From burek021 at gmail.com Sat Nov 1 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Sat, 1 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141031 Message-ID: <20141101010503.191B818A024F@apolo.teamnet.rs> [00:06] ffmpeg.git 03Michael Niedermayer 07master:25a360286db8: avcodec/dvbsubdec: merge declaration and init [00:06] ffmpeg.git 03Michael Niedermayer 07master:adfc3b81b1ca: avcodec/dvbsubdec: use av_freep() to avoid leaving stale pointers [00:06] ffmpeg.git 03Michael Niedermayer 07master:a1cb8779e2c7: avcodec/crystalhd: use av_freep() to avoid stale pointers [00:06] ffmpeg.git 03Michael Niedermayer 07master:2a26b22a1739: avcodec/cngdec: Use av_freep() to avoid leaving stale pointers in memory [00:06] ffmpeg.git 03Michael Niedermayer 07master:fc8d59fa6f4c: avcodec/h264_parser: Use av_freep() to avoid leaving stale pointers [01:39] thanks for the ops [01:44] ffmpeg.git 03Michael Niedermayer 07master:f76cd09d466e: doc: Add documentation for the whitelist options [02:05] ffmpeg.git 03Martin Storsj? 07master:cf589faa5b7a: movenc: Add a flag for using default-base-is-moof in tfhd atoms [02:05] ffmpeg.git 03Michael Niedermayer 07master:8065a0cdbed7: Merge commit 'cf589faa5b7aed3bb38e08dcd00bd951e69686d1' [02:31] ffmpeg.git 03Martin Storsj? 07master:00c67fe1d0bc: movenc: Write a 0 duration in mdhd and tkhd for an empty initial moov [02:31] ffmpeg.git 03Michael Niedermayer 07master:61f1c96ef1be: Merge commit '00c67fe1d0bc7c2ce49daac9c80ea39d5a663b73' [02:32] https://ffmpeg.org/pipermail/ffmpeg-devel/2014-October/164753.html <-- is this a good idea, or am I on crack [02:38] rcombs : you're from usa i am going to assume. i wonder where the 'or am i on crack' phrase came from, or when the last time you've even heard the term 'crack head' . just curious because its been a while since i've heard it on tv, was it a big joke in the 90s? on in living color or something? [02:39] just seems like we dont hear about crack heads anymore [02:41] hmm, I don't actually hear "crackhead" much [02:41] but yes, I'm from the US [02:41] I'm not sure where I picked up the phrase [02:49] Compn: fresh on our minds in Canada http://www.theguardian.com/world/2013/nov/05/toronto-mayor-rob-ford-admits-crack-use [02:50] ffmpeg.git 03Martin Storsj? 07master:c55d1d382cd4: movenc: Don't write any tfdt atom for ismv files [02:50] ffmpeg.git 03Michael Niedermayer 07master:77eff7a58a43: Merge commit 'c55d1d382cd41345a79782ace41f9b43f45dca9a' [02:58] ffmpeg.git 03Martin Storsj? 07master:aae6b3b918b4: movenc: Don't write any iso brands in ismv files [02:58] ffmpeg.git 03Michael Niedermayer 07master:dd2f86864491: Merge commit 'aae6b3b918b4133b8cc2d1631196c1d406d0351a' [02:59] rcombs : i say "crack head" all the time to my friends. just a weird phrase that isnt in pop culture anymore [03:00] kevmitch : lol , do you think his brother is going to be toronto's mayor ? [09:57] lll [09:58] test [10:06] 'morning [10:45] kk [12:09] ffmpeg.git 03Michael Niedermayer 07master:d15a94ba202d: configure: fix escaping in xcb cflags / libs [13:00] michaelni ramiro Tried submitting tests to the server and command spat out "Broken pipe". After I ran the tests/fate.sh command again on the fate_config.sh script and it spat http://paste.kde.org/p4gnak8af out [13:13] amalia the script detects probably that this checkout has already been tested and submission attempted and skips doing that again until the next version [13:15] I changed the fate_config.sh build-only variable which held "yes" to hold nothing. Thought that made the script different from the previous one [13:15] ramiro told me to do another test without the build-only option [13:33] amalia, didnt ramiro ask that 2 days ago, you needed 2 days to remove "yes" from the config ? ;) [13:33] and its not the script that "differ" or doesnt its the git checkout revission [13:34] I was hustling for ubuntu to install on virtual box [13:34] ahh, ok [13:35] did you succeed installing it ? [13:35] Yeah I did [13:36] Just Internet disturbing dure to rains and thunderstorms here [13:38] and both host and ubuntu guest are stuck with the same error ? [13:38] or one works ? [13:39] host gets stuck. Setting fate tests on guest ubuntu already [13:41] ffmpeg.git 03Michael Niedermayer 07master:0c6c5d7bcb1b: avcodec/mjpegenc: use av_freep(), avoid leaving stale pointers in memory [13:42] amalia, try the host again, theres a new git revission at HEAD. also it should be possible to just delete the file in which the revission is stored when you want to rerun it on the same revission but the server might then still reject it [13:43] I am updating ffmpeg right now - doing a git pull michaelni [14:48] amalia: remember to reply by e-mail since I'm not always on irc. [15:19] heh google docs [15:28] kierank, i loathe gdocs [15:28] olol its a patch [15:56] oh, it's from my opw applicant... [16:01] wait what, she finished the actual task before the qualiofication task? [16:02] no, she didn't finish any tasks. she just took my old code, mangled it, and sent it in for review. [16:16] par for the course [23:15] ffmpeg.git 03Michael Niedermayer 07master:d5633dfc288b: doc: document -dump_separator [23:42] ffmpeg.git 03Michael Niedermayer 07master:7f7facdedaf2: doc/formats: document libavformats bitexact flag [23:47] I'm compling ffmpeg using --enable-static, but when I try the executable on a pc that's not my development box I get that my system is missing libiconv-2.dll [23:47] I'm using mingw to compile [23:48] my entire configure is: [23:48] ./configure --extra-cflags="-I/usr/local/include" --extra-ldflags="-L/usr/local/lib" --enable-static --disable-shared --disable-doc --disable-ffplay --disable-ffserver --enable-memalign-hack --enable-gpl --enable-nonfree --arch=x86 --enable-w32threads --enable-libx264 --enable-libx265 --enable-libfdk-aac --enable-libass --enable-libfreetype --enable-fontconfig [23:49] --enable-static only refers to the libs FFmpeg libs that will be built [23:49] If you want gcc to prefer static libs when linking you need to use -static [23:50] Where do I put that? In --extra-ldflags? [23:50] Yes, I think [23:51] Okay I will give that a try [00:00] --- Sat Nov 1 2014 From burek021 at gmail.com Sat Nov 1 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Sat, 1 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141031 Message-ID: <20141101010502.10DAC18A024E@apolo.teamnet.rs> [00:00] I'm trying to cut out parts of a video with -ss and -t, but FFMPEG isn't honoring the -t and encodes past it. [00:42] protip: run your script with bash -x :) [01:11] Hi, is there a way to rotate/transpose a AVFrame programmatically using libavfilter? [01:13] I'm gonna go out on a limb and say "yes". https://www.ffmpeg.org/doxygen/2.4/vf__rotate_8c.html http://www.ffmpeg.org/doxygen/0.9/vf__transpose_8c-source.html [01:51] how do I play playlists (streaming, m3u) with ffplay? [02:03] Hello [02:03] I have some problems with ffserver [02:04] i think there is an bug in the new versions of ffmpeg (ffserver) [02:40] go on [05:17] Can someone tell me what this means, and how i'm able to fix this issue? "[abuffer @ 042989e0] Value inf for parameter 'time_base' out of range [0 - 2.14748e+009][abuffer @ 042989e0] Unable to parse option value "(null)" as sample format" [15:44] say, is there some kinda trick to converting something that's riff (little-endian) xvid/mp3 .avi file? [15:44] as in, something that doesn't mess with the bitrate/format too much [15:45] i've tried it several different ways and when i play it on the tv it's kind of studdering at parts [15:50] so you're saying you have mp3 inside your riff file and you want to convert it into an avi? [15:52] riff as a format is so vague that it's hard to say what's inside without having a look. [15:52] (anyway, I know nothing knotwurk, so please to wait for people with more bits in the brein.) [15:53] this is what the file command shows it as: "RIFF (little-endian) data, AVI, 720 x 536, 25.00 fps, video: XviD, audio: MPEG-1 Layer 3 (stereo, 48000 Hz)" [15:53] for some reason the tv thing doesn't play it, and it has something to do with the riff/little-endian thing [15:53] and what does your tv support? [15:53] oh a whole bunch of formats, usually there's no problem w/ avi/mpeg [15:54] maybe the bitrate is too high for your tv. [15:55] yeah maybe i should go back to the manual [15:55] or if it doesn't like the xvid then I think you're going to need to re-encode. [15:56] yeah that's what i've been doing [15:56] it might be a good idea to just reencode part of it so i can experiment [15:56] nod [15:56] I gotta go. happy debugging! [15:56] cya [16:48] knotwurk: what does ffprobe say about the file? [16:53] c_14: http://dpaste.com/278E0G3 [16:59] You could try converting to non-xvid mpeg4... [17:00] Not sure why the TV would stutter on that though. [17:14] hm, mpeg2video looks better for some reason [17:29] hi [17:41] hello [17:52] When providing ffmpeg with a URL where the content is, is it capable of taking advantage of HTTP ranges to seek the file and locate the moov atom? [22:35] Hi, anyone have any experiance with dvdauthor? Why does it not add chapters where I ask? I encoded my video with ffmpeg, and then dvdauthor tried to read it and adds chapters at 0, about 60s, and then repeats the 60s time for all chapters (which makes for a broken dvd) [00:00] --- Sat Nov 1 2014 From burek021 at gmail.com Sun Nov 2 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Sun, 2 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141101 Message-ID: <20141102010501.A90942AD60CA@apolo.teamnet.rs> [00:23] alright, I found my problem! and it's ffmpegs fault... I think, my output mpeg only had VOBUs for the first 60s or so... why? [00:24] why is that, and can I do DVD muxing with ffmpeg? [00:24] mplex worked fine [01:14] Hello everyone, I'm hoping to find someone to point me in the right direction about ffmpeg pipes? [01:15] ||||||||| [01:15] Pipes. [01:15] Pipes everywhere. [01:15] What exactly are you trying to do? [01:16] So...in short...pipe all thumbnails generated from a video over to another process [01:16] So that the secondary process can crunch the images real time. [01:16] just use pipe:1 as the output file [01:16] so...the command would be something like this...? [01:16] Not sure if that works for images. [01:16] hmm [01:17] let me see [01:17] ffmpeg -i input.avi image2 -vf fps=fps=1/60 pipe:1 "other command?" [01:18] -f image2pipe [01:18] And it'l bee pipe:1 | other command [01:18] eh -e [01:19] the secondary process is actually the facedetect module available here http://www.thregr.org/~wavexx/software/facedetect/ [01:19] what's the -e? Could you give me a complete example? [01:19] That was just me correcting my mispelled bee [01:21] so for every image that ffmpeg dumps...it would send that .png or jpeg over to stdout [01:21] ? [01:21] it should [01:21] ok [01:21] Not entirely sure what output it defaults to though... [01:21] so for that example I would leave off the image%3d.jpg? [01:21] yes [01:22] Or else you'll be writing to the filesystem again. [01:22] ok [01:22] one more question [01:22] would ffmpeg run faster from the pre-compiled package on ubuntu, or compile from source? [01:23] The compiled from source package will presumably be newer and have more fixes/improvements than the pre-compiled one. [01:23] Whether or not that makes it faster depends. [01:24] if I put that package in /usr/local/bin can I run it from anywhere? [01:25] If /usr/local/bin is in your PATH, yes. [01:27] sweet...lemme give this a whirl, lol. Thanks! [01:35] so if the second command would normally go like this...'facedetect -q image.jpg' ...how do I tell it that the info is coming from stdin? [01:43] c_14? [01:44] Usually with - [01:44] facedetect -q - [01:47] lemme see here [01:47] so this is what I have thus far...and it isn't working quit yet [01:48] ./ffmpeg -i input.avi -f image2 -vf fps=fps=1/600 pipe:1 [01:48] oops [01:48] hit enter too soon [01:49] ./ffmpeg -i input.avi -f image2 -vf fps=fps=1/600 pipe:1 | facedetect -q - | echo $? >> testlog.txt [01:50] It writes the exit code one time in the log file [01:50] And while it is running it states ignoring invalid SAR: 0/0:00.00 bitrate=N/A [01:53] I think you might need image2pipe [01:53] I tried that in place of the image2 [01:53] -f image2pipe [01:53] same result [01:54] does facedetect actually support "-" as meaning read from stdin? [01:56] I don't know that for sure [01:56] it would normally take a path to an image [01:57] any idea of another command I could test as the secondary process? [01:57] you could try facedetect -q <(./ffmpeg -i input.avi -f image2 -vf fps=fps=1/600 pipe:1) [01:57] that might be a bash thing [01:57] it might not even work with something other than text [01:58] if this is it http://www.thregr.org/~wavexx/software/facedetect/ it doesn't even look like it can handle more than one image at a time [01:58] that is it [01:59] Could ffmpeg be slowed down to handle it or something? [01:59] Drop frames on purpose I guess? [01:59] That site shows some bash script to batch process, but that's about it [02:01] Or maybe another way to fork a process that crunches the images as ffmpeg dumps them I guess? [02:01] I'm not that great with bash yet. [02:01] More into python...but weak at that one too, lol. [02:03] if you could get ffmpeg -i "input.mov" -an -f image2 "output_%05d.jpg" to output the name of each image after it's written [02:04] hmmmm...maybe echo out to a file or something? [02:04] you could pipe that into "| while read file;do [02:04] facedetect $file [02:04] done [02:05] can you be specific with the syntax? I'm kinda rough on bash [02:05] my apologies [02:05] I do appreciate the help [02:05] irc's are amazing [02:05] ./ffmpeg -i input.avi -f image2 -vf fps=fps=1/600 "output_%05d.jpg" | while read file;do [02:05] facedetect $file [02:05] done [02:06] but that probably won't work because I don't think ffmpeg does output just the image file name to stdout [02:06] you need to get it to do that [02:07] this seems like something that would be kind of common...that image2pipe call makes me think there has to be some way to write the image and pass it to a process? [02:07] I could settle for running another command (facedetect) and have it cleaning up as ffmpeg dumps images [02:08] two processes I guess. But I'm not sure that ffmpeg would need to fork or anything...just not sure how to launch two calls at once or what have you.. [02:08] yes, but that process has to accept data from stdin, but it looks like facedetect only reads single files [02:09] so could facedetect crunch on a ls *.jpeg at the same time or something? [02:09] It reads a bit slower than ffmpeg dumps [02:09] from what I've seen at least [02:11] not sure about the performance vis a vis ffmpeg+facedetect, but you might get a cleaner implementation using opencv, which can both decode video to images and perform face detections on individual frames. python bindings make it relatively easy [02:12] really? I looked at that library rather quickly and can admit it was huge...and facedetect looked so simple, :-) [02:12] taking the easy way out doesn't always pay... [02:13] yeah, take a look here [02:13] http://docs.opencv.org/trunk/doc/py_tutorials/py_gui/py_video_display/py_video_display.html [02:14] you just need to put a face detect on each image in the loo [02:15] just give a filename to VideoCapture rather than a number [02:15] wow...that doesn't seem bad at all...awesome...as always...thank you very much! [02:17] so if I want to make ffmpeg widely available to the opencv library [02:17] I have downloaded a precompiled version and have been doing the ./ffmpeg from within that downloaded folder [02:18] do I copy the folder or just the contents over into /usr/local/bin so that other processes that depend on it can use it? Im on ubuntu [02:18] and Ubuntu's package of ffmpeg doesn't have the necessary filters to do the image dumping at predefined intervals [02:22] if you install ffmpeg into an isolated prefix with configure --prefix="$BUILD"/build_libs [02:23] then configure opencv with PKG_CONFIG_PATH="$BUILD"/build_libs/lib/pkgconfig ./configure [02:23] or something like that [02:23] k [02:23] thanks again [02:23] or you could just make install ;) [02:23] and opencv would presumably just find it [02:23] i had repeated failures building ffmpeg from source [02:23] would make it all the way to the end and it would fail on the part where ffmpeg wraps it altogether [02:23] no clue wh [02:24] why [02:24] oh, so you have a partially completed build ? [02:25] if so, opencv will have a hard time using that [02:25] in any case, the above example should work fine with both opencv and libav inthe ubuntu repositories [04:38] Hrm, why is when I add -target ntsc-dvd when encoding a video it wont work with -ss or -t set as well, removing -ss/-t or -target makes it work [09:52] question - I am looking around the ffmpeg website for list of REQUIRED libraries for compile. (Something in a recent Suse update broke my ffmpeg adn I'm trying to find out what so I can roll it back) Is there a place on the site that provides sucha list? [09:53] cjasonb: broke it how? can you pastebin the error? [09:58] no error in the code output. FFmpeg was being used with the following command ffmpeg -f alsa -i hw:1 -f v4l2 -i /dev/video0 (etc. etc) to capture from a pinnacle dazzle usb device. Was working flawless. Now getting the video as green video, with very chopy sound. Not sure if issue is v4l2 or some underlying render lib. Hoping to find a list so I can compare updated libs with required libs [09:59] could also have something to do with data strem from the usb port I suppose. [10:00] can you capture with something else? Have you read https://trac.ffmpeg.org/wiki/Capture/Webcam ? [10:01] have not read, will chaeck now. Unfortunately dazzle device is all I own. Project is converting 100+ VHS tapes for my local church to DVD. [10:02] oh, and I am the only person in the world who does not own a web came, so loop back isnt going to work atm [10:04] should point out I had to revert to a 2.3.3 ffmpeg to even get it to work in the first place. Right now I am looking for more of a "where is the error" so I know what route to take in seeking a fix, than an actual solution (unless someone knows one) [10:05] what happened when you tired 2.4.2? [10:07] maybe suse updated your kernel and the newer driver is causing the issue. If you can try booting into an older kernel if it's still around. [10:07] hmmmmm [10:07] kernel worth a look [10:07] hadnt thought of that [10:09] the thing that realy bugs me is this. I installed suse 13.1 (I know 13.2 comming out in days), delayed it's post install update, got ffmpeg doing what I wanted, and then updated, now its broke again. Logging off to reboot an older kernel. BBL if still issues [10:27] well, kernel does not appear to be the issue. Previous motv output (motiff based tv front end) was flawless, now it is black screen and realy scratchy audio. FFMPEG now locks up until I press ctrl-c right after identifying Input #0 (the audio portion). Video is locking. Jumping to pastebin to set up a link [10:31] pastebin output http://pastebin.com/BJ81qkaf [10:32] the parameter error I am unsure of. That EXACT line worked previously (copy pasted it to chell) [10:32] ^^shell [10:34] if i am guessing right, the issue isnt ffmpeg, but a sys lib or a dependancy of v4l2 [10:38] cjasonb: pastebin the output of v4l2-ctl -L [10:38] brb [10:43] sorry for delay, real life jumped in and bit me http://pastebin.com/kUwP8QJu [10:44] did you try viewing the video with anything else, like "mpv tv://" ? [10:45] only motv installed and that was zilch. trying to grab a new viwer...brb [10:46] sounds like this isn't an ffmpeg issue [10:47] im stck behind dsl...these others a re big and will take a while. [10:47] I tend to agree with you [10:47] was simply asking here first since I figured faster answer here [10:48] I would google for your capture device + linux and sort by most recent. [10:48] been trying that for a few days now (more) [10:49] got rid of ubuntu for suse, and being unfamiliar have done a lot of googling. The device is trash for linux support unfortuantely. BUt it is all the church had, so Im stuck with it [13:12] Hi everyone. Im using ffmpeg as my encoder. this is the command I use: [13:12] (well, avconv to be precise) [13:12] avconv -y -f alsa -i hw:2,0 -f video4linux2 -i /dev/video1 -vcodec libx264 -b:v 200k -b:a 50k -strict -2 -acodec aac -f flv rtmp://79.175.176.66:1935/live/myStream [13:13] my problem is that video and audio are out of sync. audio is ~3 seconds delayed [14:28] can ffmpeg do 1 operation on a bunch of files? [14:29] not by itself, write a script [14:58] better yet use xargs [14:59] ls lots*of*files | xargs -i ffmpg -i '{}' [16:10] anyone by accident happens to have ffmpeg built for raspbian? [16:27] kevmitch: Would that run a separate ffmpeg for each file? Because otherwise it's not going to work. [20:45] What is BFF and TFF in [Parsed_idet_0 @ 0xb0eb780] Single frame detection: TFF:0 BFF:0 Progressive:49 Undetermined:0 [20:46] can any one link me documentation, I searched [20:48] Bottom Field First vs Top Field First [20:48] c_14: thanks [00:00] --- Sun Nov 2 2014 From burek021 at gmail.com Sun Nov 2 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Sun, 2 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141101 Message-ID: <20141102010502.B11B12AD60CB@apolo.teamnet.rs> [00:03] So when I do that it says it can't find my libs [00:04] Was I supposed to static compile the libs? [01:07] ffmpeg.git 03Lukasz Marek 07master:fe72622819d3: lavd/alsa: implement get_device_list callbacks [01:07] ffmpeg.git 03Lukasz Marek 07master:ed1f8915daf6: ffserver_config: postpone codec context creation [02:12] ffmpeg.git 03Lukasz Marek 07master:d2d97b34a0a8: ffserver_config: fix compilation warning [02:26] Okay so this is what I got now [02:26] ./configure --target-os=mingw32 --extra-cflags="-I/usr/local/include/fontconfig -static" --extra-cflags="-I/usr/local/include/fdk-aac -static" --extra-cflags="-I/usr/local/include/ass -static" --extra-cflags="-I/usr/local/include -static" --extra-ldflags="-L/usr/local/lib/pkgbuild -static" --extra-ldflags="-L/usr/local/lib -static" --extra-libs=-lstdc++ --enable-static --disable-shared [02:26] --disable-ffplay --disable-ffserver --enable-memalign-hack --enable-gpl --enable-nonfree --arch=x86 --enable-runtime-cpudetect --enable-w32threads --enable-libx264 --enable-libx265 --enable-libfdk-aac --enable-fontconfig [02:26] Finally builds, but now it's complaining that it needs libx264-142.dll [02:26] What am I missing to make this thing completely static? [02:27] probably a static libx264 to link against [02:27] Shouldn't pgkbuild handle it regardless if it was built static or not? [02:28] I mean I have libx264.a in my /lib [02:28] pkgbuild? [02:28] er [02:28] pgkonfig [02:28] pgkconfig* [02:29] also, you don't have -L/lib in there and I'm not sure if that's in by default on mingw [02:29] in mingw everything goes to /usr/local/lib [02:29] why do you need it completely static anyway? [02:30] Mostly wanting to try it out and I'd like to be able to carry the exe with me on a flash drive without everything else. [02:31] hmm [02:31] I goofed a bit [02:31] use `make V=1` to figure out what the actual linker command for ffmpeg_g.exe is [02:31] --extra-ldflags="-L/usr/local/lib/pkgbuild -static" should be --extra-ldflags="-L/usr/local/lib/pkgconfig -static" [02:32] Okay [02:32] uh, why are you -L-ing your pkgconfig dir? [02:32] that's generally not how that works [02:33] idk, found it online somewhere about fontconfig. It fixed my issue [02:33] also, do multiple instances of --extra-*flags all apply with ffmpeg's configure? (I'm not sure) [02:34] I really don't know either. I can try to get rid of them to see if I have issues [02:34] looking at the code; I _think_ that should work [02:35] either way, see exactly what gets passed to the linker [02:35] Okay. [02:38] Hmm, I feel like I should have stdout this to a file [02:45] Why do you have so many --extra-cflags options? I don't think they work like that at all. [02:45] J_Darnley: apparently each one appends to the variable, so they _should_ work (but my instinct was also "wat") [02:45] Oh they do? [02:46] yeah, unless I'm _really_ bad at reading shell [02:46] and I'm pretty sure I'm not [02:48] Well it would certainly be clearer to merge them all into one. [02:48] yup [02:48] and -L-ing a pkgconfig dir smells nasty [02:57] Hi. Sorry for my shitty hacks, make v=1 is done. What am I looking for? [03:01] The error? [03:01] There is no error during make [03:02] Then what's the problem? That you are linking to some shared libs? [03:02] It's just that it's not staticly building the external libs into the exe [03:02] Yah [03:02] Did you install a static version of each lib you want? [03:03] I didn't as far as I know [03:03] I'm new to this [03:04] I was under the impression that pkgconfig would handle it [03:05] No. That just handles dependencies. [03:05] So I need to rebuild all of my dependdencies? [03:06] Maybe. For some, like libx264, static should be the default. [03:07] If it's static I should get a .a file correct? [03:08] only a .a file, yes [03:09] I think all of my stuff was just giving me .a and .pc [03:09] So I think they're all static [03:11] So the make install is giving me a whole bunch of dlls [03:11] I don't even know. I have to go, but I'll try later. [03:11] Thanks [10:39] was thinking of adding telecine detection to vf_idet [10:41] initial tests looks like finding repeated fields is relatively reliable [10:43] does adding a tally of frames with repeated fields sound sane? [11:37] ffmpeg.git 03Michael Niedermayer 07master:cc769931ab0c: avcodec/parser: use av_freep() to avoid leaving stale pointers in memory [11:37] ffmpeg.git 03Michael Niedermayer 07master:842745fe1705: avcodec/pthread_frame: Simplify code by using av_reallocp_array() [11:37] ffmpeg.git 03Michael Niedermayer 07master:e5054c8eed33: avcodec/pthread_slice: Use av_freep() to avoid leaving stale pointers in memory [13:21] ffmpeg.git 03Anton Khirnov 07release/2.4:de31f857077a: hevc_mvs: initialize the temporal MV in case of missing ref [13:21] ffmpeg.git 03Anton Khirnov 07release/2.4:0b41eeac45fb: hevc_mvs: make sure to always initialize the temporal MV fully [13:21] ffmpeg.git 03Michael Niedermayer 07release/2.4:13ecdb06f850: Merge commit 'de31f857077a52714f3a2f2e92ac037d42d37769' into release/2.4 [13:21] ffmpeg.git 03Michael Niedermayer 07release/2.4:0ddcee172ec8: Merge commit '0b41eeac45fb7f7ad6d3f4fc846b00d108824b0b' into release/2.4 [13:28] ffmpeg.git 03Vittorio Giovara 07release/2.4:e443165c3234: imc: fix order of operations in coefficients read [13:28] ffmpeg.git 03Timothy B. Terriberry 07release/2.4:ca8c62d187fd: resample: Avoid off-by-1 errors in PTS calcs. [13:28] ffmpeg.git 03Michael Niedermayer 07release/2.4:3b7db9c4f55d: Merge commit 'e443165c323406d01da7e7930f042d265d01fb35' into release/2.4 [13:28] ffmpeg.git 03Michael Niedermayer 07release/2.4:20071ff1a4ba: Merge commit 'ca8c62d187fdca13979379fb2ab172ed662aa2f8' into release/2.4 [13:35] ffmpeg.git 03Michael Niedermayer 07release/2.4:81b38caf21fc: swresample/swresample: fix sample drop loop end condition [13:35] ffmpeg.git 03Christophe Gisquet 07release/2.4:f3d34cff7681: utvideoenc: properly set slice height/last line [13:35] ffmpeg.git 03Karl Kiniger 07release/2.4:71af22097d33: vf_drawtext: add missing clear of pointers after av_expr_free() [13:35] ffmpeg.git 03Michael Niedermayer 07release/2.4:5a1efc7b8585: postproc/postprocess: fix quant store for fq mode [13:35] ffmpeg.git 03Michael Niedermayer 07release/2.4:bf7ee2524b8d: postproc: fix qp count [13:35] ffmpeg.git 03Michael Niedermayer 07release/2.4:2185103bcdd2: avformat/mxfdec: Fix termination of mxf_data_essence_container_uls [13:35] ffmpeg.git 03Lukasz Marek 07release/2.4:e4d921dc71cd: lavd: export all symbols with av_ prefix [13:35] ffmpeg.git 03Christophe Gisquet 07release/2.4:30a0622a5dbf: avcodec/tiffenc: properly compute packet size [13:35] ffmpeg.git 03Michael Niedermayer 07release/2.4:045670a6f7e4: avcodec/hevc_ps: Check default display window bitstream and skip if invalid [13:35] ffmpeg.git 03Michael Niedermayer 07release/2.4:ca47574e16ca: avcodec/sgidec: fix linesize for 16bit [13:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:73c6520c096b: avcodec/sgidec: fix count check [13:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:8cba067fe52a: avcodec/diracdec: Use 64bit in calculation of codeblock coordinates [13:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:8e95ddbe82c4: avcodec/diracdec: Tighter checks on CODEBLOCKS_X/Y [13:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:3f3e5f8f60ef: avcodec/dirac_arith: fix integer overflow [13:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:c7b7e0790c7b: avcodec/dxa: check dimensions [13:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:66fcf1fa404e: avcodec/dnxhddec: treat pix_fmt like width/height [13:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:24d725f45574: avcodec/utils: Align dimensions by at least their chroma sub-sampling factors. [13:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:1f636a697f6b: avcodec/g2meet: check tile dimensions to avoid integer overflow [13:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:35bc67503e80: avcodec/cook: check that the subpacket sizes fit in block_align [13:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:63523485f4d3: avcodec/svq1dec: zero terminate embedded message before printing [13:40] ffmpeg.git 03Michael Niedermayer 07release/2.4:04aa2ffbcf70: Update for 2.4.3 [14:33] ffmpeg.git 03Christophe Gisquet 07release/2.3:c3c8857263d4: avcodec/tiffenc: properly compute packet size [14:33] ffmpeg.git 03Michael Niedermayer 07release/2.3:4a03c31728c8: avcodec/sgidec: fix linesize for 16bit [14:33] ffmpeg.git 03Michael Niedermayer 07release/2.3:5c791b1c9ce0: avcodec/sgidec: fix count check [14:33] ffmpeg.git 03Michael Niedermayer 07release/2.3:1ec6a3c768b2: avcodec/diracdec: Use 64bit in calculation of codeblock coordinates [14:33] ffmpeg.git 03Michael Niedermayer 07release/2.3:ad98b2891cea: avcodec/diracdec: Tighter checks on CODEBLOCKS_X/Y [14:33] ffmpeg.git 03Michael Niedermayer 07release/2.3:e26fd791efaa: avcodec/dirac_arith: fix integer overflow [14:33] ffmpeg.git 03Michael Niedermayer 07release/2.3:7b7d12ea0448: avcodec/dxa: check dimensions [14:33] ffmpeg.git 03Michael Niedermayer 07release/2.3:e3275571c7e2: avcodec/dnxhddec: treat pix_fmt like width/height [14:34] ffmpeg.git 03Michael Niedermayer 07release/2.3:0db579445f52: avcodec/utils: Align dimensions by at least their chroma sub-sampling factors. [14:34] ffmpeg.git 03Michael Niedermayer 07release/2.3:bfee1e90725a: avcodec/g2meet: check tile dimensions to avoid integer overflow [14:34] ffmpeg.git 03Michael Niedermayer 07release/2.3:25d2a4dde724: avcodec/cook: check that the subpacket sizes fit in block_align [14:34] ffmpeg.git 03Michael Niedermayer 07release/2.3:9395a3a96bf8: avcodec/svq1dec: zero terminate embedded message before printing [14:34] ffmpeg.git 03Michael Niedermayer 07release/2.3:b44506c393b1: update for 2.3.5 [14:47] ffmpeg.git 03Michael Niedermayer 07fatal: ambiguous argument 'refs/tags/n2.3.5': unknown revision or path not in the working tree. [14:47] Use '--' to separate paths from revisions [14:47] refs/tags/n2.3.5:HEAD: avcodec/pthread_slice: Use av_freep() to avoid leaving stale pointers in memory [15:25] ffmpeg.git 03R?mi Denis-Courmont 07release/2.2:c7caed88a035: h264: Always invoke the get_format() callback [15:25] ffmpeg.git 03Michael Niedermayer 07release/2.2:2d1d053c5d2c: Merge commit 'c7caed88a03567e8777a606f4bd42f093c6b302c' into release/2.2 [15:40] ffmpeg.git 03R?mi Denis-Courmont 07release/2.2:0989a120f1de: mpeg12: Always invoke the get_format() callback [15:40] ffmpeg.git 03Vittorio Giovara 07release/2.2:787a6156a2d8: imc: fix order of operations in coefficients read [15:40] ffmpeg.git 03Timothy B. Terriberry 07release/2.2:72ed4166a647: resample: Avoid off-by-1 errors in PTS calcs. [15:40] ffmpeg.git 03Michael Niedermayer 07release/2.2:0f4c03cd6340: Merge commit '0989a120f1dec400c54fcb54670cb84bba36d99b' into release/2.2 [15:40] ffmpeg.git 03Michael Niedermayer 07release/2.2:7ac50d846c8f: Merge commit '787a6156a2d887bb1d65c1233a94a61741e7af7c' into release/2.2 [15:40] ffmpeg.git 03Michael Niedermayer 07release/2.2:3bb4c3e74dee: Merge commit '72ed4166a64714952777fb028b546a52e5b4e2c2' into release/2.2 [15:44] ffmpeg.git 03Reimar D?ffinger 07master:46353759cb3c: mpeg4vdpau: Fix priv data size. [15:45] ffmpeg.git 03Carl Eugen Hoyos 07release/2.4:857e39169728: Stop demuxing wtv on eof. [15:45] ffmpeg.git 03Michael Niedermayer 07release/2.4:39518589e73e: avformat/options_table: add FF_COMPLIANCE_UNOFFICIAL [15:45] ffmpeg.git 03Michael Niedermayer 07release/2.4:a8a6cdfcd7b6: avformat/matroskadec: do not trust the default duration to be the real 1/timebase if its less than 5fps [15:55] what a hack [15:56] ffmpeg.git 03Michael Niedermayer 07release/2.3:4f515913a205: avformat/matroskadec: do not trust the default duration to be the real 1/timebase if its less than 5fps [15:56] ffmpeg.git 03Michael Niedermayer 07release/2.3:10464ca0eba3: avformat/options_table: add FF_COMPLIANCE_UNOFFICIAL [15:56] ffmpeg.git 03Michael Niedermayer 07release/2.3:63ed7e09dd1e: avformat/mpegts: Improve probe heuristic by considering the overall frequency of 0x47 headers [15:56] ffmpeg.git 03Cl?ment BSsch 07release/2.3:193b949f715e: avcodec/mjpegdec: Fix chroma width rounding [15:56] ffmpeg.git 03Michael Niedermayer 07release/2.3:19ccc06d8b61: avformat/mp3dec: Improve seeking frame sync code [15:57] ffmpeg.git 03Michael Niedermayer 07release/2.3:4e2e997fafc6: avformat/mpeg: increase score for short mpeg-ps by 1 [15:57] ffmpeg.git 03Michael Niedermayer 07release/2.3:48b586ca4e8c: ffmpeg: Copy extradata if it has been initialized later from the encoder [15:58] ffmpeg.git 03Christophe Gisquet 07release/2.2:aa40f11b815a: utvideoenc: properly set slice height/last line [15:58] ffmpeg.git 03Michael Niedermayer 07release/2.2:13b7962aae30: postproc/postprocess: fix quant store for fq mode [15:58] ffmpeg.git 03Michael Niedermayer 07release/2.2:17f0581e0dd5: postproc: fix qp count [15:58] ffmpeg.git 03Michael Niedermayer 07release/2.2:6505eb45bcf4: avcodec/diracdec: Use 64bit in calculation of codeblock coordinates [15:58] ffmpeg.git 03Michael Niedermayer 07release/2.2:81e1b5f5fe5b: avcodec/diracdec: Tighter checks on CODEBLOCKS_X/Y [15:58] ffmpeg.git 03Michael Niedermayer 07release/2.2:45361d8aa300: avcodec/dirac_arith: fix integer overflow [15:58] ffmpeg.git 03Michael Niedermayer 07release/2.2:557e8bd58968: avcodec/dxa: check dimensions [15:58] ffmpeg.git 03Michael Niedermayer 07release/2.2:635215381116: avcodec/dnxhddec: treat pix_fmt like width/height [15:59] ffmpeg.git 03Michael Niedermayer 07release/2.2:16a4aef34574: avcodec/utils: Align dimensions by at least their chroma sub-sampling factors. [15:59] ffmpeg.git 03Michael Niedermayer 07release/2.2:f6499563c306: avcodec/g2meet: check tile dimensions to avoid integer overflow [15:59] ffmpeg.git 03Michael Niedermayer 07release/2.2:f00ec3307b5f: avcodec/cook: check that the subpacket sizes fit in block_align [15:59] ffmpeg.git 03Michael Niedermayer 07release/2.2:9b8b35910ffb: avcodec/svq1dec: zero terminate embedded message before printing [15:59] ffmpeg.git 03Michael Niedermayer 07release/2.2:e812a089f549: avcodec/svq3: Dont memcpy AVFrame [16:02] ffmpeg.git 03Michael Niedermayer 07release/2.2:41ee9a44955b: update for 2.2.10 [16:13] ffmpeg.git 03Michael Niedermayer 07release/2.2:c9659dfd2942: avformat/mpeg: increase score for short mpeg-ps by 1 [16:13] ffmpeg.git 03Michael Niedermayer 07release/2.2:114e4b970e0a: avformat/mp3dec: Improve seeking frame sync code [16:13] ffmpeg.git 03Michael Niedermayer 07release/2.2:64624c56784d: avformat/matroskadec: do not trust the default duration to be the real 1/timebase if its less than 5fps [16:14] michaelni: what if the file is actually, say, a slide-show? [16:15] wm4, ehm, what exactly do you talk about, what file ? [16:15] about this commit: avformat/matroskadec: do not trust the default duration to be the real 1/timebase if its less than 5fps [16:18] then the generic code will calculate the frame rate from the timestamps [16:19] why not always do this? [16:19] it needs to read a bit more of the file/stream than otherwise [16:20] ffmpeg.git 03Reimar D?ffinger 07release/2.4:25fc3deed800: mpeg4vdpau: Fix priv data size. [16:20] also i wonder, is the default duration in general set "correctly" for slide shows in matroska [16:20] maybe ideally there should be a separate API function to retrieve such information [16:21] yes [16:28] thardin: I failed to backport the fix for ticket #4040 to origin/release/2.4 [16:28] Do you have time to look at this? Michael wants to do a 2.4 release soon. [16:32] ffmpeg.git 03Michael Niedermayer 07release/2.4:70f6d553d98e: Move get_avc_nalsize() and find_start_code() to h264.h [16:32] ffmpeg.git 03Michael Niedermayer 07release/2.4:5405ba7b635b: avcodec/h264: simplify find_start_code() [16:33] ffmpeg.git 03Michael Niedermayer 07release/2.4:9a641b909cb8: avcodec/h264_parser: rewrite the parse_nal_units() loop logic based on h264.c [16:41] ffmpeg.git 03Thomas Volkert 07master:8d9277c3c01c: avformat/rtpdec_h261: code aligned to the HEVC code [16:51] I am not able to send a mail on ffmpeg-devel list. It have attached some sample reference input images and their output. [16:52] has* [16:52] I* [16:56] arwa: why are you comparing to scale2x? [16:56] it's not the same algorithm as xbr [16:57] but they have xbr implemented [16:57] oh, alright [17:00] arwa, i'm leaving, i'll check it later tonight [17:00] should I send the source code also? [17:00] okay [17:01] arwa, we have a size limit on the mailing-list, so if you add big attachment the message will be moderated [17:02] So, should I mail it to you on your mail id? [17:03] arwa, no, can you put it on a server and post the link? [17:03] also, for small images it shouldn't be a problem [17:03] how big was the attachment? [17:04] the mail already got through [17:04] 2.5 mb zip file [17:04] arwa, that's huge [17:05] I will put the results on a server. [17:08] yup [17:08] upload the images on imgur or lut.im or wherever [17:08] arwa, but yes the email got through so i can see the samples [17:08] anyway, time to leave, see you later [17:12] ffmpeg.git 03Michael Niedermayer 07release/2.3:ab43652c67a4: Move get_avc_nalsize() and find_start_code() to h264.h [17:13] ffmpeg.git 03Michael Niedermayer 07release/2.3:48bf926bad57: avcodec/h264: simplify find_start_code() [17:13] ffmpeg.git 03Michael Niedermayer 07release/2.3:4b8cb3fe51eb: avcodec/h264_parser: rewrite the parse_nal_units() loop logic based on h264.c [17:41] ffmpeg.git 03Michael Niedermayer 07release/1.2:e5cf4d16c60e: avformat/mpeg: increase score for short mpeg-ps by 1 [17:41] ffmpeg.git 03Michael Niedermayer 07release/1.2:1abb0e563918: avformat/matroskadec: do not trust the default duration to be the real 1/timebase if its less than 5fps [18:12] ffmpeg.git 03Kevin Mitchell 07master:2847843868c9: avfilter/idet: add metadata to "current" frame instead of "next" frame [18:12] ffmpeg.git 03Kevin Mitchell 07master:ae6118de19a5: avfilter/idet: add current frame classification to metadata [18:39] ffmpeg.git 03Michael Niedermayer 07release/2.2:667fe8c75b0b: Move get_avc_nalsize() and find_start_code() to h264.h [18:39] ffmpeg.git 03Michael Niedermayer 07release/2.2:76587eea6486: avcodec/h264: simplify find_start_code() [18:39] ffmpeg.git 03Michael Niedermayer 07release/2.2:969aee07e68c: avcodec/h264_parser: rewrite the parse_nal_units() loop logic based on h264.c [19:05] ffmpeg.git 03Kevin Mitchell 07fatal: ambiguous argument 'refs/tags/n2.2.10': unknown revision or path not in the working tree. [19:05] Use '--' to separate paths from revisions [19:05] refs/tags/n2.2.10:HEAD: avfilter/idet: add current frame classification to metadata [19:39] ffmpeg.git 03Rodger Combs 07master:ae437c7ce704: avformat/assenc: Add ignore_gaps option [20:16] ffmpeg.git 03Michael Niedermayer 07master:6d64a14e6dea: avformat: add webp muxer [20:16] ffmpeg.git 03Michael Niedermayer 07master:5aaf5df06de6: avcodec/libwebpenc: support "P" frames in webp animations [20:27] Hi, I have question. Do codecs have any other common options beside these defined in options_table.h? [20:38] lukaszmluki: It is possible that options with identical names exist for different codecs, and I expect such options to exist [20:39] but you mean private options right, per specific codec ony? [20:53] Yes, private options [21:04] ffmpeg.git 03Michael Niedermayer 07release/1.2:7ddf252c7e1e: Move get_avc_nalsize() and find_start_code() to h264.h [21:05] ffmpeg.git 03Michael Niedermayer 07release/1.2:aa9d70587150: avcodec/h264: simplify find_start_code() [21:05] ffmpeg.git 03Michael Niedermayer 07release/1.2:9282c96071be: avcodec/h264_parser: rewrite the parse_nal_units() loop logic based on h264.c [00:00] --- Sun Nov 2 2014 From burek021 at gmail.com Mon Nov 3 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Mon, 3 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141102 Message-ID: <20141103010502.D837E18A024F@apolo.teamnet.rs> [00:08] ffmpeg.git 03Lukasz Marek 07master:e9899ca3ddc1: ffserver_config: dont pass pointer to enum as pointer to int [00:45] ffmpeg.git 03Carl Eugen Hoyos 07release/1.2:a8fd50bb4eeb: Fix compilation after aa9d7058. [00:51] michaelni: "ffmpeg -i input output.webp" still defaults to image2 [00:51] is that intended or you forgot to remove the webp extension from image2 muxer? [01:01] ffmpeg.git 03Michael Niedermayer 07master:f5abd12c4d6c: avformat/img2enc: remove webp extension as we have a dedicated webp muxer [01:25] thanks jamrial , good spot :) [01:58] jamrial, thx ! [02:08] ffmpeg.git 03Michael Niedermayer 07master:547fce95858e: avcodec/h264_slice: Clear table pointers to avoid stale pointers [02:15] no prob :) [02:42] ffmpeg.git 03Michael Niedermayer 07release/2.4:043f32606046: avcodec/h264_slice: Clear table pointers to avoid stale pointers [03:08] ffmpeg.git 03Michael Niedermayer 07fatal: ambiguous argument 'refs/tags/n2.4.3': unknown revision or path not in the working tree. [03:08] Use '--' to separate paths from revisions [03:08] refs/tags/n2.4.3:HEAD: avcodec/h264_slice: Clear table pointers to avoid stale pointers [11:13] ugh [11:13] that ticket [11:14] someone should just say that x264opts passes options straight to libx264, and that libx264 doesn't do any ref calculation if level is set. only x264cli does :P [11:16] you want me to copy paste that to the ticket JEEB ? (honestly cant tell if serious or not) [11:17] I tried to write something but I hadn't saved my trac password to my passphrase chain [11:18] x264opts is meant for straight passing of options to libx264 without lavc's extra intervention, other options can have specific logic within lavc :P [11:18] JEEB : use login bug/bug [11:18] anonymous login :) [11:18] Not anymore [11:18] no ? [11:18] or maybe that was mplayer trac login [11:18] As in now everyone knows it. [11:19] I guess it is still anonymous though. [11:19] but doesn't work :) [11:19] sorry that was mplayer trac login, i got confused [11:19] too early, damn daylight savings time [11:24] JEEB: THere is an email about this "bug" on ffmpeg-user if you want to send an email there. [11:25] I don't want to touch that. I get enough measles and lose SAN points from being on #ffmpeg [11:26] fair enough [11:33] ffmpeg.git 03Michael Niedermayer 07release/1.2:25a9823eb44a: avcodec/h264_slice: Clear table pointers to avoid stale pointers [11:33] ffmpeg.git 03Michael Niedermayer 07release/1.2:42669252e8f5: postproc/postprocess: fix quant store for fq mode [11:33] ffmpeg.git 03Michael Niedermayer 07release/1.2:23b3fe00bb0d: postproc: fix qp count [11:33] ffmpeg.git 03Michael Niedermayer 07release/1.2:c36f5df34aca: avcodec/diracdec: Use 64bit in calculation of codeblock coordinates [11:33] ffmpeg.git 03Michael Niedermayer 07release/1.2:341cf9ed8167: avcodec/diracdec: Tighter checks on CODEBLOCKS_X/Y [11:33] ffmpeg.git 03Michael Niedermayer 07release/1.2:35cf24eee996: avcodec/dirac_arith: fix integer overflow [11:33] ffmpeg.git 03Michael Niedermayer 07release/1.2:d59804bafbaf: avcodec/dxa: check dimensions [11:33] ffmpeg.git 03Michael Niedermayer 07release/1.2:6e45a99a7cc4: avcodec/dnxhddec: treat pix_fmt like width/height [11:33] ffmpeg.git 03Michael Niedermayer 07release/1.2:b92ccfefc3a3: avcodec/utils: Align dimensions by at least their chroma sub-sampling factors. [11:33] ffmpeg.git 03Michael Niedermayer 07release/1.2:96dac432f790: avcodec/svq1dec: zero terminate embedded message before printing [11:36] ffmpeg.git 03Michael Niedermayer 07release/1.2:7e8b4506e0aa: update for 1.2.10 [11:53] ffmpeg.git 03Michael Niedermayer 07master:70b7cf9c39d0: avformat/webpenc: removed unused variable [13:54] nodejs fate eh [13:56] u wot mate [13:56] https://github.com/TimothyGu/fateserver-node [13:57] it's certainly more readable [13:57] i have wondered for a while now if programmers 20 years from now will look back at js like js people look back at 90s cgi perl [19:29] ffmpeg.git 03Michael Niedermayer 07master:a52cb42ba662: avformat/matroskadec: use gmtime_r() for thread saftey [19:29] ffmpeg.git 03Michael Niedermayer 07master:013c3eb05cbb: avfilter/vf_drawtext: use gm_time_r() for thread saftey [20:18] ffmpeg.git 03Michael Niedermayer 07master:bab09864b423: avcodec/jacosubdec: use time_internal.h, simplify code [20:18] ffmpeg.git 03Michael Niedermayer 07master:32a2876b12f6: avformat/segment: use time_internal.h, simplify code [20:18] ffmpeg.git 03Michael Niedermayer 07master:76886589eea7: avformat/wavenc: Use localtime_r() for thread saftey [20:18] ffmpeg.git 03Michael Niedermayer 07master:5ece4f8b7362: avformat/sbgdec: Use localtime_r() for thread saftey [20:18] ffmpeg.git 03Michael Niedermayer 07master:63e62cfbe23d: avformat/img2enc: Use localtime_r() for thread saftey [21:40] from frame.h: "Presentation timestamp in time_base units (time when frame should be shown to user)." on top of pts doxy [21:40] what time_base is it refering to? [21:40] AVStream.time_base? [21:41] if so, we have AVPacket.pts == AVFrame.pts, never a rescaling? [21:41] (i'm trying to follow the same model for subtitles) [21:42] yes AVStream [21:42] and yes [21:42] the decoder doesn't even know the time_base [21:42] or at least that's how it _should_ work [21:43] alright, so there is no automatic rescaling system in place in the decoding part? [21:43] i was wondering if some users would need that [21:43] (requesting all the frame for every media to be rescaled homogeneously) [21:44] frame ts* [21:44] the subtitle converter API actually needs the time_base (breaking all expectations), so I set a fake time_base [21:44] the AVSubtitles are in ms tb or something currently [21:45] and yeah it sucks, i removed all the rescaling in the decoders, it's insane :p [21:45] anyway, ok, thanks [21:57] i wonder if that time_base is not relative to the AVFormatContext.time_base [21:57] oh well.. [22:05] isn't time_base per AVStream? [22:07] i meant AVCodecContext sorry [22:07] (the sub decoders rescale using AVCodecContext->time_base) [00:00] --- Mon Nov 3 2014 From burek021 at gmail.com Mon Nov 3 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Mon, 3 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141102 Message-ID: <20141103010501.CB14918A024E@apolo.teamnet.rs> [01:00] Hi [01:31] Is it possible to only keep video parts containing movement? [03:10] How to rip dvd with chapters? [09:07] yo all [09:08] rtmp://$OPT:rtmp-raw=rtmpe://1.1.1.1/live playpath=toni1 swfUrl=http://ukcast.tv/player/player.swf live=1 pageUrl=http://www.ukcast.tv/ [09:08] how to open this shitty url? [09:34] Hi guys. I use ffmeg to encode to rtmp to a wowza server. the problem is, it seems that some older flash players (i dont know which versions) do not show the vide, they only display audio [09:34] any idea what could be wrong? [09:34] avconv -y -f alsa -i hw:2,0 -f video4linux2 -itsoffset 4.8 -i /dev/video1 -vcodec libx264 -b:v 220k -b:a 40k -strict -2 -acodec aac -f flv rtmp://... [09:34] this is the command I use [09:35] they only *play* the audio of course :p [09:36] emilsedgh: lol maybe they hate H.264? try to use profile baseline i think default is using high [09:37] Action: emilsedgh searches for 'profile baseline' [09:37] thanks DelphiWorld [09:37] emilsedgh: ;) [09:40] woah, soo many options and settings and i dont know a thing about them. I have a lot to learn! [09:41] emilsedgh: lol [09:55] emilsedgh: pm? [10:34] michaelni: congratulation for the debianisation of FFMpeg! [11:06] hey [11:07] is it just me or does ffmpeg not support EA vp61 ? [11:15] Action: DelphiWorld slaps emilsedgh around a bit with a large trout [11:16] Action: emilsedgh gets slapped [11:16] Action: feliwir wants response [12:09] how can i check if a container is seekable? [12:24] saste? [12:26] feliwir: if you're seekable then he is :P [12:30] h?? [12:30] i don't understand that sentence :D [12:37] feliwir: DUDE kyding [12:37] :D [12:39] ilove11ven: spamy:P [12:40] ha? [12:40] ilove11ven: in out ;) [13:00] Hi Guys [13:01] hi [13:01] Is anyony up for some discussion and feedback for an ffmpeg setup? [13:01] can someone tell me if ffmpeg has ea vp6.1 support [13:01] it keeps crashing on me [13:02] anyone* [13:02] where's the difference in that case? [13:08] hi [13:46] Hi, is there any way for the flv format to have a video with multiple pixel aspect ratios? [19:31] Hello, i noticed that in some broken RTMP streams when i set timeout=10 ffmpeg hangs. When i set timeout=2 it works. Why that? [19:56] is here someone who understands the source code of ffmpeg? [20:24] I installed ffmpeg on ubuntu but i can't find ffserver does anyone know how to install it? it is supposed to be part of ffmpeg package [20:37] The ffmpeg package on ubuntu is (probably) not actually FFmpeg. [20:39] You'll probably have to build from source. [20:40] Wasn't it even about to be removed from ffmpeg itself? [20:42] I think the general consensus is: don't get rid of it while it still works for some people. [21:46] is tir supported by ffmpeg? [00:00] --- Mon Nov 3 2014 From burek021 at gmail.com Tue Nov 4 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Tue, 4 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141103 Message-ID: <20141104010502.D82BA18A00C9@apolo.teamnet.rs> [00:23] ubitux: ping [00:23] jamrial: pong [00:24] http://pastebin.com/MmWcCwhP [00:24] :) [00:25] "pslld xmm, [memory], xmm" like you suggested doesn't work :P [00:25] too bad :) [00:26] Does the above look better than the first version? [00:26] i like it better but that's your code [00:28] It may simplify adding indep8 at a later time, so if you like it more i'll go with it [00:28] i'll send a new patch in a few minutes [00:34] ffmpeg.git 03Lukasz Marek 07master:2121e3e13061: ffserver_config: improve error handling [00:34] ffmpeg.git 03Lukasz Marek 07master:9c097f1cfc18: ffserver_config: improve AVOption handing [01:07] ffmpeg.git 03Carl Eugen Hoyos 07master:e6b7246a688b: lavf/movenc: Write G.726 bitrate to make the files decodable. [01:07] ffmpeg.git 03Carl Eugen Hoyos 07master:d457478fb028: Silence warnings for fic files with zero-sized cursors. [02:14] ffmpeg.git 03Reynaldo H. Verdejo Pinochet 07master:200270cc8b28: ffserver_config: fix line lengths [02:15] ffmpeg.git 03Reynaldo H. Verdejo Pinochet 07master:17cc78505c33: ffserver_config: simplify some if true conditions [02:15] ffmpeg.git 03Reynaldo H. Verdejo Pinochet 07master:33aacb775013: ffserver_config: add fixme on buffer_aggressivity/eq deprecation [03:20] ffmpeg.git 03Reynaldo H. Verdejo Pinochet 07fatal: ambiguous argument 'refs/tags/n1.2.10': unknown revision or path not in the working tree. [03:20] Use '--' to separate paths from revisions [03:20] refs/tags/n1.2.10:HEAD: ffserver_config: add fixme on buffer_aggressivity/eq deprecation [03:51] when viewing stream information using ffmpeg -i command, the stream info is showed "Stream #0:0, 0, 1/90000: Video: h264 (libx264), yuv420p, 640x360 [SAR 1:1 DAR 16:9], *0/1*, q=30-100, 1000 kb/s, 25 tbr, 90k tbn", what is 0/1 and what does it mean? [09:22] ffmpeg.git 03Cl?ment BSsch 07master:4f4de7f49e0d: README: fix 2 typo in the doc/examples sentence [09:59] ffmpeg.git 03Stefano Sabatini 07master:f0158e6f0cc1: lavf/flvenc: fail in case the muxed packet is too big [12:00] ffmpeg.git 03Martin Storsj? 07master:2f221b6a9365: movenc: Define the flag bits using shifts instead of as decimal numbers [12:01] ffmpeg.git 03Michael Niedermayer 07master:ea0b9218f4a6: Merge commit '2f221b6a9365aa400061e16266f2d1242f7169f8' [12:10] who is thomas volkert on irc again? [12:11] Does udp-lite actually have any real-world use? [12:11] I mean do you actually get packets which are bitflipped as opposed to just being lost [13:29] ffmpeg.git 03Michael Niedermayer 07master:206c98f303e8: avcodec/options_table fix min of audio channels and sample rate [13:40] ffmpeg.git 03Benoit Fouet 07master:50138ea4f7b3: configure: add xcb cflags and extralibs to cflags and extralibs. [17:33] ffmpeg.git 03Kevin Mitchell 07master:fe6f5f2908ae: avfilter/vf_idet: add a "half_life" option for statistics [17:33] ffmpeg.git 03Michael Niedermayer 07master:5d590d87b30c: avfilter/vf_idet: fix rounding of av_dict_set_fxp() [17:33] ffmpeg.git 03Michael Niedermayer 07master:4bbd8f05f7c3: avfilter/vf_idet: use av_rescale() [17:33] ffmpeg.git 03Michael Niedermayer 07master:898635ad9ec9: avfilter/vf_idet: use exp2() [19:08] ffmpeg.git 03Michael Niedermayer 07master:bd0f866731ac: doc: Better documentation for the bitexact flag [21:01] ffmpeg.git 03Michael Niedermayer 07master:716674b151c1: avcodec/libwebpenc: add quality option [22:16] creating chapters is not possible yet with ffmpeg.c, right? [23:32] ffmpeg.git 03Michael Niedermayer 07master:a6593f7cc6e4: avformat/mpegts: Do not add pid if its already there with add_pid_to_pmt() [23:32] ffmpeg.git 03Michael Niedermayer 07master:a7f25979dd56: avformat/utils: Leave skip_clear enabled until after estimate_timings() [23:32] ffmpeg.git 03Michael Niedermayer 07master:db0471c40f0f: avformat/mpegts: Continue parsing PMTs during duration estimation [00:00] --- Tue Nov 4 2014 From burek021 at gmail.com Tue Nov 4 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Tue, 4 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141103 Message-ID: <20141104010501.C53BF18A00B6@apolo.teamnet.rs> [00:37] hello.. is there a way to embed subtitle to avi permanently? [00:41] with a filter -vf "ass=subtitle.ass" - it must be a subtitle file with a .ass extension, so make sure it is that format. [00:42] oops -vf "subtitle=file.ass" [00:43] maybe it was the first one - I'll check, it's been a while [00:44] blippyp could you tell me the hole command please? [00:46] ya, it's ass=subtitle.ass [00:46] so if you had a subtitle called subs.ass [00:46] and you wanted to join it to a video called vid.avi [00:46] ass=ass.ass [00:46] right [00:47] ffmepeg -i vid.avi -vf "ass=subs.ass" out.avi [00:47] obviously you want to add a little more in there for codecs and whatever [03:02] Surfer: https://trac.ffmpeg.org/wiki/HowToBurnSubtitlesIntoVideo [03:03] thanks relaxed ;) [03:04] relaxed is it a re-enconde? [03:29] Surfer: yes, hardsubbing requires re-encoding [03:29] relaxed no success :/ [03:32] relaxed http://pastebin.com/MPJeKEGF [03:34] Surfer: your ffmpeg verison is too old. Either compile a recent version or try my build http://johnvansickle.com/ffmpeg/ [03:34] relaxed i would like to reburn the video with subtitle [03:34] not just add subtitle like a second file [03:34] ok [03:35] Surfer: then do like I said earlier with the -vf "ass=subtitle.ass" filter, but you will need to convert your .srt file to a .ass file [03:36] right [03:36] hold on - I'm looking how to do that now (it's been a while) [03:36] ok [03:36] oh it's easy [03:37] just run ffmpeg -i subs.srt subs.ass [03:37] then run a command like ffmpeg -i input.avi -vf "ass=subs.ass" -c:v libx264 -crf 18 -c:a copy output.mkv (mp4, whatever...) [03:38] it will take a long time though [03:38] Surfer: is there any reason you wanted an .avi? [03:43] Surfer: Before you waste your time though, your other file was working perfectly (well, the part I watched was) - Your problem was with your player - I haven't used VLC in a long time, but in order to get it to use the subtitles I had to tell it to use it each time I played a file. The options are there, you just need to find them. [03:43] relaxed just because it's smaller than mkv [03:44] blippyp yes.. files were working perfectly as you said [03:44] Surfer: No it wouldn't be - the codecs are what determines the sizes - avi, mp4, mkv etc.. are just containers and won't really affect your file size. [03:44] but i don't want to add a subtitle file in a avi file [03:45] actually mkv should be smaller than avi [03:45] i want to rebuild the video with the sub [03:45] relaxed i see more avi files to download [03:45] That's fine - people have their reasons - but it will take a lot longer to re-encode it than it will to simply re-mux the subtitle stream into it... [03:45] i prefer avi [03:46] when viewing stream information using ffmpeg -i command, the stream info is showed "Stream #0:0, 0, 1/90000: Video: h264 (libx264), yuv420p, 640x360 [SAR 1:1 DAR 16:9], *0/1*, q=30-100, 1000 kb/s, 25 tbr, 90k tbn" [03:46] what's 0/1 in that report? [03:46] blippyp i prefer to re-enconde [03:47] blippyp is there any way to do that with ffmpeg? [03:47] No biggy - I sometimes burn them on two (not for full movies) but to each his own - Was only curious, it's a time consuming process. (especially for me - my systems aren't exactly made for video editing) [03:47] two=too/also [03:48] Surfer: the command I gave you last will re-encode the video like you want [03:49] The only thing that won't get re-encoded is the audio - the command I gave you told ffmpeg to copy it instead (since you weren't altering it) [03:49] ----> Surfer: then do like I said earlier with the -vf "ass=subtitle.ass" filter, but you will need to convert your .srt file to a .ass file [03:49] this? [03:49] no [03:49] just below that [03:49] I wrote out the entire command for you - you just have to copy/paste [03:50] alter the filenames accordingly [03:50] ffmepeg -i vid.avi -vf "ass=subs.ass" out.avi [03:50] ? [03:50] no - hold on [03:50] did you already convert the srt at ass? ake sure your version can do that first. [03:50] first convert your .srt file with: ffmpeg -i subs.srt subs.as [03:51] to, not at [03:51] then run: ffmpeg -i input.avi -vf "ass=subs.ass" -c:v libx264 -crf 18 -c:a copy output.avi [03:51] i'm installing a new version [03:52] right [03:52] the crf is the compression ration 18 is pretty good (I usually notice a difference between 17 and 18 0 is no compression, default is 21 or 23 or something. [03:52] version 2.2.1-65.el6 works fine? i'm using repositories [03:53] of CentOS [03:54] a lot of conflicts [03:54] no idea - I'm on Arch, doesn't hurt to try [03:54] conflicts? [03:55] one repository conflicts with other [03:55] and break installed packages [03:56] no idea, like I said I don't use CentOS (never have) I would just try it with whatever you already have installed and see how that goes first [03:56] you should never add external repos for this very reason [04:01] blippyp http://pastebin.com/L52cfVsD [04:03] probably has to do with what's specified in your subtitle (like fonts and stuff) - open it up in an editor and put something more generic in it... [04:03] http://sprunge.us/VebT [04:04] I use this as a base when I'm playing with subtitles - it should be pretty generic - check the V4+ Styles section - I had to modify that [04:04] (I'm not an expert on subtitles - someone else might be able to look at your file and tell you exactly what to change) [04:04] Could also be a whole other issue as well [04:05] But I would change that section to match mine and see what happens [04:16] Surver: there is also a static build you can download - I think the bot in here will tell you how to get it... Don't remember the command, I haven't chatted in here in a long time [04:17] there ya go [04:17] you can just download that and copy it somehwere (maybe rename it to ffmpeg-static and stick it in your /bin folder or something) [04:17] it should be up to date I think [04:27] I wish to edit a bunch of v422/pcm_s16le files I wish to edit. Should I convert these first to another format? [04:28] pardon the repetition... [04:28] why would you need to do that? [04:30] unless your source is raw (in one way or another), everytime you convert it/re-encode it you will likely degrade the quality [04:31] I'm not sure if I need to, just wondering if another format is better to work with in a linear fashion. They are raw format currently. [04:31] if your editor will open them, then keep them like that... ;) [04:32] if you have to convert them, then keep the quality at 100% when doing so. [04:34] Thanks blippyp. Pretty much what I expected for an answer :) One more newbie Q: Any recommendation for the best current linear editor under Linux? I currently use VDub under Windows but would like to use Linux... [04:36] I have around 70 files I need to combine and crop from original camcorder footage. [04:37] no idea - I use ffmpeg for everything believe it or not [04:37] I would ffmpeg - you could easily script it if you know anything about that [04:39] It might be the way, just that cropping out the unwanted noise and "Is the camera still running?"-looking at the ground footage would be easier to do visually. [04:39] blippyp it's working! [04:39] thank you! [04:39] :D [04:39] http://81.4.100.163/2.avi [04:40] Surfer: awesome - glad to hear it [04:40] blippyp was just a version problem [04:41] *ffmpeg [04:41] downloaded the static build I take it? [04:41] but subtitles are still getting error [04:41] did you change the styles section like I told ya? [04:41] http://johnvansickle.com/ffmpeg/releases/ [04:42] which section? [04:43] the styles section - it's probably just complaining about the font [04:43] it's working though - I'm watching it now [04:43] and my subtitles are disabled, still showing them [04:44] cool ;D [04:44] you might have to watch it through once to verify that the subtitles didn't bug-up on ya [04:44] was your fault haha [04:44] nope [04:44] couldn't tell ya why it's buggered up - I'd watch it to verify you're happy with it [04:44] just kidding ;P [04:45] the first times the hardest - next time you try this it will be easier... ;) [04:45] subs can be a pain [04:45] i hope so [04:45] yes, they are [04:46] i see now lol [04:46] you learned a lot though - you must have [04:46] yeah sure [04:46] seems interesting [04:46] it is - now type man ffmpeg-filters [04:46] ffmpeg is powerful [04:46] and play with a video - you'll be blown away [04:46] you have no idea - I stopped using all other editors because of it [04:47] yeah, it can take time to get use to (doesn't everything)??? It's worth it imo... [04:47] i'm going to install microsoft fonts [04:47] especially if you're like me and would rather run everything thru a script [04:47] yeah [04:47] script is the key! hahaha [04:48] when I had problems with the subs before I'm pretty sure it was font related also [04:48] [Parsed_ass_0 @ 0xaf8cc20] fontconfig: cannot find font 'Arial', falling back to 'Liberation Sans' [04:48] [Parsed_ass_0 @ 0xaf8cc20] fontconfig: cannot find font 'Arial', falling back to 'Liberation Sans Italic' [04:48] new error [04:48] wait till you see how crazy those scripts can get [04:48] yeah - font issues [04:48] but i think i can fix it [04:48] no need - if you're happy with the font [04:48] it eventually found one to use [04:49] the fonts aren't an issue anymore because now it's 'burned' into the video itself - it no longer needs the font files [04:49] you don't need to specify a subtitle when viewing it either in a player, they will always be there [04:49] yeah i know [04:50] you encoded that fast [04:50] I wish I had your system [04:50] but default font is horrible [04:50] haha [04:50] It would have taken me forever to encode that [04:50] learn how to pull a section of video to play with - it will save you tons of time [04:50] it's a low profile [04:50] frame=12153 fps= 13 q=23.0 size= 102959kB time=00:08:26.97 bitrate=1663.7kbits/s [04:51] ah, still would have taken me longer [04:51] I usually do everything with -qp 0 (like -crf 0, but better) - no quality loss, it encodes as fast as it can [04:51] when I'm happy, then I encode with a crf 17 [04:52] you also have presets to use - ultrafast, slow, veryslow, etc... [04:52] man ffmpeg [04:52] it will tell you more about the basics [04:52] ok [04:52] i'll take a look [04:52] happy ffmpeggin' [04:52] ;) [04:53] i'm trying to install microsoft fonts [04:53] lol [04:53] thanks =] [04:53] no problem [04:55] I'm gonna go watch The Walking Dead now - but I'll be back in an hour if you have any more questions - just post them, either someone else will answer them or I'll try to help you when I get back [04:55] ok [04:55] thanks blippyp [04:56] thanks for the help bro ;) [04:56] no problem, always glad to help (and receive help myself!) [09:05] can ffmpeg.exe remux mp4 by direct stream copy video and reencode audio lpcm->aaC? [09:19] <__jack__> dongs: if ffmpeg.exe is nothing by ffmpeg in PE format: sure you can [09:19] <__jack__> ffmpeg -i source -c:v copy -c:a aac output.mp4 [09:19] ok, lemme try that. [09:22] it worked [09:27] how do I make it do same thing on *.mp4 [09:29] I have a webcam stream and a static image. How can I switch between the webcam and the static image and output a live stream? [09:30] dongs: maybe http://ss64.com/nt/forfiles.html ? [09:32] <__jack__> bash loop :D [09:32] i use a proper os so "bash loops" is not an option [09:33] install cygwin! [09:33] then use xargs [10:02] Action: DelphiWorld burp emilsedgh [10:19] is there a way to find out the available audio codecs for ffmpeg? [10:19] reason I ask - ffmpeg tells me "Unknown encoder 'libmp3lame'" [10:20] shevy: offcource, can you ask ffmpeg? :P [10:20] but I could swear it worked before (I just recompiled ffmpeg) [10:20] well, how? [10:20] shevy: ffmpeg -codecs :-P [10:20] cool [10:21] shevy: ;) [10:21] hmm [10:21] oh [10:21] --enable-libmp3lame enable MP3 encoding via libmp3lame [no] [10:21] that means it must be specifically enabled right? [10:21] the [no] part [10:22] shevy: you must configure it [10:23] okay! [10:27] shevy: :P [10:28] I thought I did that but I think I was multitasking and forgot it [10:29] shevy: multitasking? you should have double taped the home button to see it ;-) [10:29] yeah [10:29] about 10 open kde konsole tabs [10:29] about 20 tabs in firefox [10:30] woooooooof [10:30] running editor... 5 open pdfs ... mplayer is playing some video ... 3 docs open... [10:30] the best part is the bookmarks feature in browsers [10:30] I bookmark stuff like "hey I will look at this later again" - most of the time, I never look at it again hahaha [10:30] shevy: go to sleep a bit... your engine died;) [10:30] I can't! [10:30] I need to work faster! [10:31] shevy: ok, then you should buy oel from me! :P [14:31] hi all, I have a problem with filter_complex, i would be grateful if anyone can help me out, my command has [0:v]trim=404:414,setpts=PTS-STARTPTS[tr1];[0:v]trim=410:414,setpts=PTS-STARTPTS[tr2];[tr2]setpts=0.25/PTS[v2] and then use concat, (full command and output here http://pastebin.com/CQTn72h8 for some reason whenever i use the same part of a clip in another track the results are not what i expect, see the video from this command [14:31] can anyone please guide me to a solution? [14:53] is there anyone around who can help me find a solution? [14:57] hi all, I have a problem with filter_complex, i would be grateful if anyone can help me out, my command has [0:v]trim=404:414,setpts=PTS-STARTPTS[tr1];[0:v]trim=410:414,setpts=PTS-STARTPTS[tr2];[tr2]setpts=0.25/PTS[v2] and then use concat, (full command and output here http://pastebin.com/CQTn72h8 for some reason whenever i use the same part of a clip in another track the results are not what i expect, see the video from this command [16:40] hello [16:41] I have installed ffmpeg in /usr/local/bin/ffmpeg [16:41] i can see that by typing the path that its installed [16:41] but clip bucket does not recognize it [16:41] any suggestions? [18:21] I have a .MOV recorded on an iphone 6+ and i'm trying to extract a still from it. ffmpeg reads it just fine when I give it the file, but if I cat it to ffmpeg and tell ffmpeg to read from stdin, I get "Unable to parse option value "-1" as pixel format" [18:24] ah shit [18:24] http://ffmpeg.org/pipermail/ffmpeg-user/2012-November/011084.html [19:19] hi, I am trying to stream hls segment directly to ftp server, While hls is still writing segment, it updates m3u8 manifest [19:20] I want hls to update m3u8 only when it has done writing to file [19:31] Hey folks! I was curious there are known issues using -acodec copy -vcodec copy and having the auto / video become slightly out of sync. I'm taking a large video and splitting it up into smaller ones. The first small video is fine, but the other ones are off a second or so. Any thoughts? [19:38] splitting when using codec copy mode is inherently inaccurate, because it can only split on keyframes; but I'd expect it would set the pts such that it would play back synchronized... [19:41] kepstin-laptop: Thank you for the thoughts. So, if it doesn't actually end up playing back synchronized (there's that 1 second delay), would that imply that increasing the number of keyframes could help? [19:42] what are you using as the output container? that might make a difference. And what player? [19:45] kepstin-laptop: As the container, I'm using a .mov (though I've tried .mp4 as well). For a player, I've tried a few - QuickTime, Screenflow, Chrome. [19:45] hmm. mov and mp4 are basically the same thing. [19:46] you could try something like mkv, just to see if it behaves differently. [20:22] Wanting the latest version of ffmpeg on my Mac, instead of using homebrew (which seems to be defaulting to ffpmeg 2.4.2) Im compiling manually but getting ERROR: libfdk_aac not found - this after installing libfdk_aac. Have tried adding export PATH=$PATH:/Users/[my_user_name]/ffmpeg/fdk-aac to my shell profile; same result. My pastebin of CLI commands and output: http://pastebin.com/78Csa1qK [20:25] cobadger: have you tried using brew's --HEAD option? [20:25] might be easier for this situation [20:26] rcombs: no, this is kind of new to me. so [20:26] brew install ffmpeg --with-fdk-aac --with-ffplay --with-freetype --with-frei0r --with-libass --with-libvo-aacenc --with-libvorbis --with-libvpx --with-opencore-amr --with-openjpeg --with-opus --with-rtmpdump --with-schroedinger --with-speex --with-theora --with-tools HEAD [20:26] ? [20:27] LGTM [20:28] rcombs: thanks, giving it a go [20:32] gotta go; good luck! [20:35] rcombs ++ [20:51] hey all [20:51] https://www.youtube.com/watch?v=lypIoOlBDdA [20:52] mostly it went ok! turns out we didn't turn off autofocus on the webcams though, so we got a lot of good shots of the judges' heads. [20:53] oh well. live and learn. people seemed reasoanbly pleased to get something at least. :-D [20:55] thanks to everyone here for putting up with me for the past week. :) [21:26] Is there any way to make drawtext fit the fontsize to a predefined box size? [22:31] I've installed ffmpeg on OS X using homebrew, but in the build the sftp protocol is not supported. Is there a way for me to compile it manually so I can use that feature? [23:52] bees: did you add --enable-libssh? [23:53] not sure what homebrew would call that. --with-libssh? [23:56] it wasnt listed when i ran [23:56] brew info ffmpeg. but i guess i will try it anyway [23:56] Action: llogan ignant about brew [00:00] --- Tue Nov 4 2014 From burek021 at gmail.com Wed Nov 5 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Wed, 5 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141104 Message-ID: <20141105010502.23E982AD60CA@apolo.teamnet.rs> [01:32] how do i include mulitple vf filters [01:32] i tried -vf 'scale=-1:360,ass=[HorribleSubs]_Trinity_Seven_-_01_[720p].mkv.ass' but it doesnt work [01:35] !pastebin ashlee [01:35] er... what was the command again? [01:35] anyway, pastebin your command with ffmpeg's complete output [01:35] how do i chain two vf filters [01:35] or just include two vf filters [01:35] the way you did looks not wrong to me [01:36] so post your output [01:36] it says cant initialize [AVFilterGraph @ 0x7fc910503ae0] Error initializing filter 'ass' with args '' [01:36] probably shell escaping then [01:36] [Parsed_ass_1 @ 0x7fc9106004a0] No filename provided! [01:38] ok thanks [01:38] oh also, you can pass an mkv file to the ass-filter [01:38] you don't need to extract it first [01:38] which subtitle track will it pick [01:38] the first one i guess? since when did horriblesubs release more than 1 subtitle track? [01:39] (except for monogatari s2) [01:39] im not just dealing with one file [01:39] ah [01:39] i have other files which have multiple subtitle tracks [01:39] i will stick to this method for now [01:40] are there docs on how that works [01:40] not sure, when in doubt, test it out [01:40] yeah i rather not [01:41] since i know this way works lol [01:41] read man ffmpeg-filters too [01:41] ah, if you use subtitles instead of ass you can select the stream index [01:42] for example: subtitles=video.mkv:si=1 [01:47] So I am totally new to ffmpeg (but not handling audio/video in general), but I have a clear goal. I want to take a variable number of frames at random from a directory full of video files, and render them together into one file. Where would be a good place to start looking for resources? [01:48] https://trac.ffmpeg.org/wiki/Create%20a%20video%20slideshow%20from%20images [01:48] maybe take a look at the image2pipe input format so you can pipe random images into ffmpeg [01:49] the select filter may also be worth a look. [01:49] (would be like: cat *.jpg | ffmpeg -i - [other stuff] out.mp4 [01:49] ) [01:51] i see [01:53] That is great, I'll start taking a look at that. [01:56] Hmm... I can see now how I could use images to achieve this, but I wonder how I could pull just small segments of the videos. Kind of like a quick random montage from a selection of videos. [03:26] Does anyone have a good boiler plate to convert avi to h.264 web ready video? [03:30] d3m0n: `ffmpeg -i in.avi -vcodec libx264 -crf 18 -acodec libfdk_aac -faststart 1 out.mp4` [03:31] Thanks rcombs, I'll give it a try. [03:32] Alski: If I were you I would probably make an EDL file using mplayer instead of pulling 'random' clips, then I would use the time stamps from the EDL file to create your montages. (just a thought). [03:57] rcombs: Unrecognized option "faststart" [03:58] d3m0n: erm, -mov_flags +faststart [03:59] Unrecognized options "mov_flags" haha [04:03] Ahh I think it's -movflags [04:33] Hello everybody [04:33] iv got into trouble when I'm building ffmpeg-php module [04:34] can anyone help ? [04:40] >php [04:41] yep php [07:03] How can I increase/decrease the quality when creating webm: [07:03] >ffmpeg -i input.avi -vcodec libvpx -acodec libvorbis output.webm [07:06] 4chan? [07:06] huh? [07:07] just thought you wanted it for 4chan [07:07] webm is widely O_O [07:07] used* [07:07] ya, I know - but most people want it for 4chan... ;) [07:07] Nah for my site [07:08] Here's a script I made for 4chan, it will show you lots I think - https://github.com/BlippyPippins/4ChanWebm [08:29] avi to webm keeps screwing up my color [08:29] how can i set the best output [08:30] I already tried -quality best [08:30] https://trac.ffmpeg.org/wiki/Encode/VP8 ? [08:30] "ffmpeg -i desktop-src.avi -vcodec libvpx -quality good -cpu-used 0 -acodec libv [08:30] orbis desktop.webm" [08:32] And what do you mean screwing up your color? [08:32] It flickers differnt hues [08:32] post your code [08:32] That my source doesn't have [08:32] I did above? [08:33] "ffmpeg -i desktop-src.avi -vcodec libvpx -quality good -cpu-used 0 -acodec libvorbis desktop.webm" [08:33] oh [08:33] Set a bitrate and -q modifiers for libvpx [08:34] did it switch the format for you for some reason? maybe post your output from running the command to pastebin or whatever [08:35] One sec, just ran the command agian with "-minrate 1M -maxrate 1M -b:v 1M " [08:35] See how that looks [08:43] Ahh yea adding bitrate helped, I'll have to play around a bit more. Thanks [08:43] Whats the default bitrate if you don't set it? [08:44] don't remember, low - you have to configure it right [08:44] that script i gave you will work great - just set the file size limite and the time limit for your videos [08:45] there's another method - in my script it's a two-pass method since 4chan has a limit on the file sizes [08:45] I never did learn how to do it the other way [08:45] I am using windows D: [08:45] it's not as easy as setting x264 [08:46] ah [08:46] still - the script is pretty easy to read - the science is still there - there's math to determine your bitrates [08:48] your bitrate=(mbLimit*8192/timeLimit) [08:48] follow that to determine your bitrate and you will always get a good encoding [08:48] awesome, thanks. I'll try that [08:48] the ressult will be in K [09:59] hey [09:59] is there away to output the current audio level in real-time? [10:07] I have seen volume-detect however that is at the end [11:46] hey guys. I ended up ripping out my ioctls from the v4l2 code in ffmpeg master and made a standalone tool instead to twiddle the C920 H.264 settings instead. It's a little simpler this way. [11:46] you can find the code at http://pastebin.com/MJdVUyuC [14:07] Hello, does anyone here have experience with AV Foundation screen recording? [14:17] how do I get more info about a .mp4 file without playing it? [14:20] with ffprobe [14:23] cool [14:25] that works nicely [14:25] now I have another question [14:25] I get something like this here: [14:25] "Input #0, mov,mp4,m4a,3gp,3g2,mj2," [14:25] does this mean it has all these codecs in use? [14:25] this is a .mp4 file [14:26] Video: h264 [14:26] Audio: aac [14:26] hmm [14:26] I guess it uses only h264 and aac... not sure what that line next to "Input:" tells me [14:27] The format [14:28] hmm [14:29] And since mov,mp4,m4a,3gp,3g2 and mj2 are closely related it's almost impossible and usually unnecessary to know which exactly it is. [14:29] aha, ok [14:31] don't parse that info dump [14:31] use the -show_* options [14:32] <__jack__> -print_format is also usefull [14:37] Action: DelphiWorld slaps emilsedgh around a bit with a large trout [14:37] yo all [14:37] guys [14:38] i am having a bonch of streams [14:38] ffmpeg -re -i ... -i ... -i ... [14:38] and muxing them in a single mux [14:38] how to include a audio track? [14:39] <__jack__> beuh, just mix it as a video steam ? ffmpeg -i source1 -i source2 -map 0:0 -map 1:0 .. -c .. output.mkv [14:39] __jack__: lol. if you see my cmd... you'lle explode [14:40] __jack__: http://paste.debian.net/130232/ [14:41] DelphiWorld: I believe you're missing a '-i' in front of $URLIN_1 [14:42] c_14: what do you think of the command? ;) [14:43] Other than that it's huge? [14:43] c_14: allmost working [14:43] c_14: but its saying the loop option not found [14:43] for the audio file how to make it repeat [14:44] loop forever? [14:45] c_14: yeah [14:45] write/find a program that'll loop it for you and output to stdout/a named pipe and use that as an input for ffmpeg [14:46] c_14: mmmmmm, ffmpeg itself couldn't repeat? [14:46] then what's -loop [14:46] I do not know of a way to infinitely loop input with ffmpeg (for non-picture files). [14:46] Action: DelphiWorld love ffmpeg... [14:46] -loop _should_ probably do what you want, but it was never implemented completely [14:46] c_14: i added after -i flacfile.flac [14:47] for example -i my.flac -loop 1 [14:47] am i correct? [14:47] If it were functional, it'd have to be in front of the -i my.flac [14:47] let me try [14:51] c_14: ok removed loop & gav up on it [14:51] but still audio not up http://paste.debian.net/130234/ [14:53] No audio at all? [14:53] c_14: the command dont evean run [14:53] i dont know how to capture the log for you [14:56] '| curl -n -F 'f:1=<-' http://ix.io' <- put that behind the command exec [14:57] might need |& [14:58] c_14: but now no reply from curl? [14:58] how to give you the url [15:00] <__jack__> c_14: awesome! [15:00] <__jack__> DelphiWorld: the url is printed [15:00] no, its not printed for me. remember i am runing this long command from a script! [15:01] Press q to quit ffmpeg or kill it manually [15:01] Using pkill ffmpeg or something [15:01] ffmpeg is not evean runing [15:01] it refuse to run at all [15:01] and print all my long cmd on the terminal [15:01] <__jack__> run ./script.sh | blabla, that will give you http://ix.io/doh [15:02] <__jack__> or |& [15:02] lol __jack__ doh:P [15:04] here goes c_14 http://ix.io/bD [15:05] but empty... [15:05] Did you try with |& ? [15:05] yeah and retrying now [15:05] if i do & i dont see the url [15:11] c_14: i got the issue i think [15:11] need --map [15:11] the audio is --map 0:0 and the others --map 0:1 [15:11] am i correct? [15:17] The audio is -map 0, the others should be -map whatever the final output pad[s] of your filtergraph are [15:18] c_14: ffmpeg -re -i c14.flac --map 0.0 -i c14.avi --map 0.1 [15:18] correct? [15:20] -i c14.flac -map 0:0 -i c14.avi -map 1:0 [15:20] ouch [15:22] nop [15:22] still dont work [15:22] Action: DelphiWorld give up [15:23] The -map has to be an output option though. [15:23] So at the end [15:23] -i c14.flac -i c14.avi -map 0:0 -map 1:0 [15:23] etc [15:24] c_14: but after all the filters / blabla befaure the output url? [15:24] yep [15:25] oh... c_14 a big issue now [15:25] the urls am i using for the input have option [15:26] at the end i'm adding -an [15:26] but the flac file is mine [15:26] so no idea what to do [15:26] You'll have to get rid of that -an [15:26] in OUTPUT_ARG [15:26] c_14: then how to kick the audio out of the urls am i using for the input [15:27] Get rid of the implicit maps and make them explicit. [15:27] how? [15:27] IE add a named output pad to your filtergraph and use -map 0:0 -map '[namedoutput]' [15:27] Assuming the audio file you want is the first stream of the first input. [15:28] mmmmmmmmmmmmmmmmmm [15:28] Basically just make this '[tmp29] [E_F] $OVER:x=850:y=480' this '[tmp29] [E_F] $OVER:x=850:y=480[v]' and add -map 0:0 -map '[v]' [15:29] still a bit complicated [15:31] c_14: could you add it to my script and past? [15:31] http://ix.io/f2r [15:31] Should be it. [15:31] is it pocible to download... [15:32] curl -o script.sh http://ix.io/f2r [15:33] nop c_14, i got Stream map ''[v]'' matches no streams. [15:34] Get rid of the '' around [v] [15:34] http://ix.io/f2t [15:34] like so [15:34] c_14: i love that pb man [15:35] Have to go do something. Will be back in about 30min or so. [15:35] lol c_14 :P [15:35] back to the old issue i has:) [15:40] Action: DelphiWorld slaps emilsedgh around a bit with a large trout [15:58] hi all, I have a problem with filter_complex, i would be grateful if anyone can help me out, my command has [0:v]trim=404:414,setpts=PTS-STARTPTS[tr1];[0:v]trim=410:414,setpts=PTS-STARTPTS[tr2];[tr2]setpts=0.25/PTS[v2] and then use concat, (full command and output here http://pastebin.com/CQTn72h8 for some reason whenever i use the same part of a clip in another track the results are not what i expect, see the video from this command [16:10] DelphiWorld: What old issue? [16:10] c_14: no stream... [16:11] Hi all, I have a series of .dat file contained with a folder called dir00000 which appears to be from an unbranded 16 channel h264 cctv system. I have managed to play a single dat file via ffmpeg but i how do I string all the dat files together and is there anyway of selecting which feed to view rather than the footage rotating between all 16 channels? [16:11] DelphiWorld: Is FFmpeg giving that error? [16:12] c_14: yeah [16:14] Can you add -report and then pastebin the ffmpeg-[date].log. Set that anywhere in the commandline, or set FFREPORT to any value. [16:15] ridders24: https://trac.ffmpeg.org/wiki/How%20to%20concatenate%20(join,%20merge)%20media%20files and what do you mean with rotating? [16:15] sruli_: you might have to supply the input file numerous times, once per trim command where you are cutting over the same part [16:21] c_14: cheers I'll have a look at that. What I mean by rotating is that each .dat appears to contact a very brief recording from each of the 16 channels. So you'll briefly get footage, then grey where it moves on the the next need but no cam is connected, then moves on to another say cam 3 thats connected and so on [16:22] ridders24: you'll need to build/find something that'll parse that into separate streams [16:33] i am looking for an ffmpeg expert to trouble shoot a small issue of audio/video syncing when Concatenating 3 video files. [16:33] probably a simple fix, but i am not a programmer [16:50] is there away to output the current audio level in real-time?, I have seen volume-detect however that is at the end [19:33] If I was to string a series of videos together, I would probably use mkfifo right? [19:38] No. [19:38] http://ffmpeg.org/ffmpeg-all.html#concat-1 http://ffmpeg.org/ffmpeg-all.html#concat-2 or http://ffmpeg.org/ffmpeg-all.html#concat-3 should work for that. [19:42] Awesome, I'll start looking in the docs for more like that. Just wondering though, my goal is to make a script that can take a series of videos, pull a small segment from each of them, and then concatenate them together. Kind of a "datablast" of video clips. Is this something I can achieve by just going through the ffmpeg docs or do I have long road ahead of me? [19:55] Alski: do you want to re-encode them or just mux them together? [19:57] What would be the functional difference? [19:57] unless there all in perfectly matching video codecs and you're doing all splitting on keyframes, that'll really need a re-encode. [19:57] (sorry, new to ffmpeg) [20:27] you cannot streamcopy with filters, in order to re-mux the clips you would have to use bitstream filters, which basically means that re-muxing would depend on the codecs used in the source files. [23:02] im using ffmpeg to convert an mp3 file to wav using the pcm_s16le codec. i am attempting to lower the bitrate to 13kb/s using the option "-ab 13k", however the output bitrate states 128kb/s. has anyone run into this issue before? [23:03] nray: setting the output bitrate on a raw pcm encoding is meaningless [23:05] kepstin-laptop: i must be going the wrong direction then. what i need is to convert mp3 files to a Microsoft PCM format (readable by our Asterisk system). any suggestions/direction? [23:06] PCM sounds like .wav nray [23:06] nray: ffmpeg -i file.mp3 -c:a pcm_s16le test.wav will do that, can you give the command you're running and the output you get? (in a pastebin?) [23:07] nray: if you need it for asterisk, it might require a particular sample rate. In that case, you'd have to use an audio filter to resample, I think. [23:07] You can usually just use -ar [23:11] keptstin-laptop let me see if i cant find a safe call recording i could get in pastebin. is there a filter available within ffmpeg to do the resample? [23:12] You can usually just use -ar [23:12] If that doesn't work aresample [23:21] c_14 that definitely accomplished the bitrate changes, however the audio is now coming out extremely muffled and inaudible. [23:27] You could try the aresample filter, see if that gives better results. [23:27] Though the audio quality will deteriorate. [23:27] That will always happen when downsampling. [23:29] You can also try dithering or switching to the SoX resampler. [23:29] See man ffmpeg-resampler [23:43] well, if you're downsampling to narrowband or something for voip, you lose audio quality. no way around that [23:43] it'll always sound muffled, since the high frequencies are lost [00:00] --- Wed Nov 5 2014 From burek021 at gmail.com Wed Nov 5 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Wed, 5 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141104 Message-ID: <20141105010503.2BD3318A0250@apolo.teamnet.rs> [00:42] ffmpeg.git 03Michael Niedermayer 07master:09711545f5d2: avformat/rtmpproto: Do not fail when the length cannot be determined for live streams [00:59] are github pull requests encouraged/discouraged? [01:01] encouraged [01:01] i think [01:01] put it another way, we like pull requests [01:01] good to hear. patches/email are so cumbersome [01:01] kevmitch: patches sent to ffmpeg-devel mailing list is the best way, mostly because some pull requests have been forgotten on occassion, but we've gotten better i think. [01:02] yes, if you want more review, patches to list are good, and can send with git send email [01:02] which is easier ... i'm not sure how easy than git pull request though [01:02] only a few people look at the pull requests, AFAIK. [01:02] yeah, but i have to confdigure my MTA first [01:03] which is another error prone pain [01:03] you could send as an attachment made from git format-patch. [01:03] yeah, that's what I've been doing [01:03] gmail is not designed for it though [01:04] gmail is not designed for * [01:04] lol [01:04] if you *must* use pull request we can live with it [01:07] I'll just figure out my MTA and use git-email. I kind of got the impression that the turnaround might be longer on github. [01:13] If you can't get git send-email to work we also tolerate attached pacthes from git format-patch [01:28] kevmitch / llogan : we could make our own email server for people who use gmail but want to send patches [01:29] i mean... free email server for devels and git-send-email :P [01:29] gmail people could still then reply on list [01:29] just git would use our servers [01:42] Compn: it should be possible to git-send email through gmail, so I don't know if that'd be necesssary [01:43] although it would make it easy to provide a very general set of instructions for setting up git send-email I guess [01:45] of course it's possible. just configure their smtp server on git [01:49] The only problem I saw with using gmail was that the password appears to be stored in plain text. [01:50] then your password should be password = "Enter password here" [01:51] nobody would suspect it [01:51] :) [01:52] I just abuse my ISP's mail server which apparently lets me send mails without a password and with lying about the from address. [01:54] is there a wtv file in fate suite samples? i'm downloading but it's slow here in 2001 network tech land. i accidentally deleted my samples dir. [03:54] llogan : theres some wtv samples on samples repo [03:54] i'm not sure but i think the fate samples are http browseable somewhere [10:09] 'morning [10:28] what would be the old way of doing ffmpeg -i in.flv -af aresample=first_pts=0 -y out.wav? [10:29] did we have such option in the past? [10:29] (in order to pad with silence at the beginning) [10:36] does libavfilter support the concept of 'subtitle' filtering, for example: to strip or add coloring [11:28] ffmpeg.git 03Michael Niedermayer 07master:3dae05f4f7a8: avformat/mpegts: also print PMT version in av_dlog() [12:49] BuxiNess_: https://github.com/CodecCentral/roger [12:52] The author said he tries to have an alpha version this spring [12:53] cbsrobot, many thks [14:02] pross: no [14:29] ffmpeg.git 03Benoit Fouet 07master:dc351e138190: id3v2: prefer TDRC for date over TDRL. [14:44] pross: wouldnt converting ass to srt automatically take out a lot of style ? also mplayer -dumpsrtsub :P [14:44] not even sure that works anymore [18:14] Compn: thanks. nah i want to manipulate the text and add my own colors [18:37] Hi guys [18:39] areany of the devs online this evening [18:41] nope [18:42] it is supper time in Europe [18:42] I'm in europe too lol [18:42] But okay, I'll try later [18:48] ChrisHan_: just ask your question, so people can see it and answer if/when they can. [18:49] A new version of x265 has been released today that significantly speeds up encoding. I was wondering when this will be included into FFMpeg. [18:50] Then, if I hopefully get an answer to this question, I will then ask the Handbrake devs when they will include the new FFMpeg version in their SVN builds on their IRC. [18:50] ffmpeg uses it as external library. if there are no api changes, it should just work. [18:51] So if I compiled FFMPEG with the new x265 build it should just sort of slot in and build? [18:51] x265 does releases? or they just merged some speed up changes? [18:51] they do releases yes [18:51] they do a release every 4 to 6 weeks [18:51] it was on their Facebook page today they have released 1.4 [18:51] the api has been stable [18:51] you can build ffmpeg with it [18:54] Great! thanks [18:54] ChrisHan_: I've already tested x265 1.4 in HandBrake. There is a review request currently under review for it. It is *not* significantly faster. [18:54] OH, I was under the impression it was, my fault. [18:54] j45, rd mode 4 and 5 are significantly more parallel now [18:54] when you use pmode [18:54] but thats only going to help if you have more than 10-16 cores [18:55] I'm encoding at 0.7FPS atm on Handbrake SVN [18:55] was hoping for at least 1FPS lol [18:55] is that HD? [18:55] 1080p [18:55] ok. I have 8 cores on my best machine and I've only tested a few presets [18:55] j45, its mostly useful for placebo in data centers [18:55] not normal people using handbrake [18:55] ChrisHan_: https://reviews.handbrake.fr/r/773/ [18:56] Thanks j45 [18:56] Avatar is 3 hours long and looks shitty on x264 at 3000kbps. Hence x265 [19:00] h264 3000kbps is kinda low even for 1280x720 [19:00] 6000kbps is fine for normal films. I'm squeezing them onto a 4.7GB DVD [19:01] also, unless you have a haswell CPU you wont see any real speed gain from this new x265 version [19:01] Dual core ivy bridge in laptop [19:01] last time i checked, most development was on avx2 simd [19:01] It's got a GeForce 660M but they don't have CUDA or OpenCL in it yet [19:02] then you're not going to see much difference compared to 1.3. like the guy in that handbrake review [19:02] How gutting. Guess my encode can wait a week lol [19:03] i was testing on a quad haswell laptop. but haven't really compared the slower presets much yet. [19:03] ffmpeg.git 03Michael Niedermayer 07master:e4f8a973aa8b: swresample: Fix swr_drop_output so it does not flush the buffers [19:03] ffmpeg.git 03Michael Niedermayer 07master:97da68172aa7: avfilter/af_aresample: split flushing code out [19:03] ffmpeg.git 03Michael Niedermayer 07master:09024fe681e1: avfilter/af_aresample: Limit data per inserted packet [19:03] I CBA splashing out on a bluray writer so trying to cram into regualr DVDs [19:04] anyway thanks for the help [19:43] ffmpeg.git 03Michael Niedermayer 07master:786594184a17: avformat/mpegts: fix iteration count in add_pid_to_pmt() [21:00] ffmpeg.git 03Michael Niedermayer 07master:66b9e60af0b8: ffmpeg_opt: store canvas size in decoder context [21:08] ffmpeg.git 03Vittorio Giovara 07master:351d0f8b7a6e: get_bits: remove unused assignment [21:09] ffmpeg.git 03Michael Niedermayer 07master:a28313b09e73: Merge commit '351d0f8b7a6ecce411ae75fb3511573c34317218' [21:17] ffmpeg.git 03Vittorio Giovara 07master:240b22afe14e: motion_est: remove dead code [21:17] ffmpeg.git 03Michael Niedermayer 07master:567ea2d64c63: Merge commit '240b22afe14ef477da1b439b9ed7bca6cc7d6c26' [21:37] ffmpeg.git 03Vittorio Giovara 07master:c442190a6bfd: error_resilience: initialize prev_* variables [21:38] ffmpeg.git 03Michael Niedermayer 07master:e1bcbca9987e: Merge commit 'c442190a6bfd8036f6c32b78e1e96ff3b830f8f0' [21:52] ffmpeg.git 03Vittorio Giovara 07master:f52b8717617e: celp_filters: don't use filter lenght as loop bound [21:52] ffmpeg.git 03Michael Niedermayer 07master:d9103e848dae: Merge commit 'f52b8717617e94da90a45afdfff23e94f9ecf35c' [22:00] ffmpeg.git 03Vittorio Giovara 07master:9f6f407463ff: aacsbr: treat 1-d arrays as such [22:01] ffmpeg.git 03Michael Niedermayer 07master:ce63cb4ff217: Merge commit '9f6f407463ff8b7681befd04b6655bb7c6d9b3e1' [22:31] ffmpeg.git 03Michael Niedermayer 07master:930ffd46e1e7: aacsbr: change order of operation to prevent out of array read [22:31] ffmpeg.git 03Michael Niedermayer 07master:df82125acba5: Merge commit '930ffd46e1e742674aa7cc1c2450020c63b5015b' [22:46] ffmpeg.git 03Vittorio Giovara 07master:77ab341c0c6c: aacdec: add default case in channel layout [22:46] ffmpeg.git 03Michael Niedermayer 07master:ae4bb6c4884f: Merge commit '77ab341c0c6cdf2bd437bb48d429e797d1e60da2' [23:14] can anyone tell whether the bayer pixdesc formats are correct [23:14] the 8 bit formats all say that the depth is R=2, G=4, B=2 [23:15] so that amounts to 8 bit per pixel? [23:15] instead of 8 bit per component? (with components skipped) [23:17] e.g. AV_PIX_FMT_BAYER_BGGR8 is documented as having "BGBG" components (for odd lines), with 8 bit per sample, so that means B actually has 4 bit per pixels, not 2? [23:18] hm I guess it becomes 2 bit by average [23:18] but that really seems... surprising, in context of pixdesc as generic description [23:27] (though I bet these bayer formats completely break drawutils) [00:00] --- Wed Nov 5 2014 From burek021 at gmail.com Thu Nov 6 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Thu, 6 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141105 Message-ID: <20141106010503.0FC4D18A0254@apolo.teamnet.rs> [00:44] wm: for BGGR8, odd lines have 'BGBG' components, even lines have 'GRGR' components. There are no blue on the even lines. [00:44] pross: Note that scaling fails badly: https://trac.ffmpeg.org/ticket/4053 [00:45] noted. why on earth would you want to scale bayer! [00:47] but i agree it should not crash. [00:49] If it refuses to scale, FFmpeg will happily choose yuv420p, if it crashes, this is difficult... [00:49] But as said, I still don't understand how to get useful output at all;-( [00:49] Do you need samples? [00:52] I have two samples with two output tiff files: How can I produce similar looking output? [00:58] cehoyos: thats the white balance issue right? [00:59] bayer data is (usually) unprocessed: it is the raw data returned by the image sensor. [01:03] I guess it is the white balance issue: I think we should document how to get useful output from Phantom Cine input (if this cannot be done automatically). [01:38] ffmpeg.git 03Carl Eugen Hoyos 07master:f2ffaae9ac93: Use -fno-optimize-sibling-calls on parisc also for gcc 4.9. [02:09] ffmpeg.git 03Michael Niedermayer 07master:77f1199e8fd9: avcodec/mpeg12dec: do not trust AVCodecContext input dimensions [02:17] ffmpeg.git 03R?mi Denis-Courmont 07master:4ff670d99beb: hwaccel: Deinitialize hardware acceleration early enough [02:17] ffmpeg.git 03Michael Niedermayer 07master:43fee7ad920d: Merge commit '4ff670d99bebd97429322719089363d83143477d' [02:26] ffmpeg.git 03Lou Logan 07master:37425fcb0474: doc: clarify -frames options behavior [09:58] ffmpeg.git 03Michael Niedermayer 07release/2.4:cd57d608a412: avcodec/mpeg12dec: do not trust AVCodecContext input dimensions [09:58] ffmpeg.git 03Michael Niedermayer 07release/2.4:f9ca1fecb0fd: ffmpeg_opt: store canvas size in decoder context [10:32] ffmpeg.git 03Michael Niedermayer 07release/2.4:a5cc8775cf1d: avcodec/h264_sei: ff_h264_decode_sei: dont try to parse trailing zeroes [10:32] ffmpeg.git 03Carl Eugen Hoyos 07release/2.4:9798dc8061c9: Use -fno-optimize-sibling-calls on parisc also for gcc 4.9. [10:34] ffmpeg.git 03Martin Storsj? 07master:a490391157dc: rtmpproto: Ignore errors from the getStreamLength method [10:34] ffmpeg.git 03Michael Niedermayer 07master:513d57cc4d33: Merge commit 'a490391157dcf4dc6b65352ec3eea2781dd0a404' [11:25] ffmpeg.git 03Thomas Volkert 07master:07c3a4f69336: avformat/udp: UDP-Lite (RFC 3828) support added [18:11] ffmpeg.git 03Kevin Mitchell 07master:fdf22f973d41: avfilter/vf_idet: add a repeated field detection [19:00] ffmpeg.git 03Michael Niedermayer 07master:a604de4fd89e: avutil/time_internal: do not attempt to override *time_r() macros [19:43] ffmpeg.git 03Andrey Utkin 07master:4fd70679877b: v4l2: support MPEG4 compressed streams [20:08] ffmpeg.git 03Reimar D?ffinger 07master:a26d0ffa18db: vdpau_mpeg4: Do not fail on unknown profile. [21:05] i think i will still keep a N dialogues per subtitles [21:05] it can be useful in some situations [21:06] typically if you have a markup that can't be expressed in a single dialog [21:06] any objection? [21:11] ffmpeg.git 03Nicolas George 07master:90cdec5e2698: ffmpeg: init sub2video.last_pts. [21:27] "Consider using a tool like VirtualDub or avidemux to fix it." [21:27] ffmpeg can't do it? [21:29] at least the ydidnt say nandub [21:35] maybe we should mention mencoder [21:40] fix it with dd and a hexeditor [23:58] What's the usual command to run to create configure script for an autotools project? ./autogen.sh? [23:59] if there is an autogen.sh then yeah [23:59] in most projects you can use "autoreconf" [00:00] --- Thu Nov 6 2014 From burek021 at gmail.com Thu Nov 6 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Thu, 6 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141105 Message-ID: <20141106010501.F318C18A0252@apolo.teamnet.rs> [02:27] I'm trying to copy an RTSP stream but getting an error: http://pastebin.com/kNFELARn [02:35] danomite: doesn't look like it likes your audio track? I could easily be wrong though, try to re-encode just the audio to aac to ac3 or something and see what happens... [04:01] hrm, so git ffmpeg crashes trying to encode some stuff using aac.... do I just post a bug on trac? I got the debug backtrace [10:23] Action: DelphiWorld ugh emilsedgh [10:38] I am using this as my webm encode: "ffmpeg -i desktop-src.avi -c:v libvpx -quality best -cpu-used 0 -qmin 0 -qmax 62 -minrate 1M -maxrate 1M -b:v 1M -c:a libvorbis -vf scale=1280:720 -threads 6 desktop.webm" [10:38] But it changes the color of my source [10:38] How can I preserve that [10:57] Can you pastebin the output from ffmpeg as well as upload a picture showing the color corruption? [10:59] Yea sure one second please [11:06] c_14: http://pastebin.com/2ahYsfv0 / http://i.imgur.com/1lUufEP.jpg?1 [11:07] That banding is killing me too [11:14] Hmm [11:16] First of all, qmin and qmax shouldn't do anything because you're using constant bitrate. Have you tried increasing the bitrate? [11:17] Does minrate and maxrate take effect if using b:v also? [11:17] I figure 1M is pretty good [11:17] I can try 2M [11:17] And how about -quality best [11:17] Does it make a difference [11:20] Increasing -minrate 1.5m -maxrate 2m -b:v 3m seems to make quality worse [11:26] d3m0n: It could be the color conversion. Try, `ffmpeg -i desktop-src.avi -t 10 -pix_fmt yuv420p -c:v ffvhuff output.avi` and see if it looks any better. [11:33] Is there a command I can use to make it near lossless and degrade from there [11:34] ffvhuff is lossless, except for the color conversion done with -pix_fmt yuv420p [11:35] Yeah that kept my color! But I am hoping for webm D: [11:35] Did you add -pix_fmt yuv420p? [11:35] I just used your full command [11:37] oh, well then we can assume it's not the color conversion. Did you see https://trac.ffmpeg.org/wiki/Capture/Desktop and https://trac.ffmpeg.org/wiki/Encode/VP8 ? [11:37] Well, when I used your command it kept my original color. [11:37] ANd yeah I've been trying to use the examples from that wiki [11:38] This was my command I came up with from that command " ffmpeg -i desktop-src.avi -c:v libvpx -quality best -cpu-used 0 -qmin 0 -qmax 62 -minrate 1M -maxrate 1M -b:v 1M -c:a libvorbis -vf scale=1280:720 -threads 6 desktop.webm" [11:38] But as c_14 mention I removed the qmin/qmax [11:39] right, omit those and try -crf [11:40] including b:v/maxrate/min? [11:41] no, try -crf 10 [11:44] encoding now [11:46] "ffmpeg -i desktop-src.abi -c:v libvps -crf 10 -c:a libvorbis -threars 6 desktop.webm" produced quite a poor quality [11:46] Looks like it kept the right color though, but I can see some shaded flickering [11:48] -crf 5 -b:v 3M [11:51] Haha yea, this is driving me nut. The quality is still seems fairly poor. The banding on the colors is bad [11:55] Maybe I am just expecting more than I should [12:35] hi all [12:35] people I have problem wit compile my application which is using ffmpeg I got following error: libavcodec.a(tiff.o): undefined reference to symbol 'lzma_code@@XZ_5.0' [12:36] I have linked m and z (-lz -lm) [12:45] Vardan: it's liblzma [12:45] so try adding -llzma [12:53] cool thanks :) [12:56] I'm trying to understand one thing, I'm using ffmpeg and linked ffmpeg libs, now as I understand g++ what that I also link ffmpeg related libs too, is it possible to say g++ to link all ffmpeg related libs? [13:01] g++ is used for C++, ffmpeg and libs are C only. Ignoring that, gcc/g++ can be used to call the linker and do the linking. [13:30] hi, I'm having an issue with ffmpeg on android, some rtsp streams are very choppy [13:31] seems it's caused by androids bionic c library does not support pthread_cancel [13:32] in libavformat/udp.c it seems this is needed when circular buffers are used, which seem to be the case, but I'm not sure [13:32] can anyone point me in the right direction as to how to debug this? [14:52] hi, I fail to get nv12 conversion working. http://pastebin.com/sr5ZAVPb Any clue? [14:53] avconv -pix_fmts |grep nv12 lists the nv12 format as available for both in and out. [14:55] avconv is part of libav, see #libav for help or use ffmpeg from FFmpeg [14:56] c_14: okay, thanks, I wasn't aware of the difference. I'll give ffmpeg a go. [14:56] Be aware if you're under ubuntu and non-testing debian the standard ffmpeg binary/package is actually also libav. [14:57] What you can do is compile from source or use a static build. [14:57] jonand: add -f rawvideo as input option [15:00] ubitux: that solved my problem, thanks! [16:09] relaxed: did you update the archive for ffmpeg-2.4.2-64bit-static.tar.xz? [16:09] i computed a sha1sum on it a few days ago but now they don't match [16:29] Hey [16:29] I need help making a video from images [16:29] I am doing this now [16:29] ffmpeg -framerate 1/5 -i img%03d.png -c:v libx264 -vf fps=25 -pix_fmt yuv420p out.mp4 [16:30] and what happens is that the first image is shown for 2 seconds [16:30] then it flashes the next image and the video stops [16:30] What I want to happen is for it to show each image for 2 seconds in a movie [16:30] Hello I'm new to FFmpeg and I'm trying to create a point to point connection from one to another computer. The host is working with Debian and the client with Ubuntu. I've tried to connect with a normal UDP connection and it worked but with a delay of about 5 or 6 seconds. The client recieved the streaming with VLC but now I'm trying to decrease the delay. How I'm going to do this? I also tried it with rtp but I'm not really getting it working. Hope you can help [16:30] so if there are 2 images, the movie is 4 seconds long [16:31] Any help would be greatly appreciated [16:32] joon: ffmpeg -i img%03d.png -c:v libx264 -r 0.5 -pix_fmt yuv420p out.mp4 [16:33] Blippy! Thank you so much!! [16:33] You are awesome! [16:34] no problem [16:34] does ffmpeg support animations [16:34] on these images? [16:34] you can improve your quality by adding a -crf 17 to it [16:34] what do you mean by that? [16:34] like to slide the image in from the left [16:34] or zoom in on an image [16:35] Schipper, ibex1101: http://www.irchelp.org/irchelp/irctutorial.html [16:35] *slowly zoom in [16:35] yes look at the man pages: man ffmpeg-filters [16:35] sweet [16:35] Thanks again [16:35] glad i could help [16:36] Hello I'm new to FFmpeg and I'm trying to create a point to point connection from one to another computer. The host is working with Debian and the client with Ubuntu. I've tried to connect with a normal UDP connection and it worked but with a delay of about 5 or 6 seconds. The client recieved the streaming with VLC but now I'm trying to decrease the delay. How I'm going to do this? I also tried it with rtp but I'm not really getting it working. Hope you can help [16:41] Hello is someone there? [16:42] ibex1101: be patient anyone who can answer your question and has the time to will likely be glad to help you. [18:56] hello, is there something that can be done to this: Warning: data is not aligned! This can lead to a speedloss ? it started after i added ,scale: to my command [18:57] both the cropping resolutions and final resolutions are divisible by 16 [18:57] I don't think it's the resolution, but rather the fact that the memory isn't aligned to specific alignments that let people do certain things faster [18:58] what kind of input are you using because on all systems nowadays what lavc and friends do internally is mostly allocated with memaligned malloc [18:59] it had no errors with only cropping, adding scaling gave the warning [18:59] its analog av in to x264 + aac audio [19:02] blippyp, thanks! [19:04] varikonniemi, well you probably had the nonalignment all the time then, but only the video filters check it [19:04] *check for it [19:04] and analog av really sounds like it [19:04] since not lavc or whatever would be allocating the memory in that case [19:05] the crop filter did not say anything, so it must be only the scale filter that notices it? [19:06] probably, or the crop filter introduces the lack of alignment by just moving pointers and some other structures around [19:07] that sounds plausible [19:11] Is there an option to output to a different file every half hour? [19:12] look at the segment muxer [19:13] thanks [19:30] <[twisti]> would it be ok to ask a generic video encoding question ? im encoding a bluray file, and im having trouble making an educated choice about bitrate/file size. i dont really know much about video encoding, does anyone know if there is some sort of primer or tutorial or guide about bitrates and image quality ? [19:33] if you are using x264 it's all rather simple [19:33] (because it provides the simplest tools to achieve certain things) [19:33] it has a rate control mode called CRF that is the closest thing we have to "constant quality" right now [19:34] and it has a simple-to-use preset system to set your preference between speed and more compression efficiency [19:35] <[twisti]> im more asking about what they mean [19:35] also I hope you mean your input is a blu-ray transport stream because if you are actually encoding something meant to go on a blu-ray you are much more limited (so in a way you just don't have any alternatives regarding what to do), and ffmpeg can't output you compliant files out of the box (unless you use raw H.264 streams) [19:36] <[twisti]> no, no, its something i copied and pasted off a BD [19:36] [twisti], and I'm asking if you are using x264 to encode, because then I could just give the steps to find the balance for a type of source you have on hand [19:36] <[twisti]> my goal is to reduce file size ideally without losing quality (although i understand that any encoding will lead to SOME amount of quality loss) [19:36] because there are no "written on the wall" bit rates [19:36] <[twisti]> yes, im using x264 [19:36] ok [19:37] <[twisti]> if it helps, im trying to encode a tv show, 43-ish minutes per episode [19:37] now, first of all limit your encode to a few thousand frames at first of content that your source is mostly of (-t and -ss will probably be of use) [19:37] then start with just -preset medium (this is the default) and -crf 23 (this is actually too) [19:37] then encode it, and see if it looks good [19:37] if it does, raise the crf value [19:38] this will make the encoder use higher quantizers, and effectively make a smaller file (and theoretically, worse quality) [19:38] if it looks bad, then you lower the value [19:38] after a couple of tries you should have found the highest crf value that still looks good to you, while gives you good compression [19:38] <[twisti]> im having trouble 'telling' the quality with 'just my eyes', i know that sounds dumb, but i just cant tell right away [19:39] that means you are way too low on the CRF scale :) [19:39] and you will get used to it [19:39] <[twisti]> so higher crf = worse quality ? [19:39] rather call it higher compression (unrelated to compression efficiency) [19:39] because it not necessarily looks worse [19:40] after finding your CRF value of choice, you then start poking the presets [19:40] and pick the slowest that is still fast enough for your use case [19:41] now, the CRF *is* affected by such things as the video's frame rate, resolution and the internal settings (poked by the presets), so you might have to adjust the CRF a bit after finding the preset matching your needs, but generally these changes are not big [19:42] for example if you encode the exactly same clip with the same crf value with, say, preset slow and preset placebo, the placebo one will end up most probably a bit bigger [19:42] this is due to the encoder "seeing" more things, and you would have to adjust the crf value to compensate if you really care [19:43] is it possible to have a copy stream playable immediately ? [19:43] on the other hand if you compare one of the faster presets and a slower preset, then the compression efficiency gets better so much that the slower preset's output will most probably end up quite a bit smaller [19:44] anyways, point being that settings and content affect the CRF so you might have to adjust it depending on various things :) [19:44] danomite: mostly depends on the output format/muxer. e.g. .nut should do it. [19:45] <[twisti]> life was easier back when i could just click a torrent and have things :( [19:45] <[twisti]> thanks for your input [19:45] <[twisti]> seems like i have some trial and error ahead of me [19:45] that said, if you encode a single TV series with similar content then a single CRF value should be just fine for it if you are encoding it at the same resolution, after you have found your preferred preset and CRF value [19:45] I need the output to be mp4 guess i'll just wait to be a segment behind [19:45] yeah, first pick the crf value, then tweak the preset and finally do some final tweaking of the crf in case you feel you need such [19:46] [twisti], now imagine how hard this was when we didn't have something as useful as CRF or such a simple system to set different compression efficiency/encoding speed ratios [19:47] pulling bit rates out of your arse and then trying to tweak those [19:47] and then having to look at all the settings in an encoder and trying to find a good spot with compression efficiency and speed [19:55] greetings :) I have a legacy real media RV40 file that no media player I've tried on linux can play successfully. Would someone here maybe be kind enough to just have a short look on whether this is a serious problem? The file is at http://www.datafilehost.com/d/63a8ce3f (2.2MB), 57 seconds of black and no audio [19:58] debianuser: oh, with mplayer these files actually seem to play... I tried mpv, assuming it would be the same :-/ [20:01] but audio doesn't work, ffcook seems to have problems :-/ [20:03] <[twisti]> JEEB: once i picked CRF, preset will only affect file size and encoding duration, not quality ? [20:05] [twisti], "the CRF *is* affected by such things as the video's frame rate, resolution and the internal settings (poked by the presets), so you might have to adjust the CRF a bit after finding the preset matching your needs, but generally these changes are not big" [20:05] as I quote myself [20:06] the CRF depends on the internals that are switched around by the presets, but that tweaking is generally small, if any is needed [20:06] <[twisti]> just making sure i understood correctly [20:08] basically since the crf value's results can differ between presets due to the internal behavior changing, you might need to tweak the crf value afterwards a bit. but yes, preset controls compression efficiency [20:09] so in theory the most optimal way is to first pick the preset and then the CRF, but if you are OK with slow encodes then even testing the CRF values becomes slow :P Thus it's better to get a feel with the medium preset (the default) first, and then tweak if tweaks are needed. [20:11] Is there an ffmpeg feed and ffserver example for streaming mp4 [20:53] hi. is it true that subtitle conversion (hdmv_pgs_subtitle (pgssub) -> mov_text (native)) is not possible with ffmpeg? is it from bitmaps to text? [20:55] Yes, it's bitmap to text. [20:55] You'd need an OCR program. [20:56] ok thanks. i guess what i could try is to extract the subtitles into a separate file and use it if the player supports [21:01] sacarasc: any suggestions if there are standalone bitmap subtitle formats? [21:02] my problem originally is, i have some matroska files with dual-audio, dual-subtitle stuff (from bluray). they're too slow to play on raspberry pi for some reason. without subtitles no problem [21:02] SubRip and SubtitleEdit will apparently do it on Windows... Dunno about any other OS. [21:04] i guess i could try without ocr'ing the subtitles. ffmpeg just doesn't seem to output this format [21:04] maybe there's no standalone subtitle format for this [22:32] I can use the -ss option before the input to modify audio, but it has no effect when I use it before video. Why? (Log: http://pastebin.com/7tshimna) [22:34] ObsequiousNewt: first off, you're using libav; not ffmpeg - does it behave any differently when you use ffmpeg? [22:34] I'm on ubuntu; I can't install it so easily [22:34] it's still not deployed on ubuntu? [22:35] no, it's only in debian unstable [22:35] it's in debian/unstable nowadays [22:35] Nope. [22:35] mmh i wonder how long it will take until they sync [22:35] ubitux: won't be until *at least* the next ubuntu release. [22:35] so april next year at the earliest [22:35] well, they don't have an unstable as well? [22:36] *sigh* Okay, I'll try installing it. [22:41] ObsequiousNewt: you can request for support on #libav for avconv [22:41] Ah, thank you [22:43] i think i already know the answer to this, but stupid questions will be : Does ffmpeg support an kind of screenrecording on macintosh? like x11grab? [22:47] I'm trying to do faststart in ffmpeg but I keep getting a moov atom error: http://pastie.org/private/llufzyknaz8mfiqdz045w [22:48] Does it work if you do it to a file? Also, don't you need to do segments when streaming MP4? [22:50] sacarasc, my goal is to stream mp4, let me try with a file [22:50] danomite: the '-movflags faststart' requires rewriting the file after encoding is complete; It can't be done streaming [22:50] (mp4 is inherently unsuitable for any sort of live streaming) [22:51] I need to play a live stream in a browser what's a good format? [22:51] danomite: I think http://ffmpeg.org/ffmpeg-all.html#segment_002c-stream_005fsegment_002c-ssegment might work for MP4 streaming... [22:52] at the moment? it has to be a segmented format, like HLS or DASH, browser support varies [22:52] or you could do webrtc, I suppose [22:52] or there's always flash with rtmp. [22:54] segmenting sounds like a can of worms, would it be easy enough to to transcode the format to flv and use flash? [22:54] you'd have to run a flash media server to handle doing the rtmp; then you'd have ffmpeg send the stream via rtmp to the flash media server which rebroadcasts it. [22:56] keep in mind that you only need this if you're doing stuff live; if you don't need live then just pre-process the file before uploading it and everything will be happy. [22:57] Live is key, i'm not married to any format just want it to be easy to setup and play [22:57] I don't want to re- serve it again after ffmpeg [22:57] err ffserver [23:01] btw encoding to file works [00:00] --- Thu Nov 6 2014 From burek021 at gmail.com Fri Nov 7 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Fri, 7 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141106 Message-ID: <20141107010502.942502AD60D6@apolo.teamnet.rs> [00:19] ffmpeg.git 03Michael Niedermayer 07master:817663897e59: avcodec/mpegaudio_parser: fix off by 1 error in bitrate calculation [01:20] ffmpeg.git 03Michael Niedermayer 07release/2.2:e25e0903ab60: avcodec/mpegaudio_parser: fix off by 1 error in bitrate calculation [01:21] ffmpeg.git 03Michael Niedermayer 07release/2.3:a0605792c2d4: avcodec/mpegaudio_parser: fix off by 1 error in bitrate calculation [01:21] ffmpeg.git 03Michael Niedermayer 07release/2.4:c7b64a904ae3: avcodec/mpegaudio_parser: fix off by 1 error in bitrate calculation [10:57] ffmpeg.git 03Anton Khirnov 07master:874792641ea4: vorbis_parser: use a dedicated AVClass for logging [10:58] ffmpeg.git 03Michael Niedermayer 07master:a0617025dd26: Merge commit '8747926' [11:16] ffmpeg.git 03Anton Khirnov 07master:6896f95b2483: vorbis_parser: add an AV prefix to VorbisParseContext [11:16] ffmpeg.git 03Michael Niedermayer 07master:5f7887ca8d05: Merge commit '6896f95b2483e52e717e2c75a4fd24fcb0e14b67' [11:39] ffmpeg.git 03Anton Khirnov 07master:5e80fb7ff226: lavc: add a public API for parsing vorbis packets. [11:39] ffmpeg.git 03Michael Niedermayer 07master:44fa2671e0db: Merge commit '5e80fb7ff226f136dbcf3fed00a2966bf8e9bd70' [13:08] ffmpeg.git 03Anton Khirnov 07master:2f3fadfbe3c6: lavc,lavf: switch to the new vorbis parse API [13:08] ffmpeg.git 03Michael Niedermayer 07master:7ffdc7bef2f9: avcodec/export av_vorbis_parse_frame_flags() [13:08] ffmpeg.git 03Michael Niedermayer 07master:f74be3669daa: Merge commit '2f3fadfbe3c6ad52fad5c614b6067c5401227959' [13:08] ffmpeg.git 03Michael Niedermayer 07master:4b2763cd1354: avformat/oggparsevorbis: return proper error code from vorbis_header() [13:09] ffmpeg.git 03Michael Niedermayer 07master:c04c43b3e423: avformat/oggparsevorbis: Check that initialization succeeded before declaring the end of headers [13:22] ffmpeg.git 03Anton Khirnov 07master:91e8d2eb1f7b: lavf: use the format context strict_std_compliance instead of the codec one [13:22] ffmpeg.git 03Michael Niedermayer 07master:042eba52a5ff: Merge commit '91e8d2eb1f7bf3af949008b106ec1ca037b88b0e' [13:41] ffmpeg.git 03Anton Khirnov 07master:56dc46a18932: riffenc: do not fall back on AVCodecContext.frame_size for MP3 [13:42] ffmpeg.git 03Michael Niedermayer 07master:a51eb6d34ced: Merge commit '56dc46a1893251e74be1ad63e54fb38d754bb1fe' [13:53] ffmpeg.git 03Anton Khirnov 07master:7784f47762d5: lavf: stop using avpriv_flac_parse_streaminfo() [13:53] ffmpeg.git 03Michael Niedermayer 07master:b6a99563962c: Merge commit '7784f47762d59e859b4d0f74b3e021ad9368ee2c' [14:09] ffmpeg.git 03Anton Khirnov 07master:c070a8751597: lavc: make avpriv_flac_parse_streaminfo() private on the next bump [14:10] ffmpeg.git 03Michael Niedermayer 07master:cfef947f7f1f: Merge commit 'c070a8751597e3aa1b443e88464da785d8966b14' [14:17] ffmpeg.git 03Anton Khirnov 07master:e839de0f8515: oggenc: accept only STREAMINFO extradata [14:18] ffmpeg.git 03Michael Niedermayer 07master:94fe404c25f9: Merge commit 'e839de0f851535b5e19256b52f9865f0cb768a7c' [14:23] ffmpeg.git 03Anton Khirnov 07master:acc897e6b157: lavc: make avpriv_flac_is_extradata_valid() private on the next bump [14:23] ffmpeg.git 03Michael Niedermayer 07master:cffd2713e9bb: Merge commit 'acc897e6b15776ed438b88ffe330ec48f6b50e48' [14:37] ffmpeg.git 03Anton Khirnov 07master:05e59135b353: nutdec: do not set has_b_frames [14:37] ffmpeg.git 03Michael Niedermayer 07master:0de64082a967: Merge commit '05e59135b3539062465b5005b6d46ec0418a5fc4' [15:06] ffmpeg.git 03Michael Niedermayer 07master:e591608753ed: avcodec/xiph: make extradata argument const [15:06] ffmpeg.git 03Michael Niedermayer 07master:470e116e3fda: avformat/flvenc: Use AVFormatContext strict_std_compliance instead of AVCodecContext [15:06] ffmpeg.git 03Michael Niedermayer 07master:da8cb1c36134: avformat/mmf: Use AVFormatContext strict_std_compliance instead of AVCodecContext [15:06] ffmpeg.git 03Michael Niedermayer 07master:d5999b7f282a: avformat/yuv4mpegenc: Use AVFormatContext strict_std_compliance instead of AVCodecContext [15:06] ffmpeg.git 03Michael Niedermayer 07master:8afaa03c53d6: avformat/riffenc: move MP3 LSF threshold to the midway point between the 2 [15:54] Ticket #4089 is spam- someone needs to ban this person. [15:56] ffmpeg.git 03Giorgio Vazzana 07master:be0356ca0537: lavd/v4l2: use pixel format variable names consistently [16:59] ffmpeg.git 03Michael Niedermayer 07master:d70e503ebcd4: avfilter/af_aresample: remove unused variable [17:19] ffmpeg.git 03Vittorio Giovara 07master:e0a1d0a2b04e: mpegvideo_enc: rework direct mode check [17:19] ffmpeg.git 03Michael Niedermayer 07master:5241f9005814: Merge commit 'e0a1d0a2b04eb5220d00fc7ce46a57cc5e3c7118' [17:28] ffmpeg.git 03Vittorio Giovara 07master:0a6664706168: mpegvideo_enc: factor out denominator and explicitly cast operands [17:28] ffmpeg.git 03Michael Niedermayer 07master:c96a44d0ddb0: Merge commit '0a6664706168dc1049967bec311970d720581625' [17:34] ffmpeg.git 03Vittorio Giovara 07master:37b3361e7553: mpeg12enc: factor out check in encode_dc [17:34] ffmpeg.git 03Michael Niedermayer 07master:98d6f0ffef30: Merge commit '37b3361e755361d4ff14a2973df001c0140d98d6' [18:16] ffmpeg.git 03Vittorio Giovara 07master:5d29efe4b015: mpeg12dec: simplify context duplication [18:17] ffmpeg.git 03Michael Niedermayer 07master:f663734738c0: Merge commit '5d29efe4b0154ce305d66fed2ac23e5842439256' [18:23] ffmpeg.git 03Luca Barbato 07master:ac4a5e3abd8a: pthreads_frame: Do not leak on failure path [18:25] ffmpeg.git 03Michael Niedermayer 07master:ad0cf8e063b8: Merge commit 'ac4a5e3abd8a022ab32245ad527ffc37eabab8b1' [18:43] ffmpeg.git 03Vittorio Giovara 07master:199d9f995da5: mjpegdec: fix undefined shift [18:43] ffmpeg.git 03Michael Niedermayer 07master:c53a1507aa09: Merge commit '199d9f995da53fe2507821f6d96bbc692574e1a9' [23:31] ffmpeg.git 03Thilo Borgmann 07master:93ab6693d8cf: lavd/avfoundation: Update documentation to mention audio capabilities. [00:00] --- Fri Nov 7 2014 From burek021 at gmail.com Fri Nov 7 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Fri, 7 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141106 Message-ID: <20141107010501.8EBBF2AD60D4@apolo.teamnet.rs> [00:13] Hello. I am trying to get a background video to show up in Safari, but Im having trouble encoding it correctly. Im trying to match the codecs to a video I know works. The video that works has code h264, my code says h264,AAC. Im wondering how I can get rid of the AAC part. [00:13] -an [00:14] I'm converting live rtsp mp4 stream to ogg/theora stream getting this error a bunch: http://dpaste.com/0M1D1W5 [00:14] That should just be a warning. [00:15] wait [00:15] any way to eliminate it? [00:15] The source is mp4? [00:15] yes [00:15] What's your commandline? [00:16] http://dpaste.com/1JMPZAF [00:17] Does it give you the error if you output to a file^? [00:17] -? [00:17] -^ [00:17] +? [00:18] making a file now [00:19] c_14: okay, that got rid of the AAC, but the video is still not rendering properly in Safar. The working video has a color profile of SD (6-1-6). How can I add that to my mp4? [00:20] Can you pastebin the output of ffprobe on both files? [00:20] c_14: sure [00:21] c_14, no errors [00:21] danomite: what's your ffserver.conf? [00:22] c_14: https://gist.github.com/gabeodess/01e11b8b8070d5cdcc35 [00:23] http://dpaste.com/34E6CE5 [00:23] gabeodess: add -profile main -pix_fmt yuv420p [00:24] danomite: you have NoAudio twice, shouldn't break anything but it's a noop. ditto global_header, Are you sure you want your framerate to be 1? [00:25] The global header is in there because i was trying mp4 streaming, but I'll remove the doubles for both. My source frame rate is 1 and that's intentional [00:25] Maybe get rid of that option completely (the global header one). [00:26] c_14: Works like a charm :) thanks! [00:27] If that doesn't work, try getting rid of preroll and/or startsendonkey. Those three options touch keyframes which could be causing the warning. [00:28] removed global header completely and duplicate no audio, giving this a test run [00:29] gabeodess: add "-movflags +faststart" too if you haven't yet [00:29] llogan: what will that do? [00:29] Does anyone know whether the QuickTime File Format specification means signed or unsigned when it just says "32-bit integer"? [00:29] move some shit around in the file so it can begin playback before it is completely downloaded [00:30] llogan: perfect, thanks! [00:30] show your command and console output if you want more suggestions/flames. [00:32] llogan: https://gist.github.com/gabeodess/b6b0729decd600fc3916 [00:32] ffplay has an error: http://dpaste.com/12QDZRR [00:33] llogan: also, its a silent clip, so I dont know if I should remove the audo codec or what to slim it down. [00:33] gabeodess: -an removes the audio, -acodec is a no-op [00:34] c_14: ah. cool. [00:35] danomite: is ffserver running and accepting content [00:35] gabeodess: "Please use -profile:a or -profile:v, -profile is ambiguous". probably doesn't matter if you have no audio [00:36] llogan: if it doesnt matter should I just remove the flag? [00:37] does your device or borwser require Main profile? [00:37] llogan: I dont know what that means [00:37] then just leave it or perform tests. instead of -s 640x360 you can be lazy and do "-vf scale=640:-2" [00:38] llogan: whats the benefit there? [00:38] ffmpeg will calculate the value for you [00:38] llogan: ah, to retain aspect ratio? [00:38] yes [00:39] cool [00:39] how did you come up with the value for -vb? [00:40] llogan: honestly, I just copied this off a blog. [00:40] llogan: I really dont know what Im doing [00:41] remove the -vb and its value. the defaults may suffice here. [00:41] c_14 yes [00:41] danomite: what does curl say? [00:41] llogan: note also, that speed is not a convern for me. This is a one time conversion that Im doing. [00:42] then you can add "-preset veryslow" [00:43] c_14 curl says nothing [00:43] llogan: what will that do& besides make it very slow [00:43] https://trac.ffmpeg.org/wiki/Encode/H.264 [00:44] danomite: no output from ffserver/ffmpeg ? [00:44] there is will capture [00:44] llogan: ah, cool. Why not use placebo? [00:44] it's a waste of time. [00:45] c_14 real quick when I added a global header line, ffplay started to work [00:45] llogan: oh yeah& it says that in the next sentence [00:45] but continues with the warnings [00:45] danomite: tried getting rid of the other two options I mentioned earlier? [00:46] just felt another earthquake... [00:46] or a fat kid was jumping. old building. [00:47] VideoFrameRate is intentional but haven't removed it, NoAudio is only done once now [00:47] danomite: I meant the other two. [00:47] Startonkey or whatever it was [00:47] preroll/startsendonkey [00:48] ah [00:48] trying with those 2 commented [00:49] c_14 same warnings [00:50] And without global_header as well? [00:50] Oh, and if that doesn't work I'm out of ideas. [00:50] With 1 global headers, trying without now [00:51] llogan: thanks for all the suggestions [00:51] Hello, i have a question. I'm so confused with the licenses that i don't really know what to do in my case. I have make my own COMMERCIAL project in PHP and i'm using FFmpeg. I have compiled it statically and i'm sharing it to my clients. However i provide the static build for FREE without paying. What i have to do more to be 100% ok with this? [00:51] ffplay refuses to play, Invalid data error [00:51] danomite: Then I suggest you try living with the warnings as long as they don't explicitly break things... [00:52] k, then the next question is the video quality drops out periodically [00:53] BlackDream: did you see this? http://ffmpeg.org/legal.html [00:54] Yes, but it's a lot confusing for me [00:55] how did you compile ffmpeg? i mean which configure options did you use? [00:55] Actually it says that i can't use the patented algorithms that FFmpeg has, i dont even know for which algorithms they are talking about [00:55] w8 to find it [00:56] Here's the Feed errors: http://dpaste.com/3E27N0E [00:56] ./configure --disable-debug --disable-shared --enable-static --extra-cflags=--static --disable-ffplay --disable-ffserver --disable-doc --enable-openssl --enable-gpl --enable-pthreads --enable-postproc --enable-libass --enable-gray --enable-runtime-cpudetect --enable-libfaac --enable-gnutls --enable-libmp3lame --enable-libopus --enable-libtheora --enable-libvorbis --enable-libx264 --enable-libxvid [00:56] --enable-bzlib --enable-zlib --enable-nonfree --enable-version3 --enable-libvpx --enable-libxavs --disable-devices --enable-librtmp --extra-libs='-lrtmp -lnettle -lhogweed -lgmp -lssl -lcrypto -lz -lc -ldl' [00:56] note that you should use a pastebin service for stuff that is longer than one line [00:57] :( [00:57] BlackDream: you have "--enable-nonfree" there, that build probably cannot legally be redistributed. [00:57] What impact will disable non free will have? What it is included there? [00:58] in your case the nonfree is needed for libfaac support [00:58] all the nonfree flag does is allow you to build ffmpeg with combinations of options that have conflicting licenses. [00:58] libfaac isnt the same as aac right? i can still use aac? [00:59] If so, this is the only real problem? Is the libx264 ok? [00:59] libfaac is an encoder. ffmpeg can support 4 AAC-LC encoders. the native aac encoder isn't the best, but it really isn't too terrible with enough bits [00:59] if you disable libfaac, you can still use the built-in aac encoder. libx264 is lgpl2, I think? that should be ok, provided you take care of finding out if you need patent licensing, etc (get a lawyer if unsure). [01:01] and one question. What happens if they found out that i am using the non-free codecs and i distribute them? Just curius. I am not telling that i want to be illegal. Just what they can do. If the libfaac is onyl the real problem here i'm going to remove it [01:02] libx264 requires --enable-gpl which results in GPL 2, so you will have to comply with that license [01:02] well, if you build with a combination of options that has compatible licenses, and abide by the terms of the licenses (such as making source code available), then none of the open-source developers will go after you. [01:02] It also says in the ffmpeg legal that all the patented algorithms are depending on where i live [01:03] license compliance and patent issues are two different things. honestly, the FFmpeg project only cares if you comply with the license. [01:05] you also have --enable-version3 in there, but i don't think any of your other enabled external libraries need that [01:06] and --enable-librtmp is probably just extra baggage since ffmpeg has native rtmp support [01:07] Anyway the real problem here should be the enable-nonfree i think. I have to find out what impact will have other than the libfaac [01:09] nothing else in your current configure requires that [01:09] BlackDream: you won't be able to distribute the binary if you build it with non-free [01:09] and libfaac isn't really that good anyway [01:10] OK So perfect [01:10] i will remove it [01:12] Do i have to release the source code of ffmpeg? I didn't change anything on it. Should i make the binary public for free even if they dont get my software? [01:13] If someone requests it you have to provide it [01:13] So i will have it for free for everyone [01:17] And when exactly i wil notice the difference between libfaac and aac? [01:23] BlackDream: you'll just have to compare them yourself [01:23] not that you can use libfaac anyway [01:57] hi [01:57] i have a .mov file [01:57] i nee to rotate it 180? and reseize [01:58] to 1024*... [01:58] how can i do? [02:00] wallbroken: please use a pastebin service to show some info about the input: ffmpeg -i input.mov [02:00] show the complete console output [02:05] llogan: http://pastebin.com/NdWkAT9w [02:10] wallbroken: ffmpeg -i input -vf "hflip,vflip,scale=1024:-2" -metadata:s:v rotate="" -c:a copy output.mov [02:10] or you could use transpose or rotate instead of hflip,vflip [02:10] what -2 is? [02:11] this command re-encodes the video? [02:11] -2 read https://ffmpeg.org/ffmpeg-filters.html#scale-1 see width and height options [02:11] yes, filtering requires re-encoding [02:12] if you just want to get rid of the rotation metadata: ffmpeg -i in -metadata:s:v rotate="" -c copy out.mov [02:12] and where compression algorighm is defined? [02:12] H.264 will be video format by default for .mov output (if your ffmpeg has been configured with --enable-libx264) [02:13] i want to re-encode the video according to metadata orientation [02:13] that is what my example should do [02:13] and then removes the metadata orientation fields? [02:13] if not, some player will be rotate it again [02:13] there is no "automatically rotate based on rotation metadata" feature yet [02:14] quicktime has it [02:14] *does [02:14] yes, my example removes the rotation metadata [02:14] i need to do this because my vlc doesn't read metadata orientation [02:15] you can manually tell VLC to rotate [02:15] how? [02:15] oh no [02:15] i need that vlc reads metadata [02:16] like any software image viewer does with jpg [02:16] some players use the metadata and others do not. [02:16] so you will have to decide what you want to support [02:16] llogan, rotation tag is for all the containers? [02:16] mpeg, avi, mkv, mp4, mov ? [02:16] i don't know what supports it [02:17] i think is a good thing [02:17] because re-encoding a video means loss of quality [02:20] in VLC: tools > effects and filters > video effects > geometry. then use transform or rotate [02:20] yes i know that [02:20] you asked "how?" earlier [02:21] yes, but immediatly then i said "oh no" :) [02:21] because i knew what you were talking about [02:21] can i set only the orientation tag of videos with ffmpeg? [02:22] only that... [02:22] no other things [02:22] see the second example i gave you [02:22] ffmpeg -i in -metadata:s:v rotate="0,90,180" -c copy out.mov [02:22] ? [02:22] that does not make sense [02:22] into " " i need to specify the angle? [02:22] yes [02:23] why does not make sense? [02:23] you have 3 values in your example [02:23] oh yes [02:23] it's only to figure out [02:23] ok [02:23] one last think i'd like to ask you is: [02:24] (maybe is off-topic) [02:24] it's about VLC [02:24] if you leave it empty then it removes rotation metadata completely [02:24] sometimes the video i play has a bigger resolution than my display, this means that i lost the bottom-right angle to resize with the mouse [02:25] and the only thing i have to do is to close VLC [02:26] is there some other thing that i can do to shrink the VLC window size? [02:26] sorry, but i don't really know much about VLC. you'll probably get an answer in their channel. [02:26] ok, thank you for your help [02:26] in #videolan [02:27] i hope the video i'm re-encoding will be more smaller [02:27] it's a 5 min video [02:27] and it's 1,2 GB [02:27] it's made with an iphone [02:28] see https://trac.ffmpeg.org/wiki/Encode/H.264 [02:28] maybe reducing the resolution will be 50 - 100 mb? [02:28] or is false? [02:28] it depends on many factors. read about the -crf option in that link. [02:29] i know that, i already used on virtualdub [03:45] If compression was used in this example http://www.reddit.com/r/movies/comments/2lectl/the_size_of_our_70mm_imax_copy_of_interstellar/clu69ss what would the resulting size be? [03:48] That depends on the "compression" used. [03:48] And also on the content. [03:48] c_14: x264, or x265 [03:53] I don't know how he came up with 18000x12500 per frame... Thats 225 megapixels... [03:53] I'm not even sure H.264 groks more than 4k video... [03:53] 4k as in 4096x* [03:54] There's no way that resolution can be right... I've never heard of a 200+ megapixel sensor [03:54] And I think H.265 only groks 8k [03:55] According to the Levels anyway. [03:58] Holy cow it is correct... http://www.slashfilm.com/film-interview-imax-executives-talk-the-hunger-games-catching-fire-and-imax-misconceptions/ it really is 18000x12500 per frame [04:29] My current video is brg24, I am trying to encode webm. Would there be a color conversion going on? My output keeps changing color. [04:36] I used this command but the color still changed: [04:36] >ffmpeg -i desktop-src.avi -c:v libvpx -qmin 1 -qmax 51 -b 809600 -s 1980x1080 -aspect 16:9 -pix_fmt bgr24 -c:a libvorbis -threads 12 desktop.webm [04:36] Specifically -pix_fmt bgr24 [05:04] d3m0n: I don't think vp8 supports bgr24 [05:09] fudge [05:09] relaxed: What would be your suggestions on getting the closest conversion [05:25] with webm you want yuv420p [05:26] Does it always automatically select that? [05:26] Even without -pix_fmt [05:33] yes [07:22] hi [07:50] has anyone encountered .lpd files? supposedly it is a flash format but ffmpeg / ffprobe does not seem to like it: [07:50] 2010.lpd: Invalid data found when processing input [10:32] anyone familair with nginx rtmp module? uses ffmpeg i think [10:34] is there away to output the current audio level in real-time?, I have seen volume-detect however that is at the end [11:58] Hey I'm trying to start a live desktop stream from an Ubuntu System to a Debian System. The following command is used: ffmpeg -re -f x11grab -r 10 -s 1680x1050 -i :0 -map 0:v -f mpegts -vcodec libx264 -crf 18 -preset:v ultrafast -pix_fmt yuv420p rtp://192.168.100.126 -vcodec h264 and it's giving the error "av_interleaved_write_frame(): Invalid argument". Why this error appears? [13:46] if I have a single input video stream and want to encode multiple output video streams (for multi-bitrate mp4s, one file with multiple streams, not multiple files), how would I do this? [13:50] ecraven, tee muxer [13:50] saste: thanks [14:19] Hi!! [14:20] I'm trying to create simple transcoding program using ffmpeg libraries, input(any)=>Output(h264/aac). Unfortunately, I can't call ffmpeg via command line from my program. Then, I need to use ffmpeg libraries in my code. I compiled this example http://ffmpeg.org/doxygen/trunk/doc_2examples_2transcoding_8c-example.html.But, When try transcode a file, a error occours: "broken ffmpeg default settings detected. Use an encoding preset" [14:21] What I should to do? [14:55] saste: sorry, cannot get this to work.. i have three mp4 files, each containing one video, one audio and two hint data streams. I'd like to mux all of these together into one file (for multi-bitrate-streaming). I've tried ffmpeg -i .. -i ... -i .. -acodec copy -vcodec copy out.mp4, but that only copies one audio and one video stream :( [14:56] try using -map 0 -map 1 -map 2 [14:56] does mp4 support multiple videostreams? [14:57] klaxa|work: that gives me "data stream encoding not supported yet (only streamcopy)", which is what I want anyway :) [14:57] add -c copy [14:58] and omit -acodec copy and -vcodec copy [14:58] klaxa|work: now "Could not write header for output file #0 (incorrect codec parameters ?): Invalid argument" [14:58] but it looks better :) [14:58] can you pastebin your command with complete output? [14:58] a second [15:01] http://paste.lisp.org/display/144293 [15:01] the first is from a file I have that has the structure I want to replicate [15:02] it's for the Helix streaming server -- unfortunately I have no choice on the server to use :( [15:04] hmm... no clue to be honest, my best guess would be that mp4 doesn't support some of the hinthandler streams [15:04] those were generated by -movflags rtphint in the first place :-/ [15:04] I'll try to generate the hints while muxing, not before [15:06] I'm trying the following command: ffmpeg -re -f x11grab -r 10 -s 1680x1050 -i :0 -map 0:v -f mpegts -vcodec libx264 -tune zerolatency -b 900k -crf 18 -preset:v ultrafast -pix_fmt yuv420p udp://192.168.100.127:1234 -vcodec h264 -flags nobuffer How can I reduce latency on this stream? [15:07] generating the hints while muxing doesn't work either :-/ [15:08] ah, wrong order of arguments [15:20] ffmpeg seems to add handler_name metadata, can I somehow prevent that? [15:25] see -map_metadata http://ffmpeg.org/ffmpeg.html [15:25] i think using -map_metadata -0 -map_metadata -1 should do the trick [15:52] I'm trying to capture the screen with audio [15:52] http://dpaste.com/04MZAKK [15:52] getting input/output error for the audio [15:52] how can I find the audio device [15:53] maybe just use -f pulse -i default and use pavucontrol to move it to the monitor [15:53] that's what i do [15:53] and while you're at it, maybe select the monitor as the default recording device [15:54] klaxa|work: could it work without pavucontrol? [15:54] I'm trying to make it as simple as possible [15:54] you can use pacmd list-sinks or something and grep [15:55] Or just pass the index of your monitor source instead of default [15:55] Or it's device name, might be a bit more reboot-safe [15:55] -' [16:09] BtbN: how can I find the index of the monitor source? [16:10] should be somewhere in pavucontrol [16:10] or pacmd/pactl [16:31] BtbN: "arecord -l" gives me the device [16:31] and "if alsa -i hw:0,0" capture audio [16:45] is it true that xvid/divx = h263 [16:46] xvid/divx are mpeg4 asp, which i think is a superset of h263? [16:47] sorry what is "superset" [16:49] pinkette: anyways, short answer to your question is "no, xvid/divx and h263 are not the same thing". [16:49] (but they are similar) [16:50] is avc = h264 ? [16:53] "mpeg4 avc" and "h264" are the same, yes. [16:53] does ffmpeg support h265 [16:54] ffmpeg can decode h265 (aka "hevc") natively, and can encode it via x265. [16:56] Hi. I am trying to output xml data from an rtsp stream. How can I do that with ffmpeg [16:56] The stream has #1.0 (h264 video) and #1.1 (Data: [0][0][0][0] / 0x0000) [16:56] I only want the data part [16:57] Actually, #0.0 and #0.1 [16:58] ffmpeg -i rtsp://foobar -codec copy -map 0:0 -f data data1 -map 0:1 -f data data2 [17:07] c_14: thanks, but I got "Unable to find a suitable output format for 'copy'" [17:08] is there a way to demux/extract all streams from an mkv (an arbitrarily long # of streams) ? [17:08] tiago: pastebin your complete commandline please [17:08] Kip: demux to separate files? [17:09] c_14, yes [17:09] None I know of. [17:09] Without programmatical help, that is. [17:09] c_14: ffmpeg -rtsp_transport tcp -i "rtsp://10.10.9.168:554?meta=1"?-codec copy -map 0:0 -f data data1 -map 0:1 -f data data2 [17:09] Kip: not automatically; you'd have to e.g. write a script that uses ffprobe to identify streams then ffmpeg to extract them. or just use 'mkvextract' from mkvtoolnix, maybe [17:09] I will pastebin the program output [17:10] c_14: http://pastebin.com/raw.php?i=QET30tb6 [17:11] hmm, looks like mkvextract doesn't have an 'extract all' mode either, you have to specify which tracks to extract there, too. [17:11] tiago: what shell is this? [17:12] Try replacing the "" with '' [17:12] kepstin-laptop, what would be the ffprobe command? [17:13] c_14: bash [17:13] Kip: probably "man ffprobe" ;) - it has a lot of options for output format, etc, so what your command will look like depends on what exactly you're doing. But 'ffprobe filename.mkv' is a good start :) [17:14] tiago: try replacing the double quotes with single quotes [17:14] c_14: same result [17:15] try: ffmpeg -rtsp_transport tcp -i "rtsp://10.10.9.168:554?meta=1" "" -codec copy -map 0:0 -f data data1 -map 0:1 -f data data2 [17:15] Unable to find a suitable output format for '' [17:16] Get rid of the quotes entirely. [17:16] None of those characters should be bash-special [17:16] is it true that h263 and mpeg4-asp is not same thing? [17:17] c_14: Now I got "Data stream encoding not supported yet (only streamcopy)" [17:17] got it, added -codec copy before the second map [17:17] thanks! [17:17] I still don't understand what bash was doing there though... [17:18] maybe because it's not gnu bash [17:18] I'm on mac (maybe I should've mentioned it) [17:18] Still, that shouldn't be happening. [17:18] c_14: it looks like that command had a set of empty quotes "" which caused an empty argument to be passed to ffmpeg, and ffmpeg treated it as an output filename [17:18] kepstin-laptop: Not that one, http://pastebin.com/raw.php?i=QET30tb6 <- this one [17:19] For whatever reason it decided that -codec was part of the rtsp url... [17:19] yeah, weird [17:20] c_14: the space between the end quote and the next argument is a utf-8 encoded non-breaking space instead of an ascii space [17:20] I think you can type those accidentally on a mac with a shortcut [17:20] Action: kepstin-laptop downloaded the pastebin and opened it in a hex editor. [17:21] Oh boy. [17:21] That explains it. [17:22] U+1F4A9 [17:25] no, it's actually a U+00A0 NO-BREAK SPACE, encoded as 0xC2 0xA0 [17:25] I think? [17:25] yep, that's correct. [17:26] Yes, but U+14A9 is unicode pile of poo. [17:26] *F4 [17:26] oh, afraid my head is running an older version of unicode which didn't recognize that ;) [17:27] yup [17:51] c_14: f4a9 is private use, 14a9 is ?, canadian syllabics y-cree moo [17:51] 1F4A9 [17:51] 1f4a9: ???? :) [18:12] *oops chattin' the wrong window - my bad... haha ;-P [18:19] anyone familiar iwth nginx rtmp? [19:02] hi [20:04] Hey. Could somone check if I'm not the only one getting framedrops from this command: http://pastebin.com/DtA3Mxf6 [20:05] I'm trying to do the same in C, and I'm getting the same lagspikes there. [20:06] It's a twitch.tv stream btw. [20:21] How do you skip forward a certain number of frames, like -ss does with time? [20:29] edakiri: maybe with something like -vf select="gte'(n,240)'" [20:29] may or may not work [20:31] llogan: I'm trying to -c copy , so I guess -vf won't work together with that. [20:31] Can I copy multiple excerpts from the input to the output file? [20:35] you could try to make individual segments and then use the concat demuxer [00:00] --- Fri Nov 7 2014 From burek021 at gmail.com Sat Nov 8 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Sat, 8 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141107 Message-ID: <20141108010502.5D76718A0251@apolo.teamnet.rs> [00:56] ffmpeg.git 03James Almer 07master:b385c4c6a316: x86/swr: replace sse4 instructions in pack_6ch with sse ones [03:39] ffmpeg.git 03Michael Niedermayer 07master:2daae445c09a: avcodec/aacdec: print extension type when startcode debugging is on [03:39] ffmpeg.git 03Michael Niedermayer 07master:b1c99f4c5f8f: avcodec/aacdec: Add table of profile names [04:17] ffmpeg.git 03Michael Niedermayer 07master:c11f73176856: avformat/webpenc: preserve single image VP8X flags [12:45] ffmpeg.git 03Rong Yan 07master:6a6c16cbcc27: libavutil/ppc/util_altivec.h : fix unaligned_load() vec_unaligned_load() add macros VEC_LD() VEC_MERGEH() VEC_MERGEL() VEC_ST() for POWER LE [12:45] ffmpeg.git 03Rong Yan 07master:79e0255956bc: libavcodec/ppc/hpeldsp_altivec.c : fix ff_put_pixels16_altivec() put_no_rnd_pixels16_xy2_altivec() put_no_rnd_pixels8_xy2_altivec() avg_pixels8_altivec() avg_pixels8_xy2_altivec() put_pixels16_xy2_altivec() put_pixels8_xy2_altivec() ff_avg_pixels16_altivec() for POWER LE [12:48] looooooong commit message [15:02] wm4, line breaks are for newbs [15:25] ffmpeg.git 03Michael Niedermayer 07master:1a1a98f644d6: avutil/ppc/util_altivec: make src pointers const, fix warnings [15:26] ffmpeg.git 03Michael Niedermayer 07master:ddac3053cd28: avutil/ppc/util_altivec: add () to VEC_LD macro arguments [15:48] ffmpeg.git 03Michael Niedermayer 07master:960c573cc549: avcodec/mjpegdec: support pix fmt id 0x22111111 [19:06] ffmpeg.git 03James Almer 07master:edff061fb0db: x86/swr: add ff_float_to_int32_a_avx2 [22:39] ffmpeg.git 03Henrik Gramner 07master:4981baf9b803: avstring: Mark some character handling functions av_const [22:39] ffmpeg.git 03Michael Niedermayer 07master:ed736890d6a7: Merge commit '4981baf9b803f3c4866b2e97fdadb008c62dc7ad' [23:16] ffmpeg.git 03Michael Niedermayer 07master:1384df641994: lavf: Add an option for avoiding negative timestamps [23:16] ffmpeg.git 03Michael Niedermayer 07master:66e49ff3b923: Merge commit '1384df641994bf3d6cb51084290aa94752737bae' [23:45] ffmpeg.git 03Michael Niedermayer 07master:897d5c3a4296: lavf: Print a warning if failed to avoid negative timestamps when requested [23:45] ffmpeg.git 03Michael Niedermayer 07master:0d71e825dbb3: Merge commit '897d5c3a4296f3da80b8699d1487328ca2de8e55' [00:00] --- Sat Nov 8 2014 From burek021 at gmail.com Sat Nov 8 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Sat, 8 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141107 Message-ID: <20141108010501.5A0482AD60D6@apolo.teamnet.rs> [06:08] edakiri: you can split by frames using mkvmerge [06:09] look at --split in `man mkvmerge` [10:40] hi [10:40] i'm trying to generate animated gifs from 10 second preview videos, the resulting outputs are like 19mb [10:40] does anyone know how to shrink the size down a bit? [10:50] ffmpeg -i input.mp4 output.gif [10:51] fairly simple [11:01] Mayumi: "and the COMPLETE console output" [11:01] i need to check which version you're using and which pix format it's using, etc [11:02] ubitux: http://dpaste.com/0E0Y9QW [11:04] Mayumi: does -gifflags -transdiff helps? [11:04] i don't know of those arguments [11:05] just add that as output option [11:05] also, you might want to try different dithering [11:05] i'm trying it out [11:06] like, with -vf scale=sws_dither=a_dither or something [11:06] not sure how it will affect compression [11:08] -vf scale=sws_dither=a_dither reduced by 2mb [11:08] these are for web and they really need to be like, hundreds of kb, not mb, so i'm really not sure what i'm going to do to compress that much. [11:09] try reducing framerate, scale it down, ... [11:09] gif never was a good "video" format anyway [11:09] what's a good frame rate for gifs anyways? [11:09] dunno, 15 fps might do it [11:09] ok [11:10] ffmpeg is not the best "gif" encoder out there [11:10] yeah [11:10] it's just decent enough depending on the input [11:10] i tried outputting as png frames [11:10] then piping those to imagemagick [11:10] but the gif was liek 50mb so idkk [11:11] yeah [11:11] there are all kind of techniques possible [11:11] the problem is that it's really hard to predict a good method [11:11] gif really is shitty :) [11:11] i kno [11:11] you generally have to try different method and pick the best [11:11] but like, these ultimately need to show up on tumlr dashboard [11:11] yeah [11:12] dropping the frame rate to 10 reduced to 8mb so thats something [11:12] i think these have to be under 512k [11:12] no idea how i'm going to do that [11:12] scale it down [11:12] yeah [11:12] whats the arg to scale down? [11:12] its like [11:12] -vf something i think [11:12] -vf scale=320:240:sws_dither=a_dither [11:13] you might want to try different dithering [11:13] like x_dither maybe [11:13] and pick the one that looks less like shit [11:14] ok so reduced to 320:-1 (maintaining aspect ratio with -1?) it seems to be 1.3mb [11:15] doesn't look too horribly bad [11:18] damnit it has to be under 1mb for tumblr [11:18] haha [11:18] is there a way to skip frames? [11:18] or something [11:18] is it really 10 seconds? [11:18] every nth frame? [11:18] yeah [11:18] yeah well that's what fps should do [11:18] and its 1.3mb [11:19] yeah i guess it should [11:19] eh i wonder if i drop it more how bad it will look [11:19] can you provide the avi? [11:20] http://dii87kfw16t1x.cloudfront.net/videos/erotic/test-sp.mp4 [11:22] yeah hardly compressible, let me check [11:22] i kno [11:23] a lot going on in the video [11:23] so what's your current cmdline? [11:23] your best setings [11:24] well i got it to 752kb using 5fps (terrible) and a max width of 300px [11:24] just show me the complete cmd line [11:24] ffmpeg -i in.mp4 -vf scale=320:-1:sws_dither=a_dither -r 5 out.gif [11:24] er, 320px not 300 [11:25] -rw-r--r-- 1 ux ux 1005K Nov 7 11:25 out.gif [11:25] damn [11:25] yeahh [11:25] ffmpeg -i test-sp.mp4 -vf scale=300:-1:sws_dither=a_dither,fps=8 -gifflags -transdiff -y out.gif [11:25] idk how everyone else is making gifs so small [11:25] i'm using this [11:26] they use better encoders :p [11:26] lol [11:26] 8 fps looks like shit though [11:26] when i tried making it from png frames it looked like crap [11:26] er [11:26] it looked good, [11:26] dither makes encoding worse [11:26] it was like [11:26] 50mb [11:27] it would probably be better with a few smart palette changes and no dithering [11:27] but we don't support that yet [11:28] give ImageMagick's `convert` a whirl [11:28] i tried it [11:28] it was like 50mb [11:28] idk [11:28] oh :) [11:29] http://dtbaker.net/random-linux-posts/convert-video-avi-mpeg-mp4-to-a-gif-animation-on-linux/ [11:29] hmm [11:35] it's getting late, i'm going to have to mess with this more tomorrow [11:35] thank you guys for your help i appreciate it, at least i got it dropped down quite a bit. [13:50] hello @ all! how can i see the parameters any filter needs? like a -h or man pages ? [14:49] hi. what would be the best command to encode video at 29.97 fps? (ntsc) [15:01] to answer my question, -r 30000/1001 seems good [15:14] Could someone here explain the difference between the new API modes? with and without reference counting. [15:27] Hello guys i have a very strange HLS stream that ffmpeg can not restream. It's HLS, in my vlc it works perfectly. Here is thje log: http://pastebin.com/XtyUsyf2 [15:27] I also don't understand why it says to enter -bsf:v h264_mp4toannexb filter because the 2 containers are exactly the same [15:27] but even if you enter the filter, the restream doesnt start [15:28] The stream is public so do any tests you may want [15:53] c_14 : i think you can help on this one :P [15:54] I found a way of going around the issue... [15:56] Use one ffmpeg process with -codec copy and -f matroska, pipe the output into a second ffmpeg process with -codec copy -bsf:v h264_mp4toannexb -f mpegts test.ts [15:56] Don't ask me why that helps, but it does. [15:57] In short, this: ffmpeg -i "url" -c:a copy -c:v copy -f matroska pipe:1 | ffmpeg -i - -c copy -bsf:v h264_mp4toannexb -f mpegts test.ts [15:58] Yeah but it isnt a bug? [15:58] should i report it? [15:58] If the same issue occurs with a recent git build, it's probably a bug, yes. [15:59] ok i will download the latest static build from git and i will try [15:59] wait a moment [16:01] Yeah doesnt work either [16:03] In that case it's probably a bug. [16:08] If you're going to make a bug report, be sure to mention that remuxing to matroska inbetween allows you to mux to mpegts with the bitstream filter. [16:12] ok the bug its for avcodec component? [16:20] You can just leave it unset, someone will come around and set it once they've looked into it. [16:24] c_14: https://trac.ffmpeg.org/ticket/4091#ticket [16:40] hello, I would like to test out the new webp support that was added yesterday : https://trac.ffmpeg.org/ticket/3760#comment:7 [16:40] however, when I try to transcode a video to a webp file, I get the error : Encoder (codec webp) not found for output stream #0:0 [16:41] is there some compile option I must use to build in this encoder because from what I can tell in the compile spam, its already there....? [16:41] triune: does it work if you use -c:v libwebp ? [16:42] kepstin-laptop: nope, different error : Unknown encoder 'libwebp' [16:42] hmm, I may have gotten that wrong :) [16:42] triune: if not, try http://johnvansickle.com/ffmpeg/ [16:43] I don't mind compiling myself... what's the option to enable libwebp ?? [16:43] no, the encoder name is 'libwebp', that error means that your ffmpeg wasn't compiled with libwebp support [16:44] sorry, figured that out myself... --enable-libwebp will give it one more go after it compiles ;) [16:45] huh, I would have thought that would be autodetected if you have the pkgconfig files present [16:47] ... i guess not. [16:48] you got me kepstin-laptop :) [16:55] a new era in animated images for the internet has begun! goodbye animated gif! [16:55] https://386dce635fec2a779f110d9c778f9acf6f7bde64.googledrive.com/host/0B7-WGlkadL3zcVRGSTFLRFZBQWM/other/u5.webp [16:55] works beautifully :) [16:56] hmm, I guess I have to figure out how to get it to loop now in the webp [16:59] Action: kepstin-laptop is unclear what benefit an animated webp provides over a webm [17:02] many sites like Google+ support having webp images auto-play like gifs in posts [17:03] ffmpeg -h encoder=webp [17:04] looks like there is no encoder option to loop :( [17:04] it would be an option on the muxer, not the encoder [17:04] presumably looping is just some flag in the container [17:04] webp doesn't have a muxer tho like gif does...? [17:04] at least not in what I just compiled [17:05] ffmpeg -h demuxer=gif [17:05] shows me gif muxer options [17:05] ffmpeg -h demuxer=webp [17:06] says unknown format [17:06] Action: triune slaps self across face [17:06] sorry, its there in the muxer options [17:06] ffmpeg -h muxer=webp [17:09] and now it loops [17:09] https://386dce635fec2a779f110d9c778f9acf6f7bde64.googledrive.com/host/0B7-WGlkadL3zcVRGSTFLRFZBQWM/other/u5.webp [17:09] Action: triune is a happy camper [17:11] thanks for the help guys [17:44] Question: I read a lot of LGPL & GPL and now i want to ask something. It clearly says that if an app is using GPL libraries i must provide the source code. I use FFmpeg in a commercial program. What happened in this case? GPL asks me to release my source code or the ffmpeeg source code since the ffmpeg uses the libraries? [17:44] it may be a dump question but i'm really confused [17:45] BlackDream: it depends on whether your application is linking to ffmpeg or could otherwise be considered to derive from ffmpeg. [17:45] i'm just using a static build of ffmpeg [17:46] i only use the ffmpeg from command line [17:47] in that case, you usually only need to do the ffmpeg source code. But consult a lawyer if unsure, of course. [17:49] BlackDream: do also read through https://www.gnu.org/licenses/gpl-faq.html [17:49] (in particular, the "Distribution of programs..." section) [18:08] I need some help getting my timestamps to be accurate, currently they appear to be 2.54 times real time speed. Any suggestions for resources [18:08] As in the video is playing too fast? [18:10] Im using the following command to recored a timestamp onto a video. The timestamp seems to record to fast. ffmpeg -i testinput.mp4 -vf "drawtext=expansion=strftime:fontfile=arial.ttf:text='%a %d\.%m\.%Y %H\:%M\:%S': r=29.976: x=(w-tw)/2: y=h-(2*lh): fontcolor=white: box=1: boxcolor=0x00000000 at 1" -y out1.mp4 [18:15] try with text=%{localtime\:%a %d\.%m\.%Y %H\:%M\:%S} and without expansion=strftime [18:18] And get rid of the r=29.976 [18:32] I'm using windows version of ffmpeg. Tried this "ffmpeg -i testinput.mp4 -vf "drawtext=:fontfile=arial.ttf:text={localtime\: %a %d\.%m\.%Y %H\:%M\:%S}: x=(w-tw)/2: y=h-(2*lh): fontcolor=white: box=1: boxcolor=0x00000000 at 1" -y out1.mp4but no text on output [18:54] Get rid of the : before fontfile [18:54] and you need a % in front of { [18:54] *{localtime [19:05] Got the syntax right now "ffmpeg -i testinput.mp4 -vf "drawtext=expansion=normal:fontfile=arial.ttf:text=%{localtime}: x=(w-tw)/2: y=h-(2*lh): fontcolor=white: box=1: boxcolor=0x00000000 at 1" -y out1.mp4" Output still not correct when outputing to a file but when to the screen it is ok. [19:10] Watching the console output of ffmpeg the first few seconds are longer than 1 seconds, framerate is high, then as framerate drops seconds get shorter than 1 second [19:10] Yeah, I just noticed a flaw in that command. [19:15] if that drawing the system realtime? If so, it will depend entirely on how fast the video encoder is. If the video encoder is faster than realtime, then the timestamps will be "slow" in the final video [19:16] Yes, that's the problem with the command... [19:16] I'm trying to wrap my head around the timecode thing. [19:16] s/thing/option [19:18] dv8inpp: You want to apply the current timestamp to the file and then increase the timestamp by 1s per second of video? [19:18] yes. I want to record a timestamp over a webcam/ IP cam input [19:19] If you're using a realtime input, then the command I gave you will work. [19:19] It just won't work if you get content faster than realtime. [19:19] You can add -re as an input option to test with a file if you want to. [19:19] s/I gave you// [19:59] Thanks for that. I might have solved the problem another way by using an IP camera with timestamping built in. [20:09] hey all, any ideas how i cant get the frei0r plugins working? [21:57] hi, i'm trying to cut a portion of a video and i managed to do so with -ss and -t. How can i merge the subtitles i have in sub/idx ? [22:01] You have the subs in a separate file? [22:15] yes i have both the sub and idx files [22:16] And you've already cut the video? [22:16] yes [22:17] ffmpeg -i video -ss seek_time -t duration -i subfile -c copy -c:s srt outvideo [22:17] I'm pretty sure ffmpeg can seek in subtitles anyway. [22:20] i got : "Subtitle streams other than DivX XSUB are not supported by the AVI muxer." , i also used a slitly different position for the -ss and -t options : " ffmpeg -ss 1958 -i \[10\]\ Zatoichis.Revenge.1965.DVDRip.DivX.avi -i \[10\]\ Zatoichis.Revenge.1965.DVDRip.DivX.sub -c:s srt -codec copy -t 25 tragoudi-miso-a.avi " [22:21] You need the ss either before the sub file or after the last input file. [22:21] And don't use avi... [22:21] Just going to throw that one out there. [22:23] what should i use instead of avi? [22:23] Depends, with what are you planning on playing the file? [22:23] i plan to upload it to facebook [22:24] ugh, just use mp4 [22:24] ok [22:24] upload to facebook? you might need to burn the subs in then (hardsub with the 'subtitle' video filter) [22:25] er, 'subtitles' filter [22:26] " https://trac.ffmpeg.org/wiki/HowToBurnSubtitlesIntoVideo " says subtitles video filter is for srt while picture-based subtitles should use the overlay video filter [22:27] wasn't sub/idx text? [22:27] huh, interesting. I didn't realize they showed up as a video stream [22:27] no, sub/idx is vobsub - dvd subtitles [22:28] meh, there are too many subtitle formats [22:28] tsester: yeah, then use the overlay filter [22:29] just with 2 inputs,and [0:v][1:s] (assuming the second input is the subtitle file. [22:33] how do I extract dvd_subtitle to a .idx/.sub? [22:38] Kip: doesn't look like ffmpeg can do that, but you should be able to e.g. stream copy them to an mkv [22:41] c_14: can you help me some more? i tried " ffmpeg -ss 1958 -i \[10\]\ Zatoichis.Revenge.1965.DVDRip.DivX.avi -i \[10\]\ Zatoichis.Revenge.1965.DVDRip.DivX.sub -filter_complex "[0:v][1:s]overlay" -codec copy -t 25 tragoudi-miso-a.mp4 " but got the error " Streamcopy requested for output stream 0:0, which is fed from a complex filtergraph. Filtering and streamcopy cannot be used together." [22:42] tsester: that error message tells you exactly what the problem is [22:42] i'm way over my leage [22:42] you have to get rid of the -codec copy and reencode the video [22:43] ok [22:43] Also, move the -ss, you're seeking the video not the subtitles [22:50] kepstin-laptop, ffmpeg.exe -y -i h.mp4 -map 0:4 -c copy sub.mkv gives Could not write header for output file #0 (incorrect codec parameters ?): Invalid data found when processing input [22:51] Kip: I have no idea what you're doing, but if your input is an mp4 then you certainly don't have dvd image subtitles in it. [22:53] I think nero can/used to actually mux mp4s with vobsubs... [22:54] hmm [22:54] ... could anything other than nero software actually play them? [22:55] vlc & mpc-hc plays my mp4 with vodsub inside [22:56] Kip: huh. Well, you could try using -codec dvdsub [22:57] kepstin-laptop, haha that did it! [22:58] thanks [22:58] is extracting .sup still broken? [22:59] I remember pgs muxer patches... [23:01] Hmm, doesn't look like they've hit the tree yet. [23:02] You could probably apply the patch from the ML if you need it. [23:03] c_14, I've had bad luck whenever I compile myself ;P [23:05] what's the codec called for .sup? [23:06] What's .sup? (Nothing much, what's up with you?) [23:06] :) [23:07] Presentation Graphic Stream iirc [23:09] c_14, i didn't manage to hardcode the subs,. any idea? , http://pastebin.com/raw.php?i=2gtXfcA8 [23:10] Didn't you want to seek the subtitles to the correct position? [23:11] yes, but i tried to put the -ss after the first input file but the output was from the beginning of the movie [23:11] ffmpeg -i video -ss time -i .sub -filter ? [23:13] no, this plays the video from the beginning to "time" [23:14] Is time > or < 25 ? [23:14] 25 [23:14] but [23:14] i want to play from 1958 to 1983 i think [23:18] And if you leave off the -ss the subs don't show up at all? [23:18] neither with -ss or without [23:20] now i also have a new problem, i tried to cut the video from 1958 onward but the video is out of sync with the audio:( [23:30] Hmm, looks like ffmpeg doesn't like seeking in dvd subtitle streams... [23:32] hmm [23:33] The only thing I can think of off the top of my head would be to burn the subtitles in for everything from 0 to the final time you want, use -t or -to to stop there, and then cut the beginning part you don't want in a separate step. [23:34] but even without the -ss option i didn't manage to hardcode the subtitles [23:36] c_14, can you please link me the .sup bug tracker? [23:42] Kip: The only ticket I can find is this one: https://trac.ffmpeg.org/ticket/2208 . But if you mean the patches, this would be the latest iteration https://ffmpeg.org/pipermail/ffmpeg-devel/2014-September/163398.html [23:44] c_14, thanks! [23:52] here's the video without subtitles : https://drive.google.com/file/d/0B-7D6SNJJCcuNjRtQjU3TVNNSW8/view [23:52] bye bye [23:52] Action: c_14 is currently testing on one of his own videos. Will report if I get something working. [00:00] --- Sat Nov 8 2014 From burek021 at gmail.com Sun Nov 9 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Sun, 9 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141108 Message-ID: <20141109010503.0546518A0259@apolo.teamnet.rs> [00:32] ffmpeg.git 03Martin Storsj? 07master:1d8a0c1b43e5: movenc: Allow to request not to use edit lists [00:32] ffmpeg.git 03Michael Niedermayer 07master:2893d1b36d17: Merge commit '1d8a0c1b43e58332a3a15c67d4adc161713cade8' [00:41] ffmpeg.git 03Martin Storsj? 07master:e7d20f12c5ef: movenc: Remove a now redundant check [00:41] ffmpeg.git 03Michael Niedermayer 07master:872c0bcd3c04: Merge commit 'e7d20f12c5eff5570cd897f3ce3a88456024036b' [01:00] ffmpeg.git 03Carl Eugen Hoyos 07master:e2e36a739d3a: Fix make checkheaders for libavcodec/vorbis_parser.h. [01:11] ffmpeg.git 03Martin Storsj? 07master:9cbf70fa0e44: movenc: Write correct presentation timestamps in tfra [01:11] ffmpeg.git 03Michael Niedermayer 07master:c9a9a7a1fd7e: Merge commit '9cbf70fa0e44613590b019cef1fe99aa3f3c5d9d' [01:43] ffmpeg.git 03Xiaohan Wang 07master:33301f001747: Fix read-after-free in matroska_read_seek(). [03:05] ffmpeg.git 03Lukasz Marek 07master:2b14593148e0: ffmpeg_opt: free incorrect priv_data for feed stream [10:56] Can anyone help me out with color blending masks for RGB32 format? [11:03] ffmpeg.git 03Martin Storsj? 07master:95a449d3ce8e: movenc: Remove an outdated comment [11:03] ffmpeg.git 03Michael Niedermayer 07master:0bdc5db520ba: Merge commit '95a449d3ce8e15522df47a80a8a4593ea5c2b1bb' [11:16] ffmpeg.git 03Martin Storsj? 07master:8cb7b7b461b5: movenc: Avoid leaking locally allocated data when returning on errors [11:16] ffmpeg.git 03Michael Niedermayer 07master:4342b346d273: Merge commit '8cb7b7b461b52898765b38e3eff68c0ce88347f3' [12:05] cehoyos, pretty much since that commit (Jan 2014) [12:05] That commit? [12:06] cehoyos, http://git.videolan.org/?p=ffmpeg.git;a=commit;h=b1ad9312331759679a9c956233716a67ae681d89 [12:07] So it is impossible to fix both bugs? [12:10] with both you mean the ones reported in the ticket? if so, yes [12:12] The bug that b1ad931 tried to fix and the one in ticket 3517 [12:12] I could not reproduce the one federico reported on the commit message, but in any case that change does not seem correct to me, so it's better to revert [12:12] Of course, I just wanted to ask. [12:13] the bug that b1ad931 tried to fix will probably still be there, but I need more info or a way to reproduce to understand why it happens [12:14] Perhaps this should be added to the commit message? [12:16] sure [12:26] ffmpeg.git 03Martin Storsj? 07master:9a5ac36b69ed: movenc: Require samples before trying to write edts [12:26] ffmpeg.git 03Michael Niedermayer 07master:9ebfe38f3873: Merge commit '9a5ac36b69ede4563e9ecd734141b12ea3280fbc' [12:41] ffmpeg.git 03Carl Eugen Hoyos 07master:4436a8f44ded: Remove fminf() emulation. [12:41] ffmpeg.git 03Michael Niedermayer 07master:b0ed88b4c0e9: Merge remote-tracking branch 'cehoyos/master' [12:48] ffmpeg.git 03Carl Eugen Hoyos 07release/2.4:2be7d565bb2a: Remove fminf() emulation. [12:52] ffmpeg.git 03Michael Niedermayer 07master:6f21fb793238: Revert "v4l2: setting device parameters early" [12:55] Should this commit get backported? [13:21] cehoyos, yes, if possible it should be backported [14:27] Holden: Done [14:28] ffmpeg.git 03Michael Niedermayer 07release/2.2:418e9a6113a7: Revert "v4l2: setting device parameters early" [14:28] ffmpeg.git 03Michael Niedermayer 07release/2.3:b152305bb34f: Revert "v4l2: setting device parameters early" [14:28] ffmpeg.git 03Michael Niedermayer 07release/2.4:56e11cbe3237: Revert "v4l2: setting device parameters early" [15:02] cehoyos, thanks [16:11] I'm writing a demuxer for Android with ffmpeg for mpeg ts files. And I found a problem that I cannot read the first keyframe after start or seek. anyone known why? [16:30] hello, would it be possible to make the content-type of output format avaliable to a protocol? [16:51] wtf http://git.videolan.org/?p=ffmpeg.git;a=blob;f=libavfilter/vf_interlace.c;h=e07963f79a4688410f07be836649553b0a2dbfb9;hb=HEAD#l207 [16:52] lol [16:53] never trust a pts [18:58] any dev here? [18:59] nope, nobody here but us chickens [18:59] ePirat: don't ask to ask, just ask. [18:59] I already did [19:00] well, you got the answer too. [19:00] hmm I got nothing& maybe due to connection to my znc timed out& sorry then if I missed it [19:01] nope, nobody here but us chickens [19:01] I meant: [16:30:17] hello, would it be possible to make the content-type of output format avaliable to a protocol? [19:02] well, I guess nobody knowns... [19:02] as in, MIME type? [19:02] yes [19:03] would be good for http protcol and would increase usability of icecast protocol a lot [19:03] I don't think ffmpeg uses MIME types for much of anything [19:03] most supports requests we get are due to wrong or not set mimetype [19:04] (or libavformat in particular) [19:04] but apparently it does at least have them listed [19:08] it should be possible to add a mimetype field to URLContext and initialize that [19:09] HTTP already supports it via AVOptions [19:11] each protocol which wants it also could have a private context with a mimetype instead thats set for the muxing case via AVOptions [19:11] could have ffmpeg_opt.c automatically set that option if not set via CLI [19:12] something in libavformat would be better than ffmpeg so it works for all applications [19:18] so, it'd need to be provided by avio_open2 [21:44] ffmpeg.git 03Michael Niedermayer 07master:29b1af40f3a0: avformat/mpegts: add scan_all_pmts option [21:44] ffmpeg.git 03Michael Niedermayer 07master:41ad87ad8efe: avformat/mpegts: improve first valid PMT heuristic [21:46] dorek666 [22:18] Action: Daemon404 wonders why 4436a8f44dedc83767b3d9da9beb85d1fae2ca30 wasnt on the ML [22:22] Do you want me to revert? [22:23] considering it was only recently added, i assume there was actually a reason [22:24] Yes, the reason was code that was removed the very same day the workaround for code added a day before was removed: 867c02acd [22:26] well ok [22:26] Action: Daemon404 notes none of this stuff is mentioned in any of the commit messages for any of the related commits [22:29] I tried hard to add all relevant information to my commit... [22:29] (I edited it twice.) [22:30] michaelni: Shouldn't ffmpeg (the application) set "-scan_all_pmts 1"? [22:32] cehoyos, that might make sense, yes, want to post a patch ? [22:33] I will try;-) [23:18] michaelni: I added "av_dict_set(&o->g->format_opts, "scan_all_pmts", "1", 0);" to open_input_file() and this works to set the option but how can I find out if the user had already set the option? [23:24] Solved using AV_DICT_DONT_OVERWRITE. [23:56] ffmpeg.git 03Michael Niedermayer 07master:f9fa560597cf: avcodec/aacenc: check input for NaN [23:58] michaelni: I don't think that was ticket 3762;-) [23:58] argh [23:58] 4084 [23:59] Shouldn't the encoder ignore the data instead of failing? [23:59] Please ignore the patch on the mailing list, I will send an update. [00:00] --- Sun Nov 9 2014 From burek021 at gmail.com Sun Nov 9 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Sun, 9 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141108 Message-ID: <20141109010501.E84CF18A009B@apolo.teamnet.rs> [00:02] hmm, looks like burning in dvdsub subtitles is broken... [00:54] c_14: is it a regression? [00:55] Hmm, good question. Let me finish what I'm doing and then I'll do a git bisect to see if I can find out. [02:26] Hmm, can't seem to find a version where it works. [02:27] It does seem to work if I use the idx file though. [02:34] Ye, with the idx seeking and everything else seems to work as well. [03:00] Not sure if that's a bug or not though. I mean, technically the idx is part of the "subtitle stream" as such if I understand correctly, but you can watch using only the sub... [04:38] Question on technical feasibility of AVC lossless crop, regardless of whether ffmpeg supports it http://pastebin.com/puVkC5ji [08:40] what is the usual extension for uncompressed video? [08:53] .y4m or .yuv [08:59] thank you [12:58] Action: DelphiWorld ugh emilsedgh [12:59] \o/ [15:42] hello, is it possible to remove embedded chapter marks in mp4 files? thanks in advance [15:43] -map_chapters -0 might do it [15:45] alright [15:46] so {ffmpeg -i in.mp4 -map_chapters -0 -c copy out.mp4} will do it? [15:47] Maybe, probably, hopefully. [15:47] Never tried it, but it sounds like it might. [15:52] alright thanks, will report back lel [16:04] anyone? [16:06] I'm writing a demuxer for Android with ffmpeg for mpeg ts files. And I found a problem that I cannot read the first keyframe after start or seek. anyone known why? [16:45] hello [16:45] i would like make a question [16:46] i wanna try conect a wii u gamepad with a raspberry pi (using the raspi hardware encoder) [16:46] but i am not sure if i can [16:46] (the developers have create a patch for software x264 codec ffmpeg code) [16:48] only if the raspi hw encoder creates streams that are compatible [16:49] i dont know where ask this : here or searching a rasperry pi chat (i am searching now) because on libdrc seem no many activity [16:49] hello [16:49] yes ,this is my problem [16:49] ( i know no many about video encoder,only basics) [16:50] how can i know (and how config raspi hw ) be able to create this kind of streams [16:50] (can i post here links?? [16:52] this is how wii u gamepad "understand" the streams: [16:52] http://libdrc.org/docs/re/sc-vstrm.html [16:53] and they use ffmpeg for building the lib [16:53] http://libdrc.org/docs/installation.html#system-requirements [16:54] they created a patch for ffmpeg (but software version) [17:00] I want to dump audio from mp4 file to uncompressed fomat to edit it in audacity, what should I use as "-acodec" option ? -acodec wav ? [17:01] just set output file name to something-dot-wav [17:01] will that "wav" be losless format ? [17:02] I want to conserve all audio quality [17:05] I presume yues [17:09] .wav is audio only .avi . it is mostly used for PCM that is raw format, so it is lossless [20:08] Hello, i read this https://trac.ffmpeg.org/wiki/Creating%20multiple%20outputs on how to creadte multiple outputs and it worked. However i want to use duplicate output when i transcode so that all the outputs so that the transcode will be done once. I read that this can be made using the -f tee [20:08] however, i'm currently using -f segment and i have no idea how to write it in the tee format [20:10] "The tee pseudo-muxer was added to ffmpeg on 2013-02-03, and allows you to duplicate the output to multiple files with a single instance" [20:10] it says to multiple files, so it can't be done in segment format? [20:12] what do you mean by -f segment? [20:12] tee mux? [20:12] ok let's say that i'm using this: [20:12] BlackDream: -f tee "[segment]filename|otherfoo" <- should do it afaik [20:12] ffmpeg -i "url" -acodec copy -f segment -segment_list out.list out%03d.nut [20:13] yeah but where do i specify the segment time and list? [20:14] Let me say it again. I'm currently using -f segment to output the same FFmpeg INPUT to MPEGTS & FLV. So we have double output here. [20:15] However, if i want to transcode the input stream, how i can write the command so that the encode will be made with one instance of ffmpeg and output again to segments in mpegts & flv [20:15] eh, [f=segment:segment_list=out.list:other_option=foobar] [20:15] I think that should work. [20:15] Didn't test though. [20:15] ok let me try [20:16] Slave '[f=segment:segment_list=out.list]': error writing header: Muxer not found [20:18] What's your complete command line? [20:19] maybe you don't use the =out.list, try replacing the '=' with a space?? [20:20] ffmpeg -i "URL" -codec copy -f segment -segment_format mpegts -segment_list mpegts.list -segment_time 10 out%03d.ts -codec copy -f segment -segment_format flv -segment_list flv.list -segment_time 10 out%03d.flv [20:20] eh, the new one [20:21] The one with -f tee [20:21] and in this example, i also use -codec copy twice, so how i can do the copy/transcode using it only one time but output it using duplicate output ;p [20:21] ah [20:21] sorry [20:21] i actually didn't make it so i don't know [20:22] What command were you using when ffmpeg gave you the Slave error. [20:23] ffmpeg -i "URL" -codec copy -f tee -map 0 "[f=segment:segment_list=out.list]" [20:27] try: ffmpeg -i URL -codec copy -f tee -map 0 "[f=segment:segment_list=out.list:segment_format=mpegts:segment_time=10]out%03d.ts|[f=segment:segment_format=flv:segment_list=flv.list:segment_time=10]out%03d.flv" [20:27] wow you made it [20:27] <3 love you [20:50] nick sky3 [22:40] Can anyone point me in the right direction? I'm trying to figure out the best way to utilize both processors / all cores to encode to x264 (from a lossy source mpg). [22:42] rsully: With recent ffmpegs, it should do that. [22:42] Automagically. [22:43] I last tested using default install from ubuntu 12.04. Is there an official repo for ubuntu to get the most up to date version? [22:44] Ubuntu doesn't even ship ffmpeg, it ships a fork called libav. If you want ffmpeg, try looking here: https://trac.ffmpeg.org/wiki/CompilationGuide/Ubuntu (or a couple of links in a second). If you want to use libav, go to #libav. [22:44] sacarasc ah that explains the warning i got [22:47] any pro/con of ubuntu multimedia ppa vs jon's ppa? [22:47] I've never used either, so can't comment. [22:57] looks like jon's ppa is super out of date [22:58] and the multimedia ppa is on 2.4.2 [23:03] I'll just grab a static build for now and figure it out later [23:07] Good afternoon, I'm trying to reencode/copy a part of movie that downloaded in torrent, thus having just a little piece of it. The idea is to download some continous pieces from the middle of the torrent, maybe with the header (mp4) and "fix" them with ffmpeg, but I couldn't make it work... I would appreciate any ideas :) [23:16] Q11 so your goal is to only download say 10% of the middle of the torrent + 5% of the beginning with the header and then use that to reconstruct the 10% middle to be viewable? [23:23] Q11: basically you want seekable torrents [23:34] Hello71 ooh thats a neat way to look at it. but I think he is looking at it from the perspective of only wanting a specific timestamp of the video so he can save bandwidth [23:41] Hello71: What do you mean? If I do calculate the part of the torrent I want to download, it still most likely starts in middle of something, which causes to ffmpeg to fail, as far as I can judge... [23:55] Thanks all, g2g... [00:00] --- Sun Nov 9 2014 From burek021 at gmail.com Mon Nov 10 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Mon, 10 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141109 Message-ID: <20141110010502.792B318A009B@apolo.teamnet.rs> [00:02] i dont think encoders should ignore their input and that certainly cannot be encoded [00:04] replacing Nans by silence or something else shouldnt be done in each encoder [00:07] The input was concatenated aac: Is it not valid to concatenate aac? [00:39] the output from the decoder is not valid and cannot be encoded as such [00:40] So is there a bug in the decoder? [00:40] maybe [05:39] so all the filter-pixfmt failures when yuv2planeX_8 asm is enabled happen to be on 12 and 14 bits fmts, apparently because they are trying to run said 8bit function [05:43] changing the "default:" into "case 8:" in that line fixes the failures with the funtion enabled [07:40] exit [11:02] Hello, I applied to FFmpeg for an internship through GNOME OPW. I wanted to say Hello. [11:05] My qualification task was to restore daemon mode in FFserver. I would like to thank everyone who had been supportive to accomplish the task. [11:09] ? [11:09] how did this OPW thing end anyway [11:10] wasn't november 1st the deadline? [12:06] ffmpeg.git 03Michael Niedermayer 07master:55d592f7d984: avcodec/aacdec: Skip processing channel elements which have not been present [12:11] wm4, there where various deadlines for various things, the next thing to happen is that selected applicants will be posted by OPW on Nov 12 2014 [12:44] ffmpeg.git 03Michael Niedermayer 07master:d1970929b5f8: avformat/librtmp: fix swfurl [16:44] ffmpeg.git 03Marton Balint 07master:631ac655c00e: ffplay: implement separete audio decoder thread [16:44] ffmpeg.git 03Marton Balint 07master:cc4741888d7f: ffplay: fix indentation after last commit [16:44] ffmpeg.git 03Marton Balint 07master:7ba727777587: ffplay: only output null packet once on EOF [16:50] ffmpeg.git 03Marton Balint 07master:86476c510ebd: avfilter/avf_showwaves: fix off by one error in loop condition [16:56] kierank: Is the "Extended ancillary data syntax" part of the transport stream or part of the mpeg audio stream? [16:56] C.4.2 of http://www.etsi.org/deliver/etsi_ts/101100_101199/101154/01.11.01_60/ts_101154v011101p.pdf [16:56] in the mpeg audio bitstream [16:57] Do you know where in the libavcodec mp2 decoder I can find the data? [16:57] at the end of the frame i believe but I don't know if any encoders actually write it [16:58] So it is absolutely possible that it makes no sense to read it (because it isn't used)? [16:58] The reason I ask is that users asked for PLII decoding which MythTV supports but [16:58] I wonder how to detect that two channel audio is PLII encoded. [16:59] you can't afaik [16:59] you just apply the matrix [16:59] Except if the Extended ancillary data syntax is used you mean? [17:00] i've never seen it actually used, i tried to identify such streams some time ago as well [17:00] although when i read up about decoding PLII i gave up on that too, its annoying [17:00] nevcairiel: Does your software support PLII upmixing? [17:00] no [17:01] but people asked for it, but i just cba [17:01] MythTV source file el_processor.cpp contains a GPL implementation [17:01] it requires fancy convolution filters which i didn't have the baseline code for [17:02] encoding is so nice and easy, just a mixing matrix and done [17:05] Did you ever "test" the FFmpeg PLII necoder? [17:05] There is an unapplied patch that claims to fix a bug. [17:05] it seemed to follow the spec when i looked at it [17:06] i never tried to actually decode it again [17:10] I unfortunately cannot find the patch... [17:10] Only the ticket: https://trac.ffmpeg.org/ticket/3455 [17:14] thats how DPLII is supposed to be mixed, it doesn't sound perfect when you listen to it without decoding it again [17:14] left gets the back channels subtracted, right gets them added [17:14] so right is slightly louder [17:38] Is there a reason why AVCodecContext has no AVOptions for width and height? [17:38] Is there any problem I add it? [17:40] why would it need that ever? [17:40] aren't width and height either set by the decoder, or by the demuxer? [17:40] never the user [17:42] for encoding by the user [17:42] encoder will not encode when not set [17:43] even for encoding, the size is dictated by the input AVFrames, right? [17:46] look like it is respected what is set [17:47] I found the patch for PLII encoding: http://thread.gmane.org/gmane.comp.video.ffmpeg.devel/175700 [17:48] it doesn't sound perfect when you listen to it without decoding it again -> I think he meant it doesn't sound perfect after decoding it [17:48] Or maybe not... [17:57] ffmpeg.git 03Lukasz Marek 07master:db2caf0a80d2: lavc/options: fix shallow copy context [17:57] ffmpeg.git 03Lukasz Marek 07master:457204ee1592: lavu/opt: document av_opt_copy function [17:57] ffmpeg.git 03Lukasz Marek 07master:4e179436b6c8: lavu/opt: copy dict in av_opt_copy [18:06] wm4: i checked and for example libx264 copy AVCodecContext.width into x264_param_t.i_width [18:06] and it produces frames with such dimensions [18:06] so my question is still opened [18:13] Is your question how to set it? You only need a setter function for fields that do not exist in avconv, width and height are old fields. [18:17] I asked why they are have no corresponding AVOption to set them [18:17] and if I may add it [18:18] why do you need avoptions for those, just set them directly [18:26] Hard to explain without a context, but such option would simplify some things I'm working on [18:26] but nevermind for now [18:27] for now setting manually is ok [18:54] kierank: will you send an updated patch with the change i mentioned above? [19:43] ffmpeg.git 03Carl Eugen Hoyos 07master:e971eef8c0d2: Set -scan_all_pmts 1 in ffmpeg, ffplay and ffprobe if not set by user. [19:47] jamrial: what change? [19:47] 4:39 AM <"jamrial> so all the filter-pixfmt failures when yuv2planeX_8 asm is enabled happen to be on 12 and 14 bits fmts, apparently because they are trying to run said 8bit function [19:47] ah [19:47] default -> case 8 [19:58] Should the fixes for ticket 3762 and ticket 4084 be backported to release branches? [20:08] nevcairiel: I don't remember if you are using the libavformat mpegts demuxer: The sample that you provided for tickets 2441 and 3762 (announced in 2186) needs "-scan_all_pmts 1": You may want to make this default. [21:12] jamrial: sent but i think it got stuck in the ml queue [21:22] it just went through [22:57] michaelni: what is avoidable in the c api [23:15] the data tables in the C API are structured for C, the SIMD API structures them for SIMD, filter coefficients already replicated and in the right format for SIMD consumption, also indexes to he source lines and end identifer are in there IIRC these reduce the number of needed registers as well as needed instructions [23:18] is it normal that ffplay doesn't quit anymore with 'q'? [23:21] ubitux: see the ml [23:22] i have the problem in 2.4 as well [23:22] michaelni: yes but the majority of yuv2yuvX functions are c api. [23:22] kierank: what are you refering to? [23:22] can we just not add a check for bitexact and use that function [23:22] yuv2yuvX_8 takes up 33% of my cpu doing a 422 to 420 conversion [23:23] ubitux: ffplay thread discussion [23:24] mmh ok [23:24] thanks [23:24] and maybe bitexact should be default [23:24] it's not 1998 any more [23:26] all the yuv2packedX functions are SIMD API its only the 2plane one which are not IIRC [23:26] also we have existing asm for this [23:27] thats actually why that patch fails [23:27] the existing asm is overriden but the API switch is not [23:28] please see c->yuv2planeX = RENAME(yuv2yuvX ); and the yuv2yuvX_* [23:28] oh well I'll just have to patch swscale locally [23:28] enjoy 2x slowdown [23:29] wouldn't setting use_mmx_vfilter to 0 if the yasm versions are used solve this? [23:29] michaelni: for the non-bitexact mode I don't use... [23:29] jamrial, yes but you still get the 2x slowdown [23:30] why do you want to replace fast asm by slow asm ??? [23:30] do a yuv422p to yuv420p conversion and tell me that's fast... [23:30] when it uses yuv2yuvX_c [23:31] not to mention my hac^h^hpatches to stop pointless (x * 2^n) >> n in luma [23:31] kierank, if you disable asm it is slow yes [23:31] no asm is not disabled [23:31] because of that commented out line it uses c [23:31] for me anyway [23:32] kierank, if you disable bitexact then the optiized yuv2yuvX is disabled [23:32] I'm using bitexact [23:32] but we can enable that C API function ofr bitexact that should be ok [23:32] need to test [00:00] --- Mon Nov 10 2014 From burek021 at gmail.com Mon Nov 10 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Mon, 10 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141109 Message-ID: <20141110010501.6FF3D18A008D@apolo.teamnet.rs> [11:09] I have a large .mp3 file - is there a way to radically compress it? I don't mind losing audio quality [11:16] shevy: try, ffmpeg -i input.mp3 -q:a 9 output.mp3 [11:37] Hello there! [11:38] cool relaxed [11:38] from about 184 MB to 44 MB [11:40] Could someone please help me? I want to convert a whole folder (and the other folders inside!), with some .avi files, into .webm [11:41] I'm a bit lost into ffmpeg documentation, as I'm a noob on file conversion and I don't understand so much English ^^' [11:43] Maybe a commnd line, or a good GUI for ffmpeg... ? [11:49] I found this one : ffmpeg -i input.avi -vcodec libxvid output.avi [11:50] but That's only for ONE file... And I have dozen of videos to convert [11:51] Is it working with *.avi ? [12:12] NasUser: hold on a sec [12:15] NasUser: In the top level directory run the following: [12:15] find . -name "*.avi" -print0 | xargs -0 bash -c 'for i; do ffmpeg -i "$i" -c:v libvpx -crf 10 -b:v 2M -c:a libvorbis "${i%.*}".webm; done' - [12:20] it may take a while ;) [13:04] Hi guys. I use ffmpeg as encoder for an rtmp stream from my webcam. All is fine with flv. but when I use mp4, it says: http://paste.kde.org/pwscrfrac [13:04] I'd appreciate any help [13:06] "muxer does not support non seekable output" [13:15] relaxed: which means? [13:15] (please note that when I use /tmp/1.mp4 instead of rtmp:// it works fine [13:17] rtmp is FLV only [13:18] FLV can contain various video and audio formats, but rtmp is an adobe-only, FLV container only streaming format [13:19] and the error means that mp4 cannot be written a->b (it wants to seek in the end and write a header, which of course is impossible with a non-seekable output) [13:33] JEEB: ooh, i didnt know rtmp is FLV only [13:34] thanks [13:34] my main problem is that I play baseline h264 flv through rtmp but some clients dont get any video output [13:36] avconv -y -f alsa -i hw:2,0 -f video4linux2 -itsoffset 4.8 -i /dev/video1 -vcodec libx264 -vprofile baseline -b:v 220k -b:a 40k -pix_fmt yuv420p -strict -2 -acodec aac -f mp4 rtmp://... [13:36] this is the command I use. I read somewhere that vprofile baseline and pix_fmt yuv420p will help [13:36] but they didnt [13:40] you meant -f flv ? [14:51] Hey folks [14:51] just wanna find out what's the best setup to do live transcoding [14:51] Xeon , i7 ?! [14:52] Any recommended hardware [14:53] newest arch intel cpu, as fast as your budget lets [14:54] so you think i7 is better at transcoding with FFmpeg rather than Xeon ? [14:54] anyone has experience with FFmpeg in production [14:54] I would appreciate it [14:55] many people do, but it's basically down to your exact use case and budget as well as quite a few other things [14:55] well the budget is not so restricted as this is for a business [14:55] and no idea about "standard" cpus versus xeons, although I guess the xeons could be a bit faster with a larger cache and all [14:55] not for home use [14:56] well, yes [14:56] but you still have one [14:56] and you then have the use case [14:57] we are pushing a multicast to one interface on the transcoder and getting a transcoded stream out of another [14:57] well few streams I should say [14:57] coz the multicast has about 7 of them [14:57] Transcoder == x86 server really [14:57] running ubuntu and FFmpeg [14:58] with a script to manage the streams and a web interface [14:58] basically the best way to test is to have some kind of hardware and see how well it fares, and then seeing what should be gotten next. Also some hints can be gotten by what kind of resolution/frame rate of content it is, and the preferred compression efficiency (I will guess you will use libx264) [14:58] I want to find out if there is an optimal hardware config that can give the best performance hence allowing me to squeeze in more streams per transcoder [14:59] we do have few xeon servers [14:59] :) [14:59] there's no random recommendation really [14:59] the streams are not all HD [14:59] some are SD [15:00] We are currently producing about 80 streams so we have few servers running but we are thinking about scaling up [15:01] so it makes sense to find the best optimal configuration [15:01] just that AMD doesn't cut it these days, newest architecture is good and that quad/nice amount of cache is good. RAM always helps, but it gets more useful after a few gigs after the resolutions get higher or when the compression efficiency is pushed further (more reference frames and lookahead in memory etc) [15:02] multiple streams on a single machine of course means that more cores still helps, and cache gets more important, as well as RAM [15:02] xeons offer more threads than i7s, so I would go with those if it's in the budget. [15:02] and everything else [15:03] the latest Gen maybe [15:03] we currently use HP 160 G6 with dual xeons [15:03] and also some Dells [15:03] but they are not cutting edge [15:03] I would say 2008 or 2009 [15:04] maybe the new Xeons will do better [15:04] yeah, if you are going to be using libx264 you'll get a nice boost with newest generation stuff [15:04] so its pure CPU [15:04] no GPU needed [15:04] yes, unless you like expensive heaters [15:04] correct [15:04] haha [15:05] those can be helpful for the cold winters yh [15:05] aye, but there's cheaper ways to do it ;) [15:05] indeed unless the Russian decide to cut our (UK ) Gas supply [15:05] then we start using GPUs :P [15:05] lol [15:05] but yh anyway [15:06] so latest Xeons maybe [15:06] with good Ram [15:06] yeh [15:06] If I was a rich man... http://ark.intel.com/products/75290/Intel-Xeon-Processor-E5-4657L-v2-30M-Cache-2_40-GHz [15:07] 12 cores, 24 threads [15:07] that isn't exactly current gen [15:07] and lacks f.ex. AVX2 [15:07] so $5000 [15:08] for the processor [15:08] also depending on the use case 12 cores can be too much, as depending on your settings x264 doesn't use all of those cores well [15:08] I need to look into full server solutions to be honest [15:08] if you have multiple encodes going then of course that will work better [15:08] yh I do [15:08] I have a multicast full of Mpeg transport streams [15:09] that needs to be picked one by one and transcoded [15:09] but still, I'd rather favor raw speed and newer architecture instead of pure amount of cores [15:09] great thanks for the advice [15:12] oh, v3 is the new hotness. Cheaper too http://ark.intel.com/products/81059/Intel-Xeon-Processor-E5-2697-v3-35M-Cache-2_60-GHz [20:48] can someone suggest if IRC servers are having issues? [20:49] can't join channels on servers [20:49] Maybe you should ask in #freenode? [20:50] maybe he can't join that channel [20:50] skyman3, if you are in many channels already try parting one [20:51] usim in this channel only [20:51] im in * [20:51] Can you join #freenode? [20:51] facing the issue since lastday [20:52] u mean irc.freenode.net? [20:52] Nah, the channel #freenode [20:53] no [20:53] im now in #ffmpeg [20:53] I can see that. [20:53] welcome [20:54] /j #freenode [20:55] ok [20:55] yes joined freenode relaxed [20:55] thro the CMD u provided [20:57] i wanted to join #linuxmint-help etc channels vide Greenbean.GeekShed.net or Neptune.GekShed.net which i cant [20:57] good. maybe you need to read up on your irc client's manual [20:57] which client are you using? [20:57] hexchat [20:57] im on Mint Mate17 [20:58] /j #hexchat (for support) [20:58] had okie [20:58] okie* [20:58] never had such issue earlier [21:24] can I record just a relangular region on the screen with ffmpeg ? [21:24] https://trac.ffmpeg.org/wiki/Capture/Desktop [21:24] Use the x and y initializers along with the video_size to draw your box. [21:26] c_14: great [21:26] another question... can I record an specific window ? [21:28] Not with the current implementation. [21:28] Other than adjusting the box. [21:29] ok [21:29] thanks [21:37] Is there a way under linux to grab a specific window as input with the "x11grab" device using the window ID (or something else) without taking anything else ? and if not, is there a way to achieve such a goal ? i'm trying to stream a game, and i'd like my stream to stay on the game even when i'm looking at something else (with alt-tab). [21:43] None I know of. [21:45] leopardb: I don't think it's possible [21:45] I think using another display would be interesting for you [21:46] yeah that's the only way i see :/ isn't it more taxing on the GPU to start another display ? [21:48] I think it is [21:48] duh >_< [21:48] but I don't think it will affect your recording at all [21:49] I use a ffmpeg command to record skype in [21:49] skype out [21:49] and my emulator area [21:49] works great [21:49] Morning/afternoon. I help maintain a project that uses...well, I lost track on what version of ffmpeg at this point. However, I think it was definitely before version 1.x.x came out. I'm now seeing that 2.4.3 is out, which sounds like a major leap. How stable is it right now, relatively? [21:52] Relative to a pre 1.xx version? [21:53] Very stable. [21:53] More features. [21:53] More bugfixes. [21:53] Are there any API programming changes either myself or another member would have to worry about updating on our end? [21:55] Probably. [21:56] They should all be in doc/APIchanges [21:56] if it goes back old enough, I may have a chance. I don't know if I'll be able to just say "we're using a more stable/consistent ffmpeg build" to the team, though. [21:57] The APIchanges file goes back to 0.6 [21:57] The oldest date is 2009-03-01 [21:58] I'll have to hope my instinct is wrong, then. I have to wonder if we use 0.4.8 or some fork of it. Checking our repo now to better determine that. [21:59] ......holy crap. I'm actually behind on the times. We have moved on to 2.1.3. [22:12] Anyone know what could cause this error: Assertion len >= s->orig_buffer_size failed at libavformat/aviobuf.c:452 [22:16] anyone have any idea on how to duplicate the first audio stream and re-encode as an aac stream to be added as an extra stream? (the rest of the file remains unchanged) [22:21] I've got " -map 0:v -c:v copy -map 0:a -c:a copy -map 0:a -c:a:1 libvo_aacenc -b:a:1 320k" so far but it seems to duplicate all the audio streams [22:23] ffmpeg -i input -map 0:0 -c:a:0 aac -map 0:0 -c:a:1 copy [etc] [22:23] assuming the 0th stream is thet audio [22:23] You can use -map 0:a:0 as well [22:23] That'll be the first audio stream. [22:24] The important part is the -c:a:0 and -c:a:1 [22:24] ah okay, and that would copy the stream and encode it in a different format? what I've had happen is it just re-encodes the first stream rather than leaving it untouched [22:26] It will do both. [22:26] Ie encode the first stream to aac and copy it untouched. [22:29] it strangely seems to be duplicating the video stream [22:29] Stream mapping: [22:29] Stream #0:0 -> #0:0 (h264 (native) -> h264 (libx264)) [22:29] Stream #0:0 -> #0:1 (h264 (native) -> h264 (libx264)) [22:30] Because the first stream in your input file was the video stream. Either change the -map 0:0 s to the stream id you want or use -map 0:a:0 [22:32] ah okay - I see what you mean [22:32] assuming the stream was for example; Stream #0:0(eng): Video: h264 (Main), yuv420p(tv, bt709), 1920x800 [SAR 1:1 DAR 12:5], 23.98 fps, 23.98 tbr, 1k tbn, [22:32] 47.95 tbc (default) [22:32] Stream #0:1(eng): Audio: dts (DTS-HD MA), 48000 Hz, 5.1(side), fltp, 1536 kb/s (default) [22:32] Stream #0:2(eng): Audio: ac3, 48000 Hz, stereo, fltp, 192 kb/s [22:33] I did -map 0:1 but that just re-encodes the first stream in-place [22:33] You need -map 0:1 -c:a:0 aac -map 0:1 -c:a:1 copy [22:37] Oh, and whatever you do, don't add another -c:a whatever, that'll (probably) override everything again. (though it might work if you have that one first and the other ones later). [22:37] ah okay, awesome - appreciate your help c_14 [22:38] annoyingly though, it's not copying the stream, just re-encoding it [22:38] Stream mapping: [22:38] Stream #0:0 -> #0:0 (copy) [22:38] Stream #0:1 -> #0:1 (dts (dca) -> aac (native)) [22:38] Stream #0:1 -> #0:2 (copy) [22:39] oh I see, so, stream 0:1 is copied to 0:2 - how would I be able to copy "other" audio streams directly? Assuming, I don't know how many streams there are [22:39] but, I do know I always want to take the first audio stream [22:41] -map 0:a:0 -map 0 -c:a copy -c:a:0 aac <- might work [22:44] that's kind of there, i seem to be duplicating the video stream though; Stream mapping: [22:44] Stream #0:0 -> #0:0 (copy) [22:44] Stream #0:1 -> #0:1 (dts (dca) -> aac (native)) [22:44] Stream #0:0 -> #0:2 (copy) [22:44] Stream #0:1 -> #0:3 (copy) [22:44] Stream #0:2 -> #0:4 (copy) [22:44] my command so far; -map 0:v -c:v copy -map 0:a:0 -map 0 -c:a copy -c:a:0 aac -strict -2 [22:44] turn -map 0 to -map 0:a [22:44] Or get rid of -map 0:v [22:45] -map 0 will map every stream [22:45] video, audio, data, subtitle [22:45] etc [22:46] ah awesome - i think that worked :D [22:46] Stream mapping: [22:46] Stream #0:1 -> #0:0 (dts (dca) -> aac (native)) [22:46] Stream #0:0 -> #0:1 (copy) [22:46] Stream #0:1 -> #0:2 (copy) [22:46] Stream #0:2 -> #0:3 (copy) [22:46] however, is there a way i can push the re-encoded aac stream to the last stream ? [22:46] (apologies for all the questions) [22:47] Only if you know exactly how many audio streams you have. [22:48] yeah, let's say i do (trying to automate this, got a php script so i have a count of audio streams) [22:48] i'm actually doing this as the Roku doesn't support DTS unfortunately [22:49] Let's assume 3 audio streams, -map 0:a -map 0:a:0 -c:a copy -c:a:3 aac [22:49] should do it [22:49] so, -c:a:3 would put it as the 4th stream? would that be right? [22:49] yes [22:50] Well, it won't "put it", it takes the 4th mapped audio stream and encodes it to aac. [22:50] Since -map 0:a maps the first three, -map 0:a: maps to the fourth [22:51] i think that worked; [22:51] Stream mapping: [22:51] Stream #0:0 -> #0:0 (copy) [22:51] Stream #0:1 -> #0:1 (copy) [22:51] Stream #0:2 -> #0:2 (copy) [22:51] Stream #0:1 -> #0:3 (dts (dca) -> aac (native)) [22:51] c_14 - you are awesome, appreciate your knowledge and help on this! [22:52] np [23:21] Does ffmpeg have some functionality to continously record/restream a stream, and display a static image when the source is down, while it continously retries to open the source stream? [23:21] not internally, no [23:22] It's a hls source, so my idea was to just make one 2 second .ts segment with the static image, and insert that every time i can't get a segment in time. But that propably causes chaos because of timestampf [23:22] *timestamps [00:00] --- Mon Nov 10 2014 From burek021 at gmail.com Tue Nov 11 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Tue, 11 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141110 Message-ID: <20141111010501.820042AD60DF@apolo.teamnet.rs> [00:13] you still here c_14 ? [00:15] ye [00:16] if you're interested in what you helped me with earlier, or i guess if anyone ever asks the same question - i put the code up on GitHub; https://github.com/andrew-s/roku_audio_encoder [00:18] Just a nitpick, it's FFmpeg, and ffprobe is all lowercase. [00:19] (for the readme) [00:19] :P updated [00:20] i guess it's a fairly niche thing i would imagine anyway [00:20] needing to add an AAC track to all your files i mean [00:21] I personally haven't had anybody with that issue besides you so far. [00:21] lol, fair enough [00:21] Action: blippyp looks around the room to see if anyone is there... [00:22] oops - wrong channel... ;) [00:52] hi. how can I determine the number of frames in a video? [00:53] (other than, multiplying FPS by length in seconds) [00:59] hello all ... quick question, i'm trying to convert subtitles from one format (ASS) to another (SRT) .. [01:00] i'm pretty sure the ASS is ok, but when converting to SRT it retains stuff that i don't really want ... was wondering if there's a way to handle it. (example, tags with {?}?) [01:40] heyy guys [01:42] hey frost [01:49] how can i get the raw frame data information after using av_read_packet and av_decode? [02:04] scoofy: ffprobe -show_streams -count_frames <- the nb_frames var [02:08] c_14: thanks [02:09] does anybody know a decent tutorial for recording audio generated by applications with ffmpeg and alsa and without pulse? (As a bonus, recording mic input and audio output at the same time) [02:10] I tried a few things with snd_aloop but without luck, and at the same time the .asoundrc syntax is confusing me [02:13] I don't know of any tutorials, but I can give you some pointers. [02:13] as a C programmer, I like pointers too [02:14] lel [02:14] https://trac.ffmpeg.org/wiki/Capture/ALSA <- this covers the basics as well as mic input [02:15] yeah I saw that page [02:15] Applications should be as easy as modprobe snd-aloop pcm_substreams=1 and recording from hw:Loopback,1,0 and outputting to hw:Loopback,0,0 [02:16] thanks, let me try that [02:17] asoundrc should be pcm.!default hw:Loopback,0,0 (probably) [02:21] c_14: just like that, without quotes and braces? [02:21] because then it gives me an Input/output error when I try to record [02:22] c_14: http://sprunge.us/WYhX [02:23] let me test over here [02:25] anything - just newer - even if it doesn't completely work, just something that shows 'good code practice'etc..interetesting design, but not so old and irrelevant... [02:26] I really need to stop doing that - sorry guys [02:28] c_14: ah, if I use pcm.!default "hw:Loopback,0,0" it works [02:28] well, almost [02:28] I got that far before already: it records the audio from the application but I can't hear it while recording [02:29] If you want to hear the audio while you're recording it, you'll either need to listen to the ffmpeg output (so you'll have a delay) or set up dmix [02:30] Wait, no [02:30] Not dmix, you need to duplicate the streams somewhere [02:30] ehh [02:31] This works, somehow. [02:31] Ye, with routing [02:31] https://raw.githubusercontent.com/vehk/dotfiles/master/asoundrc/asoundrc_multi <- look at the pcm.multi part [02:32] Have one of the slaves be the hw:Loopback,0,0 and the other the output device you want to listen on. [02:33] what's a slave? a device where audio is routed to? [02:33] pretty much [02:33] cool [02:34] computer lingo tends to have a lot of masters, slaves, servers, children, parents etc etc. [02:35] It's always great when you're talking with people about a process killing children and then people walk by. [02:36] heh [02:41] what does the ttable stuff mean? [02:42] and the bindings? [02:44] The ttable duplicates the stereo input channels [02:45] Ie turns the 2 channels into 4 so the route can split the 4 into 2x2 [02:45] The bindings tell the route with channels go where [02:46] The only part you should need to change are the slaves [02:50] now I get something about "channels count not available" whenever I use aplay [02:51] or, actually [02:51] if I route to "hw:0,0" I get "Device or resource busy", if I route to "hw:0,3" I get "channels count not available" [02:52] Can you pastebin your asoundrc? [02:52] sure, one sec [02:53] c_14: http://sprunge.us/eETe [03:02] Right [03:03] >pcm.!default { type plug slave.pcm "multi" } [03:03] replace your pcm.!default with that [03:04] hrm [03:04] And you _might_ need to replace the slaves identifiers with similar constructs ie slaves.a.pcm "loopin"; where loopin is pcm.looping { type plug slave.pcm "hw:Loopback,0,0" } [03:05] still "Device or resource busy" if I route to "hw:0,0" [03:05] ah, okay [03:05] why is this so much more complicated than recording without immediate playback :P [03:05] Because ALSA is rather eh complex. [03:07] Yeah, I got it working on my end. [03:07] congrats :) I didn't... [03:07] Want me to pastebin my asoundrc? [03:08] with "hw:0,0" still "Device or resource busy", with "hw:0,3" an assertion error [03:08] that would be great [03:09] http://ix.io/f6Q [03:09] Then record with ffmpeg -f alsa -i loopout [03:09] While audio is playing [03:13] I replaced Headset with the name of my card --> "Device or resource busy" [03:14] why? nothing else is playing audio at the same time... [03:14] fuser -v /dev/snd/* ? [03:15] returns nothing [03:16] what's your current asoundrc? [03:17] http://sprunge.us/agSb [03:18] MID is the HDA Intel MID card [03:18] can you pastebin aplay -L ? [03:19] http://sprunge.us/RTYY [03:20] if I try HDMI instead of MID I get: aplay: main:722: audio open error: No such file or directory [03:21] (the same file plays perfectly fine when there is no .asoundrc) [03:22] Maybe use aplay -l to get the direct hardware device numbers? [03:22] If that doesn't work, you might want to ask #alsa for this port [03:23] http://sprunge.us/gYWW <-- aplay -l and arecord -l [03:26] try hw:0,0 or hw:0 instead and switch the headset pcm to a type plug with pcm.slave as that hw:0 thing, like with loopin and loopout [03:27] no luck [03:28] whenever hw:0,0 is involved at any point, the device is busy [03:28] Ask #alsa ... or sett up a dmix in front of hw:0,0 [03:28] *set [03:28] *ehem* as expected of alsa [03:30] well thanks anyway for that effort! [03:30] *looks at clock* wow, it's been 1.5 hours since I joined [03:31] if you could lend me your soundcard too :P [03:31] I kinda need that, sorry. :) [03:40] hw: devices are by definition not multiplexed [03:40] Hello71: is that a problem in my configuration? [06:29] I'm having issues with encoding from a v4l compliant Capture Card. [06:31] It seems any AVIs or mp4's I generate play a green screen on my Raspberry Pi based xbmc device, or play video with no audio under mplayer. [07:57] hi, say I have: ffmpeg -i seq.mp4 -vf "movie=output.mkv[inner];[in][inner]overlay=main_w-overlay_w-10:main_h-overlay_h-10[out]" -map "[out]" output.mkv Why doesn't the -map "[out]" work ? I get this error "Output with label 'out' does not exist in any defined filter graph, or was already used elsewhere." [07:59] (in this case if I don't write the -map, the command works, but I'm trying to put the map because I want to add an audio track as another -map [08:23] It seems any AVIs or mp4's I generate play a green screen on my Raspberry Pi based xbmc device, or play video with no audio under mplayer. [11:38] what is the best(smaller size-good quality) speech encoder? http://en.wikipedia.org/wiki/List_of_codecs#Voice [11:42] gcl5cp: probably libopus [11:42] http://www.opus-codec.org/ [11:51] im really at my wits' end to figure out that that i can't join other channels on IRC [11:52] sky2: #hexchat is where you need to ask about this. [11:53] ok relaxed [11:54] thank relaxed, i will test it [12:16] i was only able to join #ffmpeg with sky2 [13:39] I have added some feature in ffmpeg, I want to enable it only when I say --enable-feature=foo but its codec file is compiled even I dont say enable-fearure=foo [13:53] buenos dias ffmpeg [13:53] I have a problem when using filter_complex on windows to concatenate some files [13:54] I have the error "Error while filtering" [13:55] The command I am running is as follows: http://hastebin.com/tihabagute.vhdl [14:16] The problem was that I was using was from March last year [15:19] hey guys [15:19] I'm programming a GUI for ADRone and receiving an udp stream [15:20] I want to put this stream in ffmpeq and display it in my JavaFx window [15:20] but the delay is 13s [16:42] Hello, i have 2 TS videos that i want to combine them into one. Why this doesnt work? ffmpeg -i "concat:1.ts|t.ts" -codec copy -f mpegts /tmp/combined.ts [16:42] it only adds the first ts [16:42] can someone explain these options for libmp3lame: -q:a 5 -ac 2 -ar 44100 [16:44] vbr setting 5, channels 2, audio rate 44100 [16:45] `man lame` to read about the vbr ranges [16:46] thansk :) [16:49] blackdream - try ffmpeg -i 1.st -i t.ts -vf "concat" -c:v copy combined.ts [16:50] Filtergraph 'concat' was defined for video output stream 0:0 but codec copy was selected. [16:50] Filtering and streamcopy cannot be used together. [16:50] you should also be able to just cat the files together as well like: cat t1.ts t.ts>combined.ts - but this probably isn't a great solution. someone with a little more know-how might be able to tell you why. [16:51] Yes but i want to combine mp4 movies into one without creating a new file [16:51] but if you plan on just re-encoding them again later, it probably doesn't hurt [16:51] you have to [16:51] https://trac.ffmpeg.org/wiki/How%20to%20concatenate%20(join,%20merge)%20media%20files [16:51] what you want to try to avoid though is re-encoding it [16:51] but here it is using concat without creating new file [16:52] checking out link [16:52] BlackDream: or mp4 use `mp4box -cat 1.mp4 -cat 2.mp4 -new combined.mp4` [16:53] ya - I don't see a 'join' command... or anything like that - the fastest you're going to get is by remusing it. [16:53] for, not or [16:53] remuxing* [16:53] relaxed yes it's ok with that. However why the above concat command does not work. It successfully starts the 1.ts but completly ignored the t.ts like i didnt write it at all [16:54] I don't think you used the command properly - do it like I showed you [16:55] this one: ffmpeg -i 1.ts -i 2.ts -vf "concat" -c:v copy combined.ts [16:55] yes [16:55] that should work [16:55] Yes i get [16:55] Filtergraph 'concat' was defined for video output stream 0:0 but codec copy was selected. [16:55] Filtering and streamcopy cannot be used together. [16:55] oh yeah - I forgot that [16:56] then use a bitstream filter on them [16:56] hold on I'm trying to find it [16:56] https://www.ffmpeg.org/ffmpeg-bitstream-filters.html [16:56] oh, try ffmpeg -i concat:1.ts\|2.ts -c copy output.ts [16:57] I thought you said mp4 [16:57] relaxed yes i will use it with mp4 later. [16:58] with mp4 you have to use the bitstream filter - the link I showed you last [16:58] Ah guys i think i know what is the issue with the above command and concat. It only adds the 1.ts because when ffmpeg tries to combine the 2.ts it says: [16:58] [mpegts @ 0x427d760] New video stream 0:2 at pos:2052660 and DTS:1.4s [16:58] [mpegts @ 0x427d760] New audio stream 0:3 at pos:2109248 and DTS:1.45778s [16:58] [mpegts @ 0x427d760] PES packet size mismatch [16:58] unless it errors out you're fine [16:59] yes they are with yellow [16:59] those are just inormative [16:59] informative* [16:59] for reason the output file is larger than 1.ts&.ts but it only plays the first ts [17:00] the first TS is 8 seconds and then it stops [17:01] tsmuxer might be a better tool or this job [17:01] anyway, with mp4 the above command will work? i will try now [17:01] for (my f key is flakey) [17:02] blackdream - just cat them together like I said - I've done it lots [17:03] it is ok to cat 2 mp4 files together? [17:03] but like I said - I usually re-encode them again later [17:03] no [17:03] no - you have to use the bitstream filter like I said [17:03] ok i see [17:03] h253_mp4toannexb [17:03] w8 [17:03] oops - h264_mp4toannexb [17:04] it's fast though [17:04] and with .ts it may work but you're destroying time stamps [17:04] yeah it's 'wonky' which is why I re-encode them [17:04] I would never try to distribute the 'cat/joined' versions [17:04] it works because .ts is like a 'raw' format or something - can't remember the details [17:05] also works with mpg's iirc... [17:05] blippyp you suggest this one: ffmpeg -i 1.ts -i 2.ts -vf "concat" -c:v copy -bsf:v h264_mp4toannexb combined.ts [17:05] ? [17:06] ya - something like that - you're o the right track now [17:06] Filtergraph 'concat' was defined for video output stream 0:0 but codec copy was selected. [17:06] Filtering and streamcopy cannot be used together. [17:06] again [17:06] change -c:v copy to -codec copy [17:06] It seems any AVIs or mp4's I generate play a green screen on my Raspberry Pi based xbmc device, or play video with no audio under mplayer. [17:06] use my command with that bitstream filter [17:06] sorry - you have to convert them first [17:06] don't concat them [17:07] ffmpeg -i 1.ts -codec copy -bsf:v h264_mp4toannexb out1.ts [17:07] ffmpeg -i 2.ts -codec copy -bsf:v h264_mp4toannexb out2.ts [17:07] other conversions to avi, and mp4 have had anomolies like slowed or sped up playback. [17:07] then cat out1.ts out2.ts>combined.ts [17:08] then bitstream them back to mp4 format [17:08] i cant use this filter because the containers is the same for both [17:08] I use to do this all the time with the video from my hauppauge pvr [17:08] i get a n error [17:08] ffmpeg -i concat:1.ts\|2.ts -c copy -bsf:v h264_mp4toannexb output.mp4 [17:08] stop concatenating the [17:08] the* them [17:09] first you convert them [17:09] then cat them together [17:09] then convert the joined file back to mp4 [17:09] blol blippyp [17:09] why not do it all at once? [17:09] because it won't let you [17:09] at least I've never found a way [17:10] it's fast - it basically just copies the files [17:10] yes it will, my command doesn't use the filter [17:10] oh - sorry -wasn't paying attention... [17:10] cool - I'll have to try that [17:10] relaxed didnt combine the ts, i run the video and again i'm only see the first one [17:11] blippyp: it uses the concat protocol- confusing, I know [17:12] BlackDream: hmm, can you pastebin the command and output [17:12] ya - thanks - I will definately try that next time [17:12] which *should* work with mpeg containers [17:12] ok hold one for a minute. Forget the TS files. I will use mp4 for now as that was my job from the start. So i have 1.mp4 and 2.mp4 i have download some samples files [17:13] I tried your command changing the ts to mp4 and again i was able to see only the first one [17:13] also ffmpeg stopped when frame= 166 fps= 30 q=-1.0 Lsize= 377kB time=00:00:05.56 bitrate= 554.5kbits/s [17:13] video:315kB audio:56kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 1.595011% [17:13] at 5 seconds, but both mp4 files are 10 seconds [17:13] fucking pastebin [17:13] ok ok okok [17:14] you can't spam random, cherry picked output and expect us to know what's going on [17:14] http://pastebin.com/2TwSH3B7 [17:16] BlackDream: The concat protocol doesn't work with mp4s. Use the mp4box command I gave you earlier [17:17] ok wait [17:17] gotta go - you need that bitstream filter though - mp4's need them when joining/splitting, that mp4box that relaxed is suggesting is likely doing exactly the same thing [17:17] good luck [17:29] i have an avi that has audio delayed by a second or two, can anyone help me with a command to fix it? [17:32] Lac3rat3d: ffmpeg -i input.avi -itsoffset -2 -i input.avi -map 0:v -map 1:a -c copy output.avi [17:33] thanks [17:33] wrong :( [17:33] ?? [17:33] too many -i input :) [17:33] ffmpeg -i input.avi -itsoffset -2 -c:v copy -c:a copy output.avi [17:33] ? [17:33] no, it's right :) You have to use the same input twice [17:33] oh [17:34] -2 == delay audio by 2 seconds? [17:34] we hope [17:34] ... o_O [17:34] Action: relaxed is having pints and giving tech support [17:35] ok [17:35] i'll try it out. thanks. [17:40] BlackDream: did that work out for you? [17:41] :D [17:41] Pardon me, wrong window [17:41] Lac3rat3d: you need the -map(s) too [17:41] ya, i see what you did there :) [17:42] video from file one, audio from file 2 (with delay) [17:42] i'll test it out in a bit [18:35] hey all. I'm interested in taking H264 nal units and creating an MPEGTS format context via libavformat. Has anyone done this? Or should I ask on the mailing list [18:42] When I export to mp4, the first second of the video is low resolution, is there a flag that will prevent this? I assume it is low res to allow fast start on internet streaming. [18:42] Command / output : https://gist.github.com/Tolmark12/817a44ce048079355c8f [18:42] relaxed: works, thanks :) [18:43] heyy guys [18:43] how is data organized inside an AVFrame? [18:45] http://shots.delorum.com/client/view/Screen%20Shot%202014-11-10%20at%2010.44.35%20AM.png (first second on left, after a few seconds on the right) [18:47] tolmark12: pastebin your entire ffmpeg invocation and its output [18:48] https://gist.github.com/Tolmark12/817a44ce048079355c8f [18:49] uh, 550kbps is pretty low [18:49] what's a better bitrate? [18:50] if you care more about quality than bitrate, try using x264's CRF-mode [18:52] bitrate is important since it will be sent over the internet, but maybe I just need to bump up the quality [18:53] often CRF-mode with the `maxrate` parameter is a good choice, then [18:54] or, if average bitrate is absolutely essential (hint: it's probably not), 2-pass encoding [18:55] cool, that's helpful [18:57] do you have a suggestion of a good maxrate value I should start testing at? [18:58] depends on your intended audience; 1Mbps might be a good place to start if -b:v 550k looked largely OK [18:59] cool, thanks for the help! [18:59] welcome! [18:59] oh, also, if speed doesn't matter, set -preset veryslow [18:59] ok [19:00] and if you want it to playback on mobile devices and such, you may want to set an H.264 profile and/or level (see the x264 docs) [19:00] I seem to get more reliable results from Libreoffice once I run over the file with the ODF Integrator Converter, rather than simply opening the file directly in Lo 4.3 [19:00] ohhh, I was wondering about that, thnks [19:01] On the topic of this channel. [19:02] I'd like some advice to Digitizing VHS Tapes of family movies. [19:02] ack ack ack ack ack analog [19:05] My goal is to burn them to DVD and make them availible via our Raspberry Pi based devices. [19:06] @rcombs thanks again (!) that did the trick, looks great [19:07] huzzah! [19:36] Zombie: if you want to burn them so that they can be read by a dvd player, just use -target dvd (or pal-dvd or ntsc-dvd depending). [19:37] I'm having trouble encoding them. [19:39] ffmpeg -i tv:// -tv driver=v4l2:device=/dev/video0:input=1:norm=ntsc:adevice=/dev/audio2 Mom_2nd_attempt.mp4 Here is the command [19:40] What's the problem? [19:40] It seems any AVIs or mp4's I generate play a green screen on my Raspberry Pi based xbmc device, or play video with no audio under mplayer. [19:41] ffprobe $file [19:41] And pastebin it [19:54] http://pastebin.com/YHtsfJ5g [19:56] try adding -pix_fmt yuv420p [19:57] Not sure why mplayer isn't playing the audio though. Unless, it's been built to dislike mp3s... [20:02] I think that has more to do in this case with my Capture Equipment. [20:02] This is a VCR Connected to the card with Composite Cables. [20:02] So? As long as ffmpeg doesn't throw errors, the output should be playable. [20:06] What would the ideal method for capturing from this VCR be? [20:07] I've made many attempts with mencoder, vlc and ffmpeg [20:07] tbh, whatever works [20:11] Nothing I have done so far has truly worked to my complete satistfaction. [21:48] Hi, I have a mpegts(mpeg2+ac3) video with a 10s hole in the audio...transcode to mpegts(h264+aac) removes that hole(thus, sound is out of sync) - what can I do to prevent this? [21:49] probably use copyts parameter? [21:49] only thing that works is -copyts, but I can't use that(other problems) [21:49] yep... [21:49] isn't there any other way? [21:50] what other way would you want? [21:50] other than copyts [21:50] either you leave same timestamps which will keep the hole [21:50] or you generate new ones which wont [22:32] how can I cut a part of some video with ffmpeg ? [22:32] can anyone link me ? [22:36] got it [22:36] :P [00:00] --- Tue Nov 11 2014 From burek021 at gmail.com Tue Nov 11 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Tue, 11 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141110 Message-ID: <20141111010502.8882118A00B6@apolo.teamnet.rs> [00:09] kierank, tested, works, pushed [00:10] ffmpeg.git 03Kieran Kunhya 07master:b546023b9319: swscale: fix yuv2yuvX_8 assembly on x86 [00:14] thx [01:03] ffmpeg.git 03Luca Barbato 07master:e3e317e0c015: lavc: Compact the side-data passthrough [01:03] ffmpeg.git 03Michael Niedermayer 07master:0960cc4cc60f: Merge commit 'e3e317e0c015b164b6c2eb8913e393216d78de23' [01:25] ffmpeg.git 03Tristan Matthews 07master:a1a259881fa7: v4l2: Use av_strerror [01:25] ffmpeg.git 03Michael Niedermayer 07master:d99653c9e444: Merge commit 'a1a259881fa7b23e2ffc0c2a43d4923fe42d0478' [01:40] heyyy guys [01:40] how can i get the raw video information after using av_read_packet and av_decode? [01:43] ffmpeg.git 03Luca Barbato 07master:09f25533a564: v4l2: Preserve errno values [01:43] ffmpeg.git 03Michael Niedermayer 07master:715ccc2bc479: Merge commit '09f25533a564eab743f258d168697a11122914c4' [02:51] ffmpeg.git 03Martin Storsj? 07master:b44a242c3dfa: libfdk-aacdec: Support building with the latest version of fdk-aac [03:00] ffmpeg.git 03Michael Niedermayer 07master:9889b29ff4b0: avcodec/utils: make sidedata remapping table static const [03:01] ffmpeg.git 03Michael Niedermayer 07master:f6a8c77afb86: avcodec/utils: Fix warning: comparison between enum foo and enum bar [-Wenum-compare] [04:18] ffmpeg.git 03Arwa Arif 07master:19d0949d31b0: lavfi: add xbr filter xBR [04:18] ffmpeg.git 03Michael Niedermayer 07master:064a23766923: avfilter/vf_xbr: Uppercase first letter of context type for consistency [06:21] so xbr was pushed without the reference to the ticket, and without test? [06:22] it could get various macro cleanup too [06:22] but oh well.. [06:23] the @file is also not up-to-date [06:23] since x3 and x4 are now present. [06:25] ffmpeg.git 03Cl?ment BSsch 07master:ae65a84517cf: Changelog: explicit that xBR scaler is implemented as a filter [06:30] ffmpeg.git 03Cl?ment BSsch 07master:fc0a91e3cdf5: avfilter/xbr: consistent use of @see [06:30] ffmpeg.git 03Cl?ment BSsch 07master:548a5f7ef619: avfilter/xbr: fix TODO entry [06:33] ffmpeg.git 03Cl?ment BSsch 07master:937ebb843579: avfilter/xbr: fix filter description field [06:34] ffmpeg.git 03Cl?ment BSsch 07master:30466cac9d59: avfilter/xbr: drop yet another x2 reference [06:38] ffmpeg.git 03Cl?ment BSsch 07master:ff9b21776b8d: doc/filters: explicit and complete xBR documentation [10:57] there's something about Mary ^^ [11:11] Indeed [11:14] It seems deleting tickets puzzles fflogger [11:15] it breaks the rss stream somehow [11:15] i suppose [11:18] ubitux: Is ticket 3491 still reproducible? [11:20] cehoyos: try it and tell me :) [11:20] it might be fixed [11:27] I saw he sent this patch: https://ffmpeg.org/pipermail/ffmpeg-devel/2014-April/157159.html [11:27] Didn't you commit the change as part of a larger patch? [11:29] that code is no longer in master [11:29] That's what I mean, yes. [11:30] also, the codec ids ASS and SSA don't mean what the patch implies they mean [11:30] they are both for ASS subtitles, just different internal representation [11:30] yeah that code is not present anymore [11:30] and if you look below the ifdefery [11:30] it fallbacks on what he wants [11:30] _SSA is not used at all anymore now, in any case [11:31] (the codec id) [11:31] if have a local branch dropping it [11:31] what about the avi thing? [11:31] in which i removed all the timestamp scaling in lavc go away [11:31] wm4: it works actually [11:32] (iirc) [11:32] because it seems it actually creates a buffer like if it was an independant file or something like this [11:32] i need to check again [11:33] well anyway... i'll figure that out later [11:33] nevcairiel: Did you see my comment about -scan_all_pmts? [11:33] we should probably make a helper to untangle this instead of a decoder [11:36] Bye, gtg [11:38] ffmpeg.git 03Timothy Gu 07master:e1ee0521a698: tests: Fix test name for pixfmts tests [11:38] ffmpeg.git 03Michael Niedermayer 07master:a7d451c1dd75: Merge remote-tracking branch 'github/master' [11:47] ffmpeg.git 03Martin Storsj? 07master:28396d17ff1c: libfdk-aacdec: Support building with the latest version of fdk-aac [11:49] ffmpeg.git 03Michael Niedermayer 07master:a602f88e2dea: Merge commit '28396d17ff1c1493b78d6eece484ffc27ed86d0d' [11:58] saste: i'd like to rework the macro mess in xbr if you don't mind [11:58] but i want the fate patch to get applied first [11:58] ubitux, sure [11:58] ubitux, makes sense [11:58] ffmpeg.git 03Stefano Sabatini 07master:73f74f6b166a: lavfi/xbr: apply misc cosmetical fixes [11:59] we need to factor out the LUT code btw [12:05] michaelni: why exactly was cea3a63ba3d89d8403eef008f7a7c54d645cff70 needed? [12:08] michaelni: if get_pool() fails, doesn't that mean simply no buffer was available? [12:11] michaelni: http://people.videolan.org/~jb/tmp/update [12:13] j-b, thanks, ubitux theres the hook if you want to update it, else ill try and post a patch [12:22] wm4, if get_pool() is run by 2 threads at the same time one of them will "always" fail no matter how many buffers are available [12:25] but get_pool has a retry-loop [12:26] hmm [12:26] holy shit... so get_pool removes the entire buffer list? [12:26] and readds it later? [12:26] it takes it to modify it, since its lock-free, afaik [12:27] that's much worse than using a mutex [12:30] yeah this whole lock-free approach is super complex, i never saw the real benefit [12:31] don't malloc/free acquire global libc locks anyway? you can't change the refcount without allocation [14:26] michaelni: i won't patch it, too much pain to test [14:58] ubitux, posted a patch, please review, ill wait a bit for reviews before i mail it to root at vlc [14:58] michaelni: i was looking at it, but why didn't you drop the other .mak entries? [14:58] > common.mak|library.mak|subdir.mak [15:00] they cover more cases, like commonumak or subdirkmak i think, i think if they are removed that could/should be seperate [15:02] well they're obviously specific *.mak patterns, and you add a generic case [15:02] but well, whatever [15:03] also i wanted to keep the change minimal as i didnt setup a full test environment, just tested the command outside a hook [15:04] but, if you want to write a better patch, iam not against that at all [15:06] i don't care much [15:06] as long as i can push my fate patch... :P [15:06] oh.. ifo parsing incoming in ffmpeg... nice. [15:14] yay another shitty hack [15:15] ? [15:15] what do you propose? [15:15] Nicolas suggests a -ifo_file option, which is what i agree with [15:18] dvdread support? [15:23] wm4: yeah... right [16:23] stupid question of the day: can't we hide data in the mov/mp4 stsz table? [16:23] (with a boxsize > sample_count*4) [16:24] michaelni : should we put git hooks in ffmpeg-web repo ? [16:25] i dont think it has to be connected , just so people can send patches [17:31] is anyone working on edit lists? [17:32] (in mov/mp4) [17:34] i am guessing not [17:35] ask on ml for better guesses [18:03] ffmpeg.git 03Michael Niedermayer 07master:e981de81fea7: avcodec/lagarith: fix chroma plane width & height [18:42] heyy guys [18:42] how is data organized inside an AVFrame? [18:43] see the Doxygen page for that struct? [18:44] rcombs: i already looked at it [18:45] what's unclear? [18:45] rcombs: i am wondering how pixels are inside the data parameter [18:45] that depends on the pixel format [18:45] rcombs: each pointer is one component right? [18:45] see the documentation on individual pixel formats [18:45] No [18:45] RGB [18:45] packed YUV [18:46] jfmcarreira: look at pixfmts.h [18:46] err, pixfmt.h [18:46] some are planar, some are packed; some have 1 byte per sample, some have 2 (or maybe more, I'm not sure) [18:46] jfmcarreira: AVFrame.format is one of AV_PIX_FMT_... (for video) [18:46] wm4: i know the pel formats. but i am trying to read a file using libavformat and i am not being able to copy the pixels right from the AVFrame [18:47] i used av_image_copy [18:47] copy where? [18:47] and for progressive formats like yuv420p it worked [18:47] wm4: i am creating a video playuver. and i need to copy the pixels to my class [18:48] read the source code of av_image_copy [18:48] I think it's relatively readable [18:48] ok [18:48] basically you need to know how planes and linesizes work [18:49] because i need then or in separate pointer (one for each component) or in the raw format it is on the bitstream [18:50] wm4: but linesizes are only used to copy specific lines [18:51] well, that's how data is organized [18:51] i want it whole. i was using mem cpy from the beginning of avframe->data[ch] using the number of pixels ( normally width * height) [18:52] but that might change depending on the pixel fmt [18:52] that's wrong [18:52] and there is not necessarily one plane per component [18:52] stride may not equal width [18:52] that too [18:52] rcombs: i see :) [18:55] it is working for yuv420p a planar format [18:55] but not for yuyv422 for example [18:55] it'll stop working if your width != stride [18:56] yuyv is packed. [18:57] so why different pointers in AVFrame? that what i am wondering [18:58] because that's more efficient/simpler for planar formats [18:58] i know yuyv422 is packed but it is also packed inside an AVFrame? [18:58] Of course. Otherwise it would yuv442p [18:59] but the only different there is in the bitstream not representation [19:00] so ffmpeg does not have a way of representing all the YUV formats in one common way? [19:00] AVFrame [19:00] jfmcarreira: there is no common way. [19:00] no, because they're _different formats_ [19:00] if you need a specific format use swscale [19:01] nope. i want the original pixels. my software supports different pel formats [19:01] what is pel? [19:01] a typo of pixel [19:01] palette? [19:01] so is there any way of getting the pixels from the AVFrame in raw format like they would be in a bitstream? [19:01] jfmcarreira: libavcodec always outputs the most "original" format it supports [19:02] "bitstream" [19:02] wtf does this even mean [19:02] wm4: what you mean by most "original" [19:02] you can get the original undecoded data as AVPacket [19:02] wm4: i meant in the file [19:03] are you reading raw images with libavformat? [19:05] oh, pel as in halfpel, qpel [19:09] live i am reading raw and other filetypes like avi or bmp or png [19:10] iive: and after demuxing using the ffmpeg and decoding the frames i need to copy the pixel information to a struct i know in my software [20:25] git rejects commits that not ends with new line, even for files in tests/ref/fate/, is it bug or is it intended. I see there files that not ends with \n [20:26] I meant commits with files that not ends with new line [20:35] ffmpeg.git 03Michael Niedermayer 07master:7656c4c6e66f: avcodec/utvideodec: fix assumtation that slice_height >= 1 [20:47] lukaszmluki, see http://people.videolan.org/~jb/tmp/update if you want to fix it [20:53] OK, I will try [21:01] is there a problem with forcing this \n? [21:01] i think even git warns about it by default [21:17] it is not a problem, but you need a dummy printf("\n") in c file to produce that new line [21:17] ffmpeg.git 03Thilo Borgmann 07master:48c29883fcb2: lavd/avfoundation: Use internal av_strtok instead of std lib strtok [21:31] egrep -v '/$|^tests\/ref\/|\.diff$|\.patch$' in check_ending_newline should fix it [21:31] i dont mind adding this new line, but probably updates for existing files will be rejected too [21:54] ffmpeg.git 03Jon Morley 07master:8c28a39c2c38: options_table.h: min value for colorspace is 0 (AVCOL_SPC_RGB) [22:18] is there any opposite to punpcklbw m1, m0, [zero]? [22:22] packuswb? [22:31] yes it seems so [22:55] kierank: your patch doesn't apply cleanly [22:55] of course it doesn't [22:56] cf "This file is part of Libav." [22:56] dunno why it's diverged so much though [22:57] oh it's libavfilter [22:57] of course [22:58] that's quite different [22:58] what was the thing about tinterlace vs interlace already? [22:58] licenses? [22:58] features? [22:59] both seem to be gpl [23:00] random junk in tinterlace it seems [23:08] jamrial: ah didn't know about pavgb [23:10] kierank: isn't it easy to add slice threading as well? [23:12] probably yes but i've never done that before [23:14] does it help to have [pw_1] in a reg? [23:14] pavgb is better [23:14] i wasn't aware of that [23:14] well i assume it's better [23:14] ah right [23:15] oh wait pavgb does a >> 1 [23:15] iirc it rounds [23:16] ah mmh forgeet [23:16] forget it* [23:25] you might use ssse3's pmadduswb with proper coeffs, doing the unpack and sum in one step [23:25] well, not exactly [23:26] bah, there are 3 elements in the sum, probably not going to help [23:49] ffmpeg.git 03Michael Niedermayer 07master:5dcb99033df1: avcodec/wmaprodec: Fix integer overflow in sfb_offsets initialization [00:00] --- Tue Nov 11 2014 From burek021 at gmail.com Wed Nov 12 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Wed, 12 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141111 Message-ID: <20141112010502.80B2A18A02DA@apolo.teamnet.rs> [00:10] lukaszmluki, please provide a tested patch, not a untested grep line [00:14] or maybe the grep line is tested but if i copy and paste that into the hook the result may or may not be what you had in mind and thus would not be tested [00:31] OK, just wanted to know [00:33] lukaszmluki, why does your grep line escape the / ? [00:36] hmm, it works both way, escape is not needed [00:37] I used too much sed recently :P [00:39] np, iam just trying to make sure iam not missing something or messing the test up [00:39] Action: ePirat is too stupid for git send-email [00:39] I'm sorry. [00:48] lukaszmluki, ubitux posted a new patchset for the hook, please review and test if you like. I will submit it to root at vlc tomorrow or so [00:50] ok, thanks [01:05] ffmpeg.git 03Lukasz Marek 07master:758a66dc16eb: ffserver_config: drop requirement video size being multiple of 16 [01:07] kierank: you can't use pavgb, but you can use pavgw [01:08] jamrial: but the shift is >> 2 ? [01:08] pavgw, then psrlw mN, 1 [01:09] oh that's true I guess [01:10] just checked, you save two instructions and the need to use pw_1 [01:10] pavgb would be much better, but there's no "psrlb" [01:16] ok it's a bit faster [01:54] ffmpeg.git 03Carl Eugen Hoyos 07master:685f7227dc83: tests/Makefile: New try to fix fate-ffprobe with --target-path. [03:46] ffmpeg.git 03Michael Niedermayer 07master:35dcc8a04057: avcodec/lagarith: fix integer overflow [03:48] ffmpeg.git 03Michael Niedermayer 07master:48efe9ec86ac: avcodec/utvideodec: Fix undefined behavior in shift [09:06] 'morning [10:42] ffmpeg.git 03Martin Storsj? 07master:b776113e5d4a: v4l2: Unify one instance of reading/storing errno [10:42] ffmpeg.git 03Michael Niedermayer 07master:a83af3fc33a6: Merge commit 'b776113e5d4a56759615196de98efe802e95a6b6' [11:43] ffmpeg.git 03Rong Yan 07master:bb38cb14ccf4: libavcodec/ppc/me_cmp.c : factorize little and big endian code [11:43] ffmpeg.git 03Rong Yan 07master:cfaa2339629d: libavcodec/ppc/me_cmp.c : support little endian in sse8_altivec(), hadamard8_diff8x8_altivec() and hadamard8_diff16x8_altivec() [11:56] ffmpeg.git 03Changjiang Wei 07master:6f2068e626ee: avcodec/hevc.c: for big negative mvy value, should wait line 0 of ref frame due to edge extending [14:54] ffmpeg.git 03Marvin Scholz 07master:5e08b54f47e8: Icecast: always send a content-type [16:56] ffmpeg.git 03Marvin Scholz 07master:17dc39e76baf: Icecast: Use 100-continue if possible for proper error handling [17:24] anyone know if i can build a 32/64 dual binary for OSX using configure/make out of the box, or do i have to concat the binaries myself after? [17:26] I believe you need to lipo them yourself [17:26] (is anyone still on 32-bit OS X?) [17:27] apparently some idiots are still using first gen mac minis [17:27] some even use ppc still [17:27] weren't those PPC? [17:27] but why do you care [17:27] rcombs, no [17:27] mac mini was never ppc [17:27] huh [17:27] well, fuck 'em [17:27] i was told those are intel core (solo) processors [17:28] practically the last 32-bit cpu intel made [17:28] my ideology: if the OS vendor doesnt support it, dont bother [17:28] I mean, they still make 32-bit Atoms and such [17:28] my ideology: fuck users [17:28] I'm with Daemon404 [17:28] unfortunately I'm not always following my feelings [17:29] the only current case where it makes sense to make fat binaries for Darwin is on iOS, between armv7 and arm64 [17:29] and in a couple years you'll be able to go arm64-only on there [17:30] not looking forward to building every god damn external library twice, oh well [17:30] highly recommend saying "fuck it" [17:30] you know a(n annoying) guy sent patches for out of the box fat binary support before [17:30] (repeatedly, as a part of your morning routine) [17:30] and he got flamed [17:30] and he wrote a butthurt article about it [17:30] the good old days (before my time0 [17:31] no sense having it built-in when every dep needs to be built twice manually anyway [17:32] seen PPC+x86+x86_64 fat binaries? [17:32] (they're a thing) [17:32] http://www.hackerfactor.com/blog/index.php?/archives/300-Ten-Little-Endians.html [17:32] the guy himself is insufferable and his blog is full of half truths, but its still good troll material [17:32] well his case was for x86+ppc, which is even more abstruse [19:41] ffmpeg.git 03Marvin Scholz 07master:2c0bf76bb335: MAINTAINERS: Add myself as maintainer for Icecast protocol [19:42] poor guy [19:57] < rcombs> and in a couple years you'll be able to go arm64-only on there <-- unikely [19:58] they went through three 32-bit instruction sets in four years [19:58] so I highly doubtly they'll stop just because they switched to 64-bit [21:04] ffmpeg.git 03Martin Storsj? 07master:52f954da7594: libavcodec: Unconditionally build xiph.o [21:05] ffmpeg.git 03Michael Niedermayer 07master:81e3f819bd1b: Merge commit '52f954da7594c31ad94c9bcb54290145b59b27f5' [21:16] ffmpeg.git 03Michael Niedermayer 07master:fe27aeaeab07: mpeg12enc: increase declared size of block function argument [21:16] ffmpeg.git 03Michael Niedermayer 07master:aeb1621d1f17: Merge commit 'fe27aeaeab07142b1acd2690c64ee6973bdd7eba' [21:24] ffmpeg.git 03Vittorio Giovara 07master:898e9a24ef13: mpegvideo: check mpv return value [21:24] ffmpeg.git 03Michael Niedermayer 07master:9311026ec7a0: Merge commit '898e9a24ef13d8c56b4abf4ee0af09cdb0343e2d' [21:32] ffmpeg.git 03Vittorio Giovara 07master:f349f4b5502c: mpegvideo: fix size of array [21:32] ffmpeg.git 03Michael Niedermayer 07master:d3127691665c: Merge commit 'f349f4b5502c94943c30001b8a4d75daded3281c' [21:35] ffmpeg.git 03Vittorio Giovara 07master:db71c4926d94: mjpegenc: fix argument size in encode_mb [21:35] ffmpeg.git 03Michael Niedermayer 07master:8d6b51b18206: Merge commit 'db71c4926d948717ce3b74253eb385dc43dcb14d' [21:37] ffmpeg.git 03Michael Niedermayer 07master:7c61e4b1a0f9: hpeldsp: Increase put_no_rnd_pixels_tab[][] size [21:37] ffmpeg.git 03Michael Niedermayer 07master:76bd8e16af90: Merge commit '7c61e4b1a0f9e9f5b7bc08879e9e101eb90b19ea' [21:46] ffmpeg.git 03Vittorio Giovara 07master:0e1ebfebc832: 4xm: drop unnecessary check [21:47] ffmpeg.git 03Michael Niedermayer 07master:7edb5eec4fd4: Merge commit '0e1ebfebc8326069732795698a82f3fea0742a54' [21:53] ffmpeg.git 03Lukasz Marek 07master:bb60142f562e: lavu/opt: check for NULL before parsing [21:53] ffmpeg.git 03Lukasz Marek 07master:4efc79649b95: lavu/opt: add support for binary defaults [21:55] ffmpeg.git 03Lukasz Marek 07master:b54effba4f17: lavu/opt: update tests [21:55] ffmpeg.git 03Lukasz Marek 07master:db6be5416caf: fate: add opt-test [22:04] ffmpeg.git 03Vittorio Giovara 07master:9e9be5a20c0b: hevc_mvs: prevent unitialized use [22:04] ffmpeg.git 03Michael Niedermayer 07master:4dd85093f119: Merge commit '9e9be5a20c0b36dce1cae11f5f5957886231a764' [22:04] ffmpeg.git 03Michael Niedermayer 07master:a6defd1f5b1b: hevc_cabac: decrease CABAC_MAX_BIN [22:04] ffmpeg.git 03Michael Niedermayer 07master:261d2eaa7f32: Merge commit 'a6defd1f5b1bffcea7aa00ff379a6602cdaf2d05' [22:07] ffmpeg.git 03Jernej Fijako 07master:1bdd21d97528: dvbsubdec: add missing break [22:07] ffmpeg.git 03Michael Niedermayer 07master:db36b6e2ac51: Merge commit '1bdd21d97528d870fbb4388e837abaf390f2f7d7' [22:28] ffmpeg.git 03Vittorio Giovara 07master:2383323661f3: dvbsubdec: improve error checking [22:31] ffmpeg.git 03Michael Niedermayer 07master:6229f7823e2a: avcodec/dvbsubdec: Add return code to save_subtitle_set() [22:31] ffmpeg.git 03Michael Niedermayer 07master:4809ac75fcef: Merge commit '2383323661f3b8342b2c4d356fcfe8c5d1b045f8' [22:31] ffmpeg.git 03Michael Niedermayer 07master:2813dabdd496: avcodec/dvbsubdec: use av_freep() for saftey [23:18] ffmpeg.git 03Vittorio Giovara 07master:443502aed8b8: dvbsubdec: move shared codepath [23:18] ffmpeg.git 03Michael Niedermayer 07master:e0c36f58257f: avcodec/dvbsubdec: Fix 8bit non_mod case [23:18] ffmpeg.git 03Michael Niedermayer 07master:287eb69973f8: Merge commit '443502aed8b814d883825e52e91e4f018955aa66' [23:18] ffmpeg.git 03Vittorio Giovara 07master:60b055133485: png_parser: fix size of chunk_lenght [23:18] ffmpeg.git 03Michael Niedermayer 07master:70cbb473608b: Merge commit '60b055133485891405722bc9722e2c74fc9764b8' [23:19] ffmpeg.git 03Michael Niedermayer 07master:4cb9f1a77432: h264_cabac: fix one fill_rectangle() indentation level [23:19] ffmpeg.git 03Michael Niedermayer 07master:69b2d43d6afe: Merge commit '4cb9f1a77432de6f368df69bebbc082368a88c86' [23:54] ffmpeg.git 03John Stebbins 07master:1b667269062e: h264_parser: don't stop on SPS_EXT in split [23:54] ffmpeg.git 03Michael Niedermayer 07master:940f5c08e512: Merge commit '1b667269062eb6aec0b8726393ea91b7f7f57fde' [00:00] --- Wed Nov 12 2014 From burek021 at gmail.com Wed Nov 12 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Wed, 12 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141111 Message-ID: <20141112010501.770FB2AD60C8@apolo.teamnet.rs> [00:29] Hey guys, how can I convert MKV with SSA (or ASS) to MP4? [00:36] Sagi_ ffmpeg -i input.mkv -c copy out.mp4 [00:40] if there any way to disable libdvdcss key caching at compile time? [00:41] under mingw/windows, if fontconfig is enabled and saving to appdata mplayer crashes [00:41] not sure why [01:54] is there a simple way to extract aac audio without decoding? [01:54] form a vid? [01:54] and then I simply want to 'copy' that into antoher vid, without re-encoding (in other words, i want to copy audio from one vid to another) [01:55] ffmpeg -i video1 -i video2 -map 0:a -map 1:v -c copy outvideo [01:57] if video2 exists as still frames, would this work? [01:57] ffmpeg -i video1.mp4 -i video%04d.ppm -map 0:a -map 1:v out.avi [01:57] sure, you'll have to encode the video though [01:57] well, better would be if i had the audio extracted separately [01:58] then i could just 'add' that to encoding the second vid as audio stream [01:58] mplayer -dumpaudio video.mp4 supposedly dumps 'raw' aac stream [01:58] You can do that. [01:58] can I 'copy' that into the second vid as I encode it? [01:58] yes. [01:58] using -i video.aac ? [01:59] Either with video.aac or with the actual video file. [01:59] ok, thanks. i'll try. [01:59] Just map the audio stream from the one video and the other video stream. [02:00] so, if i also wanna encode frames, something like: [02:00] ffmpeg -i video1.mp4 -i video2%04d.ppm -map 0:a -map 1:a -vcodec libx264 out.avi [02:00] ? [02:01] -map 0:a -map 1:v -c:a copy -c:v libx264 [02:01] ok, thanks for the tip. [02:01] (i'm still a n00b in ffmpeg) [03:28] I am using https://github.com/carsonmcdonald/HTTP-Live-Video-Stream-Segmenter-and-Distributor/tree/master to generate TS files for HTTP Live Streaming. I am using a command like ffmpeg -i %s -f mpegts -acodec libmp3lame -ar 32000 -ab 48k -s 480x480 -vcodec libx264 -b 110k -flags +loop+mv4 -cmp 256 -partitions +parti4x4+partp8x8+partb8x8 -subq 7 -trellis 1 -refs 5 -coder 0 -me_range 16 -keyint_min 25 -sc_threshold 40 -i_qfactor 0.71 [03:28] -b:v 110k -maxrate 110k -bufsize 110k -rc_eq 'blurCplx^(1-qComp)' -qcomp 0.6 -qmin 10 -qmax 51 -qdiff 4 -level 30 -aspect 1:1 -r 10 -g 30 -async 2 - | %s %s %s %s %s to generate TS segment at varying bitrates. [03:28] The problem is that they all end up being larger than the specified bitrate by 35%-75% and the peak bitrate is 135%-160% over the target. [04:00] Sprry, got disconnected. What is the best way to ensure the birate stays as close to the target rate as possible with as little over in peak as possible? [04:16] minrate maxrate 2pass ? [04:27] i tried the minrate maxrate, but didnt try the 2 pass. i will go try that [04:57] c_14, you here? i tried your command: [04:57] ffmpeg -i video1 -i video2 -map 0:a -map 1:v -c copy outvideo [04:57] i get: Unrecognized option 'c' [04:58] i won't use pastebin sites for 1 line [04:58] 1 line is the command. [04:58] Then there's version stuff. [04:58] and 1 line is the output. [04:58] Then there's the error. [04:58] There's more than 1 line. [04:58] the error is 1 line. [04:59] Unrecognized option 'c' [04:59] Action: sacarasc goes away, then. [05:00] version number: ffmpeg version 0.8.16-6:0.8.16-1 [05:00] i think all the other data is irrelevant [05:03] 0.8 is ancient, upgrade. [05:06] In fact, 0.8.16 isn't even ffmpeg... It's Libav. #libav is their channel, but you can install/use ffmpeg on (I will presume Debian-like) https://trac.ffmpeg.org/wiki/CompilationGuide/Ubuntu or you can use one of these links: [05:08] yea, this is cr**ppy libav stuff... so i'll need to download the 'real thing' [05:08] for some reason, when i try 'apt-get ffmpeg' it says its at last version [05:09] well it's ffmpeg package from libav [05:10] Previous Debian and Ubuntu releases packaged libav as ffmpeg, but it's not ffmpeg. [05:15] I tried the minrate maxrate and 2pass, but when I was trying for 200kbps video and 48kbps video, the result was still about 320kbps total. [05:18] sacarasc: unfortunately :/ [05:19] i'll dl update [05:38] Does anyone know the best java library to use with ffmpeg? [05:44] I can't find anything modern on the subject. 2009 is the last post about it. [07:58] What is a good strategy for improving framerate of x11grab capturing? 6 of my eight cores are inactive (even specifying -thread 8 doesn't help) [07:58] It doesn't matter whether I steam or record to RAM with a tmpfs or to my ssd, I cap out at 33 FPS [07:58] -threads 8* [07:59] I've tried a variety of presets -- no difference, ultrafast and medium result in the same framerate [08:16] hi everyone, tried to convert basically : ffmpeg -i video.mp4 video.mpg , I had sound but no image, is this more complcated than that? [08:17] maybe... -vcodec ? [08:27] Video: h264, yuv420p, 1920x1080, 23.98 tbr, 23.98 tbn, 47.95 tbc [08:28] scoofy: is that not standard for an mp4 file? [08:28] it is [08:28] ok... [08:29] so I'm stuck [08:29] ;) [08:29] what is your goal? [08:32] my computer is too "light" to read without lagging the initial mp4 video [08:33] so I try to degrade it a little bit [08:33] you probably need to re-encode changing the profile type so your system can handle it [08:34] maybe you need to speicify some output codec? using -vcodec [08:34] simply re-muxing into a different container isn't going to help you [08:34] -vcodec mpeg4 or something [08:38] maybe I need to read more things about video encoding, althoug I only wanted to quickly convert ;) [08:39] p4plus2 - pastebin your command [08:40] what? [08:40] FabTG: you need to re-encode or use a faste device probably. I wouldn't switch from x264 personally - You would need to change your profile to baseline or main or something (its probably on high) [08:42] blippyp: http://p4plus2.pastebay.net/1520675 [08:42] blippyp: ok [08:42] this is one of many attempts but since all resulted in identical frame rate, that should do [08:43] also my configuration: ./configure --enable-gpl --enable-nonfree --enable-pthreads --enable-libx264 [08:44] also output from a few seconds of recording: http://p4plus2.pastebay.net/1520678 [08:45] p4plus2: you're streaming it - good luck with that - I've personally had very little success myself with that... [08:46] blippyp: it doesn't make a difference whether I record or stream [08:46] but my hardware sucks, which is what I think the problem(s) mostly come from for me [08:47] I don't think mine is hardware (or when steaming network related, I have a 128/15mbps connection) [08:47] if you remove all that extra garbage and just do a -crf 24 -preset ultrafast, it should work fine I think - even on my crappy hardware I can do it. [08:47] my cpu is a 3.8GHz octocore [08:47] but obviously you will want to re-encode it afterwards [08:48] there can be many bottlenecks - you simply might just need a hardware accelerator - and I don't think that's supported, but I could easily be wrong [08:49] I guess I'll just have to keep trying [08:49] p4plus2: this is one I use on my netbook: [08:49] ffmpeg -s 1024x600 -f x11grab -i :0.0 -r 30 -c:v libx264 -crf 17 -preset ultrafast -an out.mp4 [08:50] its odd because I can even record at medium or higher without a change in framerate [08:50] I'm even using a crf 17 - but the ultrafast is what makes it happen I think [08:51] I have also streamed it as well, but the results were less than ideal for me [08:52] blippyp: dropping the -r just causes massive frame drops -- like 55 of 60 per second [08:53] -threads don't work with x264 - it's uses autodetects how many or only uses one or something I believe [08:53] I already dropped -threads too [08:53] I've seen my audio configuration DESTROY my capture - maybe you need to play with that [08:53] the threads won't make a difference, just remove that option [08:54] I have removed it, also I'm not even capturing audio [08:54] you're also doing a large video - you should try dropping it to at least 720p first and try to get that working I think [08:55] changed to 720p -- no difference [08:55] huh: Stream #0:0: Video: rawvideo (BGRA / 0x41524742), bgra, 1920x1080, 29.97 fps, 29.97 tbr, 1000k tbn, 29.97 tbc [08:55] I wonder why....? [08:56] you need -framerate 30000/1001 [08:56] that -g might be hurting you - not sure, never used it, but from what I just read in the manual, I'd bet it's really high [08:56] and -r 30000/1001 [08:57] scoofy: I am trying to record ideally at 60fps -- the problem is I am only getting 30 [08:57] but my system isn't really being taxed so I feel like I am missing an option [08:58] you're trying to record screen grab? [08:58] x11grab, yeah [08:58] and stream it [08:58] the stream part can come later [08:59] since it doesn't matter whether I save it to RAM/SSD or stream -- all three have the same frame rate limit of 38 [08:59] that's weird [08:59] why do you need such high frame rate? [09:00] I was doing a bit of game streaming and unfortunately they look rather choppy at 30fps [09:00] had enough complaints to try and do something about it [09:01] and frankly even I notice it to an extent [09:01] by going 60 you are literally requiring twice the resources man [09:01] not worth it in my opinion for the 2/10 guys who can 'apparently' notice the difference [09:01] blippyp: considering 720p and 1080p are both limited to 38FPS.... [09:02] that tells me the problem is elsewhere [09:02] I agree, something 'fishy' is happening for you I think [09:02] wait and ask c_14 - that guy really knows his shit [09:02] using the native ffmpeg on my system is even worse at 17FPS [09:03] he's usually on during the day (for me) and it's 3 in the morning right now here [09:03] recompiling it from scratch hit 38 right away [09:04] I don't know what to tell you - even with my crappy systems I've played with 60fps a couple of times for shits and giggles and as far as I know it worked fine - the files stated so - i couldn't see a difference [09:04] apparently the source files I used were 60fps to begin with so unless that was bullshit....??? [09:04] as far as I'm concerned anything over 30 is only useful if you want to slow-motion your video [09:05] human eye can hardly perceive much more than that [09:05] 'apparently' [09:05] all the 'cool' kids are claiming otherwise these days [09:05] I call bullshit myself, but whatever - don't really care... ;) [09:05] i wouldn't give 2 shits for streaming video [09:06] being "only" 30 fps... [09:06] hell - I'd drop it down to 24 just cause it's easier on resources.... haha [09:06] scoofy: I'll be honest I feel its more to do with the vsync isn't picking "every other frame" so to speak [09:06] could be - not sure - I never use that option either [09:06] iirc NTSC is at ~60 FPS, that get halved via interlacing so you get ~30 FPS [09:07] (but this is not NTSC) [09:07] just that also works at 60 fps i think [09:08] yeah - never considered that - maybe the source I used to 'play with 60 fps' was interlaced.... no idea... [09:09] maybe that's why I never noticed - could be limited by my crappy monitors as well... haha [09:09] I know my tv goes to 60 though, and I've never noticed a difference between watching videos on the computer compared to the tv.... [09:10] I've always just assumed I was part of the lucky 99% who can't tell either way... ;) [09:10] I generally don't notice a difference between 30 and 60 for the most part [09:11] FabTG: you still around? [09:11] but it does make a difference in the recordings I've done, which makes me think the "30 of 60" frames its not picking frames evenly so to speak [09:11] does that make sense? [09:12] have you used -framerate instead of -r? [09:12] I'd say your system can't handle recording that much myself [09:12] -framerate is for input [09:12] -r is for output [09:12] iirc -framerate is supposed to be used with x11grab, it will duplicate frames though if your throughput is too low [09:12] isn't it? [09:12] klaxa: I haven't, I'll try it [09:12] p4plus2, -framerate specifies input rate... so if your *input* rate is 60, change -r to that [09:13] he might need to specify both input and output frames though... [09:13] yep [09:13] -framerate 60 -r 60 [09:13] oh wow [09:13] assuming he wants 60 fps output [09:13] would you look at that [09:13] or -framerate 60 -r 30 [09:13] working now? [09:13] ^ for 30 fps output (but that wouldn't make sense for 60 fps screen grab) [09:13] that indeed did the trick [09:13] significantly better now [09:13] better check the encoded video, like i said, it might duplicate frames [09:14] i had a similar issue with an NTSC video (audio out of sync) [09:14] good call klaxa [09:14] specifying both -r and -framerate fixed it [09:14] klaxa: I will but this time the "stream 0" says 60FPS when before it was always 30 [09:14] so it was incorrect [09:14] until you specified -framerate 60 [09:14] could be 60 times the same frame per second [09:15] scoofy: correct [09:15] not sure how to tell the difference though [09:15] pull two consecutive frames out and md5 them??? [09:15] i think i sometimes get a "duplicate frames" warning automatically [09:15] or was it in mencoder? not sure [09:16] isn't there a flag in the codecs to join duplicates? [09:16] not sure x11grab reports duplicate frames in that case, mostly i only get dropped frames [09:16] I have dup=0 [09:16] if he ran one of thoseont he file afterwards and it dropped back down to 30 that would be a good sign [09:16] so I think I am safe [09:16] wouldn't that result in variable frame rates? i wouldn't trust a streaming service with that [09:18] also, checking for duplicate frames might actually be easier by hand, by stepping through images, checking md5 hashes will probably not work since the codec might change the frame even if the input didn't change unless the codec was lossless [09:18] visually it looks much better, plus I show dup=0 [09:19] drop=10 after five minutes or so now which is reasonable enough [09:19] yeah - that's why I brought up the option with the codecs stripping duplicates - I also doubt the md5 method would work either [09:20] how are your resources - are they maxing out? [09:20] they look fine [09:20] if they're not - then you're still running into some kind of conflict somewhere probably [09:20] it shouldn't drop frames [09:20] its not [09:20] (00:20:52) p4plus2: visually it looks much better, plus I show dup=0 [09:20] you just said it dropped 10 frames...??? [09:21] 10 over five minutes [09:21] that could be easily attributed to a random lag spike [09:21] dup=0 is probably a good sign [09:21] where are you writing to? SSD? maybe writing to ram would be faster [09:21] maybe check your io while recording [09:21] ya, I suppose the lag could account for it... [09:21] you're streaming it also right? [09:21] klaxa: I'm writing to a stream, but tmpfs works just as well now [09:22] ah [09:22] 128 down and 15 up so I think thats pretty solid for streaming [09:22] well, either way much better - cool [09:22] tell everyone you couldn't figure out how to switch to 60 and watch them all still complain about it... [09:22] hahaha [09:22] yeah this problem has been bugging me for quite a while, I don't know how I never came across -framerate [09:23] blippyp: I'll probably try that and see what happens [09:23] I would - just to see if they actually noticed [09:23] plus I enjoy screwin' with people... ;-P [09:23] I may say nothing to see if anybody says something first [09:23] seriously you should - just as a social experiment [09:24] but there will likely be someone looking at the numbers who will speak up first [09:27] anyways thanks for the help everybody, much appreciated [09:27] yw [09:28] FabTG: I'm taking off - gonna go do some programming, but if you read this - You don't have to encode your entire video to see if that profile changing will work for you (although I'm betting it will) - You can just record a short clip (like a couple minutes) to test and see how your system handles the difference. [09:30] > my computer is too "light" to read without lagging the initial mp4 video [09:31] i would suggest using a player that supports caching and if available hardware decoding [09:32] that's what I do - my systems are crap for playback these days also (everything on a high profile now) - and I couldn't be bothered to re-encode every movie I download - but my blu-rays all play them fine luckily... :) [10:18] can I use ffmpeg to create a virtual camera from my desktop? [10:30] hi, I have some questions about https://github.com/FFmpeg/FFmpeg/blob/master/LICENSE.md#incompatible-libraries [10:31] I built ffmpeg binaries, which are statically linked against the ffmpeg libraries and some 3rd party libraries, including libx264 (GPL) and libfdk-aac (nonfree) [10:32] would those ffmpeg binaries really be unredistributable? could I get in trouble for redistributing them? or am I overreacting? [10:46] hans_s: yes, because libfdk-aac is non-free [10:47] (yes to the first two questions) [10:48] yes, it's marked nonfree, but I didn't find any explanations, why. also I'm a noob when it comes to licenses, that's why I'm asking [10:49] google the license for each and then google why they're not compatible [10:50] for example, I have no idea what license I'm breaking (ffmpeg/libx264/libfdk-aac) [10:51] ok, thought maybe someone already knew the answer, before I start researching [10:52] the license info should be in the source dir for each. [10:56] in short the libfdk-aac's license isn't compatible with the gpl [10:56] so, I'm violating GPL, if I redistribute that binaries? [11:07] yes [11:12] ok, thanks a lot, relaxed [11:29] hi [11:30] I am trying to wrap my head around ffmpeg in general, so as a start I am trying to get the following to work: [11:30] ~$ ffmpeg -f v4l2 -i /dev/video1 - | mplayer - [11:30] of course this doesn't function because it needs to know how to format the stuff off the camera [11:31] the source is this: http://paste.scsys.co.uk/438728 [11:31] I want to get the h264 output and *without processing* send it to mplayer [11:31] is this possible? [11:58] I have a stereo mp3 file that contains voice recording, it's stereo and I want to have it converted to mono, is https://trac.ffmpeg.org/wiki/AudioChannelManipulation#a2monostereo the right approach? [11:58] I'm at the same time transcoding it to AAC using -c:a libfdk_aac -b:a 48k [11:58] should I do it in two steps? [11:59] sorry I meant https://trac.ffmpeg.org/wiki/AudioChannelManipulation#stereomonostream [11:59] that is just -ac 1 [12:09] I have this output for option at start of transcoding "Stream #0:0: Audio: mp3 (libmp3lame), 11025 Hz, mono, s16p" for the mp3 and "Stream #0:0: Audio: aac (libfdk_aac) (mp4a / 0x6134706D), 11025 Hz, mono, s16, 48 kb/s" for the AAC [12:09] yet the mp3 is roughly half the size of the m4a [12:09] ribasushi: read https://trac.ffmpeg.org/wiki/Capture/Webcam first, but something like, ffmpeg -f v4l2 -i /dev/video1 -c:v libx264 -f flv - | mplayer - [12:10] oh I figured, I was using -q:a 8 for the mp3 encoder and that resulted in an average bitrate of 24kb/s [12:11] relaxed: I actually got things to record fine with: [12:11] yes, the range is 0 - 9, 0 being highest quality [12:11] now using an explicit -b:a 48k [12:11] relaxed: ffmpeg -report -s 1920x1080 -f v4l2 -vcodec h264 -i /dev/video1 -copyinkf -vcodec copy test.mp4 [12:12] works in the sense it does not consume any CPU [12:12] and writes at about 3mbit/s [12:12] relaxed: is 11025 Hz/mono/48 kb reasonable for a talked podcast? [12:12] Action: ribasushi will wait for anddam to get help ;) [12:14] anddam: should be, what do your ears tell you? [12:14] my ears say that mp3 24kbps is a bit shitty [12:14] AAC 48kb sounds very good [12:14] I mean both mono 11Khz [12:15] ribasushi: it's because you're stream copying, which is optimal [12:15] relaxed: I got more questions though, just didn't want to cross-talk ;) [12:16] is MPEG the appropriate container for an AAC encoded audio stream? [12:16] I guess that's actually "MPEG-4" [12:16] anddam: sure [12:16] ribasushi: shoot [12:17] relaxed: so the above still doesn't flow into mplayer correctly (I get nothing on the screen), but even if I just save to a file and play it after - the video is twice as fast [12:17] what do I need to adjust for that? [12:19] ffmpeg -report -s 1920x1080 -f v4l2 -vcodec h264 -i /dev/video1 -copyinkf -vcodec copy -f flv - | mplayer - [12:19] the mp4 container has to be written completely before you can play it [12:19] right, I get capture (camera is on, stats are rolling) and no videoscreen [12:20] err I am aware of that [12:20] I mean once I write out the container [12:20] (ffmpeg stops) [12:20] the resulting mp4 is twice the speed [12:20] oh, does ffplay play it? [12:20] let me check... [12:20] you may have to set the input frame rate [12:21] ok wtf... things work now [12:21] Action: ribasushi chalks up to a fluke moves on [12:21] so next question: [12:24] relaxed: https://etherpad.mozilla.org/BFPkw4sg6n [12:24] (I'll be updating it as I go) [12:24] why the error / what am I doing wrong [12:25] https://trac.ffmpeg.org/wiki/Capture/ALSA [12:30] relaxed: this got me on the right track, thank you very much [12:30] still fiddling with settings a bit... [12:36] you're welcome [12:44] weird, "-strict 2 -vn -sn -c:a libfdk_aac -b:a 32k -ac 1 -ar 11025 scaled.m4a" results in "ID_AUDIO_BITRATE=32000" [12:44] while "-strict 2 -vn -sn -c:a libfdk_aac -b:a 32k -ac 1 -ar 11025 scaled.aac" results in "ID_AUDIO_BITRATE=29888" [12:50] hello all, I am writing a script that runs ffmpeg to convert a video from mpeg2 to mpeg4 and have added a function to remove the file after successful completion, but when I quit the transcoding no error is returned and my file gets deleted. Is there a way to tell if ffmpeg has been told to quit rather than just finished successfully. Thanks in advance for any help. [12:52] btw I am checking the exit code (with $?) to do this [12:56] you could compare duration of the files maybe? [12:57] Ahh that's a good catch I will try that Thank you [12:58] checking whether or not ffmpeg was stopped because the file ended, it received a "q" or received a SIGINT seems rather hard [12:58] Yes, I am finding that out :) [12:58] wow what's with my english... i'm really tired [13:02] relaxed: the alas-capture page showed me how to get a stable name for the audio [13:02] relaxed: how do I get the same for v4l2 (I don't want to use -i /dev/video1) [13:02] (I did look on the wiki - either I am blind, or stupid, or it isn't there ;) [14:15] bye, thanks [14:34] relaxed: just wanted to let you know that I got what I wanted so far [14:35] ffmpeg is awesome [14:35] ffmpeg -y -report -f v4l2 -vcodec h264 -video_size 1024x576 -i /dev/video1 -f alsa -itsoffset 0.3 -ar 44100 -ac 1 -i "default:CARD=Intel" -c copy -f matroska - | netcat 192.168.0.11 30000 [14:46] yo [15:09] yo c_14 [15:10] i have a ts stream where there's 4 substreams [15:10] 1 video 2 audio and one subtitle [15:10] how to pickup one audio using -map? [15:15] dont wory i got it [15:15] mkv is so dick:P [15:15] can accept any codec! [18:09] hi o/ [18:22] i have a small problem [18:22] anyone can help me ? :D [18:22] [Parsed_subtitles_1 @ 0x686b80] fontconfig: cannot find font 'Arial', falling ba [18:23] https://zerobin.net/?66ceb91390d077b5#9IMXbv+cH2hjmNlcbCvj5x+ff94xCzuJZroifMFI/OA [18:30] Will: And? [18:30] and ... [18:30] it's not a error ? [18:30] No. [18:30] It's a warning. [18:31] It's going to use a different font. [18:31] before i don't have this error [18:32] (i reinstate ffmpeg today) [18:34] this warming * [18:57] Hello people [18:58] I was wondering, if I capture a dshow source from FFMPEG in 720x576 25fps, but the source connected to that capture device is acutally 288p 50fps progressive scan, how do I make the output be the same 288p 50fps format again? (Like DSCaler's "Old Game" filter) [19:00] (for additional information: I'm capturing some SNES stuff) [19:02] I believe the correct description of what I'm trying to do would be "Each field gets its own 720x288 frame" [19:02] -vf yadif=send_field [19:04] that's yet another de-interlace filter, and this option make it output one frame for each field. [19:07] that was quick [19:07] thank you :) [19:23] when i use -acodec opus it works fine normally, but when i am trying to use it while concat-ing two files i get this: Unable to find a suitable output format for 'opus' [19:23] opus: Invalid argument [19:42] trying libopus instead of opus doesn't help [19:44] oh my bad, it was due to bad syntax [19:55] is there any way to have ffmpeg start all inputs etc, but to actually begin the output recording on a signal or somesuch? [19:55] I am writing a script that starts recording from multiple sources (one of them is not ffmpeg based), looking for ways to sync them at record time [20:20] Hello people... If I try to reencode a video, in middle of which there are zeroes instead of frames, the output has replicated frame... How can I just omit the missing frames thus jumping to the next good frame? Thanks [21:45] ok this is aggravating [21:46] I have one file with a perfectly synced audio/video track, and I am trying to PIP it with another video-only file [21:46] like so: [21:46] http://paste.scsys.co.uk/438821 [21:46] the problem is that the resulting audio is out of sync even with the image it was synced with [21:46] (by a large ~1 sec margin) [21:46] what am I doing wrong? [22:34] hi, if I want to create a video composed of photos , I can use a command like this: ffmpeg -framerate 0.33 -pattern_type glob -i '*.jpg' -c:v copy out.mp4 [22:35] the problem is that the framerate should be higher, in order for the video to 'refresh' faster that the photos change from one to another. [22:35] how can I achive that ? And what's the name of such operation ? [22:35] What? [22:36] Can you explain? [22:37] f.i. if I play mplayer out.mp4 , the video is somewhat problematic, becasue f.i., pressing arrow left or arrow right , does not render correct images [22:38] maybe it is lacking an index, but maybe also the framerate should not be the same frequency at which I want the photos to alternate in the video [22:38] On mplayer arrow left and arrow right will jump n seconds to the nearest keyframe. [22:38] stop hl please [22:38] xD [22:39] harovali1: https://trac.ffmpeg.org/wiki/Create%20a%20video%20slideshow%20from%20images [22:39] You can increase the fps of the output if you think that'll help. [22:47] c_14 tanks [22:49] sometimes I get '100 buffers queued in output stream 0:0, something may be wrong.' , however the video gets generated. What does this signal? [22:49] I'm not entirely sure, but you can probably ignore it. [22:52] anyone has ideas about my question with the desyncing audio? [22:53] No clue, sorry. Might want to wait around or ask the ml. [22:58] aight [23:08] I've got an error about non-monotonous DTS http://dpaste.com/0MPXDF6 , any suggestions on how to fix it? [23:10] That's a warning, not an error. [23:12] ok, I still need to fix it, the video stalls when the error appears [00:00] --- Wed Nov 12 2014 From burek021 at gmail.com Thu Nov 13 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Thu, 13 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141112 Message-ID: <20141113010503.4556A18A02F2@apolo.teamnet.rs> [00:03] ffmpeg.git 03Luca Barbato 07master:c6074a30ba3b: opt: Fix the documentation mentioning av_set_string3 [00:03] ffmpeg.git 03Michael Niedermayer 07master:1e16492b989e: Merge commit 'c6074a30ba3b5fb4319ee6ee599656d58548cdc8' [00:12] is anyone opposed to the fact that i take over xbr code? [00:20] ffmpeg.git 03Luca Barbato 07master:c9c7d59b7d26: tiff: Use av_mallocz_array [00:20] ffmpeg.git 03Michael Niedermayer 07master:dadc1f5ee9d9: Merge commit 'c9c7d59b7d26f0328d612995dd9256337ae1cbfb' [00:23] ubitux, you should ask arwa, as she ported the code otherwise no objections and thanks! [00:23] ok [00:33] ffmpeg.git 03Vignesh Venkatasubramanian 07master:597d826123dc: lavf/webmdashenc: Representation IDs should be unique. [00:50] damn, we can do so much better than this code... [00:50] gonna take me a little while though [00:54] Action: Daemon404 pats ubitux [01:05] ffmpeg.git 03Marton Balint 07master:eaf4ab9802d5: ffplay: signal the frame queue before closing audio [01:05] ffmpeg.git 03Michael Niedermayer 07master:12aab852c58d: Merge remote-tracking branch 'cus/stable' [01:07] ubitux, btw, that huge rgb/yuv table is ugly ... [01:07] i know [01:08] again, i did that to be bitexact with the (modern) reference [01:08] its not bitexact to the xbr reference code i saw [01:11] and i dont think the refernce makes that much sense [01:38] michaelni: yes, we can probably replace it with direct computation [01:38] last time i tried (using the colorspace macros) it was a bit slower though, but well maybe not worth the lut [03:40] ffmpeg.git 03Michael Niedermayer 07master:9d6ad68fa445: avcodec/h264_parser: Avoid adding SEI to the global header [07:00] valgrind doesn't like the new opt test [08:44] where to find the code that the parameter "-re "? is set? [08:45] I have debugged and can'f find where the "-re" parameter works [08:45] I guess it maybe at the function open_input_file [08:46] thanks in advanced [08:46] git grep '"re"' [08:46] start from here [08:47] the git grep rate_emu [08:47] and you'll end up in ffmpeg.c, 2 occurences [08:48] ubitux, thank you very much!!! [08:48] I find it! [10:56] 'matin [13:00] ffmpeg.git 03Reimar D?ffinger 07master:5e8e2f3861df: configure: Hack to treat x32 as x86_64. [13:00] ffmpeg.git 03Carl Eugen Hoyos 07master:0ea54d698be6: lavd/avfoundation: Remove unused -frame_rate option. [13:53] ffmpeg.git 03Carl Eugen Hoyos 07master:1f4bce894a15: lavf/tcp: Clarify that the -timeout option takes microseconds. [14:59] > Edits are not restricted to fall on sample times. [14:59] damn mp4 [15:00] of course now [15:00] not [15:01] if i'm understand it well, it means you need to decode in order to slice a & v to match these edit list [15:01] yes [15:01] it's not something for a demxuer to do. [15:01] awesome& [15:01] similar to orderec chapters [15:01] same deal. [15:01] does that accuracy matters in practice? [15:02] depends how inaccurate i geuss? [15:02] i dont know why you would ever want to put timeline stuff in a demxuer though [15:02] application of it, at least. [15:02] well, it's supposed to be at format level [15:02] why? [15:03] timeliens arent format level [15:03] theyre metadata [15:03] i wonder how this can be honored properly in ffmpeg design [15:04] it cant be, at a demuxer level [15:04] the least bad idea i can think of is exporting it somehow [15:04] same for matroska [15:04] similar to other metadata (replaygain?) [15:05] so we would associate a timeline field in the AVStream? [15:05] it's ugly regardless [15:05] it gets uglier still [15:06] i don't really see a clean solution [15:06] cause currently, the mov/mp4 demuxer applyes the first edit (and no others) [15:06] which is really "Special" behavior [15:06] so exportign edits would be a "regression" [15:06] i was under the impression that the edit list could be honored at demuxer level by dropping packets (and scaling timestamps - there is a rate factor too) [15:07] in my matroska demuxer i seek around the stream, dropping wouldnt cover enough cases [15:07] if youre trimming by time, it's fairly obvious that you cant just drop packets [15:07] well i mean, building the internal index differently [15:08] like ignoring some referenced samples etc [15:08] i really think putting it in lavf is a terrible idea [15:08] ubitux: IMO there should be a high level API which handles such things [15:08] >high level [15:08] lavc and lavf are awfully low level anyway [15:08] >libav* [15:08] dohohohohoho [15:08] wm4: i'm fine with that TBH [15:09] most people just want video and audio frames from a file (or the reverse) [15:09] they don't want to connect dozens of components [15:09] i lvoe writing the 50 boilerplate loops, wm4 [15:09] dont hate [15:09] and such an API could handle ordered chapters and edit lists with ease [15:09] i like having control of how threading and buffering works for the loops [15:10] doubt any high level api would make me happy :D [15:10] so we could say that we add an avformat option, telling demuxers to export an AVStream->timeline and not try to drop the packets themselves [15:10] it would make many API users happy [15:10] and let the highlevel api deal with that? [15:10] ubitux: sure why not [15:10] that seems fairly reasonable to me [15:10] does mp4 do that per stream? [15:10] i wonder how this is going to affect... stream copy for instance. [15:10] wm4: yes [15:10] wm4, yes [15:10] disgusting [15:11] i guess some format do that at presentation level? [15:11] ubitux, well it should simpyl be unsupported for stream copy, no? [15:11] i guess... [15:11] ubitux, mp4 and matroska are at presentation level [15:11] ? [15:11] im not aware of what otehr formats have timelines [15:11] mp4 is at stream level [15:11] not presentation [15:11] oh [15:11] i thought you meant application [15:11] yeah it does it per stream [15:12] matroska is even more special [15:12] because of editions [15:12] and stuff [15:12] we don't have an opw minion to work on that? [15:12] i have enough projects currently :( [15:13] im not sure there is any opw candidate who *could* [15:13] like, actually be able to. [15:13] it's not easy [15:13] we need to know how much stuff we could put into these timelines [15:15] i suppose we have: external file reference, time ranges (anything using chapter references?), rate, loops?, anything else? [15:15] i cannot rememebr if its possible, but i thought you could mix N video streams into one (this might be out of spec) [15:15] ok [15:15] ffmpeg.git 03Rong Yan 07master:e74e14608fa3: libswscale/ppc/swscale_altivec.c : fix hScale_altivec_real() yuv2planeX_16_altivec() yuv2planeX_8() for little endian [15:15] well, let's export this as libavfilter filtergraph. [15:16] Action: wm4 barfs [15:16] that might actually work... :D [15:16] ubitux, btw i really like when the pts/dts is vfr AND the rate in edits varies [15:16] double vfr! [15:16] :) [15:16] i really wonder if i could just have a char *timeline with a libavfilter filtergraph to handle these timeline [15:16] hatst errible [15:16] you sure are going to hate me for that though [15:16] api users will kill you [15:17] :D [15:17] matroska has segment UID, and a list of segments: each with start and end time in nanoseconds, a segment UID reference, and the edition [15:17] i really wonder if i could just have a char *timeline with a libavfilter filtergraph to handle these timeline <- fuck no [15:17] Daemon404: it would even work for matroska! you could movie=OP.mkv into the filtergraph @_@ [15:17] Action: Daemon404 does not use lavfi api [15:17] nobody wants to give that lavfi shit too much control [15:18] be honest and design a high level API [15:18] well, we can imagine different export mode [15:18] the only it could be worse tahn exporting a filter string would be exporting tha tstring via av_log [15:18] only way* [15:18] lol [15:18] :D [15:19] actually i don't even need a new field [15:19] i can use the metadata system [15:19] Action: wm4 cries [15:19] :D [15:19] metadata? or sidedata? [15:19] i get confused [15:19] Action: ubitux laugh evily [15:19] Daemon404: av_dict_set(&st->metadata, "timeline", "trim=...") [15:20] hurr durr [15:20] btw ubitux [15:20] yes? [15:20] there is already a hack^Wavoption in the mov demuxer [15:20] to disable edit application [15:20] (even though it only applies teh first edit anyway...|) [15:20] yeah i'm on it [15:20] i started moving it in a data area [15:21] and move the adjustments into mov_build_index() instead [15:21] i'm probably going to try that experimental timeline thing just for the lulz [15:27] I suppose you could try something like the concat demuxer, which tries to do sample accurate cutting by using certain avpacket side-data [15:28] AV_FRAME_DATA_SKIP_SAMPLES [15:32] that works for audio [15:32] what about video with not much keyframes? [15:32] what about trailing samples for audio too? [15:33] yeah [15:33] ubitux: that side data handles trimming too [15:33] for video, no idea, maybe a flag whether the frame should be shown? [15:34] although I don't like this approach much, it'd allow for easy integration into existing code [15:34] clearly, exporting a lavfi timeline as stream metadata is a better solution [15:44] does anyone have samples with crazy edit lists? [15:44] i only have one here, and it's not that obvious [15:45] typically samples with stream timescale... [15:45] + edit lists [15:45] highlight Daemon404 [15:46] i have lots [15:46] not directly on me though [15:46] there's one iOS app that generates crazy-as-fuck edit lists [15:46] i can get you a file [15:47] lol an iOS app? [15:47] yeah [15:47] it recoeds 1 second of video every day [15:47] and stiches it into a video over N days [15:47] and it uses edit lists to sync stuff [15:47] woah [15:47] can you tell me the name of the app? [15:48] uh [15:48] not off the top of my head [15:48] let me ask the community person who told me [15:48] might be an hr or so [15:48] theyre in EST [16:06] ubitux, http://1secondeveryday.com/ [16:06] thats it [16:06] thank you [16:07] only the ios version does edit list stuff i think [16:10] yeah probably, it's the iOS framework which is responsible for these edit lists [16:11] yea [17:01] Daemon404: I can't imagine how I would feel if, after recording 7 years of 1-second-a-day videos, I miss /one day/. [17:55] ffmpeg.git 03Michael Niedermayer 07master:39dfe6801a3f: avcodec/dvbsubdec: Cleanup on *malloc failure [18:42] ffmpeg.git 03James Almer 07master:84ccc317cecf: x86/flacdsp: separate decoder and encoder dsp initialization [20:53] ffmpeg.git 03Michael Niedermayer 07master:d03867c24831: avcodec/dvbsubdec: av_assert* instead of assert() [23:36] ffmpeg.git 03Lukasz Marek 07master:5dc0f607e795: lavu/opt: fix memleak in test [23:36] ffmpeg.git 03Lukasz Marek 07master:5aed6f56d91b: ffserver_config: report not closed last tag [23:36] ffmpeg.git 03Lukasz Marek 07master:173d51c982f1: lavu/opt: fix av_opt_get function [00:00] --- Thu Nov 13 2014 From burek021 at gmail.com Thu Nov 13 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Thu, 13 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141112 Message-ID: <20141113010502.345A82AD60E6@apolo.teamnet.rs> [03:47] can ffplay show the frame rate? [10:11] for anyone interested - I solved my problem from yesterday by simply switching to matroska as intermediate containers (nothing changed in the codec/quality/pipeline) [10:11] with flv sources PiP desyncs stuff [10:11] with mkv sources - it works flawlessly [10:12] weird ;) [11:13] ribasushi jap? [11:13] or deu? [11:16] neither. [11:45] Hi guys! Using libvorbis  if it has anything to do with it  are the options '-aq' and '-qscale:a' have the same effect? I can't find any authoritative doc on this, hope you can help me. [11:50] arkonova: yes, use "-q:a" and look at --quality in "man oggenc" for the quality range. [11:51] relaxed: Great, thank you [12:59] Hello. Can someone explain the "-c copy" part in "ffmpeg -i www.sample.com/file.m3u8 -c copy file.ts" please? :) [13:00] Safa_[A_boy]: it means that ffmpeg won't do any processing on the audio/video/subtitles etc.. it would just demux the input , copy the packets verbatim and then mux the output [13:00] I didn't find the -c parameter in the man page.. [13:01] Thanks :) [13:01] if you want to do any processing , e.g. on the video (adding a text,box, deinterlace, etc) ffmpeg should first decode the video , the process the raw image and then encode it. [13:01] if you want to change the codec, you also need to decode and encode with the new codec. [13:02] hum... the full syntax is something like -c:v -c:a -c:s, for video, audio, subtitles etc... -c just assumes all of them. [13:02] it is in there, it is short for the -codec option. [13:03] Ok thanks very much for these great information! :) [13:04] Safa_[A_boy]: welcome from BH to FFMPEG! [13:04] lol i thought i'm the only one from the mena ;) [13:06] I am having some issues when muxing to HLS, after calling avio_close the file descriptors to the .ts files are still open, the file descriptor to the m3u8 file has been closed. Is there a known bug here? [13:09] DelphiWorld: O.o [14:31] is "MX Player" for android based on MPlayer ? [15:06] can someone please tell me what parameters I need to produce this output stream? "Video: h264 (High) (avc1 / 0x31637661), yuv420p(tv, bt709" [15:29] hmmmm have a tangentially-ffmpeg-related question [15:29] I have an rfbproxy capture [15:30] the following combo reproduces the colors *perfectly*: rfbproxy -l -p rec_vnc.rfb & xtightvncviewer 127.0.0.1:5910 [15:30] ggVGc: -c:v libx264 -profile:v high -pix_fmt yuv420p [15:30] however this pipeline generates "dull" colors (likely 601/709 loss or something): rfbproxy --framerate=30 -x rec_vnc.rfb | ppmtoy4m -v 1 -F 30:1 -S 420jpeg | yuvplay [15:31] my question is - how/if can I use ffmpeg's yuv4mpegpipe to reverse the incorrect colors [15:31] why don't you pipe directly to ffmpeg to simplify what's going on [15:32] relaxed: thanks, I figured that out. But will that be completely lossless? I would like basically -crf 0, but not 4:4:4 [15:32] relaxed: I am: rfbproxy --framerate=30 -x "$WORKDIR/rec_vnc.rfb" | ppmtoy4m -v 1 -F 30:1 -S 420mpeg2 | ffmpeg -y -hide_banner -f yuv4mpegpipe -i - ..... [15:32] relaxed: the "disassembled pipeline" is because I am trying to debug where the color goes wonky [15:32] lose ppmtoy4m [15:33] oh? how can I do that? [15:33] clearly there's a codec I missed experimenting with... [15:34] try ... | ffmpeg -vcodec ppm -f image2pipe -r $framerate -i - ... [15:34] hm, seems profile:v is not compatible with lossless encoding [15:35] lossless is high anyway [15:36] for visually lossless use -crf 16 (or lower) [15:36] relaxed: indeed this works to remove ppmtoy4m (thanks for that): ... -f image2pipe -c:v ppm -r 30 -i - ... [15:36] the color problem is still there however [15:37] relaxed: yeah, I know, but this is source material for video editing, so I am trying to keep all quality. But after effects will not import h264 with profile High 4:4:4, so I guess I won't get all that I wished for [15:37] anyway, works not with profile:v High [15:37] thanks [15:37] works now* [15:37] ggVGc: high444 is the lossless profile [15:37] yep, I read tht [15:37] but yeah, after effecrs doesn't accept it [15:38] ribasushi: can you have rfbproxy output to a file? If so, run "ffmpeg -i" on it and paste the results. [15:39] effecrs? [15:41] relaxed: is this what you asked? http://paste.scsys.co.uk/439411 [15:42] can you view rfbexport with ffplay? [15:42] for the record - rfbproxy doesn't really have many knobs to tweak, -x is pretty much all of it [15:42] Do the colors look right? [15:43] does "ffmpeg -i -i rfbexport" return the same thing? [15:44] minus an -i :) [15:44] relaxed: ffmpg -i rfbexport errors [15:44] relaxed: still trying to see how to make ffplay to run... [15:44] relaxed: I can upload the rfbexport somewhere, sec... [15:45] (compressing, will take a bit...) [15:49] ggVGc: oh, after effects! It's hard to believe it doesn't take lossless h264 [15:50] relaxed: I find that hard to believe also, but as far as I can tell, it doesn't support pixel format other than yuv420p, and it doesn't support h264 files with the profile "High 4:4:4", but with the profile High it works. [15:50] i.e it refuses to load files with High 4:4:4, and if it's the wrong pixel format, it loads them but the image is completely garbage [15:51] but I am not well versed in video formats, so I might be misunderstanging something or doing simething wrong [15:51] ggVGc: did you use the .mov container? [15:52] hm, no, tried .m4v and .mp4. It loads them if I used High, but not Hith 4:4:4 [15:52] why would the container matter? [15:53] I might not matter, but it's wortk a shot. [16:00] :( didn't work [16:02] ggVGc: Ut Video is lossless, try ffmpeg -i input -codec:v utvideo -codec:a pcm_s16le output.avi [16:02] thanks, will give it a shot [16:02] and it should work according to google [16:03] nice, thank you! [16:03] I am just now trying a .mov with vcodev huffyuv [16:04] relaxed: I pm-ed you a complete pipeline that plays for me, and all colors are desaturated [16:04] the magenta, the greens, etc [16:04] ggVGc: it supports the following pixel formats rgb24 rgba yuv422p yuv420p, so use the one you need [16:07] thanks, I guess yuv244p is better then, since the real source is yuv444p [16:07] hm, that utvideo didn't work either :( [16:08] ggVGc: add -pix_fmt yuv422p after the input [16:10] trying that now, but I don't think this was a pixel format error. When the pixel format was wrong before, it loaded but didn't show up right [16:10] this didn't even load [16:10] "cannot be opened, unsupported format" etc. [16:10] didn't work with the pix_fmt either [16:22] ggVGc: hmm...my last guess, ffmpeg -i input -c:v rawvideo -pix_fmt yuv422p -codec:a pcm_s16le output.avi [16:25] yeah, I was just trying rawvideo [16:25] will see how it goes [16:26] can I make ffmpeg use one video as guide and output a new one with exactly the same stream config? [16:26] i.e read stream properties from video A to convert video B to C.out [16:28] wis "MX Player" for android based on MPlayer ? [16:29] relaxed: so, I exported a lossless video from after effects, and it gave me a stream with this. "rawvideo, bgr24, 960x576, 809758 kb/s, 60 fps, 60 tbr, 60 tbn, 60 tbc" [16:29] so if I can reproduce that in ffmpeg I should be okay [16:30] what does "SAR 1:1 DAR 5:3" mean+ [16:30] ? [16:32] Sample Aspect Ratio 1:1 (square pixels) and Display Aspect Ratio 5/3 [16:33] waressearcher2: it's based on ffmpeg https://sites.google.com/site/mxvpen/download [16:34] ggVGc: did that inport work? [16:34] yeah, I figured that out. How can I make it convert that? I am close now, but my source material is a a h264 yuv444p .mkv, and the destination needs to be "rawvideo, bgr24, 960x576, 809758 kb/s, 60 fps, 60 tbr, 60 tbn, 60 tbc" [16:34] relaxed: no, but if I can make it match what I wrote above it will work [16:35] this almost gets me there "ffmpeg -i grab.mkv -pix_fmt bgr24 -codec:v rawvideo -s 960x576 -d 960x676 -acodec none test.avi" [16:35] but I get "rawvideo, bgr24, 960x576, 800891 kb/s, SAR 1:1 DAR 5:3, 60 fps, 60 tbr, 60 tbn, 60 tbc" [17:06] hi, everyone, how to detect a video' s watermark position? [17:11] relaxed: .... so in the end I manged to get it to work with h264, .mp4 container, and -crf 1 instead of -crf 0 [17:11] :( [17:11] I am so stupid [17:11] since -crf 1 gives the profile High, but is still basically lossless [17:30] Attempting to take an infinite stream of PNG files (png files being produced on the fly by a C program) and transcode them into video compatible with V4L2 devices. I'm confident that ffmpeg can do this, but could someone point me in the right direction? [17:32] what does "compatible with V4L2 devices" mean? [17:33] I need to be able to write the video to a video device in /dev/ [17:33] Such a v4l2loopback device [17:34] The ultimate goal is to get a stream of images to appear as the output of a webcam device [17:37] cprogram | ffmpeg -vcodec png -f image2pipe -r $framerate -i - (encoding options) output [17:38] I'll give it a try [17:38] Thats close to what I came up with on my first go [17:38] but let me verify [17:39] Isweet: are you trying to use another v4l2 program to read the loopback stream? (just curious) [17:40] Yeah, I test it with VLC/MPlayer [17:40] on Ubuntu 12 or 14 or something [17:41] I dont have much experience with different encodings [17:42] Off the top of your head, do you know what video encoding might be compatible with v4l2? Usually theres an explicit encoding option but I dont think ffmpeg has one [17:42] let me check [17:43] it depends, chould be mjpeg or h264 depending on the device [17:43] coulr* [17:44] ha, I can't type [17:46] Isweet: https://trac.ffmpeg.org/wiki/Capture/Webcam [17:48] Yeah, I'm trying to write rather than read [17:48] Sorry for the delay, trying to get ffmpeg to install on 14.04 [17:49] right, but I thought you might get some ideas for your output there. [17:49] relaxed: yes I actually wanted to ask "based on ffmpeg" but didn't replaced word MPlayer after asking same question in #mplayer [17:49] relaxed: but how it is "based" if ffmpeg is not a player ? is it based on ffplay ? [17:51] waressearcher2: without looking at it I assume it uses ffmpeg's libs to decode, demux, etc, just like ffplay and mplayer [17:53] Currently installing ffmpeg from source for 14.04 [17:53] I'l test when I'm done [17:53] I think the problem I was running into before is that the video encode wasn't compatible with my v4l2 loopback device [17:53] Relaxed: Thanks [19:46] is "MX Player" for android supported here ? [19:54] waressearcher2: we only provide support for FFmpeg stuff here [19:58] relaxed: Almost got the image -> video transcoding for v4l2 figured out [19:58] Using: cprogram | ffmpeg -vcodec png -f image2pipe -r 1/25 -i - -f rawvideo /dev/video0 [19:59] But I'm getting "File 'dev/video0' already exists. Exiting." [20:00] But I'm getting "File '/dev/video0' already exists. Exiting."* [20:01] Is there a distinction between writing to a file and creating it? [20:08] can someone confirm this is the right command to capture video in linux: "ffmpeg -threads 0 -async 30 -f alsa -i pulse -f x11grab -s 640x480 -r 30 -i :0.0+44,47 -vcodec libx264 -preset superfast -crf 16 -acodec libmp3lame -f mp4 /tmp/1.mp4" ? [20:19] I'm having trouble getting my hls output to use the command line parameters: http://dpaste.com/3T97K4D [20:29] waressearcher2: looks about right [20:31] danomite-: what appears to be the issue? [20:31] ie, what ffmpeg is doing looks fine from where I'm sitting. [20:32] c_14, the output of hls isn't honoring the command line args, for example hls_time is set to 2 seconds but the output is well over that [20:49] waressearcher2: do y-u really need the -async? i don't know what -threads 0 as an input option for x11 is supposed to do. you may want to add -pix_fmt yuv420p as an output option depending on your supported players [20:49] *you [20:50] and use -framerate instead of -r for x11grab input option [20:50] danomite-: add -force_key_frames 'expr:gte(n,n_forced+2)' as an output option [20:50] Hello. How can I capture playback using alsa or pulse? [20:50] https://trac.ffmpeg.org/wiki/Capture/ALSA [20:50] http://ffmpeg.org/ffmpeg-devices.html#alsa [20:51] http://ffmpeg.org/ffmpeg-devices.html#pulse [20:51] I saw all of that. [20:51] There was nothing. [20:51] The Nothing took them. [20:51] Do you mean playback from an application? Ie audio output? [20:51] Yes. [20:52] Either by using snd_aloop with alsa or by pushing funky nobs in pavucontrol [20:52] the wiki article could use an update mentioning that sort of stuff, IIRC [20:53] I could add the ALSA stuff to the wiki page, I just went through it with somebody a couple of days ago. Don't know squat about pulse though. [20:54] same here. [20:54] and what's the difference between -f alsa -i pulse, and -f pulsa -i default or whatever? [20:54] i don't ever see people usign -f pulse [20:55] Is it not so obviously? [21:05] Modprobed snd-aloop, used hw:1 for capture. [21:05] Nothing. [21:05] What should I do? [21:05] Sec, will have the wiki article up shortly. [21:05] Thanks. [21:06] You'll be the first victi-- I mean volunteer. [21:09] c_14, any pointers on reducing the stream delay with hls output? [21:11] xanal0verlordx: right, added 2 examples at the bottom of the page [21:12] Capture/ALSA? [21:12] danomite-: no clue, sorry [21:12] Thanks. [21:12] yep [21:12] If something doesn't work, ping me. [21:13] What is pcm_substreams? [21:14] c_14: thanks. you could remove the ALSA suggestion from https://trac.ffmpeg.org/wiki/ArticlesForCreation when you're satisfied with the update [21:18] Does not work without those substreams. And I can't rmmod it, it's in use. [21:21] c_14, thanks, the key frame option worked [21:23] Currently using % cat abcd.png | ffmpeg -y -vcodec png -f image2pipe -r 1/25 -i - /dev/video0 [21:23] To try to write images to a video device on linux [21:24] Getting "Unable to find a suitable output format for '/dev/video0'" [21:24] Any solutions? [21:25] xanal0verlordx: modprobe snd-aloop index=1 pcm_substreams=1 <- try that [21:26] isweet: add -f rawvideo [21:26] xanal0verlordx: you'll have to adjust the asoundrc though [21:27] Because you're using a different (new) loopback device [21:27] c_14: Tried that, yields a different error: "av_interleaved_write_frame(): Invalid argument" [21:28] isweet: you can try setting -f v4l2 as an output device, but I'm not sure ffmpeg'll like that either [21:29] c_14: "Unknown V4L2 pixel format equivalent for rgba" [21:30] What pixel formats does your device accept? [21:31] xanal0verlordx: or try `fuser -v /dev/snd/*' to find the processes using the sound output, kill them, `rmmod snd-aloop' and `modprobe snd-aloop pcm_substreams=1' [21:31] Thats a question over my head. Its a video loopback device (https://github.com/umlaeute/v4l2loopback) based on v4l2. I assumed it would support at least the standard pixel formats listed under the V4L2 spec. Is that an incorrect assumption? [21:32] hmm, add -pixel_format yuv420p [21:33] % cat abcd.png | ffmpeg -y -vcodec png -f image2pipe -r 1/25 -i - -f v4l2 -pix_fmt yuv420p /dev/video0 [21:33] like so? [21:34] ye [21:34] Brilliant! No errors that time [21:34] Let me actually read the devicenow [21:34] now* [21:35] It totally works. [21:35] c_14: Thanks so much. Been struggling with this for a few days. [21:35] The pixel format stuff is out of my wheelhouse [21:55] c_14: will try now. [21:56] Killed pulseaudio, rmmoded snd_aloop. [21:56] How to add this parameter to /etc/modules? [21:56] Just write it with snd-aloop? [21:57] I think so. [21:58] Though you might have to add that to modprobe.d [21:59] with 'options snd-aloop pcm_substreams=1' [21:59] Do not care right now. It still does not work. [22:00] $ cat .asoundrc [22:00] pcm.!default { [22:00] type plug slave.pcm "hw:Loopback,0,0" [22:00] } [22:00] Right? [22:00] ye, try adding a newline between plug and slave [22:01] llogan: what option should I add to configure ? --enable-x11grab ? or is it only to grab desktop ? will it grab say fullscree opengl programm ? [22:02] waressearcher2: yep, you should configure ffmpeg with --enable-x11grab. [22:02] And some sound libs. [22:03] Maybe, lame. [22:03] It will grab everything you see. [22:04] http://sprunge.us/eEcQ these are options for configure I use, any other I should add ? [22:04] c_14: still does not work. [22:05] xanal0verlordx: by 'lame' you mean '--enable-libmp3lame' ? [22:05] Yes. [22:06] xanal0verlordx: are you into pr0n ? [22:06] Wut? [22:07] Can you pastebin your .asoundrc and the output of aplay -l and the output of ffmpeg ? [22:11] https://clbin.com/0GtWa [22:11] c_14: [22:13] news.microsoft.com/2014/11/12/microsoft-takes-net-open-source-and-cross-platform-adds-new-development-capabilities-with-visual-studio-2015-net-2015-and-visual-studio-online/ [22:13] OHWOW [22:14] Today is 1st of April? [22:17] So what exactly isn't working? [22:17] I do not hear any sound. [22:17] When playing the mp3? [22:18] Yes. [22:18] Probably because your audio device is still set to the loopback device? [22:18] xanal0verlordx: do you have 2 sound cards ? [22:18] one Intel and other Yamaha ? [22:18] waressearcher2: dunno. [22:18] xanal0verlordx: lspci [22:18] c_14: I hear any other sounds. [22:19] You have pulse installed? [22:19] Yes. [22:20] I can think of 2 reasons it might not be working. 1: it's an avconv issue (avconv is from libav and is a fork of ffmpeg [and technically not supported here, if you're on debian you can get ffmpeg from unstable, otherwise you'll need to build from source]) or 2: pulse is being a bitch and breaking things [22:21] I think that libav devs did not breaked that stuff, so maybe it is pulse. [22:22] Okay, do you know how to do loopback with pulse? [22:23] waressearcher2: ~$ lspci | grep Audio [22:23] 00:1b.0 Audio device: Intel Corporation 7 Series/C210 Series Chipset Family High Definition Audio Controller (rev 04) [22:23] Sadly no, I know it's possible but I don't know the details. [22:26] xanal0verlordx: you have one card and its Intel [22:27] waressearcher2: I see. [22:28] now you know [22:30] Will try tomorrow and tell you about the results, if you want to make your wiki better. [22:30] Bye. [22:48] Yet another question. [22:49] Is there a way to grab video using Android? [22:49] Smth like -f androidfb. [22:51] Wait, it uses Linux fbdev. [22:51] So I can grab video directly from /dev/fb0? [00:00] --- Thu Nov 13 2014 From burek021 at gmail.com Fri Nov 14 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Fri, 14 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141113 Message-ID: <20141114010502.C7CFB18A02EC@apolo.teamnet.rs> [00:06] ffmpeg.git 03Michael Niedermayer 07master:2f6bb86f8588: swscale/utils: support bayer input + scaling, and bayer input + any supported output [02:32] ubitux: why does ASS get AVFMT_NOTIMESTAMPS but SRT doesn't? [03:07] ffmpeg.git 03Michael Niedermayer 07master:3e1ac1034535: avcodec/utils: Add ATRAC3+ to av_get_audio_frame_duration() [08:28] ffmpeg.git 03Reimar D?ffinger 07release/2.4:3e0802e42b07: configure: Hack to treat x32 as x86_64. [08:30] rcombs: i don't know... i'm not actually aware of the effects of that AVFMT_NOTIMESTAMPS [08:31] but ASS muxing definitely needs them [08:31] thought, the way the dts are used might come into play here [11:37] damn the edit lists are twisted [11:38] like, first entry can be relative to whole presentation while others entry are relative to the stream itself [12:00] ffmpeg.git 03Vittorio Giovara 07master:68a35473ed42: 4xm: more thorought check for negative index and negative shift [12:00] ffmpeg.git 03Michael Niedermayer 07master:64e12aec9654: Merge commit '68a35473ed423a14731c418939fba7913647979a' [12:01] ffmpeg.git 03Michael Niedermayer 07master:29234f568181: vp7: fix checking vp7_feature_value_size() [12:01] ffmpeg.git 03Michael Niedermayer 07master:8e6084eebe86: Merge commit '29234f56818135faf2f1868ab324c073abd28fbd' [12:11] ffmpeg.git 03Vittorio Giovara 07master:b09cf8afc519: libopusenc: check return value [12:12] ffmpeg.git 03Michael Niedermayer 07master:3c6e148e843e: Merge commit 'b09cf8afc5199d359ac985ad7cea72a6a9f20e4e' [12:52] ffmpeg.git 03Vittorio Giovara 07master:8dd0a2c5cf40: libopusenc: prevent an out-of-bounds read by returning early [12:52] ffmpeg.git 03Michael Niedermayer 07master:aa7c429c4e8e: nellymoserenc: fix array element ordering [12:52] ffmpeg.git 03Vittorio Giovara 07master:bdcb5794f0c2: nellymoserenc: fix array index [12:52] ffmpeg.git 03Michael Niedermayer 07master:7b824e8b25cb: Merge commit '8dd0a2c5cf40a8a49faae985adc11750b6429132' [12:52] ffmpeg.git 03Michael Niedermayer 07master:0f29e92091d4: Merge commit 'aa7c429c4e8e561009176d51b7dcb626c85eb276' [12:52] ffmpeg.git 03Michael Niedermayer 07master:946a5cb64d58: Merge commit 'bdcb5794f0c2d74371152303bffe4172671af264' [13:04] ffmpeg.git 03Vittorio Giovara 07master:b1b1a7370e14: display: fix order of operands [13:04] ffmpeg.git 03Vittorio Giovara 07master:e3f50f247155: dnxhdenc: check negative index [13:04] ffmpeg.git 03Michael Niedermayer 07master:76fa78911f58: Merge commit 'b1b1a7370e141c912e3d0bbaa668dcee05c3ad67' [13:04] ffmpeg.git 03Michael Niedermayer 07master:7f8ef7876efe: Merge commit 'e3f50f247155216229e34f165bae8c329d5a001e' [13:04] ubitux: one effect is allowing negative timestamps [13:05] ubitux: if it (and AVFMT_TS_NEGATIVE iirc) are not present, the first frame gets shifted to start at 00:00:00 if its timestamp is negative [13:06] ubitux: which can be really confusing behavior for e.g. using ffmpeg to shift times in plain SRT files [13:31] ffmpeg.git 03Vittorio Giovara 07master:2c98dc75f280: vc1dec: always initialize tx and ty [13:31] ffmpeg.git 03Vittorio Giovara 07master:28d82b7675be: vc1dec: refactor check with missing parenthesis [13:31] ffmpeg.git 03Michael Niedermayer 07master:94cf1ef008c2: Merge commit '2c98dc75f2802a2fe91922d4a11b698b66420e5b' [13:31] ffmpeg.git 03Michael Niedermayer 07master:ef4e54e0df44: Merge commit '28d82b7675bea76a1349070a3cdd737d964d4775' [13:35] ubitux: weirdly, the behavior is different between seeking on input and output [13:36] ubitux: try `ffmpeg -ss 25:00 -i longfile.srt outfile.srt`, where 25:00 is between 2 subtitles [13:36] I'd say it doesn't give the expected behavior [13:40] rcombs: aren't you looking for -itsoffset? [13:40] if I'm trying to shift without clipping, yes [13:40] (right?) [13:40] yeah, i guess [13:41] it's at least somewhat weird that -ss as an output arg gives the expected behavior, but as an input arg it doesn't [13:41] it's probable this stuff is a bit buggy, i admit i haven't look closely [13:41] ffmpeg.git 03Vittorio Giovara 07master:4dda5e9b0829: sunrastenc: mention missing break [13:41] ffmpeg.git 03Vittorio Giovara 07master:e9a6ae775dab: dpxenc: mention missing break [13:41] ffmpeg.git 03Vittorio Giovara 07master:5aa710f46119: vorbisenc: add missing parenthesis [13:41] ffmpeg.git 03Michael Niedermayer 07master:ccfae038152b: Merge commit '4dda5e9b0829b119c17d950906c61d3ebffc494f' [13:41] ffmpeg.git 03Michael Niedermayer 07master:b882079b1265: Merge commit 'e9a6ae775dabef3942632e8d4ef95fff94a1b310' [13:41] ffmpeg.git 03Michael Niedermayer 07master:198e55bfa791: Merge commit '5aa710f46119bb9c1c38542f80f5338eb8b5ffb2' [13:41] also make sure you don't involve the decoding chain in your test [13:42] like, try if it's different with -c copy [13:43] i remember fixing some stuff so ffmpeg -ss 123 -i in.mkv -c copy out.mkv work fine [13:43] but if transcoding is involved, timestamps might be messed up [13:43] (because of the timestamps still in the decoded form blabla you know the story) [13:43] no difference [13:43] ok [13:44] i don't know... [13:44] ffmpeg.c itself shifts the timestamps so the first frame starts at 00:00 [13:44] (to avoid a negative timestamp) [13:44] ok.. [13:44] I'm not sure why it works differently for an output seek [13:45] might have to do with the lack of a `strim` filter to insert [13:45] please open a ticket if you can summarize the weird behaviour(s) and inconsistencies [13:45] sure [13:58] ffmpeg.git 03Vittorio Giovara 07master:6abe7edabb7d: ffv1: fix out-of-bounds read [13:58] ffmpeg.git 03Michael Niedermayer 07master:e266e186cf12: Merge commit '6abe7edabb7d57e82d7ea6312d30cf05d2192c5b' [14:07] ffmpeg.git 03Vittorio Giovara 07master:2ffb0598dbdb: mlpdec: check for negative index [14:07] ffmpeg.git 03Vittorio Giovara 07master:1a9c1333b5d7: escape124: explicitly set get_bits1 variable [14:07] ffmpeg.git 03Michael Niedermayer 07master:9b424accbe10: Merge commit '2ffb0598dbdb81c40650952aa9299fa02fa5e834' [14:07] ffmpeg.git 03Michael Niedermayer 07master:2c4d5d3497e9: Merge commit '1a9c1333b5d70b427c82cb98f383aa2fa9b2b319' [14:29] ffmpeg.git 03Vittorio Giovara 07master:d5d2d6c3b8cf: dcadec: initialize variables before use [14:29] ffmpeg.git 03Vittorio Giovara 07master:8e104619a627: shorten: check for return value [14:29] ffmpeg.git 03Michael Niedermayer 07master:785f71fcd508: Merge commit 'd5d2d6c3b8cff61eb26c18bbd977881cf6d5524a' [14:29] ffmpeg.git 03Michael Niedermayer 07master:ecb748866e00: Merge commit '8e104619a627fcf5f4c2bd3c09d0c2d323aae745' [14:45] ffmpeg.git 03Vittorio Giovara 07master:c6d7c201dfa8: indeo3: check ff_set_dimensions return value [14:46] ffmpeg.git 03Vittorio Giovara 07master:2b5c1efa1465: g2meet: check ff_set_dimensions return value [14:46] ffmpeg.git 03Michael Niedermayer 07master:dfa0800c414a: Merge commit 'c6d7c201dfa80502cb6cefbee7dc9160cedb5187' [14:46] ffmpeg.git 03Michael Niedermayer 07master:b697a3314e84: Merge commit '2b5c1efa1465d8646f8be525cace7a21404e40ad' [15:03] ffmpeg.git 03Vittorio Giovara 07master:c7384664ba0c: avs: check ff_set_dimensions return value [15:03] ffmpeg.git 03Vittorio Giovara 07master:994ab1804b8b: ansi: check ff_set_dimensions return value [15:03] ffmpeg.git 03Vittorio Giovara 07master:59846452af76: svq1enc: check ff_get_buffer return value [15:03] ffmpeg.git 03Michael Niedermayer 07master:04f6a5230db8: Merge commit 'c7384664ba0cbb12d882effafbc6d321ae706cff' [15:03] ffmpeg.git 03Michael Niedermayer 07master:057b74d19c0f: Merge commit '994ab1804b8bf532f44876927b07b51f1f63247f' [15:03] ffmpeg.git 03Michael Niedermayer 07master:45660c7d1b76: Merge commit '59846452af762f6af5ced4399e8dcd709ca50fcd' [15:26] ffmpeg.git 03Vittorio Giovara 07master:a2448cfe167a: jpeg2000: do not compute the same value twice [15:26] ffmpeg.git 03Michael Niedermayer 07master:7a79c055e378: Merge commit 'a2448cfe167a4cd4eb631318550d4eef38fca24a' [15:32] why is there so much time shift craziness in the mov demuxer because of the edit list? [15:32] i mean, it's supposed to be a start time, why not use st->start_time [15:32] and let the application deal with that? [15:38] ffmpeg.git 03Peter Ross 07master:2093c1dc51ee: cinedec: report white balance gain coefficients using metadata [15:38] ubitux, youll likely not find out [15:39] most of that stuff in mov*c is from The Time Before Commit Messages [15:40] everything starts at baf2ffd3297b707dbb5794ec568c61091acf5c0c according to git log -p --reverse --full-diff -Gtime_offset libavformat/mov.c [15:41] ive found usually when i look at mov.c it goes back to some commit form baptiste with a useless message, and like 5 changes in one commit [15:41] from svn times [15:48] even today, commit messages are neglected [15:49] and my patch that fixes the big Libav memory leak still wasn't reviewed yet [15:51] ok i think i get it... [15:52] so basically, we support the presentation start_time and the and eventually an initial skip in the stream itself [15:52] initial skip? [15:53] and instead of having them treated separately (like using st->start_time for presentation and the ts adjustement for the initial skip), the same logic is used, afaiu [15:53] wm4: yeah, basically, you have a concept of start_time (so relative to presentation) [15:53] and then the other entries specifies how each segment can cut/skip parts of the stream itself [15:54] basically we support both that concept of start_time and the skip of the first segment [15:54] and it's kind of "merged" into the same logic [15:54] not sure if I get it [15:55] ok. so... edit list in mp4 are kind of weird, they have 2 usages [15:55] so you have edit list for each streams [15:56] and if for example you have a presentation of 40 seconds [15:56] audio st is 40 seconds, and video st is 30 seconds [15:56] you want video st with a start_time of 10 [15:56] what you do is that you create an edit list in the video st [15:56] so the first 10 seconds have no video or what? [15:56] and the first entry is {.time=-1, duration=10} [15:57] this particular "empty" edit list entry basically translates to this start_time [15:57] wm4: yeah well, don't ask me, maybe static frame or dunno [15:57] you can take the audio example if you prefer [15:57] like audio starting later [15:58] mpv and ffplay would in this case keep reading packets from the demuxer to find the next video frame [15:58] anyway, you can have such empty entry (which you can identify by time=-1) [15:58] and then you can have other entries [15:58] which are relative to the stream itself [15:58] (this actually happens with mov files that contain "slide shows", which very low number of video frames) [15:58] *with [15:58] like, you can say to skip some frames in the stream itself [15:58] ok [15:59] like you can have {.time=-1, duration=10} followed by {.time=5, duration=whatever} [15:59] so basically the video stream starts after 10 seconds in the presentation [15:59] but you actually have to skip 5 seconds of video [15:59] so currently, we seem to support that [16:00] but that's the most complex pattern we support [16:00] and both this start_time and skipping is handled with the same logic [16:00] (setting some kind of time_offset and messing with pts/dts/whatever) [16:00] all at demuxer level [16:00] that's how i understand it [16:00] but i might be wrong :) [16:01] what we don't support is having more than 1 skip/duration entry [16:02] or cutting up frames [16:02] yeah right [16:02] anyway, i think the start_time should be handled outside of stream skipping [16:03] so, in theory, you could implement everything in the demuxer, and use side-data to make the decoder cut up the frames [16:03] basically by just setting st->start_time, assuming that's possible [16:03] i don't know... maybe [16:03] so packet side data saying "please ignore this?" [16:04] a high level API of some sort (NOT lavfi), which handles this kind of stuff, would be cleaner [16:04] yes [16:04] like AV_FRAME_DATA_SKIP_SAMPLES [16:04] that is used for gapless stuff, and to trim frame padding [16:05] sounds more clumsy that exporting a timeline information along with the stream [16:05] and let the application handle it the way it fits [16:05] but i dunno [16:05] s/that/than/ [16:06] (in ffmpeg/ffplay we would translate with timeline to a filtergraph which we would auto insert) [16:06] also, i wonder how the timestamps should be adjusted with all these skips [16:14] ideally, a demuxer would just export the edit list, and another piece of code would turn it into a linear stream of some sort [16:15] which could be via lavfi, some high level API, a demuxer that passes through the packets and uses tricks like AV_FRAME_DATA_SKIP_SAMPLES... [16:26] yeah maybe [16:27] anyway, i'm curious about samples that make use of st->start_time [16:27] for both audio and video [16:28] you could even use the libavdevice "lavfi" demuxer to turn a lavfi graph into something that can actually be used by applications [16:28] although you'd lose things like hw decoding [16:28] (my conclusion is that lavfi is not a playback engine) [16:29] i like the idea of exporting a timeline through C API, and let user deal with that for now [16:29] and in ffmpeg we will use our lavfi system [16:30] but honestly, designing such monster api just to fix playback issues is kind of a pain [16:36] specs: "If this field is set to 1, it is an empty edit. The last edit in a track shall never be an empty edit." [16:36] first sample i'm checking: [16:36] [mov,mp4,m4a,3gp,3g2,mj2 @ 0x7ff45c000920] st 0: duration=2437071 time=200 rate=1.000000 [16:36] [mov,mp4,m4a,3gp,3g2,mj2 @ 0x7ff45c000920] st 0: duration=2392423 time=-1 rate=1.000000 [16:37] great. [16:37] thank you. [16:37] ("this field" ? time field, obviously) [16:38] oh wait my bad... [16:38] not actually the stream. derp. [16:40] lol [16:41] I wouldn't be surprised if that actually happened [16:41] apparently it's normal for specs [17:52] ffmpeg.git 03James Almer 07master:3cec54b7d72b: x86/flacdsp: add SSE2 and AVX decorrelate functions [18:01] ffmpeg.git 03Michael Niedermayer 07master:dae7e4e63da3: tests/tiny_psnr: remove redundant initialization [19:06] ffmpeg.git 03Michael Niedermayer 07master:4001fc426798: avcodec/4xm: remove duplicate assert [21:13] ffmpeg.git 03Michael Niedermayer 07master:5c805d69a49a: avcodec/nellymoserenc: fix sign error [21:49] can i push the xbr test? [21:55] ubitux, y [21:56] thanks [22:07] michaelni: are you ok with fate test update? [22:15] ffmpeg.git 03Carl Eugen Hoyos 07master:34288651633a: lavc/flashsv2enc: Fix encoding resolution error message. [22:29] lukaszmluki, will look at it in a few minutes, am working on some other patches atm [22:31] does stefano still wander about here? [22:31] you mean saste? [22:31] yes! [22:34] saste: where did you get inspiration for gammaval() [22:39] pross, i don't remember [22:40] i remember it was proposed on the mailing-list during review, so you should be able to track it [22:40] ffmpeg.git 03Aleksey Vasenev 07master:df8248f66e36: avfilter/vf_interlace: more accurate pts calculation [22:40] ffmpeg.git 03Aleksey Vasenev 07master:8349001638fb: avfilter/vf_tinterlace: fix frame rate [22:45] thanks [22:46] ffmpeg.git 03Cl?ment BSsch 07master:57688aecbd72: fate: add xBR filter tests [22:47] pross, saste: Maybe this thread: http://thread.gmane.org/gmane.comp.video.ffmpeg.devel/121563 [22:47] lukaszmluki, you need accessor functions for recommended_encoder_configuration, see MAKE_ACCESSORS [22:48] access to it from outside libavformat should use only these accessors [22:53] lukaszmluki, maybe the opt serialize should start with something like serialize_format_version=123, not sure, also can be done later [22:53] about the fate update, that should be ok [22:57] i forgot about this accessor [22:57] but i have to go, see ya later [23:04] i propose to add a new lut gamma function iaw rec 709. what should this be called? gammaval709()? [00:00] --- Fri Nov 14 2014 From burek021 at gmail.com Fri Nov 14 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Fri, 14 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141113 Message-ID: <20141114010501.AAAB118A0239@apolo.teamnet.rs> [03:28] Can't repackage MP4 with MPEG4 to a TS. `./ffmpeg -i source1.mp4 -c:copy test.ts` -- executes without warnings or errors, while resulting TSfile video track is unreadable. http://pastebin.com/yTChS3QV [03:48] Mista_D: try -c copy -mpegts_m2ts_mode 1 [03:53] nevermind, m2ts doesn't support mpeg4 [04:04] Mista_D: I think you should file a bug report [04:20] Mista_D: why did you remove the ffmpeg version from your console output? [04:21] ah, i see it in the additional paste below [04:22] llogan: its the latest release I think... [04:23] relaxed: will file a bug with a file sample shortly, thanks [04:31] Mista_D: I was able to reproduce it [04:35] relaxed: with other mpeg4 file? [04:37] yes [07:52] hey guys i just did a source compile of a new setup, and im trying to do a 2 pass on a video and getting the following [07:53] http://pastebin.com/QJi161Zt [07:53] Could not find codec parameters for stream 0 (Video: h264 (avc1 / 0x31637661), none, 1920x1080, 20998 kb/s): unspecified pixel format [07:54] im wondering if this movie got corrupt possibly on xfer off the cell phone [07:55] aixenv: yeah, it looks like a decoding issue. [07:55] any way to verify if the file is corrupt? [07:55] if i do a "file" i get [07:56] shells_cell_5572.mp4: ISO Media, MPEG v4 system, version 1 [07:56] so im a bit perplexed how it could be corrupt yet still look like an mp4, but again im not sure, this just seems weird [07:57] aixenv: can you play it? [07:57] no i was not able to play it via windows, (which is why i was going to try and see if i could make it happy via linux/ffmpeg) [07:57] i tried in like 3 dif players too [07:58] vlc player, wondershare player, and windows media player [07:58] that's a pretty good sign it is. [07:58] crap, its kinda important for her work, any way i could fix this? [07:59] grab it off the phone again [07:59] she xferred and nuked off phone :( [07:59] since that's the default [07:59] lame default btw [08:00] oh well, ill just tell her the files are hosed, that stinks, one other questoin [08:00] i just upgraded (src compiles and what not) all my ffmpeg stuff, everything looks to be working and happy [08:00] i have a home project where i encode my videocam videos, and put them ona streaming site i made (just home use) [08:00] hmm, try ffmpeg -pix_fmt yuv420p -i shells_cell_5572.mp4 -c copy output.mp4 [08:01] i havent updated/tweaked my ffmpeg command in about 3-4yrs, anything you'd adviseme do dif than this? [08:01] ok ill try that too , here's the commmand im using currently [08:01] I saw it in the pastebin [08:01] ffmpeg -y -i "$file" -vcodec libx264 -r 30000/1001 -deinterlace -s 1024x576 -crf 21 -maxrate 2M -bufsize 6M -vpre slow -threads 0 -acodec libfaac -ar 48000 -ab 128k ${file%.*}-1.mp4 >> "${file%}"-encoding.out 2>&1 [08:01] dif command [08:01] that was me just trying to get this video working [08:01] thats my "normal" encoder script command [08:02] yeah, -vpre is outdated. read https://trac.ffmpeg.org/wiki/Encode/H.264 [08:02] go with preset now? [08:02] yes [08:03] ok thanks, anything else youd change ? [08:03] btw that command gave "Option pixel format not found" [08:03] -pixel_format yuv420p [08:04] same [08:06] its ok if its hosed its hosed ; would that pastebin command give better quality/size ratio than my above command? if so ill just switch to that 2 pass command, i notice that uses libfdk_aac versus libfaac too [08:06] and i grabbed the 2 pass example from that link you had given me :) [10:32] is ffserver still activley developed? [10:34] I don't think it's been actively developed for a long time. [10:39] Is there a reccomendation for an open source alternative? [10:56] Are there any examples of using the fps filter in C ? I'm trying it but for some reason the output from buffersink returns duplicate frames, even though the source is 50 fpsfps and target is 25fps [12:38] Why does using the fps filter with avfilter_graph_parse_ptr give duplicate frames ? I get the same frame repeated for 25 seconds instead of 25 frames every 1 second [12:42] Try inverting the number? [12:45] c_14: I'm doing fps=fps=25/1 .. Is thought thats right according to the docs [12:45] In the api? [12:45] I'm pretty sure it's inverted in the API iirc. [12:46] Using the C API .. Thanks for the tip , lemme try it your way [12:49] Thats slows down the fps to 1 frame every 25 seconds .. Doesn't fix the issue I've been experiencing where the fps filter throws out duplicate frames when it's supposed to drop them [12:55] I have no clue. Never really used the api. [13:01] Hello [13:01] c_14: Thanks anyway, thats one thing that I didn't try [13:02] Two things: 1) I forgot the link of the ffmpeg binary& where it is? ^^" 2) how to shrink some audio stream by 24/23?? [13:03] http://johnvansickle.com/ffmpeg/ <- this one? [13:04] Yes thanks ^^ [13:04] And what do you mean shrink by 24/23? [13:06] Err, 25/23* [13:07] c_14: I got a movie at 25fps which is ~\frac{25}{23} times bigger than the same movie in French at ~23fps [13:07] s/at/with/ [13:07] s/with/of/ [13:08] Found monitor in pavucontrol. [13:08] Will try. [13:17] Yeah. [13:17] I made that. [13:18] relaxed: ok, I got to the bottom of the issue (but don't yet know how to solve it) [13:18] http://i.imgur.com/Z9nIk4T.png <--- normal color [13:18] wget -qO- http://i.imgur.com/Z9nIk4T.png | ffplay -hide_banner -f image2pipe -i /dev/stdin <--- color loss [13:18] this is the basic effect I am trying to avoid, and not sure how :( [13:20] lots of my google hits suggest that this has to do with crappy rgb->yuv conversion, but even things like h264rgb do not help even though they should [13:37] ffmpeg -f x11grab -framerate 25 -video_size `xrandr | grep '*' | cut -d' ' -f4` -i $DISPLAY -f pulse -i `pactl list sources | grep 'Name:' | grep 'monitor' | cut -d' ' -f2` [13:37] c_14: is it right? [13:37] Without output yet. [13:45] Whoops, micro and playback does not mix. [13:54] pavucontrol says that he see 2 streams. [13:54] ffmpeg too. [13:54] Stream 0 from pulse default and stream 1 from pulse monitor. [13:55] I hear my voice with this variant. [13:55] If I put monitor at first, it records only the music. [13:58] xrandr | awk '/*/{print $1}' [13:59] relaxed: how much faster it will be? [14:01] you won't notice a difference, but it looks more elegant [14:02] Okay. [14:04] the more I read, the more I am convinces this is the result of a shit rgb->yuv conversion, and what's more irritating is that the yuv gamut is *wider* than rgb, so logically the trip ought to be lossless [14:05] aside from that I wonder how come this isn't a FAQ with a good answer :( [14:05] ribasushi: I downloaded the png- which command shows the issue? [14:06] relaxed: just ffplay-ing it through image2pipe [14:06] wget -qO- http://i.imgur.com/Z9nIk4T.png | ffplay -hide_banner -f image2pipe -i /dev/stdin [14:06] or ffplay -hide_banner -f image2pipe -i [14:07] so the png's color is already off? [14:07] it isn't [14:07] look at the png in a browser, or gimp [14:07] compare to what ffplay shows you [14:09] relaxed: how can I saw awk that I need both entries? [14:09] Name: and monitor. [14:09] I know only or, |. [14:09] & and && does not work. [14:12] Found. [14:12] /Name/,/monitor/ [14:12] No? [14:12] Ergh. [14:12] ribasushi: ffmpeg -i Z9nIk4T.png ffmpeg_encoded.png; feh -d Z9nIk4T.png ffmpeg_encoded.png [14:13] ribasushi: there's no difference ^^ [14:15] xanal0verlordx: I don't have `pactl` so I can't help. If what you have works, use it. [14:15] But it's awk releated question. [14:16] As you said, it looks better. And it really looks better, so I want to change my second command to smth like this. [14:20] sounds like a good exercise to learn awk [14:23] Found. [14:23] /smth/&&/smthelse/ [14:28] ribasushi: ffplay doesn't playback the file in rgb24, try mpv instead [14:30] relaxed: I understand and see that [14:31] relaxed: what I want is to produce videos that will get *common players* to display the correct color (this includes ffplay, because vlc uses the same codebase) [14:31] relaxed: and given that rgb is a subset of yuv (gamut-wise) I believe I should be able to do this somehow, just haven't found the magic yet [14:34] I still have a problem. [14:34] I hear only 1 stream on my video. [14:34] Should I mix them both into one? [14:35] Audio streams. [14:36] 1 video from x11grab, 1 mono audio from micro, 1 stereo from playback. [14:36] ribasushi: mpv uses ffmpeg's libs too. ffplay is far from a "common player" [14:48] I have ffmpeg-2.3 and I run that command to grab screen: "ffmpeg -f x11grab -s 640x480 -framerate 25 -i :0.0 -vcodec libx264 -preset superfast -pix_fmt yuv420p -y -f mp4 /tmp/1.mp4" and it gave me "Segmentation fault" [14:48] what option should I use to get verbose output ? -v debug ? [14:49] relaxed: ok let's try it from a completely different angle [14:49] waressearcher2: try to check where it fails. [14:49] relaxed: given the very same png, I am doing this to create a "movie" out of the image: ffmpeg -y -hide_banner -f image2 -loop 1 -i out.png -t 5 -f matroska -c:v libx264 -preset ultrafast -qp 0 5_second_still.mkv [14:50] the resulting file has lost color on all players (mpv, vlc, ffplay, mplayer), and has the same colorloss when sent to youtube [14:50] xanal0verlordx: I changed codec x264 to mpeg4 and it works, could be a problem with x264 library [14:50] relaxed: that is what I am ultimately trying to fix [14:52] relaxed: even worse - if I take the resulting video, and extract a frame from it - the color is *fine*: ffmpeg -hide_banner -y -i 5_second_still.mkv -f image2 -vframes 1 out_frame.png [14:54] What's the difference between -filter and -filter_complex? [14:55] With first I can apply only 1 filter? [14:56] xanal0verlordx: yes [14:56] And that is all difference? [14:57] -filter_complex is used for miltiple inputs [14:58] In order to use amix I should use filter_complex, right? [15:10] how to grab sound also ? [15:14] xanal0verlordx: save yourself the headache and just always use -filter_complex... ;) [15:15] if I add options: "-f alsa -i pulse -vb 2000k -ab 128k -ar 44100 -ac 2 -acodec libmp3lame" it says "cannot open audio device pulse (No such file or directory)" [15:15] So you do not have PulseAudio installed. [15:15] Are you a retard? [15:15] Use hw:0 instead. [15:15] blippyp: I'm trying to understand it all. [15:23] I can't use "-f alsa -i pulse", is it possible to grab audio without having "pulse" library ? [15:24] waressearcher2: sure; try either the 'default' alsa device or one of the hardware devices. [15:26] kepstin-laptop: I run that command: "ffmpeg -f x11grab -s 1024x768 -framerate 30 -i :0.0 -vcodec mpeg4 -vb 2000 -f alsa -i default -vb 2000k -ab 128k -ar 44100 -ac 2 -acodec libmp3lame -y -f avi /tmp/1.avi" and then started "wolfenstein 3D" and when I watched video there was no audio from wolfenstein, so "-i default" doesn't working, and "-i hw:0" [15:26] also doesn't working [15:26] oh, you want to capture the audio that an application is playing? That's harder. [15:27] kepstin-laptop: yes, but what it is capturing otherwise ? from microphone ? [15:27] He is trying to do same as me. [15:27] Maybe. [15:27] kepstin-laptop: harder but possible ? [15:27] you'll either have to set up an alsa loopback (aloop) device or set up pulseaudio [15:27] I what to capture some gameplay [15:27] But I do not want to work with bare ALSA. [15:27] PulseAudio is shit, but it works. [15:28] So I am using PulseAudio. [15:28] waressearcher2: you can use JACK. [15:28] yeah, if the application is playing back audio into pulseaudio (either natively or via the alsa plugin), then you can record from the pulseaudio 'monitor' device to get the game audio [15:30] any ideas how can I capture only one window ? I can use "-i 0:0+40,50" to capture specific place on screen but if I move window it will not be in focus anymore so is there a way to capture specific window ? I know there is option "-i title=RecordWindow" for windows version but is there similar thing in linux ? [15:31] waressearcher2: xwininfo. [15:33] waressearcher2: no, nothing like that in the X display capture. For the moment, just don't move your window :/ [15:35] Yes, the only solution is just not moving window. [15:59] balls [16:47] Hi there. We are developing a client for ar drone 2.0 and want to transcode and save the video stream with ffmpeg. If I use ffplay for video playback the is no problem. But with the transcoding I get a huge delay of ten seconds and more. [16:48] This is what I tried so far: [16:48] ffmpeg -re -i sourceFile -acodec copy -threads 8 -vcodec libx264 -b 160k -f mpegts udp://127.0.0.1:5000 [16:48] How can I speed this up? [16:49] btw sourceFile is a tcp steam [16:50] I suspect that x264 is buffering frames, for optimal average bitrate encoding. To rule it out, try -vcodec mjpeg -vqscale=4 [16:51] ops... vqscale 4 [16:51] i think. [16:51] I made that. [16:51] ffmpeg -f x11grab -framerate 25 -video_size `xrandr | awk '/*/ {print $1}'` -i $DISPLAY -f pulse -i `pactl list sources | awk '/Name:/ && /monitor/ {print $2}'` -f pulse -i default -filter_complex amix -f flv -c:v libx264 -pix_fmt yuv420p -g 50 -keyint_min 50 -b 600k -minrate 600k -maxrate 3500k -bufsize 1000k -crf 18 -preset ultrafast -c:a libmp3lame -ar 44100 -ac 2 -ab 96k test.flv [16:52] santa: there is x264 option -tune zerolatency, that you might want to try. [16:52] Are all the parameters right? [16:52] Especially video encoding. [16:52] also have in mind that thread encoding is usually frame based, so with 8 threads there would be 8 frames lag. [16:53] I already tried "zerolatency" but it doesn't work. [16:55] well, try with different codec, to rule encoder lag. afaik mpegts is not going to buffer a lot of frames on its own. [16:56] also, try without sound (-an) [16:56] Why my bitrate is not constant? [16:58] Well, we have to decode the video with h264. Is there any better way? [17:04] santa: i ask you to test that, not run it in production. [17:04] also... the lag might be caused by buffering in the video player [17:07] yo [17:07] is it pocible to aply audio effect using FFMpeg? [17:09] man ffmpeg-filters the top half shows the audio filters you can apply [17:10] blippyp: i can't use man at all due to missworking with tts [17:12] https://www.ffmpeg.org/ffmpeg-filters.html [17:12] it's the man page in html... ;) [17:18] blippyp: :P [17:45] is there a way to generate two different files for video and audio ? instead of one video_audio.avi to get two video.avi and audio.mp3 ? [17:52] waressearcher2: https://trac.ffmpeg.org/wiki/Creating%20multiple%20outputs [18:08] hi, anyone is able to tell me if it's possible to run ffmpeg on kobo arc without troubles bcs i tried thungs like ffmpeg4android and it's crashing during h264 encoding same with ffmpeg media encoder for android [18:30] hey spidy [18:30] oops [20:17] Is ffserver being activity developed? [20:18] somewhat [20:29] is ffserver still recommended for new projects? [20:31] danomite, last I heard - no [20:32] Is there a recommendation for new projects? [20:53] look at "git log ffserver.c" for development (or via the git browser at git.videolan.org) [21:16] https://github.com/FFmpeg/FFmpeg/graphs/contributors [21:16] There does not appear to be much activity [21:19] I really just want to get a live flash stream going, working on a paste [21:22] ffserver is basicaly dead, it's only left there because it works for some people. It was/is a candidate for removal. [21:22] If you want an rtmp server, use nginx-rtmp [21:29] BtbN, thanks for the straight answer [22:05] I'm trying to run the command "ffmpeg -y -threads 1 -i kwt2.wmv -b:v 1200000 -b:a 128000 -acodec libvorbis -aq 100 -async 1 kwt2.mp4" to convert a wmv to mp4. I can convert any other format to mp4 using this method including most wmv files, but I have two wmv files that fail. There's no error message other than it saying "Killed" at the end even when using verbose logging. ffprobe for the file doesn't show anything useful other [22:05] than "Flip4Mac WMV Export Component for QuickTime (Mac)" and "Stream #0:0(eng): Video: wmv3 (Main) (WMV3 / 0x33564D57), yuv420p, 1920x1080, 1233 kb/s, 25 fps, 25 tbr, 1k tbn, 1k tbc" [22:05] It's interesting to note that even if I convert it successfully to ogv I can't then convert the ogv to mp4. I get the same "Killed" message. [22:06] Sirisian|Work: you can use "-b:v 12000k" [22:09] (Also this isn't related to audio. Any audio format produces the same issue.) [22:10] Sirisian|Work: can you provide an input sample file? [22:10] One moment, I will. [22:14] llogan, https://drive.google.com/file/d/0B1SdFF_bw3xManlpRkpyaDZfSk0/view?usp=sharing [22:14] That should work ideally [22:15] I'm thinking this might have something to do with the flip4mac. Both the files I can't convert were made with it. [22:15] if I use option "-vcodec mpeg4" but add option "-vtag xvid" what will it change ? the coder stays the same "mpeg4" [22:15] ? [22:19] Sirisian|Work: works for me. upgrade your ffmpeg. [22:19] ooh good. Yeah mine is like 5 months old. [22:19] and vorbis in mp4? [22:20] do you need async? [22:20] llogan: wth? vorbis work with mp4? [22:21] Seems fine. I use the same command to convert to ogv and mp4 for the web. Works on every device I've tested. [22:22] or it ignores it and does its own thing. [22:22] ok, but you should refer to the specs [22:22] most people use AAC audio [22:22] yeah faac. I saw that one being used a lot for mp4. [22:22] Sirisian|Work: use libfdk_aac [22:23] Is that -c:a libfaac ? [22:23] git rid of the bitrates, (the default rate control should be fine), and add -movflags +faststart [22:24] since you did not include the complete ffmpeg console output from your command we can't know you how your ffmpeg has been configured [22:25] http://pastebin.com/6HmGq5Dd [22:25] it would be -c:a libfdk_aac, but you don't have support for this [22:26] well I'm rebuilding it. I can include it. [22:26] so use "-c:a aac -strict experimental" instead (or compile with libfdk_aac support) [22:26] hey llogan [22:26] when i do ffprobe [22:26] on some video or streams [22:27] sometime i dont see the audio bitrate if they use aac [22:27] what do that mean [22:27] i don't know [22:28] :P [22:39] what is "-bf" option ? [22:44] > -bf E..V.... set maximum number of B frames between non-B-frames (from -1 to INT_MAX) (default 0) [22:46] how is it improve quality if I set "-bf 1" ? [22:47] You'll get more p frames and less b frames. [22:47] Whether or not that improves quality depends on the vidie. [22:47] *video [22:47] c_14: can ffm send to icecast2? [22:47] or you're unstreamable:P [22:48] also NUM in that option: "-q:v NUM" goes from 0 to 100 ? I set it to "-q:v 4" and its good quality [22:49] DelphiWorld: https://ffmpeg.org/ffmpeg-protocols.html#Icecast [22:49] waressearcher2: qscale depends on the codec [22:55] c_14: I use "-vcodec mpeg4" [22:57] For mpeg4 qscale goes from 1 to 31 where 1 is the highest quality. [23:03] c_14: is it the same as "-q" option ? [23:03] yep [23:04] -q is an alias for -qscale [23:40] llogan, Interesting. I ran it once and it worked and created a file I could play. I tried to run it a second time and it failed. This was after rebuilding ffmpeg from scratch. So I uninstalled it then tried again and it still fails. I think I'm going to format this machine and try again. [23:41] Sirisian|Work: are you using ffmpeg from current git master? [23:42] yeah [23:42] Are there tags or something? [23:43] git clone --depth 1 git://source.ffmpeg.org/ffmpeg [23:43] that's fine. [23:43] remove the --depth 1 if you're interested in checking out older revisions [23:44] such as if you want to use git bisect to find a regression [23:44] or i guess tools/bisect-create to be more accurate [00:00] --- Fri Nov 14 2014 From burek021 at gmail.com Sat Nov 15 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Sat, 15 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141114 Message-ID: <20141115010503.3725218A008D@apolo.teamnet.rs> [00:41] random news: $100 SPI donation incoming (right i know that's not that much) from a marketing company who was happy to get help to concatenate files at format level [00:46] neat [00:58] hi. reasking a question from #ffmpeg. are there any plans to support tiled OpenEXR files ? [01:12] I have no idea what that is but from the name I reckon it is already supported except in a rather inefficient manner. [01:13] ubitux: "use the concat demuxer kthx" [01:13] "that'll be $100" [03:09] ffmpeg.git 03Peter Ross 07master:b186b7131e16: avfilter/vf_lut: gammaval709() [06:20] is it just me, or is VideoLAN git being a bit unreliable? [08:41] j-b: Hi, is it known that the git server refuses connections? [09:46] cehoyos: no [09:47] The following does not work here: [09:47] git clone git://git.videolan.org/ffmpeg.git test [09:48] Same with: git clone git://source.ffmpeg.org/ffmpeg.git test [09:53] j-b: Can you reproduce? [09:55] not at all [09:55] The error message shown is "fatal: read error: Connection reset by peer" [09:56] Reproducible from servers located in other parts of the world. [09:58] j-b: Does cloning work for you? [09:58] yes [09:59] Can you try from another server: I tried from a Debian test server and from gcc buildfarm and from ffmpeg.org [10:02] Same for "git clone git://git.videolan.org/vlc.git" [10:04] cehoyos, same problem here [10:04] Grazie! [10:05] cehoyos: try. [10:06] It seems to work now, thank you! [10:06] Yes, works now! [10:07] Why did it work for you? [10:48] rcombs: i actually wrote a 15 lines shell script, but yeah that's the idea [11:16] is it possible to set the encoding speed? [11:16] speed? [11:16] -re option does not release the cpu [11:17] I want to set the encoding speed to 25fps [11:17] the input is a file on my disk [11:18] I think if the speed exceeds 25fps, the cpu should be release to do other work [11:19] av500, do you think so? [11:20] I'm debugging melt to set the preview speed. [11:28] maybe we could introduce a -vf speed that does some sleep() :-? [11:28] hi, is there any way to get the current sar in a write_packet muxer function? [12:25] pomaranc, don't think so [15:48] ffmpeg.git 03Benoit Fouet 07master:98abb98cb1c7: avcodec/pngdec: rename decode_frame to decode_frame_png [16:15] ffmpeg.git 03Benoit Fouet 07master:1523d1484d2e: avcodec/pngdec: create a function to decode IHDR chunk. [16:15] ffmpeg.git 03Benoit Fouet 07master:3f1eaf590ca2: avcodec/pngdec: create a function to decode pHYs chunk. [16:39] hullo. got a question that's not specifically about ffmpeg, but requires some knowledge of video file formats, so I figure it's the best place to ask. [16:40] I'm trying to create a "datamoshing" effect (glitches) from video samples. If I were to take, for example, a pair of identically encoded, identical-ish length, identical resolution videos with different content, then xor the two together and preserve the header, am I likely to find that it's renderable? [16:44] probably not [16:44] Probably not. Different content will lead to differently sized coded frames. [16:44] interesting [16:44] is that the case for pretty much all formats? [16:45] (barring uncompressed AVI) [16:45] Most common "consumer" formats, yes. [16:45] bugger. [16:45] maybe for pro codecs like ProRes, but I wouldn't bet on it [16:46] my normal trick is just to find/replace various sequences in a hex editor to generate glitches, then screencap them back off of VLC [16:46] 99% of the time it renders just fine and has the usual blocking glitches [16:50] gsuberland: hmm in that case, you could go to the same length with ffmpeg. decode the 2 videos -> combine them to your creativity in the filter graph -> encode the resulting video [16:50] daemon404: I can't help but think this is our own doing <- I don't think so, why didn't you ask for the usecase / the sample? [16:51] because we enabled everyone to encode prores [16:51] what do you think all the horrible chinese things use to encode? [16:51] beastd - I'm not really familiar with that terminology. [16:51] they hack up ffmpeg. [16:51] also i have the sample. [16:51] i work for the source [16:51] Could you provide it? [16:51] not without asking the user first [16:51] (I believe I asked for such a sample a few times.) [16:52] i will ask the community team to ask. [16:52] Unfortunately the patch means that such samples will be created which I thought you would consider hackish. [16:52] Thank you. [16:52] oh [16:52] hmm [16:52] Could you provide ffmpeg -i output? [16:52] yeah i never though of that [16:52] that's bad [16:53] it's a shame the muxing and demuxing is tied like that... [16:53] ah well, nm. at least I know it's a bad idea to mess with data so heavily ^^ [16:53] thanks for the info [16:53] gsuberland: well i talk about using ffmpeg to do what you like, but we should talk in #ffmpeg [16:53] Just run "ffmpeg -i", I am very curious. [16:53] oh well, too late... [16:56] ffmpeg.git 03Benoit Fouet 07master:b35fa041521c: avcodec/pngdec: create a function to decode IDAT chunk. [16:56] ffmpeg.git 03Benoit Fouet 07master:4f313a50ee78: avcodec/pngdec: create a function to decode PLTE chunk. [16:57] cehoyos, i need to wait for vittorio to un-timeout [16:57] the user deleted the file, he has it somehwere. [16:58] Thank you. [16:59] Any chance that you forward such issues to FFmpeg in the future? Anything we can do to make this happen? [17:01] Did anybody read the HAP readme? It claims GPU support but I only see CPU encoding and decoding... [17:02] i usually give things to michaelni [17:02] most of the weird jpeg fixes have been from my file [17:02] s [17:03] He has the prores file? [17:03] no [17:04] not in this instance [17:04] just most instances in the past. [17:04] i dont forward files without asking the user [17:04] unless it is under a proper creative commons license [17:04] in which case i can. [17:06] Of course not. For many issues the "ffmpeg -i" input might already be sufficient. [17:07] most of the issues are more subtle [17:07] or output useless stuff [17:07] like "pixel format '-1'" [17:07] which is my favourite useless message [17:10] ffmpeg.git 03Benoit Fouet 07master:6499e63f7b9e: avcodec/pngdec: create a function to decode tRNS chunk. [17:10] ffmpeg.git 03Benoit Fouet 07master:c25b6ae8a2b9: avcodec/pngdec: fix some indentation/whitespaces [17:19] voting +1 for accepting string args in places that currently take confusing int (enum) args in vf_scale and other ffmpeg.c swscale CLI stuff [17:19] ffmpeg.git 03Benoit Fouet 07master:24ca2ffad826: avcodec/pngdec: use else if instead of if for small bpp handling. [17:19] ffmpeg.git 03Benoit Fouet 07master:8cab24df0780: avcodec/pngdec: create a function to handle small (<=4) bits per pixel values. [17:33] ffmpeg.git 03Benoit Fouet 07master:00df32f6a9c4: avcodec/pngdec: split frame decoding in its own function. [17:36] ffmpeg.git 03Carl Eugen Hoyos 07master:cde0ad5ea780: tests/Makefile: Fix path for creation of ffprobe-test.nut. [17:36] benoit-: why not "png_decode_frame" instead of "decode_frame_png" to be consistent with the other function in the file? [17:36] ^ that also seemed weird to me, looking at the commit message [17:37] without knowing what other functions are in the file, but just for consistency with a lot of other lav* stuff [17:37] in any case, nice cleanup... why are functions in libav* always so damn large [17:38] wm4: same reason ass_render_event is so damn large [17:38] one day, maybe I'll know what that reason is [17:39] rcombs: I have a PR that splits it up [17:39] it was ignored [17:40] benoit-: also too bad the return code are not forwarded [17:40] it's most of the time invalid data anyway, but enomem could happen [18:45] if any ffmpeg developers are here, I was wondering if you could cut a new 2.0.x release with the CVE-2014-527[12] fixes. We need to do a security update for Mageia 4 and it would be much better to have a new release. [18:48] I noticed that no patches have been committed to the 2.0 branch since 2.0.5, so I guess it would just be adding those CVE fixes [18:49] I see there's also a CVE-2014-854[1-9], not sure if any of those are relevant to the 2.0 branch [18:59] Luigi12_work, ok, ill try to make a new 2.0.x [19:01] ffmpeg.git 03Thilo Borgmann 07master:204533423945: lavd/avfoundation: Introduce device alias 'none' to allow the user to record only audio or video. [19:01] ffmpeg.git 03Thilo Borgmann 07master:d525e662e4f9: Changelog: Mention AVFoundation screen capturing. [19:01] ffmpeg.git 03Thilo Borgmann 07master:e6e614963046: doc/indevs: Rework and update documentation of AVFoundation device. [19:04] michaelni: thanks [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:dd72df9845d0: avformat/utils: do not wait for packets from discarded streams for genpts [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:c2af6b500bcf: avformat: add av_stream_get_parser() to access avformat AVParser [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:2428b02bb4be: ffmpeg: Use av_stream_get_parser() to avoid ABI issues [19:29] ffmpeg.git 03Anshul Maheswhwari 07release/2.0:3d10235b8352: v4l2enc: adding AVClass [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:8d48223a329a: avcodec/dvdsub_parser: never return 0 when the input isnt 0 [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:c7fac44ef83e: avcodec/dvdsub_parser: Check buf_size before reading 32bit packet size [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:65f581472c48: avcodec/dvdsub_parser: print message if packet is smaller than the packet size field [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:bcb10b99f4e8: ffmpeg_opt: Use av_guess_codec() instead of AVOutputFormat->*codec [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:04973b02c3aa: avformat/tee: flip assigment direction [19:29] ffmpeg.git 03Anton Khirnov 07release/2.0:c56d6f3552a0: cdgraphics: do not return 0 from the decode function [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:656f930160db: avcodec/iff: check pixfmt for rgb8 / rgbn [19:29] ffmpeg.git 03Christophe Gisquet 07release/2.0:b469fce85d8a: proresenc_kostya: remove unneeded parameters [19:29] ffmpeg.git 03Christophe Gisquet 07release/2.0:57a6cd8ab1c2: proresenc_kostya: report buffer overflow [19:29] ffmpeg.git 03Christophe Gisquet 07release/2.0:caf08defa69f: proresenc_kostya: properly account for alpha [19:29] ffmpeg.git 03Christophe Gisquet 07release/2.0:2958b8b86e36: wavpack: report if there is no bits left [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:9fc7de8d8030: avcodec: fix aac/ac3 parser bitstream buffer size [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:d4253b3a5b74: avcodec/utils: add GBRP16 to avcodec_align_dimensions2() [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:d76c9b5665a8: avcodec/snow: check coeffs for validity [19:29] ffmpeg.git 03Michael Niedermayer 07release/2.0:d7d29f0d433a: avformat/swfdec: Use side data to communicate w/h changes to the decoder [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:2fdb02693bbd: avformat/swfdec: Do not change the pixel format [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:6e70816f7191: avcodec/mpegvideo: Use "goto fail" for all error paths in ff_mpv_common_frame_size_change() [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:2b415192fd9a: avcodec/mpegvideo: check that the context is initialized in ff_mpv_common_frame_size_change() [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:beb83f0a4070: avcodec/mpegvideo: Set err on failure in ff_mpv_common_frame_size_change() [19:30] ffmpeg.git 03Katerina Barone-Adesi 07release/2.0:47fe68eec821: apetag: Fix APE tag size check [19:30] ffmpeg.git 03James Almer 07release/2.0:992ce9777c7e: x86/dsputil: add emms to ff_scalarproduct_int16_mmxext() [19:30] ffmpeg.git 03Gianluigi Tiesi 07release/2.0:f238de199053: avcodec/libilbc: support for latest git of libilbc [19:30] ffmpeg.git 03lvqcl 07release/2.0:2eda0e705a28: avutil/x86/cpu: fix cpuid sub-leaf selection [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:52fd0cda2cdd: avcodec/ac3enc_template: fix out of array read [19:30] ffmpeg.git 03Reimar D?ffinger 07release/2.0:8622618839c6: configure: add noexecstack to linker options if supported. [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:3f9a148022ef: avcodec/jpeglsdec: Check run value more completely in ls_decode_line() [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:49d69844f500: avcodec/mjpegdec: check bits per pixel for changes similar to dimensions [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:ac82e318bb81: avcodec/utils: Add case for jv to avcodec_align_dimensions2() [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:16775c7aaa07: avcodec/mmvideo: Bounds check 2nd line of HHV Intra blocks [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:abbfc4d87ec4: avcodec/tiff: more completely check bpp/bppcount [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:ae81d9a7da10: avcodec/pngdec: Check bits per pixel before setting monoblack pixel format [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:db48767d7c16: avcodec/pngdec: Calculate MPNG bytewidth more defensively [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:97fdbd12f909: avcodec/cinepak: fix integer underflow [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:be7105dff61e: avcodec/gifdec: factorize interleave end handling out [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:d36bb362a673: avcodec/qpeg: fix off by 1 error in MV bounds check [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:b4e0acfa043c: avcodec/smc: fix off by 1 error [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:ff59edb6dcd7: avcodec/svq3: Dont memcpy AVFrame [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:128b0510e159: avformat/mpegts: Check desc_len / get8() return code [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:0e7173826216: avcodec/h264: Check mode before considering mixed mode intra prediction [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:322470e60633: swresample/swresample: fix sample drop loop end condition [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:d7906baa8597: postproc/postprocess: fix quant store for fq mode [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:9b1673531cf9: postproc: fix qp count [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:70402f6ee786: avcodec/diracdec: Use 64bit in calculation of codeblock coordinates [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:c0be9d726440: avcodec/diracdec: Tighter checks on CODEBLOCKS_X/Y [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:2cf83677b314: avcodec/dirac_arith: fix integer overflow [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:808b0ccc03dc: avcodec/dxa: check dimensions [19:30] ffmpeg.git 03Michael Niedermayer 07release/2.0:85cac770bdb0: avcodec/dnxhddec: treat pix_fmt like width/height [19:31] ffmpeg.git 03Michael Niedermayer 07release/2.0:bde9e859b3c7: avcodec/utils: Align dimensions by at least their chroma sub-sampling factors. [19:31] ffmpeg.git 03Michael Niedermayer 07release/2.0:e2865d931618: avcodec/g2meet: check tile dimensions to avoid integer overflow [19:31] ffmpeg.git 03Michael Niedermayer 07release/2.0:8efb06c8735f: avcodec/cook: check that the subpacket sizes fit in block_align [19:31] ffmpeg.git 03Michael Niedermayer 07release/2.0:694c3dab363f: avcodec/svq1dec: zero terminate embedded message before printing [19:31] ffmpeg.git 03Michael Niedermayer 07release/2.0:0140f11c3bad: avcodec/h264_slice: Clear table pointers to avoid stale pointers [19:31] ffmpeg.git 03Carl Eugen Hoyos 07release/2.0:2389309d4836: lavc/utils: Make pix_fmt desc pointer const. [19:31] ffmpeg.git 03Michael Niedermayer 07release/2.0:05e5d785fa6a: avcodec/options_table fix min of audio channels and sample rate [19:31] ffmpeg.git 03Michael Niedermayer 07release/2.0:f8675743c40a: avcodec/utvideodec: fix assumtation that slice_height >= 1 [19:31] ffmpeg.git 03Michael Niedermayer 07release/2.0:0baeb59307e6: avcodec/wmaprodec: Fix integer overflow in sfb_offsets initialization [19:31] ugh, bloody release spam [20:01] ffmpeg.git 03Michael Niedermayer 07release/2.0:3d91569c5e39: update for 2.0.6 [20:36] J_Darnley: about those EXRs... tiled EXR are generated by most rendering software, and they are not properly handled by ffmpeg [20:37] i can provide example files if anyone is interested [20:39] Then I suggest you create a new issue on the bug tracker. [20:43] ok. I will. thx [20:47] some oddball effects: http://effectv.sourceforge.net/colstreak.html [20:47] maybe some could be ported as filters [20:48] yeah, oddball effects is totally what we need [20:53] at least they're in C [23:39] ffmpeg.git 03Thilo Borgmann 07fatal: ambiguous argument 'refs/tags/n2.0.6': unknown revision or path not in the working tree. [23:39] Use '--' to separate paths from revisions [23:39] refs/tags/n2.0.6:HEAD: doc/indevs: Rework and update documentation of AVFoundation device. [23:56] michaelni: thanks again for the 2.0.6 release. I'm guessing it contains fixes for CVE-2014-527[12] and CVE-2014-854[1-8] [23:58] yes, see http://ffmpeg.org/security.html [23:59] michaelni: ahh I had to shift-reload so it didn't reload from cache :o) [00:00] --- Sat Nov 15 2014 From burek021 at gmail.com Sat Nov 15 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Sat, 15 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141114 Message-ID: <20141115010501.E6E2A2AD60C0@apolo.teamnet.rs> [00:19] hi. wanted to ask about support for tiled openEXR ... [00:20] is it a planned feature ? [00:26] Rexor: if a contributor or a developer is interested in implementing it, then I guess it may be planned [00:26] you could submit a feature request on the bug tracker if ffmpeg does not currently support this (I know nothing of tiled exr) [00:27] you can even offer a bounty to get it implemented [00:32] thanks llogan [00:32] maybe I can ask this on #ffmpeg-devel ? [00:33] most 3d rendering software produces tiled EXRs [00:33] so they're a standard in animation and VFX [00:36] What is the pause before ffmpeg starts publishing a stream called and how can I reduce it? [01:11] I have tried every think I can think of after pouring over the documentation in order to do what I thought would be a simple task: Extract mulitple segments from a .mp4 file. It contains trailers footers and commercials I want to eliminate. I used ffprobe to find the breaks but I'm just not getting how to specify the start and stop times. I can do a single segment with "-ss 60 -to 120" for [01:11] instance. I want to put multiple segments into the output. any help greatly appreciated. [01:16] randomNumber: you can output each segment then use the concat demuxer to concatenate them, or use the trim and concat filters if you want to re-encode. [01:56] thanks. I did try that but for some reason the 2,3,4th segments seem to hang the program. It just sits there (NOT) blinking at me. eg no status updates for progress. do you know if the program takes a while to read the file up to the various entry points? [01:57] here are the lines I am using [01:57] ff -y -i tlr.mp4 -ss 37.72900 -to 386.8130 tlr1.mp4 [01:57] ff -y -i tlr.mp4 -ss 448.4490 -to 741.3250 tlr2.mp4 [01:57] ff -y -i tlr.mp4 -ss 833.3030 -to 1158.035 tlr3.mp4 [01:57] ff -y -i tlr.mp4 -ss 1188.252 -to 1595.430 tlr4.mp4 [01:58] maybe I'll look at the concat filter some more. [02:04] randomNumber: -ss after the input decodes to that point instead of seeking, so it takes longer [02:53] good morning guys [02:55] is it posible to detect and overlay a moving objects on Live Stream ?? [02:58] afaik, not without diving into the api and implementing it yourself [03:03] if someone can make a easy to use solution for me, i will pay the work ! [03:04] probably opencv [03:10] yes I 've heard about it, but if some one can make a solution, if will pay the work for that, please contact me on skype: pro-tech-ex [03:24] RedHat maybe this can help http://www.lavrsen.dk/foswiki/bin/view/Motion [03:34] pentanol thank for link, but this doesn't work for me [03:35] this is a simple motion detection... [04:30] RedHat do you want change live stream? [05:21] I am trying to call avconv from another program. avconv complains that [05:21] Error initializing filter 'select' with args 'eq(pict_type\,I),setpts=N/(25*TB)' [05:21] Error opening filters! [05:21] but the command can executed without problem directly in the console [05:22] what cause the difference [05:22] the command I am using is [05:23] avconv -i news.avi -vf select='eq(pict_type\,I),setpts=N/(25*TB)' -q 1 news/%09d.bmp [05:27] xieyi: for avconv support ask in #libav [05:29] ok [08:19] Hi all [08:21] i was wondering is it possible to get RAW H264 data from a camera received by a socket and decode it on the server side ? [12:59] it possible to live stream a video to html5 video in desktop browsers? [12:59] mobile has HLS. does desktop has something similar? [13:10] Trying to compile ffmpeg 2.4.3 with the "new?" MSVS 2013 community edition (The one that was made free the other day) [13:10] just installed MSYS [13:10] and getting "cl is unable to create executable file" [13:11] config.log says in the end [13:11] link -o ./ffconf.QPprqeef.exe ./ffconf.LGVmgfGf.o [13:11] link: invalid option -- o [13:11] Try `link --help' for more information. [13:11] C compiler test failed. [13:14] It should be /out:XXX.exe even if i think -out:XXX.exe will work (It's specified in the latter way in probe_cc in the configure script) [13:18] hah [13:18] it's assuming it's the shell link [13:20] (ie making a file link) [13:23] ok [13:23] setting ld_default="cl" fixed it seemingly [13:25] (Under the toolchain detection) [13:52] argh [13:52] now linking dll's fails [13:52] Action: jonaslund deletes msys link [13:54] and revert configure change [14:28] Is there a ffmpeg command line option to make the program pause after the output on errors? Whenever there is an error, it flashes by and disappears and I never know what was wrong. [14:28] ffmpeg appears to fail to convert a 56 MB PNG to JPEG. Other, smaller PNGs can be converted with the exact same command. Why does it fail at 56 MB? (Latest stable) [14:28] (I don't want this to happen when it has success, though, but I can't find a way to get either to happen.) [14:52] martyj-o: ffmpeg halts on errors, but you can send all its output to stderr if you need to review it. ffmpeg input output 2>ffmpeg.log [14:53] er, the console output is sent to stderr [14:59] pastebin what you're doing for more help [15:07] hi. is there support for the quicktime hap codec (https://github.com/Vidvox/hap-qt-codec), maybe in development? does anyone know of any other means of hap compression using linux? [15:15] relaxed: Well, I just run a ffmpeg.exe command through a Windows context menu shortcut, and it usually works, but if it fails, it just flashes by (command prompt that doesn't pause). [15:16] flavioberetti: that code looks like it's all 3- or 2- clause msd or mit licensed, should be ok to use to to make an ffmpeg codec from. But I haven't heard about anyone doing it... [15:17] martyj-o: either wrap the windows command through an batch file or add the pipe relaxed described to the launch command [15:17] that way you'll get the output in another file [15:17] in the batch file you can end it with a pause command [15:17] No batch file possible in this context. [15:18] kepstin-laptop: ok thanks. i will ask the ffmpeg-devel list then in case noone else has an idea. [15:18] And it must not pause on success. [15:18] you can write a batch file that only pauses on error [16:54] GOOD morning! [16:54] I'm experimenting with using ffmpeg to resize animated gifs, and if I read and write to disk it works just fine [16:54] BUT [16:55] I'm trying to do it over stdio and I'm getting CRYPTIC FAILURES https://gist.github.com/jesusabdullah/551889ce44bb4e112ec4 [16:55] ideas for troubleshooting? [17:00] i'm not familiar with gif or avformat it general, but [17:01] some formats need seeking backward and you can't do that over pipe [17:01] also using stdin/out is quite bad idea, ffmpeg have a bad habit of reading input and printing stuff.. [17:02] you might try a named pipe `mkfifo` [17:03] yessch [17:03] that won't do at all [17:05] my instinct is that gif does not need seeking [17:05] but that's just instinct [17:08] jesusabdullah: windows ? [17:08] ah no [17:08] nvm then [17:10] jonaslund: osx [17:11] I *wish* I could blame windows [17:11] FUCKING WINDOWS 98 GET BILL GATES IN HERE [17:12] jesusabdullah: Your commandline works for me. For 2 random gifs. [17:13] hmm [17:13] interesting [17:13] jesusabdullah: Could depend on your input or on your platform [17:13] yeah [17:13] now I'm wondering which [17:13] you might want to file a bug report if you can provide the input sample [17:14] hmmm [17:14] yeah maybe [17:14] you know where the bug tracker is offhand? [17:14] also, completely unrelated: what's the difference between ffmpeg and avconv? [17:15] jesusabdullah: https://ffmpeg.org/bugreports.html [17:16] hello, Can anyone point me to where I can find out what is "fd" in the ffplay output? google&documentation gave me no info :-x [17:17] thanks c_14 [17:19] Mansor: is->frame_drops_early + is->frame_drops_late [17:19] https://git.videolan.org/?p=ffmpeg.git;a=blob;f=ffplay.c;h=f79161dd00601c4e1f20892fa07a73ce1ae2db3c;hb=HEAD [17:20] cool thanks! [17:24] beastd: what's your version? Maybe whatever comes with homebrew is different [17:31] hi [18:36] <_Vi> How do I `setpts` without transcoding (i.e. with `-vcodec copy`)? [18:37] you don't [18:38] <_Vi> What tool to use instead? [18:38] You can't speed up or slow down a video stream without reencoding. (that I know of) [18:39] <_Vi> I want to make slow and long 1 FPS video from usual one, without transcoding? I.e. multiply each PTS by 20 without changing the encoded data. [18:40] <_Vi> Workaround way: 1. Convert to matroska (without transcoding); 2. Convert matroska to XML (using my tool); 3. Fix timestamps using textual tools; 4. Convert back to matroska; 5. Process matrosksa with FFmpeg to mux it normally. [18:40] if any ffmpeg developers are here, I was wondering if you could cut a new 2.0.x release with the CVE-2014-527[12] fixes. We need to do a security update for Mageia 4 and it would be much better to have a new release. [18:44] _Vi: you can try changing the container level timestamps/frame rate, but that won't always work. The best choice for something like that would be mkvtoolnix or mp4box depending on your format [18:44] Luigi12_work: Try asking in #ffmpeg-devel if someone can backport the commits to the 2.0.x release branch [18:46] c_14: thanks. The commits may very well be in the 2.0 git branch already (I haven't checked), so it might be as easy as cutting a new release. I asked in #ffmpeg-devel though. [18:47] c_14: ok I just checked, no commits in 4 months, so yes they need to be backported too [19:09] Hi, does anyone know how to record 32bit audio from alsa with ffmpeg ? [19:11] As far as I'm aware, the same way you'd record audio with a different bit depth from alsa. [19:12] I tried to set acodec pcm_s32le but that still records 16bit, and -sample_fmt did not work either :/ [19:23] c_14: ran some tests again, so with "avconv -acodec pcm_s32le -f alsa -i hw:1,0 test.wav" I get "[pcm_s32le] invalid PCM packet" [19:23] Maybe it's big endian? [19:24] c_14: actually, I was not getting this before. If I add -ar 96000 it was just silently recording 16bit [19:24] c_14: arecord works with the little endian [19:33] ffmpeg -f alsa -c:a pcm_s32le -i loopout -c:a pcm_s32le out.wav <- works for me [19:33] loopout is my device name [19:34] If that doesn't work for you, try compiling ffmpeg: https://trac.ffmpeg.org/wiki/CompilationGuide/Ubuntu and then try again with that [20:09] c_14: This does not work on the laptop (libav) but works here on my workstation, so that might be an error introduced by the fork [20:25] ffmpeg appears to fail to convert a 56 MB PNG to JPEG. Other, smaller PNGs can be converted with the exact same command. Why does it fail at 56 MB? (Latest stable) [20:26] Is there a ffmpeg command line option to make the program pause after the output on errors? Whenever there is an error, it flashes by and disappears and I never know what was wrong. [20:26] (I don't want this to happen when it has success, though, but I can't find a way to get either to happen.) [20:26] I just run a ffmpeg.exe command through a Windows context menu shortcut, and it usually works, but if it fails, it just flashes by (command prompt that doesn't pause). [20:33] hrm [20:33] I wonder if old DVD players will play 16:9? [20:33] or just 4:3? [20:38] https://trac.ffmpeg.org/ticket/3797 was this bug only fixed for hls output? [20:43] It would appear so. Can you reproduce the issue on git master HEAD with something else? [20:54] i can reproduce it with 2.4.3 with flv output [20:55] I'm gonna try to get master HEAD to build [21:05] c_14: It works with recent ffmpeg, thanks for the hint [21:22] martyj-o: are you using a recent build? [21:32] Hello again. [21:33] Can somebody say me that what's wrong with AAC? License, patent or what prevents it to be included into binaries? [21:37] Mainly licensing issues. [21:37] c_14, i've got the test running, hopefully it shows up [21:37] The native encoder is incnluded though. [21:38] *included [21:43] Linux streaming is easy. [21:43] Any ideas about what can I stream? [21:44] Dota 2 is everywhere, so it's bad choice. [21:46] Dota 3 :P [21:47] Gaben does not know what is "three". IceFrog still makes DotA. [21:47] We will never see it. [21:50] Visual novels. [21:50] Yes. [21:56] I'm getting segfaults when trying to burn timecode with -filter_complex drawtext ... works well on plain text ( text="foo" ), and the same filter graph used to work well in previous versions of ffmpeg. [21:56] anyone else see this issue? [22:01] What version are you running? What version did it work with? What's your commandline and complete output (pastebin)? Can you run a git bisect? [22:09] Doesn't work with version 2.4.1, worked previously with 2.2.1, http://pastebin.com/D5DzuvT1, can't run git bisect at the moment (but will be able to later tomorrow) [22:11] c_14: updated pastebin to include output: http://pastebin.com/Fq4AfskW [22:12] Hm, probably a bug. [22:12] Can you make an issue with the output from that pastebin, the core and if you can the commit where it breaks (using a git bisect). [22:13] c_14 the warning occurred with head, i'm going to prepare a paste [22:15] sorry, false start here, i didn't capture the commandline [22:17] Question: Using avcodec, I'm reading in an audio media file. I'm trying to get seconds elapsed. Using AVPacket.pos returns the byte position in the stream. I tried using AVPacket.pos / frame->sample_rate but it is inaccurate. What am I missing and or overlooking? [22:20] HSD, all packets have timestamps [22:20] pts [22:21] you will have a timescale (aka how many ticks is one second), as well as an X/Y timestamp [22:23] ah, so I'm reading the wrong thing? [22:24] I noticed on AVStream->time_base is a very small number. 1/1441..... I forget [22:24] is that what you mean by timescale? [22:26] checking doxygen for pts doesn't give much info. (unless I'm looking in the wrong spot) [22:26] via the AVPacket Struct Reference [22:29] c_14 here's the build script i made and used using git master/head: http://dpaste.com/1811QR2 [22:30] c_14 here's the ffmpeg command line and output: http://dpaste.com/2VCNT54 [22:31] llogan: Latest stable, as mentioned. [22:34] danomite-: do those warnings appear when outputting to a file? [22:35] c_14, i'll give it a whirl, should I do file only or file and stream publishing? [22:36] do file only [22:52] JEEB, thanks! I got it figured out. packet.pts * av_q2d(container->streams[stream_id]->time_base) [22:53] :) [22:54] c_14 if the error never happens with a file what does it mean? [22:55] That the error is in this case (assuming you used an flv output file) in the rtmp protocol implementation. You used the same commandline with exception of the output file, right? [22:56] correct -f flv -vcodec copy -an out.flv [22:57] Right, if you use that output file as the input for your stream to rtmp, do you get the error? [22:57] will test that now, [23:05] Is there a ffmpeg command line option to make the program pause after the output on errors? [23:06] martyj-o: not internally, no. if ffmpeg throws an error it dies [23:06] I'm not sure why but the stream does not want to play properly: http://dpaste.com/3K0FSX2 [23:06] :( [23:07] *file [23:07] ffmpeg appears to fail to convert a 56 MB PNG to JPEG. Other, smaller PNGs can be converted with the exact same command. Why does it fail at 56 MB? (Latest stable) [23:08] martyj-o: do you have such a png I can test with? [23:12] c_14: Not feasible to share it. [23:12] It's year 2014 and no secure way to send files still [23:12] Also, it's kind of private. [23:13] I can try upscaling something... let me see [23:30] c_14, would it be worth filing a bug report at this point? [23:31] yep, and if you can upload the out.flv you used as input for the 3K0FSX2 command so the devs can test [23:32] c_14, I'm not even sure that video has the problem but i'll give it a whirl [23:32] It says failed to update header with correct duration, so at the very least something is wrong [23:33] yep, thanks [23:49] c_14, what could I use to host the video file? [23:49] Either upload it directly to the tracker, or to the ffmpeg ftp [23:49] https://ffmpeg.org/bugreports.html [23:49] See submitting sample media [23:53] I'm not sure that the 10MB would be useful, i'll put it on my google drive [23:58] They'll much prefer it on the ftp, if it isn't useful they can always delete it. [23:59] I'll try to get it on there after I make the bug report intelligible. [00:00] --- Sat Nov 15 2014 From burek021 at gmail.com Sun Nov 16 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Sun, 16 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141115 Message-ID: <20141116010502.095B22AD60C0@apolo.teamnet.rs> [00:13] c_14, here's the bug report: https://trac.ffmpeg.org/ticket/4110 [07:05] I'm using deshake=edge=0 and I get a green margin - how can I make it black? [08:49] How do you troubleshoot a codec appearing in the configuration list, but no the -codecs list? [08:51] I.e. when I do ./configure --enable-libfdk_aac, it tells me libfdk_aac is enabled, but ffmpeg -codecs | grep fdk doesn't show anything, and trying to encode with it says "Unknown encoder" [09:04] vity: make sure you're using the correct binary. what does "which ffmpeg" return? [09:05] relaxed: /usr/local/bin/ffmpeg , but I get the same problem running ffmpeg from the compilation directory [09:07] ffmpeg -codecs|grep aac [09:07] shows aac and aac_latm [09:07] only aac has the encode flag though [09:09] run configure again and post your config.log somewhere [09:12] http://privatepaste.com/download/af6ddcbf7a [09:13] ok, now run make [09:14] then run "./ffmpeg -codecs|grep aac" (the ./ is important) [09:15] that's what I was doing when I said "from the compilation directory", but make just returns a list of HTML and MAN files since I didn't clean it after the last time [09:16] It still just has aac and aac_latm [09:17] The one message I get that might help, is up at the top it says "WARNING: library configuration mismatch", but googling that makes it sound inconsequential [09:24] vity: run "LD_LIBRARY_PATH=/usr/local/lib ./ffmpeg -codecs|grep aac" [09:26] relaxed: still just aac and aac_latm, but without the configuration warning [09:26] are you following this guide? https://trac.ffmpeg.org/wiki/CompilationGuide/Ubuntu [09:27] vaguely. I'm in arch instead of ubuntu, but otherwise pretty much the same [09:28] follow the guide [12:54] hi [12:55] hi all, i want to serve an offline video file on apache, for some reason loading the file in my videoplayer works fine but cannot seek unless the video file buffers in the player, any idea how to provide a seeking (youtube like where you can jump in into different time frame and play from that point in time) ? [12:56] pagios: which video container? [12:56] relaxed: mp4 [12:57] did you use "-movflags faststart" or qtfaststart? [12:57] yes [12:57] do you have any mp4 file samepl file working with seeking? [12:57] i can then compare [12:58] hmm, are you using a flashplayer? [12:58] or html5? [12:59] relaxed: flash[player for RTMP and html5 for hls [13:00] hmm http://pl.wikipedia.org/wiki/HTML5_video#cite_note-3 is this real? i need only mp4 (H.264) to cover all browsers? [13:01] Hi guys - I'm trying to run ffmpeg on modulus.io (node.js), and I keep running into an error "spawn ENOENT". From what I understand this means that the path to ffmpeg can't be found. Is this the *only* reason that I would be getting this error? Thx in advance :) [13:02] (*it works fine on my local machine btw) [13:03] relaxed: do you have any working mp4 sample file to share? [13:04] pagios: I do not, sorry. Streaming isn't really in my wheelhouse. [13:04] I have ffmpeg outputing an rtp stream via multicast, on the other end I have rtpdump listening and outputting info about the stream, I have found several instances where the timestamp goes in reverse, does it seem to be related to this bug: https://trac.ffmpeg.org/ticket/4110#comment:6 [13:14] hello [13:14] can ffmpeg stream rtsp from rtmp in live mode ? [13:15] yes [13:15] then I don't need any temporary huge files on disk right ? [13:15] but it's not a space problem [13:16] I've only been able to get ffmpeg to publish stream to an rtmp server, no intermediate files [13:16] it's a bit convoluted i had to use all kinds of steps, basically, the RTMP stream comes from the web, and i'd like to play it on a nokia phone with accepts only rtsp with plain-mp4 [13:17] well i have to set up a RTSP server myself right ? .. total newbie when it comes to this [13:18] althought i've done some research with HSL because i am still not able to find a program to dump those stupid adobe flash segmented streams [13:18] I've only used rtmp successfuly with the nginx plugin [13:18] maybe you have an idea about that ? [13:18] oh btw im on win7 [13:18] Are you serving from win7? [13:19] can you give me a few pointers what should I search for ... what's this Wowza thing i keep getitng across [13:21] those m3u8 streams I haven't found a way to download, it says it's a 2GB file, it let's me only download a few megs, or about 14 seconds, and i've sniffed for some settings that kind of streaming and tried to use URL tricks to change parameters but i was unsuccessful at changing the timing that controled the 14 sec cutoff [13:21] segmented whatever, but that's unrelated [13:21] no I'm not streaming anything at the moment [13:21] I just got into this, been planing for a few days [13:22] here's the deal, the source is RTMP and it's from the web, externally [13:23] i need to caputre that, conver it to a visual-mp4 that older nokia phones can handle and stream that via RTSP from my Win7 PC over my Wireless router [13:23] I haven't been successful at serving rtsp [13:23] the conversion process would also involve resizing ... etc to fit nokia mp4 specs [13:25] i was just trying to find out if there's an easier solution, the obvious it's just to dump the RTMP with RTMPDUMP, conver the live with ffmpeg, stream with another program, but the problem is it has to be live, ffmpeg would need to keep reading the file as new data is being written [13:25] and i don't think it'll recognize that unless it's assigned a protocol to get it into live mode [13:25] the source is live [13:25] it's not vod, it's not rewindable [13:26] I've run into live555 proxy server which can serve rtsp but I've only worked with it a couple hours today on linux [13:26] but it does have like a 5 second cache [13:26] did you ever just consider using hls paramters to limit the size of your local files? [13:27] Sorry I don't understand that question [13:27] that HLS talk is unrelated sorry i mixed it up with this [13:27] Just an FYI, HLS can create a "loop"ed playlist [13:30] not sure what that means [13:30] i am not the owner of the RTMP stream, it's external [13:30] it's a show on the web, not my stuff [13:31] i'll show you what rtmpdump writes into the file [13:32] http://pastie.org/private/sfvymcsi77gkuy3vktsw [13:36] Is that the original source of the stream? [13:36] ofcourse [13:38] at the time the show is live, i usually go exercise and I don't like to be sitting on the PC doing nothing just listening for 3 hours ... that's why I use a cell phone to get internet radio via wireless [13:39] but the audio quality is better on the video stream so I thought I might just use some good software to get me some RTSP on that phone, it would be easier if I had a smartphone, but I don't like them and even if I would I don't have the extra money since I spend it on more imporant things [13:41] nokia is E72 .. i think it's a 16:9 screen so that would be easier to convert than to square one that I got previously [13:43] oh, by the way, it's live only once a day, even tho the streaming is 24/7 - those are all rebroadcasts [13:43] so you get the video now but it just a repeat [13:47] it's the Real Player that handles RTSP on nokia, i'll fetch those specs, but maybe i come across a alternative video player for the device but that's a low chance [13:47] the codec is Visual-MP4 , not AVC (h264) [13:50] actually i might have found some conflicting info, i only got this phone a few days ago as I had N79 for many years [13:52] some overview http://www.techradar.com/reviews/phones/mobile-phones/nokia-e72-680833/review/7 [13:53] not sure google search throws out some other supporting formats [13:53] i need to check directly with realplayer on there for what's actually the deal [14:01] more specifically the mp4 subtype is MP4V-ES [14:02] that's the format is says it is on nokia, the video from web i converted with Any Video Converter Ultimate specifically setting to "nokia MP4" profile [14:26] Anyways can anyone tell me what's the point of HLS and this whole Segmented streaming ? [14:27] that makes dumping, DLing a rewindable stream basically impossible without some utilities I may not know about [14:30] dumping an hls stream is easy: ffmpeg -i http://foobar/baz.m3u8 out.mpegts [14:31] it's also nice for adaptive streaming, mobile devices, connection roaming, etc. [14:32] And for not being flash [14:33] thanks [14:33] do you guys have any idea on the rewindable streams, i mentioned it before, one of them had a 14 hour cache ... news networks like to use this [14:34] can I dump that 14 hours all in one file too ? [14:34] well that's where the problem many times [14:35] it's that it isn't always identified as a m3u8 stream imo, not in the url at least [14:36] im mostly using GetFLV to get the source links, worked fine so i haven't been using anything else [14:36] one of the links looked like this: http://livestream-f.akamaihd.net/9033656_3137612_c50787c5_1_678 at 16885?v=3.0.3&fp=WIN%2015,0,0,189&r=YSMHH&g=OHXIDQNYJVMU&seek=187&transferToken=oL+uZ4oZ&session=AQBWOiKmBRa/10jkSlRWwG2mj8RfyuPrV8TBdMcw/3UBzWkOC0wIllvOOy8va1pi4DauBPcj [14:37] that particular one is Komo News [14:49] I found the m3u8 for that stream. [14:49] found those specs [14:49] http://developer.nokia.com/community/wiki/Recommended_video_encoding_settings [14:49] But they have some weird auth-shit going on. [14:49] well you'd used some kind of utility right, i couldn't get through, i tried manipulating URL tags to start downloading from the beginning of the cached stream [14:50] couldn't get past the 15 sec segment cuttoff, the downloaders report filesize as 2GB, but only downloads like 4 Megs before "size mismatch" [14:51] this pesky streaming is usually on the large CDN providers etc, and akamai ofcourse [14:57] Hello i'm rrying to embed subtitles in a movie ith this command: [14:57] i forgot to say you might get rid of that token and session id because it's like a week old and i think that's part of the auth stuff even tho the links worked without the tokens and session id when just raw DLing [14:57] ffmpeg -i movie.avi -sub_charenc UTF-8 -i subtitles.srt -scodec mov_text -acodec copy -vcodec copy -f mp4 output.mp4 [14:58] The process is working fine, no errors, however when i start the movie in my player subtitles arenotworking and the video codec is messed up a bit [14:58] the token and session seem to be added after stream start playing, the first link GetFLV throws out doesn't have much of those params after ?v [14:58] Wader8: I went directly to the komo news site, clicked on live streaming and went looking for the url from there, but haven't been able to make the auth work correctly. [14:58] yeah exactly ofcourse i should have gave you that link too [14:58] Popara: what player are you using? [14:59] VLC [14:59] does the mov_text only works when the input movie is mp4? [14:59] cause i tried with an mp4 movie and it worked [15:06] It should work regardless. [15:06] wait [15:07] guys [15:07] trying with VLC [15:07] streaming RTSP [15:07] http://pastebin.com/sk4rw5dp [15:07] what do I need to enter in out URL [15:08] rtsp://:8445/mystream [15:08] i only want to use it on my own home network [15:08] rtsp://localhost:8445/mystream [15:08] is that OK ? [15:08] because im under router, do i need port forward ? [15:08] VLC doesn't specify the full url [15:09] or I just put the ip of my DCHP [15:10] dhcp [15:10] Put the address where you want to get the stream to go. (probably) [15:11] Popara: hmm, can't see anything wrong with that [15:11] Yes told you, i dont see any errors either [15:11] i'll try my internal IP [15:11] but in the target video subtitles are not working w8 i will take a screenshot [15:11] Can you try with mplayer/mpv ? Or some other media player? [15:12] ok w8 [15:16] no subtitles are not working with any other player either [15:17] When is the time to show some subtitles then the video freezes [15:17] then the same again... [15:17] i will convert the movie to mp4 and then i will try the above command again, i think with mp4 it worked somehow wait [15:19] no the same happened [15:20] :/ [15:24] I just tested with a random avi and a random srt file I had laying around on my pc and it worked. [15:24] can i give you a movie and the srt files to test it? [15:24] sure [15:31] damn VLC gets an encoder error when i try to stream a file that's being dumped by rtmpdump such a pity [15:33] VLC plays the file normally though [15:33] on it's own [15:37] well im a bit closer seems like there's some connection, ddwrt reports destination ip as the gateway [15:37] connection is still UNREPLIED, but nokia at least gives me "cannot play file" [15:38] ... and it's getting time for me to hit the bed ... ah i won't be finishing today :( [17:33] hi, I've been reading online but haven't found a concrete solution: is it possible to add an "fps" watermark to an encoded stream with filters, and if so, how? [17:33] thanks in advance [17:41] wickwire: you mean frame count? [17:42] yes that's it [17:42] I'd just like to be able to add a frame count [17:42] every second perhaps [17:42] hmmm now that I think about it, maybe I shouldn't be doing it on the actual video, [17:43] I mean, [17:43] I'm streaming the video [17:43] and I'm comparing LAN and internet performance [17:43] considering the network CPE upload [17:43] to the internet [17:44] perhaps if I write the actual fps on the video, it won't be totally reliable reaching the player...? [17:44] sounds like that would be the player's job [18:34] hi im trying to use ffmpeg on my android and i used ffmpeg media encoder and ffmpeg4 android and both are crashing after a while during h264 encoding and iget no error message [18:34] my android is 4.1.1 and processor is omap [18:35] is it impossible to run ffmpeg on the omap processor without crashing troubles? [20:06] anyone have working code to stream a webcam I want to show it on a web page for remote viewing application [20:06] Ubunut 14.04 [20:07] ffmpeg -f video4linux2 -s 640x480 -i /dev/video0 http://localhost:8090/feed1.ffmcat is getting close [20:08] except is say Missing audio stream which is required by this ffm [20:10] I don't care about the sound is there a way to provide a null sound input? [20:10] well do you have audio defined in your ffserver configuration? [20:10] set NoAudio. [20:12] Yes I'm using https://www.virag.si/2012/11/streaming-live-webm-video-with-ffmpeg/ [20:13] So in the Stream block one line command NoAudio [20:13] or is it set NoAudio [20:16] hmmm I set [20:16] # This is the input feed where FFmpeg will send [20:16] NoAudio [20:17] same error [20:31] biofool: add -f lavfi -i aevalsrc=0 [20:34] c_14: --- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x2862418} --- [20:34] +++ killed by SIGSEGV +++ [20:34] Segmentation fault [20:34] Really explosive :^) [20:34] In ffmpeg or in ffserver? [20:35] ffmpeg -f lavfi -i aevalsrc=0 -f video4linux2 -s 640x480 -i /dev/video0 http://localhost:8090/feed1.ffm [20:38] fflogger: that's the exact command. output was those 2 lines [20:39] Didn't even print a version? [20:39] And ffloger is a bot. [20:39] my bad. No there was more [20:39] too many windows [20:40] http://pastie.org/9721628 is command and output [20:43] Try with -af aevalsrc=0 instead [20:47] tnx. I rebuilt from the latest source now "ffmpeg -f lavfi -i aevalsrc=0 -f video4linux2 -s 640x480 -i /dev/video0 http://localhost:8090/feed1.ffm" [20:47] works. [20:52] except if I stream ogg I get a player but no video shows up [20:54] where do you "add '-strict -2' if you want to use" The encoder 'vorbis' is experimental but experimental [20:56] like so? ffmpeg -f lavfi -i aevalsrc=0 -f video4linux2 -s 640x480 -i /dev/video0 -strict -2 http://localhost:8090/feed1.ffm [20:57] Now: Stream mapping: [20:57] Stream #0:0 -> #0:0 (pcm_f64le (native) -> vorbis (native)) [20:57] Stream #1:0 -> #0:1 (rawvideo (native) -> h263p (native)) [20:57] Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height [21:01] does this look right VideoCodec h263p [21:01] VideoSize 720x576 # Video resolution [21:01] VideoFrameRate 9 [21:01] 0:0 is the vorbis stream [21:03] guys is VLC using internal ffmpeg for it's streaming ? [21:04] or i'll have to go ask them [21:04] im doing the RTMP to RTSP thingy [21:04] I've made at least some progress [21:04] vlc uses ffmpeg only as optional de/encoder plugin. [21:06] I dumped RTMP with RTMPDUMP, Streamed the file using VLC as a Visual MP4, got it to work, but the audio's not working, even tho i tried the exact same source configs, and then tried the MP4 Nokia Specs and it didn't work ... I did select AAC sound, but the Nokia Specs has MP4 AAC LC, that LC thing im not sure what is [21:06] and VLC doesn't have the LC thingy at the end [21:06] but anyway i'll have to go there and see what I can do with VLC and maybe i'll just have to come back to ffmpeg if it doesn't work [21:07] because, i can see, it's disconnecting every 20 seconds [21:07] without sound i couldn't figure out of it's repeating or it's continuing where it left, but disconnecting every 20 secs is unacceptable [21:12] c_14: now I see "Codec bitrates do not match for stream" from ffserver [21:12] pastebin your ffserver.conf and commandline please [21:15] http://pastie.org/private/umpi69jv9eyfpyhmia [21:16] your ffserver.conf file? [21:18] repasted + confi http://pastie.org/private/rsf6o7weio6llz28ieibow [21:19] Well, you commented out the audio settings. [21:20] right when I was debugging the audio issue [21:21] Also, you should probably get rid of the +global_header [21:22] And h263p won't work in webm [21:26] OK. what codecs can I use in webm? [21:27] vp8, vp8, vorbis, opus [21:27] vp9 [21:27] Aaaand, that's basically it. [21:28] Error reading configuration file '/etc/ffserver.conf': Invalid argument [21:28] I meant /etc/ffserver.conf:25: Unknown VideoCodec: vorbis [21:29] vorbis isn't a video codec [21:29] /etc/ffserver.conf:25: Unknown VideoCodec: vp9 [21:29] You might have to write libvpx-vp9 [21:30] And vp9 isn't really that good for streaming. [21:31] Because it's reeeeally slow. [21:32] I'd be so glad to get something working that I could tweeklater I'm fine with slow [21:33] so ffserver -codecs|grep vp8 [21:33] so ffserver -codecs|grep vp9 [21:33] shows ffserver version N-67694-gdcb10ef Copyright (c) 2000-2014 the FFmpeg developers [21:33] built on Nov 15 2014 11:40:19 with gcc 4.8 (Ubuntu 4.8.2-19ubuntu1) [21:33] configuration: [21:33] libavutil 54. 11.100 / 54. 11.100 [21:33] libavcodec 56. 12.101 / 56. 12.101 [21:33] libavformat 56. 12.103 / 56. 12.103 [21:33] libavdevice 56. 3.100 / 56. 3.100 [21:33] libavfilter 5. 2.103 / 5. 2.103 [21:33] libswscale 3. 1.101 / 3. 1.101 [21:33] libswresample 1. 1.100 / 1. 1.100 [21:33] D.V.L. vp9 Google VP9 [21:33] Did you not build with libvpx ? [21:34] I did a default build [21:34] You should probably rebuild with libvpx and preferably also libvorbis [21:35] Hello, i'm trying whole day to combine multiple video files together and i can't. It must be FFmpeg issue. I'm trying the concat demuxer using ffmpeg -f concat mylist.txt out.mp4 and it just fails whenever it goes to the second video no matter what the video is [21:36] do you know the specific options or where I find them? [21:36] Pastebin your mylist.txt the ffmpeg output and an ffprobe of the first 2 videos [21:36] ok [21:36] biofool: hmm? --enable-libvpx --enable-libvorbis [21:37] vorbis is listed as an enabled decoder from ./configure [21:37] so is vp9 [21:37] You want the encoder. [21:37] Not the decoder [21:39] http://pastebin.com/Jkdt1MCX [21:39] The 1.mp4 & 2.mp4 are different, the first one is ~23:06 minutes and the other one is only some seconds. You will notice that FFmpeg at the end it says 44 minutes both files which is out of range [21:40] ERROR: libvpx decoder version must be >=0.9.1 so I should build from https://chromium.googlesource.com/webm/libvpx? [21:40] the 1.mp4 and 2.mp4 are using exactly the same codec as i transcoded them first [21:40] In the last command if i use the '-re' i get many errors at the end which are not shown there for some reason without the 0re [21:42] biofool: probably [21:42] Popara: 2 things, the fps isn't the same in both videos. That might be an issue. Also the first audio stream is stereo the second is 5.1, that might be an issue. [21:43] Test with -an and then with -vn [21:43] See if either works. [21:44] like that? ffmpeg -f concat -i mylist.txt -an out.mp4 [21:44] yep [21:44] what -an does and -vn remind me, [21:45] no audio/no video [21:45] ah yea [21:45] ok i have to wait a bit until it finish [21:48] c_14: with -an it said again time=00:44:08.08 which is too much [21:48] But did it finish? [21:49] yeah no errors [21:50] c_14: i just cp the 2.mp4 (the 1minute video) 2 times, so i had 2.mp4 and 3.mp4 [21:50] i then used the concat and it worked fine [21:50] so they have to be exactly the same in their codecs right? [21:51] yep [21:51] well thats an issue now because i want to join more than 10 files. Is there any way to transcode them all in the same way somehow [21:51] yes [21:52] ffmpeg -i video -r 25 -s [size] -c:v [codec] outfile [21:52] That should do it. [21:52] Probably. [21:52] Maybe. [21:52] eh [21:52] You'll also need the audio foo [21:53] cant this be done in the concat command? For example [21:53] ffmpeg -f concat -i mylist.txt -r 25 -vcodec libx264 -acodec aac -strict -2 out.mp4 [21:53] and do it there and not before? [22:02] c_14: there is a concat filter which says that combine video files with different codecs [22:02] https://trac.ffmpeg.org/wiki/How%20to%20concatenate%20(join,%20merge)%20media%20files [22:02] I assume this is what we want [22:02] i will try now [22:03] i like the way it started [22:03] Stream mapping: [22:03] Stream #0:0 (h264) -> concat:in0:v0 [22:03] Stream #0:1 (aac) -> concat:in0:a0 [22:03] Stream #1:0 (h264) -> concat:in1:v0 [22:03] Stream #1:1 (aac) -> concat:in1:a0 [22:03] concat:out:v0 -> Stream #0:0 (libx264) [22:03] concat:out:a0 -> Stream #0:1 (libfaac) [22:07] OK have libvpx installed and now ffserver is OK with VideoCodec libvpx [22:07] but [22:07] ffmpeg -f lavfi -i aevalsrc=0 -f video4linux2 -s 640x480 -i /dev/video0 -strict -2 http://localhost:8090/feed1.ffm [22:07] gives the same error message [22:07] [vorbis @ 0x340ac00] Current FFmpeg Vorbis encoder only supports 2 channels. [22:07] use libvorbis [22:07] Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height [22:09] c_14 wohoo it worked! And i was also able to combine different type of containers like avi,mp4,mpg with different type of codecs [22:09] i dont know what that does exactly but it worked [22:24] any idea if ffmpeg makes MP4 VISUAL subtype as ES when streaming RTSP automatically [22:24] streamed MP4 Visual is usually identified as MP4V-ES [22:25] c:a is copy, but i need c:v conversion from AVC to MP4 Visual [22:37] Is there a way to filter the -codecs list by what pixel formats they support? [22:38] rather, I want to filter by codecs supporting mono or grey pixel formats [22:41] what exactly does the -bufsize do [22:41] i forgot [22:41] my video is 640x360 with 416k bitrate [22:42] but it's live streaming, in case of network lag id like any kinds of buffers to be sufficient enough [22:42] i set mux_delay to 10 secs [22:43] c_14, I'm using rtpdump to get info off a stream from the same source we were talking about yesterday, it does go backwards periodically [22:44] The timestamp, would it be valuable to include it on my bug report? [22:45] Including it shouldn't hurt in the worst case and in the best case it'll be helpful, so sure. [22:46] i get no explanation on what -bufsize is on ffmpeg documentation [22:53] anybody [22:54] im trying to figure out whether i should use vf scale or -s [22:54] seems like -s as and output adds scale the same way [22:54] i do want as output [22:55] but the whole parameter ordering makes me a headache [23:01] -vf scale and -s do the same thing, the only difference is that with the filter you can choose where in the filtergraph to put it [23:04] ah okay, i don't use any filters here ... [23:07] mpeg4 profile level in nokia specs says Simple Profile 4a ... [23:08] im looking some docs and it says 4 is Advanced, and there's no 4a [23:08] but oh well that stuff is from 2007 i found [23:08] man so much obsolete stuff [23:44] hi all [23:46] I'm new to using ffmpeg. I want to convert an AVC file to PNG image sequences. I get that working, but I also want to crop the frames. [23:47] -vf "crop=in_w:819:0:131" gives me file with a 820px height [23:47] why? [23:47] I need actually 818px height, but -vf "crop=in_w:818:0:131" will give me a file with 816px height... [23:49] the original file is 1080p [00:00] --- Sun Nov 16 2014 From burek021 at gmail.com Sun Nov 16 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Sun, 16 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141115 Message-ID: <20141116010503.1AB0C2AD60D6@apolo.teamnet.rs> [00:11] BadClown: Samples are always welcome, you can just send an email to the user mailing list. [00:38] kierank, thardin: Do your players show a correct ratio for the sample from ticket 4107? http://samples.ffmpeg.org/ffmpeg-bugs/trac/ticket4107/ [00:38] I don't have any player [00:38] but it's probably avci-50 where the SAR is wrong [00:38] you have to "fix" it and make it 4:3 [00:40] Do you mean it is wrong in the file? [00:40] Or in the extradata we add? [00:40] It also shows incorrectly if I remove the code that adds extradata. [00:54] it's wrong in the file at least [00:55] probably also in the extradata [00:55] kierank: I just looked in the mxf header and it sets dar to 16:9 [00:55] yeah but the avc SAR is 3:4 [00:55] Is it likely that this is how it is meant to be read: codec SAR is wrong, container DAR is what should be used? [00:56] no [00:56] it's just a mistake in the encoder [00:56] So (as you already explained to me months ago) the SAR in the H264 stream is really set to 3:4, there is nothing to be fixed? [00:56] either you respect the container or you detect it's avci-50 and "fix" the sar [00:56] or both [00:56] But generally, we prefer the container aspect over the codec aspect, to this should still be played correctly. [00:57] Unfortunately, DAR is useless for avcodec context... [00:57] Or AVStream in this case. [00:57] You mean AVC-50 has always SAR=4:3? [00:58] AVCI-50 [01:11] it shoul [01:11] d [01:52] ffmpeg.git 03Vittorio Giovara 07master:4b39cc1a093c: riff: support ProRes in avi (APCN fourcc) [01:52] ffmpeg.git 03Michael Niedermayer 07master:720a8d2b7518: Merge commit '4b39cc1a093c239412ded522c4a899744e7f2008' [01:54] ffmpeg.git 03Thilo Borgmann 07master:e4cb6abb2f46: bgmc: fix sizeof arguments [01:54] ffmpeg.git 03Michael Niedermayer 07master:ac1735e76dc9: Merge commit 'e4cb6abb2f46910c72178e2f987a0198f0fd10b1' [02:00] ffmpeg.git 03Vittorio Giovara 07master:3a6ddfb8745e: exr: check return value [02:00] ffmpeg.git 03Michael Niedermayer 07master:a5adeff45745: Merge commit '3a6ddfb8745e4b306a5637927fb057f630345e2f' [02:13] ffmpeg.git 03Vittorio Giovara 07master:60e0ee7ca25b: lpc: always initialize ref and err [02:13] ffmpeg.git 03Michael Niedermayer 07master:d4065a9f4755: Merge commit '60e0ee7ca25bd3bea54043b0607efe4cd51acaf3' [02:13] ffmpeg.git 03Michael Niedermayer 07master:85929b9caa90: avcodec/lpc: remove unneeded {} [02:26] ffmpeg.git 03Vittorio Giovara 07master:d16ec1b6db25: atrac3plus: always initialize refwaves [02:26] ffmpeg.git 03Michael Niedermayer 07master:53ab7846eeb6: Merge commit 'd16ec1b6db25bc348b0d4800c9a0c9b7070e3710' [02:44] ffmpeg.git 03Kieran Kunhya 07master:2e1704059ae8: vf_interlace: Add SIMD for lowpass filter [02:44] ffmpeg.git 03Michael Niedermayer 07master:6f373d75e892: Merge commit '2e1704059ae8625beda2ffde847ad22c5ba416dc' [02:51] ffmpeg.git 03Michael Niedermayer 07master:05e0ea60507d: avfilter/vf_tinterlace: Fix output top field first flag for MODE_INTERLACEX2 [02:51] ffmpeg.git 03Mark Reid 07master:933eca91e605: libavformat/mxfdec.c: refactored resolving timecode component [04:18] ffmpeg.git 03Michael Niedermayer 07master:f043965cd514: avfilter/vf_tinterlace: fix linesize vs. width [04:18] ffmpeg.git 03Michael Niedermayer 07master:9d548fce2405: avfilter/tinterlace: split context definition into seperate header so it can be used by future optimizations [04:18] ffmpeg.git 03Michael Niedermayer 07master:18b46ecc93bf: avfilter/tinterlace: Move lowpass_line to a separate function and call it through a function pointer [04:18] ffmpeg.git 03Michael Niedermayer 07master:fb3eb573699e: avfilter/tinterlace: add Support for ff_lowpass_line_avx() & ff_lowpass_line_sse2() [04:18] ffmpeg.git 03Michael Niedermayer 07master:05e4b25e9b0a: avfilter/x86/vf_interlace: rewrite asm [10:03] /usr/bin/ld: /usr/local/lib/libavcodec.a(hevc_cabac.o): relocation R_X86_64_PC32 against symbol `ff_h264_cabac_tables' can not be used when making a shared object; recompile with -fPIC [10:03] Action: funman finds ticket 17143 [10:03] 1713* [11:51] ffmpeg.git 03Marvin Scholz 07master:3a6bb9735053: Icecast: Send 100-continue header if possible [11:51] ffmpeg.git 03Michael Niedermayer 07master:f3c324a0fefd: Merge commit '3a6bb9735053c453f806ceab1d91124648d90aca' [12:04] ffmpeg.git 03Marvin Scholz 07master:8562c1483ba6: Icecast: Send content-type in all cases [12:04] ffmpeg.git 03Michael Niedermayer 07master:42c8db69b686: Merge commit '8562c1483ba647f562e4c1df68a9231274b80e6b' [12:12] Daemon404: Hi, any news about ffmpeg -i output for the Prores sample? Thank you. [12:12] ffmpeg.git 03Mika Raento 07master:b08fd7ea7879: mov.c: fix handling of seek return in read_mfra [12:32] The following sample plays with gray bars with FFplay but black bars with vlc: [12:32] http://streams.videolan.org/issues/12764/Test_mov_MJPEG_pcm_s24be_320x240_3151Kbps_24fps.mov [12:32] Which one is correct? [12:40] whatever quicktime says is correct :P [12:42] Could you test? [12:46] I'm at work and Windows 7/WMP shows black bars, for what it's worth. [12:47] Thank you! [12:51] ffmpeg.git 03Shin-ichi Toyama 07master:12630fa821ea: avcodec/dvdsubdec: New option for obtaining global palette from .IFO file (experimental) [13:13] michaelni: what's the point of pxor m0, m6 ? [13:17] oh m6 is now all 1 [13:22] relaxed: Does wmp play this file? http://streams.videolan.org/issues/12767/Test_mov_h261_pcm_s16le_352x288_251Kbps_15fps.mov [13:22] It shows many error messages with FFmpeg. [13:22] Or QT? [13:29] j-b: Can you provide the sample for videolan ticket 12768 ? [13:31] I would also like to test 12770 [14:40] ffmpeg.git 03Brandon Lees 07master:ffaf2074ebbc: Fix the timeout option not working when connecting to a HTTP url that requires authentication. [14:41] ffmpeg.git 03Michael Niedermayer 07master:deccb4d827d6: avformat/http: simplify chained_options copying [14:44] cehoyos: sorry, I'm no longer at work. [15:39] funny, with a ffmpeg produced file, seeking in ffplay will print [mpeg4 @ 0x7fffd80015e0] Failed to parse extradata [15:45] why does this happen? [15:45] http://pastie.org/private/31z6minb9clhjusookmgrg#13 [15:46] ah --disable-avformat does the trick [16:01] keirank: Are you searching for --disable-network or --disable-all? [16:03] You can do "--enable-filter=drawtext,format,interlace" since several years [16:04] Do you really have mp1 samples? There were difficult to find iirc... [16:06] kierank: Are you searching for --disable-network or --disable-all? [16:07] cehoyos: all broadcast audio is mp1 [16:48] kierank: Is this audio that is sent to TV stations for broadcast? Or do I misunderstand? [16:49] in all the channels I transmit they are MPEG 1 Layer II [16:49] as opposed to MPEG 2 Layer II [16:49] which is for 32khz and below iirc [16:51] But MPEG1 Layer II == "mp2" in FFmpeg terms [16:51] "mp1" is MPEG1 Layer I [16:51] Which you are enabling in your configure line and for which we have one sample iirc [16:51] ah [16:52] And I believe --disable-all will make your configure line simpler [16:52] --disable-everything is more a debug tool to make bisecting faster [16:53] wm4: Does mpv play this sample? http://streams.videolan.org/issues/12767/Test_mov_h261_pcm_s16le_352x288_251Kbps_15fps.mov [16:54] cehoyos: ah possibly I've always been looking for --disable-all [17:00] cehoyos: as "well" as ffplay [17:08] aka, not well [17:08] It works fine here with ffplay, only MPlayer fails... [17:09] fine? [17:09] you don't have the grey bar? [17:09] j-b: Do you have samples for tickets 12768 and 12770? [17:09] Yes, the same grey bar as for every mov file with cropping. [17:09] But MPlayer fails completely [17:09] cehoyos: in ffplay I see black borders which don't seem to belong there, and " qscale has forbidden 0 value" is spammed [17:10] and a grey bar [17:10] cehoyos: 770 is 2GB, so no way to upload it. [17:10] 768, yes, I do. [17:10] and for me 767 is annoying because I have seen no player playing it correctly [17:10] so either the sample is broken, or everyone is based on ffmpeg [17:11] I don't understand: You can use mplayer -vc -ffh261, [17:13] But was is incorrect with libavcodec? [17:21] is that 767 like http://www.cccp-project.net/beta/test_files/h263_adpcm_lolborders.mov ? [17:21] grey bars and mov reminded me of this [17:22] yes, exactly [17:22] lolborders [17:24] seen lots of files like that, mostly h263 [17:27] so how are they cropped? [17:38] ffmpeg.git 03Michael Niedermayer 07master:18fcdc098162: avcodec/mpeg4videodec: forward return code in ff_mpeg4_decode_picture_header() [17:38] ffmpeg.git 03Michael Niedermayer 07master:10411afdff10: avcodec/mpeg4videodec: replace some return -1 by more specific error codes [17:38] ffmpeg.git 03Michael Niedermayer 07master:7d37e45f6bac: avcodec/mpeg4video_parser: fix spurious extradata parse warnings [17:59] JEEB: yes, same lolborder [18:23] wm4 : the video file (not sure if both container and codec) has metadata that says crop x:y , and the player is supposed to do it automatically [18:23] probably i misunderstood the question :P [18:25] [16:27] <+wm4> so how are they cropped? <-- at work we have a special check for them, and crop if need be. [18:25] pure class. [18:25] Daemon404: how? [18:25] i'd guess dump crop metadata, have script pass -vf cropdetect to ffmpeg [18:25] mediainfo provides "original" and "cropped" res for it [18:25] abs(orig-cropped)==1 [18:27] s/cropdetect/crop [18:27] interesting [18:27] derp [18:27] is it in the mp4 format? [18:27] mov or mp4 or some bastardization of it yes [18:38] if the codec has that meta, I'd expect it to be done at the decoder level [18:38] Daemon404: any news about the Prores sample? [18:38] no [18:39] rcombs, ffmpeg makes the decision to not crop it [18:39] cause it is yv12 [18:39] and this returning an extra line is deemed better than one too few [18:39] (since it was originally an odd number of lines) [18:39] rcombs: that was my question. Is the codec or the demux aware? [18:39] what's the problem with cropping the right/bottom borders [18:39] j-b: dunno, I'm just expressing expectations [18:39] wm4, you lose one legitimate line [18:39] technically. [18:40] lavc can return odd width/height with jpeg and 4:2:0 [18:40] well i dunno why h263 or avc doesnt then [18:41] is it in the mp4 format? <- I think it's another resolution noted in the container or so [18:42] i think it might be noted in the bitstream [18:42] but dont quote me. [18:42] I remember looking into it once [18:42] basically it's a flag in mov to scale the picture to that res [18:42] but yes, it's lavc doing it [18:42] since it gets the data as the resolution and starts decoding the picture into it [18:42] which is why you get grey [18:43] boxdumper will probably tell you moar [18:43] because what QT does is take the actual picture and scale it to that size [18:44] qt crops the line off [18:44] lavc will take the resolution as a resolution, and then decode into it a smaller picture [18:44] not scale [18:44] lemme find the comparison pic for my file [18:45] right now im trying to find out something for mp4 that i *havent* solved yet [18:45] i.e. how the shit do i detect the ios6 "slo-mo" shit in mp4/mov [18:45] QT can [18:45] boxdumper had nothing stand out to me [18:45] http://i4.minus.com/iyF0ALNoDz61c.png [18:46] JEEB, i usually only see 1px [18:46] so it's cropping and scaling [18:46] it would seem [18:46] because the AR is clearly different [18:46] who needs NLE project files? [18:46] just use ALL OF THE MOV FEATURES [18:47] :D [18:47] why is mov so terrible [18:47] the idea of atoms/boxes itself is pretty sane [18:47] the spec is insane though [18:47] well specs. [18:47] and Way Too Much STuff [18:48] still. ill take boxes/atoms over EBML any day. [18:48] ebml isn't much different after all... [18:48] Yeah, the Way Too Much Stuff is the real problem [18:48] no-one read it properly [18:48] vague-as-hell specs dont help [18:49] and at least matroska has a proper concept of interleaving [18:49] and you need to read X,Y,Z specs to implement shit properly [18:49] specs that are fragmented over like 100 different specs [18:49] all of which cost like 200chf each [18:49] yup [18:49] well at least there are specs [18:49] i guess [18:49] yeh [18:49] Action: Daemon404 stares at mkv [18:50] maybe I should reopen that one document I had on "marumoska" [18:50] [qcelp @ 0x7f8040c5c660] Frame #1, IFQ: bitrate cannot be determined. [18:50] JEEB: I suspect for your sample the mov container sets the display size (as done for many other samples), the sample in question does not contain such data afaict. [18:50] k [18:50] you can check with boxdumper [18:50] comes from the L-SMASH project [18:50] https://github.com/l-smash/l-smash [18:51] a convenient way of getting a text dump of an mp4/mov/whatever file [18:51] Is your sample publically available? [18:51] he linked it above iirc [18:51] is that 767 like http://www.cccp-project.net/beta/test_files/h263_adpcm_lolborders.mov ? [18:51] grey bars and mov reminded me of this [18:51] do we have a slowmo sample ? [18:51] Daemon404: Can you copy the line? [18:52] i saw that on the teeeveee but never saw a sample of it... [18:52] Compn: The ticket is only missing the necessary info, sample is available. [18:52] cehoyos : which ticket is for slowmo sample ? [18:54] JEEB: The tkhd atom contains all necessary information, no problems here [18:54] the h261 sample may not contain the nececssary info, did anybody test with QT? [18:54] k [18:54] Compn: Search for a "new" important ticket (a regression) [18:54] [17:51] <@cehoyos> Daemon404: Can you copy the line? <-- ? [18:54] i dont follow [18:54] Compn, yes. anyone with an iphone can also make one [18:54] I missed the lolborders sample link. [18:54] it's an OS feature [18:55] Not necessary, we already have a sample for this regression. [18:57] the mediainfo informations are not enough for crop [18:57] cehoyos : http://www.cccp-project.net/beta/test_files/h263_adpcm_lolborders.mov [18:57] I have it three times now, thank you all! [18:57] lol [18:58] :) [18:58] ffmpeg.git 03Arwa Arif 07master:a2f05d33373e: lavfi : change xBR filter to LGPL [18:59] Width : 320 pixels [18:59] Original width : 352 pixels [18:59] Height : 230 pixels [18:59] Original height : 288 pixels [18:59] that looks like enough info to crop, j-b [18:59] (from JEEB's sample) [18:59] it's always bottom and right. [19:03] Daemon404: Any news about the ProRes sample? The patch looks very bad to me. [19:03] you asked about 3 times today [19:03] i replied already [19:04] [17:38] <@cehoyos> Daemon404: any news about the Prores sample? [19:04] [17:38] <@Daemon404> no [19:04] So where can I find the information? [19:04] what? [19:04] I thought two hours are sufficient to test one sample, sorry! [19:04] ... [19:04] the problem is not testing [19:04] it's GETTING it [19:04] it's sitting on a coworker's pc [19:04] I thought you know somebody who has it? [19:04] because the user deletec the original [19:04] yes and people tend to be away [19:04] and busy [19:04] on weekends [19:04] GO FIGURE [19:04] Actually who you gave it to, iiurc. [19:05] Daemon404: sure, but try the sample from the ticket [19:05] it says 180x [19:05] but if you crop 180x you loose half the picture [19:05] j-b: Does it work with QT? [19:05] cehoyos, what? [19:06] What what? [19:06] cehoyos: not for me [19:06] So the sample was either remuxed or is supposed to by played as vlc does currently? [19:06] [18:04] <@cehoyos> Actually who you gave it to, iiurc. <-- i dont follow [19:06] the lolborder one is correct [19:06] Or do you mean it does not play at all? [19:07] the one reprocessed by lavf is butchered, of course [19:07] Of course, vlc reads the tkhd atom. [19:07] it's lavf [19:07] No? [19:08] With --demux=ffmpeg, the sample plays incorrectly, with default demuxer it work fine. [19:08] Or do I misunderstand? [19:09] for me, in VLC all are wrong [19:10] using l-smash, lavf, or our demuxer [19:10] If I play h263_adpcm_lolborders.mov it works fine with vlc: The borders are correctly cropped. [19:10] cehoyos: wut? [19:10] cehoyos: indeed, works in 2.1 [19:11] j-b, since when does vlc have l-smash support [19:11] Why shouldn't it: This was always a feature of vlc missing in (for example) ffplay [19:11] Daemon404: :D [19:11] Reading tkhd display resolution. [19:11] Daemon404: I did not say anything [19:12] j-b, i am an l-smash API user as well, FWIW [19:12] ;) [19:17] cehoyos: thanks a lot. It's a VLC 2.2.0 regression. [19:18] Can you add any information about 12768? Is it reproducible with FFmpeg? [19:18] not before monday [19:18] and yes, it was tested with FFmpeg. [19:18] So it crashes FFmpeg? [19:19] yessir [19:20] From your pov, how is FFmpeg supposed to know about such issues? [19:22] no idea. [19:22] when someone reports the bug [19:22] lol ;D [19:22] Ok, I read the bug report and I requested the sample: What else should I do? [19:22] ffmpeg.git 03Luca Barbato 07master:74d7db586a2e: dv: Drop a spurious check [19:22] ffmpeg.git 03Michael Niedermayer 07master:89e705cd5c94: Merge commit '74d7db586a2e9aeb107e357739c7e4dde0b6991c' [19:22] wait for me to upload it? [19:22] and test again HEAD [19:22] cehoyos : wait monday and j-b will upload it with more info [19:22] What kind of into? [19:23] cehoyos: there are other bug reports, lately reported that are still reproduced on HEAD [19:23] gdb i assume [19:23] Could you point me to one you mean? [19:25] michaelni: why is that message still in copyright boilerplate? (xbr) [19:26] arwa: just saw your messages; you were looking for guidance for mp filters? [19:27] arwa: Here are two commits from ubitux which should help with porting uspp and fpss: [19:27] http://git.videolan.org/?p=ffmpeg.git;a=commitdiff;h=a2c547ff [19:27] yeah [19:27] http://git.videolan.org/?p=ffmpeg.git;a=commitdiff;h=852f74bd [19:27] fspp [19:28] so we're really going to port uspp and fspp? [19:28] ubitux, well people refuse to delete the mplayer shim until theyre ported [19:28] so clearly someone thinks theyre Very Important [19:28] i see 2 problems for each of them [19:29] fspp is basically a simpler version of spp which is faster, but i wonder if it's any relevant nowadays [19:29] and uspp has a libavcodec dependency [19:30] dont other things in lavfi have that too [19:30] which is technically not really a problem but it needs to be considered [19:30] yeah sure [19:30] i think uspp is the simpler to port though [19:31] ubitux, because it was in the patch and maybe someone reading it wonders why its LGPL when the code its based on was GPL [19:31] I think Reimar commented on fspp several times already. [19:31] iam perfectly fine with remocing it [19:31] michaelni: well it's in the commit message... [19:31] uspp has been requested a few times. [19:31] michaelni: yeah i think that would be better [19:32] arwa: I believe the eq and eq2 filters are more difficult to port (but I may be wrong), a patch exists though. [19:32] ffmpeg.git 03Michael Niedermayer 07master:2fa6d21124bd: on2avc: Fix out of array access [19:32] ffmpeg.git 03Michael Niedermayer 07master:dcb10ef9bff3: Merge commit '2fa6d21124bd2fc0b186290f5313179263bfcfb7' [19:32] no they're not "hard" actually [19:32] the problem is how consistently we can integrate them [19:32] saste: ping [19:33] J_Darnley: ping too, how is eq/eq2 going? [19:33] anyway, i can give guidance for the uspp port [19:33] ubitux, pong [19:33] saste: remember what was decided for eq/eq2? [19:34] like how to merge them into one feature wise etc? [19:35] btw, unrelated but we need something in the documentation to compare all the pp filters [19:36] because no user is ever going to understand why we have pp, spp, fspp, uspp, ... [19:37] ubitux, yes [19:38] about eq/eq2, I remember some comments from paul [19:38] >why anyone cares about mplayer code in ffmpeg anyhow [19:38] also, you mentioned that J_Darnley was working on that [19:38] between the missing filters, I agree eq/eq2 seems the most useful [19:39] So is my task is to port eq/eq2 filters? [19:39] arwa_, when do you plan to start with your project? OPW program will officially start in December [19:40] arwa_, we're discussing what filters we should port [19:40] can we start a coop edit document of some sort to summarize the state of each remaining filter to port? [19:41] I have exams in a week, I will not be able to do much work. But at least I can get some idea. [19:42] so here is the thread for eq/eq2: http://lists.ffmpeg.org/pipermail/ffmpeg-devel/2014-September/162930.html [19:42] ubitux, i was thinking about the same thing [19:44] current state for eq/eq2 @ http://lists.ffmpeg.org/pipermail/ffmpeg-devel/2014-September/163005.html [19:45] kierank: about mp=ilpack, so is swscale able to do what you want now? [19:45] (and if so... how?) [19:45] not 100% sure but I'll find out in the next few weeks [19:45] it is on my todo list to write an interlaced upsampler [19:45] ok so ilpack in stand-by [19:47] softpulldown in standby? http://ffmpeg.org/pipermail/ffmpeg-devel/2013-May/143262.html [19:48] so uspp is a requested pp filter, using lavc/snow and no asm; could be ported relatively easy but needs some serious cleanup [19:49] pp7: no comment. [19:49] state of the art in 2001 [19:49] thread about pp7 & fspp http://lists.ffmpeg.org/pipermail/ffmpeg-devel/2013-June/144740.html [19:50] and... done? [19:51] wm4 : should vaporsynth be ported to ffmpeg then ? :) [20:04] ubitux, done with my tests (for now atleast). Are you free now ? [20:04] akira4: sure [20:04] right. So where do we start? [20:05] mmh first, since i have no idea how the dvd libs work, maybe you should start figuring that out [20:05] akira4: cehoyos told me that libdvdnav is probably what we should look for instead of libdvdread [20:06] oh. Also what about the dvdsubdec.c in libavcodec? [20:06] this is the dvd subtitles decoder, but we're interesting in the demuxing part [20:06] like, extracting the undecoded "chunks" out of dvds [20:07] (is freenode drunk again or...?) [20:07] hmm. The patch you told me look into uses libdvdnav I think. [20:07] ah, well then you should probably use that [20:08] i don't know the relationship between dvdnav and dvdread, but dvdnav probably depends on dvdread [20:08] hmm. Do I need to externally download them? [20:09] you could get the sources from videolan if you want to look at them [20:09] but anyway, i guess the first thing would be to port the experimental dvdnav protocol reader as a demuxer [20:09] do you have dvd images to test? [20:10] nope. couldn't find any :( [20:10] with subtitles I mean [20:11] mmh, i could give you various copyrighted crap but that might not be wise for now [20:11] let me check if i find something free [20:11] okay. [20:14] ffmpeg.git 03Stefano Sabatini 07master:aea7616d39bb: lavfi/xbr: remove relicensing notice from copyright header [20:22] seems i can't find anything so far [20:22] damn. [20:23] ah, found this: https://www.vhsretrostyle.com/creative-commons.html [20:23] akira4: "Nasty.Old.People_2009.dvd.iso [3.7 GB]" [20:23] mmh subtitles seems to be external [20:24] still, it can be useful for now for dvd support [20:24] okay. I'll try testing with it. [20:26] umm ubitux , my university's proxy won't let me download the iso :-/ [20:27] I'll try downloading from somewhere else I guess. [20:30] akira4: you can't borrow a real dvd from someone? [20:31] I can try. [20:40] Action: wm4 still has a dvd image from ubitux [20:40] ah? :D [20:53] grrr closed wrong window [20:53] ohwell [21:07] ffmpeg.git 03Cl?ment BSsch 07master:08bb6f919c0a: avfilter/xbr: do not pass unchanging r2y to macros [21:07] ffmpeg.git 03Cl?ment BSsch 07master:086487b633de: avfilter/xbr: localize some filtering variables [21:07] ffmpeg.git 03Cl?ment BSsch 07master:9f9c74177138: avfilter/xbr: avoid unecessary macro redirections [21:07] ffmpeg.git 03Cl?ment BSsch 07master:55f05ac0f1bf: avfilter/xbr: use different macro names for each dimension [21:07] ffmpeg.git 03Cl?ment BSsch 07master:e07048404043: avfilter/xbr: simplify width overread checks [21:07] ffmpeg.git 03Cl?ment BSsch 07master:6bf9786a9bd0: avfilter/xbr: mark source pointers as const [21:07] ffmpeg.git 03Cl?ment BSsch 07master:18e4bf4f54ca: avfilter/xbr: refactor the 21 pixels definition into a macro [21:07] ffmpeg.git 03Cl?ment BSsch 07master:7e91f77547b8: avfilter/xbr: refactor src/dst pointers definitions into a macro [21:07] ffmpeg.git 03Cl?ment BSsch 07master:a3c3ee697398: avfilter/xbr: misc cleanup in FILT[234] macros [21:07] ffmpeg.git 03Cl?ment BSsch 07master:fda209b74179: avfilter/xbr: simplify left/up conditions [21:07] ffmpeg.git 03Cl?ment BSsch 07master:a99004a926b3: avfilter/xbr: misc style fixes [21:07] ffmpeg.git 03Cl?ment BSsch 07master:d1529273fb2f: avfilter/xbr: make xbr[234]x a bit more consistent [21:07] ffmpeg.git 03Cl?ment BSsch 07master:9a796f7f18e4: avfilter/xbr: consistent copyright header [21:08] spam [21:08] yeah sorry [21:08] damn cosmetics [21:08] still haven't figured a sane way of refactoring the logic of FILT[234] but should be good enough for now [21:09] maybe i should make a relevant change now [21:23] ffmpeg.git 03Cl?ment BSsch 07master:c4fb79a3db1a: avfilter/xbr: remove FATE test entry from @todo [21:23] ffmpeg.git 03Cl?ment BSsch 07master:be96201e5bf1: avfilter/xbr: use function pointers for xbr[234]x [21:23] ffmpeg.git 03Cl?ment BSsch 07master:454b71428368: avfilter/xbr: add video and filtering flags to options [22:13] ffmpeg.git 03Michael Niedermayer 07master:bd7d8604bd27: avcodec/nellymoserenc: update comment to match 5c805d69a49a1f32a7a8a1b16fb3d631d85ca56d [22:16] ffmpeg.git 03Cl?ment BSsch 07master:8bc1553cdb59: avfilter/xbr: add slice threading [22:16] finally a "useful" commit [22:16] anyway, back to more relevant stuff [22:22] oh, not yet. got an idea for a relevant refactoring.. [22:37] ffmpeg.git 03Cl?ment BSsch 07master:bca3c2cfc02c: avfilter/xbr: refactor xbr[234]x into a single function [22:46] ffmpeg.git 03Cl?ment BSsch 07master:7eece0693424: avfilter/xbr: clarify default "interpolated" pixels assignments [22:48] ffmpeg.git 03Cl?ment BSsch 07master:77204f7366d4: avfilter/xbr: fix style in FILT4() calls [22:53] ubitux: the progress of eq hasn't gone any futher than in the mailing list message you linked to. [22:53] J_Darnley: i think it would make sense to have it in "hue" [22:53] (and that filter probably needs to be renamed) [22:54] (or aliased) [22:54] typically, to "levels" or whatever [22:54] now it's always a bit tricky because of the eval level [22:54] (to enable optims or not) [23:27] about the [FFmpeg-devel] [PATCH 4/4] web/style.less: Separate out .table-bordered from .table [23:29] I don't really care [23:29] it looks good with or without the border [00:00] --- Sun Nov 16 2014 From burek021 at gmail.com Mon Nov 17 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Mon, 17 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141116 Message-ID: <20141117010502.8B60118A0146@apolo.teamnet.rs> [01:04] https://www.youtube.com/watch?v=NAsM30MAHLg&index=1&list=PL0INsTTU1k2UYO9Mck-i5HNqGNW5AeEwq [01:04] fun stuff fourier related [01:06] ffmpeg.git 03Martin Storsj? 07master:c00365b46d46: aarch64: Make the function pointer tables position independent [01:06] ffmpeg.git 03Michael Niedermayer 07master:f3cba01ccefe: Merge commit 'c00365b46d464ce47716315c1801818d811bdb9a' [01:11] ffmpeg.git 03Michael Niedermayer 07master:08ee02deca81: avfilter/vf_tinterlace: remove unused variable [01:13] ubitux: I still believe dvdnav should be neither a protocol nor a demuxer but an in-device. [01:13] yeah sure whatever, it's the same thing [01:13] ffmpeg.git 03Lukasz Marek 07master:c727006616ef: lavc/options_table: set min to -1 for timecode_frame_start [01:14] ffmpeg.git 03Lukasz Marek 07master:c544ffd2aeab: lavc/options_table: add pixel_format and video_size options [01:14] ffmpeg.git 03Lukasz Marek 07master:4a30277a59ef: lavc/options: set timecode_frame_start to -1 as option default [01:14] ffmpeg.git 03Lukasz Marek 07master:01974a58df40: lavc/options: initialize pkt_timebase [01:14] ffmpeg.git 03Lukasz Marek 07master:eec693328a09: lavu/opt: introduce av_opt_is_set_to_default() [01:14] ffmpeg.git 03Lukasz Marek 07master:bee5844ddd4e: lavu/opt: introduce av_opt_serialize() [01:14] ffmpeg.git 03Lukasz Marek 07master:6690d4c3f53b: lavf/ffm: store/restore private codec context [01:14] ffmpeg.git 03Lukasz Marek 07master:2657f00d3f99: ffmpeg_opt: use codec private context in ffserver streams [01:14] ffmpeg.git 03Lukasz Marek 07master:a38e06c1aaac: ffserver_config: handle codec private options [01:14] ffmpeg.git 03Lukasz Marek 07master:745730c9c208: lavf/ffm: use AVOption API to store/restore stream properties [01:20] ubitux: nice find [01:50] Action: ubitux wonders what's the point of that threshold filter given that they provide an equivalent with lut filter... [02:03] ffmpeg.git 03Lukasz Marek 07master:ec012837527c: lavf/ffmdec: reident after last commit [03:03] ffmpeg.git 03Michael Niedermayer 07master:ce80f9fee971: avformat/segment: export inner muxer timebase [04:39] ffmpeg.git 03Michael Niedermayer 07master:530eb6acf8ee: avformat/hlsenc: Free context after hls_append_segment [17:51] ffmpeg.git 03Michael Niedermayer 07master:1de786777e0f: avcodec/dvdsubdec: Check all fseek()s return codes [18:19] ffmpeg.git 03Michael Niedermayer 07master:374c907fb35f: avcodec/vorbis_parser: Move vp check [19:10] ffmpeg.git 03Michael Niedermayer 07master:62eca2f827d4: avdevice/xcbgrab: Fix/remove unneeded NULL checks [19:34] do we have any filter with slice threading and computation in each of them? [19:34] like, typically each slice returns a score, and each of these needs to be summed at the end [19:35] i'm wondering about allocating a tab of the number of jobs but i wonder if there is another way [19:35] oh mmh there is a ret thing.. [19:37] ffmpeg.git 03Michael Niedermayer 07master:51ddaf65496b: avformat/mpeg: fix memleak of sub_name on error [20:26] btw, stupid question but... what's exactly the point of the arg in threading execute function, given that filter have access to the private context anyway? [20:28] everything in the thread data arg could be in the private context [20:28] since AVFilterContext is passed [22:33] something happened with git server? [22:35] nvm, my dns is not working for some reason [22:48] strange, i'm having dns problems too... [23:26] michaelni: didn't libswscale have floyd steinberg dithering at some point? [23:28] some cases support error diffusion dither [23:37] found it, thanks [23:38] ramiro: x264 has FS [23:41] kierank: grep floyd -ir gives nothing. but i see Sierra-2-4A error diffusion on depth.c [23:41] ah ok [00:00] --- Mon Nov 17 2014 From burek021 at gmail.com Mon Nov 17 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Mon, 17 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141116 Message-ID: <20141117010501.83B6E18A00C0@apolo.teamnet.rs> [00:05] that's encoder limitation [00:06] it won't give you uneven output [00:06] probably goes deep into how computers work imo, can't you see everything is even, 64, 128, 1024, 4096 ... etc [00:07] but im not the expert, might be another reason [00:10] It will give me 820 and 816px, but not 818? strange [00:11] Might have to be divisible by 4 [00:12] hm okay, that might be true [00:17] hi all, i have problem with ffmpeg compile on opensuse 13.2 x64. Here is some log: http://pastebin.com/ZfMx3Ai2 [01:06] gi again guys [01:06] this doesn't work for me, says Unable to Connect [01:07] ffmpeg -loglevel debug -re -i rtmp://xxx.xxx.xx:xxxxx/xxxxxxxxxxx/_definst_/ -c:v mpeg4 -vf scale=640:360 -r 29.97 -g 60 -pix_fmt yuv420p -minrate 368k -maxrate 432k -bufsize 576k -c:a copy -f rtsp -muxdelay 3 rtsp://192.168.1.101:8554/ajshow [01:13] trying to convert an wma to mp3 - what does it mean "Queue input is backward in time"? I am getting skipping in the resulting mp3 [01:16] also: "Application provided invalid, non monotonically increasing dts to muxer in stream 0:" [01:17] does ffmpeg actually create a rtsp server or it's just feeding it ? [01:26] oh i just had a malformed URL, i had to add in the title as well that wasn't part of the url in the tracing but it had to be specified in rtmpdump separately [01:26] works with VLC now too [01:28] althought I have the same problem with a disruption after around 59 seconds , i think that might be the real player's issue on the nokia itself [01:28] not 20 secs as i previously said [02:41] I have a problem that I think ffmpeg hangs while reading /dev/video0. After 4 or 5 seconds of running it stops responding to keystrokes and the status screen stops updating [02:44] frame= 30 fps=7.5 q=0.0 size= 56kB time=00:00:01.00 bitrate= 458.8kbits/s dup=9 drop=0 [02:46] frame= 107 fps=3.8 q=0.0 size= 128kB time=00:00:03.56 bitrate= 294.0kbits/s dup=76 drop=0 [02:46] then no updates [02:52] http://pastie.org/private/u5l8umtgdfmavp0h3f0nw [02:53] What's the specs of the machine you're doing this on? [02:54] Core i5, 16G ram [02:55] 17:55:13 up 8:58, 5 users, load average: 1.59, 1.33, 1.24 [02:55] Something is weird, then, because doing under 4 frames per second on an i5 is slow... And probably causing it. [02:56] it slows down [02:56] After 30 frames, it's going at 7.5... [02:57] then it keeps dropping [03:00] strace shows it still running [03:00] in this loop [03:00] poll([{fd=4, events=POLLOUT}], 1, 100) = 1 ([{fd=4, revents=POLLOUT}]) [03:00] sendto(4, "fm\0\0\0\5\7\360/x\36`\0\314,H\350\331\263..85*\254\206\357wy\356w\271"..., 4096, MSG_NOSIGNAL, NULL, 0) = 4096 [03:00] poll([{fd=4, events=POLLOUT}], 1, 100) = 1 ([{fd=4, revents=POLLOUT}]) [03:00] sendto(4, "\r\n", 2, MSG_NOSIGNAL, NULL, 0) = 2 [03:00] but the time almost always shows 3.56 seconds [03:01] Yep frame 107 & 3.56 seconds [03:01] it's not frozen I just typed q and then about 5 seconds later it quite [03:05] strace output shows this line 7000 times in succession [03:05] ioctl(3, VIDIOC_DQBUF, 0x7ffff8a3b070) = -1 EAGAIN (Resource temporarily unavailable) [03:11] red herring. That was an older strace [03:12] strace output shows this line 7000 times in succession [03:12] so now I dont get that issue [03:15] frozen at frame= 101 fps=4.4 q=0.0 size= 112kB time=00:00:03.36 bitrate= 272.5kbits/s dup=74 drop=0 [03:29] hey folks. Having a slight problem concating mp4 files with the same codecs. There's a very slight delay between each file and I'd like it to be seamless [03:29] just using a very simple command structure of: ffmpeg -f concat -i list.txt -c copy output.mp4 [03:32] why does -vcodec libx264 cause "Unknown encoder" [03:32] Did you compile with x264 support? [03:32] ah because ffmpeg -encoders shows D.V.LS h264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 [03:34] Not specifically I have ./configure --enable-libvpx --enable-libvorbis [03:34] Then you don't. [03:34] So I'm curious it I can combine multiple mp4's to have no audible gaps between them [03:36] Ok before I go build something I don't need I"m just trying to stream my local webcam /dev/video0 to a local web page using ffserver [03:36] I found an example that The following command will stream media from your the system's webcam and microphone. Note that the High profile and level 4.0 is used. Also tested and working is Medium profile with level 3.1. [03:36] ffmpeg -f video4linux2 -s 320x240 -r 16 -i /dev/video0 -f oss -i /dev/dsp \ [03:36] -g 52 -strict experimental -acodec aac -ab 56k -vcodec libx264 -vb 452k \ [03:36] -profile:v high -level 40 -r 16 \ [03:36] -f flv "rtmp://example.com:8081/publish/first?secret" [03:37] so I thought I'd try a different encoding to see if that helps [03:40] so 264lib comes from git clone git://git.videolan.org/x264.git [03:49] jewt: sounds like it should work, have you tried? [10:22] I've been trying all day to get ffmpeg and ffserver to stream the video from my webcam to html5 is this remotely possible? [14:20] I encode FLV videos with x264 and FFmpeg built-in AAC encoder. Sometimes sound are not in the position the video in. [14:20] It's because this AAC encoder is experimental? [14:32] https://trac.ffmpeg.org/wiki/EncodingForStreamingSites [14:32] Fix -video_size $(xwininfo -root | grep 'geometry' | awk '{print $2;}'). [14:32] Using grep and awk together is stupid. [14:40] fixed [16:11] llogan: I have a one liner to x11grab windows if you'd like to add it to the wiki: x=($(xwininfo|awk '/Wid/||/Hei/||/Cor/{gsub(/+/,",");print $2}') );ffmpeg -video_size "${x[0]}"x"${x[1]}" -framerate 60 -f x11grab -i :0.0"${x[2]/,/+}" -t 30 -c:v rawvideo -pix_fmt yuv420p output.mkv [16:36] relaxed: http://trac.ffmpeg.org/ticket/2563 [17:01] ubitux: is -x11select covered by xcbgrab's -x and -y ? [17:02] it's not fixed yet :P [17:02] just an idea.. [19:19] Hello. I'm trying to concatenate two mp4s using the concat demuxer. Things aren't quite working because I think the two mp4s are encoded a bit differently. I'm willing to re-encode one of the pieces to make them match, but what exactly needs to be set in order for concatenation to work? Thanks in advance. [19:24] afaik, frame rate, codec, (profile where applicable), sar, dar, resolution, audio codec, audio channels, audio sample rate [19:25] Some of those might be extraneous, but making them all match won't hurt. [19:25] hrmm, okay [19:25] does codec include codec_time_base? [19:26] No, frame rate includes codec_time_base (according to my subjective little organization) [19:26] Okay, but regardless both codec_time_base and time_base should match for it to work? [19:26] I think so, yes. [19:29] And what are sar and dar? Sorry I'm somewhat new to all this. [19:30] Eh, Source Aspect Ratio and Display Aspect Ratio [19:30] (Not sure if it's source, it's something with an s though.) [19:30] Basically the DAR is 16:9 or 4:3 or whatever. [19:30] ahh, looks like sample [19:30] And the SAR is either 1 or something fucked up. [19:30] from the ffprobe output [19:30] 0:1 [19:31] I think? Anyway, looks as if for my two mp4s everything matches except time_base (based on the ffprobe -show_streams output). [19:31] is there a way to change just the time_base without changing the codec_time_base? [19:33] Hmm, try `ffmpeg -i file -copytb 0 -c copy outfile' [19:37] hmm [19:37] I still seem to get the same time base [19:37] the first video is at 1/30000 [19:37] the second at 1/90000 [19:38] with -copytb 1 ? [19:38] oh, with -copytb 0 [19:38] Nah, I meant "Can you retry with -copytb 1?" [19:38] But was too lazy to write the whole thing. [19:38] oh, I see. I'll try [19:38] all good [19:41] looks like it's the same. [19:41] Damn, was worth a shot though. [19:43] yeah, thanks. [19:43] At any rate, it seems like if I can just figure out how to change the time_base without changing the codec_time_base then I'll likely be in business. [19:43] Try reencoding with -r 'fps' [19:43] everything else seems to match [19:43] okay [19:43] ffmpeg might pick something suitable [19:45] hrmm, still has the same time base [19:47] Try reencoding with -copytb 0 [19:47] okay [19:50] still the same [19:50] :( [19:50] Do the files have audio streams? [19:50] yes, they do [19:51] they're both stream copied [19:51] so they concatenate fine [19:51] the result of the concatenation [19:51] is that the first piece of the video plays fine [19:51] then the second piece doesn't have video, but the audio continues just fine [19:52] the first piece was re-encoded from one mp4, and it has the time_base 1/30000. I used the libx264 codec. [19:52] The second piece was just stream copied. [19:52] Try the reencoding with -copytb 0 -an [19:52] and it has a time_base of 1/90000. [19:52] okay [19:52] If that doesn't work, I'll need to take a look at the ffprobe outputs [19:54] okay, sure. [19:54] Same as before. This time the one I used -an on doesn't have an audio stream. [19:54] but time_base is the same [19:54] Where should I paste ffprobe outputs? [19:55] Some sort of pastebin [19:55] sprunge.us ix.io pastie.org pastebin.com etc [19:55] kk. [19:56] Why does my video sometimes blink? [19:57] http://ovrload.ru/t/40125_test.flv [19:58] But it normalizes when there is some move on screen. [20:02] cs_14: http://pastebin.com/q2HjbAU9 [20:02] Those are the ffprobe outputs with -show_streams for the two pieces. [20:07] also looks like the bitrate on the first one is lower. [20:08] Bitrate shouldn't matter. [20:08] It's because of zerolatency tune, I think. [20:09] Changed it to film, all looks good. [20:09] skwaap: try muxing to mkv with no audio: ffmpeg -i mp4 -c:v copy -an out.mkv [20:09] Just want to overwrite that pesky value. [20:11] when I do that to either one, it ends up with time_base 1/1000 [20:11] so that might work [20:14] You'll have to concat the audio and mux it back in later though. [20:15] Unless it also works without -an [20:15] that shouldn't be too bad. I'm still having issues concatenating, however. Let me play with this a bit, though. [20:15] well, it gives the same time_base without -an [20:16] but when I concatenate I get the same artifacting in the video [20:18] wait, this is with -an. No audio. I'll take a look at it a bit more, not sure why it's not concatenating right. I'm using ffmpeg -f concat -i inputs.txt output.mp4 The inputs.txt file just has the two mkv files. I'll take another look at the ffprobe outputs on the mkv files. [20:20] The only differences I see are has_b_frames 1 versus has_b_frames 2. [20:20] and start time, etc. [20:21] I get this message when I concatenate: Non-monotonous DTS in output stream 0:0; previous: 0, current: 0; changing to 1. This may result in incorrect timestamps in the output file. [20:22] That _should_ be fine normally. [20:24] hrmmm. weird. [20:24] well, I'll play with it some more on my own. At least now I can get time_base to match with mkv. Might be something stupid I'm doing on my end. Thanks a lot for the help! [20:28] c_14: why does screen blink on still images using zerolatency tune? [20:29] On x264, ofc. [20:30] I don't know the exact reason, but the zerolatency tune can reduce quality. [20:33] It basically kills the lookahead. [20:33] Why killing the lookahead causes those distortions is beyond me. [20:53] c_14: what should I use for fast encoding and good quality? [20:53] In games. [20:54] Do you notice any latency without the zerolatency tune? [20:54] No. [20:55] And what's the default tune in x264? [20:55] none [20:55] the defaults should be generally fine [21:09] without lookahead you cannot average bitrate, so you spend too much bitrate on easy scenes and too little on hard ones [21:16] It works. [21:16] Thanks all. [21:17] Now I need a streaming service. [21:17] Twitch lags a lot. [21:17] hitbox too. [21:35] ffmpeg -i "$FILE" -acodec libmp3lame "$FILE".mp3 >>log.txt [21:35] why this does not work ? [21:36] the log part? [21:36] use &> instead of > [21:36] instead of the output on the screen ... I just want to save it to the log file [21:36] yep [21:36] so you get both stdout and stderr [21:36] ffmpeg in general outputs most of its stuff into stderr so that you can use stdout for piping [21:36] common thing with command line video encoding things [21:37] hello! [21:37] JEEB: so, I can not do that ? [21:37] quick question: is there any way in FFmpeg to change pitch without changing speed? [21:37] what? [21:37] I just told you to also redirect stderr [21:37] ie. the pitch filter in SoX [21:37] I haven't found that [21:37] just stderr is 2> and both stdout and stderr together are &> [21:37] :P [21:38] JEEB: I'm not very good at bash,.... I just started yesterday [21:38] link me please for some wiki thing [21:38] > redirects stdout (the default, no number or anything) [21:38] 2> redirects stderr [21:38] &> redirects both stdout and stderr [21:39] and if you want it to append instead of clear the file, then you add another > just like you've done in your example :P [21:39] hmmm let me try [21:41] haha [21:41] JEEB: great ! [21:41] thanks man =) [21:44] LanDi: because ffmpeg writes all into stderr. [21:47] I got it [21:47] I see, yes. Came here 3 minutes ago. [21:50] How do I stream the URL at http://pastebin.com/RndJCf2T to a local file _with_ subtitles? ffmpeg says there are 3 streams: audio, video and unknown. I suspect the "unknown" stream is the subtitles. [22:04] >This paste has been removed! [22:31] exit [23:08] Does ffmpeg support interleaving for mp4? Can't find any way to achieve this except using mp4box [23:21] brainopia, you mean faststart ? [23:25] faststart will only move moov atom upfront [23:26] Chocola4: but my problem is related to a big size of moov atom (it slows down start of streaming) [23:26] I was looking for solution to separate moov or shrink it somehow [23:27] I've tried fragmentation but chrome won't play fragmented mp4 [23:27] I've heard about interleaving but it seems I was mistaken it has nothing to do with moov size [23:28] so the last thing I can do is to shrink moov size somehow [23:31] the CTTS atom in my files are pretty big, is there a trade-off to get rid of them? [23:53] another question: cfr + maxrate are supported with each other? [23:55] you mean crf? [23:55] if so, yes. but you also need to provide bufsize [23:55] because maxrate is always calculated within a bufsize [23:56] match the bufsize between your client/player and the encoder, and you shall not need to buffer ever again :) [23:56] (unless x264 screams at you hard and tells you it was forced to break VBV, although by now that is very, very rare) [23:57] and if your player or whatever sets or lets you set the amount of buffer in seconds, then you use maxrate*. For example, "two seconds" worth of buffer with maxrate of 2000 is 4000 [00:00] --- Mon Nov 17 2014 From burek021 at gmail.com Tue Nov 18 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Tue, 18 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141117 Message-ID: <20141118010502.1074018A008D@apolo.teamnet.rs> [00:10] JEEB: thx for advice with bufsize [00:11] it really helped! [00:11] no problem [00:12] also don't use minrate, that one you shouldn't touch ;) [00:12] just maxrate and bufsize [00:12] that lets the encoder use rate control as well as possible within your limitations [00:12] (plus of course the rate control mode, which in your case would be CRF I guess) [00:31] brainopia it's possible to force specific moov size but that may not end well [00:33] I would probably interleave with mp4box, it spreads the moov over a specified amount of time making it possible to download all of it to start the playback. Usually about 55-60% of total [00:33] to not download** [00:35] does someone know how i can enable ffplay when compiling ffmpeg? [00:36] i tried with --enable-ffplay but there is nothing being installed called "ffplay" in /usr/local/bin/ [00:37] my configure command is: ./configure --enable-gpl --enable-nonfree --enable-ffplay [00:44] eh, nevermind. forgot to install libsdl and googled for "compile ffmpeg ffmpeg" instead of "compile ffmpeg ffplay" [05:41] I had a question: I have two identically-formatted but different length audio files (.ogg) with one stream and two channels each and i'd like to merge them into one ogg file with two streams (each stream still containing its own two channels as it did before). Is there an easy way to do this in ffmpeg? [05:57] hello [05:57] is there any onre [05:57] one* [06:32] Hi, in 2011 installed ffmpeg version CVS. After converting an avi file to mp3 at 64k, I could split that file into 0.5 second clips and then concat any group of them into a file on the fly and have good quality. I just brew installed the latest release and get choppy results. The complete version description I had was: [06:33] FFmpeg version CVS, Copyright (c) 2000-2004 Fabrice Bellard [06:33] Mac OSX universal build for ffmpegX [06:33] configuration: --enable-memalign-hack --enable-mp3lame --enable-gpl --disable-vhook --disable-ffplay --disable-ffserver --enable-a52 --enable-xvid --enable-faac --enable-faad --enable-amr_nb --enable-amr_wb --enable-pthreads --enable-x264 [06:33] libavutil version: 49.0.0 [06:33] libavcodec version: 51.9.0 [06:33] libavformat version: 50.4.0 [06:33] built on Apr 15 2006 04:58:19, gcc: 4.0.1 (Apple Computer, Inc. build 5250) [06:33] usage: ffmpeg [[infile options] -i infile]... {[outfile options] outfile}... [06:33] Hyper fast Audio and Video encoder [06:33] I'm not sure how to return to that same configuration. Any advice? [06:45] i got an error ERROR: freetype2 not found [06:45] tried many to fix it [06:45] but could not [06:46] any one has idea about it that how it get fixed [06:46] :( [09:58] I have this problem on CentOS: http://stackoverflow.com/questions/23537277/ffmpeg-unknown-input-format-lavfi [09:58] any solution? [09:59] I'm trying to build ffmpeg from source [10:14] I'm using deshake=edge=0 and I get a green margin instead of black which is what I would expect "Fill zeroes at blank locations" how can I make it black? https://www.ffmpeg.org/ffmpeg-filters.html#deshake [11:11] hi [11:12] need a little help about bit rates on capturing webcam with ffmpeg [13:32] ffmpeg is hanging ffmpeg -f lavfi -i aevalsrc=0 -f video4linux2 -s 640x480 -i /dev/video0 -strict -2 http://localhost:8090/feed1.ffm [13:33] do you have suggestions to diagnose this? [13:48] hi [13:49] what do i miss to get a non pixelled movie with this : ffmpeg -f video4linux2 -i /dev/video0 date-`date +%d-%m-%Y--%H-%M-`.avi [13:51] Bouib: add "-q:v 3" after the input [13:55] Hello all [13:55] I got a quick one, I need to resample audio streams [13:55] and reading the doc, I just found to ways to do it : AVAudioResampleContext or SwrContext [13:56] does anybody know what is the diff/advantages using one or the other ? [13:59] filed a bug https://trac.ffmpeg.org/ticket/4117#ticket [13:59] very reproducible [15:30] Converting an avi to mp3 at 64k, the quality is fine. Using MP3::Splitter to divide the mp3 into 0.5 second fragments, concatenating the smaller files back together causes choppiness. What ffmpeg library do I need to look into? [15:39] How can I obtain FFmpeg version CVS? [15:43] By traveling back quite a few years in the past [15:57] fangmai: are you looking for source code? [16:00] hi! My problem could simply lack of codecs. The last version of ffmpeg that worked for me was FFmpeg version CVS. However, I'm reading that it might have something to do with mplayer-cvs. [16:01] I just need to be able to concatenate very small files on the fly after they've been split up after being converted to mp3. [16:03] Well for support you have to build with the latest code so you should try that [18:59] Sorry! Will do! [19:00] No problem. It's just hard to help unless we can see what you're doing. [19:19] Hi I have a wide angle camera with 180 degree view, But that video is kind of radially distorted when I save my image [19:20] Does ffmpeg have any filter for this kind of problem or use wideangle camera [19:24] How do I get back to this exact installation on osx? http://pastebin.com/fahpFnWZ [20:04] Is it possible to override the timestamps of a source to workaround this bug? https://trac.ffmpeg.org/ticket/4110 [20:08] relaxed: sorry, i was afk over the weekend and my girlfriend got hit by a car (ok, besides a broken nose and knee). feel free to add whatever you like to the wiki. [20:10] fangmai: why would you want to use something so ancient? [20:30] llogan: all I know is that I could convert an avi to mp3, split it up, and then concat some of the files back together and there was no choppiness. After installing ffmpeg with brew and all the options...it doesn't convert with the same quality. [20:43] michaelni: Just a quick question. Apparently ffplay doesn't have the code to pull remaining bi-directional frames off the stack? Can you offer wisdom? https://github.com/chelyaev/ffmpeg-tutorial/issues/7 [20:48] fangmai: how can the issue be duplicated? [20:52] llogan: I use ffmpeg to convert avi to mp3. MP3::Splitter splits mp3 into 0.5 second or about 4.kb fragments. Using Cat 0.mp3 0.5.mp3 1.mp3 > frag.mp3 results in choppiness instead of a clean sounding file. I'm not familiar with this technology, but I think it has something to do with the codec used to convert the avi to mp3. Working it... [20:54] troy_s, i dont know what you mean by pull and stack [20:59] fangmai: you should provide enough information so anyone can attempt to duplicate whatever issue you're experiencing. this includes all commands and any sample files that are required. [20:59] oh, excuse me. You're right. I need to think about that. [20:59] how do you know the problem isn't with MP3::Splitter (whatever that is)? how do you know it's not your players? [21:01] fangmai: and at the very least you should show your ffmpeg command and the complete console output (use a pastebin service and provide the link). [21:06] llogan: I know that it is the conversion from avi to mp3 and not the splitter or player. I have the mp3 file which was converted with the old ffmpeg build with: ffmpeg -i in.avi -ab 64k out.mp3 [21:06] I can split that file now and then concat it back together perfectly. [21:06] I also have the original avi. Converting it to mp3 with the same command on the current ffmpeg build and then splitting and concatentating back together gives me a choppy file. [21:08] michaelni: The frames that remain. See link? [21:10] right, sorry! moment [21:12] also see quesiton #4 in http://lame.sourceforge.net/tech-FAQ.txt [21:14] troy_s, i just looked at the first message until the list of dts and pts of b frames, i wont read the rest [21:16] the timstamps and frame ordering make no sense [21:17] llogan: Here the command followed by output: http://pastebin.com/dcy9KVQV [21:18] Actually, I should add the splitting to that. moment. [21:22] troy_s, i basically dont know what this is about and i dont have the time to find out [21:24] Hi guys, could someone please dedicate a few mins to explain how to loop png watermarks? [21:24] I'm using the following pattern https://gist.github.com/maokomioko/41fbe9e89c7ab940026a for single placement [21:25] you should always include the complete console output [21:25] Aha, 1 min [21:26] updated [21:28] is it possible to have the transcoded output of one file be copied to another? [21:28] https://trac.ffmpeg.org/wiki/Creating%20multiple%20outputs#Teepseudo-muxer [21:29] llogan: here is the whole process, from conversion to splitting, to concatenating http://pastebin.com/EwjsFp8T [21:30] thanks c_14 [21:31] maoko: so you want [logo] filterchain to repeat indefinitely? i don't really understand what you want to do. [21:32] llogan, within a certain time interval, yes [21:32] Each n-minutes or frames [21:34] maoko: maybe with timeline editing http://ffmpeg.org/ffmpeg-filters.html#Timeline-editing and expression evaluation http://ffmpeg.org/ffmpeg-utils.html#Expression-Evaluation [21:36] llogan, looks like it! It should be placed right after overlay=10:10, right? [21:36] Hmm. But it could interfere with the fade in/out settings [21:37] i don't have a good recommendation or example at the moment [21:37] Trying it now [21:38] overlay supports timeline editing, but fade does not (refer to "ffmpeg -filters" [21:39] fangmai: you could use the atrim and concat filters in ffmpeg and omit the perl thing and cat. [21:39] does using mp3splt instead of MP3::Splitter work for you? [21:40] llogan: that looks hot, I'm going to try that [21:42] will all of your inputs contain MP3 audio as your example shows? or will they vary? [21:43] Usually it's avi or flv to mp3 [21:45] your avi already contains mp3 audio. and your flv may too. [21:47] llogan, thanks for clarifying. Just got your message [22:02] But it could be possible by removing the loop and inserting the same overlay twice [22:02] Trying it now [22:23] Hi all, would anyone be able to help me better understand how to get codecs compiled from source into an ffmpeg build? [22:46] compstomp, did you look at the compilation guide aleady [23:05] yes i have, thank you [23:08] Just to break down the problem into smaller parts, I tried compiling just with libfdk-aac. I wrote a tiny bash script based off the compilation guide to do so: http://pastebin.com/P18YEVDq [23:09] The similar script I wrote for compiling ffmpeg with several components is here: http://pastebin.com/F9FCFb7c [23:10] The error was that libfdk-aac could not be found (i tried with other components as well they, too, could not be found). I'm assuming I am somehow not properly getting the compiled codecs linked into ffmpeg [23:18] http://pastebin.com/ApphQM1E [23:19] compstomp: Search for the line "If you ever happen to want to link against installed libraries [23:19] " and follow what it says. [23:34] First, thank you for your help. I see that quote shows up in many references online. Along side it, libtool also is always mentioned. As I'm sure its clear, I'm not very well versed in linking together libraries/binaries. Are you suggesting that I somehow compile my codecs into .lo objects and then somehow reference them during the ffmpeg compilat [23:34] ion process? [23:35] No, I am saying set where it looks for libraries. [23:35] It says what to do further along in that section. [23:37] compstomp: AFAICT you gave the library a prefix of --prefix="$HOME/Software/ffmpeg/ffmpeg_build", but your ffmpeg --extra-*flags are pointing to $HOME/ffmpeg/ffmpeg_build/include [23:38] hi [23:38] i want to embed some srt subtitle to a video, I used -vf subtitles=foo.srt but.. [23:38] the sub appears too small [23:38] ah great, thank you both! [23:38] llogan is the best. [23:39] no, i just don't want to do any paid work today. boring. [23:39] is there a way to increase it? can I only increase if I convert the sub to ASS and then change the size there? [23:40] diegoviola: can you share the srt file, or just pastebin the contents? [23:42] and also... [23:42] http://ix.io/fek [23:49] after changing it to --extra-cflags="-I$HOME/ffmpeg/ffmpeg_build" I still get the same error :/. I don't expect you guys to bother with my debugging, but might you be able to help me better understand the structure / syntax of how an ffmpeg compilation pulls in libraries? [23:50] compstomp: Did you recompile and change your --prefix? [23:51] any ideas [23:51] how to increase subtitle font [23:51] I recompiled (ffmpeg) and changed my cflags to be the ffmpeg_build. Was that incorrect and instead the --prefix should have referenced ffmpeg_build/include? [23:51] diegoviola: i was still waiting for your ffmpeg command and complete console output [23:54] ffmpeg -i input.avi -target ntsc-dvd -q:a 0 -q:v 0 -vf subtitles=foo.srt out.mpg [23:56] using -q:a and -q:v defeats the purpose of using -target [23:56] you must have seen that on the arch wiki? [23:57] yes [23:57] ok I will remove it [23:57] i removed that from the wiki but reverted it after getting several wiki overlord messages [23:58] so it can remain the shitpile that it is [23:58] just add a discussion on the wiki or something [23:58] i wasted enough time on it [23:59] always other things to do [00:00] --- Tue Nov 18 2014 From burek021 at gmail.com Tue Nov 18 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Tue, 18 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141117 Message-ID: <20141118010503.2843718A00CB@apolo.teamnet.rs> [00:02] ffmpeg.git 03Michael Niedermayer 07master:00672d2ce513: avcodec/rl2: clear freed pointers [00:02] ffmpeg.git 03Michael Niedermayer 07master:7ababb85f963: avcodec/snow: clear freed pointers [00:03] ffmpeg.git 03Michael Niedermayer 07master:20bf91f8322f: avcodec/roqvideoenc: clear freed pointers [00:27] ffmpeg.git 03Reynaldo H. Verdejo Pinochet 07master:33d6f90e3e02: ffserver: drop pointless explicit !=0 checks [00:28] ffmpeg.git 03Lukasz Marek 07master:1a054bd18767: ffserver_config: do not store preset name [00:28] ffmpeg.git 03Lukasz Marek 07master:3f07dd6e392b: ffserver_config: fix possible crash [01:20] ffmpeg.git 03Michael Niedermayer 07release/2.3:80473805149b: avformat/segment: export inner muxer timebase [01:20] ffmpeg.git 03Michael Niedermayer 07release/2.4:cc9c74ea8773: avformat/segment: export inner muxer timebase [04:14] ffmpeg.git 03Michael Niedermayer 07master:4be03a7a6c39: cmdutils: Exit in case of faulty stream specifiers [05:11] ffmpeg.git 03Michael Niedermayer 07master:9421d974bc79: ffmpeg: Fix last newline at log level less than "info" [08:42] 'morning [11:19] ffmpeg.git 03Anton Khirnov 07master:920bca3e2332: hevc: do not store pcm_flag in the context [11:19] ffmpeg.git 03Michael Niedermayer 07master:7cbe1e044701: Merge commit '920bca3e2332dced9c78bd14cfc2ff138188bd57' [11:27] ffmpeg.git 03Anton Khirnov 07master:2c6a7f934837: hevc: do not store rqt_root_cbf in the context [11:27] ffmpeg.git 03Michael Niedermayer 07master:f1b20930f206: Merge commit '2c6a7f9348378f887066fb1669c46b9485e8ef3e' [11:36] ffmpeg.git 03Anton Khirnov 07master:84b946398408: hevc: remove a redundant line [11:36] ffmpeg.git 03Anton Khirnov 07master:16c01fb43473: hevc: remove an unused function parameter [11:37] ffmpeg.git 03Michael Niedermayer 07master:e7fdfbdc58bd: Merge commit '84b9463984083f4e83948c73c1a5dbaf596ff3f7' [11:37] ffmpeg.git 03Michael Niedermayer 07master:c192be196839: Merge commit '16c01fb4347312b6d29a6498dad627665b96a20e' [11:46] ffmpeg.git 03Anton Khirnov 07master:de1f8ead8993: hevcdsp_template: templatize transquant_bypass [11:46] ffmpeg.git 03Michael Niedermayer 07master:30156eab6d81: Merge commit 'de1f8ead8993512925a3ee6c7491473414419e55' [12:00] ffmpeg.git 03Anton Khirnov 07master:a7b365ae191f: hevc: reduce code duplication in hls_prediction_unit() [12:00] ffmpeg.git 03Michael Niedermayer 07master:e078549421b2: Merge commit 'a7b365ae191f45a0d7ed7b34033d5d0cbdd47139' [12:06] ffmpeg.git 03Anton Khirnov 07master:84c0ece5fd95: hevc: further reduce code duplication in hls_prediction_unit() [12:06] ffmpeg.git 03Michael Niedermayer 07master:c23d7de22ec0: Merge commit '84c0ece5fd9569c0f31804f02a199ecd0e7d13d8' [12:13] ffmpeg.git 03Anton Khirnov 07master:eb335f3c5ce3: hevc: reduce variable scope [12:13] ffmpeg.git 03Michael Niedermayer 07master:0d5af820f7be: Merge commit 'eb335f3c5ce37f2b93c993e28404d113bee843bc' [12:18] ffmpeg.git 03Anton Khirnov 07master:eac3ac1fe077: hevc: eliminate an unneeded intermediate variable [12:18] ffmpeg.git 03Michael Niedermayer 07master:91a9ae5b6bbf: Merge commit 'eac3ac1fe0774b65316852616b2672702dbc3f31' [12:38] ffmpeg.git 03Anton Khirnov 07master:8b573ddda759: hevc: remove superfluous assignments and checks [12:38] ffmpeg.git 03Michael Niedermayer 07master:c88ae843da4a: Merge commit '8b573ddda75980f724f779ff75aacc2ff81d9e0e' [12:58] ffmpeg.git 03Andrew Stone 07master:4d0cd5f58c89: flvenc: move metadata updates into a single function [12:58] ffmpeg.git 03Michael Niedermayer 07master:bb7be3b763d3: Merge commit '4d0cd5f58c892276716f46f4b2702915e5018215' [13:08] ffmpeg.git 03Andrew Stone 07master:c64f3615118d: flvenc: Send new metadata when FLAG_METADATA_UPDATED is set. [13:08] ffmpeg.git 03Michael Niedermayer 07master:4127d97c8de6: Merge commit 'c64f3615118d757dcf76040fe5407bf2b3883206' [13:50] Hi, does anyone know where I can get an up-to-date spec file for building ffmpeg into an RPM? I can't seem to see it in the git repo. [13:54] There is no officla one, every distro does packaging itself [14:13] Hmm, okay thanks. [15:23] ffmpeg.git 03Michael Niedermayer 07master:cc5f73154132: avformat/flvenc: remove unused variable [15:23] ffmpeg.git 03Matthew Oliver 07master:70205f179925: mpcodecs: Use _INLINE guards for inline asm. [15:24] ffmpeg.git 03Matthew Oliver 07master:0b3c23054279: configure: Enable mpcodec compilation without inline asm. [17:17] _ChunkyRGB16toChunkyYUV16 I guess that's one way to spell "packed" [17:18] lol [17:45] Daemon404: Any news about the Prores sample? [19:05] ffmpeg.git 03Carl Eugen Hoyos 07master:92c07acce736: Read (display) aspect ratio from mxf files. [21:11] ffmpeg.git 03Michael Niedermayer 07master:894d10332ca5: avcodec/lcldec: support rgb24 with width%4 != 0 [21:11] lcldec ? hmm [21:11] dont remember that one [21:12] ah zlib and mszh [21:12] got some 'convert colorspace' code in it. how many decoders are duplicating swscale ? [21:23] michaelni, Compn: did you unsubscribe peter_trompeter at hotmail from -user due to the spam ([FFmpeg-user] FW:infoe)? i'm wondering how the message got through. i'm assuming he was subscribed. [21:27] i dont watch ffmpeg-user much [21:27] so i dunno [21:43] ffmpeg.git 03Martin Storsj? 07master:0f9eb9165bb7: movenc: Include empty tracks in iods when writing fragmented mp4 [21:43] ffmpeg.git 03Michael Niedermayer 07master:b96c1cd78b85: Merge commit '0f9eb9165bb7d7982fdedf64f6bcec856f1bedd6' [21:46] llogan, no matches in logs/subscribe for peter_trompeter so i would guess he was never subscribed with that email [21:48] thanks for looking. i shouldn't have been so lazy. i don't know how it got there then. [21:53] ffmpeg.git 03Martin Storsj? 07master:2d9d6afb8d2f: movenc: Factorize adding fragment info into a separate function [21:53] ffmpeg.git 03Michael Niedermayer 07master:1fddfaa282ff: Merge commit '2d9d6afb8d2f284f5e620ecc19f643d5cd3facb8' [22:05] ffmpeg.git 03Martin Storsj? 07master:2ded57371abe: movenc: Add support for writing sidx atoms for DASH segments [22:05] ffmpeg.git 03Michael Niedermayer 07master:9e0b0c60bd7f: Merge commit '2ded57371abead879bcee56da5131e5fac0d17ef' [22:23] ffmpeg.git 03Martin Storsj? 07master:fe5e6e34c05e: lavf: Add an MPEG-DASH ISOFF segmenting muxer [22:23] ffmpeg.git 03Michael Niedermayer 07master:b5b15c4dd943: Merge commit 'fe5e6e34c05e274f98528be4f77f3c474473f977' [22:39] ffmpeg.git 03Martin Storsj? 07master:3847f3ab58b3: movenc: Add tfra entries for all tracks in a moof [22:39] ffmpeg.git 03Michael Niedermayer 07master:3fa4351d6bea: Merge commit '3847f3ab58b3b74604807394247bf68827258103' [23:09] ffmpeg.git 03Martin Storsj? 07master:40ed1cbf147d: movenc: Allow writing a DASH sidx atom at the start of files [23:09] ffmpeg.git 03Michael Niedermayer 07master:4dc305d784ea: Merge commit '40ed1cbf147d09fc0894bee160f0b6b6d9159fc5' [23:13] http://b.pkh.me/hqx-xbr-pkmn0.png [23:14] http://b.pkh.me/hqx-xbr-pkmn1.png [23:15] http://b.pkh.me/hqx-xbr-pkmn2.png [23:16] http://b.pkh.me/hqx-xbr-pkmn3.png [23:16] enough for tonight. [23:18] the wooden table is nice on -pkmn2 [23:26] ffmpeg.git 03Martin Storsj? 07master:c302d034ba26: tools: Add a sidxindex tool [23:26] ffmpeg.git 03Michael Niedermayer 07master:e60c025e7357: Merge commit 'c302d034ba2690a935df8bf7e4f5d44ed86e8d5c' [23:33] some of the tools are poorly documented (if at all). maybe they should be added somewhere in doc. or maybe just a wiki page. [23:49] http://pastie.org/pastes/9726335/text ? btw, the redundancy of drawtext args is ugly [23:49] also, using vars could be nice [23:51] and btw, wtf @ http://pastie.org/pastes/9726340/text [23:51] not sure what went wrong here [23:51] looks like a bad error handling in drawtext, because it looks fine [23:51] never seen that one before, IIRC [23:52] try with a --enable-fontconfig build? [23:53] ffmpeg.git 03Michael Niedermayer 07master:afbaa9a737b3: avdevice/oss_audio: avoid strerror() and errbuf [23:54] ubitux: ah, there it is (trying 2.4.3 in Extra repo) [00:00] --- Tue Nov 18 2014 From burek021 at gmail.com Wed Nov 19 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Wed, 19 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141118 Message-ID: <20141119010503.A6B6B18A02E0@apolo.teamnet.rs> [00:11] I used to think that hqx was the best scaler for pixel art but I'm now leaning to xbr [00:11] those images make it look very good [00:11] Except the grass texture in the first [00:12] that looks bad with both scalers [00:24] Action: gnafu has never been a fan of ultra-smooth scalers, and rather prefers nearest neighbor or similar. [00:28] J_Darnley: the rendering of the grass is actually original :D [01:06] i prefer how hqx handled the kid in pkmn0. xbr fused his eyes and hair together [01:08] yeah :) [02:55] ffmpeg.git 03Martin Storsj? 07master:9257692ac15e: lavf: Only initialize s->offset once when using avoid_negative_ts make_zero [02:55] ffmpeg.git 03Michael Niedermayer 07master:9943c5a42bdf: Merge commit '9257692ac15eff7b07540c1f61cebde0d8823fbd' [03:07] ffmpeg.git 03Vittorio Giovara 07master:a28468d0daf4: librtmp: append the correct field to the string [03:07] ffmpeg.git 03Michael Niedermayer 07master:7384ec19cf13: Merge commit 'a28468d0daf4be14761c16a3ddd33266b2380123' [03:15] ffmpeg.git 03Vittorio Giovara 07master:771656bd8541: libvpxenc: clean memory on error [03:15] ffmpeg.git 03Michael Niedermayer 07master:8426edef4cf5: Merge commit '771656bd85416cd6308b11aed6f2c69a8db9c21b' [03:32] ffmpeg.git 03Vittorio Giovara 07master:85dc006b1a82: lavc: fix bitshifts amount bigger than the type [03:32] ffmpeg.git 03Michael Niedermayer 07master:cdbebae44440: Merge commit '85dc006b1a829726dd5e3a9b0fcc6a1dbfe6dffa' [03:40] ffmpeg.git 03Vittorio Giovara 07master:2007082d2db2: mov: check ff_get_wav_header() return value [03:40] ffmpeg.git 03Michael Niedermayer 07master:4cd4a6de4f97: Merge commit '2007082d2db25f9305b8a345798b840ea7784fdb' [03:48] ffmpeg.git 03Michael Niedermayer 07master:9bb6e1175f6e: avcodec/internal: Add () to argument of FF_SIGNBIT() to ensure correct order or operations [04:34] ffmpeg.git 03Michael Niedermayer 07master:0b75b6c3cd2c: avformat/mp3dec: avoid seeking to negative positions [05:18] ffmpeg.git 03Michael Niedermayer 07master:4243415741e3: avcodec/mjpegdec: Support some subsampled GBR variants [07:27] Finally got Lollipop on my Nexus 7 2012 [11:54] ffmpeg.git 03Michael Niedermayer 07master:ade140eb7331: libavcodec/libx264: Use av_freep() avoid leaving stale pointers in memory [11:54] ffmpeg.git 03Michael Niedermayer 07master:98fbf8ef6737: libavcodec/libxavs: Use av_freep() avoid leaving stale pointers in memory [11:54] ffmpeg.git 03Michael Niedermayer 07master:bb5e1482992d: libavcodec/tiffenc: Use av_freep() avoid leaving stale pointers in memory [11:54] ffmpeg.git 03Michael Niedermayer 07master:a54a51cc9bdd: avutil/float_dsp: add avpriv_float_dsp_alloc() [11:54] ffmpeg.git 03Michael Niedermayer 07master:4eae568a0712: doc/APIchanges: Fix some wrong versions [13:53] ffmpeg.git 03Michael Niedermayer 07master:06d27428995b: avcodec/nellymoser: Use avpriv_float_dsp_alloc() [13:53] ffmpeg.git 03Michael Niedermayer 07master:aa97223f14a1: avfilter/af_amix: Use avpriv_float_dsp_alloc() [13:53] ffmpeg.git 03Michael Niedermayer 07master:9e526213a27f: avfilter/af_volume: Use avpriv_float_dsp_alloc() [17:46] ffmpeg.git 03James Almer 07master:bccae39072c4: lavf/ffmenc: fix memleak in ffm_write_header [18:47] ffmpeg.git 03Vadim Kalinsky 07master:d1d390427342: avcodec/options: Set AVCodecContext->codec upon initialization. [19:32] should be showing up in vivid (proposed) soon https://bugs.launchpad.net/ubuntu/+bug/1393522 [20:30] ffmpeg.git 03Peter Hall 07master:ea79dfbad321: avcodec/libvorbisenc: Give CODEC_CAP_SMALL_LAST_FRAME to libvorbis encoder. [20:42] ffmpeg.git 03Martin Storsj? 07master:7813e6752bda: configure: Fix enabling memalign_hack automatically [20:42] ffmpeg.git 03Michael Niedermayer 07master:30d3f9769ff3: Merge commit '7813e6752bdab38a5686c301e869ee71d97bce69' [21:00] nevcairiel: supposedly lavfsplitters has some hack to get the "correct" mkv duration (e.g. if the value in the header is wrong); what exactly does it do? [21:00] read the last packet [21:01] seeing nothing about that in matroskadec_haali.c [21:01] its in haalis parsing code [21:01] MatroskaParser.cpp [21:01] thanks [21:03] oh dear that's weird [21:03] so it uses the index to seek to the last index entry, and reads all packets [21:12] ffmpeg.git 03Cl?ment BSsch 07master:6da12d46d3d7: doc/writing_filters: use a more portable sed command in the walk-through [21:12] ffmpeg.git 03Cl?ment BSsch 07master:568f1ecccf5c: doc/writing_filters: fix reference to Lenna image [21:15] damn OSX and BSDs not having a -i in sed [21:29] ubitux: tempfile or perl it is ! [21:30] :( [21:54] j-b: i just read your gsoc 10 year summary. are you going to apply again? [21:59] llogan: no [21:59] it's a waste a time [21:59] ffmpeg.git 03Rodger Combs 07master:5f8fcdd4481b: dashenc: degrade gracefully if a stream's bitrate is unavailable [21:59] llogan: sad business... [21:59] i only ask because the summary seemed very positive [22:00] writing about the negative parts is useless... [22:01] so I decided to focus on the good parts [22:02] got a possibly broken lagarith sample on twitter if anyone wants to monkey with it https://twitter.com/cracki/status/534553384897224705 [22:04] j-b: personally, i am done with it too [22:04] llogan: sadly, that's rhe right way [22:05] but I wanted to blog about the positive parts [22:14] ...the twitter guy showed up in #ffmpeg and i asked him to submit a bug report. [00:00] --- Wed Nov 19 2014 From burek021 at gmail.com Wed Nov 19 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Wed, 19 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141118 Message-ID: <20141119010502.82AE618A0254@apolo.teamnet.rs> [00:02] Action: llogan is done waiting for console output and gets to work [00:03] so how to increase subtitle size [00:03] Just my two cents, if you're using VLC to playback you can just do so there [00:04] compstomp: If you compile fdk with one prefix, but tell ffmpeg to look in a different place, then you're going to get that problem. [00:04] That's what llogan was saying. [00:05] The prefixes for both compilations were the same though :/ [00:05] "--prefix="$HOME/Software/ffmpeg/ffmpeg_build", but your ffmpeg --extra-*flags are pointing to $HOME/ffmpeg/ffmpeg_build/include" [00:06] I realize I fixed a backup I made not the original. Will check back. Thanks again! [01:16] Hi all, I am having trouble linking x265 compiled from source into ffmpeg. Here is the complete console input / error output: http://pastebin.com/YhKSSpJh [01:19] the accompanying config.log: http://pastebin.com/Hrh6VqeZ [01:26] Does "$HOME/Software/fmpeg/ffmpeg_build/lib/pkgconfig" contain x265.pc ? [01:26] it doesn't for some reason [01:27] Well, that's your problem. [01:27] although curiously enough, a previous try at this did result in it being there [01:27] Where are you installing x265? [01:27] ie, where in the world is x265.pc [01:28] http://pastebin.com/hYzR3SfS [01:28] I must somehow be setting the flags wrong with cmake [01:29] Well, nah. It's easy then. you wrote ffmpeg with one f in the PKG_CONFIG_PATH variable [01:29] try adding another one [01:30] how many threads can ffmepg use? [01:31] gah! so simple. Thanks! will check back [01:31] diegoviola: depends on the encoder [01:32] Usually some funky fraction of the resolution. [01:32] isn't ffmpeg the encoder? [01:32] encoder as in codec [01:32] Nah, ffmpeg is a wrapper. [01:33] libx264..etc [01:33] Well, a wrapper as well as a few other nice things. [01:34] c_14, did you see anything that might have indicated why x265.pc might have absconded off into the darkness? [01:34] Did you check the missing f thing? [01:36] Yep have corrected that. But that is an a script called after the completion of compiling x265. Isn't that a biproduct of the x265 compilation? [01:40] Try with -DLIB_INSTALL_DIR=$HOME/Software/ffmpeg/ffmpeg_build/lib [01:41] And keep all existing flags or drop one? [01:41] Keep them. [01:41] Shouldn't hurt. [01:42] kk, much appreciated. Will take quite awhile for my old laptop to compile again but will be back with the results [01:44] Before trying again, is it perhaps wring to be setting the current binary dir? Should I instead be setting the EXECUTABLE_OUTPUT_PATH? [01:48] Just use the bin dir one [01:48] does the -DLIB_INSTALL_DIR= not require the :PATH= syntax? [01:49] system is 32 bit Linux Mint 17 if that is relevant [01:49] eh, it might [01:50] tbh I usually just use the funky ncurses ui thingy [01:56] Probably worth trying that route just to see if I can get it to work, but really would like some zero interaction scripts for posterity. Cmake just now starting so probs another 25 min + of compilation. If you're not still around then, thanks for your help! [02:39] Would anyone familiar with the cmake compilation process for the videolan x265 codec have an understanding why I'm getting this strange, but not complex, error? http://pastebin.com/cRmD0Ad1 Line 489 should read: /home/sauser/Software/ffmpeg/ffmpeg_build/lib/pkgconfig/x265.pc [02:47] eh, hrm. I might have found the issue. Get rid of the -DLIB_INSTALL_DIR and the -DCMAKE_CURRENT_BINARY_DIR [02:48] The cmake actually uses the CMAKE_CURRENT_BINARY_DIR as the source location of the x265.pc file [02:48] So if you change it, it won't find the pc file and won't copy it to the pkgconfig folder [02:49] Set BIN_INSTALL_DIR instead [02:49] My cmake foo is not very strong. [02:53] Hey you got me to the point that after just moving the .pc file its actually compiling! [03:41] hello [03:42] im trying to do RTMP to RTSP live-reconversion [03:42] but the server doesn't start and there's no error [03:59] http://justpaste.it/i19n [05:12] I'm trying to change the font of an ASS subtitle [05:12] the ASS file has something like this: [05:12] Style: Default,Arial,16,&Hffffff,&Hffffff,&H0,&H0,0,0,0,1,1,0,2,10,10,10,0,0 [05:12] I change Arial to something else and it still looks the same [05:12] any ideas? [05:16] nvm got it to work [08:48] Hi all [08:49] i got something really strange, when i set the fps on 30 VLC shows me that the frame rate is 1000 ? how is this possible ? [09:26] i'v find out the problem, the streaming fps is not constant [09:26] it will going down... how more frames it sends how lower it goes [10:07] i have make some logs https://gist.github.com/anonymous/b2af1c5c63b95361dcd1 [10:52] Is there someone who can help me with this ? [12:43] Anyone happen to have a copy of the CEA-708-E spec? [13:04] hi [13:04] https://www.ffmpeg.org/ffmpeg-filters.html#Examples-40 lists "a quick emboss effect" [13:04] when trying to get it going I get what seems like a syntax error [13:04] ffmpeg -filter_complex format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2' -f matroska - | ffplay -i - [13:04] what am I doing wrong? [13:05] (I am currently just testing it to see if it applies - I want to convert a small graphic to an embossed-like image in the top right corner, something like a watery-watermark) [13:13] ah I see... [13:54] in the geq filter [13:54] why does lum=p(X,Y) work as expected (leaving the image intact) [13:55] yet lum=lum(X,Y) doesn't? [14:29] in a complex filter how can I ask for the size of a particular input (not part of the current filter chain) [14:36] you guys don't talk much do you ;) [15:48] grumper: there is no way, you need to rely on external scripting [16:45] Hi all is it possible to set the 1k tbn, 15 tbc ? [16:45] now the input tbn and tbc is different from the output [16:53] encoding with libx264: is it possible to use one crf value for the first X seconds of the video, then another crf value for the remainder? [16:57] rsully: you can encode the X seconds at one crf, cut the recording at that point, encode the rest of the video at another crf and then concat [17:02] what is crf ? i'm using it but don't know what it is [17:02] constant rate factor [17:02] Basically a mechanism to ensure that the "visual quality" of a video is constant over time. [17:03] thank you c_14 [17:03] c_14: do you know how to set the tbn ? [17:04] The input is 1200k tbn and the output is 1k tbn so i think thats not good right ? [17:06] Armada found this post from 2 years ago http://ffmpeg.org/pipermail/ffmpeg-user/2012-November/011391.html [17:06] Aartsie * sorry [17:08] rsully: oke so it is not possible to change them :) [17:09] yeah unless you're doing some super critical video editing it doesn't look like there is any need to modify them either [17:10] rsully: ok :) the problem i have is some latacy of 13 seconds [17:10] rsully: but i don't know how to improve that on a rtmp stream [17:11] what is your use of ffmpeg? I mostly just use it for thumbnails and converting videos to mp4, so I've never dealt with latency [17:13] you also might want to check this guide https://trac.ffmpeg.org/wiki/StreamingGuide [17:13] i have a raspberry pi camera so it will converting from RAW H264 to x264 and stream it by rtmp [17:13] yes thank you i've already see that [17:14] how are you serving the rtmp stream? [17:15] rsully: /opt/vc/bin/raspivid -t 0 -fps 15 --width 1280 --height 720 -v -o - | ffmpeg -framerate 15 -re -i - -vcodec copy -r 15 -g 30 -preset ultrafast -maxrate 2400k -bufsize 960k -f flv rtmp://192.168.1.53/live/test [17:16] ooh I didn't even know ffmpeg could do that [17:17] rsully: hahaha ok no problem :) ffmpeg is really nice [17:17] Aartsie: most of your output options there are completely ignored because you are using -vcodec copy [17:17] since copy isn't reencoding the video [17:18] when I encode a movie with -target ntsc-dvd I always get 16:9 res, is there a way to get 4:3? [17:18] for DVD [17:18] diegoviola: it should automatically set the output aspect ratio correctly if your input video is correct... [17:19] ok [17:19] ty [17:22] diegoviola: if your input video isn't correct, you should look at the 'setdar' video filter in ffmpeg; if your input is 16:9 and you want to convert it to 4:3, then try the crop filter. [17:28] the input aspect ratio is correct [17:52] hi, is anyone experienced with muxing raw h264 frame stream into mp4 container, i assume i have trouble with setting up the extradata as it was written at: http://www.szatmary.org/blog/25, I am using avformat from ffmpeg version 2.4.x., after setting up extradata which sps = 6 bytes and pps = 2 bytes, and succesfully container is created i use mplayer to player my file and mplayer prints errors about deco [17:52] ding some of the frames, each key frames are succesfully decoded and renderer and P/B aren't, is someone able to give me some adivces upon the issue ? [18:00] pierre__, check if your input stream is in AnnexB format [18:00] if it is, you'll have to strip all the PSS/SPS packets out of it and only write them once in the header [18:04] this is what i do, pps and sps frames comes separetly to my muxer, once i receive them i create avcc from them and put that data into contextcodec.extradata then skiping av_write_frame with those frames, next frames are placed to the av_write_frame, btw is it correct to write each frame (VCL) with annex b to mp4 container ? [18:08] i will give you more light on the issue, I have recorder which stores raw h264 frames in custom conterner, frames are stored already in annex b format, the next step is to export the recorded material and throw those frames into muxer, and this is the problem end of the pipline, the muxer [18:21] hi there, I do not see Subtitle stream here: http://pastebin.centos.org/13946/ but in the original video there is one: http://vimeo.com/43687450 how could this be? [18:32] xvzf: maybe the player overlays the subs [18:36] relaxed, so simply I did not download the subtitles, because vimeo does not let me that? [18:54] xvzf: I'm just guessing, of course, but if that is what's going on- yes [19:32] How can I remove only the first ten seconds of a directory full of tv shows so that my .srt files will be displayed properly? [19:33] for i in dir/*; do ffmpeg -ss 10 -i "$i" "${i%.*}_short.mkv"; done [19:34] fhenning09 i would consider updatting the srt files instead of the video files [19:34] well they have some dudes twitter banner is on for ten seconds so i dunno [19:35] that sucks. i would try to find the original source without that (whoever sync'd the subtitles must have had it?) [19:44] c_14: Thanks was what i was looking for just might b a while [19:45] fhenning09: you may also want to add "-c copy" [19:45] well, with -c copy it won't be able to get the time exactly correct - although it is likely that the first frame after the banner will be a keyframe, for what it's worth. [19:46] won't it likely get it incorrect anyways because the -ss is before the -i? [19:46] nah, -ss before -i should be frame accurate if you're reencoding the video. [19:47] hmm ok. i always have issues when trying to generate a png frame [19:49] http://pastebin.com/XTYK6UP5 [19:50] very detailed manpage [19:52] -ss position (input/output) [19:52] When used as an input option (before "-i"), seeks in this input file to position. Note the in most formats it is not possible to seek exactly, so [19:52] ffmpeg will seek to the closest seek point before position. When transcoding and -accurate_seek is enabled (the default), this extra segment between [19:53] the seek point and position will be decoded and discarded. When doing stream copy or when -noaccurate_seek is used, it will be preserved.-ss position (input/output) [19:53] When used as an input option (before "-i"), seeks in this input file to position. Note the in most formats it is not possible to seek exactly, so [19:53] ffmpeg will seek to the closest seek point before position. When transcoding and -accurate_seek is enabled (the default), this extra segment between [19:53] the seek point and position will be decoded and discarded. When doing stream copy or when -noaccurate_seek is used, it will be preserved.-ss position (input/output) [19:53] use pastebin [19:53] When used as an input option (before "-i"), seeks in this input file to position. Note the in most formats it is not possible to seek exactly, so [19:53] ffmpeg will seek to the closest seek point before position. When transcoding and -accurate_seek is enabled (the default), this extra segment between [19:53] the seek point and position will be decoded and discarded. When doing stream copy or when -noaccurate_seek is used, it will be preserved. [19:53] well that was useless [19:54] http://pastebin.com/FFAvysK8 [19:54] sorry [20:00] Hi folks [20:01] Is there a modern binary of FFmpeg which has libfdk_aac? [20:01] No. [20:02] :( [20:02] The libfdk_aac license prohibits binary redistribution. [20:02] it's not distributable so you can only build it yourself :P [20:02] Okay. Well, I'm trying to get this result: http://sinclairmediatech.com/building-ffmpeg-with-libx265/ [20:02] I see. [20:02] I'm just trying to avoid all the messy building process [20:02] I would like to try libx265 with libfdk_aac. [20:03] Hmm [20:03] you mean x264 probably, since x265 is not yet ready to be usable unless you are trying to do some very very low bit rate scenario [20:03] oh, i actually mean x265 (hevc) [20:04] i need low file size [20:04] well, what kind of low [20:04] enabled gpl && die_license_disabled_gpl nonfree libfdk_aac [20:04] JEEB: that's the question. that's what I want to experiment with. [20:04] because where x265 beats x264 is where both look bad, but x265 just looks better [20:05] if you just mean "as visually lossless as possible but still compressed quite a bit" then x264 is probably what you still want [20:05] I'm trying to make acceptable proxy videos for enormous DV-AVI files [20:05] for editing? [20:05] No, for searching auto-captions and viewing the results [20:05] I have hundreds of hours. like 730 GB [20:06] I haven't tested since end of september or so, but I don't think much has changed since - so if you actually use a bit rate when the x264 psychovisual opts start working, then x264 is better [20:06] It may be acceptable to sacrifice a little quality [20:06] a bit... where x265 wins is where both suck and x265 just sucks somewhat less :P [20:06] JEEB: haha [20:07] but i mean, isn't it supposed to be half the file size? [20:07] JEEB: what about encoding time? [20:07] no [20:07] llogan, x265 was way way slower [20:07] i haven't tried it in a long time [20:07] placebo+ with x265 (everything except ME or so at max), placebo with x264 [20:08] for me, file size is important [20:08] i'm seeing if i can cram these videos onto a large thumb drive [20:08] uex1fi, the specification's theoretical (tested with the reference implementation and PSNR) improvement is somewhere around 30% , but then the real world steps in [20:08] which may not be realistic, but i'm trying [20:08] real world meaning implementations [20:08] what do you mean [20:08] i know i can play them in vlc [20:09] as in, you don't just magically get the theoretical performance improvement [20:09] ok [20:09] esp. when the implementation of the previous format is just SoDamnGood [20:09] with mpeg-4 part 2 it was simpler [20:09] it had no deblock and no CABAC [20:09] (in-loop deblocking I mean) [20:10] so it was rather simple to hop onto AVC within a couple of years [20:10] all right [20:10] now we're 1.5 years or so from the specification, and the previous format's implementation is damn good [20:10] but hevc is (basically) always smaller file size at the same quality as x264? [20:11] uhh [20:11] at this point x265 (the best current HEVC implementation) only wins over x264 in visual quality in very very low bit rate scenarios [20:11] at the same size [20:11] ok [20:11] then [20:11] Have you ever used Freemake video converter? [20:12] I used that a while back to convert hundreds of videos to MP4. [20:12] (I still have the original videos, of course.) [20:12] looks like one of those GPL violators. [20:12] So -- is it possible to get a better-looking encode? [20:12] At a smaller size? [20:13] as soon as the bit rate scenario you are going for is somewhere along where x264's psychovisual opts are actually working good,, x264 wins [20:13] (at very low rates the psychovisual opts actually make the picture look somewhat worse) [20:13] JEEB what bitrate would that be? [20:13] okay. i'm not familiar with those options [20:13] rsully, there are no specific rates [20:13] so i need help [20:13] and it's mostly psy-rd/trellis [20:14] I think AQ is still useful in that part? [20:14] basically, you will notice if you go low enough :P if it generally looks good then you're not there yet. [20:14] if I can upload maybe a 20-second dv-avi, does someone want to mess with it? [20:15] anyways, if you want a realistic alternative for relatively well compressed video, just use x264 and pick the slowest preset that is still fast enough for you [20:15] uex1fi: use x264, set the preset as slow as you can handle, use crf mode if you don't need a fixed file size, and 2-pass bitrate mode if you do need a fixed filesize. [20:16] it doesn't need to be literally fixed, but generally quite small and well-compressed [20:16] which is what crf is for [20:16] ok then [20:16] start with 23 and go up if it still looks good, down if it looks bad [20:16] then you will hit the highest crf value that still looks "good enough" for you [20:16] crf is "constant rate factor", it's an arbitrary number where bigger numbers give you worse video quality/smaller files. [20:17] ok [20:17] which is the most amount of compression you will get, basically [20:17] would you recommend libfdk_aac for audio? [20:17] i'm curious how that compares to libvo_aacenc [20:18] much better [20:18] libvo_aacenc is horribly bad. [20:18] even the internal aac encoder is better than vo_aacenc [20:18] then... my video source is interlaced. [20:18] ffmpeg is defaulting to libvo_aacenc for me with mp4 output [20:18] uex1fi: for best video compression, deinterlace it before compressing. [20:18] yes, because someone decided it's better in whatever crazy way :P [20:18] hmm [20:19] so an interlaced x264 file is not smaller? [20:19] smaller file size [20:19] JEEB what should i force it to use? i'm using osx binary [20:19] -c:a aac -strict experimental [20:19] or you could compile [20:19] even with brew if you prefer [20:19] llogan and then use libfdk_aac? [20:20] yes [20:20] you'd have to compile fdk too [20:20] because I assumed that interlaced is less data than progressive [20:20] how much better is libfdk_aac than the builtin one [20:20] in most cases, i would expect the original interlaced video to be slightly bigger than a deinterlaced copy, both encoded with x264 [20:20] interesting... so deinterlaced can actually be smaller [20:20] rsully: depends on how much you trust your own ears or a japanese audio fanatic [20:21] uex1fi: assuming that you have a 60 fields per second interlaced video, and deinterlace it to 30 frames per second [20:21] rsully: http://trac.ffmpeg.org/wiki/CompilationGuide/MacOSX#ffmpegthroughHomebrew [20:21] may be useful. i don't know. [20:21] llogan yeah i usually use homebrew, i was just doing some encodes on a freshly installed computer and didnt feel like downloading xcode [20:21] this is NTSC 29.97 video [20:21] llogan now i know how much of a bad idea that was [20:22] depends on how lazy you feel like being [20:22] or use the quicktime aac encoder. it's good [20:22] and then remux [20:22] uex1fi: 60 fields per second interlaced and 30 frames per second progressive is the same amount of "raw" video data, but video encoders generally can do better predicition with progressive frames [20:23] makes sense [20:24] ok, so besides deinterlacing, crf mode, and two-pass, should i mess with any other settings? [20:24] can i squeeze out more quality? [20:24] use either crf or two pass; they are different (and conflicting) [20:25] uex1fi: don't bother touching any of the x264 options other than the -preset [20:25] ok thanks [20:25] then which is better? crf or two-pass? [20:26] two-pass gives a lower file size, i believe [20:26] in crf mode you pick a quality and get filesize determined automatically; in 2-pass mode you pick a filesize and get a quality determined automatically [20:26] the relative quality of crf and 2-pass modes is effectively identical [20:27] (in fact, the way 2-pass mode works in x264 is the first pass calculates the crf needed to get the desired filesize, and the second pass is just a crf encode) [20:28] hmm okay [20:28] so crf isn't a blindly constant quality [20:29] i mean, if the screen is black for 10 seconds, will it lower the bitrate? [20:31] plain black screens are easy to encode, so x264 won't use very many bits on that section of the video, probably... [20:31] ok, thanks for the info. [20:32] that's try of pretty much any codec tho, at least when you're not doing a one-pass constant bitrate encode. [20:32] don't do a 1-pass constant bitrate encode unless you're doing live video streaming :) [20:33] yeah haha [20:33] i'll try it out [20:33] thanks everyone [20:33] see you later [20:34] I am so glad the x264 devs put that preset system in, those websites with random sets of x264 options that nobody really understood were getting annoying :/ [20:37] llogan are there any options i don't want when i compile with homebrew? should i just enable all? [20:38] just enable what you're going to use [20:39] well i didn't even know i had to specify --with-fd-aac but apparently it is so much better. just don't want to miss out on something like that agaib [20:40] --with-fdk-aac --with-freetype --with-libass --with-libvorbis --with-libvpx [20:40] and whatever you need for libx264 support [20:40] i don't know what --with-tools is [20:41] might as well enable opus too, it's lots of fun :) [20:45] alright recompiling ffmpeg now with about 5 options more than i usually use [20:45] so this should only take like an hour [20:47] rsully: what does "brew info ffmpeg" say about --with-tools? [20:52] lol 'Enable additional FFmpeg tools' [20:52] ok. thanks for looking. [20:52] i don't see any ffmpeg-* binaries [20:53] just does this https://github.com/Homebrew/homebrew/blob/5e00a9e354c8c60207591d9185655612c4f3517d/Library/Formula/ffmpeg.rb#L145-L148 [20:54] ah, ok. not typically useful for most users [20:55] i still don't know what is in that dir it installed though [20:55] ah all the ff* tools like ffprobe, ffserver [20:55] the stuff in ffmpeg/tools (in the source directory) [20:57] doesn't homebrew cleanup the src dir? can't find anything useful in /usr/local [20:58] i don't really know anything about brew [21:17] hello [21:17] does 10-bit x264 produce a smaller file size than 8-bit? [21:19] uex1fi: given roughly the same output quality, I think you can get smaller encodes with 10bit, yeah [21:20] note that the -crf values between 8bit and 10bit are not comparable. [21:20] right [21:20] are there any ffmpeg windows builds that have this? [21:21] maybe this: http://sourceforge.net/projects/ffmpeg-hi/ [21:21] uex1fi: well, that's silly, they're distributing a non-redistributable build [21:22] with fdk-aac in there, it's a license violation. [21:22] yay licensing? [21:22] ok [21:25] any significant reason not to use 10-bit? [21:26] there's basically no support for 10bit in any hardware decoders, and limited player support. You can't use it for mobile or any fixed-function hardware (bluray players, etc) [21:27] ok [21:30] all right, thanks [21:30] see you later [21:31] It proved to be a pretty big gain in the anime scene; the increased accuracy of 10bit meant fewer bit needed when using lots of reference frames, and the 10bit colour space made background gradients and stuff look nicer. [21:32] based Daiz, etc. [21:32] of course "big gain", when half the encoders started making multi-gigabyte files because they could :/ [21:33] yeah, that's interesting... [21:33] so really, 10-bit is better, but just not supported widely [21:33] ? [21:34] the drawbacks are that it's somewhat slower (i think? haven't tested this personally) and is not widely supported for playback, yeah [21:35] i don't thing there's any time when it'll give worse quality than 8bit, at any rate. [21:36] I seem to recall that it behaved better when you are doing multiple generations of lossy encoding, but dunno where I saw that. [21:40] wow, so many options to sift through [21:40] haha [21:44] see you later [21:45] JEEB llogan libfaac vs libfdk_aac ? what flags should i use? ffmpeg is now defaulting to libfaac (compiled from homebrew now, instead of binary) [21:45] fdk [21:46] faac is worse than fdk [21:46] how big of a difference are we talking? [21:49] does brew enable libfaac by default or did you add some sort of --with-faac? [21:49] if you also have fdk, then use -c:a libfdk_aac [21:50] brew enables libfaac by default [21:50] dumb [21:50] http://trac.ffmpeg.org/wiki/Encode/AAC [21:50] also, just listen to them to compare. preferrably ABX if you feel up to it [21:50] do both have the same 14kHz cut off? [21:51] i don't know [21:51] i don't use faac [21:51] i'll do an encode of both and checkout the spectrum analyzer [21:51] ok, but you should listen too [21:52] llogan aac page doesn't mention abx [21:54] i didn't provide it for abx [21:54] just general info about each encoder [21:54] examples, etc [22:06] cheers. [22:06] anyone wanna talk about lagarith? I figure this is faster than twitter. [22:07] Cracki: you should submit a bug report [22:07] include the ffmpeg command, the complete console output, and the sample file [22:07] of course you can search for an existing related report first, (not that i can think of an existing one ATM) [22:08] https://trac.ffmpeg.org/ [22:08] i've looked but google only turned up open bugs [22:08] so I figure either nobody bothers or I'm having a different issue [22:09] then submitting a bug report is the best way then [22:09] ugh now I have to make up another password. ok... [22:14] ok, about to create the ticket [22:14] thanks [22:18] if i encode an mpg to one video file and one audio file and then mux them together is it possible the end result won't be in sync? [22:21] Cracki: you forgot the complete ffmpeg console output [22:22] oh right, I'll attach that. thanks. [22:22] and please explain how it was created and what plays is correctly (as you did on twitter) [22:24] Cracki: and you may be asked to test ffmpeg from current git master since your build is about two months old [22:24] shit, how is it nov 18 already? [22:25] are nightly builds (zeranoe) ok? [22:25] yes [22:25] then i'll do just that [22:25] thanks. i'll be back in an hour if you have other questions [22:44] rsully, good question. in your place, I'd try to find videos where the audio or video have offsets (ffprobe can give machine readable output) [22:45] any examples of what i'm looking for? [22:46] Hello. I don't know if this is the proper channel to ask. I'm not an expert of "ffmpeg". I can't understand why, when I convert a video from a format into another, the quality gets worse. Am I doing something wrong? Thank you. [22:46] You're reencoding. [22:46] use -c copy [22:46] uh if you just do an ffprobe on a file, you'll see Duration: ..., start: 0.000000 etc [22:46] (@rsully) [22:47] Cracki ok i see that, so how should i compensate for that? my video input has start=0.04 and my audio file input has start=1 [22:47] francesco_, video compression is lossy because it saves bits. if you squeeze video too much, it loses information. [22:47] rsully: -itsoffset [22:48] Cracki, I'm trying to convert a FLV video into an AVI, to watch it with my DVD reader. I'm simply doing "ffmpeg -i myfile.flv myfile.avi". [22:49] yeah that's too little. ffmpeg will choose some default codec and a really low bit rate. [22:49] Cracki, what should I do? [22:49] francesco_, if you do ffprobe myfile.flv, what audio and video codecs does it say? [22:49] I was told that that was they way chosen by default to obtain the best quality. [22:49] c_14 is there a way to just say "figure it out"? [22:50] I didn't try. [22:50] Let me try. [22:50] rsully: maybe with async or vsync or one of those funky ones [22:50] It _should_ normally just work. [22:50] francesco_, definitely not. the defaults are supposed to look ugly, so people know to set appropriate values. [22:50] Cracki, I understand. [22:50] The x264 defaults are decent, but I don't think avi choses x264 [22:50] that too [22:52] *chooses [22:52] c_14 so by default it shoudl account for the start times? [22:52] Normally yes. [22:53] Cracki, I don't have ffprobe installed. I'm installing it. [22:53] if it's a random flv, i'd start with something like ffmpeg -i myfile.flv -c:a mp3 -b:a 64k -c:v mpeg4 -b:v 2000k myfile.avi [22:53] c_14 ok just wanted to check. i had encoded a bunch of mpg to mp4, and then realized i didn't use fdk_aac, so now i'm trying to take the audio from the mpg and the video (copy) from the mp4 so i don't waste time reencoding [22:55] I can't install ffprobe. I don't know why. :( [22:55] then skip ffprobe [22:56] I think it's a problem of dependencies. [22:56] when you do ffmpeg with just input and output file, what does it say on the terminal? [22:56] (it's reporting some information on the input) [22:57] It did when I tried to convert it. I didn't pay attention. It shows the conversion of the file. [22:57] before that [23:00] Cracki, which details do you need? [23:01] video and audio codec of the input file [23:01] Duration: 00:56:20.16, start: 0.000000, bitrate: 64 kb/s [23:01] Stream #0.0: Video: vp6f, yuv420p, 528x304, 25 tbr, 1k tbn, 1k tbc [23:01] Stream #0.1: Audio: mp3, 22050 Hz, stereo, s16, 64 kb/s [23:01] i don't need those. you can just convert the video to whatever. [23:01] ah yes [23:02] that's a nice small video [23:02] When I simply convert it into an AVI, the quality gets worse. [23:02] ffmpeg -i myfile.flv -c:a copy -c:v mpeg4 -b:v 1000k myfile.avi [23:02] that should do nicely [23:02] file might get bigger but it'll be as sharp as the original [23:02] Cracki, what does it mean, exactly? [23:03] -c:a is "codec:audio" and "copy the stream" [23:03] so the audio is copied without change. [23:03] -c:v asks for mpeg4 (think xvid) and -b:v asks for 1 Mbit/s which is plenty for that video resolution [23:04] I'll try so. Any idea of why I can't install ffprobe? [23:04] that's between you, your system packet manager and the support channel of your operating system ;) [23:05] either you have ffprobe already or they packaged it somewhere nobody will find it [23:06] Cracki, unrecognized option -c:v. [23:06] uh -vcodec [23:06] same for -acodec [23:07] also -b 1000k instead of -b:v, I figure [23:08] Ok. Old version of ffmpeg? [23:08] It seems it's doing fine, now. [23:08] Cracki any reason mpeg4 over x264? [23:08] dvd players might be too old to handle that [23:09] ah yeah [23:09] mpeg2 ok, mpeg4 if they're advertising "xvid/divx" [23:10] ya old versions of ffmpeg don't understand stream specifiers (:a and :v) [23:10] I think I need to update my OS with a new one, but I should free my HDD first of all the junk that I have. :) [23:11] Was just gonna let you know that that command worked like a charm earlier cut it down to where they show up as they were supposed to thanks [23:11] the subs i mean [23:18] Cracki, the quality is pretty much the same. The original one seems more "bright", though. I think I have to work on the options that you showed me, and that I didn't know. Further suggestion? [23:19] nothing so far. the brightness can be related to different level specifications. there's 0-255 and 16-235. [23:19] if you find anything on this, try it. [23:20] some player software can misinterpret levels too [23:20] Cracki, perhaps I'm confusing brightness with more definition. :( [23:21] hmhm. if the image looks faded and blacks/whites aren't black/white, then that's it. [23:21] if the picture looks blurry, that might be because that video has a really odd resolution. [23:21] you can try to have ffmpeg rescale the video. -vf scale=... [23:21] depends on what your dvd player works best with, or what display is attached [23:22] Where could I find further details about video formats. I'm not an expert. Just to know what I'm talking about. [23:23] uh I'm not aware of a decent article on that. [23:23] DVDs have some defined formats (resolution, aspect ratio, frame rate, ...) [23:24] Ok. Thank you very much. [23:26] llogan, I'm not sure how to feel about the last comment on my ticket. yes, it's variable frame rate, insofar as null frames should have been produced by lagarith. [23:49] Cracki: i didn't really get to look at it much yet. if ffplay or ffmpeg output looks different than the lagarith native decoder then to me that's a ffmpeg bug [23:50] ...usually [23:50] the person on the ticket referenced some other tickets that look like the behavior is known and patches exist [23:51] but hey, avconv does even crazier stuff, so ffmpeg is ahead ;) [23:52] cehoyos triages most tickets, and knows most about the existance of other tickets, but his comments are his opinions [23:52] (again i didn't really look yet) [23:52] i understand :) [23:54] I understand "first level" work perfectly well *g* [00:00] --- Wed Nov 19 2014 From burek021 at gmail.com Thu Nov 20 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Thu, 20 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141119 Message-ID: <20141120010502.0B9D618A02F1@apolo.teamnet.rs> [00:03] When I try to concat two .ts files the resulting output has horrible compression artifacts, what should I do differently? [00:03] What are you doing? [00:04] Note, if you want to concat ts files you can usually just cat *ts > new.ts [02:12] I'm getting some delay when I'm starting from an rtsp stream, is it possible to control the load buffer [02:46] Need a little help. Using the FFMPEG API, is it possible to chain formats together using the AVInputFormat.next pointer? [02:47] (Looking to use avformat_open_input but only on a limited subset of the file types that FFMPEG supports.) [02:52] (Or is it more prudent to register only the subset of codecs?) [04:15] is there an easy way to split a video file into two parts with ffmpeg? so if I have a 2 hours video make two files, 1 hour each [04:16] I'd like to be able to say: split into two chunks, rather than specify how long each chunk will be [04:26] I'm trying to transcode a live stream but there seems to be a 30 second startup delay, how can I reduce it? [04:50] is there a good tutorial on how to use FFMPEG? [05:18] Techstar: http://ffmpeg.org/ffmpeg.html https://trac.ffmpeg.org/wiki/Encode/H.264 [10:25] hey people [10:28] I have a problem with ffmpeg and libopus here. I have an .opus file (got on internet) with 1 audio and 2 video streams (cover arts), but when I try to do the same kind of thing by myself, just with stream mapping, ffmpeg throw an error. [10:29] Does LibOpus in ffmpeg can handle adding video streams to .opus files ? [10:33] http://pastebin.com/raw.php?i=zVyQ12G0 [10:37] Pazns: I think you want .webm [10:38] But how does the original file exist at all ? [10:38] Seems enough legit for my music player to handle it properly, and display the cover art from the video stream. [10:39] ffmpeg's output says it's ogg [10:39] aww, didn't that [10:39] see* [10:40] it's obvious now :| [10:41] Well, thanks, then. [11:31] Hm, it's me again. [11:32] http://pastebin.com/raw.php?i=w12XVYHs Doesn't work better than previously. What am I missing this time ? [11:32] The documentation is not very clear. [11:40] try ffmpeg -i input -map 0:a -c copy -f ogg test.opus [11:41] But then the video stream (holding a cover art) is not there. [11:42] humor me [11:44] Well, tried it and there is just one (audio) stream. Like requested in the command. [11:45] ok, I just wanted to see if the opus stream was the issue. [11:58] ogg only likes theora video [11:59] ie, you'll have to use -c:v libtheora [11:59] How the original file has those two video streams at first ? [12:00] Probably the muxer producing incorrect output. [12:01] Unless the ogg spec was expanded and that hasn't been included in FFmpeg yet. [12:04] It could always be that FFmpeg just never implemented that part of the spec/feature. I just know that theora works. [12:04] It could be a weird variety of ogm ? [12:06] Anyway, now I just need to figure how to dump this into a theora stream without getting just a black screen x) [12:06] Thanks for the help. [12:07] Mhm, it seems you can use FLAC's METADATA_BLOCK_PICTURE field as a vorbis comment. [12:10] And a base64 encoded image, if I understood. [12:11] Just a metadata with this name and the b64 value will do the trick ? [12:12] No, it needs some extra info as well. [12:12] There seems to be a bash script here that you can look at: https://github.com/acabal/scripts/blob/master/ogg-cover-art [12:13] A bash script is a nope for me, I'm using ffmpeg windows build. [12:14] Sure, but you can see what you have to add to the picture etc. [12:15] You could always open a feature request on trac for FLAC/ogg/opus cover art support. [12:18] In fact I found one old trac issue about that, I guess. [12:18] https://trac.ffmpeg.org/ticket/2655 [12:18] i'm not sure if this is really related to my case [12:19] >vorbiscomment: Add support for embedded cover ar [12:19] It seems relevant, let me look at the change. [12:20] Mhm, it looks like it's only in the ogg demuxer. [12:23] The feature is "read-only", then ? [12:23] Currently, yes. [12:25] Then need to play with metadata tags. Finally, no cover art for me and more free time ! [12:28] hello [12:29] im converting bunch of flv video files to mp4 but the audio is only 4kb/s. any idea how to improve that? [12:29] -b:a 9001k [12:30] https://trac.ffmpeg.org/wiki/Encode/AAC [12:32] c_14: 13 kb/s :( [12:32] c_14: i have the same video transcoded by zencoder and it has 47kb/s [12:32] Pastebin your current commandline and output please. [12:33] c_14: http://pastebin.com/THPA6xJS [12:34] Output? [12:34] c_14: of ffmpeg -i ? [12:34] Of the command you just pasted. [12:36] c_14: http://pastebin.com/be5PDfyP [12:36] c_14: first one is the original flv file, second one is the file i got transocded by zencoder, aka that's what quality i want ,and the third is the results im getting. [12:39] Try with -b:a 47k -profile:a aac_he_v2 [12:39] And -ac 2 [12:40] Or with aac_he if you want to keep it mono [12:40] But then you should use -b:a 48k [12:59] c_14: i got invalid profile for aac_he [12:59] and for aac_he_v2 too [13:00] Aah, eeh, it seems only libfdk_aac supports those profiles... [13:01] Unknown encoder 'libfdk_aac' [13:01] Yep, you'd have to compile from source to get that one. [13:01] Though I'm not sure why the default aac encoder is ignoring the bitrate you give it. [13:02] Did you ever try with just -b:a 47k ? [13:02] i got 24kb/s with this: ffmpeg -i ET2GYGBFW4.flv -q:v 0 -profile:v baseline -r 25 -vcodec libx264 -ac 2 -b:a 47k test.mp4 [13:03] Well, it's almost twice as high as it was earlier... [13:03] Let's try 49? [13:03] 49k that is [13:03] but that's for stereo, i'd like mono [13:03] i think it's twice as high because -ac 2 [13:04] Aah, didn't see that. [13:04] When you get rid of it it's back at 13k? [13:04] ok, with mono and 49k it's 13kb/s [13:04] yeah :( [13:06] try with -c:a aac -strict -2 -b:v 47k [13:06] Too many bits per frame requested [13:08] I'd try compiling with libfdk_aac and then try with the aac_he profile. [13:08] I'm not sure what's going wrong. [13:12] You can always encode the audio with a different encoder and then just mux it back in with ffmpeg. [13:47] Hi [13:47] I have question... [13:48] 42 [13:48] what is best easy way for build ffmpeg windows 32bit version? [13:48] I just want add libfdk-aac and faac on zeranoe build [13:49] I've never tried any, but there are https://trac.ffmpeg.org/wiki/CompilationGuide/MinGW https://trac.ffmpeg.org/wiki/CompilationGuide/CrossCompilingForWindows https://ffmpeg.org/platform.html#Windows and https://trac.ffmpeg.org/wiki/CompilationGuide/MSVC [13:49] msvc builds are slow though [13:49] and using libraries is a nightmare [13:49] plus, I think zeranoe has a guide in his forum [13:50] Thanks... but I tried on msys mingw... that was so terrible... some external modules get error or can't find module.. [13:50] especially fontconfig... [13:51] and librtmp... [13:51] too much get trouble when configure or build [13:53] No way for all modules enabled ffmpeg build without build external modules? [13:53] aabuild: http://ffmpeg.zeranoe.com/forum/viewforum.php?f=19&sid=c1f8fb1fd41ecffa605dc83975d4c259 <- there's a few more topics on zeranoes wiki here about building [13:55] I've got an m4v file. I'd like to convert it to both webm and ogg. However, all of my attempts yeild files with low quality. [13:55] How can I convert, but keep quality? [13:56] https://trac.ffmpeg.org/wiki/Encode/VP8 [13:56] https://trac.ffmpeg.org/wiki/TheoraVorbisEncodingGuide [13:57] c_14: this was my last attempt. But I was really just taking parameters from pages on the net. [13:57] https://pastee.org/c6njk [13:58] This method working on mingw windows? or only on linux(ubuntu... etc..) - https://github.com/qyot27/mpv/blob/extra-new/DOCS/crosscompile-mingw-tedious.txt [13:58] c_14: I'm checking outthat encoding guide now... [13:58] bobdobbs: first, you're not using ffmpeg from FFmpeg, you're using libav [13:58] Either build from source, grab the static build or if you're on debian grab ffmpeg from unstable [13:59] oh dang. I just installed it from the default repos [13:59] Or ask in #libav for help [14:02] Though the gusari builds haven't been updated since july, so if you're going to go static use the johnvansickle builds [14:03] I think I might actually retire for the night, and come back to the task in the morning [14:03] still... this evening I have learnt something! [14:04] not all ffmpeg is ffmpeg! [14:04] Action: bobdobbs shakes fist at ubuntu [14:05] ffmpeg from FFmpeg is slowly being merged back into debian, but it'll take a while before it goes stable and linking programs against it will still be a pain [14:05] hmmm [14:18] many encoding options difference with ffmpeg and avconv? [16:08] aabuild: not really [18:10] Hi all, I have a question about getting ffmpeg to compile into a binary with libraries baked in. I have successfully gotten ffmpeg to build with shared libraries (after having to add a path to/etc/ld.so.conf.d/ and then ldconfig-ing) but I would rather not have to make changes to the ld.so cache. Here is the input for my build: http://pastebin.co [18:10] m/jtknCqWg. How might I set the configure options to include the libs in the executable? Thanks! [18:11] I have a video at 352x576 [SAR 24:11 DAR 4:3], how do I change the DAR to 16:9 without actually scaling, ie. leaving the resolution at 352x576? [18:12] PovAddict: use the 'setsar' or 'setdar' filter [18:12] would -aspect work? (I *just* found it in the manpage) [18:12] Not sure, it probably does the same thing as the setdar filter. [18:13] (except it looks like you can use -aspect with -codec copy; that doesn't work with the filter) [20:46] hey guys... is ffmpeg supposed to store a cookie that is set via header on an initial request, and replay it on subsequent requests? [20:46] the source says yes [20:47] google says no [20:47] I'm talking about 2.3.4 [20:47] Trust the source? [20:47] c_14: well... its not working :S [20:47] I call http://example/playlist.m3u8 [20:47] it sets a cookie [20:48] playlist.m3u8 asks for chunlist.m3u8 [20:48] and when ffmpeg's issues a GET [20:48] the cookie is not there anymore [20:49] There was a patch about something with HTTP and auth that was applied on the 15th. [20:49] Might want to try with git. [20:50] let me try the the last precompiled binary [20:53] c_14: nope [20:53] doesn't work [20:54] Where in the source did you find it saying yes? [20:56] source/doku/wherever [21:47] http://pastebin.com/GTSVVg67 [21:47] So I have this command [21:48] pretty simple, should just concat these three files [21:48] but it's failing [21:48] throwing an error on an audio stream-- but all three inputs have audio. [21:49] Can you ffprobe all 3 files and pastebin the result? [21:49] also, can you get rid of the double quotes around [v] and [a] [21:51] also the trailing : after a=1. a pastebin of the console output would help too [21:54] http://pastebin.com/EKp2mKCq [21:54] so it looks like i don't have audio on the first and last chunk [21:55] So it would seem. [21:56] http://pastebin.com/mFd0V0tm [21:57] Eh, no wonder you don't have any audio? [21:57] You used -an [21:58] you can use "-f lavfi -i aevalsrc=0|0:s=48000" to make a silent audio source if that's what you want (or add it to your filtergraph). [21:59] How would I pass through the audio stream the original file has [21:59] you can include it as an other input and reference it in your filtergraph [22:00] or remux the video and original audio into a new file [22:17] Got it working! thanks much for sending me down the right path. [22:44] Anyone here have experience with trimming a video stream as it comes in? (Being recorded) [22:45] "trimming" ? [22:46] Well what I want to do, is have a video stream record from webcam, but erase anything older then 30 seconds, unless you manually hit a record button. This way if you see something you can press record and start to save from 30 seconds ago until your done, but it will erase anything after 30 so you dont fill up your hdd [22:47] hmm, so basically recording to a circular buffer [22:47] You can probably accomplish something with the segment muxer [22:47] you could do something like that by recording to e.g. 10s segments, and deleting old segments [22:47] You just need something that will save the segments when you "pressrecord" [22:47] @kepstin, but what happens if you say press record at 5 second, would you have to combine the 5 second clip and the 10 second clip and grab the newest 10 seconds? [22:48] zoch: the only thing pressing 'record' does is stop deleting old segments [22:49] so the history will only be approximately 30 seconds long, it'll actually be "at least 30, but no more than 30 + the length of one segment" [22:49] hmm [22:50] so would you ever end up with more or less then 30 seconds before you hit record? (trying to wrap my head around it) [22:50] you will usually end up with slightly more; never less (assumign you implement the deletion correctly) [22:50] I think that would work actually, now should I be researching circular buffer? or segment muxer like c_14 said [22:51] segment muxer would be how to do it, yeah [22:51] cool cool. That was less complicated then I thought :) [22:53] I think the segment muxer can run in two modes; one where it keeps all segments, one where it automatically deletes them; you probably want to have the segment muxer keep all segments, and have the external tool which handles the record button do the deletion or archiving of segments. [22:53] yeah that was what I was thinking [22:54] Do you know if when running ffmpeg in something other then linux (android using external libraries) if any features would be limited? [22:54] I dunno if ffmpeg can capture directly from the cameras on android devices, i don't think so? [22:55] they're not v4l, you have to go through some funky android interface. and you'd probably want to use a hardware video encoder on the device rather than software encoding in ffmpeg [22:55] (you could still use ffmpeg to do segmenting in that case, with -codec copy) [22:55] well I believe it can be used like a webcam is used on a computer now. android wraps it in a video stream [22:56] does codec copy just copy a file in realtime as it is made? [22:56] zoch: no; codec copy means "don't re-encode the video and audio" [22:56] oh [22:56] so if the android camera interface gives you h264, you could pipe that to ffmpeg and have ffmpeg write it to split files [22:57] then yeah i would want that [22:57] exactly [22:57] cool thanks alot man. Searching google was not very helpful :D [22:58] zoch, Are you using the native android mediarecorder? [22:59] The Android native camera interface gives you raw YUV frames. [22:59] I have not decided yet, I wanted to use xamarin framework so I could use ffmpeg on ios and android, but I have not gotten into sepcifics yet, just making sure i can do it first [22:59] @thedracle, would that help me? [22:59] if you have sufficient ram on the device, you could get more complicated by storing the 30s buffer in ram and only writing to disk when you hit 'record'; saves writes on the (generally low quality) flash on the phone :) [23:00] zoch, Well, MediaRecorder can create MP4 files directly on disk using the underlying hardware encoder of the system. [23:00] Yeah, and encoded video is a problem for circular buffering in this way, because you need to make sure you synch your buffer up by iFrame. [23:01] So you don't cut off the stream with a bunch of P-Frames based on a non-existent iFrame. [23:01] @Dracle, yeah but I do not believe I can do that 30 second segmenting with the medirecorder (could be wrong) [23:01] The right format for what you're trying to do would be something like MPEGTS, and setting a really short GOP size. [23:01] Interesting [23:01] MPEGTS is made for streaming, and you can easily truncate it in the way you're indicating. [23:02] I'd use libav. [23:02] and thats for android right? [23:02] I will look into it, Gotta go to work, thanks guys! [23:02] with smartphones tho, you *really* want to use the hardware encoder if you can, just due to battery life/heat issues :/ [23:02] It's just a C library. [23:03] You can use the Android NDK to build native C stuff. [23:03] zoch, If you want, I've done a lot of programming with libAV and Android, let me know if you have questions. [00:00] --- Thu Nov 20 2014 From burek021 at gmail.com Thu Nov 20 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Thu, 20 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141119 Message-ID: <20141120010503.1FA2918A02F2@apolo.teamnet.rs> [00:51] ffmpeg.git 03dedicatedbroadcastsolutions 07master:873dac50f5c8: avformat/mpegtsenc: change the min value for mpegts_start_pid to allow ATSC PIDs [01:13] nice author name [01:23] .biz [01:33] michaelni: could you pass the output from the test you send in mailing list, it working for me [01:46] ok nvm, i see [05:16] ffmpeg.git 03Michael Niedermayer 07master:8b43b0e8b6c7: ffmpeg: fix variable type for end char [05:16] ffmpeg.git 03Michael Niedermayer 07master:9f6d48d696d6: ffmpeg: better CFR frame duplication selection [09:54] https://people.xiph.org/~jm/daala/pvq_demo/ [11:19] ffmpeg.git 03Stefano Sabatini 07master:484d42a0977e: lavf/concatdec: add timestamp log [13:28] ffmpeg.git 03Michael Niedermayer 07master:61fc1cbfbd82: ffmpeg: fix printed timestamp for droped frames [13:49] ffmpeg.git 03Michael Niedermayer 07master:beec818d99e5: avcodec/truemotion2: Use av_freep() to avoid leaving stale pointers in memory [13:49] ffmpeg.git 03Michael Niedermayer 07master:b9792afad1c6: avcodec/tta: Use av_freep() to avoid leaving stale pointers in memory [13:49] ffmpeg.git 03Michael Niedermayer 07master:4ffec6d93315: avcodec/twinvq: Use av_freep() to avoid leaving stale pointers in memory [16:51] Hi [16:52] Somebody can help me? [18:26] I wonder what that guy wanted. [18:29] he was probaly drowning... too bad [18:30] Should've joined #lifeguard, then. [18:30] Or #baywatch. [18:31] lol [18:40] ffmpeg.git 03Michael Niedermayer 07master:c5092025901b: ffmpeg: Use input packet duration in vfr/cfr code if available and valid [18:40] ffmpeg.git 03Michael Niedermayer 07master:4e20e9492146: ffmpeg: Check duration for overlap and clip in fps cfr/vfr code [19:55] ffmpeg.git 03Cl?ment BSsch 07master:de8cd93a05d1: doc/writing_filters: use ffmpeg.org instead of wikimedia.org for lena image [20:24] ffmpeg.git 03Michael Niedermayer 07master:33bc81e43741: ffmpeg: skip duration cliping for passthrough & drop modes [20:24] ffmpeg.git 03Michael Niedermayer 07master:be4485648330: avformat/avienc: factor frame skip code out [20:24] ffmpeg.git 03Michael Niedermayer 07master:660a8b43abdb: avformat/avienc: write last frame duration [20:29] ffmpeg.git 03Martin Storsj? 07master:f918b8a2933a: hdsenc: Use the right filename in an error message [20:29] ffmpeg.git 03Michael Niedermayer 07master:f3dcabef3324: Merge commit 'f918b8a2933a65020cbe490ec637d5485c11a692' [20:35] ffmpeg.git 03Martin Storsj? 07master:7fd10f66b722: hdsenc: Clear the previous codec tag when setting up the chained muxer [20:35] ffmpeg.git 03Michael Niedermayer 07master:fbb6de2ad774: Merge commit '7fd10f66b722eccc2ada9128766d002f6d751f79' [00:00] --- Thu Nov 20 2014 From burek021 at gmail.com Fri Nov 21 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Fri, 21 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141120 Message-ID: <20141121010502.7371B18A0239@apolo.teamnet.rs> [00:22] ffmpeg.git 03Michael Niedermayer 07master:0dba982bb4f7: avformat/dtsdec: dts_probe: check reserved bit, check lfe, check sr_code similarity [01:23] ffmpeg.git 03Michael Niedermayer 07master:4388e78a0f02: swscale/x86/rgb2rgb_template: handle the first 2 lines with C in rgb24toyv12_*() [01:30] michaelni: I would like to backport three patches to 2.4 that do not fix regressions: [01:31] The two patches from Peter Ross that export Phantom Cine Metadata about the colours and your swscale patch that fixes the crash when scaling bayer. [01:31] Any objections? [01:36] ffmpeg.git 03Michael Niedermayer 07release/2.4:e5f5df37c8e1: avformat/dtsdec: dts_probe: check reserved bit, check lfe, check sr_code similarity [01:54] cehoyos, should be ok if they apply without conflicts and fate passes [01:54] The Bayer patch also requires to backport fba89461 which contains a minor version bump. [01:54] version bumps are not ok in releases [01:54] If it is ok to backport this patch (it works, fate is running atm), what should I do with the version bump? [01:55] Is simply not bumping ok? [01:55] fate passes [01:56] yes not bumping is an option [01:56] why is there a version bump? [01:56] Sio is adding fba89461 to 2.4 ok in your opinion? [01:56] The other three patches look safe to me but I cannot comment on this one. [01:56] An internal symbol was added. [01:57] so, nothing that breaks abi/api. [01:58] Sorry, it was a micro bump, not a minor bump [01:58] I misread the conflict. [02:00] fba89461 is ok if it applies cleanly (except the omited version change) and it works [02:01] It applies cleanly, I will test now. [02:04] ffmpeg.git 03Michael Niedermayer 07release/2.4:633a2a082fc1: swscale: support internal scaler cascades [02:04] Tested and pushed, thank you! [02:04] ffmpeg.git 03Michael Niedermayer 07release/2.4:222236317bf4: swscale/utils: support bayer input + scaling, and bayer input + any supported output [02:04] ffmpeg.git 03Peter Ross 07release/2.4:e386241d54a0: cinedec: report white balance gain coefficients using metadata [02:04] ffmpeg.git 03Peter Ross 07release/2.4:057ee3592418: avfilter/vf_lut: gammaval709() [02:06] Good night! [03:17] michaelni: you didn't add a changelog entry about the webp muxer [03:17] considering that unlike image2 it supports creating animated files i'd say it's worth a mention [04:32] ffmpeg.git 03Michael Niedermayer 07master:c661601f4583: Changelog: add WebP muxer [04:32] jamrial, done [07:45] haha, aac patch v9 [07:45] it will never end @_@ [07:51] ooh boy [07:51] meanwhile: anyone around who knows a fair bit about movenc.c? [07:51] I'm currently attempting to get dashenc to output segments starting from the _middle_ of a stream [07:53] as in, start at the beginning, client plays a few segments, then the user seeks and jumps to, say, 2:30; restart the transcoder with `-copyts -start_at_zero -ss 150` and start writing new segments at the offset, using the same MPD [07:54] (I'm using custom-generated MPDs instead of the ones dashenc.c writes, so the duration is always set to the full final duration of the output file) [07:54] current problem is that the output segments get shifted so their timestamps start at zero, despite the -copyts [07:54] so I'm playing around to try to add an option to not do that [07:55] (Fun? Fun.) [08:09] I ended up hitting the "Track %d starts with a nonzero dts" warning after a bit of screwing around [09:09] hello [11:34] ffmpeg.git 03jessejiang 07master:29d208d5d47b: avutil/arm/float_dsp_init_vfp: replace restrict by av_restrict [12:10] ffmpeg.git 03Carl Eugen Hoyos 07master:25ccf5df723c: lavf/mux: Always call write_trailer() from av_write_trailer() to avoid a leak. [12:10] ffmpeg.git 03Michael Niedermayer 07master:9266eb0c620d: Merge remote-tracking branch 'cehoyos/master' [12:22] ffmpeg.git 03Matthew Oliver 07master:e39f8fad321f: configure: Prevent icl being incorrectly detected as msvc. [14:01] ANyone knows about vdowave3 codec? [14:14] hello, I've got a problem [14:14] me at ubuntu:~/ffmpeg-2.4.3/doc/examples$ ./decoding_encoding h264 [14:14] Encode video file test.h264 [14:14] Codec not found [14:14] I have already install x264, but still get this error [14:15] help! [14:15] anyone here? [14:40] lookatmeyou: what does ffmpeg say when you try to use libx264? [14:43] later, I find I havn't build ffmpeg with --enable-x264 [14:43] thank you [15:17] something make ffmpeg mighty unhappy on windows/msvc.... [15:17] it crashes instandtly [15:17] as of today [15:17] ffmpeg.git 03Michael Niedermayer 07master:88b4c1a7316c: avcodec/mpeg12dec: Print error/warning messages on issues in mpeg1_decode_sequence() [15:17] ffmpeg.git 03Michael Niedermayer 07master:0f8908aa1b66: avcodec/mpeg12dec: do not fail on invalid frame_rate_index [15:20] loooks like matt broke icl [15:23] Daemon404: do you have more input than that? [15:25] benoit-, i sent a link to fate on the ML [15:25] Daemon404: k [16:06] running 64bit icl test [16:13] has ffmpeg moved to tim's nodejs fate btw? [17:13] ffmpeg.git 03Michael Niedermayer 07master:1852b2a0f497: avcodec/mpeg12dec: Use more specific error codes [17:13] ffmpeg.git 03Michael Niedermayer 07master:bab11fe7bf4b: avcodec/mpeg12dec: forward error codes [17:30] urg [17:30] icl instances is actually hanging my fate box [17:31] https://www.dropbox.com/s/o7omv923gsiztez/cil.png?dl=0 [17:31] lots of that [17:31] you cant disable that dialogue apparently [17:32] Good job fixing icl in that one commit huh [18:09] j-b: [FFmpeg-devel] [PATCH 3/6] Add AV_PIX_FMT_NV12T. [18:09] what [18:09] the [18:09] fuck [18:14] sounds like it should be a codec [18:29] wm4 : can be both... [18:29] raw codec and colorspace [18:30] i personally prefer not to call different methods of packing bits/bytes a "Colorspace" [18:36] WHY? [18:36] its not a colorspace [18:36] colorspace is CIElab etc.. [18:36] or RGB [18:36] and SRGB [18:38] j-b: to make something work with v4l I suppose [18:38] but I thought v4l has a transparent conversion lib for such things [18:38] which ffmpeg can use [18:39] vf_v4l2_m2m.c [18:39] maybe because of this [18:40] What's the use of the filter without the decoder [18:45] I remember they talked about memory to memory devices at kernel summit [18:45] can't remember what it does [19:07] kierank: itn't that a virtual device? aka you have a device that gives you video frames generated by another program? [19:07] dunno [19:08] Daemon404 : sounds like you want to bikeshed over a name :P [19:09] erm no [20:26] ffmpeg.git 03Matthew Oliver 07master:17b7f99d810d: configure: disable strip when using icl. [22:39] ffmpeg.git 03Michael Niedermayer 07master:8bce5c8e74b5: avcodec/ac3dec: Use avpriv_float_dsp_alloc() [22:39] ffmpeg.git 03Michael Niedermayer 07master:9018bd11979f: avcodec/atrac1: Use avpriv_float_dsp_alloc() [23:01] ffmpeg.git 03Michael Niedermayer 07master:b054054c9b8f: avcodec/dcadec: Use avpriv_float_dsp_alloc() [23:01] ffmpeg.git 03Michael Niedermayer 07master:21ded9ce67d8: avcodec/imc: Use avpriv_float_dsp_alloc() [23:34] is ffplay broken or what? [23:35] can't play mp3s anymore, is it on my side only? [23:35] oh well, just this pulse crap again probably [23:36] (only audio looks broken) [23:54] works here [00:00] --- Fri Nov 21 2014 From burek021 at gmail.com Fri Nov 21 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Fri, 21 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141120 Message-ID: <20141121010501.6692718A00B8@apolo.teamnet.rs> [00:10] hi, is there any anti-flicker in the ffmpeg? thank you [00:35] b-p: can you elaborate? [00:40] llogan what? [00:40] i need something like the vlc players anti flicker [00:50] b-p: no dedicated antiflicker filter exists. perhaps you could port the one in VLC to FFmpeg [00:51] if you are unable, then you could submit a feature resuest at the bug tracker, and evn post a bounty if you want to offer some monetary incentive [00:51] *request [02:17] Hi guys, what would be the cmake equivalent of "./configure --prefix=$TARGET_DIR --enable-static --disable-shared" Trying to compile x265 for static linking into ffmpeg. Thanks! [02:19] cmake -DCMAKE_INSTALL_PREFIX=$TARGET_DIR . [02:19] I think it builds static by default. [02:21] so presumably -DBUILD_SHARED_LIBS:BOOL=OFF and -DCMAKE_C_CREATE_STATIC_LIBRARY:BOOL=ON are fruitless? [02:22] It might do something. [02:22] i dislike cmake [02:24] I hate it so much. Would kill to just have a ./configure like every other codec's sources [02:25] Hmm, actually it looks like shared might be default. [02:25] So you might want to keep -DBUILD_SHARED_LIBS:BOOL=OFF [02:26] kk will do. Would the lib dir automagically be set to PREFIX/lib, or does it need to be explicitly stated? [02:26] Should be [02:26] great! thanks! [02:27] I assume if shared needs to be disabled it probably wants to be explicitly told to be static? Via: -DCMAKE_C_CREATE_STATIC_LIBRARY:BOOL=ON ? [02:31] I can't find that string anywhere in the cmake folder... [02:31] Or anywhere in the repo for that matter. [02:32] tbh, all I did was use the ./make-Makefiles.bash script, hit generate and started browsing the CMakeCache.txt file [02:33] hmm good idea, thanks [02:33] There seems to be an x265-static target in the Makefile though. [02:34] Worst case you could probably just call make x265-static [02:35] what would you guess the output of that would be? Just the .a ? [02:35] Following the fun cmake tree down to CMakeFiles/x265-static.dir/build.make, it seems to generate the .a [02:37] And in my current conf, the cli.dir/build.make links against libx265.a [02:37] Which looks like it's called from the cli target. [02:55] Welp, ffmpeg is having difficulties linking x265 it seems. Full console input / output: http://pastebin.com/sqnjJ3Gh config.log: http://pastebin.com/8T7XQC7s [02:55] Appreciate the help! [03:01] What does pkg-config --exists --print-errors x265 return? [03:02] Package x265 was not found in the pkg-config search path. [03:02] Perhaps you should add the directory containing `x265.pc' [03:02] to the PKG_CONFIG_PATH environment variable [03:02] No package 'x265' found [03:02] yep, do that [03:03] export PKG_CONFIG_PATH=wherever_that_was_again [03:03] x265.pc exists in my $TARGET/lib/pkgconfig [03:03] yes, then that [03:03] along with .pc's for many other codecs. Why is it that only x265 is yelling at me? [03:04] configure doesn't use pkg_config for everything [03:04] s/_/- [03:07] What is the best way to add this to the path in a bash script? [03:08] export PKG_CONFIG_PATH=$TARGET/lib/pkgconfig [03:08] Or rather [03:08] export PKG_CONFIG_PATH=$TARGET/lib/pkgconfig:$PKG_CONFIG_PATH [03:08] ahh exactly that thanks! [03:20] since i?m very bad with video codecs, video container formats, audio codecs, subtitle formats and combining them all: is there a script "here ffmpeg, look at this video file A, recode video file B like A, GO!" ? [03:23] Hey c_14, for some reason even with that PKG_CONFIG addendum it still cant find x265. I tried calling pkg-config -exists in my script but I see no output in the console. Tried capturing it to a variable and echoing that but still nada. [03:23] It won't print anything if it works. [03:23] check $? [03:24] returns 0 [03:24] then it found it [03:24] But why isn't ffmpeg... [03:24] Can you pb again? [03:24] yep [03:24] Schnabeltierchen: there might be some funky gui that can do something similar [03:26] i would prefer cmdline, because synology nas and stuff [03:27] mhm but k, any gui you would suggest for a remote running ffmpeg? [03:27] None I know of. [03:27] I tend to just use the commandline. [03:28] input-output: http://pastebin.com/FM6yLwC5 config.log: http://pastebin.com/ZAb2RMvg [03:31] Schnabeltierchen, It isn't nearly a one-click solution as you may like, but just running ffmpeg -i on whatever input file you have will tell you all you need to know. You could probably write a script to parse that output and do it all based on that. Probably more work than just practicing your codecs, subs, and containers though :) [03:32] compstomp: just making sure, but this is git head, right? [03:32] input file being the file whose attributes you want to emulate, that is. [03:32] head of ffmpeg and x265 [03:34] once given parameters would be another solution, cause i tested 1080p video, with diferent audio streams and subtitles it worked... now i need to recode all my videos like this sample video... [03:38] Unabashed self-pitch here. Here's a script I wrote a little while back that helps you recursively convert an entire directory with ffmpeg: https://github.com/srwareham/Linux-Scripts/blob/master/ffmpeg-scripts/ffmpegDir.bash Would hardly take any retooling to modify to your specifications I assume [03:40] mhm this seems just like the thing i?m searching for [03:41] mhm only the subtitle thingy :P [03:42] compstomp: just making sure, but you have x265.h in $TARGET/include ? [03:42] and x265.a is in $TARGET/lib ? [03:43] in $TARGET/lib i have libx265.a but no x265.a [03:44] I also have x265.pc in $TARGET/lib/pkgconfig [03:44] eh, right. and the header file? [03:46] yep. In $TARGET/include i have x265_config.h and x265.h [03:46] Hmm, can you try adding --pkg-config-flags=--static to the ffmpeg configure line? [03:47] Trying now. [03:51] Holy _explicative_! it appears to be compiling!! [03:51] I just love obscure options. [03:53] The more power you have, the more ways you have to shoot yourself in the foot lol. Will post back if I can actually convert something. POS old laptop this is running on so likely not tonight hah [07:12] how can i rotate a mp4 video clip using ffmpeg? [10:00] hi, I have a list of URL of different kinds of audio stream I'd like to dump, is ffmpeg suited for the job? I've seen some rtsp:// with .ra, mp3 and wma among these [10:00] I'm looking for a "swiss army knife" of dumping tool, i.e. I'd like to just toss my url as argument and be done with it [11:06] Is it possible using the latest libavformat to upload encoded packets directly to an rtmp server URL ? For i.e setting AVFormatContext->filename to an rtmp URL ? [11:28] hello - i need some utility to process mp3 file - i want to remove certain segment from mp3 file - trimming - any good tool/approach for this? [14:23] hello, anyone help [14:23] me at ubuntu:~/ffmpeg-2.4.3/doc/examples$ ./decoding_encoding h264 [14:23] Encode video file test.h264 [14:23] Codec not found [14:23] I have already install x264, but still get this error [14:23] Did you build your ffmpeg with x264 support? [14:23] yes [14:25] ./configure --enable-x264 --enable-shared --enable-gpl [14:25] I build ffmpeg using this command [14:26] ./configure --enable-shared --enable-gpl --enable-libx264 [14:26] this, sorry [14:43] I havn't built ffmpeg with --enable-x264 [14:44] thank you very much [14:53] anyone know of any info about mimicking the mp3 encoding output of adobe audition with ffmpeg? i'm automating some editing that someone's been doing with audition, and am trying to get my output as close as possible [15:12] "Only VP8 or VP9 video and Vorbis or Opus audio and WebVTT subtitles are supported for WebM." [15:12] trying to convert flv to webm [15:12] ffmpeg -y -i files/ET2GYGBFW4.flv -vf scale=320:240 -q:v 0 -profile:v baseline -r 25 -vcodec libx264 -ac 2 -strict -2 -b:a 49k files/RN8N5WT37RO.webm [15:14] Well, that's because "Only VP8 or VP9 video and Vorbis or Opus audio and WebVTT subtitles are supported for WebM." [15:14] ok, removed vcodec, my bad [15:14] Encoder (codec vp8) not found for output stream #0:0 [15:14] im getting this now [15:14] have to install it guess [15:14] So your ffmpeg isn't compiled with vp8 encoding support [15:16] oh man that's a pain to compile on freakin' mac :D [15:16] why did i switch form linux, oh way [15:16] Use macports or something like that. [15:17] im creating a service thats going to run on linux anyway. ill just move it to test server [15:26] hello [15:29] Hello everyone, i am having problem with out of sync between audio and video. I am getting mpegts stream via udp and after decode and encode finally muxer it by rtmp. In general it is working as expected without problem, but after a while i am getting out of sync. [15:30] I am using the latest ffmpeg realese [15:31] ubuntu 64 bit server i am suing [15:31] using [15:31] anoy one has a idea? [15:48] hi i needed a bit of help to convert files to divx [15:48] to Divx Home Theater profile [15:49] how to set Macroblocks [15:49] VBV Buffer [15:50] VBV Bitrate [15:50] B-Frame [15:50] how i can set those in ffmpeg [15:54] @Athideus hi [15:54] hello there! [15:54] i needed a help with converting to divx [15:55] how to set Macroblocks, VBV Buffer, VBV Bitrate, B-Frame [15:55] i am looking for a way to distribute a list videos files to many different servers for encoding that then get sent back to the 'controller' server when completeed [15:55] wow i got no idea about that [15:55] i tried googling "ffmpeg cluster" but i havent found anything extremely usefull [15:56] DivX is just an mpeg4 encoder. There is no such thing as "encode as divx" [15:56] Athideus: I don't think there is a distributed ffmpeg... [15:56] And as DivX is a closed source commercial encoder, you're not going to be able to use it with ffmpeg. [15:57] h264 with x264 has better quality anyway, so that's not a huge loss [15:57] it dosent really need to be a distributed ffmpeg install, just a way to send the files to each encoder server, tell it to encode it, and then tell it to xfer back to the controller server [15:57] mount some samba/nfs share on all of them? [15:58] well is it possible to set those options in h264 or x264 [15:58] i wanted to play the file in my dvd [15:59] Check if it support h264, and if it does, use that [15:59] +s [15:59] BtbN: hrm thats a pretty good idea for the xfering portion, now i just have to figure how to get the controller server to distribute the workload [16:00] i plays it but i wanted get the file as close as possible to DivX Home Theater profile [16:00] Athideus, check out pdsh [16:00] so you want it to look worse? [16:01] Awesome, that looks like exactly what i am looking for. Thanks BtbN [16:01] it's not about worse it about that player works better with DivX Home Theater profile [16:02] Well, if you want to encode with DivX, you have to use their software. [16:03] so you saying h264 and x264 don't got this options Macroblocks, VBV Buffer, VBV Bitrate, B-Frame [16:03] Athideus, that's exactly how clusters usualy work. One control server which has the workdirs of all nodes mounted via nfs, which then distributes the data and launches the actual job via pdsh. [16:11] @BtbN so for h264 and x264 i can't set Macroblocks, VBV Buffer, VBV Bitrate, B-Frame? [16:12] david___: see: http://mewiki.project357.com/wiki/X264_Settings [16:12] and ffmpeg --help encoder=x264 [16:13] err... ffmpeg --help encoder=libx264 [16:13] oh didn't think about that sorry [16:13] thanks for the info [16:14] well i gtg now have a good day/night [16:45] is there a site i can find the default settings for q:v (1-60) ? also i notice -s XxY has an effect on quality where as -vf "scale=X:-1" does not, is that correct? also what does -1 do? [16:49] sor_: http://ffmpeg.org/ffmpeg-filters.html#scale-1 [16:49] sor_: the range for -q:v is 1 - 31 (for ffmpeg's encoders) [16:50] thanks [16:57] relaxed, so does -s XxY have an impact on quality where -vf "... does not? [16:59] sor_: the result of using -s (as an output option) and -vf scale=... should be exactly the same. [16:59] thanks [17:00] (in fact, if you read the docs, you'll find out that the -s option is just a shortcut that adds a scale filter to the end of the video filter chain) [17:05] kepstin-laptop, ya that's what i understood the problem is that my file size is considerably smaller with -s hmmmm thanks [17:17] Are there any sort of licensing restrictions on using FFmpeg's decoders in a commercial applications? Does licensing need to be purchased to use each decoder? [17:19] Zeranoe: are yout talking about copyright licensing or patent licensing? [17:19] as long as you follow the rules of the LGPL or whatever licenses apply, you don't have to pay anything. [17:20] as far as patent licenses go, talk to your lawyer [17:22] kepstin-laptop: Because patents are held by each respect codec technology developer? [17:25] Living in the U.S., I'm assuming if FFmpeg is compiled with H.264 decoder support, patent royalties would need to be paid to MPEG LA [17:26] Action: kepstin-laptop isn't a lawyer, and isn't familiar with your juristiction, so he has nothing further to say on the matter. [17:27] yes, mpeg la takes money for people using standard video [17:29] Well that's stupid. [17:29] patents are stupid, but people demand that some mythical small inventors should be protected. [17:30] hey [17:30] I'm using a FFmpeg Windows build helper [17:31] Zeranoe: anyways, the ffmpeg devs won't do anything to you if you violate patents, that's not their concern. But do follow the *copyright* license compliance checklist on http://ffmpeg.org/legal.html [17:31] And it successfully built FFmpeg with default options [17:31] Zeranoe: there is some stuff on patents on that page too [17:31] But it doesn't understand one option: --high-bitdepth=y [17:32] Why is this an unknown option? Is there a different option for specifying bit depth? [17:33] uex1fi: I dunno what you're talking about; that's not an ffmpeg build or runtime option. [17:33] uex1fi: is that libx264 option? [17:33] iive: yes, and libx265 [17:33] (ffmpeg's decoders/encoder and stuff like swscale is built with high bitrate supportted by default if applicable) [17:34] hmm [17:34] i wonder if this is an option specific to this particular build script? [17:35] okay, but bit *depth* cannot be changed after ffmpeg is built, right? [17:36] uex1fi: ffmpeg decoders/encoders that support multiple bit depths can have the bit depth selected at run time. [17:36] x264 cannot have its bit depth changed except at build time [17:36] Any idea why "--disable-decoders --enable-decoder=*264* --enable-decoder=*mp3*" is still enabling the mpeg2video decoder? [17:37] uex1fi: i can't find the string "high-bitdepth" in libx264/5 source [17:38] libx264 supports 8 and 10 [17:38] uex1fi: interestingly, if you build two copies of x264 (one 8bit, one 10bit), and dynamically link them to ffmpeg, you can switch them out at runtime by swapping which library is used. [17:38] okay... then i think this is something specific to this build script [17:38] kepstin-laptop: that is interesting [17:41] has anyone here used the windows build helper script recently? [17:42] I'm slightly confused by the instructions for enabling options like nonfree and bitdepth [17:49] Seriously... I cannot get the mpeg2video decoder to disable [17:53] looks like the only stuff that selects mpeg2video in the configure script is the hwaccel versions of the mpeg2 decoder. [17:54] this works for me, --disable-decoders --enable-decoder=*264* --enable-decoder=*mp3* --disable-decoder=mpeg* [17:55] Zeranoe: i suspect disabling hwaccel support might fix your issue? [17:55] Passing --disable-decoders and I still have "bmp h264 mpeg2video h263 hevc vc1" enabled. [17:55] relaxed: Why did you have to disable mpeg a 2nd time [17:56] Zeranoe: that list of codecs looks suspiciously like the list of codecs that have hwaccel decoders available. Please try disabling hwaccel. [17:57] (note that 'bmp' is pulled in if you have the 'gdigrab' screen capture enabled) [17:59] I would file a bug report. It makes sense that the decoders should be parsed first. [18:00] eh, it's just a quirk of the way the configure script works [18:00] you disable all the encoders, then enable the hwaccels, then the hwaccels need some decoders so they re-enable them [18:12] Hey, when I encode a regular .mp4 file as HEVC .mp4, it plays fine in VLC. But when I do DV-AVI to HEVC .mp4, there is no video, only audio. [18:12] Can anyone suggest a fix or a workaround? [18:13] okay [18:14] kepstin-laptop: Running with '--disable-encoders --disable-decoders --disable-hwaccels' and I still get decoders: bmp, h264, vc1, h263, and hevc enabled. [18:15] relaxed: the console history doesn't go all the way back [18:16] Zeranoe: well, that got rid of your mpeg2 :) [18:16] relaxed: http://pastebin.com/dZjntU0u [18:16] kepstin-laptop: lol, still not predictable results [18:17] Zeranoe: there's probably some filters or muxers/demuxers that you have enabled which are pulling in those codecs [18:17] ah, you need to disable the parsers [18:18] the h264, hevc, vc1 parsers all pull in the respective video codecs [18:18] Jeez [18:19] Zeranoe: you're not exactly doing something that's very common, most people want their builds with this stuff enabled :) [18:20] What do the parsers do [18:20] allow some metainfo to be read from streams (stuff like frame timing, size, type, etc?) to be extracted without running the full decoder. [18:20] I think? [18:24] lol, and of course --disable-parsers doesn't actually disable all parsers and the bmp decoder is still enabled [18:27] Zeranoe: this is a windows build? [18:28] like I said before, the bmp decoder is pulled in by the 'gdigrab' screen capture input device [18:29] (that's actually an interesting bit of code; basically the bitmap datastructures are used internally by gdi, so rather than reimplementing bmp, I just made the gdigrab device pass bitmap images to the bmp decoder) [18:34] kepstin-laptop: impressive [18:34] kepstin-laptop: would that be possible/beneficial with other decoders? [18:57] Can anyone explain where all the codecs are coming from when they aren't listed in -encoders or -decoders http://paste.ubuntu.com/9127939/ [18:58] As in, how are the codecs still included if they do not have decoder or encoder support? [19:01] hello, im doing some experiments with FFServer, and i see that one of the formats for streams is HLS (Http Live Streaming), however i cant really find a lot of documentation or exmaples on FFServer options for that format [19:01] is FFServer supposed to support realtime HLS segmenting and delivering ? or am i misunderstanding this? [19:02] Zeranoe: inside the configure script, there's dependency logic where if e.g. a filter or protocol or anything requires that a codec is enabled, it can list it in a special "select" list [19:02] then if you enable that filter or whatever, it'll automatically also (recursively) enable everything in the select list too [19:05] you can trace this backwards by just looking through the configure script, searching for encoder names; e.g. https://bpaste.net/show/6c863abeb256 [19:06] hmm. In theory, I guess you could use gdigrab without the bmp decoder; you could for example use -c:v copy and save the screen capture to (a series of) bmp files. [19:07] but that sounds silly :) [19:10] i have 2 cores on this system, should i be able to get more than one core for a libx264 encoding? [19:11] 1 socket, 2 cores, 4 logical processors (i5-337U at 1.8Ghz) i can see it's only using one of them [19:11] owonoko: did you compile with pthreads ? [19:11] (it should be default) [19:11] good question, it's a windows tablet [19:12] ugh, eh [19:12] Where did you get ffmpeg from? [19:13] i just read that rendering wma2 is slow sometimes because of the wrong use of an assembler like yasm as well, my wma2 is super slow to encode [19:13] the zeranoe builds [19:13] it's been a couple of years since i did a windows cross compile from source [19:13] wonder if it's worth it [19:14] it's abotu 220 seconds to produce a 60 second wmv file at 640x480 [19:15] it has pthreads support http://ffmpeg.zeranoe.com/builds/ [19:17] basically i've got a situation where my camera on the tablet produces mjpeg at 1280x720 and i need to preview that in a flash player based user interface (after it has been recorded) then i need to output a truncated and cropped version of the recording as a wmv [19:18] so i'm going mjpeg avi container -> disk, disk -> vp, disk -> [filter] -> wmv [19:18] not sure if that's the best way to do it [19:18] Can you pastebin your current commandline and output? [19:19] sure [19:19] When piping WAVE data to stdout such as in: "ffmpeg -i image.ape -f wav - >image.wav" ffmpeg sets length fields in the RIFF header to 0xFFFF...FF values for some reason (treats input as unknown length?). Is it possible to make it set correct length (get it from APE input file)? [19:20] I'm piping to application that splits the wav into multiple files and would like to avoid temp files [19:20] pmarty: maybe for some special cases, but what if you were using an audio filter or something? [19:20] pmarty: why can't you just use headerless (raw) audio? [19:21] kepstin-laptop: hmm, that's an idea, I wonder if this thing (shnsplit) can consume it [19:22] pmarty: if you're just trying to run shnsplit on the ape, shnsplit can handle ape, you don't need to convert to pcm [19:23] c_14: yes but it expects me to have mac in my PATH [19:23] export PATH=/path/to/mac:$PATH ? [19:23] Or do you not have the program at all? [19:23] is it having any particular problem with that length value? It's fairly common for applications which are streaming wav to use that as a de-facto "no lenght set" [19:24] I wanted to use ffmpeg instead because mac has some license issues [19:24] (I don't know, it's not in debian) [19:24] mhmk [19:24] kepstin-laptop: yes it chokes on last track [19:25] Try adding -t (length-of-file) [19:25] it treats header length literally [19:25] ffmpeg might be smart enough to write the header if it knows it'll only create x seconds of output [19:27] c_14: there's no general way for it to know in advance, tho, thanks to stuff like filter chains [19:28] Not even if you explicitly set the max length for that output file? [19:28] huh, strange. it doesn't look like there is support for raw cd audio format data input in shnsplit. [19:29] yep, shnsplit wants WAVE delivered from it's format backends [19:29] I normally use 'bchunk' to do that, which does take raw audio input, but it doesn't support any fancy stuff like encoding audio for you :) [19:34] This is how I imagined my clever command to do APE image -> *.flac: shnsplit -i 'ape ffmpeg -nostats -i %f -f s16le -' -o 'flac flac -8so %f -' -f cue image.ape [19:36] s/s16le/wav/ to make it work for each except the last track [19:56] What's the best way to concat a crazy amount of .ts files(like 300k), with random format changes all over the place? [19:56] without re-encoding, with a new part each time the format gets incompatible [19:56] -f concat takes _ages_ [19:56] cat *ts > concat.ts [19:57] You should be able to concat ts files like that. [20:00] c_14, as long as they contain compatible formats [20:00] the moment it gets incompatible(resolution change, for example), it('it' beeing whatever youtube does with it) just continues with a gray video [20:02] so i somehow have to identify those spots. running ffprobe on 300k files and analyzing its output might take a while [20:04] The only way I can think of is programmatical. [20:10] -c copy? [20:19] Hello71, doesn't help at all if the ts files don't belong to the same stream anymore [20:22] the timestamps make a huge jump, and even the video format might change [21:09] how can I record my desktop screen + virtual audio + (optional: microphone audio) ? [21:10] https://trac.ffmpeg.org/wiki/Capture/Desktop [21:10] https://trac.ffmpeg.org/wiki/Capture/ALSA [21:11] hey, where can i find the logs for this channel? i want to follow up on a regression bug i had in march (really :) ) [21:12] hannes3: !pb logs [21:12] damn it [21:13] :) [21:13] http://lists.ffmpeg.org/pipermail/ffmpeg-devel-irc/ [21:13] -devel and -user are stored there [21:14] thanks [21:16] anyone have any thoughts on encoding a 64 and 16 kbps mp3 from a 128kbps mp3 vs a ~700kbps wav? the 128kbps having been encoded from that wav [21:18] how much difference would you expect in quality? this is for a talk show with intro/outro music beds [21:18] i'd use opus if you can [21:18] well, you're always gonna get better quality by reducing the number of generations of lossy encoding. But really, 64 and 16kbps mp3? that's just gonna be horrid. [21:20] ...and probably for stereo input too [21:21] it's mono [21:21] I suppose if you're doing mono, the 64kbps mp3 is probably acceptable. [21:21] i was going to say, it's pretty good [21:22] the 16kbps is certainly undesirable, but it's for low-bandwidth people [21:22] the host keeps the 128kbps as the archival quality file [21:22] there's little difference from the raw wavs [21:23] why not archive as flac? [21:23] *i* would keep the wavs, just because, but it's not my call :) [21:23] i thought about suggesting that; how do you think file size would compare? [21:24] i don't know. smaller. [21:24] the 128kbps file is ~112MB. the wavs total just under 600MB [21:24] depends; if it's mostly speech, I think the predictor in flac actually works pretty well for that. [21:24] the closer your audio is to noise, the worse lossless compression works ;) [21:25] c_14: hey mate, heres that output http://dpaste.com/038SAG5 [21:26] cyphase any reason you are using mp3 and not something like aac? [21:26] amusingly, opus compresses noise pretty well, since it just encodes "this frequency band has noise at this power level" :) [21:26] c_14: so it's only using one of the 2 cores, i've tried various -threads settings, the input avi is coming from a dump from the camera, i'd love to improve its performance at any rate since it's affecting the other outputs too (vp6, h264) [21:27] Devrim, that's what the show's been using forever. but i could talk to the host if there was a benefit i could bring to him [21:28] owonoko: the msmpeg4 encoder doesn't support threading [21:29] c_14: right that was only part of my question originally [21:29] c_14: can i do anything else to improve the transcode to wmv [21:29] c_14: re the yasm cross compile [21:31] The build you have doesn't seem to be disabling yasm, so it should be fine. [21:32] what i was thinking is that there may be some advantage to building against the features of this specific cpu [21:34] There _might_ be, but you probably wouldn't notice without benchmarking and it'll only really be worth it if you're planning on using it extensively. [21:34] this tablet is going to produce hundreds of output movies, i need to sort out the performance any way i can [21:35] Devrim, what would be the benefit of switching their archives/podcast to aac? [21:35] owonoko: why not use a different encoder? [21:35] cyphase better compression, higher quality for the same filesize [21:35] thinking about paying microsoft for their pro encoder thing, but haven't looked into it yet [21:35] (i didn't read channel history) [21:35] llogan: for coroporate internet capable wmv files? [21:35] disadvantage would be mp3 is supported on more devices (very old mobile devices might not play it) [21:35] owonoko: bleh, sounds aweful. [21:36] you can't use H.264 in MP4 container? [21:46] llogan: windows office users have a pretty limited wmp install [21:46] but that's what i want long term, for us to come to an agreement about a codec [22:13] Why can't i copy this mpegts file to another mpegts one, while flv output works fine? [22:13] I get [mpegts @ 0x221b520] H.264 bitstream malformed, no startcode found, use the h264_mp4toannexb bitstream filter (-bsf h264_mp4toannexb) [22:14] Adding said option only leads to "Failed to open bitstream filter h264_mp4toannexb for stream 0 with codec copy: Invalid argument" [22:15] Oh, yeah. For flv i still need to add -bsf:a aac_adtstoasc [23:23] anyone here uses FFServer ? [00:00] --- Fri Nov 21 2014 From burek021 at gmail.com Sat Nov 22 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Sat, 22 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141121 Message-ID: <20141122010502.8005118A0254@apolo.teamnet.rs> [01:44] ffmpeg.git 03Lukasz Marek 07master:7d75a399a4d2: lavc/options: fix rc_eq leak [01:44] ffmpeg.git 03Lukasz Marek 07master:ab922f9ef1e4: lavu/dict: add av_dict_get_string [03:44] ffmpeg.git 03Michael Niedermayer 07master:f0ae0354d3f0: avformat/avidec: fix handling dv in avi [07:59] how to display detailed info of the ffmpeg compile command ? [07:59] now it displays CC libavcodec/x86/vc1dsp_mmx.o [07:59] YASM ... [07:59] I want to display the compile command in detail [08:00] I have forgot how to modify the makefile [08:01] <[mbm]> usually "make V=1" [08:01] ok [08:56] do I understand this right that v4l now provides yet another hw decoding API?? [10:02] wm4: maybe [10:04] Yes, my codec outputs something noone can play, but it's GOOD! [10:10] arm soc something something blabla [10:11] So? [10:11] apparently socs mean you are allowed to write terrible software [10:11] There are already too many PIXFMT for no good reasons [10:11] also this is a weird format [10:11] adding one that will stay for the next 10 years, because of some obscure format got in v4l is stupid [10:11] it'd be the first tiled one [10:11] yes [10:12] and it should not be a precedent [10:12] So, basically, noone can display it, reencode or do anything with it, without the filter. [10:15] seriously... [10:18] I agree [10:18] the main problem is that it breaks too many assumptions [10:29] ffmpeg.git 03Michael Niedermayer 07release/2.1:2d0b2db27e84: avformat/avidec: fix handling dv in avi [10:29] ffmpeg.git 03Michael Niedermayer 07release/2.2:bf219a564c42: avformat/avidec: fix handling dv in avi [10:29] ffmpeg.git 03Michael Niedermayer 07release/2.3:b6ff3acafcc9: avformat/avidec: fix handling dv in avi [10:29] ffmpeg.git 03Michael Niedermayer 07release/2.4:944570906b74: avformat/avidec: fix handling dv in avi [10:34] thardin: The patch was already merged: http://ffmpeg.org/pipermail/ffmpeg-devel/2014-November/165532.html [10:34] Thank you for looking at it, I wondered how reliable the information is... [10:35] yeah, I recall doing something similar so.. :) [10:36] Do you have an idea about the sample in http://samples.ffmpeg.org/ffmpeg-bugs/trac/ticket2345/ [10:36] Iirc, you said that it is possible to detect the "JPEG 2000 pictures" string in the file [10:38] it's probably possible to detect, but I've been somewhat less than enthusiastic about supporting files from broken muxers [10:39] That's how FFmpeg works... [10:39] it may be better to mail the muxer devs if they're actually doing the wrong thing( I don't recall what the problem here was) [10:39] When demuxing I mean! [10:39] there is a degree of broken where you should just refuse instead of piling hacks on top of hacks [10:39] otherwise muxers never get fixed, since it "works" [10:40] And it is not less broken imo to store a h264 stream with a wrong SAR and expect that the player reads the DAR [10:40] but that's the way it's supposed to work [10:40] at least in the MXF world [10:40] I am not convinced this applies here: The file contains "JPEG 2000" iirc, so it can hardly be something else. [10:41] I should say that I don't actually know if the file has any problem or just that mxfdec doesn't have the corresponding codec UL added [10:42] Is it possible that there is no UL (sorry, I am not 100% sure I remember correctly) [10:42] ? [10:43] that's possible [10:43] one thing: [mxf @ 0x7fc2d40008c0] "OPAtom" with 2 ECs - assuming OP1a [10:43] j-b: I don't think it's *that* stupid [10:43] is because mxfdec doesn't derive WrappingKind from codec UL like it should [10:43] benoit-: I do think, it's extremly bad. [10:44] I've been wanting to somehow bring parts of libbmx in for that stuff [10:44] j-b: i can understand that it opens a door, which is what is bad, right? [10:44] but alas the NIH runs deep [10:44] because on itself, the format introduction is not [10:47] benoit-: yes. [10:49] (I forgot the "from your point of view") [10:51] So? [10:52] Adding total crap, because of one specific device that is already obsolete is really a bad thing [10:54] Can any of you, explain, out of his head, what is NV12 ? [10:54] I guess the answer is yes for most of you. [10:55] Can any of you, explain, out of his head, what is this format ? [10:55] I guess the answer is no for almost everyone of you. [10:55] the question _what_ this format is is slightly unrelated, but I still wonder [10:56] it's not. [10:56] i.e. the exact representation [10:56] If you cannot explain what the representation is, then it shouldn't be a non-opaque PIX_FMT [10:56] AV_PIX_FMT_UYYVYY411 is my favorite pixfmt [10:57] I wish we could at least remove the endian-swapped formats [10:58] or the 9bits formats... [10:58] AV_PIX_FMT_YUVA420P9BE [11:00] if you want to remove 9, you should also remove 10, 12 and 14 [11:00] what about the 10, 12, 14 ones [11:00] its all two-byte formats [11:01] so, apparently these odd bit widths make it easier to process the data, because you can keep them in a 16 bit register without the danger that intermediate results overflow [11:02] but as far as that argument goes, wouldn't a 14 bit format be enough? [11:03] that doesnt change their pixfmt representation [11:03] obviously there would need to be another flag that identifies the actual number of used bits [11:03] like with audio? [11:03] (especially since libav* has them in the lower bits, not the higher bits) [11:03] The 9, 10, 12 and 14 bit formats were added by x264 for performance reasons, see Reimar's explanation a few weeks ago. [11:05] Only nine and ten, sorry [11:06] nevcairiel: 10 could make sense, because of 30bits inside 32 bits [11:06] but thats packed [11:06] That's not what "10" means in pixfmt.h [11:10] and, uh, what's up with these Bayer formats? [11:11] same remark, sorry :) [11:11] but it's less bad [11:25] j-b: I really like the "less bad" view :) [15:19] ffmpeg.git 03Michael Niedermayer 07master:367c9d33d6dd: avformat: replace some odd 30-60 rates by higher less odd ones in get_std_framerate() [15:24] michaelni, ping, re: 367c9d33d6dd [15:25] i dont think tha taxtuall fixes slomo mov from iphone (QT uses metadata to detect it) [16:08] Daemon404, the commit was intended to fix the regression. if the file is intended to play differently, then i didnt realize this, i might look into that too but i think it would be ideal if these 2 issues would be seperate tickets [16:09] apple has a "special" way to flag "slomo" videos from iphone [16:09] and quicktime more or less plays ix 10x slower during playback [16:46] Daemon404, do you know where/how that metadata is stored ? [16:46] not 100%, but mostly. it's a hack from apple. [16:47] i wasnt the one who figured it out, ill wait on them. [16:47] at best it should be sidedata anyway. [17:25] ffmpeg.git 03Benoit Fouet 07master:33acebd3ccfc: avcodec/pngdec: add APNG support. [17:25] ffmpeg.git 03Benoit Fouet 07master:5d37d70b0b7c: avformat/apngdec: add APNG demuxer. [17:47] guess we can remove that from the GSoC template [17:48] oh, Compn already edited the wiki, haha [18:28] michaelni: The sample from ticket 4012 works fine with versions 2.1 and earler. Should I backport your change to 2.2., 2.3 and 2.4 (assuming it is possible) or better not? [18:28] Daemon404: Hi, did you have time to find out "ffmpeg -i" info about the Prores sample? [18:46] cehoyos, not sure, maybe wait a week or so in case it causes any regressions but if no regressions then yes backport is ok if no conflicts [18:51] ffmpeg.git 03Michael Niedermayer 07master:c05310d4699c: avdevice/pulse_audio_common: Use av_freep(), avoid leaving stale pointers [18:51] ffmpeg.git 03Michael Niedermayer 07master:883f85fa8ffd: avdevice/fbdev_common: Use av_freep(), avoid leaving stale pointers [18:51] ffmpeg.git 03Michael Niedermayer 07master:6995be43aee5: avdevice/avdevice.c: Use av_freep(), avoid leaving stale pointers [18:51] ffmpeg.git 03Michael Niedermayer 07master:2ae2c60554c2: avcodec/vp6: Use av_freep(), avoid leaving stale pointers [19:32] benoit-: why, re: comment on dash patch? [20:20] ffmpeg.git 03Michael Niedermayer 07master:d96d8e121f1e: avcodec/libspeexdec: support zygoaudio [20:20] ffmpeg.git 03Michael Niedermayer 07master:018ce902840a: avcodec/libspeexdec: more verbose error message [20:49] ffmpeg.git 03Michael Niedermayer 07master:0c3ebbf6a54d: Changelog: add zygoaudio [21:27] ffmpeg.git 03Martin Storsj? 07master:aa8b39d99958: lavc: Move the libtwolame encoder registration to the list for external libraries [21:27] ffmpeg.git 03Michael Niedermayer 07master:5af0a701a1c0: Merge commit 'aa8b39d999589154f79300de9038994d0093cd34' [21:39] ffmpeg.git 03Luca Barbato 07master:fd9badd3cb3b: xwma: Do not leak on failure path [21:39] ffmpeg.git 03Michael Niedermayer 07master:15ed7ca437a8: Merge commit 'fd9badd3cb3b60f5c54dcea35523e1ecca2f67a6' [21:41] https://launchpad.net/ubuntu/+source/ffmpeg/7:2.4.3-1 so... we're now back on ubuntu? [21:44] interesting [21:57] can we make our opw crypto student work on aes-ni :) [22:25] ffmpeg.git 03Vittorio Giovara 07master:863ee8a855b8: lavfi: clean memory on error in ADD_FORMAT() [22:25] ffmpeg.git 03Michael Niedermayer 07master:42f3cb419aa0: Merge commit '863ee8a855b8ce27ffef41479eb66da58763faed' [22:25] ffmpeg.git 03Michael Niedermayer 07master:75819fafd821: avfilter/formats: free the correct pointer in ADD_FORMAT() [22:25] ffmpeg.git 03Michael Niedermayer 07master:b9ffafbfcc0f: avfilter/formats: Alloc NULL fmts in SET_COMMON_FORMATS() [23:03] ffmpeg.git 03Vittorio Giovara 07master:a42d5c861fea: libtwolame: prevent a NULL pointer dereference [23:03] ffmpeg.git 03Luca Barbato 07master:d466d82faaf6: dvdsubdec: Do not leak on failure path [23:04] ffmpeg.git 03Michael Niedermayer 07master:ad2424e6b298: Merge commit 'a42d5c861fea8d18d997c6ba3f4a1d8aa95a288b' [23:04] ffmpeg.git 03Michael Niedermayer 07master:ac967ad8724e: Merge commit 'd466d82faaf6e0e57a3a4be5e38e3902ef251ac3' [23:11] anyone have thoughts on using dlopen/dlsym to load AVFoundation, so a binary that supports it could still be 10.6-compatible? [23:15] &actually, would that even work for obj-c stuff? [23:37] ffmpeg.git 03Luca Barbato 07master:312daa15891d: vp9: Use the correct upper bound for seg_id [23:37] ffmpeg.git 03Michael Niedermayer 07master:a82f3de053a7: Merge commit '312daa15891dc7abb77a404fe927d5ee35c52a71' [23:56] ffmpeg.git 03Vittorio Giovara 07master:1f80742f49a9: qdm2: avoid integer overflow [23:56] ffmpeg.git 03Michael Niedermayer 07master:70e3fae88d59: Merge commit '1f80742f49a9a4e846c9f099387881abc87150b2' [00:00] --- Sat Nov 22 2014 From burek021 at gmail.com Sat Nov 22 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Sat, 22 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141121 Message-ID: <20141122010501.7890B18A0252@apolo.teamnet.rs> [00:36] If I do "ffmpeg -i test.avi /tmp/out.mp4" What audio/video bandwidth settings will ffmpeg use? [00:37] bakers, ffmpeg will tell you [00:38] To a certain degree, anyways. [00:39] If you really care, read the source code. [00:42] I see it tell me AFTER the file is done. But I don't see it output any targets before it starts [00:43] When ffmpeg starts encoding it'll output a line: "Output #*, mp4, to '/tmp/out.mp4': [00:43] I have an avi that I want to convert to mp4 and I want it to keep the quality roughly the same [00:43] Everything from there to 'Stream mapping:' [00:43] Just use -c copy [00:44] (assuming mpeg4/h264 and aac) [00:44] right now I HAVE to specify an -ab/-vb [00:44] They're OLD divx files I want as MP4 [00:44] Here is my output [00:44] http://www.fpaste.org/152648/14165270/ [00:45] mp4 is a format, divx is a codec [00:45] You can put divx into mp4 just fine. [00:45] I meant I want h.264 sorry [00:45] If I look at line #27 I don't see it specifying a video bandwidth [00:46] the input is 1246Kb/s [00:46] no mention of what it's trying to do for the output [00:46] ye, x264 defaults to crf [00:46] somewhere around 23 or so iirc [00:47] Ah ok... [00:47] Where can I get an explanation of what each CRF level is? [00:47] https://trac.ffmpeg.org/wiki/Encode/H.264 [00:48] c_14: Perfect thanks. I didn't even know about CRF [00:48] Also, you can add -c:a copy [00:48] mp4 supports mp3 audio [00:48] Unrelated question [00:49] Wasn't "MP2 Audio" = MP3 [00:49] and h.264 audio = AAC [00:49] what does h.265 use? [00:49] There's no such thing as H.264 audio [00:49] is there an article that discusses the -g option ? I'm having trouble understanding it [00:49] AAC is largely associated with h.264 as the defacto standard [00:50] MP3 is MPEG-2 Layer III [00:50] AAC is often used together with H.264 video, but there is no reason for taht. [00:50] *that [00:50] Other than that both are popular codecs. [00:51] Correlation does not imply causation. [00:51] is aac still the newest audio algorithm from the MPEG guys? [00:51] or did they develop something new alongside h.265 [00:52] As far as I'm aware, AAC is the newest audio codec from MPEG, yes. [00:53] That's about what I thought. Thanks [00:53] Though there might be some obscure other codecs. [00:55] danomite: just what the option does or about the gop in general? [00:56] in general [00:58] Tried the wikipedia page? Last I checked it was rather concise and easy to understand. [00:59] Thanks, I didn't think there was article [01:06] is the ffmpeg option -g in seconds and does it imply that there will be one keyframe only? [01:06] err i see now its in frams [01:40] danomite: the latest AAC thing is probably MPEG-4 AAC ELP. [01:41] thanks but I think someone else was asking about aac [01:41] oh, duh. you're right. [01:41] that was for bakers [01:48] if i set -crf 0 and -maxrate 192k will ffmpeg deal with such conflicting values? [01:55] I think it should. [02:09] What better way to momitor the routers than via a laptop webcam over LAN cord sent via raw udp stream https://cdn.mediacru.sh/KKiTakmvmDb6.jpg [02:09] <3 ffmpeg [02:10] Maybe you could get your mother to watch... Proper momitoring. [02:10] :P [03:36] prolly a silly question, but... on ubuntu, if I use ffmpeg, am I actually using something that isn't ffmpeg? [03:36] yes [03:37] hai c_14! [03:37] thanks [03:39] c_14: ultimately, I'd like to do some conversions. I've got an m4v file that I want in ogg and webm as well. Do you think I should stick with ubuntu's libav/avconv? Or rip them out and install proper ffmpeg? [03:40] oh gawd. reading that SO post. There's some convoluted history, right there. [03:40] oh, software. [03:41] avconv/libav should probably work, but if you need help with libav/avconv you'll have to ask in #libav [03:41] gotcha. [03:41] thanks again :) [03:53] I can run ffmpeg for hours even days right? [03:53] You can run ffmpeg for as long as you like. [03:54] As long as you don't find a bug, that is. [03:54] Or the input dies. [03:54] etc [03:54] trying to do rstmp to mjpeg which it works [03:55] at least for about 30 minutes [03:56] live ip camera stream to .jpg [03:57] Does it error out? [03:57] no just stops [03:58] or "freezes" [03:58] What version are you running? [03:59] Can you pastebin commandline and console output? [03:59] ffmpeg -i rtsp://streamurl -s 640x480 -y -f image2 -update 1 -r 1 C:\xampp\htdocs\public_html\stream.jpg [03:59] yeah [04:00] latest version, i got it last night [04:01] windows 7 environment [04:01] Is it always about 30min? [04:02] random [04:02] maybe "-frames:v 1" would help [04:05] trying to copy output [04:09] c_14: http://pastebin.com/vvQpqeXC [04:10] it's my brothers IP camera, he baught some weird brand and so far I can not really do much with the settings. [07:41] how do i look at a h264 file to see what the keyframe layout looks like [07:41] i need to dump a h264 file into another format so it's easy to scrub in this editor at 1/4 of a second or so res [07:42] doesn't need to be properly playable or have sound either [08:34] ahh yep go tit [10:44] using this while transcoding flv to webm "-b:v 128k -bufsize 128k" [10:45] and resulting video has bitrate: 210 kb/s [10:47] that's a _very_ low bitrate for a video [10:48] 320x240 video [10:48] even then [10:49] Encoders propably just can't follow a that low restriction [10:49] that on the side, why isn't it enforcing bitrate correctly [10:49] BtbN: are you familiar with zencoder ? [10:50] i have video transcoded by someone else with 107 kb/s , the same video, and it's quality is ok [10:50] bitrate allways fluctuates a bit, even in cbr mode. [10:50] trying to match that but ffmpeg is not enforcing limits [10:50] <__jack___> bah, I already made 720p files (audio + video) at 1k bps [10:50] Did you even put it in cbr mode? [10:51] <__jack___> (but not with webm hihi) [10:51] for h264 that's still somewhat ok, depending on how much motion there is [10:56] BtbN: no, let me see how to do it [10:58] BtbN: how do i use cbr mode? [10:58] BtbN: btw, why it can't dynamically cap it to say 128k max? [10:59] What should it do if it needs more bitrate than that? Just gray out the frame or what? [11:00] no idea, i just see the same video transcoded with 107kb/s bitrate and it's fine [11:01] there is no visible difference in quality between the 107kb/s and 210kb/s encoded ones but the size difference is significant [11:02] <__jack___> hefest: you may try a min & max bitrate, and, of course, a 2-pass encoding [11:03] Does webm/vp8 even support that? [11:03] Also, 2pass is not cbr. 2pass is constant output filesize [11:03] but variable bitrate [11:03] <__jack___> a 107k bps video does not mean that the whole file is at 107k all the time, it's an average [11:04] __jack___: that's ok, just need to reduce the size [11:05] <__jack___> do you need some limitation on the bitrate ? some equipement cannot handle high -or low- bitrate [11:05] <__jack___> if you're only concerns about size & quality, use constant quality [11:06] __jack___: not sure, im just trying to match the output of the transcoding of the same video from flv to webm by another service [11:06] __jack___: how do i do that? [11:08] <__jack___> use crf (like there: https://trac.ffmpeg.org/wiki/Encode/VP8) [11:09] __jack___: great, thanks [11:13] ffmpeg -i in.mp3 -i image.jpg -map 0:0 -map 1:0 -metadata:s:a title="Killer Queen (Remastered 2011)" -metadata:s:a artist="Queen" -metadata:s:a album="The Platinum Collection (2011 Remaster)" -metadata:s:a track="11" new.mp3\ [11:13] any idea why none of the meta is being saved to the mp3 file [11:14] I've tried every combination of that / removing album art / etc and it wont work [11:14] the meta is even printed in the ffmpeg pre-mux synopsis [11:14] __jack___: with lowest crf possible i've managed to match the size of the file but the quality of the video is disasterous [11:15] <__jack___> I've an idea: are you using the same codec ? [11:16] __jack___: http://pastebin.com/XtjUqiz8 [11:16] __jack___: first one is the one im trying to match [11:17] even opened it with a hex editor and cant see the meta [11:17] <__jack___> another idea: are you using the same source file than him ? [11:18] __jack___: yes, the exact one [11:19] <__jack___> ok, then I guess it's about encoding option (more than bitrate stuff) [11:19] <__jack___> did you trying presets ? [11:20] no [11:20] not sure what that is :) [11:22] __jack___: ok, so it's reading configuration from file. are there default preset options for each encoder? [11:22] <__jack___> dunno, I'm not familiar with vpx [11:25] <__jack___> found that: http://ubuntuforums.org/showthread.php?t=1522381&p=9539218#post9539218, I'm looking for the preset files; you can try theses option, it will work with some luck :D [11:25] nevermind, got it [11:27] __jack___: i tried playing with -crf transcoding the same flv to mp4 and it works good. when i match the size of the other transcoder the quality is good [11:28] __jack___: yeah, i stumbled upon the same thread :) [11:30] __jack___: hmm, that's producing 3 times the size of the competing webm [11:41] ubitux: i think it was you who i wanted to send files to in april (yeah...) http://lists.ffmpeg.org/pipermail/ffmpeg-devel-irc/2014-April/002026.html username12565 was me [11:41] just wanted to say, the bug is gone now. at least i could not reproduce it [11:41] my mail account is weird atm so no mail [11:42] ok ok [11:42] no worry [11:42] :) [11:42] just went through old mail that was marked TODO [11:46] __jack___: i got more info from the "good" webm transcoding, http://pastebin.com/Uu0N5Fds [12:00] __jack___: ok, i managed to get the same quality with: ffmpeg -y -i ET2GYGBFW4.flv -vf scale=320:240 -q:v 0 -r 25 -strict -2 -c:v libvpx -qmin 20 -qmax 50 -crf 25 RN8N5WT37RO.web [12:04] __jack___: thanks! [12:58] hi, what's the command line to convert a bunch of PNG-images to a movie with a fixed framerate? [13:01] even if I use -vf "fps=120" or -r 120, ffmpeg always "detects" the input video stream (which is a bunch of png files..) as having 25fps [13:02] because the default for image files input is 25, and you are settings the output framerate [13:02] So it would take the 25 fps and try to "resample" it to 120 fps [13:03] ffmpeg options allways affect the input/output they are in front of [13:04] ok, how to change the default input framerate for images? [13:06] put the option in front of the input [13:07] ffmpeg -i input.wav -i %05d.png -c:v libx264 -vf "fps=120" out.mp4 [13:08] is my line. Where to put the "-r 120" ? [13:08] (I already tried various permutations) [13:08] in front of your png input [13:08] but after the audio input [13:08] And it's not a video filter of course [13:09] -r should just do it [13:09] so "-r 120" right after input.wav ? [13:09] ffmpeg still "detects" the video input stream as having 25fps [13:10] like i said, options allways affect the input/output which they are imediately in front of. [13:12] I'm sorry, your suggestion doesn't seem to work. [13:12] You're doing it wrong then. [13:12] could you edit the above command so I can try it? [13:13] as I said, I already tried various permutations before even asking here [13:16] gregor1255: try to put -r 120 before the -i %05d.png [13:16] gregor1255: if that fails, try also to put -framerate 120 before the -i %05d.png [13:17] aha, -framerate works [13:18] what's the difference between -framerate and -r ? [13:19] sounds like they do the same thing. (And I can't find "-framerate" in ffmpeg's help page. I don't have the man-page installed, though) [13:24] anyway, thx jonas. [13:29] gregor1255: -framerate is an option for the image2 demuxer, that demuxer reads a series of image files as a video (or reads just a single image if you wish), [13:30] wheras -r is a global option, as an input option (before an input file) it asks to ignore the timestamps from the input and generate new timestamps with that framerate, or as an output option (after the last input file) it asks to resample the video to a new framerate, throwing away or duplicating frames. [13:31] thx, very informative :) [13:31] see http://ffmpeg.org/ffmpeg-all.html#image2-1 [13:32] actually, -framerate also goes both ways, because as an output option it can be a setting for the image2 muxer [13:32] hmm no [13:32] it can't [13:32] sorry, ignore that [13:32] it's input only [17:04] hi [17:13] I'm using this command to record my webcam: [17:13] ffmpeg -f video4linux2 -s 1280x960 -i /dev/video0 out.mkv [17:13] how can I record the video in 30 fps ? [17:14] because the actual is too low [17:45] -framerate 30 [18:56] Hello. [18:57] Is there a way of determining the bit depth of MPEG-2 video in an MXF file? I've tried ffmpeg -i myfile.mxf and it produces useful information, but doesn't include the bit depth of the video: Here's what I get http://pastebin.com/zxd27e4K [18:57] yuv422p(tv, bt709) [18:57] 4:2:2, 8bit [18:58] yuv422p10 would be 10bit [18:58] I see. [18:58] This camera is a ripoff :( [18:59] Slightly surprised that it doesn't recognise the audio as stereo. [19:02] Hfuy, it's two separate mono streams, I have no idea how ffmpeg would know them to be separate sides of a stereo stream [19:05] Hfuy, also it seems like the lavc mpeg2 decoder doesn't even support >8bit [19:05] if that is even standardized [19:06] http://git.videolan.org/?p=ffmpeg.git;a=blob;f=libavcodec/mpeg12dec.c;h=78888c7309f69fa561c897d76d9dda87f4dd47a8;hb=HEAD#l1217 [19:14] JEEBcz: I'm not a software engineer. [19:14] if the 4:2:2 colorspace list only has the 8bit one there that kind of says something :P [19:14] I assume, however, that I can rely on ffmpeg's identification of the file as 8-bit, at least inasmuch as it definitely isn't 10-bit? [19:15] and the output colorspace is the only thing that will tell you what bit depth it is [19:15] greetings [19:15] earthlings [19:15] as i am an alien from a distant planet, i do not fully understand ffmpeg [19:16] in other words, either it is 8bit (if there are no artifacts on the screen when you play it) or you don't know because lavc only supports 8bit MPEG-2 :P [19:16] especially yadif and scale [19:16] I don't fully understand ffmpeg, and I'm from London [19:16] It plays fine in VLC. But I'm not sure what they're doing. [19:16] my condolences [19:16] I'm pretty sure it's 8bit then [19:16] I'm pretty sure libmpeg2 doesn't support >8bit :P [19:16] Hfuy: London? What strange word is this? [19:17] My command line is: ffmpeg.exe -v verbose -i "F:\ARCHIVES\Video\JP 55.avi" -t 5 -vf "yadif=1:1,mcdeint=0:1:10,scale=iw*sar:ih,unsharp=3:3:2.03" -codec:v libx265 -preset ultrafast -x265-params crf=28 -codec:a libfdk_aac -y "JP 55.mp4" [19:18] So I'm converting a DV-AVI to a x265 MP4. [19:18] you are doing extremely unefficient compression with libx265, yes [19:18] The only purpose of the -vf is to deinterlace the video. I think it's too complicated right now. How can I simplify it? [19:19] JEEBcz: It is slow [19:19] yes [19:19] so use libx264 [19:19] which is not slow and you will get getter performance [19:19] more bang for the same buck [19:19] and then you should just use yadif if you only want to use yadif [19:19] I was going to see how tiny I could get these video files [19:19] getting something tiny is not an issue [19:19] the issue is how well it compresses [19:20] as in, the quality on that size [19:20] But at low bitrates... isn't x265 smaller file size at same quality [19:20] ? [19:20] yes, but only at the slowest settings [19:20] I mean, the shit that drags my machine to 0.5fps [19:20] hmm, well i was also going to maybe try x264 10-bit [19:20] the machine being a 4790K [19:20] yessss... super slow [19:21] okay okay... so first of all, the yadif [19:21] I'm not sure if this is a bug, or if I don't understand it wel [19:21] but if I don't have yadif, the x265 file has no video in vlc. [19:22] and if I use a different yadif setting, it doesn't either [19:22] somehow THIS works. but why? [19:22] http://ffmpeg.org/ffmpeg-all.html#yadif-1 [19:22] I assume ffmpeg wouldn't actually be able tod ecode 10-bit mpeg-2, then [19:22] Hfuy, yes - I presume it would output yuv422p even with 10bit [19:22] that said, if VLC could play it the probability is extremely high that it isn't 10bit [19:23] But would that actually be a proper picture? [19:23] Would it be a corrupted mess? [19:23] it would be incorrectly decoded [19:23] ffmpeg -i myfile.mxf test%03d.jpg works [19:23] yup, so it is 8bit [19:23] oh -- i think if i even removed everything past the yadif, it wouldn't show video ("yadif=1:1,mcdeint=0:1:10,scale=iw*sar:ih,unsharp=3:3:2.03") [19:24] also I have no idea how freaking old lavc decoder is in your VLC re: HEVC decoding :P [19:24] And i'm wondering why i need all this fancy stuff just to see video in vlc [19:24] in any case, just "yadif" should work [19:24] the default would be one frame per two fields, and "auto" field parity [19:25] and all input will be deinterlaced [19:25] as is noted on the documentation page I linked :P [19:25] okay [19:25] i've never heard of 10bit mpeg-2, is that a new standard extension? [19:25] iive, I was expecting that [19:25] I've no idea. [19:26] because this person just expected 10bit and I had no idea if MPEG-2 Video even had it :P [19:26] I'm not expecting ten bit particularly. [19:26] Just wanted to ensure it wasn't. [19:26] so I just noted that the lavc decoder doesn't even have such a thing [19:26] Hm. [19:26] IEC 13818 Part 8 was a 10-bit extension. [19:26] "withdrawn due to lack of interest" [19:27] So, no. There isn't. [19:27] okay, so "yadif=1:1" produces no video. if i have the other vf stuff after it, it does [19:27] ? [19:27] post full command line and terminal output in a pastebin [19:27] and link here [19:28] okay [19:28] also as I said already, I have no idea what your VLC contains and what bugs it might have [19:28] right [19:28] I recommend using latest lavf/lavc and mpv for testing myself [19:28] would that be in the vlc nightly? [19:28] the fuck I know [19:29] if you're on windows then just grab lachs0r's mpv nightlies [19:29] http://mpv.srsfckn.biz/ [19:29] interesting, never used this before [19:30] the least retarded of the mplayer forks [19:30] then i will see if it plays correctly with this [19:32] mpv.exe will not run. i wonder why [19:34] it's a terminal application. you either run it from the command line, or you drag and drop a file on it :P [19:35] duh ok [19:35] also as I already said even yadif by itself should be OK, so make sure you actually need those options you are giving it [19:35] you are setting mode and parity to 1 [19:35] (the first and second parameter) [19:36] see the documentation that I linked to you way earlier in case you were just blindly copying or following something [19:36] all right [19:37] first of all, using only the "yadif=1:1" part DOES produce a video in mpv. so vlc was the problem [19:37] unsurprising [19:37] now i will see if i did yadif the right way [19:37] yup, i'll have to keep that in mind for the future [19:38] okay now -- perhaps mediainfo isn't to be relied upon, but it says that the mp4 i produced with ffmpeg has 59.940 frames, not 29.97?!? [19:38] any insight? [19:38] so yes, you had no idea what parameters you were giving to yadif [19:38] congratulations [19:38] thank you [19:38] you are right [19:38] see what value 1 does for the first option of yadif [19:39] i shall [19:39] I have linked the documentation, and surprisingly it actually tells you what it does in English :) [19:39] ooooh nice [19:39] and what the default value is [19:40] okay. if i do "yadif=0:1:0"... [19:40] i would think this is the correct settin [19:40] i think dvavi is bff [19:41] yes. i will try this [19:41] also I recommend you drop libx265 if you need to use such a fast preset. just getting something small has not been a problem with encoders ever since libx264 happened [19:42] you can encode 1440x1080 at 30fps with 128kbps bit rate, if you really want [19:42] okay, i will consider it BUT ... the preset is just so i can get the other settins right [19:42] do note that to get good results from libx265 you will want to pretty much go placebo [19:43] there is no other way [19:43] okay, not even veryslow [19:43] you should not take libx264's presets as your base in any way or form [19:43] it's a completely different thing [19:44] hmm okay, but i think i read the presets have the same names? [19:44] yes, because libx265 copied the naming [19:44] ahh [19:44] and i know crf is different [19:44] do note that x265 is developed by chinese and indian developers for the company MCW, and they just licensed the rights for the name [19:45] so it's something completely different to x264 [19:45] oh really? i thought there were actually two x265s technically, but one was much more obscure [19:46] well anyway, i do appreciate the help [19:46] this is working [19:46] and x265 can do 10-bit and 12-bit? [19:47] which would be smaller sizes? [19:47] it can do 10bit [19:47] HEVC is the same as AVC specification-wise I think [19:47] so up until 14bit as far as the specification goes [19:48] i know i can google around, and i will, but does file size get smaller as the bit depth gets greater? [19:48] depends on the format [19:48] x264 and x265 [19:48] those are encoders [19:48] not formats [19:48] x264 encodes AVC/H.264 and x265 encodes HEVC/H.265 [19:49] right? [19:49] with AVC adding bits also added bits to the in-the-middle data [19:49] okay... [19:49] I forgot the proper english word :P [19:49] and what's a "format" technically? [19:49] a video format [19:50] what do you mean? a wrapper, like a .mp4 file? [19:50] x264 encodes videos in the AVC video format (also known as H.264) [19:50] x265 encodes videos in the HEVC video format (also known as H.265) [19:50] ohhh ok [19:50] no, I would call those containers [19:51] also if I had to use the word "format" I would call containers "file format" [19:51] not "video format" [19:51] so technically, x264 is an *encoder* which encodes into the AVC video *format* [19:51] yes [19:51] it follows the AVC specification [19:51] and creates videos that AVC decoders can decode [19:52] yes, it makes sense now [19:52] "x265 supports the Main, Main 10 and Main Still Picture profiles of HEVC, utilizing a bit depth of either 8-bits or 10-bits per sample YCbCr with 4:2:0, 4:2:2 or 4:4:4 chroma subsampling." [19:52] the AVC and HEVC names are the original ones from ISO/IEC, and H.264 and H.265 names are given to them by the ITU-T when ITU takes them into usage [19:53] ok, i will have to read about that. i was wondering what the naming difference meant [19:53] now this says 8-bit and 10-bit [19:53] yes, that's what I knew x265 supported out of the HEVC specification [19:54] but you said up to 14-bit? or maybe i missed something [19:54] see the wording I picked [19:55] "supported _out of_ the HEVC specification", at least as far as I remember nuance-wise this means that the specification has more [19:55] which is most often the case [19:55] hmm okay [19:55] the specification has more things than what encoders or decoders support [19:55] each implementation picks what they support out of a specification [19:55] that is why you have profiles [19:56] main profile is 4:2:0, 8bit only [19:56] main 10 is 4:2:0, 8-10bit only [19:56] basically feature sets [19:56] i see. sooo although the spec may allow 14-bit, the x265 encoder doesn't necessarily support it [19:56] yes [19:56] which means i can't go above 10-bit [19:56] that i know of [19:57] yes, x265 doesn't support >10bit as far as I know [19:57] okiedokie [19:57] i get it. thanks [19:57] also I wouldn't be surprised if you actually don't get as much compression gain from >8bit with HEVC compared to AVC [19:58] IIRC parts of the reasons why higher bit depths helped as much with AVC were fixed in HEVC [19:58] yeah [19:58] basically, they made the inbetween values somewhere in the middle of encoding have more bits [19:58] while with AVC it depended on the bit depth [19:59] ok [19:59] so with 8bit encoding in AVC the in-the-middle value would be reduced back to 8 bits somewhere in the middle of encoding, while HEVC just defines that value to be higher to begin with [19:59] (which hardware manufacturers then proceeded to cry a river about) [20:00] ok [20:00] lots of complexities [20:00] but i think i understand what i need to know now [20:00] i'm going to go play around with it [20:00] anyways, feel free to test things out but just remember that by using a better format you don't necessarily get better results [20:01] and why, in a nutshell? [20:01] because implementations [20:01] the specifications for video formats don't specify how good of a job you must make compressing [20:02] just that your bunch of bits has to be compliant with the decoding flow [20:02] so for example, x264 has been around much longer and has been refined, so in some ways, it may still be superior to x265 [20:03] yeah, so there isn't one way to make an encoder; the end result just has to fit into the specification [20:03] hence some encoders are better than others, though they all use the same spec [20:04] yup? [20:04] okay! [20:04] i've got to go now [20:05] thanks a bunch. extremely educational [20:05] see you [21:34] hey everyone [21:58] hey guys, what's up with the fps column in this output when the input stream from the camera is 30 fps https://gist.github.com/jotham/9409e595eaf13bb82a74 [22:35] owonoko, best guess is that its the count of input frames [22:38] It's the encoding framerate, ie how many frames per second it's encoding. [22:53] Hi guys, I'm having some trouble statically compiling x265 to use with ffmpeg. I can see that CMAKE_C_CREATE_STATIC_LIBRARY unfortunately isn't in CMakeLists.txt so I tried making x265-static (which appears in CMakeLists). Do any of you have a proven method of compiling x265 statically on a UNIX-based system? Here is the console input / output: [22:53] http://pastebin.com/RLNEhTY7 Thanks! [22:54] c_14: right but it's a 30fps stream and it's not i/o or cpu bound at the moment at any level [22:58] Never mind, problem was some odd permissions error in my target directory :D Thanks anyways! [23:16] Is libvo_aacenc still a low quality AAC encoder? [23:16] I downloaded a static build of ffmpeg and it wants to use that AAC encoder [23:17] when you say "wants to use" do you mean that it is the only aac encoder included? [23:19] compstomp: when I do ffmpeg foo.avi /tmp/out.mp4 it defaults to that aac encoder. How do I see if other encoders are installed [23:19] ffmpeg -encoders | grep aac [23:19] will show you your aac encoders [23:20] I see libvo_aacenc and aac_latm [23:20] Reading this: https://trac.ffmpeg.org/wiki/Encode/AAC#libvo_aacenc [23:21] "but is a rather poor encoder compared to libfdk_aac and even the native FFmpeg AAC encoder" [23:21] I am not one of the devs here so I don't consider myself nearly an expert on the topic. But most of the people around strongly prefer libfdk_aac [23:21] I don't know if that's current though [23:21] Do you know if I can get a static build with that encoder in it? [23:23] Currently, fdk_aac is the preferred aac encoder. To find a good static, you'll just have to google around searching for something like "ffmpeg static YOUR_OPERATING_SYSTEM libfdk_aac libfdk-aac" (annoyingly, the convention on when to use - and _ isn't bulletproof [00:00] --- Sat Nov 22 2014 From burek021 at gmail.com Sun Nov 23 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Sun, 23 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141122 Message-ID: <20141123010503.0660D18A0252@apolo.teamnet.rs> [01:53] ffmpeg.git 03Luca Barbato 07master:cd975d5658a1: hevc: Spin the mv_mpv_mode calls in a stand alone function [01:53] ffmpeg.git 03Michael Niedermayer 07master:148506c965dd: Merge commit 'cd975d5658a1cbe99939df75db59d5ae9fbcb4e0' [02:18] ffmpeg.git 03TOYAMA Shin-ichi 07master:e01acd868b01: doc/decoders.texi: typo in description for option ifo_palette [02:18] ffmpeg.git 03Michael Niedermayer 07master:230aeee94c27: doc: fix the the typos [02:22] can vaapi just be used for general decoding (i.e not rendering) [02:28] kierank: yes [02:29] in fact, the rendering part of vaapi is a joke [02:29] so you're better off using decoding only [02:29] https://gitorious.org/hwdecode-demos appears to be rendering only [02:32] don't know, but reading back from the vaapi surfaces is possible [02:40] llogan: Thank you for answering so many questions on the Zeranoe forum. I really appreciate it. [03:45] ffmpeg.git 03Mark Reid 07master:08d81d0a01e4: libavformat/mxfdec.c: export source package uids and names as metadata [05:44] Looks like there is an error with DeckLink support using the latest FFmpeg and DeckLink SDK 'libavdevice/decklink_common.cpp:61:42: error: 'char* dup_wchar_to_utf8(wchar_t*)' was declared 'extern' and later 'static' [-fpermissive]' [05:54] Full error: http://paste.ubuntu.com/9164142/ [07:53] anyone help! I compile using this command: g++ deco.c -o deco -D__STDC_CONSTANT_MACROS -lavdevice -lavformat -lavfilter -lavcodec -lswresample -lswscale -lavutil [07:53] but got lots of undefined reference error [07:53] for example, deco.c:(.text+0x6a): undefined reference to `avcodec_find_encoder(AVCodecID)' [07:54] I have already installed ffmpeg libs in /usr/local/lib [07:58] forgotten extern C? [08:02] I don't quite understand what you mean [08:02] http://paste.ubuntu.com/9166377/ [08:03] this is my codes [08:05] use gcc not g++ [08:08] it works, thank you very much [10:00] Anyone help, when I run a program I got this error [10:00] me at ubuntu:~/codec$ ./deco [10:00] H264 Codec not found [10:01] But I've already installed x264 on my machine [10:02] And the ffmpeg example runs OK [10:02] root at ubuntu:~/ffmpeg-2.4.3/doc/examples# ./decoding_encoding h264 [10:02] Encode video file test.h264 [10:04] http://paste.ubuntu.com/9168165/, this is my codes [10:16] anyone please help [10:20] didn't build libavcodec with --enable-libx264? [10:21] also, didn't call av_register_all [10:22] I build ffmpeg 2.4.3 with ./configure --enable-libx264 --enable-gpl --enable-shared [10:22] let me try av_register_all [10:24] thank you, it's av_register_all. [13:19] ffmpeg.git 03James Almer 07master:14b9302f5f44: lavf/apngdec: properly skip currently unsupported in-stream tags [13:28] ffmpeg.git 03Matthew Oliver 07master:3bedd72a9ec9: lavf: fix apngdec under msvc. [14:07] ffmpeg.git 03Reimar D?ffinger 07master:79be253635fc: mxfdec: minor simplification. [14:07] ffmpeg.git 03Reimar D?ffinger 07master:c2c27e9e51c9: indeo2: move variable declarations into blocks using them. [14:07] ffmpeg.git 03Reimar D?ffinger 07master:478c61ccb2be: h264_i386: Optimize decode_significance_8x8_x86 for 64 bit. [14:07] ffmpeg.git 03Reimar D?ffinger 07master:33fc1ccb4910: h264_mb: Use smaller data type for refs in await_references. [14:07] is there a variable shift instruction [14:10] shifting left is easy but right is much harder [14:18] pmulhw with proper multiplicand ? [14:18] which might be obtained from a table using phufb ? [14:19] never tried, but might be doable for 16 bits values [14:19] for 8bits, I don't think so [14:20] help! When I decode a h264 video using avcodec_decode_video2, I got this error http://paste.ubuntu.com/9172231/ [14:21] This is the avcodec_decode_video2 output. [14:22] What's that mean? [14:22] kurosu: how do you shift right with pmulhw [14:22] kierank, you shift left and truncate the 16 MSBs ;) [14:23] ah [14:23] of the 32 bit result [14:23] but I think for some values you wouldn't get the proper result [14:23] pmullw is s16xs16 [14:23] I think there might be a variation with u16, not sure [14:24] also, maybe bmi (as in the amd insn set) has somthing [14:25] I am considering writing v210enc simd [14:26] ffmpeg.git 03Michael Niedermayer 07master:0b9a9e0e2ccd: avcodec/libspeexdec: make array const [14:27] yeah, I had figured it was about some peculiar unpacking [14:28] I'm not sure it's possible to just invert the unpack [14:28] probably not worth it, but you might want to shift only the parts that are constant shift, mask the result and gather it (not sure it works) [14:29] oh, bmi seems to be only about gprs anyway [14:32] I guess I could shuffle and shift 3 times each [15:24] how does one produce a SIMD output when using bytestream2_init_writer [15:29] hello guys [15:41] after using ffmpeg feed a ffm...the player only can not play the complete video stream [15:42] any one know the reason [15:56] ffmpeg.git 03Benoit Fouet 07master:90c9b494052e: ffplay: fix mem leak when opening input or parsing options fail. [16:37] kierank: pmulhrsw can be (ab)used for right shifting individual elements by different amounts. i did that when I wrote x264's v210 asm [16:38] Gramner: i see [17:31] pmulhrsw does rounding, but it's obviously doable if you managed to [17:31] on the other hand, that probably means it's just a sign-off away from being in ffmpeg then (if that was the intent) [17:37] kurosu: Gramner did the opposite of what i wanted to do [17:37] he did v210 -> nv20 [17:37] I want yuv422p10 -> v210 [17:37] ok [17:42] when will we get the "one true colorspace" to end them all? :P [17:42] I think I can make it work but it'll be embarassingly bad [17:42] when hardware, codecs, players all get together and pick one space / packing / byte and just get on with it. no more converting [17:43] yeah good luck with tha [17:43] "v210 is good enough for everyone" :P [17:43] ehe [17:43] it pretty much is in the yuv domain [19:02] ffmpeg.git 03Michael Niedermayer 07master:8e6a44cfc5ec: avdevice/iec61883: Use av_freep(), avoid leaving stale pointers in memory [19:02] ffmpeg.git 03Michael Niedermayer 07master:7df2981f04dc: avfilter/avf_concat: Use av_freep(), avoid leaving stale pointers in memory [19:02] ffmpeg.git 03Michael Niedermayer 07master:9146a476003c: avfilter/graphparser: Use av_freep(), avoid leaving stale pointers in memory [19:07] ffmpeg.git 03Reimar D?ffinger 07master:8437cc72060f: ffv1dec: Avoid unnecessarily large stack usage and copies. [19:10] can someone using msvc 2012 check what is wrong with the ebur128 filter? it's been broken there since forever [19:10] it's a known bug [19:10] it works in msvc 2013, so it's probably something related to the c99 converter [19:10] Ah, i see [19:10] nevcairiel reported it [19:10] and it was fixed in recent versions [19:11] its a compiler bug [19:11] msvc 2012 has an issue with some sse2 float things [19:11] they fixed it in 2013, but they are not going to fix 2012 [19:12] ye [19:45] ffmpeg.git 03Michael Niedermayer 07master:2f6550bb9a8d: avcodec/mjpegdec: fix pixfmtid 0x14111100 [19:46] oh yeah vp6f is flipped [19:46] i remember mplayer bug on that i think [20:33] ignore the patch i sent [20:33] it's broke [20:52] how do you OR two xmm registers in x86 asm? [20:53] Is there not por? [20:54] I don't believe so [20:54] por for integers or orps for floats [20:55] oh there is a POR [20:55] Yeah. POR is listed in Intel's docs [20:55] it's just a pita to grep in the manual [20:56] kierank: it may help to read the intel and the amd manuals together. they're organized differently, so they partially complement each otehr. [20:57] ffmpeg.git 03Michael Niedermayer 07master:4327088da355: avcodec/x86/lossless_audiodsp: support len %16 == 8 in scalarproduct_and_madd_int16() [20:57] ffmpeg.git 03Michael Niedermayer 07master:ce6141259bb9: avcodec/wmalosslessdec: support 24bit lossless [21:00] how do I use movhlps in x264asm? [21:00] what do you mean? [21:01] there's nothing special about that instruction [21:01] yasm moans about it [21:01] movlhps m1, [r3+r4] [21:01] are you using ymm registers? [21:01] it's only valud for xmm [21:01] valid* [21:02] so use xmN in that case [21:02] also movhlps with memory arg is not valud [21:02] oh [21:03] you probably want movq [21:03] oh, movLH. movhps then [21:03] (x86 has too many move instructions) [21:04] yeah, I don't like the whole int/float split [21:09] michaelni, regarding scalarproduct_and_madd_int16, does an order&8 mean order <= 8? [21:09] otherwise, I kind of remember we decided to padd the buffer with 0 in another case [21:10] order <= 8 (for real files, not for the format) means mmxext only is fine, otherwise ... [21:10] hmmm soo close [21:10] my luma is right but my chroma is wrong [21:11] kurosu, the fie has order == 8 and the sse* code did definitly not work for it [21:11] fiLe [21:11] kierank: can you post your code? [21:11] http://pastie.org/private/htp6vzpwhnkiekveqtna3a [21:12] probably the pshufb indices are broken [21:12] michaelni, at this point, what happens if eg order=24? it would fail/crash? [21:12] it should use the mmx code, also theres a bug in my code ill push a fix in a moment [21:13] packing is here: https://developer.apple.com/library/mac/technotes/tn2162/_index.html#//apple_ref/doc/uid/DTS40013070-CH1-TNTAG8-V210__4_2_2_COMPRESSION_TYPE [21:15] seems that Cr is broken badly [21:17] ffmpeg.git 03Michael Niedermayer 07master:d3512a0e8926: avcodec/x86/lossless_audiodsp: fix fallback code for 32bit [21:18] kierank: don't you mean to load from r1 and r2, not r2 and r3? [21:18] lol [21:18] yeah [21:18] thanks [21:18] please use named argument... [21:18] that really helps readability [21:19] I don't know how to use those [21:19] is there any documentation? [21:19] in the x264 header [21:19] http://git.videolan.org/?p=x264.git;a=blob;f=common/x86/x86inc.asm;hb=HEAD#l98 [21:19] documentation [21:19] kurosu, if theres some fixable problem with the padding and you think its better then iam happy to revert the fallback [21:21] michaelni, not really, I just misread the code and so was really trying to avoid what it did [21:21] I don't think I'll provide anything better in the foreseeable future [21:24] ubitux: [21:24] cglobal v210_planar_pack, 5, 5, 4, y, u, v, dst, width [21:24] libavcodec/x86/v210enc.asm:43: error: undefined symbol `width' (first use) [21:24] That's probably why I never used named args [21:24] use with widthq or something [21:24] am I being stupid [21:25] you need a suffix for usage [21:25] like r2 ? r2q [21:25] width ? widthd probably in your case [21:25] assuming that's int [21:25] I've done movsxdifnidn r4, r4d though before [21:25] instead of using movsxdifnidn, make width ptrdiff_t [21:26] kierank: http://git.videolan.org/?p=ffmpeg.git;a=blob;f=libavutil/x86/pixelutils.asm;h=7522f24a42de84ef84d939af58f97272380c2efc;hb=HEAD#l110 [21:26] here are some simple examples [21:27] (without any local variables but it's not much more complicated) [21:27] thanks [21:28] use m also when the ABI doesn't pass the argument in a gpr but on the stack (long arg list usually) [21:29] and, should the case arise, try avoiding using float parameters in your dsp prototype, the ABIs became wildly more protracted to handle [21:31] could well be an avconv bug but the pictures look the same but avconv only decodes 7/12 frames [21:31] who knows why... [21:56] kierank: forgot a file? [21:57] libavcodec/x86/v210enc_init.c or something [21:57] ubitux: ah true [21:57] and the asm itself... [21:57] :) [21:57] change HAVE_MMX while at it [21:57] (with ARCH_X86) [21:58] too late [21:58] but changed locally [21:58] (and coding style in v210_planar_pack_c but libav will surely point it out) [21:59] what about the style [21:59] ( style ) [22:00] ah [22:00] r0+2*widthq ? forgot to change one here [22:00] anyway, that's all for meaningless review [22:15] ubitux: thx [23:02] kierank, could you add the result of {START,STOP}_TIMER [23:10] kurosu: I need to find a file long enough to be significant [23:11] true, and that's probably only me that wants that [23:12] I had the problem when benchmarking it myself [23:14] it doesn't need to be too long to get at least a vague idea of how much faster it is [23:16] Action: kierank downloads a file from Daemon404 [23:16] ffmpeg.git 03Matthew Oliver 07master:293fee4bc235: libavcodec/tiff: Fix static linking of lzma with msvc. [23:17] hum, that's like raw ? you can probably append it to itself several times or the like [23:29] kurosu: [23:29] C: 132635884 UNITS in test, 512 runs, 0 skips [23:30] 65324749 UNITS in test, 512 runs, 0 skips [23:30] asm [23:31] on Intel(R) Xeon(R) CPU X5675 @ 3.07GHz [23:31] what are you using? odd that it used UNITS instead of decicycles [23:32] msvc? [23:32] linux x64 [23:32] using avconv btw [23:33] kierank, ok, approximately halved [23:33] I'm wondering if unrolling+interleaving twice would help [23:34] ah yeah, timer.h in libav always uses "UNITS" it seems [23:55] kurosu: about 3 times faster on sandy bridge [23:56] kierank, well that's unexpected [23:56] does that still pass fate ? [23:56] or whatever benchmark you set for yourself [23:56] don't think there is a fate test [23:56] since it's an encoder [23:57] 3 times faster asm compared to c i mean [23:57] ah ok [23:57] saste: signalstats provides this info [23:57] saste: i'm actually working on making it faster right now btw [23:57] might be useful to show it visually though.. [23:58] kierank, still 2/3 of the original asm, that sounds good (though on the high range of the gain I had expected) [23:58] saste: could you add GRAY16 support? [23:58] (I had expected at most, no gain could also have happened) [23:59] ubitux, sure [23:59] kurosu: i'm not 100% sure it's bitexact [23:59] ubitux, there is any way to use signalstat to show the computed values? [23:59] saste: probably not yeah... [23:59] the idea of tdiff is that it can be used to monitor things changing (e.g. for surveillance systems) [00:00] --- Sun Nov 23 2014 From burek021 at gmail.com Sun Nov 23 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Sun, 23 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141122 Message-ID: <20141123010501.F278118A0251@apolo.teamnet.rs> [00:17] You shouldn't find a static ffmpeg compiled with libfdk_aac. [00:18] Anybody who offers such a build is going against the libfdk-aac license. [00:18] s/going against/violating [00:31] hello [00:31] c_14, no, he's not. He's violating the GPL. The libfdk_aac license is actualy quite permissive. [00:32] i am reading about using mcdeint with yadif [00:32] the third parameter in mcdeint is qp, or quantization parameter. could someone help me to understand what this means? [00:35] BtbN: aah, I just know it's a license issue somewhere. Never really looked at the specifics. [00:36] uex1fi: https://en.wikipedia.org/wiki/Quantization_(image_processing) [00:37] "Higher values should result in a smoother motion vector field but less optimal individual vectors." [00:37] For one, what's the highest value? Also, what is a vector. [00:38] is it basically that the higher the value, the smoother the video? [00:38] and if it is set too high, the video will look too smooth? [00:39] c_14, The libfdk license is very strange. There are quite a few GPL projects which interpret it in a GPL compatible way. [00:41] uex1fi: qp ..FV.... set qp (from INT_MIN to INT_MAX) (default 1) [00:42] c_14: I don't quite understand that [00:42] i know the default is 1, but i just don't know the number range allowed [00:42] from INT_MIN to INT_MAX [00:43] i don't know that those mean.... [00:43] Usually -32767 to 32767 [00:43] okay, wow [00:43] but the examples i see just have it set to maybe 10 or 20 [00:44] Just because you can set it from INT_MIN to INT_MAX doesn't mean you have to. [00:44] Or should. [00:44] yes [00:45] ok, so it's not like a 1-10 range... i'll just have to do some tests to find the right value, then. [00:45] as for mcdeint, what's a really good deinterlace setting without going crazy? [00:46] i was thinking of doing extra_slow, but i don't know if it's going overboard [00:52] or let me put it this way. in mcdeint, does 'extra_slow' always produce a better result than 'fast'? [00:53] I've never really used mcdeint. You might just have to compare results. [00:54] yeah [00:57] hi.. is there a way to get ffmpeg to write a .srt subtitle file without the html tags? [00:57] my player can't interpret some of them, so they show up on the screen, I just want the text [01:50] Hello, I am trying to statically compile ffmpeg on Mac osx 10.9.5. I have gotten my heuristic to work on linux but now I need it on mac and I am not sure the best way to go about getting a static binary compiled. Here is my console in-out: http://pastebin.com/z5C5Pcuw and here is config.log: http://pastebin.com/xEzGX1x0 Thank you in advance! [01:54] for yadif, 2 is "Like send_frame, but it skips the spatial interlacing check." [01:55] what is this? [01:55] why would you skip that check? [07:15] when i run ffprobe on an mp3 i encoded from some wavs (and sometimes with a few mp3s thrown in), i get a non-zero 'start'; as in the following string: "Duration: 02:01:02.88, start: 0.025056, bitrate: 64 kb/s". why would it not be 0.000000, what are the effects of that, and how can i fix it? [07:52] anyone help! I compile using this command: g++ deco.c -o deco -D__STDC_CONSTANT_MACROS -lavdevice -lavformat -lavfilter -lavcodec -lswresample -lswscale -lavutil [07:53] but got lots of undefined reference error [07:53] for example, deco.c:(.text+0x6a): undefined reference to `avcodec_find_encoder(AVCodecID)' [07:54] I have already installed ffmpeg libs in /usr/local/lib [08:34] Hey, could someone help me with a question I have about image sequences -> video [08:34] ? [08:35] I would like to create a video using the mpeg4 video codec with *just* two images as an input. Id like to be able to use two arbitrary images rather than using the image-%03d.png syntax. Is this possible> [08:58] Try using the concat filter. [09:01] lookatmeyou: make sure your PKG_CONFIG_PATH is set correctly and then pas the output of `pkg-config --cflags --libs libavdevice libavformat libavfilter libavcodec libswresample libswscale libavutil' to g++ [09:02] *pass [09:02] thank you, later I use gcc instead of g++ and works [10:05] Anyone help, when I run a program I got this error [10:05] me at ubuntu:~/codec$ ./deco [10:05] H264 Codec not found [10:05] But I've already installed x264 on my machine [10:05] And the ffmpeg example using x264 runs OK [10:05] root at ubuntu:~/ffmpeg-2.4.3/doc/examples# ./decoding_encoding h264 [10:05] Encode video file test.h264 [10:05] http://paste.ubuntu.com/9168165/, this is my codes [10:16] anyone please help [10:24] solved, thanks [11:11] I have a question to ffmpeg [11:13] We do livestreaming over the internet and transmit the stream to rtmp streaming server over udp. If theres no input to udp we wanne show a picture that signals that the stream is currently unavailable. Can it be done with the videofilter complex switch? [11:25] Hi. is there any command line to tell encoder to ignore invalid data from input? [11:28] Hello, I try to convert images to video. This works if the images are named 1.png, 2.png, 3.png ... but I can not make it work with 00001.png, 00002.png, 00003.png. I am forced to change their name. Is it possible to do otherwise? Thank you for your help! [11:33] user39564: use %05d.png as your input [11:36] user39564: http://ffmpeg.org/ffmpeg-formats.html#image2-2 [11:37] relaxed: thank you, sorry for the loss of time. [11:38] viruser: pastebin the problem [11:41] relaxed: Here you go http://paste.debian.net/132952/ [11:42] background: I have salvaged a video from a scratched cd. the program filled unreadable sectors with zeros. Most players crash when they reach the zero parts. but mplayer can skip it. I want to fix the video [11:44] does mplayer handle the audio we [11:44] well* ? [11:45] yes. Like i said, it works fine. it just skips the broken areas and continues to read the unbroken parts. [11:46] I want ffmpeg to create a file without the broken area so other players can handle it as well [11:46] if that's possible [11:46] and ffmpeg stops encoding at these sections? [11:47] Well, as the dump suggests, it stops at 35kth frame [11:49] use mplayer's '-vo yuv4mpeg' and '-ao pcm' to dump the video and audio, then feed those to ffmpeg as input [11:51] I'll try that. thanks [14:34] When I decode a h264 video using avcodec_decode_video2, I got this error http://paste.ubuntu.com/9172231/ [14:34] This is the avcodec_decode_video2 output. [14:34] What's that mean? [14:35] either your code is wrong or your AVC stream is broken [14:35] as in, I hope you are using lavf to open the actual annex b stream [14:35] and then decoding with lavc [14:36] http://paste.ubuntu.com/9172489/, this is my codes [14:37] And th input file Avatar.h264 can use ffplay to play. [14:37] exactly [14:37] you are only using libavcodec [14:37] not libavformat [14:38] you need to first read the raw annex b with libavformat [14:38] and then push its output to libavcodec [14:38] OK, let me try, thanks [14:44] Could you please give me some codes to show the process? [14:46] I don't have to, there are already samples in the FFmpeg code base :P [14:46] where you most probably copied your code so far [14:48] Is it in avio_reading.c? [14:49] Hello guys, i noticed that the -f tee Mux is very buggy. Some streams are unable to be restreamed while some others are loosing synch with the video/audio. This with all the other mux doesnt happen. [14:49] How can i reproduce it to make a bug report. Its not something that is happening all the time but most of the time yes [14:52] hello [14:53] how do I stream RTSP with ffmpeg, without pipe option, thank you [14:53] http://forum.videohelp.com/threads/368465-Stream-from-RTMP-to-Nokia-72-%28RTSP%29-on-home-network?p=2358217 [14:53] the code is in the thread, but it didn't work for me [15:25] Hello Guys [15:26] I try to use a media player to play a mpg stream..but when fffmpeg finish filling up the feed1.ffm [15:26] the player stoped ... [15:27] anyone know what is wrong here [16:25] hello [16:26] my ffmpeg config does not produce a working RTSP link [16:27] http://pastebin.com/2JKAPzkH [16:27] I have already followed the tutorials I found on the web, much of the material is several years old, please assist thanks [16:43] so i find out i need ffserver for streaming [16:44] and zeranoe says it can't be compiled for windows and it's outdated [16:44] worthless [16:45] ffserver is barely used, barely maintained [16:45] if you can't do what you want with just ffmpeg, you're pretty much out of luck [16:47] Wader8 [16:47] there is one way, you can output it to HLS [16:47] no i don't want HLS sorry [16:47] and make a webserver and place the m3u8 file there [16:47] yes just saying [16:48] because my device only supports 640x360 MP4 Visual [16:48] and over RTSP [16:48] and VLC does this 99% except the breaking 1% ... there's a bug that times it out every 60 secs, impractical to watch like that [16:49] yeah, VLC is simpler to use and most probably more maintained [16:49] at least in that part [16:49] basically im trying to get video over, and video's not necessary, i get audio okay, but the cutoff 60 s is always there [16:50] there would be no issue if VLC didn't have that bug, or actually RealPlayer on Nokia didn't have that bug [16:50] either fix the issue in VLC or try to find an ffserver user that uses RTSP [16:50] (and then find out if ffserver has the same issue) [16:50] i was trying to do that but never could get ffmpeg to work [16:50] to find out if VLC is to blame or not [16:50] well, ffserver being what it is... [16:51] i've tried TVersity, very confusing program had no idea how to setup, and can't lower bitrate below 0.8 which is unacceptable that won't work over my wlan distance [16:52] does anybody know anything else, Wowza is not free and will probably run out of a few days right ? [16:52] or a few weeks [16:53] VLC is so good in this because I can take directly an external stream, transcode and publish as a stream, it's just exactly the right type of a thing, only that bug, it's working, just loses connectiong every 60 secs I have to clik Play and it's impractical to do that while doing something else [16:53] automate it somehow? [16:54] or debug the issue [16:54] to see which end has it [16:54] automate what ? [16:55] don't understand that [16:55] well im not an expert on VLC so I have to yet find out how to open logs and traces [19:15] noob question alert: is it possible to convert one audio channel of a stereo stream (dvd ac-3 codec) to a mono stream without re-encoding? [19:15] no [19:16] ok, thx! [19:16] I have a curious issue. I'm trying to remux an NSV stream, and stream copy the VP6 and AAC into an FLV container for RTMP streaming. It plays fine in ffplay, but when c:v copy are used in ffmpeg, the image produced is upside down. Horizontal flip. [19:31] I would like to create a video using the mpeg4 video codec with *just* two images as an input. Id like to be able to use two arbitrary images rather than using the image-%03d.png syntax. Is this possible? [19:31] Also, Hi everyone :) [19:32] If you have an answer, Ive posted this question on the pastebin http://pastebin.com/nnxqkytf [20:37] is there an #ifdef that will tell me wether i am on libav or ffmpeg? [20:39] https://github.com/FFMS/ffms2/blob/master/configure.ac#L123 [20:41] lol [20:41] how long do you think will the LIBAVCODEC_VERSION_MICRO trick work? [20:42] probably will never be removed [20:42] the only thing that might break it is libav going to 100 with micro, but I think they won't be able to do that :P [20:43] JEEBcz: thx a lot! [20:46] Now my little program at least compiles on ubuntu :) [20:47] But what I don't get is why we have 2 versions of ffmpeg with libav lagging behind [21:13] I want to convert audio data to 32bit float mono. Is there a function ready to use, or do i need to code it myself? [21:19] Set the codec to pcm_f32le and audio channels to 1? [21:27] c_14: found it: audio_resample seems to be what i was searching for. [21:41] Hey, how do you make webms? [21:54] ffmpeg -i video webm.webm [21:55] https://trac.ffmpeg.org/wiki/Encode/VP8 [22:46] i currently have an ffmpeg command line that takes multiple audio inputs, concatenates, and encodes to mp3. how can i specify a silence buffer to go between/after each clip? i know i could create a silent audio file and just use it as an input, and that would work well for a static buffer length (especially since the command line is being generated programmatically), but i want to be able to specify the buffer length dynamically. bonus points if i [22:46] can specify each buffer's length separately. if i don't find a solution, my fallback is to create the silent audio dynamically with ffmpeg -f lavfi -i aevalsrc=0 -t SECS silence.wav [23:24] does dca codec decode the diff part of DTS-HD MA stream for lossless output? [00:00] --- Sun Nov 23 2014 From burek021 at gmail.com Mon Nov 24 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Mon, 24 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141123 Message-ID: <20141124010501.93AAA18A008C@apolo.teamnet.rs> [00:09] Hi, I am using ffmpeg to cut part of a log video (> 1 hour), cutting 5 seconds takes ages (> 1minute): ffmpeg -ss 200 -i movie.mp4 -to 5 .. [00:10] Can I mitigate this? [00:10] I am outputting to stdout to then pipe to other unix tool, mabe it's the problem? [00:11] If you put the -ss after the -i, it should be quicker, but less accurate. [00:11] The farest the start point, the longer it takes [00:11] Well sacarasc it's the opposite [00:11] https://trac.ffmpeg.org/wiki/Seeking%20with%20FFmpeg [00:12] Then I am wrong. I always forget which way it is. [00:12] And as long as you aren't using -c copy, they should both be just as accurate. [00:13] is there a way to convert only part of a video to a sequence of PNGs? I know how to do it for the whole video, but I have a large movie where I want to only extract frames from about a 5 second interval [00:14] diginet: I am just finishing a tool that does exactly this but with animated gifs [00:14] vvo: interesting (making a gif is actually my end goal) [00:14] ahah [00:14] then we are on the same path [00:14] I tried the png way [00:14] but it ended up with bigger gifs [00:14] in the end [00:15] (mp4 > %d.png > gif) [00:15] going to push my project in some minutes [00:15] yeah [00:15] it's too bad GIF doesn't have any sort of interframe compression [00:16] I combine ffmpeg, convert and https://github.com/pornel/giflossy [00:16] And I get very good compression [00:16] I am only struggling with fast seeking, which is not as fast as I thought [00:17] vvo: is that your repo? [00:17] For screencast it's ok but for 3.2gb movies it's slow [00:17] going to push in minutes, will keep you posted here [00:17] wait [00:17] it's in node.js but from the index.js file you will be able to find all the needed commands to | pipe [00:17] because you can do it all in memory (ffmpeg | convert | gifsicle) unix way! [00:18] vvo: lol, imagine what an entire feature length movie in GIF format would look like. . .filesize-wise [00:20] no no [00:20] only a part of it of course [00:21] but even cutting from 200s to 205s takes ages [00:21] seems like it's dealing with 0-200 frames while they will be dropped [00:57] diginet: https://github.com/vvo/gifify [00:57] what do you think? [01:08] diginet: ? [10:25] Hi, I am trying to capture video and audio from webcam with v4l2 drivers. I would like to capture one video stream with libaac and another audio stream with pcm_s16le separately from the same webcam source. Tried with: ffmpeg -f alsa -ac 2 -acodec pcm_s16le -i hw:2,0 -f v4l2 -vcodec h264 -r 30 -s 1920x1080 -i /dev/video0 -f alsa -ac 2 -i hw:2,0 -map 0:0 -c:v copy -c:a copy VIDEO+AUDIO.mp4 -map 1:0 -c:a pcm_s16le AUDIO_ONLY.wav getti [12:34] Hi I have a large collection of mkv and due to some stupid resizing of volumes many of them may be damaged (missing parts) is there a way to check their integrity with ffmpeg? [16:50] rhagu: ffmpeg -i mkv -f null /dev/null <- should do it [17:10] c_14 thanks for the answer, I tried this: "ffmpeg -v warning -i my.mkv -f null -" as it was mentioned on a homepage, what would be the difference? [17:12] You might get slightly less messages. [17:17] c_14 it is running at the moment, will probably take about 2h, the last time I ran the command I mentioned, no errors showed up. Does this command check everything (audio, subtitles, video), or just video? [17:18] It'll check one video stream and one audio stream. [17:18] add -map 0 to check all streams [17:24] thanks [17:26] After a list of streams I get: "Number of stream maps must match number of output streams" using "ffmpeg -map 0 -i my.mkv -f null /dev/null" [17:27] ffmpeg -i my.mkv -map 0 -f null /dev/null [17:28] Hi, I am trying to capture video and audio from webcam with v4l2 drivers. I would like to capture one video stream with libaac and another audio stream with pcm_s16le separately from the same webcam source. Tried with: ffmpeg -f alsa -ac 2 -acodec pcm_s16le -i hw:2,0 -f v4l2 -vcodec h264 -r 30 -s 1920x1080 -i /dev/video0 -f alsa -ac 2 -i hw:2,0 -map 0:0 -c:v copy -c:a copy VIDEO+AUDIO.mp4 -map 1:0 -c:a pcm_s16le AUDIO_ONLY.wav gettin [17:29] c_14 I changed it to ffmpeg -i my.mkv -map 0 -f null /dev/null but still the same: Number of stream maps must match number of output streams [17:30] ffmpeg -f alsa -ac 2 -c:a pcm_s16le -i hw:2,0 -f v4l2 -c:v h264 -r 30 -video_size 1920x1080 -i /dev/video0 -map 0 -map 1 -c copy VIDEO+AUDIO.mp4 -map 0 -c:a pcm_s16le AUDIO_ONLY.wav <- zenny [17:34] c_14: thanks, but with your suggestion it produced "Could not find tag for codec pcm_s16le in stream #0, codec not currently supported in container" error in Input section and "Could not write header for output file #0 (incorrect codec parameters ?): Invalid argument" in the output section. [17:34] c_14 http://pastebin.com/gXADWAG6 [17:35] zenny: get rid of the -c:a pcm_s16le before -i hw:2,0 [17:36] rhagu: you're using libav, either see #libav for support or download a static build of ffmpeg/build from source. [17:36] http://johnvansickle.com/ffmpeg/ [17:36] c_14 thanks, I will google libav [17:37] c_14: with -c:a pcm_s16le before -i hw:2,0 also give the error as before (see http://pastebin.geany.org/EK4hR/) [17:38] zenny: mp4 doesn't support pcm_s16le audio [17:38] Either use a different codec or a different container. [17:39] ie switch -c copy to -c:v copy -c:a aac or .mp4 to .mkv [17:44] c_14: Thanks, this command worked: 'ffmpeg -f alsa -ac 2 -i hw:2,0 -f v4l2 -c:v h264 -r 30 -video_size 1920x1080 -i /dev/video0 -map 0 -map 1 -c:v copy -c:a libfdk_aac VIDEOnAUDIO.mp4 -map 0 -c:a pcm_s16le AUDIO_ONLY.wav' [17:45] c_14: what is the difference between '-map 0 -map 1' and '-map 0:0 -map 1:0'? [17:48] -map 0 maps the whole first file -map 0:0 maps the first stream of the first file [17:50] c_14: thanks a zillion for explanation, that is where I have been messing with. [18:29] anyone tried transcoding with ffmpeg and google native client? pnacl? [18:53] c_14: the stream capture with both and audio and video does not sync as the audio is always 4 seconds shorter than the video eventhough the streams are from the same webcam. Any clue? [18:54] Add -itsoffset 4 before -i hw:2,0 [18:54] might help [19:10] I think -itsoffset needs time stamps to work, so you'll probably have to use it on the output afterwards. [19:12] so, ffmpeg -i output -itsoffset 4 -i output -map 0:v -map 1:a -c copy newoutput [19:54] is mpeg-dash supported by all PC browsers? [20:02] Hello. I'm trying to extract a subtrack from an mp4, and I'd like it to be accurate (not just keyframe). I'd also like to avoid re-encoding the *whole* video to do this. What would be a reasonable approach for this? [20:02] My current thought is to encode a small piece at the beginning for accuracy (up to a keyframe), stream copy the rest, and put them together with the concat demuxer. Does this seem reasonable? [20:03] Kinda fiddly to do, but it should work. [20:04] Okay, thanks. I'm having trouble with the concat demuxer [20:05] when I stream copy an mp4, and re-encode a different mp4, I can't get them to concat [20:05] presumably I need to change some of the re-encoding settings so that the properties of the mp4s match. I'm not exactly sure what needs to be changed, tho. [20:14] So I guess now my problem is: I'm trying to concatenate 2 mp4s with the concat demuxer. I'm willing to re-encode one of them. What options do I need to use to make the concat demuxer work? [20:16] You're probably better using one of the other concat ways... Filter maybe. [20:17] hmm [20:17] that requires me to re-encode, right? [20:17] both pieces? [20:17] or no? [20:23] remux the mp4s to mpegts, concat those, remux back [20:25] kk [20:25] and I can stream copy when I copy to mpegs? [20:29] you can just concat the two ts files [20:30] oh, okay [20:30] so something like [20:30] ffmpeg -i vid.mp4 -codec copy -ss spot1 -t duration out1.ts [20:30] ffmpeg -i vid.mp4 -acodec copy -vcodec libx264 -ss spot2 -t duration out2.ts [20:30] then [20:31] Hello. I am trying to losslessly join some .mp4 files (which were originally mts, but converted with ffmpeg). However when I do 'ffmpeg -i "concat:file1.mp4|file2.mp4" -c copy output.mp4', output.mp4 contains only the video/audio from file1.mp4. Am I crazy? [20:31] I can just concatenate the two .ts files [20:31] rubidious: https://trac.ffmpeg.org/wiki/How%20to%20concatenate%20(join,%20merge)%20media%20files [20:32] then ffmpeg -i combined.ts -codec copy out.mp4 [20:32] something like that? [20:32] I'll try [20:35] does chrome on linux support smooth streaming? [20:40] c_14: thanks, I was not aware that the concat protocol only worked for certain formats [20:42] when I do something like: ffmpeg -i vid.mp4 -acodec copy -vcodec libx264 -ss spot2 -t duration out2.ts, the video and audio end up out of sync. [20:42] and when I stream copy I can't play the resulting .ts [20:42] is there something else I should do when remuxing to a .ts? [20:47] What I would do is do the video cuts separate from the audio cuts. [20:47] ie, cut the video, concat those (without any audio track) [20:47] right [20:47] then cut the audio track and mux it back in [20:47] then put the audio track back in later [20:47] okay [20:47] that makes sense [20:47] -an means no audio, right? [20:48] yep [20:49] Hello, i have the following command: ffmpeg -i URL -codec copy -f segment -segment_format flv -segment_time 10 -segment_list_size -segment_list test.flv test%04d.flv [20:49] thanks [20:49] skwaap: https://www.ffmpeg.org/ffmpeg-bitstream-filters.html - look at section 2.4 h264_mp4toannexb [20:49] It is possible to remove the timestamps from these segments so that the player will not end if it reads the first segment? [20:49] cause it reads the first segment and then stops because flv contains i think the timestamps data [20:49] with mpegts is fine but i want with flv [20:50] does chrome on linux support smooth streaming? [20:50] is smooth streaming delivered in a player? [20:50] any other way to deliver it? [20:52] blippyp: okay, thank you! [21:11] Hi guys [21:13] I have written a patch to fix issues with the latest ffmpeg from git and the changes to xcb shape inside ffmpeg that isn't compatible with the latest libxcb and breaks ffmpeg with the error message "xcbgrab.c:(.text.unlikely+0x52c): undefined reference to `xcb_shape_rectangles'" [21:14] when compiling [21:30] https://ffmpeg.org/developer.html#Contributing [21:38] c_14 any help to my problem plz ? :( [21:43] I know of know way to write flvs without timestamps, nor do I quite understand your problem. [21:50] Submitted my patch to the ffmpeg-devel mailing list. [22:46] Thanks to c_14 and relaxed for their inputs and I say goodnight. [23:20] is there a way to read a .mpls from \BDMV\PLAYLIST\ which is already on my hard drive? I only kept the BDMV folder [23:21] Try the bluray demuxer? [23:26] c_14, https://bpaste.net/show/e9fb6b969601 [23:27] It's looking for the index file? Do you have that somewhere? [23:27] no [23:30] Try passing it the mpls file directly? [23:33] anyone here? [23:33] hello [23:33] ping [23:34] c_14, https://bpaste.net/show/fe03e7472c1e [23:36] Kip: try -playlist 0 bluray:"foobar\Akira\" [23:39] c_14, https://bpaste.net/show/8a86104c768b [23:40] yay [23:40] Now just give it an output file. [23:40] eh wait [23:40] no [23:40] add a -i before bluray [23:41] I forgot about that. [23:46] hello, i updated to ffmpeg 2.4.2 and all the sudden my codec_describtor is NULL [23:47] From what version did you update? [23:47] from 2.2.2 [23:48] Did you read APIChanges? [23:48] c_14, https://bpaste.net/show/60608dee3076 [23:48] no, i didn't expect any major changes [23:48] did you remove the codec_describtor? [23:49] I didn't do anything, and in any case, shouldn't it be descriptor? [23:49] ye :) [23:50] i used to do the following: m_codecCtx->codec_descriptor->props ,but this doesn't work anymore all the sudden because the descriptor is NULL [23:51] Kip: I find the mix of forward and backslashes in the output delightful, not sure if that's an issue though. [23:54] feliwir: http://sprunge.us/UaTN <- first, I think it's called avcodec_descriptor and second the source says you should be using getters/setters [23:55] It should still be there though. [23:57] there is nothing called avcodec_descriptor [23:58] but i gonna try ur hint the getter [00:00] --- Mon Nov 24 2014 From burek021 at gmail.com Mon Nov 24 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Mon, 24 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141123 Message-ID: <20141124010502.A0DFE18A008C@apolo.teamnet.rs> [00:00] yeah, it's useful [00:00] though i thought we could do it somehow... [00:00] I got the feeling it could be generalized though [00:00] blend filters or something [00:00] check vf_blend [00:00] ubitux, yes, but that would imply some tweaking with timestamps [00:01] ah yeah indeed, i remember, it takes 2 streams [00:01] kierank: try adding a vsynth test then [00:01] at the moment we probably don't have filters dealing with consecutive frames from the same stream [00:02] those are used to test the bitexact output of video encoders [00:02] we could extend the filter and take a queue of frames, and compute a function of the previous frames [00:02] ok the asm matches the c version [00:03] doesn't match remuxing an existing file for some reason [00:10] on some samples it can be 4x faster [00:16] kierank, btw, you're doing 6 or 12 pixels at a time, right ? [00:16] 6 pixels at a time [00:17] does the code properly handle the tail (remaining pixels) ? [00:17] that part is in c [00:17] ok I haven't noticed that [00:17] i.e the asm does one line (apart from the end bit) [00:17] 1280 isn't a multiple of 6 so there are real world uses for non mod-6 [00:18] sure [00:18] and now I have seen the remainder of the code [01:26] ffmpeg.git 03Michael Niedermayer 07master:4b68edd08ac3: flvenc: Remove an unused variable [01:26] ffmpeg.git 03Martin Storsj? 07master:eec7f032a903: lavf: Remove a redundant include of sys/stat.h [01:26] ffmpeg.git 03Michael Niedermayer 07master:d899ea2017ef: Merge commit '4b68edd08ac352e314ae3fc701f90b081e549324' [01:26] ffmpeg.git 03Michael Niedermayer 07master:6d8dda169384: Merge commit 'eec7f032a903e06d249d1e8aa6630b65292bf40f' [01:40] ffmpeg.git 03Martin Storsj? 07master:f856d9c2f314: dashenc: Don't require the stream bitrate to be known [01:40] ffmpeg.git 03Michael Niedermayer 07master:9f0fd17c614f: Merge commit 'f856d9c2f314c493c672dfb9c876da182525da3d' [02:03] ffmpeg.git 03Michael Niedermayer 07master:65ce8f889553: avcodec/x86/Makefile: fix order [03:35] ffmpeg.git 03Lukasz Marek 07master:691f9be622c7: lavc/anm: fix mem leak in case of init failure [03:35] ffmpeg.git 03Lukasz Marek 07master:969382162f18: lavc/smacker: fix mem leak in case of init failure [03:35] ffmpeg.git 03Lukasz Marek 07master:bceabbdabab3: lavc/libvorbisdec: fix mem leak in case of init failure [03:35] ffmpeg.git 03Lukasz Marek 07master:f87a34486af0: lavc/libvorbisdec: use better error codes [06:53] ffmpeg.git 03Reimar D?ffinger 07master:458aadf8627c: lpc: Reduce stack usage by allocating LLSModel in context. [06:53] ffmpeg.git 03Reimar D?ffinger 07master:d0682b5eb0d9: svq1enc: reduce stack usage of recursively-called function. [06:53] ffmpeg.git 03Reimar D?ffinger 07master:6369a7b742bd: xface: Fix encoder crashes due to too small on-stack array. [06:53] ffmpeg.git 03Reimar D?ffinger 07master:dfc6f56c5a6a: xface: reduce table sizes. [06:53] ffmpeg.git 03Reimar D?ffinger 07master:cad3148ea3da: xface: reduce stack usage by directly storing 2 bytes data instead of pointers. [11:56] ubitux: just noting: https://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/2014-November/072772.html [12:13] wm4: yeah i just saw that [12:18] ffmpeg.git 03Cl?ment BSsch 07master:8ef46f4a0cd3: avcodec/microdvddec: support various broken form of color tags [12:19] that was quick [12:21] about the italic thing with / it's weird.. [12:21] should i support that as well? [12:21] i mean there is already a system for italic [12:21] it looks like another random markup convert from another format [12:21] "convert" [12:22] actually I've seen this often [12:23] do you have samples? [12:24] {3113}{3186}/W Londynie jest pewien reporter.|/Nazywa si Simon Ross. [12:24] thx [12:24] does the 'W' means anything? [12:25] no idea... maybe it's part of the text [12:25] is "Nazywa" a word or it's /N + "azywa"? [12:25] another one: {3618}{3669}/Facet pisze do Guardiana. [12:25] {4549}{4619}Je<9C>li za to bekn, Mark,|je<9C>li sprawa si pogorszy... [12:25] oops the <9C> etc. are from less [12:25] yeah i guessed so [12:26] so it seems / is strictly standalone [12:26] k [12:52] wm4: i support it only add the beginning of a line, right? [12:54] it can happen after | too (apparently) [12:56] yeah it's the beginning of a line [12:56] yes then it should be fine [12:58] mmh i need to support if it's written after the real tags too i suppose [12:58] like, {3618}{3669}{y:b}/fuck you sanity [12:58] and {3618}{3669}/{y:b}fuck you sanity [12:58] I have no example of such a thing [13:14] ffmpeg.git 03Cl?ment BSsch 07master:6e411d9cc7ec: avcodec/microdvddec: support non persistent italic marker '/' [13:14] wm4: i supported it anyway [13:15] nice [14:31] kierank: random example: http://git.videolan.org/?p=ffmpeg.git;a=blob;f=libavfilter/vf_lut3d.c;h=862dddeacf55bb0f934692eac52fde8e1361cf0c;hb=HEAD#l201 [14:31] was hoping it wouldn't look like that but ok [14:31] another solution is to make a function with if (bitdepth == 8) { ... } else { ... } [14:32] with av_always_inline [14:32] and then create wrapper function above [14:32] where did i do that recently... [14:32] ah, xbr [14:32] kierank: http://git.videolan.org/?p=ffmpeg.git;a=blob;f=libavfilter/vf_xbr.c;h=47e4b769ca233a0489967d01428a8f8ba07730a5;hb=HEAD#l325 [14:32] it's not bitdepth here [14:32] but you get the idea [14:33] the if (n == 2) { ... } else if (n == 3) { ... } ... aren't evaluated in the inner loop that way [14:42] kierank: and you could mix both too [14:43] last solution is #define BIT_DEPTH 8 and #include "v210enc_template.c" [14:43] but overkill for your case [15:19] ffmpeg.git 03Carl Eugen Hoyos 07master:600e38f5632e: Fix standalone compilation of the apng decoder on x86. [15:19] ffmpeg.git 03Carl Eugen Hoyos 07master:78093cf849b6: Print a warning if vp6 is muxed into flv: The output is flipped. [17:03] ffmpeg.git 03James Almer 07master:305b03097db3: lavf/apngdec: print currently unsupported in-stream tags in a more readable form [17:15] does encoding mpeg2audio/video work with ffmpeg at the moment? [17:15] seems like our application fails with the latest ffmpeg library also some green frames come up now randomly [17:26] just terrible this unstable API [17:31] afaik it's working, but if you think there's a bug open a ticket in trac.ffmpeg.org and provide a way to reproduce it (preferably with the ffmpeg command line tool) [18:18] is it possible to do a byte equivalent of CLIPW? [18:53] pmaxsb pminsb? [18:54] they are sse4.1 instructions, though [18:56] you may be able to emulate them using sse2 instructions [19:09] jamrial: is it available unsigned? [19:09] oh it is [20:54] ffmpeg.git 03Lukasz Marek 07master:331fae80a1fb: lavc/mss1: fix mem leak in case of init failure [20:54] ffmpeg.git 03Lukasz Marek 07master:4e9745fbff51: lavc/rv30: fix mem leak in case of init failure [20:59] if anyone is interested in figuring out why my 8-bit asm is broken please look: http://pastebin.com/sK5VDd6U [20:59] most of the luma is fine, just not the last 4 bytes out of 12 each time [21:07] you're using a shuf10 constant in the 8bit version. is that intended? [21:08] yes because the final shuffle should be the same [21:08] in both versions [21:25] Hi guys [21:27] I have a patch for the latest ffmpeg via git has a problem with libxcb and shape causing ffmpeg not to compile properly. (xcbgrab.c:(.text.unlikely+0x52c): undefined reference to `xcb_shape_rectangles') [21:31] typically we ask people to use gt to send a patch to the mailing list. [21:31] uh [21:31] that should be git [21:31] http://pastebin.com/egGkbj09 [21:31] BoRiS, commit and then use git format-patch to create a patch that can be applied as-is including the commit message etc [21:32] that can then be posted on the ffmpeg-devel mailing list for review [21:32] you can either use git send-email or otherwise make the content of the e-mail be the format-patch [21:33] git send-email is generally recommended since it doesn't do weird stuff to the content when sending [21:33] format-patch and copy-pasting into a mail client can lead to lolfails [21:43] jamrial: turns out it was the clipping mask [21:50] Submitted my patch to the ffmpeg-devel mailing list. [21:50] coal [21:51] ok, it's still not format-patch, but I guess someone might fix that for you when applying it :P [22:06] "fix: [22:15] Daemon404: 24-bit was claimed to work [22:16] it used to print a message saying it was unimpletented [22:17] http://git.videolan.org/?p=ffmpeg.git;a=commitdiff;h=ce6141259bb910aa7f580f55cf71e3c503f9a4fb [22:17] lol [22:18] Well, at least that isn't WMA lossless _encoding_ support! [22:18] lolwut [22:25] not even the 16bit version is truly lossless last i checked [22:41] huzzah, I got `git send-email` working! [00:00] --- Mon Nov 24 2014 From burek021 at gmail.com Tue Nov 25 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Tue, 25 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141124 Message-ID: <20141125010502.0E85C2AD6163@apolo.teamnet.rs> [00:01] ty c_14 that worked [00:23] Using AVStream.codec.time_base as a timebase hint to the muxer is deprecated. Set AVStream.time_base instead. [00:23] what is this when i use the -f segment [00:23] how can i fix it? [00:24] Ignore it. [00:24] Or try a recent git build. [00:24] Might have been fixed. [00:24] Can't remember. [00:24] Nothing important. [00:24] Well i see some different in the stream output when i use -f mpegts and when i use -f segment [00:24] thought it will be this [00:25] difference* [00:25] And a description of said "difference" [00:28] c_14, so is that a current limitation of the ffmpeg bluray protocol that it requires index file, backup, etc. and it cannot solely use the BDMV folder? [00:29] Kip: it looks to be a libbluray limitation. [00:29] At least from what I can see. [00:42] Hey guys, has anyone tried statically compiling ffmpeg with ffplay? Are including STL and any other libs specific to ffplay too much of an obstacle? Thanks. [01:23] can ffmpeg join 2 .h264 files into 1 single stream? both files were encoded with identical settings, resolution, etc.. [01:24] hotwings yes, by using concan [01:31] you should be able to just concat the actual files if it's raw bitstream, but not sure if there is any container around it [01:32] ok thanks, ill give that a shot [07:45] after upgrading my ubuntu, ffmpeg started saying "ffmpeg: error while loading shared libraries: libass.so.4" . How to make ffmpeg use the new libass.so.5 ? [07:48] vibhavsinha: abi change; you will need to recompile [07:49] vibhavsinha ln -s is not helps? [07:52] JEEBsv: cool. doing it right away. thanks [10:01] does ffmpeg use librtmp? [10:04] <__jack___> it can [10:05] <__jack___> see the configure --help : --enable-librtmp enable RTMP[E] support via librtmp [no] [10:08] __jack___: by default it usees it to output rtmp? [10:10] <__jack___> seems to no ([no]) [10:25] Does anyone know what is wrong with these commands? https://gist.githubusercontent.com/Renari/2c2e819ba0da0c9876aa/raw/cd65a4f2ce6cb88a81f4f7dd2f5feb2ea7d60cb3/gistfile1.txt [10:26] It's saying "Write error occurred on file "": The pipe has been ended. [10:26] This is while trying to use ffmpeg as an external encoder in virtualdub. [12:13] hi everyone [12:26] Hi, i?m new to ffmpeg and also ffserver, i?ve created a ffmpeg syntax to include several 3 audio inputs to 1 video (was made via map) . If you like, i can show you the code . We stremed it and it was working, but our provider has several issues so we try to use it with ffserver and do it by our self, unfortunately , i can?t find a example for a con [12:26] fig file with several audio, could someone please help? [14:58] when i run ffprobe on an mp3 i encoded from some wavs, i get a non-zero 'start'; as in the following string: "Duration: 02:01:02.88, start: 0.025056, bitrate: 64 kb/s". why would it not be 0.000000, what are the effects of that, and how can i fix it? [15:02] cyphase, probaby audio timestamps don't start at 0 so that's why you get that unmber [15:02] effects? probably none [15:02] fix it? I don't think you have to, but reencode should fix that [15:10] well, reencode just results in it happening again [15:11] Mavrik, also, i'm trying to encode a 11025hz 16k mp3 as cbr, but it's coming out vbr. any thoughts? [15:11] the same command with different sample and bit rates are cbr [15:12] probaby the encoder can't handle such a strange setting [15:12] 16k is... awfully low for mp3 [15:12] yea, i know. it's mono at least [15:13] it's a low-bandwith archive for a radio show [15:15] at those settings, HE-AACv2 will probably be miles better and more supported :/ [15:15] which mp3 encoder are you using? LAME? [15:16] yes, libmp3lame, but i have no preference if there's better [15:18] Mavrik, i may end up getting them to switch to AAC, but for now it's mp3 [15:28] Hi guys! [15:29] I have two mp4 files and both play fine: Sound and video are ok. [15:29] Now I want to concatenate them using this command: [15:30] ffmpeg.exe -f concat -i text.txt -c copy output.mp4 [15:31] In the resulting file, the part, which came from the second mp4 file, does no longer show the video. [15:31] Instead, I see a black screen with some green blocks, which are moving somehow strangly over the screen. [15:31] Thesound is still working correctly during this problem. [15:32] I have tried several different places, at which I did the cut and the problem happened at all of them. [15:32] well [15:32] What can I do to get a working result? [15:32] are your videos EXACTLY the same encoding wise_ [15:32] ? [15:32] use same SPS/PSS, same parameters, etc.? [15:32] Yes. I have recorded them with the same source,same settings, all should be identical: [15:33] Recorded with Windows Media Center (gives a wtv file), converted that to mp4 with ffmpeg, both times with the same settings... [15:34] hrmpf, alot of players very much don't like spliced video [15:34] especially MP4 one because MP4 wasn't really designed for that (also, segments support is very hit-and-miss) [15:35] The direct output file (wtv) basically contains mpeg2 video. Would it work better, if I tried concatenating that and _then_ converting to mp4? [15:36] Joergi, doing it all in one step with video filter will be by far the best solution [15:37] Do you have an example command? [15:37] I have already tried quite a few things and what I do here has always worked best... [15:38] What I have done til now is this: [15:38] ffmpeg -i 1.wtv 1.mp4 [15:38] ffmpeg -i 2.wtv 2.mp4 [15:39] Created a text file with both names in it (I can't pipe cause I'm on Windows) and then [15:39] ffmpeg.exe -f concat -i text.txt -c copy output.mp4 [15:49] I know checked the files again... [15:50] and it seems like they DO have one difference: [15:50] The Sample Aspect Ratio is different: [15:50] In the one file it is 16:11 while in the other it is 64:45. [15:50] Can that be a reason for my problem? [15:51] hi everyone [15:51] can anyone tell me how to get ffmpeg running 24/7 for streaming? does it have any memory concern? [16:09] you should have a look into OBE [16:10] open broadcast encoder [16:18] Hi everyone, i've a got a mp3 file with some metadatas, i'm able to see them with audacity, but if i try to check them with ffmpeg as : ffmpeg.exe -i" my.input.mp3", it just gave to me a few metadatas, in example : my metadata "Comment" is missing, have you got an idea to help ? where should i search ? thanks in advance [16:19] (if i missed something, feel free to tell me, i'm french and i guess my english is pretty poor ^^') [16:19] hey all :) here is one good for you: http://pastebin.com/rSEZy9cB - although it seems valid this causes a single core CPU to go nuts [16:21] actually no matter what i set - even the simplest video from a single image causes enormous CPU load [16:42] I want to change the Sample Aspect Ratio of a video, but _keep_ the Display Aspect Ratio. How can I do that? [16:50] use the setsar filter [16:56] is ffmpeg able to manage ID3v1 tag please ? [17:01] it should, iirc [17:01] Try ffprobe and various -show_streams or something [17:01] also, ffmpeg -h muxer=mp3 [17:04] c_14: I tried, but as noted at https://www.ffmpeg.org/ffmpeg-filters.html#toc-Examples-24 that also changes the DAR value and I want that value to stay! [17:05] Joergi: try setting both setsar and setdar [17:06] For example? [17:06] ok so another question ^^, how could you explain that ffmpeg is not able to find me all metatags stored in mp3 ^^ ? thanks :) [17:07] Joergi: you can also use a scale filter with iw*SAR:ih or somethnig along those lines. Can't remember the exact math. [17:07] could it be caused by differents ID3 versions ? [17:13] c_14: Setting -vf setdar=16:9 -vf setsar=16:11 does not work. It always changes _both_ values. [17:13] With other words: It does not take the first of the two settings into acount. [17:13] That's because it's ignoring the first -vf [17:14] -vf setsar=16:11,setdar=16:9 [17:15] Also does not work; same result as with my two -vf commands. [17:15] Yours changes the DAR to 20:11 [17:17] What resolution is your video? [17:17] 704x576 pixel [17:23] The setdar filter should set the correct sar for that automatically. what does setdar=16:9 give you as a sar? [17:24] 64/45. This is the old and incorrect SAR of the video [17:30] Hmm, that should only give you a dar of 1.73827 not 1.77778 [17:34] c_14: http://pastebin.com/qL47qLcx [17:35] use just -vf setdar=16/9 [17:36] I'm pretty sure you're not supposed to use a colon there... [17:37] If that doesn't work try -vf setdar=dar=16/9 [17:38] It seems to me that ffplay's 's' command (framestep) is incompatible with -skip_frame nokey option, am I right? [17:41] Both give the same result, SAR stays at the incorrect 64:45. [17:42] Joergi: you're using it with stream copy? [17:44] pmarty: I am using it like so: ffmpeg.exe -i 2.mp4 -acodec copy -vf setdar=dar=16/9 2neu.mp4 [17:45] hmm, that's a 720x576 frame; PAL video? [17:45] try using different output container (mkv) [17:45] if so, in order to get an exact 16:9 display, you have to crop to 704:576 first. [17:45] because of video standard sillyness :/ [17:47] ffmpeg.exe -i 2.mp4 -c copy 2neu.mkv [17:47] How would I dI that? [17:47] Di => do that? [17:47] Joergi: if you just want the correct sample aspect ratio, just do -vf setsar=16/11 [17:47] the display aspect ratio won't be exactly 16:9, but the video will be correct. [17:48] ffmpeg.exe -i 2.mp4 -c copy -vf setdar=dar=16/9 2neu.mkv [17:48] try this [17:49] https://en.wikipedia.org/wiki/Standard-definition_television#Pixel_aspect_ratio has a convenient list of the pixel aspect ratios (SAR in ffmpeg parlance) for the various PAL formats. [17:49] pmarty: Filtering and streamcopy cannot be used together. [17:50] That is all it says [17:51] Running [17:51] ffmpeg.exe -i 2.mp4 -vf setdar=dar=16/9 2neu.mkv [17:52] changes nothing - SAR stays at 64:45. [17:52] Am I right that changing SAR while keeping DAR is impossible? [17:53] dar = w/h x sar [17:54] so, yes they're connected [17:56] Joergi: there is certainly a way of changing dar with stream copy at the container level (not -vf) [17:56] I remember doing this [17:57] pmarty: iirc - I also remember doing this... [17:57] Joergi: -aspect 16:9 option [17:57] it will work for matroska, possibly others [17:57] that's probably it [18:29] pmarty, blippyp: My PC just had to crash [18:29] Did you write anything after "Joergi: -aspect 16:9 option" [18:30] ? [18:30] nope, have you tried using it? [18:31] Will do. The complete command would then be...? [18:33] I tried ffmpeg -i 2.mp4 -vcodec copy -acodec copy -aspect 16:9 output.mp4 [18:33] => SAR incorrect. [18:34] You said that SAR and DAR would be connected like this: [18:34] dar = w/h x sar [18:35] How can I then, with the same width, height and DAR, have two different SAR values? [18:53] ARGH!!! [18:53] Need something to bang my head against...afk [18:53] *bang* [18:54] *bang bang* [18:54] Some simple math shows what the problem was: [18:54] Although I used the same device with the same settings, the same programme and the same commands to modify the files... [18:54] ... [18:55] they had A DIFFERENT RESOLUTION [18:55] I cannot say why this is the case, but fact is the resolution is different. [19:54] Hello all, [19:54] I'm trying to create mpeg videos containing just 2 images, but I have many pairs that I need to do this to. Very few if any have sequential names, and it would be so much easier for automating if I could specify the input filenames for the ffmpeg command without relying on the image%01d.png type syntax. Is there some known way to cherry-pick images for a movie sequence in ffmpeg? Am I missing something in the docs? Thanks so much for any help you guys can [19:54] provide. [19:55] Something like: ffmpeg -f image2 -i '12-small.png' -i '921-small.png' -vcodec mpeg4 -b:v 800k -r 24 -y output.mov [19:57] gabriel: maybe something like ffmpeg -i 'first-image.png' -i 'second-image.png' -filter_complex 'concat=n=2:v=1:a=0' -vcodec mpeg4 -b:v 800k -r 24 -y output.mov [19:57] Let me try that [19:58] Action: kepstin-laptop wonders why you want a 2-frame 1/12 of a second long video, but whatever. [19:59] Im trying to use the mpeg encoder to compare how simillar two images are. So something like (mpegSize(x, y) + mpegSize(y, x))/(mpegSize(x,x) + mpegSize(y,y) -1) [20:00] That worked! Thanks so much [20:00] that's got to be the craziest idea i've ever heard of. [20:00] Hahaha crazy like a fox? [20:00] but sure, I suppose it would work [20:01] I got it from here: http://www.cs.ucr.edu/~eamonn/SDM_insects.pdf [20:41] Is there a way to make ffmpeg output the resulting fielsize in Bytes instead of KB? [23:21] I am trying to track down a problem with MP4 encoded with the latest ffmpeg, they play fine on my desktop browsers but not properly on BB10 or Andoid 5 [23:22] if i encode them with qtfaststart flag then it wont work at all, otherwise they play for a few seconds and then stop [23:22] its weird and cant find why it would happen [23:51] http://pastebin.com/DPWztTur [23:53] I'm extracting part of .wmv file like this: ffmpeg -i in.wmv -ss 1500 -c copy out.mkv . But then the pts of the first frame of out.mkv is at 1.7 sec (first frame of audio is at 0.03 sec). ffplay plays a first few frames in slow motion (A-V is negative). It hinders -f concat also. Why everything cannot simply start at 0? [23:53] I checked pts with (a)showinfo filter [23:56] TSM: has it worked with a previous ffmpeg? [23:59] llogan, no, i was running a version from back in feb previously [00:00] --- Tue Nov 25 2014 From burek021 at gmail.com Tue Nov 25 02:05:03 2014 From: burek021 at gmail.com (burek) Date: Tue, 25 Nov 2014 02:05:03 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141124 Message-ID: <20141125010503.23CCD2AD6163@apolo.teamnet.rs> [01:01] ffmpeg.git 03Michael Niedermayer 07master:8294f5042d5c: Changelog: remove 24bit lossless wma entry for now, some samples work others do not yet work [01:20] michaelni: do you still do daily merges from libav? [01:20] I have doubt about one of their patches from today [01:21] BBB, yes, which patch? i see no new commits in libav [01:22] the vp9 one [01:22] it might not have been pushed, I see it on the list [01:23] whats the subject of the mail/patch ? and what should be done with it ? should i ignore it? [01:23] [libav-devel] [PATCH 08/10] vp9: Use the correct upper bound for seg_id [01:24] Id ignore it [01:24] i merged the #define and ignored the rest a few days ago [01:24] oh ok, good, thanks [01:24] the define is fine yes [01:25] oh right I see its 2 days old [01:25] I dont check their list very often [01:25] thanks [01:25] np [01:28] heh, just glanced at their git and list to see if some DASH patches I'm waiting for are in yet (they're not) and was surprised to see one of my commits [01:28] (properly attributed, mind) [01:39] rcombs, if your work depends on / is blocked by some patches somewhere, you could include them in your patchset which you post to ffmpeg-dev. That assumes of course that these patches are "ready" and not unfinished [01:44] michaelni: I think they'll be getting a little bit more work/testing on libav's end before getting merged there pretty shortly, and I don't need the changes urgently, so I figure I'll just wait and let the process proceed as usual [01:44] these in particular, in case you're wondering: https://github.com/mstorsjo/libav/commits/dash-full [01:45] I think mstorsjo's planning to replace that last commit with something more complete first [02:05] rcombs : wbs (mstorsjo) is here too [02:30] ffmpeg.git 03Michael Niedermayer 07master:5182a2a235c3: avutil: remove FF_CONST_AVUTIL53, its no longer needed [04:17] ffmpeg.git 03Lukasz Marek 07master:efe34e87ebf5: lavc/libxvid: fix mem leak in case of init failure [04:17] ffmpeg.git 03Lukasz Marek 07master:02cb7d4c9c3a: lavc/smvjpegdec: fix mem leak in case of init failure [04:17] ffmpeg.git 03Lukasz Marek 07master:c9d39fc8c687: lavc/huffyuvdec: fix mem leak in case of init failure [05:14] ffmpeg.git 03Michael Niedermayer 07master:b4d8724ab28d: avutil/file: fix av_tempfile() documentation [05:34] Compn: thanks, I'd forgotten his nick (derp) [12:19] ffmpeg.git 03Vittorio Giovara 07master:9c12c6ff9539: motion_est: convert stride to ptrdiff_t [12:19] ffmpeg.git 03Michael Niedermayer 07master:ea41e6d63730: Merge commit '9c12c6ff9539e926df0b2a2299e915ae71872600' [12:54] ffmpeg.git 03Vittorio Giovara 07master:065923b0781b: mpegenc: prevent a NULL pointer dereference [12:54] ffmpeg.git 03Vittorio Giovara 07master:277ff7f5dc13: lavu: move internal define to the only places where it is used [12:54] ffmpeg.git 03Michael Niedermayer 07master:010adacbe236: Merge commit '065923b0781b06a2604f69f4e2c2407b7750a854' [12:54] ffmpeg.git 03Michael Niedermayer 07master:932d8d790ccf: Merge commit '277ff7f5dc134f1c2dfc4ea0ef3540340482e3d2' [13:15] ffmpeg.git 03Vittorio Giovara 07master:b99ca863506f: aacdec: avoid an out-of-bounds write [13:15] ffmpeg.git 03Vittorio Giovara 07master:30b8eb0f87b0: sol: simplify sol_codec_id() [13:15] ffmpeg.git 03Michael Niedermayer 07master:bcaef717ec0f: Merge commit 'b99ca863506f0630514921b740b78364de67a3ff' [13:15] ffmpeg.git 03Michael Niedermayer 07master:4fecf170d75d: Merge commit '30b8eb0f87b0eaefdc115ef38f8ad87dd3a6e50b' [13:24] ffmpeg.git 03Luca Barbato 07master:299d8ab104fb: cook: Make sure there is enough extradata [13:24] ffmpeg.git 03Michael Niedermayer 07master:1db2d39dfdda: Merge commit '299d8ab104fb350254eb2e6d9ecdce892a2a55b1' [13:37] ffmpeg.git 03Vittorio Giovara 07master:0562887a9843: tiffenc: initialize return value [13:37] ffmpeg.git 03Michael Niedermayer 07master:d2e0543766e0: Merge commit '0562887a984388fdc7a9b71c9374ff9c756fb4f1' [13:46] ffmpeg.git 03Michael Niedermayer 07master:55b59fab880a: roqaudio: Always use the frame buffer on flush [13:46] ffmpeg.git 03Michael Niedermayer 07master:f41d409b7a6c: Merge commit '55b59fab880a9fcdd30f97c5170af282087ac4f7' [13:50] ubitux: what happened to your patch that added a flag to demuxers whether a file is fully read? [13:58] ffmpeg.git 03Vittorio Giovara 07master:208f3abb9177: aacsbr: always initialize max_qmf_subbands [13:58] ffmpeg.git 03Michael Niedermayer 07master:96398cc12332: Merge commit '208f3abb917757743313da0da714e525e03159d2' [14:07] ffmpeg.git 03Martin Storsj? 07master:6f26f14f134e: sidxindex: Write mimeType=audio/mp4 for audio-only representations [14:07] ffmpeg.git 03Michael Niedermayer 07master:7a19a8fb696f: Merge commit '6f26f14f134e753d6168591f30815b1c08c1498b' [14:15] hi any openh264 patches around for ffmpeg? [14:15] why would you want to use openh264 via ffmpeg?? [14:19] patent fud probably [14:19] oh for encoding probably makes sense [14:19] it's openhevc that's not [14:27] wm4 for encoding only testing [14:40] A few people have asked that recently yet nobody seems to want it enough to provide a patch. [14:41] so, raw codecs don't output timestamps, but they're also not marked with AVFMT_NOTIMESTAMPS [14:41] is that as intended, or a bug? [14:44] ffmpeg.git 03Martin Storsj? 07master:fe42f94ce102: dashenc: Don't segment all video streams when one stream gets a keyframe [14:44] ffmpeg.git 03Michael Niedermayer 07master:9a5730966522: Merge commit 'fe42f94ce1023f9c2f7e86404c60afcee5b078a9' [17:29] ffmpeg.git 03Michael Niedermayer 07master:1d242f9816ac: avformat/mpegenc: assert that premux_packet is non null [17:29] ffmpeg.git 03Boris Reisig 07master:54170a33c2c9: avdevice/xcbgrab: fix undefined reference to xcb_shape_rectangles in xcbgrab.c [17:54] ffmpeg.git 03Rodger Combs 07master:39f247121ec6: ffmpeg: fix accurate seeking with -copyts [17:57] michaelni: can I get access to coverity? is it enough if I "ask" for access on the coverity site? [18:05] wm4, i just send you an invite [18:06] michaelni: thanks... seems I have access now [18:06] heh, there's not much [19:42] wm4: so i was wondering if it was correct to close the fd [19:42] because of later operations that could happen [19:42] and i think that's why i didn't applied it yet [19:42] yeah, I was worrying about that too [19:43] it's orthogonal though [19:43] the avio layer probably needs a callback for this [19:43] well you might not want to close it yourself either [19:43] so i'm not sure if that patch in itself really helps [19:45] why wouldn't I want to close it? [20:02] wasn't it the purpose of the patch? [20:02] like being able to liberate fd etc? [20:06] ubitux: yes, but you said I "might not want to close" [20:06] what's the reason? [20:37] wm4: the reason is that later in the framework we might want to access the pb again [20:41] like you see the force flushing option [20:41] for example [20:41] but this might be just for muxing [20:41] that's just an example [20:41] i'm just not that comfortable doing that change yet [20:43] you don't need to include any actual closing [20:44] and the sub demuxers definitely don't read more data after opening [20:45] can we be sure the framework isn't peaking for $randomreason? [20:48] well you just return EIO or EOF in this case, and probably nothing bad happens [20:49] but I coul check utils.c for possible issues [20:49] *could [20:50] yeah see [20:50] post read_header() [20:50] you have an avio_tell() [20:50] it's checking if the pb is set though [20:50] ffmpeg.git 03wm4 07master:f41cf2e09ec2: avformat/rawdec: raw formats have no timestamps [20:50] but well, might want to check a bit more [21:03] damn, a v9 [21:04] i wonder when it's going to get upstreamed [21:04] :D [21:05] get hype [21:06] i think he said he wanted to post a v9 with a important bugfix he had and then start splitting [21:10] i think he said that for v7 or so [21:16] ffmpeg.git 03Benoit Fouet 07master:9e1cfbd38a18: avformat/apngdec: transmit all the chunks between consecutive fcTL ones. [22:10] ffmpeg.git 03Michael Niedermayer 07master:e6ea75c5e6b6: doc/examples/decoding_encoding: fix storing all channels [23:06] Gramner: any comments on the new v210 asm sets? [00:00] --- Tue Nov 25 2014 From burek021 at gmail.com Wed Nov 26 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Wed, 26 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141125 Message-ID: <20141126010501.CA7EC2AD61A2@apolo.teamnet.rs> [00:00] the odd thing is how it plays on the BB10, it starts, knows the length of the video 27s then stops after 7 seconds, if i take qtfaststart off then BB10 complains of a codec error [00:01] another odd thing, if i do not send a Content-Length header to the device over html then the video plays correctly [00:01] all this time it works fine on desktop chrome/firefox [00:02] I have done webinspection on the BB10 and it always downloads the whole file before playing it so very odd [00:04] TSM: are you playing it "directly" on the device (uploaded to device and play with native player) or via a browser, or some other way? [00:06] via browser, if i save it then play it all is fine [00:07] the browser on the BB10 offloads the playing to the video app [00:07] it is odd [00:13] TSM: i don't know what to suggest here other than trying a simpler command using defaults and seeing if any of your options cause the issue [00:13] omit audio, etc [00:15] unrelated to the issue, but you have support for libfdk_aac, so why are you using the native AAC encoder? [00:20] dont know about the aac, i read about libfdk so based on your question it seems like it should be the better encoder [00:21] i wish the x264 encoder was faster but i suspect that is because of my lack of CPU extensions that some of the newer procs have [00:24] yes, libfdkaac is better. you could use a faster preset [00:26] superfast is a awfull on the output at 1M, fast seems to be the best tradeoff, im also using baseline to make sure that it will play on andrioid devices [00:26] i still average about 30-35fps encode speed with watermark,drawtext,scale [08:57] when we say MPEG-DASH server does it mean a plain HTTP Server that servers the MPD file? [10:04] I'm trying to configure urxvt. I've xrdb ~/.Xresources and now urxvt doesn't show with an interactive terminal but only a blank screen. [10:04] oops [10:05] any one care to assist me in assisting another user in setting up libnvenc with ffmpeg? [10:07] https://clbin.com/GFc2C << what we're getting. [14:01] Hi, I try to stream audio from a decklink card (SDI) to an audio loopback but I keep getting silence. If I stream to a wav it works fine, anyone knows what I could be doing wrong? [14:02] My commands are: `ffmpeg -f decklink -i 'DeckLink SDI (1)@8' -f alsa hw:0` and `ffmpeg -f alsa -i hw:0 foo.wav` [14:02] `ffmpeg -f decklink -i 'DeckLink SDI (1)@8' foo.wav` works fine though [15:44] morning/afternoon/evening having a real difficult time compiling ffmpeg with libnvenc enabled in Arch using a PKGBUILD. anyone else got ffmpeg to compile with libnvenc enabled? [15:46] ffmpeg does not have any nvenc support. Ask the guy who made that package what he did there. [15:48] BtbN, oh, so apparently he patched ffmpeg to be able to use nvenc? [15:49] curious why ffmpeg wouldn't want to utilize the onboard encoder chip of kepler and up GPU's [15:49] Quite surprising though, when i made working nvenc support for ffmpeg months ago, nobody cared. Now it's big news. [15:50] nvidia just released a driver that supports it like a couple fridays ago as far as i'm aware [15:50] I am using it since months [15:50] it worked since forever, they just wrote it in their changelog for some reason [15:51] ok, as i said for us normal users who don't compile our own nvidia kernel modules it just came available to us [15:51] oh ok [15:51] https://github.com/BtbN/FFmpeg/tree/nvenc could use an update to current master though [15:51] in your experience does nvenc provide any better quality than libx264? [15:52] no [15:52] just like any other hw encoder [15:52] x264 is just unbeatable in terms of quality per bitrate [15:52] hmmm [15:52] But hw encoders use no cpu power and nearly no hardware resources [15:53] the advantage i see is taking all the encoding load off the cpu for those that don't have the best cpu but have a kepler or higher gpu right? [15:53] BtbN: doesn't HVEC/x265 outperform x264 by double-digit-percentages in that department? [15:54] what? [15:54] x264 is just unbeatable in terms of quality per bitrate <--- responding to that [15:54] Nobody was talking h265, let alone h265 hw encoders. [15:54] ribasushi, i don't think rtmp services like twitch support x265 yet though [15:54] does youtube even support it yet? [15:55] twitch has trouble because users struggle to decode 720p with 60 fps because of too much cpu load. They don't even dare to think about h265. [15:55] lol [15:55] they don't, again I was just responding to BtbN's line in isolation [15:56] BtbN, so ffmpeg isn't going to be adding support for libnvenc? [15:56] I have no idea what libnvenc even is [15:56] It's not the nvidia nvenc library. [15:57] That's nvEncodeApi [15:58] well now i'm thoroghly lost. lol the people who have ffmpeg working with libnvenc enabled say they got ffmpeg to work by using --enable-libnvenc per here https://github.com/Brainiarc7/ffmpeg_libnvenc [15:58] Using that fork propably [15:58] which isn't even a fork, because only 4 commits [15:59] BtbN, ok, well he got it from this guy's github which actually adds the support for libnvenc https://github.com/agathah/ffmpeg_libnvenc/commit/3b5a7bdccd5ed6b4189f596549fb300e3d3fd6b1 [16:00] Again, i have no idea what libnvenc even is [16:01] BtbN, looking at that commit it's utilizing the nvEncodeAPI [16:01] most likely nvidia gpu accelerated encoder [16:01] Nope [16:01] That library has a diffrentname [16:01] -" " [16:02] libnvidia-encode.so [16:02] libavcodec/nvencoder.c [16:02] libavcodec/nvenc.h [16:03] That's just the avcodec part. [16:03] BtbN, maybe i should start over. :) i would like to be able to utilize my GTX 760 hardware encoder to encode video/audio with ffmpeg, how would I do that? [16:04] doc/examples/libnvenc.c [16:04] There is no supported way for that currently [16:05] it appears that he used the windows nvenc sdk, reviewed the Sample example and then implmented his own libnvenc encoder type for ffmpeg. is that a fair statement? [16:05] BtbN, apparently there is a way since it's been done by 2 people using ffmpeg [16:06] Not using the official version of ffmpeg, but some fork. [16:06] BtbN, right, he started with RELEASE NOTES for FFmpeg 2.4 "Fresnel" [16:07] BtbN, he did it back on Oct 24. https://github.com/agathah/ffmpeg_libnvenc/commits/master [16:10] BtbN, i'm a little confused on naming of these things. I don't fully understand C+ so i'm clearly in way over my head. you call it libnvidia-encode.so where he calls it libnvenc. the name is irrelavant isn't it? both use a dedicated SIP block for accelerated video processing, and as such, is independent of CUDA cores. [16:10] He doesn't call it like that, he _links_ against that [16:11] so it's not the same thing [16:12] ok but again i'm just trying to understand why he was able to get NVENC encoding enabled with his ffmpeg but i'm having troubles [16:13] my config.log is here https://clbin.com/8i0O7 [16:13] specifically common.mak:49: recipe for target 'libavutil/opencl.o' failed [16:31] looks like it's getting hung right here libavutil/atomic_gcc.h:31:5: warning: "HAVE_ATOMIC_COMPARE_EXCHANGE" is not defined [-Wundef] [16:48] is there currently plans to enable nvEncodeAPI into ffmpeg? [17:11] BtbN, i see you were the one who worked on the nvenc implmentation in windows. [17:11] BtbN, for OBS [17:40] yes [17:47] Hi, with the following webcam "uvcvideo" 3.17.0 capabilities : 0x84000001 [VIDEO_CAPTURE,STREAMING,(null)] [17:48] v4l2 or guvcview didn't find any standard [17:49] According to guvcview, the frame rate of the picture is changing on the fly when the webcam is streaming. [17:50] My goal is to record the video from the webcam together with the audio from JACK (jackdbus). [17:50] I don't know if it is possible with ffmpeg. The nearest I get is with the command: [17:51] ffmpeg -f jack -ac 2 -i jack -f video4linux2 -s 640x480 -r 8.5 -i /dev/video0 -qscale 3 -r 30 out.mpg [17:52] framerate changing on the fly is pretty common with webcams, particularly in the dark; they increase the exposure time pretty dramatically to make the image visible [17:52] but I get 2 issues: 1) the aaudio is not in sync with the video, and 2) the frame rate of the video is not 100% correct. [17:53] do not use -r as an input option, that's what's breaking the video framerate and sync [17:53] you can use the "-framerate" option to request the webcam to give you some framerate [17:54] but you'll probably get variable framerate output, that's usually ok. [17:54] note that with most webcams, you can get higher framerates by using mjpeg (or h264 if supported) input format; the limit when using raw video is sometimes the usb bandwidth. [18:03] With "ffmpeg -f jack -ac 2 -i jack -f video4linux2 -input_format h264 -framerate 3 -i /dev/video0 -qscale 3 -r 30 out.mpg", ffmpeg quit with " Cannot find a proper format for codec 'h264' (id 28), pixel format 'none' (id -1) [18:03] Assertion *codec_id != AV_CODEC_ID_NONE failed at /var/tmp/portage/media-video/ffmpeg-2.2.9/work/ffmpeg-2.2.9/libavdevice/v4l2.c:805" [18:06] well, then your webcam probably doesn't support it (you can check with the -list_formats all input option) [18:09] (as an additional note, the webcam will usually round up the "-framerate" option to its nearest supported framerate; e.g. the one in my laptop only supports 15 and 30fps) [18:12] Thanks, mine use a very low framerate, around 3 fps, 10 fps as very short peaks [18:12] Also, with -list_formats, ffmpeg bail with "/dev/video0: Invalid argument" [18:13] It look to be consistent with v4l-info output which find no standard at all [18:15] v4l-info find Video_capture "YUV 4:2:2 (YUYV)", 640x480 [19:27] How do you embed idx/sub file into mkv? I know how to do it with .srt but this idx/sub Im not sure which file I should use as the input [19:32] trying to complile ffmpeg with libvnenc enabled using a ffmpeg fork located on github but it's getting hung up on opencl, here's the config.log if anyone can help please. https://clbin.com/flKPy [19:35] ubuntuaddicted: we do not support forks here [19:36] llogan, ah that sucks. so if I try to compile regular ffmpeg and it fails at the same opencl error then you'll help me? [19:37] if i know an answer, but i probably won't since i know nothing of opencl [19:40] would opencl be required for software like kdenlive (a video editor) or simplescreenrecorder (a video capture program) or obs-studio (a livestreaming software) [19:41] or better yet, what is opencl used for? i guess i don't even know if I want it enabled. lol [19:42] for ffmpeg specificially, nothing worthwhile [19:44] i see that x264 can use it for lookahead per this http://git.videolan.org/?p=x264.git;a=commit;h=3a5f6c0aeacfcb21e7853ab4879f23ec8ae5e042 [19:46] you would have to test to see if your encodings would benefit from it or not [19:46] llogan, right. it's just that i can't compile ffmpeg because it appears like during ffmpeg compilation it can't find some opencl headers. [19:48] IIRC, opencl in ffmpeg is optionally used only by a few filters [19:49] llogan, i see it's used by x264, motionsearch, downscale etc etc. according to this anyway http://git.videolan.org/?p=x264.git;a=commit;h=3a5f6c0aeacfcb21e7853ab4879f23ec8ae5e042 [19:50] yes, that's x264 [19:57] ok, so x264 encoder uses opencl, or should I say "can" use opencl for some things like lookahead etc etc [20:25] i'm just hung up on my opencl, i see some header files located in /opt/cuda/include/CL/opencl.h AND /opt/cuda/include/CL/cl.h [20:25] but also /usr/include/gegl-0.2/opencl/cl.h [20:37] interlacing in 2014 > http://paste.debian.net/plain/133499 < do "we" agree? [20:39] Action: kepstin-laptop agrees [20:40] obviously european-centric, americans have to switch all the 25 to 24 or 30, and the 50 to 60 :) [20:40] kepstin-laptop: any problems that you see, or would you agree 100% ? [20:40] yeah, its a short doc for a specific tv station in eu [20:40] main issue is that getting cameras and production equipment that can do 50/60p is more expensive. [20:41] well, then shoot 25p, isnt that obvious? [20:41] for some reason ffmpeg isn't seeing my libOpenCL headers. [20:41] ubuntuaddicted: do you actually mean ffmpeg? or are you having issues with x264? [20:41] kepstin-laptop, ffmpeg but it's a fork so not sure you want to help me. :( [20:42] ubuntuaddicted: you don't want x264 built with opencl anyways, it apparently makes it slower and lower quality. [20:42] kepstin-laptop, im trying to get ffmpeg built with nvEncodeAPI capabilities. [20:43] hmm. there are very limited use cases for wanting to use the hardware encode blocks on graphics cards, pretty much limited to "i want to do realtime streaming, and my cpu is already being used for other stuff" [20:43] kepstin-laptop, yeap, that's it [20:43] but if that fits your use case have fun, but I don't know anything about it :) [20:45] but either way, opencl shouldn't be involved, since opencl means using the graphics card compute resources, so it would e.g. compete with gaming performance. [20:45] kepstin-laptop, trying to utilize this https://github.com/Brainiarc7/ffmpeg_libnvenc but for some reason when I try to configure ffmpeg it's not finding my nvidia.icd [20:46] kepstin-laptop, my last config.log https://clbin.com/9ykup [20:47] yeah, you should really try to contact the people working on that repo directly for that. Unless you get lucky and they're hanging around here... [20:48] Hi [20:49] kepstin-laptop, well i have the opencl-headers20 installed on my machine. i also have cuda installed. so i'm not sure why it's failing [20:49] How to merge 2 avi file [20:50] mam_: please describe what you mean by merge. do you want a single file that plays the two videos in order? do you want one overlayed over the other but both playing at the same time? [20:50] mam_, this may help http://unix.stackexchange.com/questions/43896/is-it-possible-to-merge-video-files-using-cat [20:51] mam_: ffmpeg has an actual faq entry for this: http://www.ffmpeg.org/faq.html#How-can-I-join-video-files_003f [20:52] Thanks I will Chack [21:15] Hi I checked only first video show [21:23] ubuntuaddicted: I checked only first video showed [21:24] mam_, not sure sorry. personally i use a linear video editor for joining video files but thats overkill for what you want to accomplish and i'm not sure how to accomplish it using ffmpeg and the command line. sorry [21:56] Hi there! Instead of writing a stream to a file, I'd like it to output to stdout. I see in a lot of online documentation that I can use "-" as the output parameter to accomplish this, but whenever I try I get the following error: [NULL @ 0x1be9240] Unable to find a suitable output format for 'pipe:' [21:58] http://pastebin.com/39sTt9E4 [21:58] MamuDingo: when outputting to a pipe, you have to specify the -f output option, since it can't guess from the filename [21:59] Ah ha... Okay, I'll try that. Thanks! [21:59] note that not all output formats can be sent to a pipe; for example, mp4 doesn't support it. [22:00] Is there any way to tell which formats support piping? [22:00] hmm, I don't actually know of any good lists. You can always try and see, it'll give an error if it's not supported. [22:01] Well, trial and error is definitely one way. :) [22:01] Thanks, again! [22:04] rawhide seemed to work for what I needed to do. [22:04] *rawvideo ... autocorrect. :( [22:05] Cut 'em out, ride 'em in Rawhide! [22:08] hello i need parameter for ffmpeg: add srt in video but want choose font : dejevu.tiff , but how? [22:10] "ffmpeg -i video.avi -vf subtitles=subtitle.srt out.avi" this is OK but i want chose font.tiff how...??? [22:20] best to just wait until someone in the channel can answer your question. [22:29] gokuh, i think your best bet is to convert the .srt subs to .ass and set a font there (for example with aegisub if you have a GUI) [22:44] klaxa: THANK YOU i try to do convert ok? thank you [23:18] Hello Everyone i need some help with video filter configuration in ffmpeg. I wanne show a picture if on the udp input of ffmpeg is no signal [23:19] the output of ffmpeg is published to a rtmp server [23:51] Got to say, there's one thing that I don't understand about FFMPEG: It sometimes brings my PC to ~100% utilization, but it remains completely responsive. o.O [23:51] That's some magic. :p [23:52] You can blame your Operating System's load management policies. [23:52] Or, thank? [23:52] What, you actually want a responsive system? [23:52] Ideally, yes. :p [23:54] I often need to babysit processes otherwise my machine stutters like mad. Not so with FFMPEG. Not even aware it's still running if I don't look at PowerShell or my CPU charts. [23:57] most of the time i can tell from my fans it's running [23:57] You mean the room heater? [23:58] i only have laptops, they don't heat up rooms very well [23:58] It's all about placement. [23:58] Pfft. Who wants more than a 35W block heater? :p [00:00] --- Wed Nov 26 2014 From burek021 at gmail.com Wed Nov 26 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Wed, 26 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141125 Message-ID: <20141126010502.DF9902AD61E5@apolo.teamnet.rs> [01:12] im having trouble using a new software called stream-m [01:13] has anyone heard of this? [01:15] im using a method called concat and its causing me to use 90% cpu usage when its looping [01:18] ok [01:18] well i can't see anyones replies [01:25] ok [01:26] Why would you come here for help with other software? [01:28] _Darnley it says for the development of ffmpeg , and the development of other software that uses ffmpeg clearly in the documentation [01:29] This software stream-m uses ffmpeg [01:30] This is the channel for discussion of FFmpeg development only. user support is at ffmpeg-user IRC and mailing list, but we only support FFmpeg stuff, and not 3rd party stuff (including stream-m) [01:30] im having trouble with the concat of a stream i beleive its causing me to use 90% CPU usage i was just wondering if anyone ran into a similar problem when it loops using this method... [01:31] So give us some ffmpeg command lines and terminal output, then maybe we can help you. [01:31] ...so you'll have to get help from stream-m, or if you are using ffmpeg directly then use the mailing list or IRC channel. [01:31] ok sorry llogan maybe i inturpreted the doucmentation wrong im just looking for a little help i'll search there [01:31] (ffmpeg-user that is, and not here or ffmpeg-devel mailing list) [01:35] Thanks again [01:55] ffmpeg.git 03Michael Niedermayer 07master:03a17f2bbf36: avcodec/mjpegdec: Print the number of bits in the unsupported pixel format error [01:55] ffmpeg.git 03Michael Niedermayer 07master:172d22a071b5: avcodec/mjpegdec: Add YUVA420 formats to *scale asserts [02:12] ffmpeg.git 03Uwe L. Korn 07master:40665d27e38e: flvdec: Document how the duration is retrieved at the end of the file [02:12] ffmpeg.git 03Michael Niedermayer 07master:b83beb131f0d: Merge commit '40665d27e38e6a2f65037878202bd1a398c7683e' [02:48] ffmpeg.git 03Martin Storsj? 07master:b9d08c77a443: lavf: Don't try to update files atomically with renames on windows [02:48] ffmpeg.git 03Michael Niedermayer 07master:71ecfcf2d31e: Merge commit 'b9d08c77a44390b0848c06f20bc0e9e951ba6a3c' [04:04] ffmpeg.git 03Martin Storsj? 07master:960aff379da4: lavf: Use wchar functions for filenames on windows for mkdir/rmdir/rename/unlink [04:05] ffmpeg.git 03Michael Niedermayer 07master:ecfafc5f2b63: Merge commit '960aff379da46dcaff61504a57714d4d4e758e41' [04:05] ffmpeg.git 03Michael Niedermayer 07master:60420fa3dc88: avdevice/dshow: fix build, ensure that feature enable #defines are set before includes [04:05] ffmpeg.git 03Michael Niedermayer 07master:c724c82a7828: doc/print_options: Fix build on mingw after 960aff379da46dcaff61504a57714d4d4e758e41 [04:19] ffmpeg.git 03Carl Eugen Hoyos 07master:0e3ea5b28a32: Include stddef from snow_dwt.h. [04:19] ffmpeg.git 03Carl Eugen Hoyos 07master:20f3cdf0c894: Include config.h from huffyuvdsp.h. [04:19] ffmpeg.git 03Michael Niedermayer 07master:67dbf8bec977: Merge remote-tracking branch 'cehoyos/master' [04:38] michaelni: Any chance you could commit the attached patch for this bug https://trac.ffmpeg.org/ticket/4130 [04:56] ffmpeg.git 03Carl Eugen Hoyos 07master:1f3d4788983d: avdevice/decklink_common: Fix "Cross-compiling FFmpeg on Debian for Windows with MinGW-w64" [04:56] Zeranoe, applied [05:01] michaelni: Thanks [11:17] ffmpeg.git 03Vittorio Giovara 07master:bc75b64cff37: vc1pred: remove logically dead code [11:17] ffmpeg.git 03Michael Niedermayer 07master:0aa208837e24: Merge commit 'bc75b64cff37d58f3944e2da3da45c37f35f019a' [11:55] ffmpeg.git 03Michael Niedermayer 07master:c117da9d3e0d: lpc: remove unneeded {} [11:55] ffmpeg.git 03Michael Niedermayer 07master:62e52b94e684: vorbis_parser: Move vp check to avoid a null pointer dereference [11:55] ffmpeg.git 03Luca Barbato 07master:b67138598ce1: vc1: Simplify a little setting the intra variables [11:55] ffmpeg.git 03Michael Niedermayer 07master:a63ec9d5737a: Merge commit 'c117da9d3e0db7dc311d817054988364b3ef4587' [11:55] ffmpeg.git 03Michael Niedermayer 07master:bcccb2c2986b: Merge commit '62e52b94e684491dfc5a6b7ca688bb86f7cd0f3f' [11:55] ffmpeg.git 03Michael Niedermayer 07master:ed25ca166171: Merge commit 'b67138598ce158e3083f6295a27b63e2065d5ecb' [12:06] ffmpeg.git 03Luca Barbato 07master:d25afb579fac: vc1: Set the is_intra bitfield to all 1 when needed [12:06] ffmpeg.git 03Michael Niedermayer 07master:79f77ce31c6c: Merge commit 'd25afb579facc83fd3a839f21411124d0b09f0ba' [12:26] ffmpeg.git 03Michael Niedermayer 07master:51946d2de8bd: vc1: Use the correct shift amount [12:26] ffmpeg.git 03Luca Barbato 07master:16158da9607f: hnm4: Use av_image_check_size [12:26] ffmpeg.git 03Michael Niedermayer 07master:0e85a28fa434: Merge commit '51946d2de8bd4a4aada43b6ab41340b0f5eb4ecb' [12:26] ffmpeg.git 03Michael Niedermayer 07master:ba1a19bc35e4: Merge commit '16158da9607f2f84232d3dd381406b2f2449ec74' [12:44] ffmpeg.git 03Michael Niedermayer 07master:57ed5a64feec: hnm4: change width/height to int to fix hypothetical integer overflows [12:44] ffmpeg.git 03Michael Niedermayer 07master:ca5c3ff90972: vf_interlace: x86: improve asm performance [12:44] ffmpeg.git 03Michael Niedermayer 07master:8c9945285e30: Merge commit '57ed5a64feec4af1f16f9a74c63cfa9aa8147242' [12:44] ffmpeg.git 03Michael Niedermayer 07master:3fe3c8abb15a: Merge commit 'ca5c3ff90972a5c97aabda2ace57ba72dcd7d83b' [12:44] ffmpeg.git 03Michael Niedermayer 07master:ca59b5b6eceb: avfilter/x86/vf_interlace: remove redundant instructions [13:35] ubitux: went through utils.c; I wonder why everything checks whether pb is NULL (except some code doesn't) [13:36] ubitux: I don't think the API allows the demuxer to close the pb at all, if that's what you planned [13:37] even setting to NULL might not be fine (due to the API) [13:43] <__gb__> hi kierank, you can use vaapi without rendering, yes -- even without a display server actually (using the VA/DRM backend) [13:45] <__gb__> and, if you have a kernel version recent enough (>= 3.15 to have that feature enabled by default), then you can use render nodes too for using vaapi remotely through ssh, even if another user has locally started a display server [13:46] <__gb__> if you don't want to go over elaborated libudev stuff, you could just open /dev/dri/renderD128 -- that's the render nodes companion to /dev/dri/card0 [13:47] <__gb__> and provide the resulting fd to vaGetDisplayDRM() and use the resulting VADisplay wherever needed [14:10] ffmpeg.git 03Michael Niedermayer 07master:0eecf40935b2: avcodec/mjpegdec: Fix context fields becoming inconsistent [14:56] ffmpeg.git 03Michael Niedermayer 07master:3d5d95db3f5d: avcodec/utils: Check that the data is complete in avpriv_bprint_to_extradata() [15:28] ffmpeg.git 03Michael Niedermayer 07master:8cd80b5fcbfa: avformat/jacosubdec: Cleanup when avpriv_bprint_to_extradata() fails [15:48] "Getting board?" [15:48] Action: wm4 facepalms [16:11] michaelni: I think ecfafc5f2b632780dd310def1d6b3a2c10565465 or 960aff379da46dcaff61504a57714d4d4e758e41 introduced a new compiling error http://paste.ubuntu.com/9233871/ [16:18] Zeranoe, that looks like a C++ error... weird [16:22] Daemon404: More context: http://paste.ubuntu.com/9234011/ looks like decklink_enc.cpp is being included [16:25] Zeranoe, does a cast to wchar_t* fix that ? [16:26] michaelni, a cast is the wrong fix [16:26] this is not a valid error in C [16:27] well it's compiled as C++ apparently? [16:27] that seems majorly wrong [16:27] these are just stubs for win32 [16:27] so nobody cares whether they're inline [16:27] same with win32 pthread wrappers [16:28] doesnt matter, it can still be done right in C++ [16:28] with the new operator? [16:28] let me sort it [16:28] also [16:28] #include "libavformat/internal.h" [16:28] in libavdevice [16:28] yum yum [16:29] apparently this is "ok" [16:29] because it's a compile time dep only [16:30] i think vfwcap.c is misleading [16:30] Zeranoe, were you building multithreaded? [16:30] i think ti *actually* failed in the .cpp [16:32] ok so... [16:32] https://git.videolan.org/?p=ffmpeg.git;a=blob;f=libavdevice/decklink_enc.cpp;h=6c5450f4ec5f36d1b397737b5dde9063d5ec6acc;hb=HEAD#l29 [16:32] what the fuck g++ [16:32] you should not be doing that when the func is under extern "C" [16:32] uh [16:32] extern C doesn't mean it's C code [16:32] extern C only controls linkage [16:32] nothing else [16:32] just that C mangling is used [16:33] common misconception [16:33] i see [16:33] that's kinda lulz [16:33] also you could have tortured the contributor into writing that file in C [16:33] i figured it would be smarted than to try and inline C into C++ [16:34] because inline is a hint to the compiler [16:34] not a guarantee [16:34] the standard can probably not assume that every C++ compiler cna also speak C [16:34] while turning off name mangling is easy [16:34] Daemon404: I am, pthread [16:34] Zeranoe, not what i meant, but its not relevant now [16:35] i guess a cast it is [16:35] c++/c interop sure is fugly [16:35] Daemon404: Yes to what you meant, make -j4 [16:37] https://github.com/BtbN/FFmpeg/commit/13b5cceea6300e36c2d8d7d6b8b2cfacbcdc5dc1 Is this the correct way to add a dependency on an external header file? [16:37] This configure script is extremely confusing and way too much stuff happens automagicaly based on naming [16:38] I'm still not sure if it would propperly disable my code if nvenc is disabled [16:41] Seems like it does, but i have no idea why [16:41] It also doesn't seem to matter that in allcodecs there is no compile-time removal of anything [16:42] Daemon404, michaelni: Do you need anything else from me on this? [16:42] Zeranoe, no, the fix is clear [16:43] Hm, what qualifies a library to be flagged as non-free? As far as i know, linking ffmpeg with nvenc doesn't imply any redistribution restrictions. You just can't include its header in ffmpeg. [17:01] Zeranoe, a tested patch would be best but i can try to blindly fix it and maybe after a few tries we will have found all cases th [17:01] that need changing [17:04] michaelni: I would be happy to test any patchs [17:17] ffmpeg.git 03Michael Niedermayer 07master:d0879a93eac6: avformat/os_support: try to fix build when included from a c++ file like libavdevice/decklink*cpp [17:18] Zeranoe, pushed a fix for that one line, does it build now or are there more issues ? [17:21] michaelni: running with a single make job now [17:43] michaelni: Everything seems fine now, thank you for the prompt fix [17:44] np [18:02] ffmpeg.git 03Michael Niedermayer 07master:2f1de5ca139f: avcodec/huffyuvdec: apply vertical filter in steps of 1 line for interlaced BGRA [18:27] ffmpeg.git 03Yu Xiaolei 07master:bc3d02fa88c4: lavf/avio: clarify the buffer parameter of avio_alloc_context [18:30] ffmpeg.git 03Benoit Fouet 07master:155f4dd668b8: doc: add entry for APNG demuxer where needed. [18:48] so uh.... [18:48] would a patch be accepted to disable xmp printing in ffprobe? [18:48] because some files print shit for 5+ minutes [18:48] its ridiculous [19:22] wm4: alright, so what do you propose (re:closing pb) [19:22] let the API user handle it [19:23] just export the necessary information [19:26] wm4: but, i don't see what the user is supposed to do here [19:27] ffmpeg has pull requests now? [19:27] ubitux: the user can do things with the AVIOContext, or even unset it (if lavf allows that) [19:28] so the patch as is looks ok to you? [19:28] yes [19:29] although granted I'm not perfectly sure how I'll use it [19:29] i'll push it if you're able to make use of it then [19:29] fair enough [19:29] maybe we forgot something so i don't want to change the api yet [19:54] Are github PullRequests accepted nowdays? [19:54] no [19:54] That's highly annoying... [19:55] BtbN, some of us would say github PRs are highly annoying [19:55] they have a number of disadvantages. [19:55] Setting up mails, managing changes, and getting tons of ML mail is annoying [19:56] im saying its personal preference, and the overwelming majority of people here prefer the ML. [19:56] I have that stuff flying around since ages now, and didn't send it in just because of having to do that git mail stuff [19:56] ok? [19:57] and you can attach patches too. [19:57] Makes it kinda hard to discuss some of the changes [19:57] I could send a mail pointing to my github branch, as request for comments [19:59] someone might review it, but i cant guaranatee it [20:00] I'm still not sure if it qualifies as nonfree or not [20:00] BtbN: some github pull requests have been accepted, IIRC. we prefer patches sent to the mailing list, but you can try a PR. not nearly as many people see them. [20:00] It's a new h264 encoder, needs some feedback [20:00] nvidia nvenc stuff [20:01] non-free with dlopen/loadlibrary is a massive grey area [20:01] i dont think youll ever know conclusively [20:01] There is no non-dlopen way though. That's the official way nvidia wants it to be used [20:01] ML + github PRs are the worst combination [20:01] BtbN, i know [20:02] pick either one, and ffmpeg picked ML-based workflow [22:15] ffmpeg.git 03Martin Storsj? 07master:f20141d73f08: vorbis_parser: Include stdint.h in the header, to make it work standalone [22:15] ffmpeg.git 03Michael Niedermayer 07master:23e91a1bfd7f: Merge commit 'f20141d73f08ed0c8e875bd993a7143e19b266e3' [22:18] ffmpeg.git 03Lukasz Marek 07master:e29153f414f5: lavc/avuienc: fix mem leak in case of init failure [22:18] ffmpeg.git 03Lukasz Marek 07master:2a89afb376ae: lavc/libxvid: return meaningful error codes [22:42] kurosu_ / jamrial : any more asm comments? [22:44] kierank, nope [22:45] was the unrolling actually helpful? you said it was but it looked more than I would have assumed [22:45] and I'm not sure if it breaks the C part [22:46] I mean, handling batches of 12 instead of 6 as I thought it assumed [22:47] anyway, I don't really care: if it breaks, I bet you'll be eager to fix it [22:48] for 8-bit yuv422p -> v210 you have to unroll [22:48] whereas in 10-bit mode you only have 8 pixels in your register [23:10] ffmpeg.git 03Lukasz Marek 07master:ea0d8938173e: lavu/opt: handle NULL obj in av_opt_next [23:10] ffmpeg.git 03Lukasz Marek 07master:1907ff0a67ed: lavc/utils: free private options on avcodec_open2 fail [23:19] kierank: nothing aside from what i mentioned for the 10bits patch [23:19] ok [00:00] --- Wed Nov 26 2014 From burek021 at gmail.com Thu Nov 27 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Thu, 27 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141126 Message-ID: <20141127010502.747EF2AD6107@apolo.teamnet.rs> [01:11] ffmpeg.git 03Carl Eugen Hoyos 07master:5badcdf20d98: Rename sync() functions in libavformat. [01:11] ffmpeg.git 03Benoit Fouet 07master:4acefd25215a: avformat/apngdec: account for blend and dispose operations. [01:55] michaelni: we could always ask users on twitter to come up with a name. they'll probably like that, and we'll get a free name such as "Ur mom". [02:18] ffmpeg.git 03Michael Niedermayer 07master:2ad38c6e02cc: avformat/mxfdec: dont ask for samples with field dominance 0 anymore [02:25] what's the difference between yuv2yuvX and yuv2planeX [02:32] ah the latter has accurate rounding [04:30] ffmpeg.git 03Michael Niedermayer 07master:e5c01ccdf5a9: avcodec/flacdec: Call ff_flacdsp_init() unconditionally [04:30] ffmpeg.git 03Michael Niedermayer 07master:5f30522894d1: avcodec/flacdec: fix off by 1 error [06:05] Perhaps another DeckLink issue: http://ffmpeg.zeranoe.com/forum/viewtopic.php?f=10&t=1823#p7554 [12:00] ffmpeg.git 03Martin Storsj? 07master:ee37620b6ae4: movenc: Add a flag for indicating a discontinuous fragment [12:00] ffmpeg.git 03Michael Niedermayer 07master:b78074fd13ad: Merge commit 'ee37620b6ae4783cda637408422044b2d14a688c' [13:34] ffmpeg.git 03Martin Storsj? 07master:234fb81e3145: movenc: Expose the fragment index as an avoption [13:34] ffmpeg.git 03Michael Niedermayer 07master:0df95fa327c8: Merge commit '234fb81e3145e9c9aec4ec16266676fab7dc21fa' [13:34] ffmpeg.git 03Benoit Fouet 07master:8b8cb30d1188: avformat/apngdec: use packet pts and duration instead of altering stream framerate. [14:59] michaelni: can you apply my v210 stuff [15:18] kierank, is there a patch that took care of James comments ? [15:19] 1122 23:10 James Almer (5.9K)  [15:30] michaelni: ok I will squash the two patches and resubmit [16:14] ffmpeg.git 03Michael Niedermayer 07master:79ceaf827be0: avcodec/pngdec: Check IHDR/IDAT order [17:00] michaelni: patch sent with changes [17:04] hah and now libav has a patch [17:39] ffmpeg.git 03Michael Niedermayer 07master:9a53707e86eb: avcodec/pngdec: Fix paeth prediction with small images [17:46] kierank, reply sent, seems it breaks fate [17:47] how do I run that single fate test? [17:47] make fate-vsynth1-v210 [17:48] wtf is that test doing [17:49] use V=1 [17:49] it compares a 422 file by converting it to 420 [17:49] i did use V=1 [18:08] it fails even without running the asm [18:11] arwa: http://trac.ffmpeg.org/wiki/SponsoringPrograms/OPW/SponsoringPrograms/OPW/2014-12/MiscLibavfilterExtension [18:11] fast&furious edit, I have to dig it more myself [18:12] about eq/eq2, James Darnley was working on it, you should contact him and ask for more [18:12] arwa: ^ [18:13] is an upsampling to 422 and a downsampling back to 420 meant to be bitexact? [18:14] i would be surprised [18:15] maybe if you use a point scaler for downsampling or something :p [18:25] a lot of pixels are off by one [18:26] who said we had bit exact colorspace conversion? :P [18:27] it would be interesting to see if the c version was bitexact vs the asm [18:27] basically libswscale is being liquid shit _again_ [18:28] or rather , no one is making a proper bugreport so it can be fixed [18:28] instead wasting time on complaining :) [18:55] i'm pretty sure there's an off by one somewhere [18:55] or a clipping ug [18:59] kierank, in your code or swscale ? [19:00] don't know [19:02] ./ffmpeg -pix_fmt yuv422p -s 1280x720 -i in.yuv -vcodec v210 -y out.mov && ./ffmpeg -i out.mov -pix_fmt yuv422p10 -y out.yuv && ./ffmpeg -pix_fmt yuv422p -s 1280x720 -i in.yuv -f rawvideo -pix_fmt yuv422p10 -y in10.yuv && md5sum in10.yuv out.yuv [19:02] in10.yuv and out.yuv match [19:02] but in.yuv and out.yuv don't [19:02] so there's something going on somewhere [19:03] some broken colourspace conversion [19:05] michaelni: ^ [19:07] the c seems to be ok [19:07] the asm doesn't [19:24] ffmpeg.git 03Michael Niedermayer 07master:1b5d11240692: avformat/mov: Fix memleaks for duplicate STCO/CO64/STSC atoms [19:24] ffmpeg.git 03Michael Niedermayer 07master:1d3a3b9f8907: avcodec/rawdec: Check the return code of avpicture_get_size() [20:12] Gramner: ping [20:12] pong [20:12] Gramner: do you mind being a second pair of eyes and looking at my v210 8-bit chroma shuffles? [20:12] there's something subtly wrong with the code [20:13] yeah sure, in a few minutes [20:13] thanks [20:22] ah I think it's because I don't merge the luma and chroma correctly [20:23] i mean the u and v planes actually [20:24] kierank, i found the issue [20:25] the shuffle is wrong in my asm for one thing [20:26] yes, posted correction to ml a sec ago [21:14] ffmpeg.git 03Kieran Kunhya 07master:36091742d182: v210enc: Add SIMD optimised 8-bit and 10-bit encoders [21:57] a user got in touch with me asking to add strftime to the segment muxer [21:57] the segment muxer uses av_get_frame_filename() to get the filename [21:58] would it be A Good Thing to make av_get_frame_filename() use strftime, or better to just add strftime to the segment muxer? [21:59] ffmpeg.git 03Cl?ment BSsch 07master:b424e67abf0d: avfilter/signalstats: fix different buffers for out frame if burn is enabled [21:59] ffmpeg.git 03Cl?ment BSsch 07master:c7e8f610f281: avfilter/signalstats: remove pointless sub filter init system [21:59] ffmpeg.git 03Cl?ment BSsch 07master:56b98dfc4f09: avfilter/signalstats: integrate height loop into subfilters [21:59] ffmpeg.git 03Cl?ment BSsch 07master:cc5c667eb13b: avfilter/signalstats: reindent after previous commit [21:59] ffmpeg.git 03Cl?ment BSsch 07master:fad6865748c7: avfilter/signalstats: fix repitition/repetition typo [21:59] ffmpeg.git 03Cl?ment BSsch 07master:9db78a296c37: avfilter/signalstats: add slice threading for subfilters [21:59] ffmpeg.git 03Cl?ment BSsch 07master:7acbd56a8a99: avfilter/signalstats: isolate sat hue computation metrics in a function [21:59] ffmpeg.git 03Cl?ment BSsch 07master:82dda8e4eaa6: avfilter/signalstats: add threading in compute_sat_hue_metrics [21:59] ffmpeg.git 03Cl?ment BSsch 07master:9cb1d81a60bd: avfilter/signalstats: localize a few variables [21:59] ffmpeg.git 03Cl?ment BSsch 07master:c2ea7069c489: avfilter/signalstats: re-use yuv/yuvu/yuvv vars in diff [22:35] Perhaps another DeckLink issue: http://ffmpeg.zeranoe.com/forum/viewtopic.php?f=10&t=1823#p7554 [22:42] the free of the name in that line should use SysFreeString, not free() [22:42] on win32 anyway [22:42] no idea about the other operating systems [22:43] (obviously not SysFreeString on other operating systems) [22:43] nevcairiel: thanks [22:44] I wonder why I chose free... [22:44] on osx its a CFString, whatever is required to free that, i don't know! [22:45] Zeranoe: care to make a patch? [22:47] ramiro: Pretty busy right now [22:49] nevcairiel: how about you? =) [22:50] i dont write patches i cant even compile [23:44] ffmpeg.git 03Lukasz Marek 07master:e98aced69955: ffserver_config: cosmetic: simplify functions calls. [23:44] ffmpeg.git 03Lukasz Marek 07master:d57a6d20875f: ffserver_config: cosmetic: move line_num into FFServerConfig [23:44] ffmpeg.git 03Lukasz Marek 07master:f61cb6453d22: ffserver_config: map ffserver options to AVOptions [23:45] ffmpeg.git 03Lukasz Marek 07master:6c2ed67c2f6f: ffserver_config: remove ffserver_apply_stream_config function [23:45] ffmpeg.git 03Lukasz Marek 07master:aaf6cc925f7f: ffserver: allow skip setting defaults [23:45] ffmpeg.git 03Lukasz Marek 07master:ec6e035b8b1f: ffserver: export recommented encoder configuration [23:45] ffmpeg.git 03Lukasz Marek 07master:3cb0bec6870c: ffserver: dont leak child arguments [23:45] ffmpeg.git 03Lukasz Marek 07master:3d0867917faa: ffserver: dont leak pb_buffer [23:45] ffmpeg.git 03Lukasz Marek 07master:345cfd04d093: lavc/options: fix leaks in avcodec_free_context [23:54] >HEVC >interlaced [23:54] weren't there plans to not even support that [23:57] >broadcast [23:57] >industry [23:57] :-P [23:57] ffmpeg.git 03Benoit Fouet 07master:e2b8b4caf6c0: avformat/apngdec: validate frame dimensions. [23:58] Action: michaelni has the suspicion that the plans to kill interlaced are as old as interlaced formats [00:00] --- Thu Nov 27 2014 From burek021 at gmail.com Thu Nov 27 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Thu, 27 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141126 Message-ID: <20141127010501.6FC5018A0252@apolo.teamnet.rs> [00:04] shitty-ass proprietary security camera videos... "let's take a format and shit on it so it can't be played" [00:05] Yeah... [00:23] if anyone else cares, i did end up getting ffmpeg compiled with libnvenc enabled (in Arch) using this github fork https://github.com/agathah/ffmpeg_libnvenc along with the Nvidia NVENC SDK for windows because a single header file was neeeded. [01:22] how would I find out what codec are available in my ffmpeg I have installed. when I run ffmpeg --version I get ffmpeg: error while loading shared libraries: libx265.so.35: cannot open shared object file: No such file or directory [01:23] Is libx265.so.35 in your LD_LIBRARY_PATH ? [01:24] c_14, no idea how to find that out. sorry, i'm pretty much a noob that's been getting help from others [01:25] c_14, i know that when I compiled ffmpeg I had x265 as a depends and had --enable-libx265 within the configure line [01:25] echo $LD_LIBRARY_PATH; locate libx265.so; find / -iname 'libx265.so.*' [01:26] c_14, can i narrow it down to /usr/ instead of /? [01:26] On arch, probably. [01:27] They moved everything to /usr a while ago anyway. [01:27] Though, check /home as well [01:27] if it's not in /usr [01:27] In case you compiled but didn't install libx265 [01:27] or installed in a non-standard path [01:28] c_14, /usr/lib/libx265.so [01:28] /usr/lib/libx265.so.35 [01:28] ldd /path/to/ffmpeg [01:28] And pastebin it. [01:28] so just ldd /usr/bin/ffmpeg? [01:29] And put that on a pastebin, yes. [01:29] https://clbin.com/eftgn [01:31] Is /usr/lib/libx265.so.35 a regular file? [01:32] c_14, that file doesn't exist [01:32] Didn't you just tell me it did? [01:32] ls -la /usr/lib/ | grep libx265 returned only 2 files. libx265.so > libx265.so.37 and libx265.so.37 [01:33] recompile ffmpeg against the newer version of libx265 [01:33] Or (and I don't recommend this) link libx265.so.37 to libx265.so.35 [01:33] Probably just ignore I said that. [01:34] current x265 package in Extra will provide libx265.so.35. not sure where you got 37 [01:34] c_14, my locate db was old, redoing the search reveals no /usr/lib/libx265.so.35 [01:35] Or install that package from Extra [01:35] oh, apparently that's from x265-hg [01:36] either use the package from Extra, or re-compile ffmpeg [01:36] ok, all fixed. thank you [01:36] llogan, installing x265 from extra solved it [01:37] most users should need to use any CVS type package in AUR if there is an equivalent package in the official repos [01:37] should *not* [01:42] llogan, understood. i had misread a comment on the obs-studio-git package in the AUR. so it was totally my own fault. thanks for helping me guys, i can now utilize my gtx 760 for encoding with ffmpeg. :) [01:46] I thought that video cards sucked for encoding video still? Has this changed? [02:32] anyone got idea if this works and what are the proper tags names? ffmpeg -i INPUT.mkv -metadata:s:v:0 "Matrix coefficients=BT.709" -metadata:s:v:0 "Color primaries=BT.709" -metadata:s:v:0 "Transfer characteristics=BT.709" -c:v copy OUTPUT.mkv [02:37] I'm no expert, but if the colormatrix is already BT.709, then all relevant tags should be set and if it isn't setting them will probably just break things. [02:37] Why exactly do you want to do this? [02:44] c_14: when i play such files (without tags) with mpc-hc+madvr the decoder uses BT.601 when the horizontal resolution is below 1280 i guess [02:45] so i did quick test with -x264-params colorprim=bt709:transfer=bt709:colormatrix=bt709 and the behavior is OK for hd480 resolution.. [02:49] Can you pastebin an ffprobe of the file it's playing incorrectly? [03:13] c_14: diff is yuv420p10le vs yuv420p10le(tv, bt709), i just dl zeranoe's build to check it because i dont have ffprobe on my box...thats why such delay, if you want both 1 frame files i can give you url for look... [03:14] kk, then the source file doesn't have the colorspace metadata set so the program is just guessing [03:14] s/space/matrix [03:14] yes [03:16] and it seems for 1280 and higher its ok but for lower resolutions nope (which makes sense, but not in my case) [03:18] I'm not sure you can change the colormatrix without reencoding since I'm pretty sure that's stored in the video stream and not the container. [03:18] s/change/"set" [03:19] The easiest option, especially if it's just a few videos would be to set the colormatrix in the player. [03:22] well, yeah i think i could script that in madvr easily.. but wondering about the remux way.. [03:23] did another test with -map_metadata -1 and colormatrix is still set [03:23] Ye, I don't think it's container metadata. [03:23] It's stream metadata. [03:23] so i guess i cant manipulate it so easily [03:23] you can set it using the colormatrix filter, but that'll require reencoding [03:26] You could probably set it with a hex editor/a specially crafted program, but you'd have to dig into the spec[s]. [03:26] You can always try asking on the user ml. [03:38] interresting h264_changesps bitstream filter - http://forum.doom9.org/showthread.php?t=152419 [03:42] but im off now, thx for help [04:15] hi, i disabled almost all things i do not need to achieve a smaller file size, it works for compressiong most mp4 files, but one with amrnb audio encoder failed me. i compiled ffmpeg with amrnb decoder and aac encoder, but it says i need aresample filter, but i configured it with this, but it doesn't shows up in the configure result [04:19] anyone can help me? [06:34] hi , is ffmpeg supporting cedarx vs tms320 omap optimized codes ? or davinci ? [07:20] here's some interesting reading http://on-demand.gputechconf.com/gtc/2014/poster/pdf/P4188_real-time_panorama_video_NVENC.pdf [07:42] ffmpeg [07:42] -f image2pipe [07:42] -filter_complex concat=n=2:v=1:a=0 [07:42] -vcodec mpeg4 [07:42] -b:v 800k [07:42] -r 2, [07:42] -y, pipe:1 [07:43] Can anyone tell me why the above command gives me the error: Unable to find a suitable output format for 'pipe:' [07:44] there are inputs in the actual command, I just forgot to put them in there [07:45] Add -f to the output options. [07:49] Oh, I thought it would pickup on that from the -vcodec option [07:50] that applies to the codec, not the container [07:52] Got it, thanks. That worked. You guys are like mountain sages [07:52] I always think Ive got an obscure problem and someone always seems to get it straight away [10:14] Hi all, I compiled ffmpeg for a web CMS inside a non-system directory, but a dir specifically for the CMS installation (Plumi) [10:15] It compiled successfully, but when I try to run it, I get the error that it cannot find the libraries it needs to run [10:15] bin/ffmpeg: error while loading shared libraries: libavdevice.so.54: cannot open shared object file: No such file or directory [10:17] I see that the library compiled successfully, but it is in the path ffmpeg/parts/ffmpeg-build/lib [11:42] hi all [11:43] when encoding to an H264 file, is there a way (set of parameters) to ensure that the frames in the resulting file will have decoding_order = display_order? [12:12] anyone knows hot to encode an H264 without frame reordering using ffmpeg? [13:27] I'm trying to stream a video from ffmpeg to ffserver. Why does ffmpeg exit without streaming and without saying what's wrong? http://pastebin.com/cJhvZp2C [13:58] hi [14:10] g'day [14:32] do windows phone support HLS? [14:42] hello all, guys do you know what software can i use to make stuff like this: http://www.youtube.com/watch?v=-yM8mO63yE0 i know he is professional dj, and he use sound software and video separately, but anyways, would love very much to learn how he does this [15:35] I've been searching on google but I haven't found the answer yet. Can ffmpeg record an h264 video stream from an IP camera? [15:38] Also, can ffmpeg add metadata to video files? I'd like to record the time when the video starts recording [15:38] <__jack___> dougquaid: which is the output format or your cam ? a 'raw' h264 video stream ? [15:39] <__jack___> eg with or without a container [15:39] __jack___ let me check [15:40] It says h.264 (mpeg-4) part 10/avc, as well as Motion JPEG [15:42] __jack___: Does that spec describe the output format? Based on my limited knowledge of video files I'm guessing the format is mpeg-4? [15:42] <__jack___> hum, it's not very explicit to me [15:43] <__jack___> ffmpeg should be able to handle such stream [15:59] anyt ideas why resulting mp4 video has audio out of sync? ffmpeg -y -i 0CAGS5AIZMQ.flv -vf scale=320:240 -q:v 0 -profile:v baseline -r 25 -vcodec libx264 -crf 21 -ac 2 test.mp4 [16:34] hefest, try adding -c:a copy [16:36] BtbN: Could not find tag for codec nellymoser in stream #1, codec not currently supported in container [16:37] What even is nellymoser oO [16:37] BtbN: Stream #0:1: Audio: nellymoser, 8000 Hz, mono, flt [16:37] BtbN: in source FLV file [16:38] yes, i have never heard that before. [16:38] source flv file has variable frame rate, i think that's causing issues but not sure how to solve it [16:44] man this is killing me, anyone else? any hints how to solve this? [17:20] hey guys... using the latest static build and I'm getting this... [17:20] Unknown encoder 'libfdk_aac' [17:20] trying to follow https://trac.ffmpeg.org/wiki/Encode/AAC [17:21] the static build doesn't include the fdk aac encoder, because (due to license incompatibilities) distributing a build with that enabled would not be legal. [17:21] if you want to use it, you have to build your own ffmpeg [17:21] kepstin-laptop: how can I get aac?... it doesn't seem that libfaac works either [17:22] the static build has the native aac encoder enabled; if you don't want to build your own ffmpeg, you could use that. [17:22] I will try... I have a bunch of .mkv files that don't play on XBMC on my Nexus Player, but after I re-encode them with Handbrake they do [17:23] I used ffmpeg to change containers without re-encoding... put it in a .mp4 and it still wouldn't play [17:23] so now I want to use ffmpeg to re-encocde the audio but leave the video... I'm trying to track down what the problem is [17:25] it's probably more likely to be an issue with the video encoding; for example, if the mkv used 10bit h264, that won't work on any hardware players. [17:26] title : Shaun The Sheep S01E01 540p HDTV x264-DEADPOOL [17:26] encoder : HandBrake 0.10.0 2014112300 [17:26] whoops... wrong lines [17:26] Stream #0:0: Video: h264 (High 10), yuv420p10le, 960x540 [SAR 1:1 DAR 16:9], 25 fps, 25 tbr, 1k tbn, 50 tbc (default) [17:27] is that (High 10) mean 10bit? [17:27] there you go, that's exactly the problem. [17:27] yep. [17:27] kepstin-laptop: okay... so I will have to re-encode no matter what [17:27] kepstin-laptop: I can actually play these on my Android phone / tablet with mxplayer or vlc. [17:28] If you use software decoding, it would play. but the hardware decoder wouldn't work [17:28] kepstin-laptop: okay... that must be what is going on [17:59] Does ffmpeg have switch to drop all user metadata from original input vid? [18:03] Hi everyone [18:05] I'm trying to setup a live continuous (24/7) WebM streaming w/ ffmpeg feeding to icecast server. All appears to run just fine but after some days of running, ffmpeg stops and I managed to capture the final 'debug log' which is here : http://pastebin.com/SNDqbc5s [18:07] i'm issuing the following command for ffmpeg : 'screen /bin/bash -c 'sudo ffmpeg -y -loglevel debug -f video4linux2 -re -i /dev/video0 -i usora_logo.png -i artica_small_bw.png -filter_complex "[0:v][1:v] overlay=0:H-h [V1]; [V1][2:v] overlay=W-w:H-h" -pre libvpx-720p-copy -vf scale=-1:720 -strict experimental -f WebM -content_type video/webm icecast://artica:artica9900 at usora-artica.noip.me:8000/stream.webm 2> log2.txt'' [18:07] can somebody help me figuring out what's the problem? [18:09] rflmota: you just send your user and password to us. [18:09] rflmota: artica:artica9900 [18:15] rflmota: By the way, I came here today just to ask this command that you just send. I'll study your parameters. [18:17] very quick question, is it possible to play mp4 files with FLASH RTMP? [18:18] flash can play mp4 files over http. But when you're using rtmp it's streaming the video over rtmp; no mp4 is involved anywhere :) [18:18] you should be able to use ffmpeg in codec copy mode to send the contents of an mp4 file over rtmp if you want [18:19] thanks DanielSa [18:38] Is it possible, in a live broadcast using one mic and two or more video sources, switch between the videos without stop the transmission? To be more specific, using two webcams, without any analogic tool. [18:39] If you code it yourself/can find a program, yes. [18:41] DanielSa: you could pull off side by side, or picture in picture, video with ffmpeg. [19:16] Thank you guys. [19:39] excalibr: "-map_metadata -1" might do what you want [19:40] and don't forget "-map 0 -c copy" if you want to map all streams and stream copy everything if you dn't want to re-encode [19:44] What is the aac encoder marked experimental? [19:44] the native FFmpeg AAC encoder, "aac" [19:45] there is talk about removing the experimental status, but that may have to wait for bug #2686, aka the "epic thread". [19:52] That would be great, I'd love to get rid of libvo [19:53] even now with the 'experimental' status, the aac encode is better than libvo-aacenc [19:54] kepstin-laptop: I've been using it, but it's a pain to have to type "-strict -2 -a:c aac" everytime [19:54] it'd be nice if that were the default [19:55] the experimental flag is a leftover from the times when it actually produced broken audio at relatively high bit rates [19:55] which was true up until early 2011 or so? [19:55] but no-one decided to remove the flag [19:56] I'm converting a bunch of OLD avis to mp4 so my chromecast can play them [19:56] they have .mp3 audio that's pretty crappy [20:04] Converting mp3 audio to aac audio will never make it better. [20:04] unless you want it smaller and lower quality, just leave it. [20:05] chromecast should be able to play mp3 audio in an mp4 container [20:24] kepstin-laptop: Good point... [20:25] llogan, thanks. By the way, any idea why mkv file muxed with ffmpeg is not fast-forwardable on my tv media player? It is fast-fwd able though, when I muxed it with mkvmerge [20:26] excalibr: player might need the seek index at the start of the file, I think ffmpeg writes it to the end by default (since it doesn't know how big the seek index is until the encode is done) [20:27] there's an option to reserve some space at the start in ffmpeg, you could give that a try [20:28] kepstin-laptop, what's the option/switch? [20:28] excalibr: https://www.ffmpeg.org/ffmpeg-formats.html#Options-4 [20:28] (make sure your ffmpeg is writing to a file rather than a pipe, too) [20:33] kepstin-laptop, does the option take size prefix for its arg? or it must be in bytes only [20:33] On this page: https://trac.ffmpeg.org/wiki/Encode/AAC#NativeFFmpegAACencoder [20:33] Is this still accurate? "Effective range for -q:a is around 0.1-10. This VBR is experimental and likely to get even worse results than the CBR. If ffmpeg ignores -q:a then get a newer build (see ticket #1346). " [20:33] excalibr: dunno, i'd have to look at the source coded. [20:39] bakers: i don't know. that was written a year ago, but by kamedo2 who knows more about it than i do [20:55] kepstin-laptop, just gave it a go and no dice :(. No idea what's wrong [20:56] here's the command I used: ffmpeg -i input.mp4 -i en.srt -map 0:v -map 0:a -map 1 -map_chapters -1 -map_metadata -1 -c copy -reserve_index_space 100000 ffmpeg-fastfwd.mkv [20:59] hi, I would like to add looped audio input (mp3 file) to the video and can't find any useful options.. can it be done and if so, how? thanks a lot! [21:13] hay: maybe the loop option in the amovie filter? or maybe use the concat demuxer or concat filter [21:15] there's a loop example using concat demuxer you can adapt here: https://trac.ffmpeg.org/wiki/How%20to%20concatenate%20%28join,%20merge%29%20media%20files#demuxer [21:16] llogan, thanks, will look at it [21:30] How can I do multiple files at once? * doesnt seem to work [21:30] What do you mean? [21:31] "ffmpeg -f image2 -loop 1 -i picture.png -i *.mp3 -c:v libx264 -tune stillimage -c:a copy -shortest movie.mp4" for example [21:32] That's not how it works. [21:32] What are you trying to do? [21:32] Merge or concat or do the same thing for multiple files? [21:32] I am trying to merge audio and a picture into video file format. (but not one by one) [21:33] You have 3 choices, merge the audio tracks, concat the audio tracks, create one output per audio track. [21:33] Pick one. [21:34] I have no idea what those things mean? Im not very tech savy [21:34] merge ie, imagine having 1 boombox per audio track and playing a different audio track on every boombox at the same time [21:34] concat: playing each audio track in succession on one boombox [21:35] create one output per audio track: have numerous boomboxes, play one song on one boombox then the next song on the next boombox, etc [21:35] the last one. [21:35] have multiple audio tracks in one output: have songs playing on multiple radio stations, and pick between them with a dial on the boombox [21:36] Right, that was option 4. Forgot about that one. [21:36] Do you have one picture or one picture per mp3? [21:36] one pic for them all [21:37] so you have a bunch of mp3s, one picture, and you want to make a bunch of video files where each video file plays one song and shows the picture? [21:37] Exactly [21:37] for mp3 in *.mp3; do ffmpeg -f image2 -loop 1 -i picture.png -i $mp3 -c:v libx264 -tune stillimage -c:a copy ${mp3%.*}.mp4; done [21:38] probably add some quotes in case your mp3 filenames have spaces [21:38] but that's basically it. :) [21:39] you may need -pix_fmt yuv420p as an output option depending on your target player/device [21:39] Quotes around "${mp3%.*}.mp4" ? [21:39] and around "$mp3" as well [21:41] and "-movflags +faststart" as an output option may be desired if this is for progressive download playback (such as via a browser) [21:41] Well, it is for youtube. [21:42] you won't need either of the options i mentioned then [21:42] but add -crf 18 [21:42] since youtube will re-encode it anyway [21:44] you can even add lyrics with a subtitles file (hardsub it) or other text with drawtext filter [21:44] $mp3: no such file or dir [21:44] hmm [21:46] Did you use single quotes anywhere? [21:46] ffmpeg -f image2 -loop 1 -i publicdomain.large.png -i "$mp3" -c:v libx264 -tune stillimage -c:a copy "${mp3%.*}.mp4" [21:47] dert78: you're supposed to have that wrapped inside of a for loop... [21:48] How do I do that? [21:48] exactly like I wrote up there [21:49] you could add -framerate 2 as in input option. less waiting. [21:49] dert78: what is your OS? [21:50] Oh I didnt know you mean to add for in the code. /facepalm. [21:50] Windows 7 [21:50] eh [21:50] that won't work in windows [21:50] probably [21:50] oh, then that won't work, the command c_14 gave was a linux shell command [21:50] unless you use cygwin or something :/ [21:50] You'll need a batch script. [21:53] I will install cygwin real quick and see if that works [21:53] might be easier to just use a normal batch script [21:53] "cygwin" and "real quick" don't go together in the same sentence... [21:54] How does the batch script look like then? [21:54] there are many examples on the 'net [21:54] or ask in #windows if such a channel exists [21:55] but we gave you all of the ffmpeg stuff for it to work nicely, you just need to figure out the batch part [21:56] Okay, I will do that. Thanks for all the help :) [23:09] hello. i'm using ffmpeg to convert from mp3 to ogg. My problem is that running the exact same command results in binary different ogg files. I would like the conversion to be deterministic, so that I can rely on the fact that if the input file did not change, I get the exact same binary file as output. I was wondering if somebody would have a tip for me? (http://pastebin.com/RStsM2va) [23:14] quixotic: You run that same command, on the same machine, and get different outputs? [23:14] yes [23:16] bakers: they are both legit files and sound the same, but the binary data is slightly different which causes problems in my automation process/revision control system. [23:17] how are you comparing the binary data? [23:17] Filesize? [23:17] Checksum? [23:18] git is noticing the file discrepancy, i verified via winmerge [23:19] quixotic: Wild... I just did the same thing you did 5 times [23:19] http://www.fpaste.org/154471/41704037/ [23:19] bakers: thanks for taking the time. are you getting the same file? [23:19] I don't know what to tell you, that doesn't make any sense [23:20] bakers: at least i'm not crazy [23:20] quixotic: All exactly the same size [23:20] http://www.fpaste.org/154472/14170404/ [23:20] I wonder if the ogg is the same, but there is some metadata about "encode time" or something that's unique to each one [23:21] i would think metadata is messing you up [23:21] yeah [23:21] try: -map_metadata -0 to unmap metadata from the first input [23:22] If I use ogginfo to look at the metadata [23:22] http://www.fpaste.org/154474/41704056/ [23:23] Line #5 [23:23] Each file has a different serial number [23:23] ah yes, i would need a way to force the serial number [23:24] I didn't know ogg files had serial numbesr [23:25] quixotic: https://xiph.org/ogg/doc/oggstream.html [23:25] That page mentions unique serial number several times [23:25] not sure how to set that [23:27] klaxa: not sure if i'm doing it right, but i think the -map_metadata does not help [23:28] It's gotta be that serial number, I'm not sure how you would force that [23:29] -metadata:[foo]:[bar] baz=thing [23:29] check the manpage [23:38] Hi Guys, I work for a small time service provider and we have run into an issue while transcoding live video. We tweak the settings per channel and SD channels work great, but we have run into an issue when a channel changes audio from stereo to dolby or vice versa (from what we can tell). [23:38] This causes an issue and we have to restart the ffmpeg process manually. [23:38] While checking the output stream using ffmpeg it appears to be encoding the Descriptive Video channel perfectly fine (constant stereo?), however it doesn't appear to allow the primary audio channel (stream/pid) to work properly, we suspect that it is because this channel changes. [23:38] The crazy thing, volumedetect SHOWS that volume is working, yet it isn't. I would LOVE to hear someones incite into this matter! [23:40] tried setting serial or "s" as metadata, but does not seem to work.. or at least i have no idea what the correct param name should be [23:51] quixotic: use the md5 muxer to test https://ffmpeg.org/ffmpeg-formats.html#md5-1 [23:52] that will require decoding to compare hashes, right? [23:53] yes, but it could also be used to test the encoded output [23:53] by adding -c:v [23:53] to avoid any differences caused by metadata [00:00] --- Thu Nov 27 2014 From burek021 at gmail.com Fri Nov 28 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Fri, 28 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141127 Message-ID: <20141128010501.5182818A02EB@apolo.teamnet.rs> [00:02] i'll try that. thanks everybody for the help. you're great [00:02] anyway, if you just do "ffmpeg -i input -f md5 -" for each input you'll see that the hash is the same [00:03] each input being your new ogg files [00:03] i see, i'll use this info in my new automation process. thanks [00:03] and you could probably specify the desired stream with -map if you want to just check a particular stream [00:06] quixotic: and make check that flac isn't the default encoder. i think it used to be default for ogg even with libvorbis supported. (next time provide the complete console output) [00:06] *make sure [00:17] quixotic: Did you solve your problem? I'm curious now [00:40] bakers: i'll work around the issue by comparing the ffmpeg hashes to see if the file is actually the same. haven't found out how to specify the serial number [01:43] How do I use your application to compress a 1GB video to 500mb ? [01:46] samthewildone: http://superuser.com/questions/4244/how-do-i-reduce-the-size-of-a-huge-mp4-video/4252#4252 [01:51] samthewildone: https://trac.ffmpeg.org/wiki/Encode/H.264#twopass [02:56] I have a movie whose source is 133 frames, but which apparently got exported as 60 fps BUT instead of staying 133 frames played over 2 seconds, it became 8025 frames played over 133 seconds. I do not have access to the source. Is it possible to use ffmpeg to keep every 133rd frame ? [03:02] ahhh. framestep filter... now to learn how to use it. [09:54] so i'm trying to compile ffmpeg on an f21 machine with an nvidia gpu and i'm having trouble with --enable-opencl [09:54] i'm getting a not found error, i have files in /usr/include/CL and /usr/include/opencl-utils [09:55] am i missing any deps or something? i;m pretty sure i have every i need installed [09:55] everything i* [09:55] What does the configure log say? [09:56] error: OpenCL/cl.h: No such file or directory [09:57] should i link one of the cl.h's in the dirs i previously mentioned to that dir? [09:58] So it's looking in /usr/include/OpenCL/, not in either of the places you have the files. [09:59] i got that [09:59] I can' help well wih a broken keboard, sorr. :D [10:00] Hello. [10:00] I'm in the process of converting an AVI to various quicktime-wrapped output formats. [10:01] In each case, black becomes grey; it looks very much like a luma range encoding glitch. [10:01] Are there any commandline switches that control this behaviour? [10:03] indieross: I believe it is possible to set where configure looks for things. [10:04] yeah, i think ill try changing it to /usr/include/opencl-utilc/CL/ [10:05] Could ffmpeg modify the timecode of a flv stream? [10:07] I've just tried -color_range but it seems to have no effect on anything. [10:10] yuss, found the missing dep [10:11] lets see how this nvenc works out [10:14] how I can send ONE stream with FFMPEG to multiple udp multicast addresses? [10:15] http://pastebin.com/dKjV8zQS here is my command line [10:16] -f mpegts -flush_packets 0 "udp://239.239.4.100:1234?pkt_size=1316" [10:16] I would like to use something like this: -f mpegts -flush_packets 0 "udp://239.239.4.100:1234?pkt_size=1316, udp://239.239.4.101:1234?pkt_size=1316, udp://239.239.4.102:1234?pkt_size=1316" [10:16] is it possible? [10:18] hi! [10:19] I'm having troubles again... [10:19] K4T: https://trac.ffmpeg.org/wiki/Creating%20multiple%20outputs#Duplicateoutputs [10:20] I have a mts as input, and an image2 with loop, with overlay. ffmpeg does not end... starts dropping [10:20] at the end of the mts, all frames go to "drop", and never ends. '-shortest' does not help. [10:20] How can I make it finish once the mts finishes? [10:20] (I'm just putting an image cover to the first 4 seconds of the mts) [10:25] sfan5, thx [10:31] shortest=1 in overlay does not help ... [10:31] weird [10:33] ah, updating ffmpeg fixed it. [10:33] \o/ [10:44] hey [10:44] I need an help compiling ffmpeg [10:44] I have configured with --enable-shared [10:45] now when I make it says: /usr/bin/ld: libavutil/display.o: relocation R_X86_64_PC32 against undefined symbol `hypot@@GLIBC_2.2.5' can not be used when making a shared object; recompile with -fPIC [10:45] what can I do? [10:46] add -fPIC to your CFLAGS [10:46] how? [10:48] either with --extra-cflags="-fPIC" in your configure line or with export CFLAGS="-fPIC" [10:48] making sure to keep whatever's already in your cflags [11:05] usr/bin/ld: /usr/local/lib/libavformat.a(rmsipr.o): relocation R_X86_64_32 against `.rodata' can not be used when making a shared object; recompile with -fPIC [11:06] anyone knows what is this problem when I compile vlc? [11:10] Hi [11:11] guys where can I found quality scale for codecs? some man page or so? [11:13] Hi, I'm using the Ubuntu PPA, and am trying to stream to UDP: ffmpeg -f x11grab -s 1600x900 -r 50 -vcodec mpeg2video -f mpegts udp://10.0.0.55:9002 - as per an example I found on the web. [11:13] However, I get: Output #0, mpegts, to 'udp://10.0.0.55:9002': Output file #0 does not contain any stream [11:15] add -map 0 before -f mpegts [11:16] Are any questions ever actually answered in here? I keep seeing people ask, but very few responses. [11:16] depends on the time/question [11:17] c_14, it gives me: Invalid input file index: 0. [11:18] c_14, fair enough: http://pastebin.com/eTDE3zYs [11:19] I'm looking at this web page for my examples, BTW: http://www.waitwut.info/blog/2013/06/09/desktop-streaming-with-ffmpeg-for-lower-latency/ [11:23] More googling reveals commentary about an audio stream. Don't have/want that.... could it be a problem? [11:47] do I have to start the client at the same time I start the 'server' ? [12:49] davidw: you don't have a -i input option [12:49] https://trac.ffmpeg.org/wiki/Capture/Desktop [13:02] hi [13:16] damn, that's still way laggy [13:20] how does webrtc manage to be so fast -( [14:19] Ok, I wrote it up here: http://superuser.com/questions/845699/get-webrtc-like-latency-with-ffmpeg - any ideas are welcome [14:26] hi, i have ffserver running, ffmpeg doing ffmpeg -f video4linux2 -i /dev/video0 http://localhost:8090/webcam1.ffm [14:26] but I don't know how to access video feed [14:27] the remote machine connects with mplayer and sais No stream found to handle url http://192.168.255.103:8090/webcam1.ffm [14:29] Hi, I have audio/video sync problem (MJPEG -> H264) [14:29] ffmpeg -i mjpeg.mov -c:v libx264 -preset fast -crf 13 -pix_fmt yuvj420p -c:a libfdk_aac -b:a 192k output.mp4 [14:30] audio is ahead of video, there is about 0.2sec cut [14:31] if I encode just audio (MJPEG -> AAC) it is fine [15:15] Can anybody help me with a/v sync problems? [15:19] guys, new here and I need help converting rtsp to rtmp and stream via web [15:20] ffmpeg -i "rtsp://url/" -vcodec copy -f flv -r 25 -s 1920x1080 -an "rtmp://127.0.0.1:1935/live/livestream11" [15:20] I am using that in the web script [15:20] the error is Press [q] to stop encoding [15:20] [flv @ 0x15c2560]st:0 error, non monotone timestamps 1 >= 1 [15:20] av_interleaved_write_frame(): Operation not permitted [15:20] can somebody help please? [17:44] is there a way to make ffserver simply reroute to an ffmpeg that is outputting rtp already? I want to reduce latency [18:47] hi ! i m building my ffmpeg now but it won t build ffplay ! (i added the --enable-ffplay) [18:47] an idear? [18:51] You need sdl. [19:20] Hi there :) [19:21] It's possible to stream line-in input to icecast server? I use now darkice to do it, but i looking for method to decrease delay [19:22] So i thought I can try encode audio input via ffmpeg and stream it icecast on my vps [19:22] https://trac.ffmpeg.org/wiki/Capture/ALSA [19:22] https://ffmpeg.org/ffmpeg-protocols.html#Icecast [19:26] I already reading it, but I;m not sure about syntax, ffmpeg -f alsa -i hw:0,2 icecast://[username[:password]@]server:port/mountpoint will works? [19:31] should [19:41] c_14: very nice tip about ALSA [19:45] hello i have an avi movie that i convert it to mp4 using -codec copy -f mp4 test.mp4 [19:45] the new movie has dropped frames and its a bit laggy and id ont understand why,i will post the command output to pastebin though i didnt see something strange there [19:47] http://pastebin.com/QmdnxHLR [19:48] well, that 'pts has no value' message looks strange, and seems like it would cause stuff like dropped frames or laggy playback [19:48] Action: kepstin-laptop has no idea why that would happen or how to fix it [19:48] the whole movie is laggy though [19:57] do i see that right that ffserver only supports flv? [19:58] is webm even suitable for expanding live streams? because i have no problem when i stream it from a file with ffserver.. [19:58] but when i stream something that is live encoded it doesn't work... [19:59] the only thing that always works is flv [20:04] You can stream with webm [20:04] Oh sorry, I misunderstood [20:05] no... webm is complete crap... it is too slow... and even when i use a server that takes way too much electricity (so that webm encodes in realtime)... the encoded stream still stutters [20:05] the only case where i have seen webm working was when it was streamed directly from a file... without any touching.. [20:06] Well, I have it working fine on my end [20:06] can you pm me the [20:06] part of your config? [20:06] I wrote it up some time ago [20:06] sec [22:07] hello, i try to open some rtsp links served by my adsl box exemple rtsp://mafreebox.freebox.fr/....... when opening it i have tcp error can t resolve hostname mafreebox.freebox.fr [22:07] what do i miss in my ./configure build options to fix this ? [22:08] (same link works well in vlc) [22:09] dude told me i'd had to adjust preroll and leys [22:09] wrong channel [22:11] smo__: try installing nscd and see if that works [22:12] As in, see if it works once it's installed. [22:12] ok i try [22:13] yup it works [22:13] hummm soo it s os relevant ? [22:15] I don't know the details, I just know that that might help... [22:15] ok thanks but strange :p [23:34] The site really cleaned up [23:35] Is the h.264 decoder closed source? [23:35] No? [23:36] I asked because [23:36] I was unsure [23:57] hey guys, does anyone have any experience decrypting hls ts files? [00:00] --- Fri Nov 28 2014 From burek021 at gmail.com Fri Nov 28 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Fri, 28 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141127 Message-ID: <20141128010502.5766218A02F3@apolo.teamnet.rs> [00:01] there is no interlaced coding mode [00:15] hevc doesn't have interlaced coding mode?! Now i'm liking it even more. [00:17] it means non standard interlacing [00:17] vendor specific [00:17] bwahahaha [00:25] Compn: http://www.reactiongifs.com/r/jZ2J5Yv.gif [01:05] kierank, i didnt mean hevc, but in general that i think as soon as interlaced was invented it was hated ... [01:06] but i could be wrong, iam no archeologist ... [01:07] michaelni: yes exactly [01:07] that's why the sample is just 2x 1920x544 [01:07] there is no mbaff [01:07] nor paff (mv chroma compensation part) [03:02] ffmpeg.git 03Michael Niedermayer 07master:9f9440bd8122: avcodec/hevc_ps: Check return code from pps_range_extensions() [10:57] http://lists.libav.org/pipermail/libav-devel/2014-November/065317.html any idea from where this was taken? oO [10:57] git log -p --author=dericed libavformat/mov.c raises nothing [10:59] the change isnt in ffmpeg [10:59] yes [11:00] and knowing Dave a bit, he probably didn't do that on libav tree [11:16] Do we have some parity check function that use processor power to do it rather then using bitwise operator [11:27] Which C standard does ffmpeg use? Can i use variable-length automatic arrays? As in C99 [11:28] No you can't use vla [11:29] for eg in x86 it use pushf for parity flag [11:29] hm, that makes the code quite a bit more inefficient [11:30] ffmpeg.git 03Martin Storsj? 07master:9326d64ed1ba: Share the utf8 to wchar conversion routine between lavf and lavu [11:30] ffmpeg.git 03Michael Niedermayer 07master:097de4d1d68f: Merge commit '9326d64ed1baadd7af60df6bbcc59cf1fefede48' [11:43] ffmpeg.git 03Martin Storsj? 07master:79fd186a5035: lavf: Use MoveFileEx instead of rename/_wrename on windows [11:44] ffmpeg.git 03Michael Niedermayer 07master:cc663bd13a97: Merge commit '79fd186a5035cf16fc0ab288d8f59da8b1ba2c0e' [12:35] ffmpeg.git 03Martin Storsj? 07master:675ac56b7ee0: Revert "lavf: Don't try to update files atomically with renames on windows" [12:35] ffmpeg.git 03Michael Niedermayer 07master:4ae1d6021be9: Merge commit '675ac56b7ee0f204963fde55295197c5df80aa91' [12:35] ffmpeg.git 03Rong Yan 07master:22e557917d08: libavutil/ppc/util_altivec.h : fix load_with_perm_vec() add marcos vcswapi2s() vcswapc() VEC_SPLAT16() VEC_SLD16() for POWER LE [12:50] ffmpeg.git 03Matthew Oliver 07master:0167fa006044: msvc: Fix compilation errors due to header include order. [14:06] Does ffmpeg realy run the init_static_data function just when enumerating the encoders? I'm not quite sure if i should use it for my nvenc encoder, or just make assumptions about which format(s) are supported. [14:15] good question, but i think many devels are asleep right now [14:15] BtbN : and i dunno the answer [14:15] :) [14:16] The problem is, finding out the supported pixel format involves initializing CUDA and NVENC, and is quite costly and can cause troubles. [14:17] Just assuming it's NV12 works for all hardware that currently exists, but it's far from ideal. [14:17] And i can't see any other option [14:22] why far from ideal? [14:22] All hw decoders and encoders are using it. [14:22] x264 uses it internally. [14:22] Because NVENC does not guarantee that it's supported [14:23] oh lol [14:23] It's the assuming that's not ideal. NV12 itself is fine [14:23] It's propably fine to just hardcode it to NV12, as they won't ever drop it. [14:30] WTF is this license [14:30] are they stupid or what? [14:31] Which license exactly? [14:32] nvenc one [14:33] There are 2 which apply to the header [14:33] https://developer.nvidia.com/nvidia-video-codec-sdk-license-agreement [14:34] Yeah, that one's pretty annoying [14:34] The one in the headers also is [14:38] wow, indeed [14:38] Now the question is.. does it make the ffmpeg binaries nonfree or not? [14:41] BtbN: yes, it does [14:43] ffmpeg.git 03Stefano Sabatini 07master:29208e6dcf94: lavu/imgutils: remove redundant and wrong check in av_image_fill_arrays [14:51] Is it ok to call ff_lock_avcodec? Can't see it beeing used anywhere [14:53] what are you trying to do? [14:54] I have to do non-threadsafe initialization [14:54] are they stupid or what? <- it's nvidia [14:54] Like loading the cuda and nvenc library [14:54] BtbN: what for? usually this is only needed for stupid global tables [14:54] Yes, i load the nvenc library globaly [14:54] is the library always the same? or can the user specify which one to use? [14:54] With a reference counter, to avoid loading it multiple time [14:55] It's allways the very same library [14:55] the OS should be doing the refcounting for you [14:55] The OS doesn't know how many times the avcodec codec is in use [14:56] BtbN: seems ff_lock_avcodec is already held during init [14:57] so codec->init is always restricted to 1 thread in the process [14:57] BtbN : where is the api header license ? [14:57] wm4: I hate hardware vendors when they do software. It makes no sense, their headers are crap, the API very bad, and their licenses are always "you're not allowed to use this". [14:57] Compn, in the nvEncodeAPI.h [14:57] and at https://developer.nvidia.com/nvidia-video-codec-sdk-license-agreement [14:57] but they are not the same [14:57] thx [14:57] j-b: isn't that normal? they got to protect their IP!!!!!!11111111111111111111 [14:58] j-b, the NVENC api is actualy quite nice [15:00] probably we should just ask nvidia to relicense this header for us :P [15:00] since we used to have contacts with them... [15:06] i wonder whats the use case of nvenc? [15:06] isnt it quite bad [15:06] It's encoding with nearly no cpu usage. [15:07] of course [15:22] the quality is OKish on recent GPUs [15:22] for realtime streaming things or so [15:24] its been a todo for me to offer nvenc and quicksync for real-time transcoding uses [15:25] but the timetable doesnt have that in it until next year, soo [15:25] The quality is the exact same on all GPUs [15:25] Maxwell just got faster, quality is unchanged [15:26] QSV on linux is horrible [15:26] 1hm oh well, then it was just intel that increased quality as well [15:26] There is only libva, and libva has the most horrible api i have ever seen [15:26] You have to manualy generate the h264 nal [15:26] i only need windows for now [15:26] but only parts of it [16:01] ffmpeg.git 03Michael Niedermayer 07master:57e5812198aa: avcodec/hevc_ps: More complete window reset [16:01] ffmpeg.git 03Michael Niedermayer 07master:98e8a9e2f238: ffmpeg: Print a debug message if the frame parameters mismatch the context [16:20] Is it ok to av_assert0 after av_malloc, or does the code have to continue eben when oom? [16:22] don't assert, just properly quit [16:22] assert should be for impossible code-path. [16:23] Hm [16:23] so i need to refactor half my code, just because of that [16:52] BtbN: beware of http://xkcd.com/292/ [16:52] Well, i'm using goto anyway. Best way to do error handling. [16:53] Better than a growing list of deinitializations in every error case [16:53] it is standard in most c codebases to use goto error [16:54] including ours [16:54] It's basicaly the C version of exception handling [16:55] At least i got rid of any non-constant static global variable while i was on it [16:55] yeah, goto is only really considered harmful in languages that _do_ have higher-level exception handling [16:55] i wouldnt call it the same as exception handling [16:56] I'm not saying you can't use goto, just be careful with it [16:56] Just don't goto upwards [16:56] i.e. you dont accidentally let a goto slip through 3 calls up [16:56] and be uncaught [16:56] ;) [16:56] goto fail is common practice in C [16:56] Could be avoided if you realy want to, but that would involve ugly wrapping of every function [16:56] so you return from the inner function instead of goto error [16:57] and the outer function does the cleanup [16:57] nevcairiel: which makes it quite funny to hear people talk about "the goto fail bug" (and mean that one Apple TLS one) [16:57] rcombs, yeah but goto didnt cause that one [16:57] allowance of lack of { did [16:57] Well if you forget the actual error handler... [17:05] rcombs: "goto considered harmful" is often mistaken as a complete damnation of goto [17:05] rcombs: but actually this was just to contrast goto spaghetti programming to structured programming, when structured programming was still new [17:06] my understanding was that it was referring to how people learned to program in older basic variants which had very limited flow control, and mostly relied on goto [17:06] there's nothing wrong with simple gotos, they're probably less error prone than "break" or "continue" [17:07] yes it refers to loops using goto ala BASIC [17:07] and such [17:07] anyone who takes any essay as Pure Truth should probably not be programming anyway. [17:09] Daemon404: you should write an essay on that :3 [17:09] Action: wm4 goes writing a function that uses 10 gotos [17:09] it'd be delightfully meta [17:09] wm4: no, that's a different strip, http://www.xkcd.com/148/ [17:09] um [17:09] sorry, I mena [17:09] rcombs: that's a different strip [17:43] hey ubitux [17:43] i got your question on the other chan [17:43] i took the additional metadata info from https://github.com/l-smash/l-smash/blob/master/lsmash.h#L3861 [18:07] koda: what do you mean "additional"? i don't even see what was the original commit [18:08] the original commit is the one you pointed me? or i don understand the question [18:08] the author is dericed in your patch [18:08] so i'm wondering where this comes from [18:10] oh the other one, right [18:10] i think i saw something on a github pull request that got ignored for a few years [18:12] ok [18:20] ffmpeg.git 03Cl?ment BSsch 07master:5ab882d7283f: avformat/mov: strengthen some table allocations [18:20] ffmpeg.git 03Cl?ment BSsch 07master:92fa1d9231b8: avformat/mov: change conjugation for "Duplicate" [19:21] Hi! Has anybody here ever compiled x265 on msys? How is it supposed to work? [19:23] ow [19:24] i think you mean mingw via msys, and its on the wiki [19:24] https://bitbucket.org/multicoreware/x265/wiki/CrossCompile [19:25] I don't want to crosscompile, I want to build on msys [19:25] its the same whether you compile natively with mingw or cross-compile from linux [19:25] ^^^^^^^6 [19:25] Ok, so I have cmake in my path and gcc.exe [19:25] the gcc is able to compile ffmpeg (and passes fate) [19:26] you might not need a toolchain file for windows, iirc there is an option in cmake-gui if you compile it natively, to use mingw [19:26] If I run ./make-Makefiles.sh it finds gcc.exe but claims "is not able to compile a simple test program." [19:26] eh [19:26] x265 has a script to run cmake with the correct arguments depending on you using x86 or x86_64 [19:26] to quote myself from earlier today, elsewhere [19:26] check build/msys [19:26] [15:36] < Daemon404> dont use their shell script [19:26] [15:36] < Daemon404> ever. [19:26] [15:36] < Daemon404> just use cmake [19:26] [15:36] < Daemon404> their shell script is for the benefit of their outsourced developers. [19:27] I originally wanted to build x86_64 (I have a compiler that succeds FFmpeg fate) but atm, I can neither build x86 nor x86_64 [19:27] What means "just use cmake"? (Sorry, I never used it) [19:27] cmake-gui.exe is pretty straight forward [19:27] this all looks way more complicated than it should have to be [19:27] you point it at a build and source dir [19:27] and select your toolchain [19:28] rcombs, the gist is: people dont like learning new tools (cmake) [19:28] (also cmake is bad, but thats orthogonal) [19:28] Who invented this... (sorry): What is the "source code" is it /x265? or /x265/build? or /x265/source? [19:28] the latter [19:28] x265/source [19:29] `cd build; cmake "-GUnix Makefiles" ../source; make; make install` [19:29] alternately, -GNinja if you've got that [19:29] Action: Daemon404 barfs [19:29] well no, not for msys [19:29] cmake -G "MSYS Makefiles" [19:29] oh, is that a thing [19:29] why is it different :| [19:30] Action: jamrial shrugs [19:30] i dont know [19:30] i hate cmake. [19:30] a lot. [19:30] well, either MSYS or MinGW or UNIX depending on fuckall, try until one works [19:31] or just cross-compile and maintain some semblance of sanity [19:31] i run nightly builds of x265, but with msvc and icl [19:31] cross-compiling with cmake sucks [19:31] cmake-gui now asks for a generator: Which one should I choose: mingw or msys? [19:31] jamrial, i used to work on openembedded as my job. [19:31] guess why i hate cmake? [19:32] cehoyos, good question. it's... unclear [19:32] haha [19:32] I hate MinGW/MSYS [19:32] "it's like cygwin, but half-assed!" [19:32] Error in configuration [19:33] rcombs: try msys2. it's like cygwin, but not half assed [19:33] Ok, I chose "mingw" which was wrong, how can I change what I chose? [19:33] clean the build folder [19:33] jamrial: maybe one day the differences will be documented! [19:34] well, for starters, it's in active development :p [19:35] cehoyos, it worked here by using msys makefiles and choosing to specify native compiler location [19:35] and pointing it at gcc and g++ [19:35] and specifying yasm iin the config of course [19:36] I chose "Msys makefiles" now and I get the identical error as with the shell scripts. [19:36] gcc.exe not able to compile a simple test program. [19:36] "paths are POSIX-style except when they're not!" [19:36] maybe your msys environment is broken [19:36] "drive letters may or may not be required!" [19:37] rcombs, thats untrue [19:37] http://pastebin.com/vJ7MKyU9 [19:37] ffmpeg.git 03Michael Niedermayer 07master:970a8f1c256f: avcodec/mjpegdec: Fix integer overflow in shift [19:38] cehoyos, works here with msys+mingw for me... [19:39] What did you do? [19:39] I mean, did you run cmake-gui from msys? [19:39] Or from cmd.exe? [19:39] i double clicked it. [19:40] x265 [info]: build info [Windows][GCC 4.9.1][64 bit] 8bpp [19:40] does it not work if you run 'cmake -G "MSYS Makefiles" ' first, then 'cmake-gui '? both on msys and in the build folder [19:40] How did you pass the path to the toolchain to cmake-gui [19:40] ? [19:41] Assuming you are using Hendriks compiler or aren't you? [19:41] shouldnt matter [19:41] you select 'specify native toolchain' in the wizard thing [19:41] The toolchain doesn't matter? [19:41] and you get file open dialogs to specify teh paths to gcc and g++ [19:42] jamrial, that should work too [19:43] jamrial: I get the identical output as above: gcc is not able to compile a simple test program. [19:43] See http://pastebin.com/vJ7MKyU9 [19:44] then there's something wrong with your gcc toolchain or msys environment [19:44] The same toolchain passes fate for FFmpeg... [19:44] Of course, you are right, I just wonder what I should do. [19:45] My msys is antique, I just wonder how this affects cmake-gui... [19:45] it works without gui too [19:45] running cmake -G "MSYS Makefiles" in msys works fine [19:47] The following shows if I run cmake-gui with a double-click: [19:47] CMake Error: CMake was unable to find a build program corresponding to "MSYS Makefiles". CMAKE_MAKE_PROGRAM is not set. You probably need to select a different build tool. [19:47] ... is gcc even in your path [19:47] actually it doesnt need to be [19:48] because you specify the toolchain path [19:48] It is in my msys path, if I double-click cmake-gui, how can gcc be in my path? [19:48] cmake doesn't recognize "MSYS Makefiles"? [19:48] Or do I just misunderstand? [19:48] Where can I specify the toolchain path (I believe I asked that above)? [19:49] there is an option to specify it when you set the generator [19:49] but as i said, you can just 'cmake -G "MSYS Makefiles"' from the same msys prompt you bult ffmpeg in [19:52] Daemon404: http://pastebin.com/cXvyDcyC [19:53] your environment seems broken then... [19:53] Ok (of course), but how can I proceed? [19:53] And why does fate succeed? [19:54] i have no idea how your stuff is set up [19:54] A very old msys installer [19:55] did you successfully compile ffmpeg on this environment? [19:56] he said he did [19:56] No, I mean yes, as said above, fate passes both for the native (old) compiler on my msys environment and with Hendrik's current compiler. [19:56] I have a second Windows box, which version of msys should I install there? [19:56] idea: make sure your build directory is empty [19:57] especially no *.cache files [19:57] I removed it after every try. [19:57] latest msys using their installer, or just use msys2 [19:57] That is my question. [19:57] both should work. but if you want something in active development, go with msys2 [20:01] Allow me to try again: What did you use when you did a successful install last time, and would you recommend it? [20:01] Daemon404: Would you mind uploading a 64bit x265.a with 10bit support? [20:03] i used msys until a month or so ago, then switched to msys2. i compiled x265 successfully with both [20:04] and i'd say use msys2 [20:05] Ok, thank you. If you can upload a x265.a (64bit, 10bit), I would be grateful. [20:08] ffmpeg.git 03Vincent Bernat 07master:5269cef4087f: avformat/udp: Allow to specify DSCP class [20:08] ffmpeg.git 03Vincent Bernat 07master:d0f8b94b432e: avformat/rtpproto: Allow to specify DSCP class [20:16] cehoyos, why 10bit? [20:17] x265 RD currently has issues in 10bit mode [20:17] TherAk [20:17] ? [20:18] wrong window [20:18] what is rd? [20:18] rate distortion [20:18] basically, 10bit is not prod-ready. [20:18] So there is no reason to test it? [20:19] i dont think so, yet [20:19] Ok, Thank you. [20:19] its on their list to work on [20:19] Any news from the prores file? Where you able to run ffmpeg -i on it? [20:19] Were... [20:20] koda has it, i never remembered to poke him [20:20] Perhaps you could do so now? [20:20] he's not around, it's dinner time here [20:20] ill just leave a message. [20:20] I will ask again, thank you! [20:20] gtg [20:35] maybe in 2025 the aac patch will be pused [20:51] is there any need for more samples that have trouble on the current AAC encoder but work with the patch (and don't give assertion failures)? [20:51] 'cause I've got one [21:12] We don't need any proof that the old one sucks, we know! :P [21:12] nevcairiel: do you need more easily-noticeable proof that it's better now? [22:22] ffmpeg.git 03Anton Khirnov 07master:2443e522f005: lavu: add wrappers for the pthreads mutex API [22:22] ffmpeg.git 03Michael Niedermayer 07master:476027800721: Merge commit '2443e522f0059176ff8717c9c753eb6fe7e7bbf1' [23:00] ramiro: Any progress on that DeckLink issue? [23:21] ffmpeg.git 03Lukasz Marek 07master:568853b8f533: lavf/ffmdec: add common options to recommended encoder configuration [23:21] ffmpeg.git 03Lukasz Marek 07master:500d6088618a: lavf/ffmenc: store recommended encoder configuration [23:22] ffmpeg.git 03Lukasz Marek 07master:3ff39901049f: ffmpeg_opt: make use of recommended encoder configuration [23:30] ffmpeg.git 03Lukasz Marek 07master:f00e9c4b10f5: lavu/opt: add consts where possible [23:42] ffmpeg.git 03wm4 07master:fbd6c97f9ca8: lavu: fix memory leaks by using a mutex instead of atomics [23:43] ffmpeg.git 03Michael Niedermayer 07master:6db8cd8f37b7: Merge commit 'fbd6c97f9ca858140df16dd07200ea0d4bdc1a83' [23:59] ffmpeg.git 03Kieran Kunhya 07master:96fda42a8f9b: vf_interlace: get rid of useless loads [00:00] --- Fri Nov 28 2014 From burek021 at gmail.com Sat Nov 29 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Sat, 29 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141128 Message-ID: <20141129010501.6D67518A02F4@apolo.teamnet.rs> [00:06] Jack64, You mean mpeg2 files from a dvd? [00:08] Jack64: well, if you have the encryption keys, that would be a good start... [00:18] I mean from a stream [00:18] I was just doing a small test and it worked [00:19] so I'm trying to interface with an app that delivers tv streaming [00:19] i login to the service and get the m3u8 file [00:19] and the key [00:19] and the ts streams on the m3u8 file [00:20] then I put them on my local webserver [00:20] and I create a new m3u8 file pointing to the files + key file [00:20] and then i load the m3u8 file on vlc and it worked [00:21] well the point is to make a small relay on my server so i can watch the stream on my tv with XBMC [00:22] because I can't do the authentication there [00:30] Hi [00:34] have you heard about PJPG and 0511? I have two cameras with these "Pixel Format", according to "v4l2-ctl --get-fmt-video" command. And I'm trying to use both at "-input_format" and I'm always getting this error "No such input format: PJPG". Any idea? Thank you in advance. [00:36] what is the point of ffserver when i can only stream 1 file per ? [00:36] i keep failing to stream more than one file using ffserver over the same stream... [00:36] i thought it transcodes everything into the same stream... [00:37] but ffserver doesn't do that... it doesn't even keep clients connected when the stream stops... [00:37] or push new frames to the clients when the broadcast is up again... [00:38] it does absolutely nothing than just existing and /not/ showing the least bit of functionality.. [00:39] ffserver has no functioniality other than being a glorified proxy... why that has even to be configured.. i seriously dont know... [00:40] at least a way to configure a timeout before it breaks over ffmpeg reconnecting would make it 100% more useful than it is now... [00:40] imagine you'd using it to livestream something... [00:41] how fucking often that would go offline... people would /hope/ to see at least a commercial... [00:42] someone should actually try using ffserver... like i have tried the last days... you'd see what a disfunctional piece of software it is.. [00:42] sometimes it just crashes(!).. and goes offline(!)... what server does that? :O [00:45] all i want is ffmpeg to feed completely different sources to ffserver... and connect to ffserver with a client that does play everything (transcoded to the same format and size(s))... [00:45] *it* *does* *not* *happen* [00:46] I heard ffserver is a candidate for removal because it's not good and no one cares about it. [00:47] ffserver, at least, would have to send a blank screen to the playback client until ffmpeg connects with another source... but it doesn't do that... [00:47] sacarasc, if that is so... that would explain *everything* i have gone through... [00:47] i wonder what people use instead... [01:14] I'm getting a blackscreen when I'm trying to re-encode an rtmp stream with libx264 [01:14] any help? [01:29] anyone? [01:31] c_14: http://pastebin.com/iFpTaWD8 [01:32] And the output? [01:32] just audio [01:32] The console output [01:33] c_14: is it because I"m trying to reencode x264 and it says in ffmpeg -codecs that only encoding is available with x264? [01:34] wait, what [01:34] It doesn't show decoding for h.264? [01:34] c_14: I was unable to see the console output due to the command being run by nginx [01:34] can you exec with a &>file ? [01:34] for h264 yes [01:34] but not for libx264 [01:34] libx264 doesn't decode [01:34] libx264 is only the encoder [01:35] the decoder is ffmpeg internal [01:35] ok [01:37] c_14: I'm not sure how I can see the output when the command is being run by another program [01:37] worst case: exec sh -c 'ffmpeg command &>/path/to/file' [01:39] Hi again everyone. Can anyone tell me why I might only be seeing the first frame of an image sequence using image2pip as the input format, mpeg4 as the vcodec, and m4v as the output file? I ame using -framerate 25 and -r 25 [01:40] If I do the same exact parameters but set the output format to mov, I can see both frames [01:40] The only thing I can think of is a bug in the muxer. [01:43] hold on [01:43] almost done with printing output [01:45] Ok, if it helps Im using this exact command: ['/usr/local/bin/ffmpeg', '-framerate', '25', '-i', 'test_rand/11-small.png', '-i', 'test_rand/25-small.png', '-f', 'image2pipe', '-filter_complex', 'concat=n=2:v=1:a=0', '-vcodec', 'mpeg4', '-b:v', '800k', '-r', '25', '-f', 'm4v', '-y', 'pipe:1'] [01:45] can I just state -vcodec with no other options for video? [01:46] MrSavage: normally yes [01:47] gabriel_: does the same problem occur when writing to a file [01:47] Have you tried setting -framerate for the second input as well? [01:47] Yeah, if the output format is the same [01:47] ohhhhh [01:47] interesting [01:47] let me try that [01:49] c_14: this is not writing a log exec sh -c 'ffmpeg -i rtmp://localhost:1935/$app/$name -vcodec libx264 -acodec copy -f flv rtmp://localhost:1935/small/test &> /usr/local/nginx/logs/ffmpeg.log'; [01:50] hmm, is the file being created? [01:51] Does the nginx process have write permissions there? [01:51] c_14: like this? ['/usr/local/bin/ffmpeg', '-f', 'image2pipe', '-framerate', '25', '-i', 'test_rand/11-small.png', '-framerate', '25', '-i', 'test_rand/25-small.png', '-filter_complex', 'concat=n=2:v=1:a=0', '-vcodec', 'mpeg4', '-b:v', '800k', '-r', '25', '-f', 'm4v', '-y', 'pipe:1'] [01:52] What's that first -f image2pipe doing there? [01:52] unless 11-small.png is a fifo [01:53] Also, I don't think you can't write mp4 to a pipe [01:53] c_14: it should have permission. I'm executing as root [01:53] MrSavage: is the file created? [01:53] c_14: no [01:54] try replacing the command with exec sh -c 'touch /usr/local/nginx/logs/ffmpeg.log' [01:54] c_14: basically Im only interested in the resulting filesize, so I dont want to write the resulting file to disk. I thought the -f image2pipe was neccesary to get the resulting bytes into memory [01:55] You're using the image2pipe as an input format, that should only be necessary when you're actually receiving images from a pipe [01:55] c_14: still no file [01:55] Im able to pipe the results to output.m4v and play it with vlc but thats all I know [01:55] Ok so I just want -f image [01:55] 2 [01:56] MrSavage: so either nginx is jailing the exec or it's failing outright, what about exec touch /usr/local/nginx/logs/ffmpeg.log [01:56] gabriel_: ye, but it should normally autodetect that in most cases [01:57] c_14: All I care about is using mpeg compression, the output file doesnt matter because I just want the resulting filesize. Do you have any tips? [01:57] just pipe it out raw... [01:57] -f raw [01:58] Ill try that. [01:58] should work [01:58] if not, use mkv or mpegts [01:59] http://pastebin.com/2FNmHXrB [01:59] Thats interesting. Why is input#1 a png_pipe? [02:00] c_14: exec touch /usr/local/nginx/logs/ffmpeg.log; does not work [02:01] gabriel_: ehhh, no clue, but use -f mpegvideo [02:01] mpegts seemed to work [02:02] MrSavage: can you try /tmp/ffmpeg.log ? in case nginx is dropping permissions [02:02] (assuming /tmp is 777, which it usually is) [02:10] c_14: I see it in /tmp/ [02:11] This is driving me bonkers. I fixed the input error, but Im still only seeing the first frame! :/ [02:12] c_14: this doesn't give me anything in the file though exec sh -c 'ffmpeg -i rtmp://localhost:1935/$app/$name -vcodec libx264 -acodec copy -f flv rtmp://localhost:1935/small/test &> /tmp/ffmpe.log'; [02:14] MrSavage: try exec FFREPORT=/tmp/ffmpeg.log ffmpeg -i rtmp://[..] -c:v libx264 -c:a copy -f flv rtmp://[..] ; [02:16] c_14: the file isn't being created now [02:16] eh, FFREPORT=file=/tmp/ffmpeg.log [02:18] c_14: still not creating the file [02:18] gabriel_: maybe https://trac.ffmpeg.org/wiki/Create%20a%20video%20slideshow%20from%20images#Ifyourvideodoesnotshowtheframescorrectly ? [02:18] MrSavage: geh, does nginx have an errorlog? [02:18] Maybe it's throwing errors there [02:18] Action: c_14 hopes [02:20] c_14: it does but it doesn't show anything related to ffmpeg [02:22] I'm out of ideas. If I could get the ffmpeg console output I might be able to help. Without it, I can only guess. [02:23] And my guess is pretty limited to "try updating ffmpeg". [02:24] -vf/-af/-filter and -filter_complex cannot be used together for the same stream. Haha nice. I will try using filterchain [02:28] Now Ive confused myself [02:31] c_14: Do you know how do get round that error? [02:32] -filter_complex 'concat=n=2:v=1:a=0,fps=25' [02:32] Oh, doesnt that set the framerate on the input stream, not the output stream? [02:33] hmm, nah [02:51] c_14: You still there? Im not sure thats doing what I want it to do. The resulting video has no frames now. :P Hey thanks for you help [02:52] eh [02:52] that shouldn't be happening... [02:54] Yeah, let me fetch my command for you [02:56] ['/usr/local/bin/ffmpeg', '-f', 'image2', '-framerate', '25', '-i', 'test_rand/11-small.png', '-f', 'image2', '-framerate', '25', '-i', 'test_rand/12-small.png', '-filter_complex', 'concat=n=2:v=1:a=0,fps=25', '-vcodec', 'mpeg4', '-b:v', '800k', '-f', 'm4v', '-y', 'pipe:1'] [02:56] Try changing the format to mpegvideo or mpegts? [02:56] isnt mpegts for mpeg2? [02:57] Requested output format 'mpegvideo' is not a suitable output format [02:57] pipe:1: Invalid argument [02:57] geh [02:57] Use matroska then [02:58] This is madness XD [02:59] Hey were back to just the first frame with matroska [02:59] Just want to see if it's the format. [02:59] nothing else changed but the format [02:59] try adding the 2nd png as another input and change concat to n=3 [02:59] don't mind the dirty hack [02:59] :O [03:01] ['/usr/local/bin/ffmpeg', '-f', 'image2', '-framerate', '25', '-i', 'test_rand/11-small.png', '-f', 'image2', '-framerate', '25', '-i', 'test_rand/12-small.png', '-f', 'image2', '-framerate', '25', '-i', 'test_rand/12-small.png', '-filter_complex', 'concat=n=3:v=1:a=0,fps=25', '-vcodec', 'mpeg4', '-b:v', '800k', '-f', 'matroska', '-y', 'pipe:1'] [03:02] Still no luck [03:02] Still only one frame? [03:02] Yeah! so crazy [03:03] Can you try changing the -framerate 25 s to -framerate 1 s? [03:03] yeah [03:03] still a 0 second output with just one frame [03:04] I must be doing something wrong... [03:05] What version you running? [03:05] Cause my version throws an error with the command like that... [03:06] unless I messed something else up. [03:06] nvmd [03:37] Right, if gabriel_ comes back and I'm not here, could someone tell him to use the concat demuxer like so: `ffmpeg -f concat -i concat -r 25 -c:v mpeg4 -b:v 800k -f matroska pipe:1' ? (or in the hope he reads the logs...) [03:54] hello [03:55] i have problem record audio with ffmpeg [03:55] the line is 'ffmpeg -f alsa -ac 2 -i pulse -f x11grab -s 800x600 -r 30 -i :0.0 -vcodec libx264 -strict -2 ./out.mp4', have alsa and pulseaudio, all sound function correctly and i listen all (also mplayer), only have problem in ffmpeg for record [03:57] if without -strict -2 , the ffmpeg requesting obligatory [03:57] any ideas? [04:05] c_14: Did you get it to work? [04:06] (on your machine) [04:08] ffmpeg version 2.4.2 Copyright (c) 2000-2014 the FFmpeg developers [04:20] Ive upgraded to 2.4.3, ill see if that changes anything [04:23] Hmmm nothing. [04:28] Can anyone help me understand why this is happening? I would expect the output stream to have 3 frames, but it says it only has 1. http://pastebin.com/S9zEdSRH [04:29] i have a problem with record sound with ffmpeg... said most upper chat [04:30] croco: pick a different audio codec [04:30] to avoid the -strict -2 [04:31] for the audio problem, open pavucontrol and see what stream ffmpeg is recording [04:31] change it to "Analog Input" or something similar [04:31] what audio? if avoid the -strict, ffmpeg try request [04:32] i see /proc/asound/cards and arecord -l/-L [04:33] you are using pulse though [04:33] have alsa and pulse audio, i trying use -f alsa and -i pulse [04:33] so you have to open pavucontrol to control what ffmpeg records [04:33] i not have pavucontrol [04:34] can you install it with your package manager? [04:34] if you don't want to install it, you will have to use pactl/pacmd and read the manpages [04:36] klaxa, i have install the pavucontrol and see in the config, not have none of ffmpeg [04:36] start recording with your command [04:36] then open the "Recording" tab in pavucontrol [04:39] klaxa, ok, and i what checking? [04:39] on the right side there is a dropdown menu [04:40] set that to "Analog Stereo" [04:40] the record say ALSA plug-in [ffmpeg]:Alsa capture from [04:40] yes, on the right there should be a menu [04:40] just select... something i guess [04:40] try changing it until you are recording what you want to record [04:41] i have two: "monitor of audio intern" (listen) and "audio intern" (not listen) [04:41] oh, do you want to record your computer's sound? or your microphone's sound? [04:41] the record from computer sound [04:42] then select "monitor of audio intern" [04:43] klaxa, change option, and equal not record sound [04:44] huh? that is weird... can you pastebin your command and the output? (use pastebin.com or something similar) [04:44] ok [04:45] klaxa: do you have a quick second to look at the pastebin above? Im totally at a loss [04:46] http://pastebin.com/PNvmHN8m [04:48] croco, can you try using -f pulse instead of -f alsa? [04:48] gabriel, sorry, i don't know either [04:49] So weird. I dont understand it at all [04:49] maybe something with mp4 not supporting low framerates? [04:49] have you tried using mkv instead? [04:49] It doesn the same thing with PAL or NTSC framerates [04:49] and with mkv [04:49] klaxa, the pavucontrol none output in the record [04:50] ffmpeg has to be running to show up in pavucontrol [04:51] klaxa, http://pastebin.com/xjFec9jN [04:52] klaxa: do you mean m4v or do you mean mkv? [04:52] klaxa, not show the ffmpeg in pavucontrol [04:52] croco: are you making sure there is sound to record? are you playing music or something? [04:52] klaxa, yes [04:52] gabriel: mkv, i'm currently trying with some pictures myself [04:52] I dont see mkv in ffmpeg -formats, could you explain what it is? [04:53] croco: can you take a screenshot of pavucontrol's record tab while ffmpeg is running? also change the menu at the bottom right to "All Streams" before doing so [04:53] gabriel: mkv is matroska [04:53] oh, thanks [04:53] Ive tried with matroska as well. D: [04:54] klaxa, with '-f pulse' only the ffmpeg not record sound (only have #0 for Video and none for sound) [04:55] oh right, i didn't see that... i wonder why that is [04:55] ah, try -f pulse -i default [04:55] instead of -f pulse -i pulse [04:55] ok [04:58] klaxa: any luck? [04:58] none so far [04:59] klaxa, with '-f pulse -i default' function sound! :D now only have problem with phase sincronization with Audio and Video [05:02] gabriel, i also only get one frame... weird... [05:03] Do you know who might have some clue as to why this is happening? [05:03] err... not me. [05:04] however, what you could do instead would be: cat file1.jpg file2.jpg file3.jpg > ffmpeg -f image2pipe -i - out.mp4 [05:04] Oh really? [05:04] whats the -i? [05:05] wait... [05:05] i'm sure this used to work [05:05] the - is standard input [05:07] the sincro between audio and video not import so. something is that function the sound :) Very Thanks klaxa [05:07] np :) [05:08] gabriel, i'm starting to think it's filter_complex's fault [05:09] if i use the concat protocol instead of the filter, i get 3 frames [05:09] ooooo [05:09] Could you explain how to use that? [05:09] if i then add filter_complex to change the framerate, i get one frame [05:09] that sounds promising [05:10] http://pastebin.com/s9Rfa4MM [05:10] that's what my concat.txt looks [05:10] ffmpeg -f concat -i concat.txt -c:v libx264 test.mkv [05:10] that was the command used [05:11] let me try it with the mpeg4 encoder [05:14] Odd& I cant get that to work. [05:15] It runs but I get 0 frames using mkv, 1 frame using m4v [05:17] Oh my gosh. It might be a vlc bug [05:17] klaxa: what are you using to play the resulting mkv file? [05:18] mpv [05:28] is it possible to use the concat protocal from stdin? [05:28] rather than a text file [05:29] yes. [05:37] Haha yes! it feels a little hacky, but Ive managed to work around it [05:37] klaxa: do you think this is worth raising an issue over? [05:37] thanks for you help, by the way. [05:38] maybe try the user-mailinglist first [05:38] i think i would probably file a bugreport [05:38] if it's not a bug, then maybe it should be documented somewhere [05:42] Yeah, its definitely been confusing. I still dont really understand why one works over the other [05:48] Ill put something in the user-mailing list [07:53] Hello :) [08:18] man cross-compiling is hard :( [08:45] How do you set the -- if [[ "$non_free" = "y" ]]; then ; in a script ? [10:20] Hi all, I would to learn how to code with vdpau and ffmpeg api... Can you link me some examples? I didn't found anything, only documentation but it is very small [10:48] What is the recommended container to publish a live stream with libx265 ? (since FLV is not compat with this) [10:55] is there a way to make ffserver not take an input feed and just redirect to an ffmpeg that's outputing to rtp itself? [11:24] yo [12:28] hello [12:29] when ffmpeg is built with an instruction-set specific option such as --enable-mmx, does it mean mmx instructions are used whenever possible, or just that such code is compiled in and can be run if runtime cpu detection allows it ? [12:57] braunr, afaik --enable-mmx will enable compilation options in a way where MMX is always used [12:58] I think only libswscale and libx264 can do dynamic detection [13:03] Mavrik: hm, not good [13:05] braunr, what is your usecase and which encoders are you trying to use? [13:08] hi [14:03] I want to create a live event, transmit audio and video. I'm planning to use a Zoom h4n audio recorder as usb mic and a GoPro 3 Wi-Fi (I'd like to move the camera around the room), but before buying this equipments, I'm trying with the desktop mic and my D-Link DCS-930L IP Camera. Until now I just managed to write it on a file http://pastebin.com/vryFNUYP . Can anyone show me the way? Thank you in advance. [14:11] DanielSa, there are a bundle of streaming guides on the internet.... maybe try one and see where you get stuck? [14:18] davidw: Yes I did it yesterday, until late at night :D so was when I start doubt if this 'architecture' works, retransmit a video from a IP Camera to the avserver/ffserver. [14:22] Mavrik: i maintain a home-made linux distribution at a company, where userspace must be the same binary for a range of different boards [14:23] microchip_: for now it's only about decoding and playing videos [14:23] woops [14:23] Mavrik: ^ [14:24] Mavrik: my current idea is to build ffmpeg libraries with all optimizations along with runtime cpu detection, so that the same binaries work everywhere with optimized instructions [14:26] As I said, I think detection only works for swscale, but don't hold my word for it, I've checked it in version 1.2 last time :P [14:28] Mavrik: i understand :) [14:30] DanielSa, somehow or another, these things do work [14:30] you just have to pound your head against the tablle a lot [14:41] anyone got time to help out an unexperienced user with an explanation? [14:41] waNNaBe: just ask [14:41] (and be patient) [14:44] ok, thx. i'm struggling with a simple task. converting an .avi containing a single cvid video stream with rgb24 color model to a .mp4 with h264 encoding [14:44] for some reason ffmpeg won't let me use any h264 profile other than high444 [14:45] seems like with every profile i use, the encoder tries to use yuv444p color model (which is then a wrong parameter to e.g. baseline profile) [14:46] my question to this is, whether it might be a bug or I am just lacking some insights into this stuff [14:53] waNNaBe: it's quite specific, you should also try the mailing list i guess [15:01] I'll try, thx. [15:18] question, if i am capturing audio from a composite in card and video from an ipcam (network) will there be a out of sync between the audio/video? assuming the network is a local network with good latency? i mean i dunno if data trasnfered over the audio cable would be faster than the ipcam which is encoding and sending video over the network cable [15:19] depends on the camera and microphone. There is no general answer to that. [15:20] BtbN: what do you mean [15:20] how can i make sure they are in sync [15:22] measure the difference and add an apropiate offset. [15:44] let me make it a general question, how can i make capturing an audio stream over analog and a videostream over digitial be in harmoney and synch'd [15:46] measure the difference and add an apropiate offset. [15:57] BtbN: the difference would be constant? [16:00] maybe [16:00] like i said, there is no general answer to that. [16:11] Hi everyone, I'm having issues trying to encode a stream's video into libx264 [16:12] as it's giving me a black screen, but I can receive the stream fine and hear the audio but only see black [16:12] I'm putting in this ffmpeg -i rtmp://localhost:1935/$app/$name -c:v libx264 -c:a copy -f flv rtmp://localhost:1935/small/test [16:13] it's giving me, Failed to set value 'libx264' for option 'c:v' [16:14] i guess i'll try avonv [16:14] avconv* [16:18] well i'll try to get it to work with ffmpeg since there's no #avconv channel [16:18] So far my command is this and still a blackscreen: exec ffmpeg -i rtmp://localhost:1935/$app/$name -vcodec libx264 -acodec copy -f flv rtmp://localhost:1935/small/test [16:19] check ffmpeg -encoders [16:19] if you actualy have the x264 encoder [16:20] BtbN: I have it [16:20] also here's my log output so far http://sprunge.us/WZIc [16:20] That's libav, not ffmpeg [16:21] what do you mean? It says ffmpeg at the top [16:21] and the command is ffmpeg [16:21] it says Libav in the very fist line [16:23] but isn't it still ffmpeg? [16:23] no [16:23] just read the first few lines it print [16:23] libav is a fork, that's similar but not the same [16:23] OH [16:24] Thanks [16:26] Apart from that, that log looks successfull [16:26] only problem might be that you specified no bitrate, and 0.4Mbit/s is _very_ low for 720p [16:27] BtbN: would the maxbitrate command be sufficient? [17:03] Can I ask regarding the audio/sync issue in MP4 (x264)? [17:12] hi [17:13] I would like to convert lossless mkv 67 minutes video to 5000kbit/s, what would the command line be for that? [17:14] right now I use something like http://www.pasteall.org/55392 to convert videos to a smaller file, but VIMEO can use 5000kbit/s at it's best [17:16] zybi1, add -maxrate 5000k -bufsize 5000k [17:16] to limit maximum bitrate to 5000k [17:16] maybe take 4800 just in case [17:17] other stuff looks ok [17:42] Anyone know where there's a webpage explaining advopts of libx264? [17:42] I'm trying to put CBR on [18:09] Mavrik: thanks! so maxrate AND bufsize 4800 or only maxrate? [18:20] hi [18:54] hi, how can I mux an h.264+AC3 streams from mkv to avchd (.mts or .m2ts)? [19:16] Mavrik: it resulted in a file with overall bitrate of only 3072 kbps and very poor quality [19:19] I would like to convert lossless mkv video (67 minutes, 29giB) to a video file of 5000-10000kbit/s, what would the command line be for that? [19:19] see http://www.pasteall.org/55395 what I tried so far [20:05] zybi1: you're missing a zero in your maxrate [20:13] Toropisco: ffmpeg -i file -c copy -f mpegts out.m2ts [20:16] hey there. I'm trying to live-transcode a transport stream from my sat-ip server (Digibit R1). This is the stream address I use as input: http://pastebin.com/24cQeVbU Unfortunately, ffmpeg doesn't seem to read the input stream. Any hints? [20:20] If you curl the stream to a file, can ffmpeg decode said file? [20:21] Yes [20:21] but ffmpeg -i http://url doesn't work? [20:22] no, it just lists all the enabled libs but then nothing happens [20:22] !ub mopp [20:28] here is the complete console output: http://pastebin.com/fMkP2Bpu [20:30] Surround the url in '' [20:30] I'm pretty sure the & is causing your shell to fork. [20:33] Yup, working now :) only a LITTLE embarassing. Thank you very much! [20:39] c_14, thanks [20:43] hello [20:44] i am the same of day after. [20:44] have a problem with sincronism between video and sound [20:45] use -f pulse -i default, and probe -async with non result. [20:47] not sincro between video and sound, i get '[swscaler @ 0x8369060] Warning: data is not aligned! This can lead to a speedloss' in output text [20:48] http://pastebin.com/V9B2Rj8P [20:48] ? [21:08] what is the problem with sincronism between audio and video?? [21:19] thanks c_14, but now I tried again resulting in only 4051 kbps again - http://www.pasteall.org/55397 [21:23] Try lowering the crf? [21:25] Also what version of FFmpeg are you using?^ [21:27] croco: can your pc encode h264 in realtime? [21:27] maybe try adding -preset fast, veryfast or superfast? [21:29] c_14, yes [21:30] use this command: 'ffmpeg -f pulse -i default -ac 2 -f x11grab -s 800x600 -r 30 -i :0.0 -vcodec libx264 -strict -2 ./out.mp4' probes with -preset with same result [21:31] the encode realtime not affect the efficiency of computer [21:31] croco: have you tried outputting to mkv? [21:31] But to me it looks like pulse is giving you weird pcm. [21:32] no, use mp4 and detect automaticaly h264 (or libx264) [21:34] the pcm_s16le is (i see) a default sound input source. [21:35] also probes change output codec in the same pcm_s16le and getting error of pcm_s16le not exist [21:36] [aac @ 0x8590ae0] Queue input is backward in time [21:36] [mp4 @ 0x858ecc0] Non-monotonous DTS in output stream 0:1; previous: 199391, current: 198759; changing to 199392. This may result in incorrect timestamps in the output file. [21:37] this various lines outputs [21:37] and get init forever this line: '[swscaler @ 0x84ef060] Warning: data is not aligned! This can lead to a speedloss' [21:38] retrace sound 0,5 second with video [21:42] and not most retrace in the time, not needed -preset, the computer is very fast and none lowest for ffmpeg, intentional retrace sound 0,5 second with the video [21:47] hmm, no clue. sorry [23:40] hey, still trying to live transcode from my Sat-IP server. However, now I can't get "x264opts videofilter" to work. Here the complete output: http://pastebin.com/56WpT8W0 [23:47] I think you need video-filter=select_every:2,1 [23:53] hm, no, now I get "Error while opening encoder for output stream #0:0 - maybe incorrect parameters [23:53] such as bit_rate, rate, width or height" [23:54] which is strange, because the stream is recognized correctly with all parameters [00:00] --- Sat Nov 29 2014 From burek021 at gmail.com Sat Nov 29 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Sat, 29 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141128 Message-ID: <20141129010502.7287418A0321@apolo.teamnet.rs> [00:00] ffmpeg.git 03Michael Niedermayer 07master:449bd3c0c33c: Merge commit '96fda42a8f9bf84beaaf7f5991d17f2a057de86c' [00:12] Zeranoe: new patch sent in your forum [00:28] ffmpeg.git 03Anton Khirnov 07master:1973079417e8: opusdec: make sure all substreams have the same number of coded samples [00:28] ffmpeg.git 03Michael Niedermayer 07master:fae545ca2a34: Merge commit '1973079417e8701b52ba810a72cb6c7c6f7f9a56' [00:57] ffmpeg.git 03Anton Khirnov 07release/2.4:12700b021952: mp3enc: fix a triggerable assert [00:57] ffmpeg.git 03Anton Khirnov 07release/2.4:46a17d886b85: lavu: add wrappers for the pthreads mutex API [00:57] ffmpeg.git 03Michael Niedermayer 07release/2.4:86b532898d8a: Merge commit '12700b0219521a5f20c8ba47b3ad7857ea9e0554' into release/2.4 [00:57] ffmpeg.git 03Michael Niedermayer 07release/2.4:15fd62110fd8: Merge commit '46a17d886b8559723c40b9f5cdf0e0c6b1c95180' into release/2.4 [01:15] ffmpeg.git 03wm4 07release/2.4:517ce1d09b5e: lavu: fix memory leaks by using a mutex instead of atomics [01:15] ffmpeg.git 03Michael Niedermayer 07release/2.4:b6691fba77af: Merge commit '517ce1d09b5e6b72afc2ef9490b5f8ca42fa6a65' into release/2.4 [01:15] ffmpeg.git 03Michael Niedermayer 07release/2.4:f783259fdb37: avutil/buffer: use the old atomics based code for the release branch [01:26] ffmpeg.git 03Anton Khirnov 07release/2.4:ca78ee73db9e: opusdec: make sure all substreams have the same number of coded samples [01:26] ffmpeg.git 03Michael Niedermayer 07release/2.4:0e216ed40789: avutil/buffer_internal: leave the buffer pool entries volatile [01:26] ffmpeg.git 03Michael Niedermayer 07release/2.4:56b84b023d70: Merge commit 'ca78ee73db9e059f501706ba6108e23902e84933' into release/2.4 [02:09] ffmpeg.git 03Timo Rothenpieler 07master:95fc80672ffe: Move extralibs variables using ldl after ldl definition [02:09] ffmpeg.git 03Benoit Fouet 07master:7dfee8d69793: avcodec/pngdec: split P frames handling to a separate function. [02:09] ffmpeg.git 03Benoit Fouet 07master:cfd83a8af6bf: avcodec/pngdec: add support for 'over' blend operation for 'none' dispose operation. [04:17] ffmpeg.git 03Michael Niedermayer 07master:ea38e5a6b757: avcodec/hevc_ps: Check num_long_term_ref_pics_sps [11:56] ffmpeg.git 03Martin Storsj? 07master:857e6667f906: rtmpproto: Clarify a comment [11:57] ffmpeg.git 03Michael Niedermayer 07master:a105c1f204bd: Merge commit '857e6667f9061ae261c0b951113e4efc4329b05e' [12:13] ffmpeg.git 03Martin Storsj? 07master:3c3b8003a13d: rtmpproto: Simplify code for copying data into the output packet [12:14] ffmpeg.git 03Michael Niedermayer 07master:42a095d095b5: Merge commit '3c3b8003a13d9c3668c0bb6d79d2376da3b2b352' [12:25] ffmpeg.git 03Martin Storsj? 07master:44127b157e9f: rtmppkt: Make pkt->data reallocable [12:25] ffmpeg.git 03Michael Niedermayer 07master:900fff89c9ed: Merge commit '44127b157e9f8acb837d4bb3a094f56b40da3ef5' [13:04] ffmpeg.git 03Bryan Huh 07master:a9d8d35e4833: dashenc: Add options to make segment names configurable [13:04] ffmpeg.git 03Martin Storsj? 07master:fcae9f212a60: dashenc: Avoid a VLA-like construct [13:04] ffmpeg.git 03Michael Niedermayer 07master:5ce070c16c72: Merge commit 'a9d8d35e4833fc4dfbf557ce73c84e9ca6224427' [13:04] ffmpeg.git 03Michael Niedermayer 07master:f001a2bd344e: Merge commit 'fcae9f212a6001d966c52dc22cd4b22e9851b428' [15:13] Action: Compn forgets timothy gu's irc nick [15:14] Timothy_Gu [15:14] shocking, i know [15:14] ah [15:14] so not online [15:14] thus, he does not exist! [15:15] i cant remember: did we move to his nodejs fate? [15:15] dunno, fate server would probably look different if so ? [15:15] no [15:15] html/css would be the same probably [15:15] it's a new backaend [15:15] fate looks diff [15:15] ah [15:15] backend* [15:25] oh carl isnt online when i finally get him is dang prores thing [15:25] OH WELL [15:29] Action: koda still is unsure what he needs it for [15:34] Daemon404 : you got the sample or what ? [15:34] where is it ? [15:34] i'll relay it to him [15:34] he wanted output from a shitty prores-in-avi sample [15:34] ill just wait [15:34] whatever prores thing is :P [15:34] ok [15:35] Compn: i got the sample, but i cannot share it [15:36] Action: koda also wonders why people are asking Daemon404 rather than the patchs author [15:43] koda : i have no idea what is going on, i was just offering my services since it seems like Daemon404 is tired of non-instant communication with some people :P [15:43] Action: koda hugs Compn [15:44] ffmpeg.git 03Rong Yan 07master:57c89c50bd2f: avcodec/ppc/h264dsp: POWER LE support for h264_idct8_add_altivec() h264_idct_dc_add_internal() h264_loop_filter_luma_altivec() write16x4() VEC_1D_DCT() weight_h264_W_altivec() biweight_h264_W_altivec() VEC_LOAD_U8_ADD_S16_STORE_U8() ALTIVEC_STORE_SUM_CLIP() [15:44] ffmpeg.git 03Rong Yan 07master:89f3043c7fa6: avcodec/ppc/h264chroma_template: POWER LE support for PREFIX_h264_chroma_mc8_altivec() PREFIX_no_rnd_vc1_chroma_mc8_altivec() CHROMA_MC8_ALTIVEC_CORE_SIMPLE() CHROMA_MC8_ALTIVEC_CORE() [15:44] ffmpeg.git 03Rong Yan 07master:8cc5a78e45d9: avcodec/ppc/h264qpel_template: POWER LE support for PREFIX_h264_qpel16_h_lowpass_altivec() PREFIX_h264_qpel16_v_lowpass_altivec() PREFIX_h264_qpel16_hv_lowpass_altivec() [15:44] ffmpeg.git 03Rong Yan 07master:bd67d0ead145: avcodec/ppc/h264qpel: POWER LE support for put_pixels16_l2_altivec() and avg_pixels16_l2_altivec() [15:44] in my capacity as a 'dev' , i pretty much have no idea what is going on 100% of the time. [15:44] ehehe [15:55] nevcairiel, looked into schannel.... my god [15:57] koda: :) [15:57] kierank: o/ [16:16] Welcome to the dark side [16:27] ffmpeg.git 03Benoit Fouet 07master:aff50ae1d176: avcodec/pngdec: do not blend on transparent black [16:27] ffmpeg.git 03Benoit Fouet 07master:af8804ac8638: avcodec/pngdec: allow for some code path optimizations. [16:27] ffmpeg.git 03Michael Niedermayer 07master:885a763cacf5: avcodec/ppc/h264qpel_template: protect unistd.h by #if HAVE_UNISTD_H [16:59] Hi all, I'm trying to learn to code with ffmpeg and vdpau API. I'm looking for some examples and found the ffmpeg_vdpau.c file in ffmpeg.org... What is it? how can I use that? [17:02] Hi all, I'm trying to learn to code with ffmpeg and vdpau API. I'm looking for some examples and found the ffmpeg_vdpau.c file in ffmpeg.org... What is it? how can I use that? [17:02] That's the vdpau code for the ffmpeg cli util as far as i know. [17:05] BtbN: do you know any basic player as ffplay.c that use vdpau? [17:05] no [17:05] using vdpau with ffmpeg is complex and not well documented [17:06] ffmpeg_vdpau.c is not very long, and should serve as an example on how to implement it more than actual useful function for ffmpeg itself [17:10] nevcairiel: yeah but there aren't comments about the use case of that functions so how can I understand it? [17:11] aleskandro_, https://gist.github.com/BtbN/517f7d11aad228f2eb9d that's from one of my projects, it uses libavcodec with vdpau. Look for the HAVE_VDPAU blocks. [18:32] ffmpeg.git 03Benoit Fouet 07master:ebf2052a6177: avcodec/pngdec: apng: fix output buffer filling when no blending is needed. [19:13] Action: ubitux wonders if there is a relationship between AV_AUDIO_SERVICE_TYPE_* and AV_DISPOSITION_* [19:13] i doubt it [19:14] AV_DISPOSITION_HEARING_IMPAIRED [19:14] AV_DISPOSITION_VISUAL_IMPAIRED [19:14] AV_AUDIO_SERVICE_TYPE_VISUALLY_IMPAIRED [19:14] AV_AUDIO_SERVICE_TYPE_HEARING_IMPAIRED: [19:14] also AV_AUDIO_SERVICE_TYPE_KARAOKE vs AV_DISPOSITION_KARAOKE [19:14] etc.. [19:15] lol awesome [19:15] i just don't understand why Anton seem to be targetting random pkt side data with this information [19:16] instead of just flagging the stream [19:16] there is probably a good reason but well. [19:16] i probably miss something [19:17] Action: Daemon404 stabs schannel more [19:17] no wonder nevcairiel never did this [19:18] I'm happy I have nfi about wtf that is [19:18] schannel? [19:18] ubitux: ask him? :) [19:18] koda: i'm on his ignore list afaik [19:19] because i'm "demotivating" him last i heard [19:19] he's not interested in talking to me and even less helping me, so i'm not going to ask him anything [19:19] i meant on the mailing list [19:20] i'm not going to force myself in here [19:21] dunno, try helping him so hell help you =) [19:21] afaik it has something to deal with the avcodecparameters [19:22] it's not important, i'll eventually have a look another time [20:02] ffmpeg.git 03Anshul Maheswhwari 07release/2.1:712b8e712828: v4l2enc: adding AVClass [20:02] ffmpeg.git 03Michael Niedermayer 07release/2.1:d5eca1651f4a: avcodec/dvdsub_parser: never return 0 when the input isnt 0 [20:02] ffmpeg.git 03Michael Niedermayer 07release/2.1:f99675627d1c: avcodec/dvdsub_parser: Check buf_size before reading 32bit packet size [20:02] ffmpeg.git 03Michael Niedermayer 07release/2.1:384be84bbe8a: avcodec/dvdsub_parser: print message if packet is smaller than the packet size field [20:02] ffmpeg.git 03Michael Niedermayer 07release/2.1:ded44bda27ec: ffmpeg_opt: Use av_guess_codec() instead of AVOutputFormat->*codec [20:02] ffmpeg.git 03Michael Niedermayer 07release/2.1:72149fcb1b79: avformat/tee: flip assigment direction [20:02] ffmpeg.git 03Michael Niedermayer 07release/2.1:7d69132775e6: avcodec/wavpackenc: Fix log2sample() result value [20:02] ffmpeg.git 03Michael Niedermayer 07release/2.1:150ae7692efc: ffserver: initialize pbuffer in prepare_sdp_description() [20:02] ffmpeg.git 03Anton Khirnov 07release/2.1:f7ed48938a33: cdgraphics: do not return 0 from the decode function [20:02] ffmpeg.git 03Michael Niedermayer 07release/2.1:f405267493c8: avcodec/iff: check pixfmt for rgb8 / rgbn [20:02] ffmpeg.git 03Michael Niedermayer 07release/2.1:c9da441eb460: avcodec/snow: fix null pointer dereference in cleanup after allocation failure [20:03] ffmpeg.git 03Christophe Gisquet 07release/2.1:e1ed566c1cda: proresenc_kostya: remove unneeded parameters [20:03] ffmpeg.git 03Christophe Gisquet 07release/2.1:92096acc0a04: proresenc_kostya: report buffer overflow [20:03] ffmpeg.git 03Christophe Gisquet 07release/2.1:407982b8f945: proresenc_kostya: properly account for alpha [20:03] ffmpeg.git 03Christophe Gisquet 07release/2.1:656bf0ca79ba: wavpack: report if there is no bits left [20:03] ffmpeg.git 03Christophe Gisquet 07release/2.1:80b6632b360a: wavpackenc: proper buffer allocation [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:6ab793c2b9bb: avcodec: fix aac/ac3 parser bitstream buffer size [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:cb8f645fafec: avcodec/utils: add GBRP16 to avcodec_align_dimensions2() [20:03] ffmpeg.git 03Timothy Gu 07release/2.1:0f8863df8640: bktr: Fix Fabrice's name [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:8c8950f982c5: avcodec/snow: check coeffs for validity [20:03] ffmpeg.git 03wm4 07release/2.1:17f30ab6a154: oggdec: fix invalid free on error [20:03] ffmpeg.git 03Mika Raento 07release/2.1:cb10e05ff06e: segment: don't access outside seg->frames array [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:255ebf3aff52: avformat/swfdec: Use side data to communicate w/h changes to the decoder [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:72f0d13802ba: avformat/swfdec: Do not change the pixel format [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:f7086be79a59: avcodec/h264: Allow partial escaping [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:7644b292222c: avcodec/mpegvideo: Use "goto fail" for all error paths in ff_mpv_common_frame_size_change() [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:31f5d5f6903c: avcodec/mpegvideo: check that the context is initialized in ff_mpv_common_frame_size_change() [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:cd190f0c3fd3: avcodec/mpegvideo: Set err on failure in ff_mpv_common_frame_size_change() [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:5bd45a1b27dd: avformat/m4vdec: Check for non startcode 00 00 00 sequences in probe [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:5e4a821b8a41: tools/crypto_bench: fix build when AV_READ_TIME is unavailable [20:03] ffmpeg.git 03Katerina Barone-Adesi 07release/2.1:36ec1c2c5519: apetag: Fix APE tag size check [20:03] ffmpeg.git 03Pascal Massimino 07release/2.1:5ecc4a644a1f: libavcodec/webp: treat out-of-bound palette index as translucent black [20:03] ffmpeg.git 03James Almer 07release/2.1:84487650e23b: x86/dsputil: add emms to ff_scalarproduct_int16_mmxext() [20:03] ffmpeg.git 03Gianluigi Tiesi 07release/2.1:34ef754854ee: avcodec/libilbc: support for latest git of libilbc [20:03] ffmpeg.git 03Pascal Massimino 07release/2.1:77367f27280c: avcodec/webp: fix default palette color 0xff000000 -> 0x00000000 [20:03] ffmpeg.git 03Benoit Fouet 07release/2.1:dc319a52f920: avformat/riffenc: Filter out "BottomUp" in ff_put_bmp_header() [20:03] ffmpeg.git 03Philip DeCamp 07release/2.1:9c61b4494113: libavutil/opt: fix av_opt_set_channel_layout() to access correct memory address [20:03] ffmpeg.git 03lvqcl 07release/2.1:8dd6075a7f28: avutil/x86/cpu: fix cpuid sub-leaf selection [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:c658f6c34d0f: avcodec/ac3enc_template: fix out of array read [20:03] ffmpeg.git 03Reimar D?ffinger 07release/2.1:d61a325a68aa: configure: add noexecstack to linker options if supported. [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:23a22b0da90d: avcodec/jpeglsdec: Check run value more completely in ls_decode_line() [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:9ec550c36483: avcodec/mjpegdec: check bits per pixel for changes similar to dimensions [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:b5298c464f05: avcodec/utils: Add case for jv to avcodec_align_dimensions2() [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:ce9d497755fd: avcodec/mmvideo: Bounds check 2nd line of HHV Intra blocks [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:ceb9d67a0ad1: avcodec/tiff: more completely check bpp/bppcount [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:5d852f338d5b: avcodec/pngdec: Check bits per pixel before setting monoblack pixel format [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:4841b2759f11: avcodec/pngdec: Calculate MPNG bytewidth more defensively [20:03] ffmpeg.git 03Michael Niedermayer 07release/2.1:96357894ff9a: avcodec/cinepak: fix integer underflow [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:05d7e92e4fdd: avcodec/gifdec: factorize interleave end handling out [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:c10c71452fff: avcodec/qpeg: fix off by 1 error in MV bounds check [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:bf2605c35610: avcodec/smc: fix off by 1 error [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:21808e218180: avcodec/vorbisdec: Fix off by 1 error in ptns_to_read [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:ea61dfe0abf4: avformat/mpegts: Check desc_len / get8() return code [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:f335275c50b8: avcodec/h264: Check mode before considering mixed mode intra prediction [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:11313263ab37: swresample/swresample: fix sample drop loop end condition [20:04] ffmpeg.git 03Christophe Gisquet 07release/2.1:3fd2ff1b4f2e: utvideoenc: properly set slice height/last line [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:a90ed80d05b7: postproc/postprocess: fix quant store for fq mode [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:95fa91d97525: postproc: fix qp count [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:917946e1c96e: avcodec/diracdec: Use 64bit in calculation of codeblock coordinates [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:824f29e4fbb7: avcodec/diracdec: Tighter checks on CODEBLOCKS_X/Y [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:dc2f4b72673a: avcodec/dirac_arith: fix integer overflow [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:8c5c45b9e1b6: avcodec/dxa: check dimensions [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:a2619a0a36db: avcodec/dnxhddec: treat pix_fmt like width/height [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:a4b6d5574f05: avcodec/utils: Align dimensions by at least their chroma sub-sampling factors. [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:ffd5ccee5df5: avcodec/g2meet: check tile dimensions to avoid integer overflow [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:3467dfed6bf3: avcodec/cook: check that the subpacket sizes fit in block_align [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:bc0a6add0aea: avcodec/svq1dec: zero terminate embedded message before printing [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:4352a971afeb: avcodec/h264_slice: Clear table pointers to avoid stale pointers [20:04] ffmpeg.git 03Carl Eugen Hoyos 07release/2.1:1987afe5a0c5: lavc/utils: Make pix_fmt desc pointer const. [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:de259f32ac12: avcodec/options_table fix min of audio channels and sample rate [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:58e673b9a345: avcodec/utvideodec: fix assumtation that slice_height >= 1 [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:d3e19509cf43: avcodec/wmaprodec: Fix integer overflow in sfb_offsets initialization [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:c4034e4e192e: avformat/hlsenc: Free context after hls_append_segment [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:51c32a535bb4: swscale/x86/rgb2rgb_template: handle the first 2 lines with C in rgb24toyv12_*() [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:b4ce4f94e781: avcodec/mjpegdec: Fix context fields becoming inconsistent [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:126ee72e47c9: avcodec/utils: Check that the data is complete in avpriv_bprint_to_extradata() [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:431d66b1d560: avcodec/flacdec: Call ff_flacdsp_init() unconditionally [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:ef8fbb4878dc: avcodec/pngdec: Check IHDR/IDAT order [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:6212758c0757: avcodec/rawdec: Check the return code of avpicture_get_size() [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:35cfb99c15dd: avcodec/hevc_ps: Check num_long_term_ref_pics_sps [20:04] ffmpeg.git 03Michael Niedermayer 07release/2.1:e4f0f854c476: avcodec/svq3: Dont memcpy AVFrame [20:17] ffmpeg.git 03Michael Niedermayer 07release/2.1:27172a5ca360: Update for 2.1.6 [20:28] ffmpeg.git 03Michael Niedermayer 07release/0.5:371403601121: avcodec/jpeglsdec: Check run value more completely in ls_decode_line() [20:28] ffmpeg.git 03Michael Niedermayer 07release/0.5:f7170c48328d: avcodec/mjpegdec: check bits per pixel for changes similar to dimensions [20:29] ffmpeg.git 03Michael Niedermayer 07release/0.5:9b4507e423e5: avcodec/mmvideo: Bounds check 2nd line of HHV Intra blocks [20:29] ffmpeg.git 03Michael Niedermayer 07release/0.5:b2f2cbdb1caf: avcodec/gifdec: factorize interleave end handling out [20:29] ffmpeg.git 03Michael Niedermayer 07release/0.5:eac21ee7ba0f: avcodec/qpeg: fix off by 1 error in MV bounds check [20:29] ffmpeg.git 03Michael Niedermayer 07release/0.5:7128f67c3f20: avcodec/smc: fix off by 1 error [20:29] ffmpeg.git 03Michael Niedermayer 07release/0.5:b8129b1a7ab9: avformat/mpegts: Check desc_len / get8() return code [20:29] ffmpeg.git 03Michael Niedermayer 07release/0.5:831416692b81: avcodec/dxa: check dimensions [20:29] ffmpeg.git 03Michael Niedermayer 07release/0.5:473b01609627: avcodec/pngdec: Check IHDR/IDAT order [20:29] ffmpeg.git 03Michael Niedermayer 07release/0.5:e74795e54108: huffyuvdec: check width more completely, avoid out of array accesses [20:29] ffmpeg.git 03Michael Niedermayer 07release/0.5:15d6b44ddcce: update for 0.5.15 [20:59] ffmpeg.git 03Benoit Fouet 07fatal: ambiguous argument 'refs/tags/n2.1.6': unknown revision or path not in the working tree. [20:59] Use '--' to separate paths from revisions [20:59] refs/tags/n2.1.6:HEAD: avcodec/pngdec: apng: fix output buffer filling when no blending is needed. [22:01] ffmpeg.git 03Michael Stypa 07master:cb58c771ade6: fix Makefile objects for pulseaudio support [22:21] ffmpeg.git 03Martin Storsj? 07master:491805636cef: rtmpproto: Fix a typo in a comment [22:21] ffmpeg.git 03Michael Niedermayer 07master:27897d2ef6d3: Merge commit '491805636cef50d3f582bd345e1460eeb739ea48' [22:57] ubitux: AV_AUDIO_SERVICE_TYPE_ lol [22:58] disposition was first! [00:00] --- Sat Nov 29 2014 From burek021 at gmail.com Sun Nov 30 02:05:02 2014 From: burek021 at gmail.com (burek) Date: Sun, 30 Nov 2014 02:05:02 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg-devel.log.20141129 Message-ID: <20141130010502.C9AD918A02D9@apolo.teamnet.rs> [00:23] ffmpeg.git 03Neil Birkbeck 07master:dd5d61795690: avfilter/vf_idet: Fixing idet for single-frame inputs. [01:08] ffmpeg.git 03Lukasz Marek 07master:0cb994dfe32e: lavu/opt: add escaping to av_opt_serialize [04:27] ffmpeg.git 03Vadim Kalinsky 07master:2db29482f1ae: avutil/bprint: C++ compatible AVBPrint definition. [05:35] http://fate.ffmpeg.org/report.cgi?time=20141128085005&slot=x86.os2.446 something broke mp2 decoding on os/2 a month or so ago [05:35] encoding, rather [10:23] pross: yeah so actually, there is this: c70a6a41ddb759a5c6e8e810ebd603e39c36a315 [12:03] ffmpeg.git 03Dave Yeo 07master:090a7801a881: libavutil/thread.h: Support OS/2 threads [12:03] ffmpeg.git 03Dave Yeo 07master:32eadfe453c3: libavutil/threads.h: correct an include to be local [13:20] ffmpeg.git 03Michael Niedermayer 07master:c299b6fd0807: avutil/buffer: Move USE_ATOMICS to thread.h to avoid it becoming out of sync with it [13:49] ffmpeg.git 03Michael Niedermayer 07master:5bf4cd8e5b77: avutil/ppc/util_altivec: add () to VEC_SPLAT16 macro [14:25] Daemon404: Hi Derek, what's wrong with pastebin (or Compn)? [14:25] Anyway, I will leave the winodw open, thank you! [15:03] ffmpeg.git 03Michael Niedermayer 07fatal: ambiguous argument 'refs/tags/n0.5.15': unknown revision or path not in the working tree. [15:03] Use '--' to separate paths from revisions [15:03] refs/tags/n0.5.15:HEAD: avutil/ppc/util_altivec: add () to VEC_SPLAT16 macro [15:43] Is there a non-convoluted way to export options for a bitstream filter ? [15:44] There's AV_OPT_FLAG_FILTERING_PARAM but no class/... way to then export said options except for libavcodec/options_table.h [15:46] passing a const char *args as a parameter to the bsf seem a dead giveaway it predates that kind of things [19:04] wow, linking ffmpeg.exe takes longer than the entire compiling, in cygwin [19:06] mingw configure script takes foreeeeeeeever [19:06] Configure runs as fast as on my linux box [19:07] But it takes like 10 minutes to link ffmpeg.exe [19:07] mingw and cygwin are diff , of course :p [19:07] static ? [19:07] yes [19:08] ffmpeg.git 03Michael Niedermayer 07master:cae851c78921: avcodec/ac3enc: Use avpriv_float_dsp_alloc() [19:08] ffmpeg.git 03Michael Niedermayer 07master:14285c3331bc: avcodec/aacenc: Use avpriv_float_dsp_alloc() [19:08] ffmpeg.git 03Michael Niedermayer 07master:b0464212bd81: avcodec/on2avc: Use avpriv_float_dsp_alloc() [19:09] linking on windows also trashes my hdd like mad [19:10] 10 min is excessive, though [19:14] whenever i install mingw and try compiling ffmpeg or mplayer, seems like i quickly pick up a bunch of errors on the disk [19:14] >windows problems [19:17] in general, running a cross-compiler on a linux VM is much faster than using a cygwin/msys environment [19:19] i rarely do it because i'm too lazy to boot up one. especially now that i started using msys2 which has git send-email working out of the box without a hassle :p [19:20] ooo [19:20] does msys2 allow building windows binary without linking to the equivalent of cygwin's DLLs ? (ie a bit like -mno-cygwin) [19:20] but yes cross compile is faster in all cases. [19:21] if msys is using mingw, then yes. because mingw uses native binary not cygwin [19:21] iirc, msys2 is based off cygwin from I had read from afar [19:22] mingw is independent from the msys/msys2 environemnt [19:22] iirc msys is just the shell, so [19:23] i dont know what msys2 is doing yet [19:23] ok but cygwin usually impacts both the environment and the toolchain, hence my question [19:23] cehoyos should force me to use msys2 , because i dont use git-send-email on windows [19:23] send patches attached to emails ehe [19:24] yeah. the same happens if you use msys gcc [19:24] depends on the number of patches [19:24] if I'm sending only one as an update to a previous post, then I do it manually [19:24] Action: Compn prefers direct commit anyway [19:24] but mingw gcc is independent regardless of what environment you use [19:24] I'm using cygwin for everything on Windows. I basicaly cross compile in it, if i want to build for normal windows. [19:24] mingw is available as cygwin package, so you use it like you would with a cross compiler on linux. [19:24] probably I could learn how to use message ID to do just that, but... [19:25] you mean "git send-email --in-reply-to= file.patch"? [19:29] yeah [19:38] Is there some example on how to use the dynarray from avutil/dynarray.h? [21:26] ffmpeg.git 03Michael Niedermayer 07master:2336c76d5a6e: avcodec/ra288: Use avpriv_float_dsp_alloc() [21:26] ffmpeg.git 03Michael Niedermayer 07master:fc9ced41e48f: avcodec/twinvq: Use avpriv_float_dsp_alloc() [21:44] https://bpaste.net/show/6755cbec2e4d Is this me, doing some weird memory corruption in my nvenc encoder, or is this an actual bug in the mkv muxer? [23:01] BtbN, no idea from this, try valgrind or asan [23:20] michaelni, already found it. Just forgot to set a variable, so i called av_malloc with a size of 0 [23:21] It's interesting that this ringbuffer actualy does squeeze out a few more fps over the old linked list [23:21] is malloc/free realy that expensive? [23:22] it can be [23:23] yeah, but i didn't expect it to actualy take some time, even if the CPU isn't busy [00:00] --- Sun Nov 30 2014 From burek021 at gmail.com Sun Nov 30 02:05:01 2014 From: burek021 at gmail.com (burek) Date: Sun, 30 Nov 2014 02:05:01 +0100 (CET) Subject: [Ffmpeg-devel-irc] ffmpeg.log.20141129 Message-ID: <20141130010501.C55F918A0259@apolo.teamnet.rs> [00:17] hm, I guess this has something to do with x264. I tried to input the http stream to a native x264.exe and it didn't work either. [04:37] How does gstreamer-ffmpeg work? It implements transcoding functionality via the ffmpeg libraries (libavcodec and libavformat, etc.) separate from the ffmpeg binary? [04:38] I have a Python software that is transcoding videos happily without even the ffmpeg binary installed on my system.... [04:48] hi, I'm trying to use ffmpeg. I have a .avi video and want to convert it to multiple resolutions and encode with H.264, is that possible? [04:53] hipsterdufus: yes [04:53] https://trac.ffmpeg.org/wiki/Encode/H.264 [04:53] https://trac.ffmpeg.org/wiki/FilteringGuide [04:53] https://trac.ffmpeg.org/wiki/Creating%20multiple%20outputs [05:10] ok, lets start smaller. How do I convert a file from .avi to mp4 with h.264 encoding? [05:16] ffmpeg -i .avi -c:v libx264 out.mp4 [05:16] That's all described very well in the first link I posted. [07:40] Hello. I've been trying to debug a SIGSEGV from ffmpeg for a few days now... [07:40] I've searched online to no avail, and while I can get it to crash with gdb attached to retrieve a backtrace, it does not crash with valgrind. [07:41] The crash occurs when I call avcodec_flush_buffers, but only about 50% of the time... [07:41] I'm typically someone who does not ask for help, but I am completely at a loss for what is causing it. I was hoping I could find some advice [07:42] oh, right [07:42] my version of ffmpeg is 2.4.3; however, it was also happening in 2.2.10, I tried both. [07:43] The exact crash happens inside of malloc.c. The final few lines of the crash are: [07:43] #0 malloc_consolidate (av=av at entry=0x7ffff1476760 ) at malloc.c:4099 [07:43] #1 0x00007ffff113bc66 in _int_free (av=0x7ffff1476760 , p=0x1f43230, have_lock=0) at malloc.c:3999 [07:43] #2 0x00007ffff6c7cc5c in av_freep (arg=arg at entry=0x7fffb9459858) at /home/me/Workspace/avidemux2/buildCore_debug/ffmpeg/source/libavutil/mem.c:239 [07:43] #3 0x00007ffff6f9fa93 in ff_h264_free_tables (h=h at entry=0x7fffb9458040, free_rbsp=free_rbsp at entry=1) at /home/me/Workspace/avidemux2/buildCore_debug/ffmpeg/source/libavcodec/h264.c:378 [07:43] #4 0x00007ffff6fa2ee1 in flush_dpb (avctx=) at /home/me/Workspace/avidemux2/buildCore_debug/ffmpeg/source/libavcodec/h264.c:1119 [07:43] #5 0x00007ffff71c017f in avcodec_flush_buffers (avctx=0x199a400) at /home/me/Workspace/avidemux2/buildCore_debug/ffmpeg/source/libavcodec/utils.c:3075 [07:43] #6 0x00007ffff7bd47d9 in decoderFF::flush (this=0x1a33c50) at /home/me/Workspace/avidemux2/avidemux_core/ADM_coreVideoCodec/src/ADM_ffmp43.cpp:349 [07:47] What is peculiar to me, is the segmentation fault is occuring when trying to free a pointer &h->non_zero_count [07:51] KoolAidPitcher: try again with git master and if it still happens open a bug report. [07:52] I have downloaded a PBS video which I would like to burn into DVD, but my DVD burner says Not-Supported-Format on most videos... and AVI files it says that the audio is not supported. Is there any way ffmpeg can convert my videos into formats that could easily be burned into DVDs ? [07:52] hmm... okay [07:52] so you have never seen this before? [07:52] fahadash: does your dvd player support mpeg4 video? [07:53] no [07:53] My DVD player supports old .VOB files [07:53] I guess thats mpeg2 , i dont know [07:54] correct, there's "-target ntsc-dvd" (and pal-dvd) [07:55] my dvd player supports ntsc [07:57] Hi relaxed [07:57] before I submit a bug report, I was curious if I could get any advice on debugging ffmpeg [07:58] our version of ffmpeg is forked-- we apply patches to it prior to using it. [07:58] KoolAidPitcher: https://www.ffmpeg.org/bugreports.html [08:02] wow [08:03] => 0x00007ffff113b0e3 : mov 0x10(%rbx),%rax [08:03] Last message repeated 1 time(s). [08:03] that's quite a moverax 0x34706d2e74736574 3778640133653489012 [08:03] rbx 0xcb8f92d18d6e233c -3778640133621931204 [08:03] rax 0x34706d2e74736574 3778640133653489012 [08:04] I'm going to look into this a little further. This might be our patches. Thanks! [11:55] I am using ffmpeg -i CAM00440.mp4; it shows "Stream #0:0(eng): Video: mpeg4 (Main Profile)". Is this mpeg4 referring to mpeg4 Part 2 [11:57] sant527: yes [11:59] relaxed: Ok. [12:00] relaxed: What are the various profiles of mpeg4 [13:32] Hi! I'm capturing and on-the-fly transcoding a dvb-s2 stream via sat-ip. Everything works fine, this is the ffprobe output https://paste.ee/p/DKNYu As you can see, the mpegts frame rate is a 50fps progressive h264 stream. 50 FPS is due to PAL standards, so the TV station just doubles every frame from 25 to 50fps. That means every other frame is redundant. Can I drop every other frame? [13:36] -x264opts video-filter=select_every:2,0,1 throws this error (complete output): https://paste.ee/p/dl44r [13:43] Is that an actual x264 option? Or is it for a different part of ffmpeg? [13:45] probably an x264cli option he is trying to use via x264opts :P [13:45] mopp, go look at the video filter documentation on ffmpeg.org [13:45] there are various filters with documentation there [13:45] you'll probably find one that matches your use case :P [14:12] "select_every" seems to be an actual x264 option. I'm using x264opts because select_every is not mapped to x264 directly. Well, I'll keep looking. Is it possible at all to drop frames from a h264 stream in order to reencode? I'm thinking of I-frames etc. [14:13] select_every is a x264CLI option of the video filter feature it has [14:13] it is not a LIBx264 feature [14:14] as I said, look at this page http://ffmpeg.org/ffmpeg-all.html [14:14] and see the documentation for the video filters [14:19] hello [14:20] anyone could tell me syntax for set icecast server as output ? [14:24] JEEB: Huh. Using x264opts was overkill. A simple -vf fps=fps=25 was enough. thx! [14:28] can I convert mp4 to NTSC DVD ? [14:34] ffmpeg -i blah.mp4 -target ntsc-dvd output.mpeg then author it and burn. [14:34] hi. I use ffmpeg to play video on -f xv :0. how do I force playback to be at the same framerate as the source video? currently ffmpeg forces fps to be that of display refresh rate. [14:35] because of this, the playback is sped up more than 2 times [14:43] hello can i use -re with -ss? [14:48] thanks sacarasc [15:17] Which CPU architecture build is recommended 32bit or 64bit ? [15:20] better match OS arch [15:33] hi all what is the difference between smil and m3u8 [15:36] Hello [15:37] I ran the command sacarasc told me to, it output-ed the audio-only file [15:38] I can't open that with windows media player [15:39] Nevermind, Windows-Media-Player sucks [16:05] Is there any documentation for libavformat aside from the doxygen docs? [16:11] how to set output of encoded mp3 from line-in to icecast? [16:12] In regards to transcoding a h264 mpegts with libx264 and -vf fps and cropping: would you advise to convert to rawvideo first or use the h264 stream directly? [16:39] can i use ffmpeg as virtualwebcam, so windows recognize its output as some webcamera and various programs capture that stream as webcamera(like skype, yahoo messenger etc) [16:47] facebook video [17:59] Alina-malina: Check something like manycam or VLC2VCam, it should support streaming sources, so you'll be able to stream ffmpeg to it [18:00] eh i need some pythonish stuff, that manycam doesnt designed as i expect [18:41] thats what i try to do http://pastebin.com/8aT9uKcU [18:41] why do i have this error [18:42] i'me trying to stream so that my iphone could watch it via safari browser [18:42] stream my webcam [19:00] Hello! Any idea why I cant transcode from stdin? http://paste.kde.org/pcqefkmzm [19:08] broken file maybe, or it's because mp4 is bad at streaming [19:08] moov atom at the end or something [19:11] klaxa: it works perfectly when not taking it from stdin. [19:12] probably the moov atom thing then [19:12] You mean that it cant seek so being that at the end it wont work? [19:13] yes, you can move it to the front though [19:13] see: http://stackoverflow.com/questions/8061798/post-processing-in-ffmpeg-to-move-moov-atom-in-mp4-files-qt-faststart [19:16] This wont probably solve my original problem. I wanted to have a way to show a progress bar when transcoding with ffmpeg. This is not a good option. The only way is to probably modify ffmpeg itself. [19:21] luc4: just take frame=$num; $num/total_frames [19:25] c_14: sorry? [19:26] what should i do with:This feed is already being received. [19:26] ? [19:26] When ffmpeg is encoding it outputs a little informational message with frame= fps= etc. Just take the current output of frame= and divide by the total number of frames [19:26] c_14: me? [19:27] nah, luc4 [19:27] droid909: kill ffserver and restart it [19:28] c_14: yes, I see that. But is there the toal amount somewhere? Also, that means modifying the ffmpeg sources. Is there a particular reason why it was never done? [19:28] c_14: tried, the same [19:29] c_14: can i post my config file, will you take a look? [19:30] c_14: config http://pastebin.com/jiSgrLZM [19:31] c_14: and i run it like ffmpeg -r 25 -f video4linux2 -i /dev/video0 http://localhost:8090/feed1.ffm [19:32] luc4: because ffmpeg doesn't know how many frames there will be until it's done [19:33] c_14: then I cant do the division... [19:33] luc4: you can find out manually with ffprobe -count_frames -show_streams iirc [19:33] though if it's cbr you can just do fps * total_time [19:34] droid909: I can't see anything wrong with that. What FFmpeg version are you running? [19:34] c_14: ah ok, so is there a particular reason why ffmpeg does not do that? I may add another param to add a progress bar& [19:34] c_14: ffmpeg version 2.4.3 Copyright (c) 2000-2014 the FFmpeg developers [19:36] luc4: because ffmpeg doesn't know if the file is constant framerate or variable framerate, and it doesn't always have the total length of the video [19:36] droid909: what's the complete console output of both ffmpeg and ffserver [19:37] c_14: mmh& concept is& if ffmpeg cannot do it automatically, I cant do that either& [19:38] luc4: if you have the whole file, you can use ffprobe to count each individual frame, grab the output from that and use that for the progress [19:38] c_14: ffmpeg http://pastebin.com/ciiRwHGQ and ffserver http://pastebin.com/rTAyNjCq [19:39] c_14: which means I can patch ffmpeg to count the frames if this is a file and create the progress bar. Correct? [19:39] luc4: yes [19:39] c_14: thanks! [19:40] c_14: maybe i should try another format? [19:41] c_14: i need any that will work in iphone's browser [19:42] droid909: what's the problem? It looks like it works. [19:44] c_14: when i access it in my google chrome it writes: This feed is already being received. [19:44] c_14: and starts downloading the feed [19:47] You're accessing http://ip/test.swf ? [19:49] c_14: ohh... [19:49] c_14: i tried feed1.ffm [19:49] c_14: it works now .. [19:50] still i need to choose something that work in safari, and swf isn't for safari, i was trying just to make it work [19:50] c_14: what format should i chose? [19:55] mjpeg? [19:55] that might ork [19:57] *work [20:07] c_14: that is my input i guess : Stream #0:0: Video: rawvideo (YUY2 / 0x32595559), yuyv422, 320x240, 36864 kb/s, 30 fps, 30 tbr, 1000k tbn, 1000k tbc [20:56] hey [20:56] this command: ffmpeg -i in.mp3 -f s16le -ar 48000 -ac 1 out.wav [20:57] should produce a valid wave file with a RIFF header, right? [20:57] or am I missing something? [21:14] In general, it seems like container formats have audio and video packets interleaved. Are there any common formats where they aren't, but rather all video at the beginning followed by audio, or something similar? [21:34] MadTBone_, I haven't heard about any: that greatly increses costs of video player hardware :) [21:34] MadTBone_, perhaps RTP which transmits audio and video on separate bands [21:38] Mavrik, that's what I figured. It would introduce lots of seeking/random i/o into what should be a mostly linear data read /write [21:41] mhm, also most formats are defined for hardware decoders where memory was premium [21:41] hence most formats have hard limits on how far audio, video and other references can be between them [22:28] I'd like to do -vf cropdetect and -vf crop in one step, in other words I'd like to parse the cropdetect output into -vf crop, in one line preferable. Is that possible from CLI without a bash script? I've only found this so far: http://pastebin.com/n8Jk3DjS but I'm on a windows box and I have very little scripting experience. [22:29] *preferably [22:50] Hi, having a problem here. How do I separate the chroma and luma of a video in ffmpeg? I tried '-vf pp=noluma' but it spat out an error. [22:50] 1 errors in postprocess string "noluma" / [AVFilterGraph @ 0x2ed7e00] Error initializing filter 'pp' with args 'noluma' / Error opening filters! [22:50] ffmpeg 2.3.3 [23:21] okay I got around it, but it looks like the postproc (pp) filter is still not working properly. [23:21] What worked was '-vf lutyuv=y=128' [23:44] how does the mergeplanes filter work? [23:45] I'm trying to merge a video with only chrominance to a video with only luminance [23:55] now I'm getting more problems [23:55] ffmpeg -i 03luminance.mkv -i 03chrominance.mkv -filter_complex [0:0][1:0]mergeplanes=0x00010210:yuv422p out.mkv [23:55] [Parsed_mergeplanes_0 @ 0x3973de0] output plane 0 width 640 does not match input 0 plane 1 width 320 [23:55] but all the inputs are 640 pixels wide. [23:57] nvm might have gotten it [00:00] --- Sun Nov 30 2014