge commit '6d98959c8a5a749592a878b9523b31eccea5d102' [14:33] <funman> michaelni: there doesn't seem to be more, I'll ask mraulet when I see him [14:36] <cone-123> ffmpeg.git 03Tim Walker 07master:49b9badcaab5: vc1: Enable the interlaced B-frame codepath. [14:36] <cone-123> ffmpeg.git 03Michael Niedermayer 07master:a067445b45fb: Merge commit '49b9badcaab513b65b79be371b1a9c1012e482ea' [14:37] <michaelni> funman, ok, thanks [14:43] <cone-123> ffmpeg.git 03Martin Storsjö 07master:99e20125232d: x86/arm: Add clobber tests to libavresample [14:43] <cone-123> ffmpeg.git 03Michael Niedermayer 07master:9ace154186d8: Merge commit '99e20125232d40ea239a5b54d26bbfe526506114' [15:14] <cone-123> ffmpeg.git 03Luca Barbato 07release/1.1:9aa22918c258: prores: Error out only on surely incomplete ac_coeffs [15:14] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/1.1:9f47f95e7087: Merge remote-tracking branch 'qatar/release/9' into release/1.1 [15:21] <mischi> hi. [15:22] <mischi> I want to suggest to add UltraStar Deluxe, a Karaoke game, to the list of projects using ffmpeg. [15:23] <mischi> With 10.000 downloads per week from sourceforge, i think it is relevant. [15:23] <av500> sourceforge still exists? [15:23] <mischi> hehe [15:23] <mischi> similar as Pascal, the language used for USDX. [15:24] <mischi> As such USDX has header translations for using ffmpeg libs from Pascal. [15:24] <kierank> av500: nowhere has better mirrors than sourceforge [15:30] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/1.1:1017b5914cec: update for 1.1.8 [15:32] <michaelni> mischi, send patch, see git://ffmpeg.org/ffmpeg-web [15:47] <cone-123> ffmpeg.git 03Michael Niedermayer 07fatal: ambiguous argument 'refs/tags/n1.1.8': unknown revision or path not in the working tree. [15:47] <cone-123> Use '--' to separate paths from revisions [15:47] <cone-123> refs/tags/n1.1.8:HEAD: Merge commit '99e20125232d40ea239a5b54d26bbfe526506114' [15:50] <Daemon404> who was it that run the vlc irc bots? funman? [15:51] <ubitux> thresh? [15:52] <Daemon404> ah right. [15:52] <Daemon404> it really should be a one line fix [15:52] <Daemon404> i see that -- error a lot [16:17] <mischi> michaelni: ok. [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:6fb0f37defbf: avcodec/h264: fix code that blindly dereferences NULL DPB [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:050d8d727ad6: avformat/rmdec: when reading audio blocks, dont leave holes when reading fails [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:4f3b0afaa236: avformat/mxfdec: check avio_read(UID) result [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:763a808c96bb: avformat/flvdec: initialize context before reading from it [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:6da213ce208d: avformat/matroskadec: check generic audio deinterleaver sub_packet_size against frame_size [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:aaaafc300df6: avcodec/wmalosslessdec: Pass on error code from decode_tilehdr() [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:cd753ee36b0f: avcodec/wmalosslessdec: deallocate uninitialized frame on decode_tilehdr() failure [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:57824d1ba617: avcodec/wmalosslessdec: shrink output on error so no uninitialized data is returned [17:29] <cone-123> ffmpeg.git 03Clément BSsch 07release/2.1:b6af89be6a2b: avformat/pjsdec: dont increase pointer when its already at the end in read_ts() [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:b3c3dc54a595: avcodec/alac: only set *got_frame_ptr when all channels have been decoded [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:706dca18d0a8: avcodec/hevc: use av_mallocz() for allocating tab_ipm [17:29] <cone-123> ffmpeg.git 03Luca Barbato 07release/2.1:738a2a04b6d8: hevc: Clip the pixel before shifting [17:29] <cone-123> ffmpeg.git 03Guillaume Martres 07release/2.1:ea21b7b68cfc: hevc: clip pixels when transquant bypass is used [17:29] <cone-123> ffmpeg.git 03Guillaume Martres 07release/2.1:e43805d40188: hevc: remove useless clip in FUNC(sao_band_filter)() [17:29] <cone-123> ffmpeg.git 03Luca Barbato 07release/2.1:39545c54826c: hevc: Reject impossible dependent tile [17:29] <cone-123> ffmpeg.git 03Luca Barbato 07release/2.1:d63476347a48: hevc: Bound check slice_qp [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:a0aa5c34a964: avcodec/hevc: Check entry point arrays for malloc failure [17:29] <cone-123> ffmpeg.git 03Carl Eugen Hoyos 07release/2.1:b336daa952d9: Fix a crash on oom when decoding hevc. (cherry picked from commit 5ab1efb9d0dc65e748a0291b67915e35578b302e) [17:29] <cone-123> ffmpeg.git 03Michael Niedermayer 07release/2.1:29353dd3f815: Update for 2.1.2 [17:52] <cone-123> ffmpeg.git 03Michael Niedermayer 07fatal: ambiguous argument 'refs/tags/n2.1.2': unknown revision or path not in the working tree. [17:52] <cone-123> Use '--' to separate paths from revisions [17:52] <cone-123> refs/tags/n2.1.2:HEAD: Merge commit '99e20125232d40ea239a5b54d26bbfe526506114' [18:55] <Daemon404> 42 [19:05] <gnafu> A likely story. [19:37] <llogan> BBB: any opinion on http://ffmpeg.org/pipermail/ffmpeg-devel/2014-January/153152.html ? [20:37] <llogan> Anssi: anything upcoming for HLS in the near future? [20:39] <Anssi> llogan: well, I'm hoping I get my patchset in shape in the next few weeks... [20:47] <llogan> Anssi: not sure how i spaced that out... [20:48] <llogan> my english sucks. i meant that I'm not sure how i forgot that patch set. [20:50] <Anssi> mostly the id3 timestamp stuff needs some thinking, I may go back to the approach I had in v1 patchset... another thing is if a cleaner approach to subtitles is feasible [21:34] <cone-644> ffmpeg.git 03Anton Khirnov 07master:c1868e7ee7b0: lavf: simplify handling of offset in av_probe_input_buuffer() [21:34] <cone-644> ffmpeg.git 03Michael Niedermayer 07master:ececbedaa86c: Merge commit 'c1868e7ee7b07b40a0fe15f50df89fe499a01a50' [21:40] <cone-644> ffmpeg.git 03Anton Khirnov 07master:8b76362836f3: lavf: use a fixed width type [21:40] <cone-644> ffmpeg.git 03Michael Niedermayer 07master:220f0927b86f: Merge commit '8b76362836f3c373c3aadc544522edcbef16dd5f' [22:35] <cone-644> ffmpeg.git 03Anton Khirnov 07master:2115a3597457: lavf: make av_probe_input_buffer more robust [22:35] <cone-644> ffmpeg.git 03Michael Niedermayer 07master:cdce0e8a506c: Merge remote-tracking branch 'qatar/master' [22:35] <cone-644> ffmpeg.git 03Michael Niedermayer 07master:6a2064820b52: avformat/utils/av_probe_input_buffer2: Fix pd.buf_size [22:35] <cone-644> ffmpeg.git 03Michael Niedermayer 07master:8c3b026a0eeb: avformat/utils/av_probe_input_buffer2: fix offset check [22:35] <cone-644> ffmpeg.git 03Michael Niedermayer 07master:05886c9d4edd: avformat/utils/av_probe_input_buffer2: fix buffer passed to ffio_rewind_with_probe_data() [23:47] <ubitux> BBB: i have 4 new samples to crash the decoder [23:47] <ubitux> interested? [00:00] --- Tue Jan 14 2014

1 0

ffmpeg.log.20140113
by burek 14 Jan '14

14 Jan '14

[01:41] <average> hello ? [01:42] <average> I did a voiceover for some small video I made [01:42] <average> then I merged audio and video I got an f1.avi [01:42] <average> the problem with f1.avi is that the audio is bigger than the video in length [01:42] <average> but ffmpeg did not complain, all fine [01:42] <average> then I had another f2.avi which was ok from that point of view [01:42] <average> and I tried to join f1.avi and f2.avi [01:43] <average> the result was a disaster [01:43] <average> ffmpeg pushed the pace on the voiceover [01:43] <average> and my voice now sounds as if I inhaled helium [01:43] <average> that was funny [01:43] <average> the error I get is the following [01:43] <average> [avi @ 0x25c95a0] Switching to NI mode, due to poor interleaving [01:44] <average> well, not an error [01:44] <average> more of a warning message [01:44] <average> but anyway [01:44] <thebombzen> Was the audio longer than the video before muxing into an AVI? [01:46] <thebombzen> becaues if so, you might have to crop it. If you mux into an avi, you can use -t to set the encode time length, and -ss to set the start time in seconds. FFmpeg will discard all data that falls outside that range. [01:46] <thebombzen> however, it doesn't work with -c:v copy and -c:a copy. [01:47] <thebombzen> if you need different start times then you could specify the -ss before the -i, or you could encode separately then mux together. [01:58] <average> thebombzen: I fixed it [01:59] <average> thebombzen: had to set the bit rate for audio to be the same [01:59] <average> thebombzen: now I have another problem [01:59] <average> joining two videos yields another with the exact same length [01:59] <average> let me explain [01:59] <average> I have v1 and v2 [01:59] <average> I join(concatenate) them and get v3 [01:59] <average> v3 is almost the same size as v1 and contains none of v2 [01:59] <thebombzen> it depends on how you join them [01:59] <average> and I get the warning I wrote about above [02:00] <average> thebombzen: I can paste the command [02:00] <average> that's not a problem.. [02:00] <thebombzen> please do [02:00] <average> but I don't know what... [02:00] <average> moment [02:00] <thebombzen> Also, this is a dumb question, but have you read: https://trac.ffmpeg.org/wiki/How%20to%20concatenate%20%28join,%20merge%29%2… [02:01] <thebombzen> it might answer your question [02:01] <average> $FFmpeg -y -i "concat:$1|$2" -c:v copy -c:a pcm_s16le -ar 44100 -b:v 2000k $3 [02:01] <average> this is what I use [02:01] <average> $1 and $2 are the videos I am concatenating [02:01] <average> $3 is the result [02:01] <average> and $3 is ok, but it only contains $1 and none of $2 [02:01] <average> that's odd.. [02:02] <average> thebombzen: read that link [02:02] <thebombzen> I'm pretty sure the concat protocol does not work with AVI containers. but I might be wrong [02:03] <albator> hi [02:03] <albator> i try to remove 5px and i m doing : -vf "crop='iw:ih/2:5:5' [02:03] <albator> but the image is really .. distorced, what am I doing wrong [02:03] <average> thebombzen: hmmm [02:03] <albator> crop 5px and width and height [02:04] <average> thebombzen: but if I go for mpeg I'll lose quality [02:04] <average> uhm [02:04] <average> so I still have to do a filter_complex.. [02:04] <thebombzen> what you can do is BASH magic [02:04] <thebombzen> you can decode the video/audio to raw video and audio and concat them together like that [02:05] <thebombzen> but that takes up too much space, so use mkfifo [02:05] <thebombzen> or even bash's temporary features [02:05] <thebombzen> for example, [02:05] <thebombzen> mkfifo rawvideo1 [02:06] <thebombzen> sorry [02:06] <bencc> is it possible to embed arbitrary meta-data in a video file? [02:06] <thebombzen> mkfifo rawvideo1 rawvideo2 rawaudio1 rawaudio2 [02:06] <average> bencc: yes, in mkv if I recall correctly, I saw it in a blogpost, forgot where [02:07] <average> bencc: some guy said he stored a lot of stuff in mkv [02:07] <albator> anyone can tell me why my cropping isn't working? [02:07] <average> bencc: but some entropy calculations can give it away to people snooping into them [02:07] <bencc> average: it doesn't have to be hidden [02:07] <average> bencc: well, ok, then mkv :) [02:07] <bencc> average: so you don't remember how to do it? [02:07] <bencc> mp4 is mkv? [02:08] <average> bencc: google for mkv, no it's not mp4, it's mkv [02:08] <bencc> ok [02:08] <bencc> I need to do it in webm and mp4 [02:08] <average> bencc: that's tough [02:09] <bencc> average: what about markers? is there only a single type of marker? chapter marker? [02:09] <bencc> or maybe there is also episode? [02:09] <thebombzen> average: if you don't want to reencode the video in temp files, you could use the concat filter. [02:09] <thebombzen> see https://trac.ffmpeg.org/wiki/How%20to%20concatenate%20%28join%2C%20merge%29… [02:10] <average> thebombzen: they have the same codecs [02:10] <average> thebombzen: they're both avi and have the same codecs, video and audio [02:11] <albator> so what s the issue? [02:11] <thebombzen> in that case, the concat demuxer is the best option. don't use the concat protocol, use the concat demuxer [02:11] <average> thebombzen: ok I will try, thank you [02:11] <average> albator: read above ^^ [02:11] <average> bencc: there's only one way to find out [02:13] <bencc> average: I searched but could only find info about chapters [02:13] <bencc> I have a feeling that there is more to it though [02:13] <albator> i m doing concat demuxer right now.. [02:13] <albator> just having final issue with encoding command [02:14] <average> albator: tell me about it [02:14] <average> .. [02:15] <albator> smt like this http://pastebin.com/Xrt9kgxv [02:41] <average> I honestly don't know what's so different between those avi files that it didn't work [02:41] <average> thebombzen: so your method worked [02:41] <average> thebombzen: uhm.. [02:41] <average> uhm still not satisfied with the result [02:42] <thebombzen> what's the issue? [02:42] <albator> reencode them first [02:43] <albator> dont see others magic options [02:43] <thebombzen> If you're talking about the mis-alignment, according to, http://ffmpeg.org/ffmpeg-formats.html#concat you can specify a duration argument as well as a file argument to the concat.txt file [02:44] <thebombzen> that allows you to crop input files' duration AFAIK. but I might be wrong [03:08] <Jinxed-> hello [03:09] <Jinxed-> anyway to configure options with precompiled ffmpeg [03:09] <Jinxed-> because the compile your own doesn't work [03:09] <Jinxed-> ./configure --enable-gpl --enable-version3 --enable-shared --enable-nonfree --enable-postproc --enable-libfaac --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libtheora --enable-libvorbis --enable-libvpx --enable-libx264 --enable-libxvi [03:09] <Jinxed-> is what I need [03:09] <Jinxed-> should be a d at the end of libxvid [03:10] <Jinxed-> I followed the guide: https://trac.ffmpeg.org/wiki/UbuntuCompilationGuide [03:10] <Jinxed-> however it gives you the wrong libavfilter.so file [03:10] <Jinxed-> .so.3 when it needs so.4 [03:11] <Jinxed-> I have prob now spent about 20 hrs working on installing vimp, mainly on ffmpeg issues [04:27] <thebombzen> Jinxed-: what do you mean "it gives you the wrong libavfilter.so" [04:27] <thebombzen> building FFmpeg on Ubuntu installs libavfilter.so.4 in /usr/local/lib [04:28] <thebombzen> if you're building something else from source, you have to add /usr/local/lib to your environment variable "LD_LIBRARY_PATH" [04:28] <thebombzen> and then you have to run "sudo ldconfig" [04:49] <Jinxed-> thebombzen: I'm starting all over [04:52] <Jinxed-> now I'm having issues with ERROR: libvpx decoder version must be >=0.9.1 [04:53] <Jinxed-> thebombzen: this is what I did: http://pastie.org/private/8sl2vrwumoqi7lycp3qqa [04:57] <llogan> Jinxed-: that's not the guide [04:58] <thebombzen> first things first: [04:58] <thebombzen> FAAC is useless [04:58] <albator> why not use git for x264 [04:58] <thebombzen> both libfaac and libfdk_aac are non-free (but open-source) but fdk_aac sounds way better for the same bitrate [04:58] <llogan> why aren't you using checkinstall since you're installing to the system so you can uninstall it in case you remove the source directories? [04:59] <thebombzen> and by nonfree I mean you have to use --enable-nonfree [04:59] <llogan> libvpx is at 1.3.0 now [04:59] <thebombzen> Jinxed-: you should build libvpx yourself, from the git master. same with x264, same with libfdk_aac [05:00] <thebombzen> also you're going to want to build libopus yourself. libopus is currently at 1.1 which is a huge improvement over the one in the ubuntu repos, which is 1.0.3 IIRC [05:00] <thebombzen> as in it sounds way better [05:00] <Jinxed-> thebombzen: I'm not sure how to modify those instructions as to be able to still install vimp [05:00] <Jinxed-> need the transcoding tools: ./configure --enable-shared --enable-pic [05:01] <Jinxed-> http://www.vimp.com/en/web/faq-installation/items/how-to-install-the-transc… [05:01] <thebombzen> you're also using an ancient version of FFmpeg [05:01] <thebombzen> 0.8.1? [05:01] <Jinxed-> 2.01 ? [05:01] <thebombzen> oops, that's commented out [05:01] <Jinxed-> that is commented out, someone suggested it [05:01] <Jinxed-> and I wasn't have success with 2.01 [05:01] <thebombzen> but yea. always build FFmpeg from the git master because it's got the most features and usually doens't have bugs (unless you're a server/system maintainer, then don't.) [05:01] <albator> git should work too no? [05:02] <albator> sure [05:02] <Jinxed-> https://trac.ffmpeg.org/wiki/UbuntuCompilationGuide [05:02] <albator> the manual from fakeoutdoor is the best [05:02] <Jinxed-> I tried that guide but it had the wrong libavfilter on it [05:02] <albator> never fail [05:02] <Jinxed-> linky? [05:02] <thebombzen> remember to always clone ffmpeg from the git master if you want a higher version number [05:03] <llogan> that is the link. but it does not include shared stuff. [05:03] <llogan> which looks like you want? [05:03] <thebombzen> that guide has many things wrong with it [05:03] <thebombzen> such as suggesting static builds, which breaks one of the whole ideas of shared libs on Linux [05:03] <Jinxed-> yeah, I modified it for the enabled shared but didn't work [05:04] <Jinxed-> thebombzen: which guide, the original one of the trac.ffmpeg.org guide [05:04] <thebombzen> that's because you have to build all the other libraries as shared also, and you need --disable-static or it will build the static library too and link to that instead [05:04] <thebombzen> as a habit, when building on Linux you should configure with --enable-shared --disable-static [05:04] <Jinxed-> alright [05:05] <thebombzen> if you run ./configure --help (this option always exists) then it provides a list of options you can use. [05:05] <Jinxed-> --prefix="$HOME/ffmpeg_build" --bindir="$HOME/bin" (what does this do?) [05:05] <thebombzen> leave that out [05:05] <thebombzen> you don't need it [05:06] <llogan> it allows you to create a "local" install as opposed to installing stuff into "the system: [05:06] <thebombzen> what it does is change the location that it will install the libraries and binaries away from the default /usr/local. but you want it to be /usr/local [05:07] <llogan> the guide used to install into /usr/local but it caused too many issues with conflicting and dependent packages [05:08] <Jinxed-> well I think I can change the path's of vimp to reflect the home location [05:08] <Jinxed-> is there any harm in it? [05:08] <llogan> it's really your choice. [05:08] <clever> main benefit i can see with the home method is that you never have to give it root, so it cant hose the entire system by mistake [05:08] <clever> and you can just rm -rf the dest dir to remove all the files its installing [05:09] <thebombzen> the libraries that you should build yourself are libx264, libvpx, libfdk-aac, and libopus. Most of the other libraries you can install from the repos. (and libaacs, libbdplus, land ibbluray if you want bluray reading) [05:09] <relaxed> you can do the same thing anywhere [05:09] <clever> easyer to remove then trying to clean it out of /usr/local/ [05:09] <clever> but every other lib that links against it must be told ffmpeg isnt in a normal loc [05:09] <relaxed> /opt/ffmpeg for example [05:09] <clever> yep [05:10] <thebombzen> yea. for an end-user, /usr/local is the best place, as long as you configured ubuntu to add /usr/local/lib to your LD_LIBRARY_PATH [05:10] <thebombzen> if you don't know how to do that, add the line [05:10] <clever> i wouldnt modify LD_LIBRARY_PATH, i would just set it in /etc/ld.so.conf [05:10] <relaxed> I disagree [05:11] <relaxed> I wouldn't even add it to /etc/ld.so.conf [05:11] <relaxed> all the guides make this mistake [05:11] <clever> it depends on why your building a custom ffmpeg [05:11] <clever> and how widely you want to force the system to use that one [05:11] <relaxed> software through you package manager most certainly depends on specific version of libav* [05:11] <relaxed> your* [05:12] <thebombzen> I don't use /etc/ld.so.conf because it has never worked for me. neither has /etc/environment. the only way I've managed to get ldconfig to search /usr/local/lib [05:12] <clever> for omxplayer, they build their own ffmpeg to a special dir, and the launch script sets LD_LIBRARY_PATH for ONLY omxplayer.bin [05:12] <clever> so it wont effect any other programs, ever [05:12] <thebombzen> is by adding export LD_LIBRARY_PATH="/usr/local/lib:$LD_LIBRARY_PATH" at the end of /etc/bash.bashrc [05:12] <clever> thebombzen: ldconfig ignores LD_LIBRARY_PATH [05:13] <clever> what your doing, is forcing ld.so to scan those directories at startup, of every single app, system wide [05:13] <thebombzen> well that's worked. I don't know why [05:13] <relaxed> I recommend writing a wrapper script that calls LD_LIBRARY_PATH=/path/to/custom/lib /path/to/ffmpeg [05:13] <clever> relaxed: thats basicaly what omxplayer does on the raspberrypi [05:13] <thebombzen> well that doesn't work because I build a lot of stuff from source [05:13] <clever> that minimizes all of the side effects [05:13] <albator> well now.. nobody agree how to build ffmpeg :D [05:13] <clever> firefox also does the same thing [05:14] <clever> it depends on how much you want to force into using the new libs [05:14] <thebombzen> I also set LD_RUN_PATH though, that might help [05:14] <clever> thebombzen: thats not even in the man page (run 'man ld.so') [05:15] <albator> before I had to export library path like you, but now ldconfig makes the trick [05:15] <albator> have no idea why.. [05:15] <clever> albator: http://www.linuxfromscratch.org/ install this on a spare system, then youll understand everything [05:16] <thebombzen> for some reason editing /etc/ld.so.conf has never worked for me [05:16] <thebombzen> I don't know why [05:16] <thebombzen> anyway I gtg [05:16] <albator> that's clever indeed [05:16] <albator> ;) [05:17] <clever> albator: basicaly, its directions on how to compile the entire system, mostly just ./configure && make && make install [05:17] <clever> along with all of the configure flags you need to make things work together right, and why they are needed [05:17] <albator> i just think ubuntu is getting easier over the years [05:17] <clever> the tricky part, is just making an isolated compiler set that lives in /tools/ and makes no reference to any /usr/lib or /lib file [05:17] <albator> and that include ffmpeg [05:17] <clever> so you can then compile everything in an isolated env [05:18] <albator> i see [05:18] <clever> that includes the dynamic linker also, which is hard-coded into every program the linker makes [05:19] <clever> normaly, everything is setup to link to /lib/ld-linux.so.2 i think [05:20] <clever> that is then responsible for checking the cache (made by ldconfig) and LD_LIBRARY_PATH, finding all the dynamic libs, and loading them [05:20] <Jinxed-> Ok, I'm a bit confused, but I tried to incorporate a few of your suggestions into this: http://pastie.org/private/jmxq95e8i7dq9nm6tglyhq [05:20] <Jinxed-> does that seem right? [05:21] <clever> i dont think hash -r is required [05:22] <clever> thats usualy needed to clear a negative hit out of the cache when you add something to PATH, but your not going to get any negative hits [05:22] <albator> it s not --enable -shared instead of disable static ? [05:22] <clever> git on line 18 wont be looked for until after youve installed it [05:22] <albator> for ffmpeg ./configure [05:22] <clever> also, why are you telling apt to uninstall everything your telling it to install? [05:23] <clever> Jinxed-: oh, and another key part that will help a ton, #!/bin/bash -e [05:23] <albator> typo :P [05:23] <clever> Jinxed-: if any command fails (like make), the script will halt [05:23] <clever> it wont keep going and hide the error behind 20 pages of output [05:24] <Jinxed-> nice [05:24] <Jinxed-> great idea! [05:25] <Jinxed-> yeah I usually run it ./script.sh | tee outputFile.log [05:25] <Jinxed-> so I can see it [05:26] <clever> also, another neat trick in bash, i think you can use it like this [05:26] <clever> exec | tee logfile.txt [05:26] <clever> if you put that at the top, it will replace stdout of itself with a pipe to tee [05:27] <clever> also, neither my method nor yours, will catch stderr [05:27] <clever> if things are obeying the rules, the errors go to stderr, and you dont get a copy of them! [05:28] <Jinxed-> hahaha [05:28] <Jinxed-> well shit [05:28] <clever> 2>&1 will redirect stderr to stdout [05:28] <clever> ./script.sh 2>&1 | tee logfile [05:28] <clever> for example [05:29] <Jinxed-> oh nice [05:29] <Jinxed-> alright, so I added the -e at the top [05:29] <clever> i often use this inside scripts i run from cron and stuff [05:29] <Jinxed-> I didn't need to make any other changes? [05:29] <Jinxed-> time to run? [05:29] <clever> exec >> /tmp/action_log 2>&1 [05:30] <clever> this will cause the script to append its stdout to the log file, and send stderr to the same place [05:30] <clever> so rather then cron going crazy and emailing me once an hour [05:30] <clever> it all goes to a simple log file [05:30] <Jinxed-> nice, I use to run a lot of my scripts in screen as well [05:31] <Jinxed-> so if I needed to check up on them live I could do so very quickly [05:31] <Jinxed-> (would log as well) [05:31] <clever> and i always keep 1 screen session on EVERY host [05:31] <clever> function loginto() { xterm -fg white -bg black -e ssh $1 -t screen -xRRU & } [05:31] <Jinxed-> so that exec line goes at the top of the bash script? [05:31] <clever> this will ssh into a given host and resume the existing screen session [05:31] <clever> yeah [05:31] <clever> so i can just loginto c2d [05:32] <clever> and i get an xterm with c2d's screen session [05:32] <Jinxed-> haha nice [05:34] <Jinxed-> alright about to run this: http://pastie.org/private/r3zd9zwvtqwx0jgdfbpiw [05:35] <clever> and tail -f is valuable in this case [05:35] <clever> since your getting no output at all now [05:35] <Jinxed-> so now I just run sudo ./myscript.sh [05:36] <Jinxed-> and tail -f the file [05:36] <clever> no need for a sudo on the outside [05:36] <clever> since it has sudo inside on the required lines [05:37] <Jinxed-> line 3: exec: exec: not found [05:37] <clever> just one exec [05:37] <Jinxed-> lol [05:38] <Jinxed-> i'm an idiot [05:38] <clever> normaly, exec takes the name and arguments of another program [05:38] <clever> and will replace the current bash with that [05:38] <clever> keeping the same pid [05:39] <Jinxed-> hahaha I didn't add -y to the script for the apt-get install [05:39] <Jinxed-> so I see the prompt in the tail -f [05:39] <Jinxed-> but then have to do it in the terminal [05:39] <Jinxed-> with no prompt [05:39] <clever> you might be able to tee it [05:39] <clever> exec 2>&1 | tee vimp_log [06:22] <Jinxed-> clever: albator thebombzen I followed the proper instructions and I get this error: ./ffmpeg: error while loading shared libraries: libavfilter.so.4: cannot open shared object file: No such file or directory [06:30] <relaxed> Jinxed-: what was your install prefix? [06:31] <Jinxed-> relaxed: http://pastie.org/private/qjevpyggniqctwhkbvqqg [06:31] <Jinxed-> that was what I used to install based on recommendations here [06:32] <relaxed> run this: LD_LIBRARY_PATH=$HOME/ffmpeg_build/lib $HOME/bin/ffmpeg [06:34] <relaxed> did it work? [06:36] <Jinxed-> relaxed: trying now [06:38] <Jinxed-> relaxed: is that one line or two lines [06:38] <relaxed> one [06:40] <Jinxed-> relaxed: didn't work [06:40] <Jinxed-> http://paste2.org/LU3P31Ja [06:41] <relaxed> DUDE copy cand paste this into your terminal --> LD_LIBRARY_PATH=$HOME/ffmpeg_build/lib $HOME/bin/ffmpeg [06:41] <Jinxed-> relaxed: I did... i pasted you the results [06:41] <relaxed> oh, it did work [06:42] <relaxed> yay me [06:42] <Jinxed-> why does ./ffmpeg -v still cause an error? [06:42] <relaxed> what is your home dir? [06:42] <relaxed> we'll get to that [06:43] <Jinxed-> /home/tom [06:45] <relaxed> ok. run this: sudo echo $'#!/bin/sh\nLD_LIBRARY_PATH=/home/tom/ffmpeg_build/lib /home/tom/bin/ffmpeg "$@"' > /usr/bin/ffmpeg-custom [06:45] <relaxed> sudo chmod +x /usr/bin/ffmpeg-custom [06:45] <relaxed> now run: ffmpeg-custom [06:47] <relaxed> The problem was FFmpeg's libs weren't in your systems search path. [06:47] <relaxed> man ld.so [06:47] <Jinxed-> tom@tom-CF-30KTPAX2M:~/bin$ sudo echo $'#!/bin/sh\nLD_LIBRARY_PATH=/home/tom/ffmpeg_build/lib /home/tom/bin/ffmpeg "$@"' > /usr/bin/ffmpeg-custom [06:47] <Jinxed-> bash: /usr/bin/ffmpeg-custom: Permission denied [06:48] <relaxed> sudo chmod +x /usr/bin/ffmpeg-custom [06:49] <Jinxed-> relaxed: the first command fails [06:49] <Jinxed-> the bash file is never made [06:49] <Jinxed-> so I can't make it executable [06:49] <relaxed> you ran it as one command, correct? [06:50] <Jinxed-> use [06:50] <Jinxed-> tom@tom-CF-30KTPAX2M:~/bin$ sudo echo $'#!/bin/sh\nLD_LIBRARY_PATH=/home/tom/ffmpeg_build/lib /home/tom/bin/ffmpeg "$@"' > /usr/bin/ffmpeg-custom [06:50] <Jinxed-> bash: /usr/bin/ffmpeg-custom: Permission denied [06:50] <Jinxed-> one line [06:50] <Jinxed-> I checked /usr/bin/ffmpeg-custom is never made [06:51] <relaxed> oh [06:52] <Jinxed-> do you just want a simple bash file [06:52] <Jinxed-> with LD_LIBRARY_PATH=/home/tom/ffmpeg_build/lib /home/tom/bin/ffmpeg [06:52] <Jinxed-> ? [06:52] <relaxed> Run this: echo $'#!/bin/sh\nLD_LIBRARY_PATH=/home/tom/ffmpeg_build/lib /home/tom/bin/ffmpeg "$@"' > /home/tom/bin/ffmpeg-custom [06:53] <relaxed> chmod +x /home/tom/bin/ffmpeg-custom [06:53] <Jinxed-> ok that worked [06:54] <Jinxed-> you want me to run the script now? [06:54] <relaxed> that script will load the require libs and run ffmpeg [06:56] <Jinxed-> doesn't work [06:56] <Jinxed-> same result [06:57] <relaxed> you're running ffmpeg-custom? [06:58] <Jinxed-> http://paste2.org/FWzGECnV [06:58] <relaxed> right, just use the wrapper script. [06:59] <Jinxed-> you mean I need to run ffmpeg with ffmpeg-custom? [06:59] <relaxed> I mean forget about ffmpeg and use ffmpeg-custom [07:00] <relaxed> you can create simliar scripts for the reset of the ff* binaries if you need them [07:00] <Jinxed-> hmm, so I'm a little worried because the whole reason I need ffmpeg and the others is vimp [07:00] <Jinxed-> I'm not sure how it will work using these wrappers [07:01] <Jinxed-> maybe it will work? I guess I can try tomorrow [07:01] <Jinxed-> enough hours spent today [07:01] <relaxed> well, you can add /home/tom/ffmpeg_build/lib to ld.so's search path [07:01] <Jinxed-> would be happy to do that [07:01] <Jinxed-> will that work? [07:01] <relaxed> and do away with the wrapper [07:02] <relaxed> sudo echo "/home/tom/ffmpeg_build/lib" >> /etc/ld.so.conf.d/ffmpeg.conf [07:02] <relaxed> sudo ldconfig [07:03] <relaxed> but I think that's a sloppy approach [07:04] <Jinxed-> tom@tom-CF-30KTPAX2M:~/bin$ ./ffmpeg -h [07:04] <Jinxed-> ./ffmpeg: relocation error: /home/tom/ffmpeg_build/lib/libavfilter.so.4: symbol sws_isSupportedEndiannessConversion, version LIBSWSCALE_2 not defined in file libswscale.so.2 with link time reference [07:04] <Jinxed-> didn't work [07:06] <relaxed> what distro is this? [07:06] <Jinxed-> ubuntu 13.10 [07:07] <Jinxed-> relaxed: the wrapper still works ffmpeg-custom -h [07:07] <Jinxed-> i'm just not sure if vimp will accept it [07:08] <relaxed> sudo apt-get purge libswscale2 [07:09] <relaxed> wait [07:09] <relaxed> run: LD_LIBRARY_PATH=/home/tom/ffmpeg_build/lib vimp [07:12] <relaxed> success? [07:12] <Jinxed-> trying [07:12] <Jinxed-> vimp isn't one command line program [07:12] <Jinxed-> its a collection of things [07:13] <Jinxed-> tom@tom-CF-30KTPAX2M:~/bin$ LD_LIBRARY_PATH=/home/tom/ffmpeg_build/lib vimp [07:13] <Jinxed-> No command 'vimp' found, did you mean: [07:13] <Jinxed-> Command 'vim' from package 'vim-gtk' (universe) [07:13] <Jinxed-> Command 'vim' from package 'vim-gnome' (main) [07:14] <Jinxed-> Command 'vim' from package 'vim-tiny' (main) [07:14] <Jinxed-> Command 'vim' from package 'vim' (main) [07:14] <Jinxed-> Command 'vim' from package 'vim-nox' (universe) [07:14] <Jinxed-> Command 'vim' from package 'vim-athena' (universe) [07:14] <Jinxed-> Command 'gimp' from package 'gimp' (main) [07:14] <Jinxed-> whoops didn't mean to paste here [07:14] <Jinxed-> i know bad form [07:15] <relaxed> Here's the deal: you have ffmpeg libs installed through you package manager that are older versions of the ffmpeg's libs you just compiled. [07:16] <Jinxed-> I don't think I have ffmpeg installed via package manager, just an outdated compiled version from following these instructions [07:16] <Jinxed-> http://www.vimp.com/en/web/faq-installation/items/how-to-install-the-transc… [07:16] <relaxed> By adding /home/tom/ffmpeg_build/lib to ld.so's search path you now have conflicting libs [07:16] <relaxed> which is why I tried to avoid doing that with wrappers calling LD_LIBRARY_PATH=/home/tom/ffmpeg_build/lib [07:18] <relaxed> they have you installing mplayer 1 too? how awful! [07:20] <Jinxed-> haha [07:20] <Jinxed-> the final product is pretty cool [07:21] <Jinxed-> a self hosted youtube [07:21] <Jinxed-> essentially [07:21] <relaxed> this is no longer an ffmpeg issue. [07:22] <Jinxed-> relaxed: a sincere thank you to you and all the others ffmpeg crew for the help tonight [07:22] <Jinxed-> i'm uploading a video with the path changed to the wrapper right now [07:22] <Jinxed-> to see if it works [07:22] <Jinxed-> damn [07:25] <Jinxed-> was unable to determine the version [07:25] <Jinxed-> probably too new for the system [07:25] <Jinxed-> but that's not an ffmpeg issue [11:55] <mohit> Hi folks [11:57] <mohit> i am encoding camera stream in android using ffmpeg.But as encoding is slow only 8-10 frames get encoded in one second. [11:58] <re-G> which codec [11:58] <mohit> h264 [11:58] <re-G> and hardware acceleration? [11:58] <mohit> hw can i improve encoding speed [11:59] <mohit> i am not using hardware acceleration is there any way of using this [11:59] <re-G> dont know [11:59] <re-G> are you tried ultrafast preset [12:00] <mohit> yes [12:00] <re-G> i guess you'll need hw accel [12:00] <JEEB> welcome to arm thingies being horrible at calculation [12:01] <mohit> actually instagram and vine android apps are recording video in square. [12:01] <JEEB> I don't think ffmpeg can use omx hw encoders, but many pieces of android hardware have a hw encoder, which is generally usable [12:01] <JEEB> off to the documentation for you, I guess :) [12:02] <mohit> yes you are right [12:03] <mohit> i am getting frames from android camera applying crop and rotate filter [12:03] <mohit> these filter makes it more slower [12:04] <JEEB> naturally [12:04] <mohit> I don,t know how instagram and vine apps are doing this [14:31] <GT1> Hi guys, I have a question. If i have a video file input with audio channel, shouldn't the itsoffset work on the audio channel too? [14:32] <ubitux> GT1: sorry i had no time to look into your issue properly, you might want to open a trac issue [14:32] <GT1> with the amerge? not a problem, with amix it's working and that's only what it matters, maybe I missunderstood the way amerege should work [14:32] <ubitux> still it's a bug :) [14:33] <ubitux> (yeah that problem) [14:33] <GT1> now I have the problem that when I use itsoffset for an input only the video file is delayed, but not the audio channel in it? [14:33] <GT1> k, I'll post it, but first I need to solve my problem at hand first :( [14:35] <Keshl> I have a video recorded and ffmpeg says the header on the audio is entirely missing, so when I try to transcode, the resulting video has no sound. I believe it'll work if I tell ffmpeg some information about the audio stream, as media players can play the "bad" file with sound without any issues. I know the audio is encoded as MP3, 44.1khz at 256 bitrate. How do I tell ffmpeg this? [14:35] <GT1> any idea how to delay the audio channel with the video channel at the same time? Because I believed itsoffset is delaying the whole input, but it seems it only delays the video channel [14:35] <Keshl> GT1: Are you delaying it with -vf? Try specifying -af, too? [14:35] <Keshl> (Just taking a random guess that -af is a thing and works like -vf.) [14:35] <GT1> no, just using -itsoffset 5 -i input.avi [14:36] <Keshl> No idea what that is. o_o Blah. xwx [14:36] <GT1> the video is dilayed correctly, but the audio is played instantly [14:36] <GT1> my command line is like this [14:36] <GT1> ./ffmpeg -i second.MOV -itsoffset 3 -i first.MOV -filter_complex "[0:v][1:v]overlay=920:0[video];[0:a][1:a]amix[audio]" -map [video] -map [audio] test.mp4 [14:36] <Keshl> Ah. [14:36] <Keshl> .. Wait, not ah. [14:37] Action: Keshl , not an expert. D: [14:37] <GT1> @ubitux do you have any idea why isn't it delaying the audio channel too? [14:38] <ubitux> no, no idea [14:38] <relaxed> GT1: you're trying to delay the audio in first.MOV? [14:38] <GT1> yea, the audio and video at the same time. This command line will delay the video channel as expected in the final render, but the audio channel renders instantly [14:40] <GT1> you can reproduce it with the console command I pasted here [14:42] <relaxed> try with a second instance of -i first.MOV and -map the video and audio from each stream [14:44] <GT1> but I need the audio to be delayed, I don't get it how would it help to use another input if I can't delay it with itsoffset [14:45] <relaxed> ffmpeg -i second.MOV -itsoffset 3 -i first.MOV -itsoffset 3 -i first.MOV -map 0 -map 1:v -map 2:a ... [14:45] <relaxed> see if that makes any difference [14:47] <GT1> i don't get it why would that help, but I'll try [14:49] <GT1> Well for some reason I get no audio at all [14:50] <relaxed> did you adjust your filter args? [14:50] <GT1> k, I just had to leave it render more, same affect, it plays the audio instantly [14:52] <relaxed> simply your command and see if you can offset it [14:52] <GT1> http://pastebin.com/twMXw5fd [14:52] <relaxed> simplify* [14:53] <GT1> same result even with ./ffmpeg -i second.MOV -itsoffset 3 -i first.MOV -map 0:v -map 1:a test.mp4 [14:54] <GT1> so basically itsoffset is only working for the video channel, WHY? :( [14:54] <relaxed> hmm, let me test [14:56] <relaxed> try using a huge offset [14:56] <GT1> rofl huge? what difference does that make? [14:56] <GT1> i'll try with 20 then [14:57] <GT1> nope [14:57] <relaxed> right, that should be fine [14:57] <GT1> same affect [14:58] <GT1> ./ffmpeg -i second.MOV -itsoffset 20 -i first.MOV -map 0:v -map 1:a test.mp4 [14:58] <relaxed> it works here with a matroska video [14:58] <GT1> The first.MOV is a 7 sec video with audio, the second.MOV is a 30 sec audio. With this command I get 7 sec of audio in the beggining of the video then nothing [14:59] <GT1> what I would want to have 7 sec of audio from 20 sec in timeline [14:59] <relaxed> put these movs up somewhere [14:59] <GT1> + the size of offset shouldn't matter, because generally I'll be needing small offsets like 1-2 max 15 sec [14:59] <grkblood13> Is there a way to pipe in hex values from a raw 16bit PCM source into ffmpeg? [15:00] <GT1> https://www.dropbox.com/s/6u762orr9drmww1/first.MOV [15:00] <GT1> https://www.dropbox.com/s/gkkijanvipma19q/second.MOV [15:00] <GT1> but it's not about video, i tried with self made video too [15:00] <GT1> and same result [15:02] <relaxed> so you want the audio from first.MOV delayed for 3 seconds? [15:03] <GT1> yes for example [15:03] <GT1> so 3 sec of silence, then the audio from the first.MOV starts playing [15:05] <relaxed> works here [15:05] <relaxed> ffmpeg version git-2013-12-21-a07d4bc [15:05] <relaxed> my command was ffmpeg -i second.MOV -itsoffset 3 -i first.MOV -map 0:v -map 1:a -preset superfast out.mov [15:05] <GT1> i have built a new version on friday or thusday, and it won't work even in that [15:06] <relaxed> I'll build git and try again [15:06] <GT1> wait< i'll copy my version number [15:07] <GT1> ffmpeg version N-59703-gbba73f5 [15:07] <GT1> no idea what that stand for... but that's my version [15:07] <relaxed> your build is from Dec 17, mine is newer [15:08] <GT1> from that version number where does it states dec 17? cuz I have no idea how to read it [15:08] <relaxed> from your pastebin it says, "built on Dec 17 2013 05:32:25 with gcc 4.6 (Debian 4.6.3-1)" [15:08] <GT1> i have a newer version wich I use on the device [15:09] <GT1> which I built on january 9 [15:09] <GT1> i just pretest the command on my linux prebuilt version, but it doesn't works on either of them [15:10] <GT1> 01-13 15:38:04.783: W/LOGTAG(5495): ***ffmpeg version N-59703-gbba73f5 Copyright (c) 2000-2014 the FFmpeg developers*** [15:10] <GT1> 01-13 15:38:04.783: W/LOGTAG(5495): *** built on Jan 9 2014 18:58:24 with gcc 4.8 (GCC)*** [15:11] <relaxed> GT1: Here's my output: http://johnvansickle.com/pub/out.mov [15:13] <Keshl> ... Blargh. Google says it's a bug ffmpeg has, not an issue with my file. Anyway, I worked around it by doing "ffmpeg -i video.flv -acodec copy blargh.mp3", which threw a warning that was ignorable, then converted the .mp3 to .flac with Audacity and then used that .flac file instead of the included audio stream when transcoding the video. [15:14] <GT1> the output you gaved isn't correct [15:14] <GT1> same effect, it palys sound in the first 7 sec of video [15:15] <GT1> maybe your player corrects it, but then it means the output file is corrupt or something. When I play it with movie play or on android it still incorrect [15:15] <Keshl> When transcoding the file, by the way, ffmpeg threw the same error as before (header missing), but the audio was there anyway. So yeah. Don't panic. I'm not sure who I'm telling this to; I just don't wanna be "that one guy who says 'nevermind, fixed it.' ". [15:17] <GT1> when I play it form web it works, but when i play it from my computer it seems wrong [15:17] <GT1> I guess then the file is corrupt and smarter players correct the issue, but default behavior is that it plays from zero timestamp [15:17] <GT1> relaxed: are you here? [15:19] <relaxed> the audio from first.mov is 7 seconds long, and with an offset of 3 that means it should end at the 10 second mark, which it does in my output. [15:20] <relaxed> tested with mpv and mplayer [15:20] <GT1> as I said, when I watch it from the web it seems correct, but when i play with from a simple player it is not, nor on android device, aka the file is corrupt, your player just corrects it [15:22] <relaxed> maybe the adelay filter is what you need. [15:24] <GT1> i'll try it out [15:33] <GT1> i don't get it, not even adelay seems to work with default players [15:36] <GT1> any idea why could this behavior be/ [15:48] <GT1> now that's sad nobody have a clue what could be the problem [15:57] <relaxed> GT1: demux the audio, create an audio stream of silence for the duration that you need, then concat the two streams [15:58] <relaxed> that should work for every player out there. [16:00] <relaxed> GT1: ffmpeg -f s16le -ar 48000 -ac 2 -i /dev/zero -t 20 out.wav [16:11] <GT1> my problem is that I need to make it work in a single command line [16:11] <GT1> because I'll use it on android device [16:12] <GT1> + as far as I know there is no dev zero on mobile [16:13] <relaxed> my nexus 4 has /dev/zero [16:14] <GT1> but does EVERY phone has dev zero? :) if yes can I make it to have a single command line? [16:14] <GT1> running ffmpeg couple of times to create zero audio then mix it together is overkill for a mobile phone [16:15] <GT1> even nexus 4 would suffer a huge performance loss launcing the exec several times, not to mention older phones [16:15] <relaxed> it sound like you're going to need a programmer [16:16] <GT1> :P I'm a programmer :) i just can't figure out why isn't ffmpeg-s itsoffset working as it should [16:17] <GT1> or you mean ffmpeg programmer? [16:38] <albator> we can only -map 2 tracks right? , how do I add subs if they are not in a srt file [16:43] <jnvsor> albator: You can map as many as the container allows, what container are you using? [16:43] <albator> mp4 [16:44] <albator> input is a mkv with 2 subs, i need to print the english one [16:44] <albator> seems to be "-filter_complex "[0:v][0:s]overlay[v]"" [16:44] <albator> trying now [16:45] <jnvsor> No, that's for video overlay. To be clear are you trying to bake the subs into the video or are you trying to add a subtitle track? [16:46] <albator> it s a "subs included in the mkv container [16:46] <albator> not a srt file or ass [16:47] <jnvsor> albator: What are you trying to achieve? [16:47] <jnvsor> Are you trying to bake the subs into the video? [16:47] <albator> to display the subs in the mp4 output [16:48] <albator> yes.. [16:48] <jnvsor> Displaying subs in the mp4 output doesn't require baking [16:48] <GT1> rofl adelay seems to work, I just didn't specified both channel [16:48] <jnvsor> albator: give us a pastebin of the command you're running [16:50] <albator> ffmpeg -re -i "file.m4v" -map 0:0 -map 0:4 -c:v libx264 -crf:v 25 -preset:v medium -profile:v main -level 3.1 -s hd480 -aspect 16:9 -c:a libfdk_aac -vbr 4 -ac 2 -ar 44100 -threads 0 -f flv [16:51] <sruz25> hi, I want to add thumbnails to a lot of mp3 files, so I need some cli interface, lame --ti worked, but it changed the filesize from 6.3MB to to 4.9MB, is there some way to add thumbnail with ffmpeg? I know it's like using canon to kill a fly, but I can't find anything better.... [16:52] <jnvsor> albator: You have no subtitle output codec (just use -c:s copy) [16:52] <albator> ok [16:52] <albator> let s try [16:53] <jnvsor> And remember to map all 3 tracks you want (video, audio, subs) [16:53] <jnvsor> sruz25: https://trac.ffmpeg.org/wiki/Create%20a%20thumbnail%20image%20every%20X%20s… [16:58] <sruz25> I don't need to extract thumbnail, I want to add them to mp3 file as album art... but I'll look around that wiki, didn't searched there yet [16:59] <jnvsor> sruz25: So you want a slideshow? [16:59] <jnvsor> sruz25: Or one still image as video for the whole file? [17:00] <sruz25> no, I have mp3 file and jpeg file and I want single mp3 file with that picture as cover art [17:01] <jnvsor> sruz25: Ah, then you want this: http://ffmpeg.org/ffmpeg-formats.html#mp3 [17:01] <angelosalatino> Good Evening everyone [17:02] <jnvsor> evenin [17:02] <angelosalatino> Hi. I have a problem with ffmpeg... [17:02] <sruz25> yeah... that's it, thanks a lot [17:02] <angelosalatino> I'm trying to take video from my webcam [17:03] <angelosalatino> using this command: {ffmpeg -f alsa -ac 2 -i pulse -f video4linux2 -r 30 -s 1024x768 -i :0.0 -acodec pcm_s16le -vcodec libx264 -preset ultrafast -crf 0 -threads 0 output.mkv}, but the frame rate is not always the same.. [17:04] <jnvsor> angelosalatino: First, use pastebin for long commands, it will make irc cleaner [17:04] <angelosalatino> when I try to play the video it seem that goes fast... is there any way to fix to a certain value the frame rate? Thank you in advance [17:04] <jnvsor> angelosalatino: Second, replace your r flag with framerate - r says "The input will be so many fps" which isn't a guarantee with a webcam [17:05] <jnvsor> angelosalatino: http://pastebin.com/TbQ4AU8h [17:06] <angelosalatino> this is my pastebin http://pastebin.com/0xYAeq8d [17:07] <angelosalatino> ok. I'm going to try it.. [17:10] <jnvsor> angelosalatino: A few other things http://pastebin.com/wMxxS58f [17:13] <angelosalatino> It works [17:15] <angelosalatino> thank you [17:39] <GT1> is there any other way of delaying input video stream for an overlay then itsoffset? because it makes my video output laggy for some reason [18:35] <thebombzen> this is more of an x264 question, but can someone tell me which tune to use for 3D animation? for stuff like Tears of Steel it's clearly film, and Spongebob it's clearly animation, but stuff like "The Incredibles" or "Big Buck Bunny" have features of both. On one hand, they have smooth gradients, but on the other, they have way less noise and a cleaner image. Which tune should I use? [19:04] <cousin_luigi> Greetings. [19:04] <cousin_luigi> What's the best way to concatenate 3 mp4/h264 clips without reencoding? [19:11] <re-G> cousin_luigi: http://trac.ffmpeg.org/wiki/How%20to%20concatenate%20(join,%20merge)%20medi… [19:13] <cousin_luigi> re-G: I'm asking because I couldn't understand that:| [19:14] <re-G> which part of that [19:14] <cousin_luigi> re-G: Whether there would be reencoding. [19:14] <cousin_luigi> ok, trying the named pipe solution [19:15] <re-G> -c copy means stream copy [19:15] <re-G> without re-encoding [19:18] <cousin_luigi> ok, thanks [19:20] <cousin_luigi> bah, having bash problems [19:31] <cousin_luigi> The fifo version failed. The other one worked. Thanks [19:52] <viric> How can I delay or advance audio, wrt video? [19:58] <jedir0x> Hi, I'm looking for some ffmpeg consulting in the Los Angeles area. [20:16] <bparker_> jedir0x: what do you need? [20:17] <jedir0x> I'm looking for help building a library using FFMpeg with ObjC and Java bindings (for iOS and Android) [20:17] <jedir0x> it's a paid job obviously - was hoping to get a hold of Baptiste as he's listed on ffmpeg as being in Los Angeles (where we're at) [20:19] <bparker_> I'd be all over it if it weren't in those languages :p [20:19] <jedir0x> :) [20:19] <jedir0x> well, it's mostly c\c++ - just need the bindings for those platforms to said library [20:20] <bparker_> right [20:21] <albator> I m trying to output 2 files, one stream and one file locally , but it only creates a 48 bytes locally while the stream is okay [20:21] <llogan> jedir0x: you can send a message to the ffmpeg-devel mailing list [20:21] <llogan> albator: use the tee muxer [20:22] <albator> encoding stugg) -f flv "rtmp://url/app/live_xxxx_xxxx" -f mp4 file.mp4 [20:22] <albator> stuff [20:22] <llogan> ok!!! [20:22] <llogan> http://ffmpeg.org/ffmpeg-formats.html#tee [20:22] <llogan> http://trac.ffmpeg.org/wiki/Creating%20multiple%20outputs [20:22] <albator> but why does it work sometimes, and not all the time.. [20:23] <llogan> unknown without more information from you [20:23] <jedir0x> llogan, thanks - I'll do that. [20:25] <llogan> jedir0x: from the point-of-view from someone looking to hire a consultant, how do you think the web site can be improved to make this "better"? [20:25] <albator> what kind of informations.. just sent my command [20:38] <albator> looks like this http://pastebin.com/zkHTNx6v [20:39] <albator> tried to ouput to a flv file too witch would be fine for me, but I get " FLV does not support sample rate 48000, choose from (44100, 22050, 11025)" while I actually chose 44100 as you can see [20:40] <albator> are you a bot? [20:40] <albator> :D [20:49] <llogan> albator: fflogger is a bot [20:49] <albator> i m back [20:49] <albator> so what do you think of my paste [20:49] <llogan> it's missing the EXACT ffmpeg command and the COMPLETE console output so it is not very useful, unfortunately [20:53] <albator> it s not missing the command [20:54] <l_r> are there tutorials about libavformat/libavcodec? [20:56] <JEEB> you can look at the examples [20:56] <l_r> hey JEEB [20:57] <l_r> was i talking with you yesterday? [20:57] <l_r> about adding timestamps.. [21:02] <l_r> is it legal to static link libavcodec and libavformat in commerical apps? [21:41] <Plorkyeran> if you build as lgpl, the user has to be able to replace your copy of the library with one they built themselves, and you have to provide the source for the versions you used [21:42] <Plorkyeran> making the library replaceable can consist of either dynamically linking it, or making all of the object files available [21:42] <Plorkyeran> the latter is generally only done when it's the only option (i.e. on iOS) because it's sort of a pain [23:09] <i_s> anyone know what kind of error this is? https://s3.amazonaws.com/uploads.hipchat.com/26765/162949/RlyvOwZ9GTWLVBr/S… [23:09] <i_s> (i encoded these with the ffmpeg libs from webcam frames) [23:10] <klaxa> wrong colorspace maybe? [23:10] <klaxa> can you pastebin the command you ran? [23:10] <i_s> well it wasn't a command, i was using c [23:11] <klaxa> ah derp [23:11] <klaxa> >libs [23:11] <klaxa> well pastebin sourcecode then, maybe someone knows what's wrong [23:11] <i_s> but maybe it is color space [23:12] <i_s> hm, i don't think i can [23:13] <l_r> if i wrap raw h264 video without timestamps to mpegts, then i transform that in flv to be sent to a remote rtmp server, does the resulting flv have timestamps . basically a pipe lik h264_device | ffmpeg mpegts | ffmpeg flv to rtmp server [23:17] <i_s> anyone know of a good explanation of the relationship between pixel format and colorspace? [00:00] --- Tue Jan 14 2014

1 0

ffmpeg-devel.log.20140112
by burek 13 Jan '14

13 Jan '14

[00:03] <ubitux> hey btw [00:03] <ubitux> for those who have seen the movie the fifth estate [00:03] <ubitux> there is a ffmpeg reference at around 1h15 [00:04] <ubitux> "or we can use ffmpeg and toss it into final cut" [00:04] <JEEB> :o [00:04] <ubitux> who haven't* seen, actually [00:05] <ubitux> we should create a page, just like nmap, with references to ffmpeg in movies [00:06] <ubitux> llogan: hey, you should tweet that [00:06] <ubitux> ;) [00:10] <Compn> haha [00:11] <Compn> ubitux : nice catch [00:11] <Compn> i havent seen it [00:11] <Compn> how is it? bad against wikileaks or ok ? [00:11] Action: Compn guesses bad [00:11] <ubitux> i haven't seen the movie, someone told me [00:11] <ubitux> look at the scene, it's the heroes so i guess it's the good side assuming you're for wikileaks [00:17] <Timothy_Gu> llogan, you should really set up a "Trivia" wiki/website page recording all these stuff. [00:23] <{V}> ubitux, nope doesn't look like they think it's positive http://wikileaks.org/RELEASE-Julian-Assange.html [00:24] <ubitux> {V}: i thought the "it" from Compn was about ffmpeg, not the movie [00:25] <ubitux> yes ofc wikileaks didn't appreciate the movie at all [00:29] <cone-780> ffmpeg.git 03Michael Niedermayer 07master:2981f6a79fac: avcodec/dct-test: correct output bias of prores idct [00:29] <cone-780> ffmpeg.git 03Michael Niedermayer 07master:65801040c648: avcodec/dct-test: reproduce 4..1019 clipping when testing prores IDCT [00:31] <michaelni> BBB, dct-test says the prores asm idct has max_err=1 for 10bit so it should still be working fine [00:59] <BBB> dunno where the code is that drops 2 bits, can't easily find it, and it's 2yrs since I last looked at it [01:00] <BBB> but anyway, the asm is not bitperfect even though libav's is [01:04] <cone-780> ffmpeg.git 03Guillaume Martres 07master:c9fe0caf7a1a: hevc: clip pixels when transquant bypass is used [01:04] <cone-780> ffmpeg.git 03Guillaume Martres 07master:b00a8b4d194f: hevc: remove useless clip in FUNC(sao_band_filter)() [01:05] <michaelni> smarter, applied your proper fix, thanks alot! [01:05] <smarter> you're welcome :) [01:05] <JEEB> fabulous alots all around \o/ [01:06] <ubitux> BBB: you're refering to 6398c0f7e19fc3637930514cd81dba7cfbbacd4a? [01:06] <ubitux> (and 370d7ef2c7a9d00ee885da7ff5dec1b879b33650) [01:06] <michaelni> BBB bit-perfect to apple or the C code ? [01:08] <ubitux> (ah, and b87d882578807e9a45848a528891bd82a5165712) [01:14] <ubitux> BBB: btw, transpose should be done, WIP available in my branch, but horizontal doesn't work yet for some reason [01:43] <BBB> ubitux: awesome! [01:43] <BBB> michaelni: c code ofc [01:45] <BBB> ubitux: and the transpose itself works? that is, you can test that by itself and it generates expected output and converges? [03:15] <michaelni> BBB, see [FFmpeg-devel] [PATCH 5/5] avcodec/x86/proresdsp_init: x86 prores IDCT is bitexact again [03:15] <BBB> \o/ [03:18] <BBB> well wait that changes the results and in some cases amounts to cheating [03:18] <BBB> 2/5 is cheating, you're decreasing the accuracy of the coeffs [03:18] <BBB> (not that I care, I mean, it's prores, I'm just saying) [03:36] <michaelni> what point do 17bit coeffs make for 10bit data ? [03:36] <michaelni> and why was 17 enough and not 64bit ? [03:39] <michaelni> also the coeffs are good enough so that mpeg P frames dont accumulate vissible errors after 300 frames being added on top of each other [03:39] <michaelni> so there should be quite a bit of headroom for one I 1 frame [03:42] <michaelni> also PSNR didnt change for the coeff change, that is 2 digits after the decimal point, in the fate test [03:43] <michaelni> if someone has suggestions for some other test that i could do to ensure that the change has no negative impact ? [04:05] <Compn> did you check if theres any difference when converting it to bt709 or bt601 ? [04:05] <Compn> thats all anyone cares about [04:12] <michaelni> i didnt check but likely there will be a pixel that is +1 or -1 in the picture [04:12] <michaelni> or actually maybe the fate tests checked that already [04:20] <cone-780> ffmpeg.git 03Michael Niedermayer 07master:2ce4543286a8: avcodec/dct-test: add support for C prores IDCT [08:50] <ubitux> BBB: i don't know if the transpose works well, but that's not the problem currently [08:51] <ubitux> i have a weird crash even without the transpose [08:51] <ubitux> in the c code, if i transpose, call vertical, transpose, it works fine and passes fate [08:52] <ubitux> now if i do exactly the same but calling the ssse3 vertical instead [08:52] <ubitux> i have a crash [08:53] <ubitux> => 0x0000000000a589c5 <+2421>: pandn xmm7,XMMWORD PTR [r10+rsi*1] [08:53] <ubitux> here i have 0x18016a8 + 384, which is only aligned on 4 [08:53] <ubitux> i'm guessing this is the problem [08:54] <ubitux> but if so, that's troublesome and i don't really know how to solve that& [08:57] <ubitux> it was working fine when i wasn't transposing in-place but on a 16*16 temporary stack buffer [08:58] <ubitux> i wonder if i should reproduce that [09:26] <ubitux> it seems using an intermediate movu solves the problem [09:30] <ubitux> or maybe not. [09:39] <rcombs> ubitux: got a spare XMM register? You could use an unaligned load [09:39] <ubitux> that's what i was trying to do, but it seems i haven't :( [09:40] <rcombs> yeah, if you're pre-AVX and you're using a memory pointer as an operand for& pretty much anything but an explicitly unaligned load/store, that has to be aligned [09:40] <ubitux> what's strange is that i didn't have the issue with the vertical filter [09:41] <ubitux> as if vertical filter was always aligned [09:41] <rcombs> what file is this? [09:41] <ubitux> (which means i could eventually replace all the movu with mova for vertical) [09:41] <ubitux> well, any fate file [09:41] <ubitux> + my misc samples [09:42] <rcombs> there's no significant difference between movu and mova in performance on modern CPUs; you'll just get better performance when you're aligned [09:44] <ubitux> ok [09:45] <cone-885> ffmpeg.git 03Stefano Sabatini 07master:d497141b8594: examples/muxing: simplify video PTS setting [09:52] <ubitux> lol, got it working [09:52] <ubitux> and my transpose was working out of the box.. [09:55] <nevcairiel> thats always suspicious [09:57] <ubitux> nevcairiel: are you questioning my awesomen^Wluck? [09:57] <nevcairiel> :D [09:57] <nevcairiel> i would never! [09:59] <ubitux> the thing is that i tested the transpose in a PoC, i only had to do the memory store [09:59] <ubitux> and luckily it was working out of the box [10:01] <ubitux> https://github.com/ubitux/FFmpeg/commit/48afad80c2b167d9d761c426a24673c2e34… [10:01] <ubitux> most ugly thing in the world [10:02] <ubitux> i have no available registers, so i'm doing 3 movu instead [10:06] <saste> what about removing ft_load_flags from drawtext? [10:06] <saste> ubitux, ^ ? [10:07] <ubitux> why? [10:07] <ubitux> 5340 decicycles in ff_vp9_loop_filter_h_16_16_ssse3, 2096476 runs, 676 skips [10:07] <ubitux> 3964 decicycles in ff_vp9_loop_filter_v_16_16_ssse3, 2096762 runs, 390 skips [10:07] <ubitux> i need to improve that& [10:08] <saste> ubitux, because it doesn't make much sense to fiddle with FT flags [10:08] <nevcairiel> h is transpose+v? [10:08] <ubitux> nevcairiel: yes [10:09] <ubitux> saste: dunno, couldn't they be used to have a bitexact output so we could add the filter to fate? [10:09] <saste> ubitux, oh i just realized your "why" was not related to my question [10:09] <ubitux> saste: it was :) [10:10] <saste> ubitux, the flags are probably too much low level to be exposed to the user [10:10] <ubitux> even the hinting? [10:10] <saste> there is something potentially useful, for example monochrome, but they can be set through a dedicated option/flag [10:11] <saste> about bitexact, that really depends on FT version [10:11] <ubitux> well maybe there is a bitexact flag in those [10:11] <ubitux> (pedantic?) [10:11] <ubitux> anyway, do they cause trouble? [10:11] <saste> ubitux, see ramiro's patch [10:12] <saste> it is messing with render, and conflicting with the way we let the user ser +render [10:12] <saste> *set [10:12] <ubitux> mmh [10:12] <ubitux> dunno, no opinion then [10:20] <ubitux> vertical definitely works assumed an aligned block all the time [11:28] <ubitux> horizontal is only 3x faster :( [11:38] <ubitux> 8.20s to 6.28s on overall decode time though [11:38] <ubitux> (46 to 60 fps) [11:38] <nevcairiel> just from one function? thats neat :) [11:40] <ubitux> no, the two [11:40] <ubitux> :p [11:40] <ubitux> https://github.com/ubitux/FFmpeg/commit/9dcbf146dfed4c9ad00ea17beb50231b4c6… [13:42] <BBB> ubitux: the destination memory is likely not aligned? [13:42] <BBB> ubitux: use LOCAL_ALIGNED if it's on the stack, or DECLARE_ALIGNED if it's in a struct that is allocated with av_malloc() [13:47] <BBB> michaelni: I care more from the theoretical point of view, you're trying to save 2 bits of coefficient precision for the sole sake of making the idct like what, 4 instructions faster per 1d or so? I don't remember the exact amount but it was something trivial like that. I don't think that's necessary anymore. But do what you wish, it's your idct, I'm not going to pretend that I care about prores, but it does yet again confirm apple's point, we [13:47] <BBB> local tests with 1 person visible checking that it "looks good to him" and then commit as the de-facto implementation to millions of users, and we don't care the least that the results changed - "oh, it still looks ok'ish", that's not how video comparisons should be done [14:05] <michaelni> BBB the C IDCT has ATM a MSE of 0.04873672, the SSE2 one of 0.02940234 [14:05] <michaelni> so my patchset actually improves accuracy [14:06] <michaelni> that is as a whole, i didnt yet test them individually [14:07] <michaelni> also using 17bit coeffs when every SIMD implementation will need to work with 15bit+1sign bit is quite inconvenient [14:08] <ubitux> BBB: are you commenting on the final patch? [14:12] <BBB> ubitux: ? [14:12] <BBB> oh new patch yay [14:12] <BBB> no, your comment yesterday (or this morning) [14:12] <ubitux> it's done and work now [14:13] <ubitux> but see the note [14:13] <BBB> movu... [14:13] <BBB> that sucks [14:13] <BBB> use mova [14:13] <BBB> and make the memory aligned ;) [14:14] <ubitux> for vertical it's mova [14:14] <BBB> oh I think I see why [14:14] <BBB> yeah that's actually expected [14:14] <BBB> it's because vertical loads from topleft block position (which is aligned) - x*stride (which is aligned) [14:14] <BBB> horizontal loads from topleft position - x [14:15] <BBB> topleft position is aligned, but x isn't [14:15] <BBB> and can't be [14:15] <BBB> so I guess that's ok [14:15] <BBB> nevermind my comments the [14:15] <BBB> you use 3 stores after the transpose right, not all 16 [14:15] <ubitux> huh? [14:16] <BBB> awh [14:16] <BBB> your transpose stores [14:17] <BBB> so the ideal situation is that the transpose doesn't just work on memory -> memory [14:17] <BBB> but the pre-lf transpose goes from memory -> mostly registers [14:17] <michaelni> BBB, IDCT_1D has 112 instructions less if i didnt miscount. that would be 14 per 8x1 dct [14:17] <ubitux> BBB: yes that's ugly, but actually there is not much i can do about that [14:17] <BBB> and the reverse for the post-lf transpose (mostly registers -> memory) [14:17] <michaelni> but possible some of this difference was from other changes, i didnt double check [14:17] <ubitux> BBB: i re-use the pre-loaded reg at first [14:17] <ubitux> BBB: but i need to store them anyway for later re-read [14:18] <BBB> all of them? [14:18] <BBB> hm... [14:18] <ubitux> i tried to remove some store but it didn't work as expected [14:18] <ubitux> there is probably a way to remove some, but it might be tricky [14:18] <ubitux> unless i'm missing something obvious [14:19] <BBB> nah it's not obvious, it's just grinding :-p [14:19] <BBB> maybe we should add a fixme comment to investigate that later? [14:19] <ubitux> sure, ok [14:19] <BBB> I don't mind committing as-is, just would like to document things we think we can improve later on [14:20] <BBB> you changed the filter14 at the end to use movu unconditionally [14:20] <ubitux> ? [14:20] <ubitux> +%ifidn %1, v [14:20] <ubitux> +%define movu mova [14:20] <BBB> euwh :-p [14:21] <BBB> make that movx please, not good to overwrite default macros [14:21] <BBB> just so it's clear it's sometimes movu sometimes mova :) [14:21] <BBB> or something not mova and movu and movh and movd and movq etc. [14:21] <ubitux> so s/movu/movx/ and a define for v and h? [14:22] <BBB> yes [14:22] <ubitux> ok [14:22] <BBB> rest looks ok [14:22] <ubitux> i'm definitely not satisfied with horizontal [14:22] <BBB> well, it's a good first try, I mean, it's really a lot faster [14:22] <ubitux> because of MASK_APPLY_FU, and the memory stores of the transpose [14:22] <ubitux> but yeah, it's like 3x faster so... [14:22] <BBB> yeah I saw the mask_apply_fu define [14:23] <BBB> one thing you could do is allocate more stack [14:23] <BBB> for the h one [14:23] <BBB> stack is guaranteed to be aligned, and then use that for temp storage [14:23] <BBB> then you can always use mova, that should help quite a bit [14:23] <ubitux> mmh [14:23] <BBB> (the initial load and final store are still movu, but the rest is mova then) [14:33] <michaelni> BBB the coeff change patch changes MSE from 0.05012422 to 0.04890000 [14:33] <BBB> you know the prores fdct? [14:33] <BBB> whoa [14:36] <michaelni> not really but these tests use the double precission reference fdct [14:36] <michaelni> iam sure the fdct used in the encoder could be iproved [14:49] <kierank> i would assume the authors of the decoder have a hook into the fdct [14:49] <BBB> sure, but the idct in proresdec should optimize for the fdct in proresenc, not some reference floating point fdct we used for academic purposes, that's not used [14:49] <BBB> e.g. the vp9 idct and fdct were definitely tuned towards each other [14:58] <michaelni> BBB, for vp9 you where in control of both fdct and idct, for prores apple is in control more or less well actually less. so there are already multiple fdcts around that are used [14:58] <BBB> ? [14:59] <BBB> I don't think apple cares the least about what ffmpeg outputs [14:59] <BBB> you should assume prores is a corporate product with - for their purposes - one fdct and one idct [14:59] <kierank> apple is in control yes, so ffmpeg must match apple [14:59] <BBB> whatever ffmpeg does is, in their words, introducing artifacts [15:00] <michaelni> yes, we should match apples fdct with our idct, sorry i misunderstood you [15:00] <BBB> brb [15:01] <michaelni> i dont have test samples from this fdct though so the reference one is the best that i can easily test against [15:01] <kierank> michaelni: you should ask the authors to tell you where in the dll the apple fdct is [15:20] <pross-au> given dll & debugger, they're usually easy to spot. [15:53] <cone-846> ffmpeg.git 03Kostya Shishkov 07release/1.1:5dcc17992430: vc1: Reset numref if fieldmode is not set [15:53] <cone-846> ffmpeg.git 03Aurelien Jacobs 07release/1.1:3e089e8f7158: matroskadec: use correct compression parameters for current track CodecPrivate [15:53] <cone-846> ffmpeg.git 03Martin Storsjö 07release/1.1:12479588d789: sdp: Check that fmt->oformat is non-null before accessing it [15:53] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:343c87ac19c8: rv30: fix extradata size check. [15:53] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:f194f2be418a: eacmv: check the framerate before setting it. [15:53] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:5e7a5dd70b51: gifdec: return meaningful error codes. [15:53] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:c5c7e3e6f7cf: gifdec: check that the image dimensions are non-zero [15:53] <cone-846> ffmpeg.git 03Michael Niedermayer 07release/1.1:ce795ac0f53e: Merge commit 'c5c7e3e6f7cf17943c04bd078f260eaf789afbc9' into release/1.1 [16:13] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:819541ff833d: gifdec: convert to bytestream2 [16:13] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:ffa83bcc4937: lzw: switch to bytestream2 [16:13] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:a8f6d93071a8: pmpdec: check that there is at least one audio packet. [16:13] <cone-846> ffmpeg.git 03Justin Ruggles 07release/1.1:24a8dfd37b45: lavr: check that current_buffer is not NULL before using it [16:13] <cone-846> ffmpeg.git 03Luca Barbato 07release/1.1:0e8ae6d10c60: mpegvideo: Drop a faulty assert [16:13] <cone-846> ffmpeg.git 03Michael Niedermayer 07release/1.1:7e3437989769: Merge commit '0e8ae6d10c609bb968c141aa2436413a55852590' into release/1.1 [16:28] <cone-846> ffmpeg.git 03Luca Barbato 07release/1.1:d6d2617d07fc: avio: Use AVERROR_PROTOCOL_NOT_FOUND [16:28] <cone-846> ffmpeg.git 03Michael Niedermayer 07release/1.1:e776a1e8f37d: ac3dec: fix outptr increment. [16:28] <cone-846> ffmpeg.git 03Michael Niedermayer 07release/1.1:cdc47c48137f: omadec: check GEOB sizes against buffer size [16:28] <cone-846> ffmpeg.git 03Michael Niedermayer 07release/1.1:35f9a0896ee6: omadec: Fix wrong number of array elements [16:28] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:51ff11647f8d: pcx: round up in bits->bytes conversion in a buffer size check [16:28] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:7b337b122959: truemotion1: make sure index does not go out of bounds [16:28] <cone-846> ffmpeg.git 03Michael Niedermayer 07release/1.1:c693ccb89a9e: Merge commit '7b337b122959b9bf634c31b549892df974f35b40' into release/1.1 [16:44] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:7c214e313c92: avidec: fix a memleak in the dv init code. [16:44] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:26221a54eca3: motionpixels: clip VLC codes. [16:44] <cone-846> ffmpeg.git 03Serhii Marchuk 07master:c917cde9cc52: mpegts muxer, DVB subtitles encoder: common DVB subtitles payload [16:44] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:cbf51c4d36af: matroskadec: pad EBML_BIN data. [16:44] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:f9f2591beb11: alsa-audio-dec: explicitly cast the delay to a signed int64 [16:44] <cone-846> ffmpeg.git 03Luca Barbato 07release/1.1:265603675722: ffv1: Assume bitdepth 0 means 8bit [16:44] <cone-846> ffmpeg.git 03Luca Barbato 07release/1.1:0358a099f8ab: indeo4: Check the block size if reusing the band configuration [16:44] <cone-846> ffmpeg.git 03Michael Niedermayer 07release/1.1:1203e92181b4: Merge commit '0358a099f8abe60230dc2e5bec59bfceb7d1be07' into release/1.1 [16:57] <cone-846> ffmpeg.git 03Luca Barbato 07release/1.1:03457cabd618: indeo4: Check the inherited quant_mat [16:57] <cone-846> ffmpeg.git 03Anton Khirnov 07release/1.1:481e55eba7a7: audio_mix: fix channel order in mix_1_to_2_fltp_flt_c [16:57] <cone-846> ffmpeg.git 03Luca Barbato 07release/1.1:1d7a453dcfe4: prores: Reject negative run and level values [16:57] <cone-846> ffmpeg.git 03Luca Barbato 07release/1.1:e361fde8b011: avi: properly fail if the dv demuxer is missing [16:57] <cone-846> ffmpeg.git 03Reinhard Tartler 07release/1.1:f53a5332b017: Prepare for 9.11 RELEASE [16:57] <cone-846> ffmpeg.git 03Luca Barbato 07release/1.1:5bbee02ae04f: shorten: Extend fixed_coeffs to properly support pred_order 0 [16:57] <cone-846> ffmpeg.git 03Martin Storsjö 07release/1.1:d149c14a2263: mov: Don't allocate arrays with av_malloc that will be realloced [16:57] <cone-846> ffmpeg.git 03Luca Barbato 07release/1.1:61057f4604eb: avi: directly resync on DV in AVI read failure [16:57] <cone-846> ffmpeg.git 03Derek Buitenhuis 07release/1.1:5ae7ed3aa4f3: nut: Fix unchecked allocations [16:57] <cone-846> ffmpeg.git 03Luca Barbato 07release/1.1:65830277d2d2: prores: Add a codepath for decoding errors [16:57] <cone-846> ffmpeg.git 03Michael Niedermayer 07release/1.1:f479c178943f: Merge commit '65830277d2d2ee3658e1f070a61044fff261ed3e' into release/1.1 [17:07] <cone-846> ffmpeg.git 03Tim Walker 07release/1.1:a0866c71293d: shorten: Fix out-of-array read [17:07] <cone-846> ffmpeg.git 03Michael Niedermayer 07release/1.1:9b89824f2022: Merge remote-tracking branch 'qatar/release/9' into release/1.1 [18:06] <ubitux> BBB: i'm going to try the stack thing [18:07] <ubitux> but reading x86inc.asm, it seems the stack align is somehow tricky, and i don't really understand [18:07] <nevcairiel> shouldnt it do this for you [18:08] <ubitux> ; and an extra register will be allocated to hold the original stack [18:08] <ubitux> ; pointer (to not invalidate r0m etc.). To prevent the use of an extra [18:08] <ubitux> ; register as stack pointer, request a negative stack size. [18:08] <ubitux> i'm a bit uncomfortable with this :) [18:08] <nevcairiel> oh yeah that always confused me as well [18:08] <nevcairiel> i dont suppose you still have a spare register eh [18:09] <ubitux> :( [18:09] <ubitux> i'm already using 8 reg [18:09] <nevcairiel> well then you are over 32-bit anyway, so one more doesnt hurt [18:09] <nevcairiel> :D [18:09] <ubitux> :) [18:10] <JEEB> aye [18:15] <cone-846> ffmpeg.git 03Yu Xiaolei 07master:842b8f4ba2e7: fix build with gas-preprocessor.pl [18:15] <cone-846> ffmpeg.git 03Michael Niedermayer 07master:6044f161d3bc: Revert "swscale: disable ARM code until its build failure with clang/iphone is fixed" [18:24] <BBB> ubitux: it does it for you [18:25] <BBB> ubitux: x86inc.asm indeed does creepy stuff, that's why we don't do it ourselves but let it handle it for us [18:25] <ubitux> so i don't have to care at all about it? [18:25] <BBB> ubitux: cglobal name, args, gprs, xmms, memory, name1, name2, etc [18:25] <BBB> exactly [18:25] <BBB> that's the whole point of x86inc.asm - don't care about complicated stuff that is annoying [18:25] <BBB> just the easy fun stuff [18:25] <ubitux> so i just add a 256 field and play with rsp? [18:26] <BBB> don't write into rsp [18:26] <BBB> :-p [18:26] <BBB> but yes [18:26] <BBB> if you add a 256 field, [rsp+0] until [rsp+256-16] is all yours to mova into [18:26] <BBB> also do you remember http://blogs.gnome.org/rbultje/2014/01/12/brute-force-thread-debugging/? I finally got to write it down [18:27] <ubitux> aha! [18:27] <BBB> I think that's now 1.5 years ago [18:27] <BBB> I was lazy I guess [18:27] <ubitux> thx :) [18:27] <ubitux> i plan to write a few things about that asm initiation [18:28] <BBB> cool [18:28] <BBB> x86inc.asm is awesome, maybe we should tell more people about it, yes [18:30] <kierank> BBB: quite interesting post [18:33] <jnvsor> Is the overlay filter supposed to drop frames from the higher fps input to match the lower one? [18:34] <nevcairiel> i would think the master image would dictate that, and the image to-be-overlayed would be dropped or repeated if required [18:34] <nevcairiel> but i have no clue how its actually implemented [18:36] <jnvsor> Well my overlay input has a variable framerate (webcam through v4l2) and when the webcam stutters so does the master input, filing a bug now [18:37] <BBB> kierank: thanks, that's why I posted it here (I don't think there's a planet ffmpeg, is there?) [18:38] <kierank> I would submit it to reddit/hacker news but I don't think it's a good idea [18:39] <kierank> you'll be told that the decoder should be in Haskell [18:40] <ubitux> :D [18:40] <ubitux> "that would have never happened with erlang!" [18:40] <BBB> feel free to submit, I don't know how that works :-p [18:41] <kierank> done [18:42] <ubitux> plussed [18:43] <kierank> BBB: dunno if it's worth mentioning the ratio of horizontal mvs to vertical mvs and how that helps (I think ffmpeg's frame threading is the same as x264's) [18:44] <BBB> nah, frame-mt is more a "how this came about", the post is more about debugging it, rather than why frame-mt helps at all [18:44] <BBB> but feel free to add a comment saying that, it's indeed helpful to know that kind of stuff as a basis for why frame-mt helps at all [18:45] <nevcairiel> i always figured frame-mt works because mvs mostly just reference one or two rows of MBs, so when those are done decoding the next frame can start decoding its top rows, and so forth [18:46] <kierank> horizontal movement dominates in a lot of motion [18:46] <kierank> 5:1 at least [18:46] <nevcairiel> guess that too [18:46] <BBB> right, vertical motion tends to be small [18:47] <BBB> so although there is some lag between progress of frame decoding in each thread, it's close enough that it works for large resolution content [18:47] <BBB> and that's when you need it most anyway [18:47] Action: kierank sends that article to someone who had slice thread heisenbugs with 4k [18:48] <BBB> lol [18:48] <BBB> frame-mt is happy now, right? [18:48] <ubitux> yes but valgrind isn't [18:48] <kierank> said person is having problems making a test case [18:48] <BBB> :-p [18:49] <BBB> valgrind valgrind [18:49] <BBB> yeah I'll look into that [18:49] <ubitux> ;) [18:49] <ubitux> BBB: btw, should i push the 3 small itxfm patches? [18:50] <BBB> I only saw one? [18:51] <ubitux> really? [18:51] <ubitux> https://github.com/ubitux/FFmpeg/compare/vp9-itxfm [18:51] <BBB> the one where you merge the %2 and %3 in mulsub_2w_2x [18:51] <BBB> oh! [18:51] <BBB> I get it know [18:51] <BBB> it shows me all 3 diffs merged [18:51] <BBB> but it's actually 3 patches [18:52] <BBB> I didn't see that :-p [18:52] <BBB> yes they're all fine [18:52] <ubitux> ok, will apply tonight [18:54] <BBB> github is funny [19:03] <ubitux> the stack thing is really going to simplify a lot the horizontal [19:04] <jnvsor> If the overlay filter got a v4l2 stream that wasn't sending frames regularly, would it block waiting for the next frame? [19:16] <BBB> ubitux: cool [19:16] <ubitux> since i can hardcode the position [19:16] <ubitux> without jungling with dst*q pointers [19:16] <ubitux> it basically removes all the lea [19:18] <BBB> only for the h [19:18] <BBB> the v still needs the leas [19:18] <BBB> and the h does need the leas for initial load and final store [19:19] <BBB> (pre-lf transpose load and post-lf transpose store) [19:19] <ubitux> yes only for h [19:19] <ubitux> but still, it's cool to see simplification in the slowest one [19:32] <BBB> yes agreed [19:35] <ubitux> yay working out of the box [19:35] <ubitux> let's bench this shit [19:39] <ubitux> 5259 4936 [19:39] <ubitux> i love you BBB ;) [19:40] <ubitux> so exactly 1k more than vertical [19:40] <nevcairiel> so, you are happy now? :) [19:41] <ubitux> yeah i guess [19:41] <nevcairiel> did the FU go away? [19:41] <ubitux> yup [19:43] <ubitux> and the file is now 666 lines [19:43] <ubitux> so i'm really happy [19:43] <nevcairiel> hehe [19:55] <ubitux> BBB: https://github.com/ubitux/FFmpeg/commit/03b6166e7ad0413e44e4aaf246f10eab199… [19:55] <ubitux> comments? [19:58] <ubitux> nevcairiel: i should have closed firefox from the beginning, i got free cycles doing so [19:58] <nevcairiel> lol [19:59] <nevcairiel> firefox the cycle eater [20:11] <cone-846> ffmpeg.git 03Luca Barbato 07master:a1f5164814ac: vc1dsp: K&R formatting cosmetics [20:11] <cone-846> ffmpeg.git 03Michael Niedermayer 07master:6bd001d76657: swscale: disable ARM code until its build failure with clang/iphone is fixed [20:12] <cone-846> ffmpeg.git 03Michael Niedermayer 07master:4daf8bc31b91: Merge remote-tracking branch 'qatar/master' [22:33] <BBB> ubitux: results look good, I don't think I had any further comments [22:34] <BBB> so ok to commit [22:34] <ubitux> cool :) [22:34] <ubitux> will push in a few minutes with the itxfm ones [22:34] <BBB> \o/ [22:34] <BBB> nice work [22:36] <BBB> poke me when it's pushed, we should be beating the crap out of libvpx at this point [22:41] <cone-846> ffmpeg.git 03Clément BSsch 07master:7c55ee6168ea: vp9/x86: merge IDCT coef macros. [22:41] <cone-846> ffmpeg.git 03Clément BSsch 07master:c9aa0b8f70b1: vp9/x86: remove reg redundancy in VP9_MULSUB_2W_2X. [22:41] <cone-846> ffmpeg.git 03Clément BSsch 07master:e11ceea68ff3: vp9/x86: factor out some code in VP9_UNPACK_MULSUB_2W_4X. [22:41] <cone-846> ffmpeg.git 03Clément BSsch 07master:af68bd1c06ff: vp9/x86: add ff_vp9_loop_filter_[vh]_16_16_ssse3(). [22:41] <ubitux> BBB: here you go [22:41] <BBB> yay [22:45] <BBB> rebase, rebuild, reblah [22:46] <ubitux> BBB: anyway, [22:47] <ubitux> BBB: thanks a lot for your teachings and mentoring :) [22:47] <ubitux> 'learned a lot with that stuff [22:48] <BBB> don't forget the 8/4 loopfilters! :-p [22:48] <ubitux> heh [22:49] <ubitux> i'm not leaving :D [22:49] <ubitux> yeah i'll do those [22:49] <BBB> \o/ [22:49] <BBB> can you measure libvpx vs ffvp9 on your computer? [22:49] <BBB> I can upload a few more files for that purpose if you want, I generated a few a while ago [22:50] <ubitux> mmh i'm going to leave soon, but feel free to share the tests you want me to run [22:52] <BBB> ok [22:52] <ubitux> don't forget to remove the assert-level option for the bench [22:52] <ubitux> :D [23:21] <jnvsor> I notice cone monitors different repos for every developer, how do you guys coordinate? [23:22] <ubitux> it doesn't [23:22] <ubitux> unless i misunderstood you [23:22] <jnvsor> Well your commits at 22:40 are on your github repo but not the main ffmpeg repos at github or ffmpeg.org [23:23] <jnvsor> Ah wait I've got them from ffmpeg.org now [23:24] <jnvsor> My bad [00:00] --- Mon Jan 13 2014

1 0

ffmpeg.log.20140112
by burek 13 Jan '14

13 Jan '14

[09:07] <tvkid> i have an embedded board with 454 MHz ARM9 processor and 64MB RAM...ethernet interface is 10Mbit it has a video input 640x480@30fps can i transcode this input video to some codec so that i can output this stream using the ethernet interface...i am not sure whether this computer is fast enough to do that can someone confirm? [09:08] <tvkid> this processor will be dedicated for this task..nothing else needs to be done [12:25] <edoardo> Hi, how do I install ffmpeg on Ubuntu? [12:25] <edoardo> Doing apt-get install ffmpeg installs avconv [12:34] <JEEB> edoardo, you can either use the static binaries provided by some folk, or you can build yourself [12:35] <edoardo> JEEB, i found a repo, ppa:jon-severinsson/ffmpeg, is it ok? [12:35] <JEEB> no idea [12:35] <edoardo> well i'll try and report then [12:35] <edoardo> see you later [12:38] <edoardo> yeah it works! [13:24] <BtbN> Does ffmpeg support indexing flv files? [14:19] <l_r> does h264 have timestamps? [14:31] <albator> hello [14:32] <albator> do you know why most of the time when i use the "filelist.txt" in " -f -i "filelist.txt" blabla , I get "impossible to open "file no such file or directory" ? [14:32] <albator> -f concat -i filelist.txt sorry [16:26] <albator> anyone? :D [16:40] <BtbN> because it failed to open some file [17:19] <albator> yes but why [17:20] <albator> i push the command in the directory where both listfile.txt and video files are located [17:20] <albator> i tried putting full path and it doesn't make any differences [17:21] <BtbN> well, it doesn't find some of the input files obviously [17:21] <albator> I got that obv.. but it doesn't explain why ;) [17:21] <BtbN> there is no "why" [17:22] <albator> ok [17:22] <albator> maybe it s an ffmpeg bug [17:22] <BtbN> just fix your filenames and paths.. [17:22] <albator> problem reading long list of caracters [17:23] <albator> my filename and path are correct.. i run the command from the folder itself [17:23] <BtbN> it wouldn't complain about it if it was [17:29] <albator> u read me? [17:48] <BtbN> albator, i can assure you, the concat demuxer is happy, even with extremely long lists of files. So the problem is somewhere on your end... [17:48] <albator> i guess so [17:57] <Turingi> I was trying to do something like this: ffmpeg -f mjpeg -i "http://camera-url" -crf 22 -c:v libx264 -preset ultrafast -tune zerolatency -y "capture.avi" -f image2 'mypipe.jpeg' (basically, create a video file from a network mjpeg stream and also spit out the individual frames into a pipe, where a serice is reading them for various processing) [17:57] <Turingi> mypipe.jpeg is a pipe and not a real file [17:58] <Turingi> s/serice/service/ [18:42] <bluebird> Hi, I'm trying to write a C program to encode some motion jpegs into an h264 video. I used doc/examples/decoding_encoding.c as a reference at it works fine. Now I am trying to make the frame rate variable but setting frame->pts or pkg.pts or pkt.duration doesn't seem to change anythin. Does someone have a simple example for encoding videos with variable frame rate? [20:00] <Mavrik> bluebird, changing frame pts should of course make variable framerate video [20:00] <Mavrik> (taking into account you have proper container and a player that can handle that) [20:59] <thebombzen> why does the topic say that "FFmpeg 2.1.1 is released" if it doesn't say that on ffmpeg.org [21:25] <jnvsor> thebombzen: Because it was. 2 months ago https://github.com/FFmpeg/FFmpeg/releases/tag/n2.1.1 [21:28] <thebombzen> Then why doesn't it say that on ffmpeg.org [21:32] <DopeLabs> hello [21:35] <jnvsor> thebombzen: http://ffmpeg.org/download.html#releases [21:35] <DopeLabs> i was curious about something... when using a still image as the source for your video stream output using -loop 1 -i image.jpg does ffmpeg continually read that file? for example, would i be able to start a stream up and just replace teh image file with a different one, if ffmpeg will then start streaming frames from the new image mid stream... so i wouldnt have to restart the command [22:02] <mikar> hey, quick question, what language is ffmpeg written in? [22:02] <viric> c [22:03] <viric> you can read, modify and redistribute it [22:35] <albator> BtbN i think i found out.. it doesn't handle well filename with space and/or punctuation symbols [22:36] <albator> despite the "" [22:44] <bencc> is it possible to embed arbitrary meta-data in a video file? [23:30] <l_r> does h264 have timestamps? [23:35] <Paranoialmaniac> l_r: yes and no. because it's optional [23:36] <l_r> Paranoialmaniac, how can i add a timestamp to an h264 video which originally does not have it? [23:36] <JEEBsv> either write your own bit/byte stream parsing and writing code, or just mux it into a container with time stamps :P [23:37] <l_r> hmm [23:37] <Paranoialmaniac> technically, add pic_timing sei into h264 stream itself [23:38] <l_r> is it possible to add it via before dd ? (consider that fortunately the source of ffmpeg is dd) [23:38] <l_r> *afterdd [23:38] <JEEBsv> I'd just say that it'd be simpler for you if you just put the H.264 stream into a container [23:39] <JEEBsv> unless you want to write your own code to deal with H.264 byte/bit streams [23:39] <l_r> but the h264 video is sent as it is to an rtmp server [23:39] <l_r> if i put the video in a container then will it work? [23:39] <l_r> i think not [23:40] <JEEBsv> I thought you could do mpeg-ts in rtmp, no? [23:40] <Paranoialmaniac> why do you think so :) [23:40] <JEEBsv> or wait [23:40] <JEEBsv> rtmp was thea dobe thing [23:40] <JEEBsv> which is FLV [23:40] <JEEBsv> so yes, you have time stamps [23:40] <l_r> hm [23:40] <l_r> wait i ll give you the command to show what s going on [23:42] <l_r> dd if=/dev/video0 bs=256k | ffmpeg -analyzeduration 1 -f h264 -r 15 -i - -vcodec copy -an -f flv rtmp://server/stream/ [23:42] <JEEBsv> so yes, your output most definitely has timestamps [23:43] <l_r> what happens there is that at night the framerate changes, then ffmpeg think it's still 15 fps instead of 10 [23:43] <JEEBsv> yes [23:43] <l_r> even if i use -re [23:43] <l_r> it does not adapt [23:43] <JEEBsv> you're giving it raw H.264 most probably without timestamps [23:43] <l_r> then resulting video is accelrated at times [23:43] <l_r> yes [23:43] <l_r> how can i add them [23:44] <JEEBsv> how do you get them in the first place? [23:44] <l_r> what is "them" ? [23:44] <l_r> timestamps? [23:44] <JEEBsv> yes [23:44] <l_r> i could write my own "dd" which gets the timestamps from the v4l2 device (i know how to do that) [23:45] <l_r> what i dont know is where to put this metadata in the blocks i am reading from the dev [23:45] <l_r> where and how/which format [23:45] <JEEBsv> so yeah, your input has them and you're just using dd for random reason x [23:46] <l_r> the video4linux2 plugin in ffmpeg is buggy [23:46] <l_r> so i have to use dd [23:46] <JEEBsv> ok [23:46] <JEEBsv> you would end up writing code in this case anyways [23:46] <l_r> yes [23:46] <JEEBsv> so you can either contribute to ffmpeg's thing [23:46] <JEEBsv> or you can make a thing that muxes that raw H.264 into a container [23:46] <JEEBsv> with timestamps [23:46] <JEEBsv> mpeg-ts or flv or so [23:46] <JEEBsv> whatever can be read on-the-fly [23:46] <JEEBsv> and has timestamps [23:47] <JEEBsv> much better than you hacking raw H.264 [23:47] <JEEBsv> esp. if your output is FLV [23:47] <l_r> and audio? what about multiplexing dd with audio directly in ffmpeg? [23:47] <l_r> to have timestamps [23:47] <l_r> audio could give timestamps [23:47] <l_r> no? [23:48] <JEEBsv> .. [23:48] <JEEBsv> I have no idea what you are trying to imply with audio regarding video PTS [23:48] <l_r> ok, forget what i said. let's switch back to your idea [23:49] <l_r> so you are saying to use a container [23:49] <l_r> so -vcodec anything instead of .vcodec copy for the input [23:49] <JEEBsv> yes, that basically leaves you with the least amount of custom implementation of stuff [23:49] <l_r> right? [23:49] <JEEBsv> uhh what [23:49] <l_r> i mean [23:50] <JEEBsv> you were talking about implementing the app that feeds ffmpeg, yes? [23:50] <JEEBsv> (if you do not want to fix the ffmpeg v4l2 thing) [23:50] <l_r> i don't care about what manipulations i do to the input. the important part is that the output of ffmpeg is the same as the above line [23:50] <JEEBsv> uhh, what [23:50] <JEEBsv> so you want it to always take it as 15fps? [23:50] <JEEBsv> wouldn't it then already work? [23:50] <l_r> no, i mean, the output format [23:50] <l_r> not really the fps [23:50] <JEEBsv> or do you mean something else with "the same as the above line" [23:51] <l_r> -f flv [23:51] <JEEBsv> I'm really not following you, we're talking about the thing that gets fed to the ffmpeg, no? [23:52] <JEEBsv> since you seemingly have no interest in fixing the v4l2 thing in ffmpeg itself [23:52] <l_r> so what does the muxing? i suppose ffmpeg, this would result in a pipe of as in dd | ffmpeg | ffmpeg ? [23:53] <JEEBsv> scroll back where I said that you will have to write your own code in any case [23:54] <JEEBsv> anyways, the route of least resistance if we forget about fixing ffmpeg itself would be to write an app that takes your H.264 stream, and then reads it with the libav* libraries. Then you pass that parsed stream into a libavformat muxer and you use the timestamps you're getting from somewhere to mux that video stream with timestamps [23:54] <JEEBsv> and then you can feed that to ffmpeg [23:54] <JEEBsv> or heck, that app of yours could do the rtmp stuff as well [23:54] <JEEBsv> since it's all libavformat [23:55] <l_r> looks a long task, not difficult but long [23:55] <JEEBsv> shorter than your original idea of "just adding time stamps to the H.264 stream" [23:56] <JEEBsv> not to mention that I have no idea if ffmpeg can parse those damn SEI messages [23:56] <JEEBsv> it's simpler because you don't have to actually implement anything [23:56] <l_r> ok [23:57] <JEEBsv> and the first step could be to just read the input stream and mux it into flv, and output that to stdout [23:57] <JEEBsv> that shouldn't be too hard [23:57] <JEEBsv> then you'd have your_app | ffmpeg [23:58] <JEEBsv> and when you get to it you could just remove ffmpeg from the thing altogether :P [00:00] --- Mon Jan 13 2014

1 0

ffmpeg-devel.log.20140111
by burek 12 Jan '14

12 Jan '14

[00:05] <j-b> E-AC3 in wav? [00:06] <ubitux> e-ac3 in wav in mxf in dcp [00:06] <ubitux> in a zip [00:06] <j-b> brrr [00:08] <j-b> ff_codec_wav_guids, that's a scary table [00:09] <ubitux> my favorite thing with wav is SMV [00:09] <JEEBsv> oh, libavformat or codec had something like tha? [00:10] <JEEBsv> because I might end up writing a demuxer for a format that jus contains captured PCM [00:10] <ubitux> which can be summarized with: [00:10] <JEEBsv> and thus it can have PCM, AC3, AAC or whatever [00:10] <ubitux> > SMV file, a wav file with video appended [00:12] <kierank> could be spdif [00:12] <kierank> has no guid [00:12] <cone-737> ffmpeg.git 03Hendrik Leppkes 07master:23bc1351ad7d: vdpau: restore compatibility with deprecated fields in AVVDPAUContext [00:21] <cone-737> ffmpeg.git 03Hendrik Leppkes 07release/2.1:70028e917ce8: vdpau: restore compatibility with deprecated fields in AVVDPAUContext [00:21] <Compn> j-b : because its better than totem/xine nonsense ? [00:21] <Compn> and its updated ? [00:22] <j-b> Compn: totem uses Gstreamer, so it gets installed. [00:22] <j-b> Compn: xine is dead [00:22] Action: smarter remembers xine [00:22] <smarter> not very fondly. [00:44] <cone-737> ffmpeg.git 03Carl Eugen Hoyos 07master:26fc6ffec45c: Cast signed libopenjpeg data values to unsigned before shifting. [00:44] <cone-737> ffmpeg.git 03Michael Niedermayer 07master:7cd46acdd5f1: Merge remote-tracking branch 'cehoyos/master' [02:24] <cone-737> ffmpeg.git 03Michael Niedermayer 07master:03fff09b3217: avcodec/wmalosslessdec: Pass on error code from decode_tilehdr() [02:24] <cone-737> ffmpeg.git 03Michael Niedermayer 07master:ae3856dcaf9c: avcodec/wmalosslessdec: deallocate uninitialized frame on decode_tilehdr() failure [02:24] <cone-737> ffmpeg.git 03Michael Niedermayer 07master:6b18a6839b43: avcodec/wmalosslessdec: shrink output on error so no uninitialized data is returned [02:36] <cone-737> ffmpeg.git 03Clément BSsch 07master:b84a7330af41: avformat/pjsdec: dont increase pointer when its already at the end in read_ts() [03:31] <cone-737> ffmpeg.git 03Martin Storsjö 07master:44a0a98f92c7: arm: Add an option for making sure NEON registers aren't clobbered [03:31] <cone-737> ffmpeg.git 03Michael Niedermayer 07master:9d5cc55f0fdf: Merge remote-tracking branch 'qatar/master' [03:35] <BBB> hm [03:35] <BBB> michaelni: ping [03:35] <michaelni> BBB, pong [03:36] <BBB> how do AVParsers signal errors? [03:36] <BBB> I'm currently returning AVERROR_INVALIDDATA but that infloops [03:36] <BBB> (parser.c comments seem to indicate negative return values aren't errors) [03:38] <BBB> just return size? [03:39] <michaelni> i think theres no way to return an error currently [03:39] <BBB> I think I fixed the last open vp9 bug in trac, feel free to review or wait for ubitux to do so [03:40] <michaelni> the idea probably was that parsers should split anything the best they can [03:40] <michaelni> but that misses malloc errors [03:42] <michaelni> so with current API thats returning size / 0 or something like that [03:46] <michaelni> an error_code field could be added to AVCodecParserContext if someone would want to export erors while keeping best effort spliting or a av_parser_parse3() ... [03:47] <BBB> nah returning size is ok [04:43] <cone-737> ffmpeg.git 03Michael Niedermayer 07master:e11983bda073: avcodec/alac: only set *got_frame_ptr when all channels have been decoded [12:19] <ubitux> BBB: isn't the valgrind stack thing in vp9 simply because you move the stack pointer assuming the data on it won't be change (and thus confuse valgrind alot©)? [12:19] <ubitux> would it be possible to use another reg as pointer to avoid the issue? [12:53] <BBB> if we confuse valgrind, we should fix valgrind [12:54] <BBB> my basic assumption has always been that if tools are broken, we should either fix them, or not use them [12:54] <BBB> (I've reported bugs like this in valgrind or asan [forgot which one] before, and they did get fixed [12:55] <Compn> might already be fixed in valgrind git? :P [12:55] <Compn> ehe [12:56] <ubitux> BBB: i'm doing some stuff on itxfm [12:56] <ubitux> i hope you aren't ;) [12:56] <ubitux> i saved one mov [12:56] <ubitux> and i might be able to save more [13:00] <wm4> save the movs [13:04] <BBB> I'm doing iadst16 [13:08] <ubitux> BBB: can you have a look to https://github.com/ubitux/FFmpeg/compare/vp9-itxfm ? [13:09] Action: ubitux goes back to lpf [13:27] <BBB> but this doesn't affect the binary right? [13:27] <BBB> x86inc.asm actually knows that a three-register instruction avx-syle instr a, b, c where a==b is the same as a two-register sse-style instruction instr a, c [13:28] <BBB> so I don't think this saves a mova [13:28] <BBB> I do like the changes, so please do commit [13:29] <BBB> oh I guess binary isn't the same because you deinterleaved the punpack/mulsub pairs [13:32] <ubitux> ah, x86inc deals with that case? too bad :p [13:32] <ubitux> indeed the binary changed because of the following change [13:32] <ubitux> do you think i should benchmark just in case? [13:34] <BBB> probably always a good idea, yes, but I think it's an ok change [13:35] <BBB> how's lpf going? is the 16x16b transpose done already? :-p [13:35] <ubitux> i'm on it :p [13:36] <BBB> \o\ /o/ \o/ [13:36] <BBB> (I'm pretty sure big transposes suck) [13:36] <ubitux> i'm trying to figure out what reg should be saved in a stashing area [13:36] <ubitux> dunno, currently it's like this: http://pastie.org/8623522 [13:37] <ubitux> (but doesn't work because of the saves) [13:39] <BBB> can you also review "vp9: fix mt-related hang a parser infinite loop."? [13:40] <BBB> ubitux: look at how transpose8x8w works on x86-32, and try to reproduce these save patterns [13:40] <BBB> (i.e. you'll need 2 stack locations and shift between the two) [13:41] <BBB> but I fully acknowledge it sucks :-p [13:41] <ubitux> the mt hang patch should be ok, but strangely after your previous patch i wasn't able to trigger that hang [13:41] <BBB> hm [13:41] <BBB> it still hung for me [13:41] <BBB> odd [13:41] <ubitux> probably 10 lucky hit in a row [13:41] <ubitux> :p [13:42] <ubitux> yes i'm trying to copy the save patterns [14:19] <BBB> error_res related, does anyone have opinions on adding a "corrupted" bit to AVFrame (or flag, or whatever) for these codecs not (yet) implementing ER, so that we can mark them as corrupt reference for future frames to resolve the difference in reconstruction of corrupt frames w/ and w/o frame-mt? [14:20] <BBB> the idea basically being that currently, most codecs will discard the reference as invalid and refuse to decode subsequent frames (or blocks) using that as a reference in single-thread mode, but will happily set progress to int-max in multithread mode (such as to not cause hangs), thus using a corrupt reference [14:20] <BBB> having a corrupt bit in the reference frame (possibly as part of ThreadFrame) would resolve this difference between frame-mt and no-mt behaviour for such codecs [14:54] <saste> #ffmpeg-meeting channel created [14:55] <BBB> maybe I''ll idle in there so I can peek in [14:55] <BBB> when does it start? [14:57] <BBB> ah, 11 est [15:25] <beastd> hi [15:31] <beastd> saste: Where is the meeting? [15:33] <ubitux> beastd: #ffmpeg-meeting [15:41] <kurosu_> ubitux: does having 50+ commits in ffmpeg qualify for joining? I don't plan to actually participate, just taking this opportunity to see what's happening [15:41] <ubitux> it's public [15:43] <kurosu_> ok [16:08] <saste> kurosu_, it's public, you can join if you like, and the log will be published on ffmpeg-devel [16:09] <Compn> everyone is of course welcome, especially with new ideas :) [16:09] <Compn> Daemon404 and wm4 usually join to troll as well [16:09] <Compn> constructive trolling that is [16:10] <wm4> in fact you are the troll [16:10] <Compn> shh dont tell everyone :P [16:17] <Cigaes> Hi. [16:18] <saste> Cigaes, hi [16:39] <nevcairiel> BBB: there already is a corrupt flag which is set for frames which are not yet fully re-constructed due to missing refs (in h264 at least), dont suppose you can use that? its primary meaning is to inform user code right now though.. [16:59] <saste> the ffmpeg IRC meeting is going to start soon, please join #ffmpeg-meeting if you want to participate [17:00] <j-b> saste: is it open to everyone? [17:00] <saste> j-b, yes [17:00] <saste> log is going to be published on ffmpeg-devel if you can't attend, you're welcome to join [17:36] <smarter> what's the purpose of writing "ALIGN 16" before a block of code in yasm? [17:36] <smarter> Like in http://git.videolan.org/?p=ffmpeg.git;a=blob;f=libavcodec/x86/h264_deblock.… [17:38] <nevcairiel> i suppose aligning code to start on a certain byte boundary is advantageous, don't ask me why :) [18:02] <kierank> smarter: it's semi cargo-cult iirc [18:02] <kierank> but basically based on benchmarking and see if it change [18:02] <kierank> changes* [18:05] <smarter> why would it change? :) [18:05] <smarter> better use of cachelines? [18:07] <kierank> instruction alignment [18:08] <smarter> is that something pengvado would know more about? :) [18:13] <smarter> oh cool, there's a section in http://www.agner.org/optimize/optimizing_assembly.pdf on code alignment [18:15] <kierank> are there any ffmpeg people at fosdem? [18:42] <BBB> nevcairiel: if we agree that's a proper use for it, sure [20:38] <cone-780> ffmpeg.git 03Michael Niedermayer 07master:d5af40071395: avcodec/alac: warn if not all channels where decoded [20:38] <cone-780> ffmpeg.git 03Michael Niedermayer 07master:0999f1613bc4: avcodec/hevc: use av_mallocz() for allocating tab_ipm [20:38] <cone-780> ffmpeg.git 03Luca Barbato 07master:5856bca360c5: hevc: Clip the pixel before shifting [21:17] <cone-780> ffmpeg.git 03Benjamin Larsson 07master:547f83453738: flv: Workaround for buggy Omnia A/XE encoder [21:17] <cone-780> ffmpeg.git 03Michael Niedermayer 07master:419787a2ffe3: Merge remote-tracking branch 'qatar/master' [21:18] <michaelni> j-b, "remote: mail: /tmp/mail.RsXXXX0vyWgz: No space left on device" on my last git push [21:19] <michaelni> i think the push itself was fine though [21:53] <cone-780> ffmpeg.git 03Michael Niedermayer 07master:859d74040e4b: avcodec/mjpegenc: pass chroma quantization matrix through as well, not just luma [21:53] <cone-780> ffmpeg.git 03Michael Niedermayer 07master:b8ff951ce03f: avcodec/mpegvideo_enc: init chroma matrix for mjpeg [21:53] <cone-780> ffmpeg.git 03Michael Niedermayer 07master:7dc0aba3fb9d: avcodec/mjpegenc: use a seperate chroma matrix when luma and chroma differ [22:26] <smarter> michaelni: 5856bca360c5 is not the proper way to fix that, I'll be sending a proper patch to libav-devel soon [22:33] <michaelni> smarter, can you also send it to ffmpeg-dev ? [22:33] <michaelni> after all the "non proper" one is already i ffmpeg [22:33] <michaelni> iN [22:34] <michaelni> BBB, ubitux should i apply " vp9: fix mt-related hang a parser infinite loop." or should i wait for some review ? [22:34] <ubitux> it's fine with me [22:34] <michaelni> ok [22:36] <cone-780> ffmpeg.git 03Ronald S. Bultje 07master:0065d2d520ca: vp9: fix mt-related hang a parser infinite loop. [22:39] <cbsrobot> what are the prores issues discussed in the meeting ? [22:41] <michaelni> cbsrobot,see http://support.apple.com/kb/HT5959?viewlocale=en_US [22:56] <kierank> michaelni: it's a FUD campaign against broadcasters [22:56] <kierank> and it's working because they are scared [22:56] <JEEB> yeah, clear FUD [22:57] <JEEB> and there's also no way they'd 'authorize' libavcodec [22:57] <kierank> I'd be curious to see if they reply [22:57] <BBB> their point is somewhat valid [22:57] <kierank> and if they attack ffmpeg I'll just say the reality of the situation is the decoder is used heavily and you have to offer a licensing programme [22:57] <BBB> prores is a 12bit format [22:58] <BBB> ffmpeg's c decodes 10bits [22:58] <BBB> ffmpeg's assembly is worse [22:58] <kierank> is prores natively 10-bit or is only one profile 10-bit? [22:58] <BBB> prores is 12bit [22:58] <BBB> prores is not 10bit [22:58] <kierank> BBB: I know for a fact the *biggest* media organisations in the world deliver using proresenc/dec [22:58] <kierank> FUD FUD FUD [22:59] <BBB> ? [22:59] <nevcairiel> last i heard prores 422 is 10-bit, only 4444 allows 12-bit [22:59] <JEEB> ^ [22:59] <kierank> BBB: broadcasters are running scared [22:59] <BBB> check our prores decoder [22:59] <kierank> because apple are saying lavc is unauthorised [22:59] <BBB> it decodes 12bit coefficients [22:59] <nevcairiel> (and yes, 4444 is not fully supported by avcodec) [22:59] <BBB> then scales them back 2bit [22:59] <BBB> then does a 10bit idct [23:00] <BBB> except the assembly, which is (in ffmpeg at least) 8bit [23:00] <BBB> I wrote a 10bit version which michael then changed back tot he 8bit version [23:00] <michaelni> huh [23:00] <BBB> I know the format surprisingly well for someone that knows nothing about video codec design or stuff like that [23:01] <kierank> 10:00 PM <"BBB> I know the format surprisingly well for someone that knows nothing about video codec design or stuff like that ??? [23:02] <nevcairiel> have a white paper; http://www.apple.com/final-cut-pro/docs/Apple_ProRes_White_Paper_December_2… ..it confirms that only 4444 is 12-bit, and 422 profiles are 10-bit [23:03] <michaelni> BBB, where is that code that drops 2 bits ? [23:03] <Timothy_Gu> Is Kostya still working on 4444? [23:05] <ubitux> Timothy_Gu: ask him (Keiler) [23:06] <Timothy_Gu> BTW I couldn't attend the meeting because I had a writing lesson and another lacrosse one back-to-back [23:23] <Compn> i know what apples problem with avcodec is [23:26] <Compn> its because apple doesnt have the rights to elvis! * Copyright (c) 2010-2011 Elvis Presley [23:26] <Compn> :) [00:00] --- Sun Jan 12 2014

1 0

ffmpeg.log.20140111
by burek 12 Jan '14

12 Jan '14

[00:10] <jnvsor> Can you give an input a name to map it with? [00:10] <jnvsor> Or are input streams always numbered? [00:13] <gusto> hey [00:13] <gusto> i am having trouble wanting to extract subtitles from a file where the sub stream does not appear instantly [00:15] <ubitux> jnvsor: you can map an input yes, but you're most likely interested in the output of the filtergraph [00:15] <gusto> http://pastebin.com/8keh27fy here trying to do the extract [00:16] <ubitux> your input has no subtitles [00:16] <ubitux> and even if it had it would probably be dvd sub, which are bitmaps [00:16] <ubitux> and you wouldn't be able to extract them with ffmpeg [00:16] <ubitux> at best remux them into something like mkv [00:17] <gusto> here the file again http://pastebin.com/eze3a4ZH [00:17] <jnvsor> ubitux: No I've gotten past that. My issue now is how to map streams when I'm not sure which number of input files come before the ones I'm looking for [00:18] <gusto> here doing remuxing into MKV http://pastebin.com/z3c090hh [00:19] <ubitux> you probably need to use some kind of ifo index [00:19] <ubitux> ffmpeg doesn't support that unfortunately, patch welcome [00:19] <jnvsor> Oh dear. Well my work just got a lot more fun :P [00:19] <ubitux> jnvsor: i don't understand at all [00:20] <jnvsor> I'm writing a script for screencapping [00:20] <jnvsor> It already supports multiple video streams, scaling, multiple audio streams etc [00:21] <jnvsor> Now I'm trying to add filter_complex support for eg overlaying webcam onto video track 1 [00:21] <jnvsor> Problem is mapping the audio streams when you don't know how many video streams came before it [00:22] <ubitux> why? [00:23] <jnvsor> Well as it is I just have the user input the amount of video streams they have, loop over it and add `-map $i` for every stream [00:23] <jnvsor> But if I'm mapping by name instead then I don't know when the audio streams start [00:26] <ubitux> -map 0:v -map 0:a ? [00:26] <jnvsor> That only maps the first stream. Ideally I'd have -map*:a [00:27] <jnvsor> first file* [00:27] <ubitux> maybe there is some kind of joker [00:27] <ubitux> like -map v, or -map :v [00:28] <gusto> ok [00:28] <gusto> i seem to have found a way how to extract it from the original DVD [00:28] <gusto> aaa, no [00:29] <gusto> does not work :-( [00:29] <gusto> w8, it works [00:30] <gusto> http://pastebin.com/Bdw5TnuB [00:31] <gusto> this way i should get at least the DVB SUB out in a MKV and will try to do something with it then [00:34] <gusto> still no luck http://pastebin.com/egBHugkc [00:37] <jnvsor> gusto: What's stopping you from just downloading fresh subs and using them? [00:37] <gusto> i do not know [00:38] <gusto> maybe its just curiosity if i d get it down from there [00:38] <gusto> i ll try VLC once again, if it can put down those subs from the DVD, thats also worth a try [00:38] <gusto> but apart from that, when it does not work, i ll stop trying [01:53] <jnvsor> From the looks of it the copy and null filters do the same thing, but is there a performance difference between them? [02:55] <reywood> hello, can anybody help with a subtitle question? [02:56] <sacarasc> Only if you ask it. [02:56] <reywood> heh [02:56] <reywood> I have a subtitle track that should be displayed by default, like when one character in a movie speaks a different language, subtitles should always be shown for them [02:57] <reywood> can't figure out how to specify that in the ffmpeg command [02:57] <reywood> the command I'm currently using looks like this: [02:57] <reywood> ffmpeg -i mymovie.mp4 -f srt -i mymovie.srt -c:v copy -c:a copy -c:s mov_text -metadata:s:s:0 language=eng mymovie-sub.mp4 [03:10] <relaxed> reywood: I'm curious why you want subs in mp4. [03:11] <reywood> relaxed: I prefer to have it all in one file. Is there a reason not to? [03:12] <relaxed> well, I think the matroska container would be a better choice. [03:13] <reywood> ideally, I'd be able to play the video on an AppleTV which doesn't support mkv [03:14] <relaxed> ugh, really? [03:14] <reywood> if it's not possible to do what I want with mp4, that's fine, but I'd like to know [03:14] <relaxed> let me see the console output from you command. [03:14] <relaxed> pastebin.com [03:17] <reywood> http://pastebin.com/s4Jxv6k2 [03:18] <reywood> if I was to use mkv, what would be the option? [03:18] <hi117> http://j00ru.vexillium.org/?p=2211 [03:18] <hi117> keep up the good work! [03:19] <relaxed> reywood: with matroska you could copy the srt along with the audio/video [03:20] <relaxed> reywood: If all else fails you can hardsub them: https://trac.ffmpeg.org/wiki/How%20to%20burn%20subtitles%20into%20the%20vid… [03:20] <reywood> yeah, I was hoping there was a way without hardsubbing [03:21] <relaxed> are you sure appletv supports loading subs in mp4? [03:21] <relaxed> you're command didn't have any errors. [03:21] <relaxed> your* [03:21] <reywood> relaxed: the subs are in the final mp4, they just aren't on by default, so If I play the movie, I have to manually select them [03:22] <relaxed> hmm, maybe there's metadata to enable them by default [03:22] <reywood> I haven't tried it on the appletv, I've played it in VLC on my comp [03:23] <reywood> I tried setting "default=1" and "forced=1" in the metadata, no difference [03:34] <relaxed> Maybe you can configure appletv to play subs by default. Other than that I'm out of ideas. [03:35] <reywood> well thanks for your help anyway [03:35] <reywood> much appreciated [06:26] <rcombs> anyone know if libavcodec does anything special when converting an SRT stream in an RTL script to ASS? [06:26] <rcombs> (i.e. does it reverse character order, or set the font encoding to Arabic/Hebrew as a hint for libass?) [09:50] <Zeranoe> My input audio and video streams have a different number of frames, how could I output with the same? [12:53] <jnvsor> If I overlay webcam into x11grab my framerate drops to ~20fps, regardless of how many threads I use. I presume something about my webcam input is bottlenecking it. Any ideas? [14:20] <albator> hi all [14:20] <albator> anybody has a little idea how to use named pipe to combine 2 files in one output? [14:58] <jnvsor> You mean like an overlay? [14:59] <jnvsor> Or are you looking for multiple tracks? [15:59] <jnvsor> Anyone here intested in testing an FFmpeg based screen capture script? :D [16:14] <Cigaes> Hi. [16:49] <Bombo> i thought i understood the commandline, but now i get an error: [16:49] <Bombo> /opt/bin/ffmpeg -i bar.mkv -c:v copy -c:a copy -c:a copy -i foo.mp4 -c:s copy -map 0:0 -map 0:1 -map 0:2 -map 1:3 baz.mkv [16:51] <Bombo> i want to take video 0:0, audio 0:1, audio 0:2 from bar.mkv, then sub 1:3 from foo.mp4, mux it into baz.mkv (copy) [16:51] <Bombo> error is: Unknown decoder 'copy' [17:00] <albator> i mean like instead of having a disconnection after sending every single video to remote rtmp, using pipe to create a single input for ffmpeg [17:01] <albator> but i don't know the proper syntax [17:02] <jnvsor> Bombo: You have the codec parts next to the inputs so it's using them to *de*code the streams, try this: [17:03] <jnvsor> Bombo: http://pastebin.com/RWaBhM3Q [17:07] <Bombo> jnvsor: oh! right, thx ;) [17:07] <Bombo> (btw no need for pastebin for just one line) [17:15] <Bombo> hmmm is it possible to delay subs? they are like 3s late [17:23] <Bombo> or i could just delay the rest ;) [17:31] <jnvsor> Hmm "Data is not aligned. This can lead to a speed loss" - any way to tweak my filters to get rid of this error [17:36] <Bombo> jnvsor: good question, i got these too [17:38] <jnvsor> It only happens when the input is uneven width/height. I get that it should be even, that's why I scale it (To stop the encoder freaking out) but I'm wondering if there's some obscure setting that will pad the data in order to align it in some way [17:41] <Bombo> not sure when i get it, i've just seen it often now, i do .mpg to .mp4 with scale=640:something [17:42] <Bombo> when the height is odd it complains so i add or remove -1 [17:43] <jnvsor> Hehe, I have a trick for that: scale=w=trunc(in_w/2)*2:h=trunc(in_h/2)*2 [17:43] <jnvsor> add that after your first scale and it rounds it off for you automatically [17:46] <Bombo> i do scale=640:in_h*out_w/sar/in_w [17:47] <Bombo> so scale=640:in_h*out_w/sar/in_w,scale=w=trunc(in_w/2)*2:h=trunc(in_h/2)*2 [17:47] <Bombo> hmmm try it next time, thx ;) [17:48] <Bombo> +i'll [17:50] <jnvsor> It's designed for something like: scale=640:-1,scale=w=trunc(in_w/2)*2:h=trunc(in_h/2)*2 [18:06] <lkiesow> albator: /quit [18:06] <albator> ? [18:06] <lkiesow> sorry, ignore that ^^ [18:07] <albator> i better ignore yes lol [18:16] <jnvsor> I'm trying to overlay my webcam onto an x11grab - that's all fine and dandy but my webcam's framerate is... sporadic... 30fps one second and 10fps the next. It's dragging my output framerate down even though x11grab is dropping frames to do it. [19:07] <albator> ffmpeg -i test1.avi -i test2.avi -vcodec copy -acodec copy \ [19:07] <albator> -vcodec copy -acodec copy test12.avi [19:07] <albator> is this supposed to encode test1 and test2 in a single file? [19:07] <albator> because I am really trying with multiples input and I can't get past file1 , it doesn't do anything on following inputs [19:25] <klaxa> albator: see: https://trac.ffmpeg.org/wiki/How%20to%20concatenate%20%28join,%20merge%29%2… [19:58] <albator> yes I saw that, it isn't working (the filelist.txt for instance) [00:00] --- Sun Jan 12 2014

1 0