Ffmpeg-devel-irc

Download

ffmpeg-devel-irc@ffmpeg.org

February 2013

1 participants
56 discussions

ffmpeg-devel.log.20130217
by burek 18 Feb '13

18 Feb '13

[00:20] <nevcairiel> michaelni: avpriv_color_frame asserts when using a hwaccel in h264, pixel format is not planar [00:25] <nevcairiel> michaelni: i sent a small patch about that [00:35] <cone-323> ffmpeg.git 03James Almer 07master:a56fd9edab57: lavu: Fix checkheaders for x86/emms.h [00:35] <cone-323> ffmpeg.git 03Hendrik Leppkes 07master:1d6f6ff4d9fa: h264: don't initialize missing pictures when using a hwaccel [00:44] <cone-323> ffmpeg.git 03James Almer 07master:2e988fd68964: lavc/tta: Use a safer check for encrypted streams [00:44] <cone-323> ffmpeg.git 03James Almer 07master:b8bb661dab8e: Changelog: Add entry for encrypted TTA stream decoding [01:55] <Compn> j-b : how goes the bittorrent demuxer? :P [01:55] <Compn> j-b : bonus points if you have it implemented into ffmpeg. that way more players can utilize ffmpeg code :) [01:56] <j-b> Compn: you _really_ think the issue is the demuxer or protocol paet? [01:56] <j-b> part? [01:56] <j-b> Of course the issue is the UI, the management and the downloading services [01:56] <Compn> hmm? i mean the overall project [01:57] <Compn> i dont know which part is the difficult . probably cache, demuxer quirks (mov/avi index at end of file) [01:57] <Compn> implementing an entire bt client ... [01:58] <Compn> j-b : but is my idea of putting it in ffmpeg good or bad ? [01:59] <Compn> or its already started in vlc ? [01:59] Action: Compn is slow with the ideas today [02:00] <Compn> j-b : its not important, just another dumb idea from me :) [02:02] <j-b> Compn: so far, we've done nothing, and we're sponsoring nothing [02:02] <j-b> so, I got no idea [02:04] <Compn> can vlc use ffmpeg's network stuff ? [02:04] <j-b> sure. [02:04] <Compn> ok [02:04] <Compn> so at least it would be possible to implement it in ffmpeg [02:04] <Compn> and vlc could use it from there [02:05] <j-b> sure [02:05] <j-b> but, once again, the hard part is NOT the networking part [02:06] <Compn> it sounds kinda similar to apple / adobe fragmented thing [02:06] <Compn> unless i'm completely missing something [02:06] <Compn> thing = http [02:09] Action: Compn keeps focusing on networking part for some reason [02:24] <iive> j-b: what is the problem with the UI? [02:24] <saste> iive, interface are usually a mess to do (correctly) [02:24] <iive> but why does it need ui at all? [02:25] <j-b> iive: well, you have media player UI, and you need to rape it, in order to have a torrent-management-system in it [02:25] <Compn> iive : limit bandwidth, specify download dir [02:25] <j-b> of course not [02:25] <Compn> j-b : cant just put it all in settings ? [02:25] <j-b> all those are trivial [02:25] <Compn> whats the hard part ? [02:25] <Compn> i mean [02:25] <Compn> what ui do you need ? [02:25] <j-b> look at Utorrent [02:25] <Compn> like utorrent status or something [02:25] <Compn> ah [02:26] <j-b> You need a playlist-like-thing to list your torrent [02:26] <j-b> because if you play just one, it will buffer for a long time [02:26] <iive> see it is network problem :P [02:27] <Compn> lolol [02:27] <j-b> moreover, the UI needs to talk to this module in a very specific way (it ain't your average demuxer) [02:28] <iive> instead of utorrent, take a look of the original bittorrent client, or even transmision... not much to click there. [02:28] <Compn> just text output? lol [02:28] <Compn> use ffmpegs ascii decoder :D [02:29] <Compn> wont be pretty [02:29] <j-b> iive: even for that, you need to control your demuxer in different way than RTSP/RTMP [02:29] <iive> the problem is elsewhere, e.g. do you want disk caching... because then you create permanent copy on your system. [02:30] <Compn> you need disk caching , for indexes at end of files :P [02:30] <j-b> So, you need a special API, and even with one-torrent-at-a-time, you need status that is a bit more than "buffering" [02:30] <iive> the biggest problem of course it bandwidth, you need to both download blocks in realtime playback and download blocks in random order to keep the seeding balanced. [02:31] <iive> ...in order for realtime playback... [02:31] <Compn> i dont think the implementation needs seeding balance [02:32] <Compn> its the same thing anyways, in order or random [02:32] <Compn> still the same amount of data [02:33] <iive> no it is not. torrent always uses random. [02:33] <j-b> you can [02:33] <Compn> torrent can be configured to set priority to rares pieces, starts/ends of files, or download in order [02:34] <iive> it is same amount of data, but if everybody have the first 10% of the movie and only 1 seed have the rest, you will get a lot worse speed when trying to get something from the 90% [02:34] <Compn> its the same speed! [02:34] <iive> no. for the 10% you have 10 seeders, so you get 10x speed. [02:35] <Compn> until everyone catches up, then its the same speed [02:35] <iive> and if you use it in video content, you are most likely to start from the start :P [02:35] <j-b> in-order is possible [02:35] <j-b> but it is bad [02:36] <iive> yes, it is bad for the swarm. [02:36] <Compn> doesnt miro already do this btw ? [02:36] <Compn> and isnt it based on vlc... [02:37] <Compn> http://www.getmiro.com/download/features/ [02:38] <Compn> or is it just a mix of the two. i dunno never used it [02:54] <highgod> Hi, all I want to ask a question. Can ffmpeg be compiled using mingw64 and create 64bits libs and dlls? thanks [03:09] <Compn> highgod : yes, it can [03:10] <Compn> oh, maybe not [03:10] <Compn> we dont have a fate box for it [03:11] <Compn> er [03:11] <Compn> yeah [03:11] <Compn> highgod : http://ffmpeg.zeranoe.com/builds/ [03:11] <Compn> zeranoe has 64bit mingw ffmpeg builds [03:11] <Compn> and instructions for doing it ^^ [03:43] <cone-323> ffmpeg.git 03Reinhard Tartler 07release/0.5:deb650c69231: Release notes and changelog for 0.5.10 [03:43] <cone-323> ffmpeg.git 03Michael Niedermayer 07release/0.5:0360dbefad8a: Merge remote-tracking branch 'qatar/release/0.5' into release/0.5 [04:24] <highgod> OK,thanks Compn. [04:29] <Zeranoe> Could anyone explain to me the license details of fdk-aac? A lot of users are asking for fdk-aac, and reading the license for this lib I find "Redistribution and use in source and binary forms, with or without modification, are permitted without payment of copyright license fees provided that you satisfy the following conditions:" Which then lists a few conditions. Is there any way to load the library dynamically and still compl [04:30] <cone-323> ffmpeg.git 03Michael Niedermayer 07release/0.5:a23a3dba2544: qdm2: check array index before use, fix out of array accesses [04:30] <cone-323> ffmpeg.git 03Michael Niedermayer 07release/0.5:fee26d352a52: roqvideodec: check dimensions validity [04:30] <cone-323> ffmpeg.git 03Michael Niedermayer 07release/0.5:13093f9767b9: vqavideo: check chunk sizes before reading chunks [04:30] <cone-323> ffmpeg.git 03Michael Niedermayer 07release/0.5:d34cfb33afb8: update for 0.5.11 [04:48] <Compn> Zeranoe : got a link? [04:48] <Zeranoe> Compn: A link for the license? [04:48] <Compn> ya [04:49] <Zeranoe> Compn: https://github.com/mstorsjo/fdk-aac/blob/master/NOTICE#L29 [04:49] <Compn> Zeranoe : thats wbs's repo, maybe he can explains it [04:49] <Zeranoe> The license looks very similar to the GPL and I'm surprised FFmpeg wont allow it in a GPL build. [04:50] <Zeranoe> Compn: Explain why FFmpeg wont allow it under the GPL? [04:50] <Compn> its very similar to gpl build [04:50] <Compn> possibly You may not charge copyright license fees for anyone to use, copy or distribute the FDK AAC Codec [04:50] <Compn> software or your modifications thereto. [04:50] <Compn> goes against gpl [04:50] <Compn> since you can charge for gpl [04:51] <Compn> not sure what a 'copyright license fee' would even be ... [04:51] <Zeranoe> Compn: If the lib cannot be included in a GPL build, would it comply with the license if FFmpeg was built to accept a fdk-aac lib file, like a .dll and use it externally when it needs it? [04:53] <Zeranoe> Compn: Since both libs allow for distribution of a binary, only thing would be the license for the fdk-aac code that FFmpeg uses to call the lib. [04:54] <Compn> you dont just want to put it in 'non-free' ? [04:54] <Compn> but i dunno, i'm no license master [04:54] <Compn> your idea sounds good [04:54] <Zeranoe> libfdk-aacenc.c is under the LGPL so it should be ok [04:55] <Zeranoe> Compn: Non-free means I cant redistribute the binary, which is my goal. [04:55] <Zeranoe> I want to provide my uses with a FFmpeg that can use the best AAC encoder lib. [04:56] <Zeranoe> users** [11:53] <cone-485> ffmpeg.git 03Stefano Sabatini 07master:ef4c71e8f83a: lavfi/unsharp: add check on matrix x/y size values oddity [11:53] <cone-485> ffmpeg.git 03Stefano Sabatini 07master:d2cadea3f085: lavfi/unsharp: directly access in-loop variables in apply_unsharp() [11:53] <cone-485> ffmpeg.git 03Stefano Sabatini 07master:64e592eef26b: lavfi/unsharp: merge definition and declaration in init_filter_param() [11:53] <cone-485> ffmpeg.git 03Stefano Sabatini 07master:89505f2c3f8e: lavfi/unsharp: add missing NULL check [11:53] <cone-485> ffmpeg.git 03Stefano Sabatini 07master:8c85a9f046fb: lavfi/mp: drop mp=unsharp filter [11:53] <cone-485> ffmpeg.git 03Stefano Sabatini 07master:7ca2f8b1130b: lavfi/mp: drop mp=kerndeint filter [11:56] <ubitux> can anyone tell me if the avcodec.h change in the subtitle character encoding patch is fine? [11:56] <ubitux> Nicolas seems to have some concern about it [11:56] <ubitux> http://ffmpeg.org/pipermail/ffmpeg-devel/2013-February/139117.html [11:56] <ubitux> "Wait for [11:56] <ubitux> confirmation about the avcodec.h part though. [11:56] <ubitux> " [11:57] <ubitux> (basically +#define AV_CODEC_PROP_BITMAP_SUB (1 << 16)) [12:02] <ubitux> oh maybe he was refering to the rest though [12:05] <ubitux> well i'll resend a patch [12:48] <ubitux> wm4: subtitles character encoding, soon. [12:48] <ubitux> (finally) [12:48] <wm4> ok [12:48] <wm4> what are you going to use? [12:48] <ubitux> it's done, i'm just waiting for a last comment [12:49] <ubitux> no utf-16 yet though, Nicolas seems to be interested to working on it [12:50] <ubitux> https://github.com/ubitux/FFmpeg/compare/master...sub-recode-nofilter [12:56] <ubitux> should i increment micro or minor when adding a field to AVCodecContext? [12:56] <nevcairiel> wasnt minor for such things [12:57] <ubitux> ok [13:01] <ubitux> saste: [13:01] <ubitux> -unsharp 7d72d2ab7b7f60159c822a097e01068b [13:01] <ubitux> +unsharp d41d8cd98f00b204e9800998ecf8427e [13:01] <ubitux> :( [13:04] <cone-485> ffmpeg.git 03Clément BSsch 07master:8732271e4025: lavc: mark bitmap based subtitles codecs as such. [13:04] <cone-485> ffmpeg.git 03Clément BSsch 07master:f796399344a1: lavc: support subtitles character encoding conversion. [13:04] <cone-485> ffmpeg.git 03Clément BSsch 07master:3a0a959dcbf8: lavfi/subtitles: split options between ass and subtitles. [13:04] <cone-485> ffmpeg.git 03Clément BSsch 07master:08d149d6889b: lavfi/subtitles: support charenc option. [13:04] <cone-485> ffmpeg.git 03Clément BSsch 07master:2a14b23484be: lavfi: rename vf_ass.c to vf_subtitles.c [13:04] <cone-485> ffmpeg.git 03Clément BSsch 07master:90fb3e9bee60: lavfi/subtitles: set a different description between ass and subtitles. [13:04] <ubitux> wm4: have fun ^ [13:07] <wm4> hm, vf_subtitles doesn't actually use this new flag? [13:08] <ubitux> yes it does [13:08] <ubitux> -vf subtitles=foo.srt:charenc=cp1251 [13:09] <cone-485> ffmpeg.git 03Clément BSsch 07master:b1e6b144ed92: lavc/utils: reindent in avcodec_decode_subtitle2() after f7963993. [13:09] <wm4> I mean AV_CODEC_PROP_BITMAP_SUB [13:10] <ubitux> mmh good point, maybe it should now [13:10] <wm4> and maybe codecs which decode to ass should be marked as well (in fact, AV_CODEC_PROP_BITMAP_SUB is entirely useless for now, because of backwards compatibility) [13:11] <ubitux> it's mainly meant to be used internally [13:11] <ubitux> notably to set sub->format (0 or 1) [13:11] <ubitux> or enable/disable charset re-encoding [13:14] <wm4> I still don't quite see how vf_subtitles prevents opening an image subtitle decoder as ass decoder [13:14] <ubitux> it doesn't [13:15] <ubitux> it will basically do nothing [13:15] <ubitux> first bitmap sub expected break the loop, do nothing [13:15] <ubitux> i'll add an error soon now that we have the flag [13:15] <ubitux> for backward compat, you can rely on sub->format [13:15] <ubitux> for the first decoded sub [13:23] <wm4> so I have to try to decode the first subtitle just to detect a property statically known before any decoding? [13:23] <ubitux> or use AV_CODEC_PROP_BITMAP_SUB now [13:24] <ubitux> i can't do much more for the old versions [13:24] <ubitux> any subtitles format not with the AV_CODEC_PROP_BITMAP_SUB prop is supposed to decode to ass [13:30] <cone-485> ffmpeg.git 03Stefano Sabatini 07master:894ed8fbb71d: tests: fix values for the unsharp test [13:39] <cone-485> ffmpeg.git 03Clément BSsch 07master:fe150a48a72e: lavc: fix codec_descriptor and pkt_timebase doxy. [13:39] <cone-485> ffmpeg.git 03Clément BSsch 07master:f3147917bf00: lavf/subtitles: error out in case of bitmap subtitles. [13:40] <wm4> ubitux: any news on fixing the ass packet format? [13:40] <ubitux> i'm going back to it slowly [13:40] <ubitux> i wanted to get done with the charset thing first [14:33] <cone-485> ffmpeg.git 03Anton Khirnov 07master:1f8f43a5b5e9: error_resilience: add required headers. [14:33] <cone-485> ffmpeg.git 03Michael Niedermayer 07master:59e46ef63ab4: Merge remote-tracking branch 'qatar/master' [16:03] <cone-485> ffmpeg.git 03Nicolas George 07master:dcc73aaaa995: doc/examples: do not allocate AVFrame directly. [16:58] <cone-485> ffmpeg.git 03Clément BSsch 07master:1e860f16688b: doc/codecs: simple sub_charenc option documentation. [17:35] <cone-485> ffmpeg.git 03Carl Eugen Hoyos 07master:cf36180143b3: Only set accelerated arm fft functions if fft is enabled. [17:37] <cone-485> ffmpeg.git 03Michael Niedermayer 07master:09ece9fa6c45: eval: print() support [17:37] <cone-485> ffmpeg.git 03Michael Niedermayer 07master:29c8619a4977: fate: add print() to the tests of eval [17:41] <durandal_1707> anyone working on lua filter? [18:06] <cehoyos> nevcairiel: If you look at the mkvalidator output in ticket 2263, is it sufficient for you to see why libavformat does not know where to find keyframes? [18:06] <cehoyos> And if you know why: How can such files be produced: I have the same version of mkvmerge installed here and if I remux the file, FFmpeg reads the resulting file fine (seeking does not take ages) [18:07] <nevcairiel> most software will probably only read the first seekhead, and without cues seeking in mkv is butt-slow [18:08] <nevcairiel> especially because the second seekhead as at the end of the file, it appears [18:09] <cehoyos> How can I check if the seekhead is really at the end of the file? [18:09] <cehoyos> And since our sample is 11G: How can I produce such a sample? [18:09] <cehoyos> ... such a shorter sample;-) [18:10] <nevcairiel> well mkvalidator says the second seekhead is at 11568150341 [18:10] <nevcairiel> that sounds like 11g =p [18:10] <cehoyos> ok, sorry.. [18:11] <nevcairiel> no idea how to produce something like this, maybe it didnt have the ability to seek in the output and then decided to write it at the end is better then not at all [18:11] <cehoyos> Something like mkvmerge -o - ? [18:12] <nevcairiel> never tried anything like this [18:15] <cehoyos> mkvmerge -o - produces a file ./- [18:17] <cehoyos> nevcairiel: Does this link describe why mkvmerge put a second seek head at the end of the file? https://trac.bunkus.org/ticket/290 [18:19] <nevcairiel> but in this case it references the second seek head from the first [18:19] <nevcairiel> so you know where it is and can parse it directly [18:20] <nevcairiel> so i can only guess that it simply failed to reference the secondary one in the primary one [18:22] <cehoyos> So basically I have to keep the file;-( [18:26] <cehoyos> Is my comment correct? https://ffmpeg.org/trac/ffmpeg/ticket/2263#comment:8 [18:27] <nevcairiel> from what i know without having seen the file, yes [18:28] <cehoyos> Thank you! [18:41] <ubitux> the print thing in eval will be very useful [18:42] <ubitux> thx :) [19:47] <cone-485> ffmpeg.git 03Carl Eugen Hoyos 07master:fdec49cbe83f: Add gray16 as a supported v4l2 input format. [19:47] <cone-485> ffmpeg.git 03Carl Eugen Hoyos 07master:fdbe7628a957: Add yvu410 as a supported v4l2 input format. [20:21] <cone-485> ffmpeg.git 03Carl Eugen Hoyos 07master:8d0757e1079f: Revert "swfenc: use av_get_audio_frame_duration() instead of AVCodecContext.frame_size" [20:23] <Daemon404> http://git.videolan.org/?p=ffmpeg.git;a=commitdiff;h=8d0757e1079f588ed69f62… <-- protip: bug ids are not a replacement for an actual descriptio [20:23] <Daemon404> n [20:24] <Daemon404> http://ffmpeg.org/developer.html#Development-Policy [20:24] <Daemon404> #3 [20:25] <Daemon404> in this case, the bug ID still doesnt have a dexription of why the patch reverts it [20:27] <durandal_1707> cehoyos: that bug do not have reproduced and analyzed set [20:27] <Compn> Daemon404 : its reverted because it fixes the bug. i think your real question is "why is that previous change causing this swf bug" [20:27] <durandal_1707> Compn: please, stop [20:28] <Daemon404> Compn, let us name all our commits "fix bug" [20:28] <Compn> hehe [20:28] <Compn> it should be fix regression :P [20:34] <durandal_1707> swf can have mp3 which is not covered by av_get_audio_frame_duration [20:35] <cehoyos> durandal_1707: I cannot reproduce it because I don't know how to feed FlashPlayer [20:35] <durandal_1707> and how you found out reverting fixes it? [20:37] <cehoyos> This is difficult to answer because I am not sure what the deeper meaning of your question is. [20:38] <cehoyos> (But the answer would be: Because I used git bisect for some time and afaict, the original test was not just unreviewed but untested.) [20:38] <cehoyos> s/test/patch [20:38] <durandal_1707> wasn't it reproduced with ffmpeg too? eg. durationts from demuxer are set to 0? [20:39] <cehoyos> Daemon404: My patch description contains an exact description why the patch was needed [20:39] <Daemon404> where [20:39] <Daemon404> its not in the commit message. [20:39] <Daemon404> how did you test this patch [20:39] <Daemon404> why does it fix it [20:39] <Daemon404> etc [20:40] <durandal_1707> also coverage for muxing lot of codecs in swf or any other container is small [20:42] <durandal_1707> cehoyos: but bug report also have also near to 0 useful info [20:42] <durandal_1707> so user might not use mp3 in swf at all [20:43] <durandal_1707> bug report is just incomplete, and fix/revert is (even if correct) pure guesswork [20:46] <cehoyos> durandal_1707: No, the duration is ok with ffmpeg -i file.swf [20:46] <durandal_1707> lol the ticket have enough info [20:47] <durandal_1707> it is just hidden in stupid log file [20:47] <durandal_1707> cehoyos: that can be from parser ... [20:48] <cehoyos> Daemon404: I have a suggetion; Could you find out which sample / bug report / hypothetical case was fixed by 9fb7e14? [20:49] <Daemon404> if it switches to a new config and fails, instead of just giving up, it will revert to the old one [20:49] <Daemon404> its pretty obvious [20:49] <Daemon404> also piitning out random Libav commits does not magically make it OK. [20:49] <kierank> oh nice, someone just *had* to try putting latm in mkv [20:49] <cehoyos> durandal_1707: I completely agree that posting logs is significantly more useful than attaching, but I tried to help the user because it was his first report afair. [20:50] <Daemon404> the amount of english comprehension in this channel is painful. [20:51] <cehoyos> Daemon404: (Do you honestly think it is a random commit? Because if yes, this whole discussion makes even less sense.) You misunderstand: I don't need an explanation what the patch does, I probably don't understand it anyway, and if I would be interested, looking at it would probably be sufficient, the point is *why* [20:52] <Daemon404> you need an explanaion of why it was reverted. "fixes bug X" with ZERO technical details is not valid [20:52] <durandal_1707> cehoyos: you should reopen bug and wait for poster to reply that issue have really been fixed (you cant or did not tried to reproduce it after all) [20:52] <cehoyos> I disagree, but that is probably not the point... [20:52] <cehoyos> durandal_1707: I disagree because OP's often never react and as you know we have too many open reports. [20:53] <cehoyos> (In this specific case, no testing is necessary, but that is beside the point.) [20:53] <cehoyos> Daemon404: Wow, I just wanted to leave because of you being quite offensive;-) [21:04] <cone-485> ffmpeg.git 03Carl Eugen Hoyos 07release/0.11:381e3e7e44f9: Revert "swfenc: use av_get_audio_frame_duration() instead of AVCodecContext.frame_size" [21:04] <cone-485> ffmpeg.git 03Carl Eugen Hoyos 07release/1.0:98d06b046dfe: Revert "swfenc: use av_get_audio_frame_duration() instead of AVCodecContext.frame_size" [21:04] <cone-485> ffmpeg.git 03Carl Eugen Hoyos 07release/1.1:6407800521d2: Revert "swfenc: use av_get_audio_frame_duration() instead of AVCodecContext.frame_size" [21:11] <Compn> kierank : no not more latm stuff :V [21:12] <kierank> :) [21:13] <durandal_1707> there is some other latm open bug? [21:42] <cone-485> ffmpeg.git 03Carl Eugen Hoyos 07fatal: ambiguous argument 'refs/tags/n0.5.11': unknown revision or path not in the working tree. [21:42] <cone-485> Use '--' to separate paths from revisions [21:42] <cone-485> refs/tags/n0.5.11:HEAD: Revert "swfenc: use av_get_audio_frame_duration() instead of AVCodecContext.frame_size" [22:14] <durandal_1707> michaelni: i sent patch for lavfi/noise to stop using rand/srand - i looked at output and quality of noise looks almost same [22:18] <peper03> Is a patch to add passing PRIVATE_STREAM_2 MPEG2 packets to the user likely to be accepted? I would only add a new codec ID and pass the packet as-is. There doesn't seem to be much point with a decoder as it's a pure data packet and from what I can see, used for different purposes (DVD, Dreamcast - Sofdec). [22:19] <peper03> I couldn't see any other data-only codec IDs so I can't base my patch on other implementations. [22:21] <durandal_1707> peper03: there is patch on ml? [22:23] <peper03> durandal_1707: Not yet. I've basically got everything prepared (I think) but I just wanted to check a couple of basic points first as I'm not subscribed to the ml yet etc. etc. [22:24] <peper03> Did I understand the guidelines correctly, indentation due to an 'if' should be submitted separately? [22:24] <durandal_1707> i'm not mpegts expert so can't really help [22:24] <durandal_1707> peper03: yes, easier to review [22:25] <peper03> durandal_1707: Sure. No problem. Again, just wanted to ensure I hadn't misunderstood. [22:26] <peper03> durandal_1707: Are you aware of any other data-only packets that are currently handled? [22:26] <peper03> Handled = passed back to the caller? [22:28] <cone-485> ffmpeg.git 03Michael Niedermayer 07master:dece584a639c: h264: avoid calling get_format() multiple times [22:29] <durandal_1707> peper03: i'm not qualified enough to answer such question [22:30] <peper03> durandal_1707: Ok, np. Thanks. [22:31] <durandal_1707> michaelni: so what I should use instead of RAND_MAX? [22:32] <michaelni> UINT_MAX IIRC [22:34] <durandal_1707> michaelni: is it relevant that rand returns int and lfg unsigned? [22:37] <cone-485> ffmpeg.git 03Carl Eugen Hoyos 07master:259603b91767: h264: don't initialize missing pictures when using VDPAU. [22:38] <michaelni> it could make a difference depending on how its used [22:42] <durandal_1707> i replaced RAND_MAX with UINT_MAX and nothing chaged, still noise on screen [22:43] <durandal_1707> rand(3) here returns positive numbers [22:44] <durandal_1707> and RAND_MAX is INT_MAX [22:48] <cone-485> ffmpeg.git 03Paul B Mahol 07master:f4e29c408cea: lavfi/noise: switch to AVLFG noise generator [22:48] Action: llogan wonders if saste will administer GSoC this year. [22:57] <durandal_1707> so g2m is comming [23:01] <Compn> someone is working on it durandal_1707 ? [23:02] <durandal_1707> Compn: codecs.multimedia.cx [23:03] <Compn> hah, someone must have trolled kostya :) [23:03] <Compn> wonder where he found the smallest go2 binary codec [23:03] <Compn> i spent a while in archive.org looking for it [23:09] <Compn> oh yes [23:09] <Compn> http://codecs.multimedia.cx/?p=431 [23:09] <Compn> that was the one i found :) [23:10] Action: Compn was the troll :D [23:45] <cone-485> ffmpeg.git 03Michael Niedermayer 07release/0.8:760929117df0: alsdec: check block length [23:45] <cone-485> ffmpeg.git 03Michael Niedermayer 07release/0.8:caeca53a09fd: qdm2: check array index before use, fix out of array accesses [23:45] <cone-485> ffmpeg.git 03Michael Niedermayer 07release/0.8:391e0fc6c90c: roqvideodec: check dimensions validity [23:45] <cone-485> ffmpeg.git 03Michael Niedermayer 07release/0.8:af343f5cddc5: eamad: fix out of array accesses [23:45] <cone-485> ffmpeg.git 03Michael Niedermayer 07release/0.8:2cac35086c9e: vqavideo: check chunk sizes before reading chunks [23:45] <cone-485> ffmpeg.git 03Michael Niedermayer 07release/0.8:e6ac11e41734: aacdec: check channel count [23:45] <cone-485> ffmpeg.git 03Michael Niedermayer 07release/0.8:41eda870483f: pngdec/filter: dont access out of array elements at the end [23:45] <cone-485> ffmpeg.git 03Michael Niedermayer 07release/0.8:377fabc9e687: Update for 0.8.13 [23:46] <Compn> j-b : is there any g2m1 samples hiding in vlc trac or ftp? :) [23:54] Action: Compn looking for webinars [23:54] <Compn> wmv3 , wmv1 , mss2 [00:00] --- Mon Feb 18 2013

1 0

ffmpeg.log.20130217
by burek 18 Feb '13

18 Feb '13

[00:06] <Xgates> klaxa: YEAH 1.1 works now :) [00:06] <klaxa> :) [00:06] <klaxa> i've messed a bit with mp3 in the last weeks, and it is very easily truncated [00:06] <Xgates> now to see if the presets in WinFF work with this version [00:07] <Xgates> you mean just to shorten mp3? [00:07] <Xgates> or cut... [00:08] <klaxa> shorten or lengthen, anything goes [00:08] <Xgates> ok [00:09] <klaxa> like i said i messed with that a bit in the last weeks, mp3 is arranged in frames, which are a fixed amount of samples (depending on the layer version) you can add and/or take out any frame of an mp3 and it's still spec compliant [00:10] <Kristian_> I need a freelancer that is working with ffmpeg and ffserver, it's about live audio streaming from usb soundcard. Anyone interested? [00:15] <Xgates> ok thanks [00:16] <Xgates> crap I forgot to compile in xvid support errr [00:16] Action: Xgates has to recompile hehe [00:16] <defaultro> saste, looks like I found something fishy. When I use the -vf lutrgb with ffplay, it works. However, when I use -vf lutrgb with ffmpeg, it creates a grey video. Most likely, the video I made last night didn't have the lutrgb option [00:44] <defaultro> saste, will do [00:45] <defaultro> saste, http://pastie.org/6199394 [00:45] <defaultro> saste, that command generated a blank grey video [00:46] <defaultro> however, ffplay shows it properly. I can see the channel gets changed [00:48] <saste> try this one: ffmpeg -f lavfi -i testsrc=d=3 -vf 'lutrgb=r=10+val:g=val-10:b=0+val' out.mp4 [00:48] <saste> also keep in mind that it may be a player problem [00:48] <saste> it is converting to yuv444p, which is not properly supported by many players [00:49] <saste> you may want to force the output to yuv420p [00:49] <defaultro> ok [00:49] <defaultro> it created out.mp4 [00:50] <defaultro> do you want the output in console? [00:50] <saste> no try to play it [00:50] <defaultro> k [00:50] <defaultro> I played it with mplayer and it's color gray [00:50] <defaultro> mplayer -fs out.mp4 [00:50] <saste> try with ffplay [00:50] <defaultro> k [00:51] <defaultro> i see it [00:51] <defaultro> very nice :) [00:51] <saste> then try this: ffmpeg -f lavfi -i testsrc=d=3 -vf 'lutrgb=r=10+val:g=val-10:b=0+val' -pix_fmt yuv420p out.mp4 [00:51] <defaultro> k [00:52] <defaultro> done [00:52] <defaultro> mplayer is able to play it [00:52] <saste> if you have an old mplayer it may not properly support yuv444p H.264 [00:52] <defaultro> got it [00:52] <saste> that's why it is better to avoid that [00:52] <defaultro> avoid mplayer? [00:53] <saste> many players/devices still need to catchup [00:53] <defaultro> got it [00:53] <saste> ipad just to name one [00:53] <defaultro> so from now on, I should use ffplay for playing [00:53] <saste> no you should avoid H.264 + yuv444p [00:53] <defaultro> ok [00:53] <defaultro> i wasn't aware [00:54] <defaultro> i don't even know what yuv444 is :) [00:54] <defaultro> so when encoding, should I use the params you used on the last command? [00:56] <saste> force the pix_fmt to yuv420p, and you got no surprises [00:56] <defaultro> k [00:56] <defaultro> i just tried the long command that was failing, it works now :) [00:58] <defaultro> Thank you so much!!! :) [00:59] <defaultro> that was one complex troubleshooting [00:59] <defaultro> :D [01:01] <defaultro> was it because the recording format was in yuv444 format? [01:02] <defaultro> recording from my video camera [01:21] <Diamonte> If I have a video that was recorded upside down. How could I rotate it in one shot with ffmpeg? [01:22] <Diamonte> from a command line [01:22] <JEEB> see the video filters listing in ffmpeg.org's documentation [01:23] <JEEB> there probably is a filter that has a name that rings a bell [01:23] <juanmabc> -vf "vflip" ? [01:25] <Diamonte> I just saw that too. [01:25] <Diamonte> I'll try it out. [01:56] <Kristian_> Anybody know a freelancer that is working with ffmpeg? Would be grateful for any kind of information. Thank You. [02:23] <saste> Kristian_, check the consulting page on ffmpeg.org [02:24] <saste> what do you need a consultant for? [06:13] <defaultro> wow, the mp4 I uploaded in youtube was 250mb. I downloaded it from youtube, it's downto 33mb. LOL [06:14] <defaultro> i ran strings then grep crf against it, it didn't have any [10:09] <alesan> hi! I have an H.264 byte stream file. This is the output of a hardware H.264 encoder. [10:09] <Kristian_> I need a freelancer that is working with ffmpeg and ffserver, it's about live audio streaming from usb soundcard. Anyone interested? [10:09] <alesan> Any idea how to treat this with ffmpeg to have programs like Vegas Video read it? [10:30] <Fjorgynn> alesan: it doesn't work now? [10:31] <Fjorgynn> alesan: $ ffmpeg -i file.mp4 -c:a copy -c:v copy file.mp4 [10:31] <Fjorgynn> or reencode it with $ ffmpeg -i file.mp4 -c:a copy -c:v libx264 file.mp4 [10:53] <t3ch> hello, why ffmpeg converting so slow command: ffmpeg -i test.mpg -f ogg test.ogg [10:53] <t3ch> what i need to do to convert little faster [10:53] <t3ch> :) [10:55] <sacarasc> Try using more threads, getting a better CPU... [10:56] <t3ch> aha ok :) [10:56] <t3ch> how to set more threads ? [10:56] <sacarasc> -threads X as an output option... Not sure how well theora does multithreading, though... [10:57] <t3ch> ok [10:57] <t3ch> will try [12:09] <pythonsnake> hi [12:09] <pythonsnake> is there a way to ffmpeg -i x.mp4 -o x.mp4? [12:15] <ubitux> only remux? [12:15] <ubitux> oh you mean same filename? [12:16] <ubitux> doesn't sound safe, especially with mp4 [12:35] <pythonsnake> ubitux: why not [12:36] <ubitux> because of seeking [12:36] <pythonsnake> ubitux: the problem i'm trying to fix is [12:37] <pythonsnake> gstreamer's discoverer can't discover a video [12:37] <pythonsnake> and im trying to use ffmpeg (remux to the same file) so it fills the missing fields [12:38] <ubitux> remux to another file, replace the old one [12:38] <pythonsnake> yeah, that's what i thought of [12:38] <pythonsnake> thanks [12:39] <ubitux> it will also reduce race risks on your side [12:39] <pythonsnake> ubitux: what happens if i dont specify extension but vcodecs and acodecs [12:39] <pythonsnake> nothing right? [12:39] <pythonsnake> as in works [12:40] <ubitux> you need to -f mp4 too [12:40] <pythonsnake> ffmpeg -i i.mpg -vcodec copy -acodec copy -scodec copy -y o [12:40] <pythonsnake> o is gonna be mpg? [12:41] <ubitux> not sure it will work [12:41] <ubitux> ffmpeg -i in.mpg -c copy -f mpeg out [12:50] <pythonsnake> thanks ubitux [13:33] <defaultro> I'm reading this thread. I don't know much about yuv. Why would someone convert yuv422 to yuv420, or viceversa? The videos that we see in youtube, what formats are they? Are they yuv420? [13:33] <defaultro> http://forum.videohelp.com/threads/316712-Where-can-i-get-YUV444-YUV422-raw… [13:34] <JEEB> pretty much all video around us is 4:2:0 YCbCr (colloquially called 'yuv') [13:34] <JEEB> and most decoders of video formats can't even take anything else but that [13:38] <defaultro> k [13:39] <defaultro> btw, the 250mb file I uploaded to youtube last night became 33mb, lol. That's a lot of image quality loss with their compression [13:40] <defaultro> i don't even know how they re-encoded it. I didn't see any 264 string from their mp4 file [13:41] <JEEB> they've not written that into their SEI for a long time :P [13:42] <defaultro> was wondering if it was possible to get the picture quality as close as our original [13:49] <defaultro> am I right that my source which is at 60fps will always be downconverted by youtube maybe to 30 fps? [13:50] <JEEB> yes, that's how it used to do it at least [13:52] <defaultro> k [13:57] <WmA> How can one feed a http ts stream into libx264? Do I need to encode from pipe? [14:04] <defaultro> JEEB, this is the video I've been playing last night. I'm trying to fix the greencast(bondi blue). I think I got it to a point that it pleases me - http://www.youtube.com/watch?v=ElZAwRgPbvQ [14:04] <defaultro> glad to hear some comments [14:45] <hid> Hi I'd like to compare videos on youtube cause I suspect somebody to have the same video as mine (a plagiat). How can I do this with ffmpeg ? [14:49] <Mavrik> hmm [14:49] <Mavrik> that's a wierd question [14:49] <Mavrik> what do you want to compare exactly? content? [14:50] <hid> yeah metadatas [14:51] <Mavrik> you see all metadata when doing ffprobe <file> [14:51] <Mavrik> youtube probably strips all of it [14:56] <hid> I checked with mediainfo and gstd was the same [14:57] <hid> evrything was the same [15:10] <Mavrik> hid: that means nothing since Youtube encodes everything to same quality [15:10] <Mavrik> hid: you should look at actual video content not metadata [15:15] <hid> Mavrik : Quality, sound, duration everthing is the same. [15:16] <hid> How can I do to erase all video/audio metadatas ? [15:16] <hid> I've never found any such tool. [15:26] <defaultro> Mavrik, I disagree with what you said that youtube encodes everything to the same quality. I've uploaded videos and pic quality has changed [15:27] <Mavrik> and where did I say that youtube encodes video to the same quality as you gave them? [15:27] <Mavrik> their outputs have same parameters and quality. [15:28] <JEEB> they don't encode for quality, they just have a set bit rate as far as I know [15:29] <JEEB> thus, they are encoding for a set average bit rate [15:29] <JEEB> not same quality [15:29] <Mavrik> *sigh* [15:30] <JEEB> as far as I know, 'tube doesn't use crf [15:30] <JEEB> some video thingies use crf with first pass log, and if it becomes too big, they use a second pass with a set bit rate [17:33] <terr1> I'm playing with ffmpeg (lastest git) and libx264 (lastest git). I'm trying to find some good generel transcoding settings for a html5 website player (progessive download - pseudo streaming). So I was playing with the different x264 presets from ultrafast to veryslow, made a benchmarking shell script to test each setting. I thought this preset setting only effected filesize (compression) not the bitrate? It seems the slower setting the lower bitrat [17:33] <terr1> e? (I'm using cfr 21 to test) [17:34] <Fjorgynn> placebo? [17:52] <xm3> Hi, could anyone help me figure out how to transcode a folder of avis & merge them into one file? [18:12] <grepper> xm3: have you looked at http://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20concatenate%20%28join,%20merg… ? [18:40] <xm3> grepper: thanks looks helpful :) [18:56] <Emmanuel_Chanel> Helllo! [18:57] <Emmanuel_Chanel> Do you know how to encode a TS file with dual monaural sounds into a mp4 with two sound tracks? [19:21] <Fjorgynn> thanks to ffmpeg I can listen to this awesome song [19:43] <creep> how much power you use your headphones on? i listen to music on my 2x40 ohm headphone at 2x1mW 0.2V/5mA rms and i don't feel like i want more [20:26] <ioni> hi guys [20:26] <Mavrik> evenin [20:26] <ioni> does anyone in here happens to use latest mplayer? [20:27] <ioni> the configure script is pulling latest ffmpeg [20:27] <ioni> yesterday mplayer started to crash when using vdpau and i'm not sure where the bug is [20:27] <ioni> in mplayer or ffmpeg [20:27] <ioni> http://paste.xinu.at/h9YC/ [20:28] <ioni> Assertion desc->flags & 16 failed at libavcodec/utils.c:511 [20:29] <WmA> blame Charly Sheen! [20:29] <WmA> ;) [20:29] <ioni> =) [20:59] <terr1> Using ffmpeg and x264, how important is customizing keyint, when doing offline transcoding to be played in html5 web player afterwards? does it matter that much? or mostly when doing live streaming ? [21:01] <Mavrik> mosty when doing live streaming [21:01] <Mavrik> if you choose low keyint you'll make compression (and by extension - quality) worse [21:07] <terr1> Ok thanks.. Ill roll with default 250 for now then... If this setting would create problems how would it manifest itself? Mostly when doing manually seeking ? [21:08] <Mavrik> terr1: by taking a long time to start after seeking [21:09] <Mavrik> also by long periods of time when blocky errors repair themselves [21:10] <terr1> Mavrik: Blocky errors.. thoes artifact squares that sometimes appear when seeking, that takes x amount of time / frames to be corrected? [21:10] <Mavrik> terr1: those only appear if there's something wrong with transmission [21:10] <Mavrik> but yeah [21:10] <Mavrik> usually they don't get repaired until next keyframe [21:11] <terr1> ah ok thanks for the information :) [22:08] <defaultro> hey guys, is it possible to add a film look color via ffmpeg? [22:10] <durandal_1707> defaultro: what film look color means? [22:11] <defaultro> let me find some videos [22:13] <Youka> What's the most simple video format to de-/encode with selfwritten code? [22:13] <defaultro> jump to 1:50 http://www.youtube.com/watch?v=Jis97Eup9_E [22:15] <defaultro> durandal_1707, this one is nice - http://www.youtube.com/watch?v=JhtHvynfcTQ [22:15] <Mavrik> Youka: probably motion jpeg or huffyuv [22:16] <Youka> thx [22:20] <durandal_1707> defaultro: i still not sure what you mean, there are some hue/equalization filters [22:23] <defaultro> ok. looks like I will have to experiment to get the desired color [22:24] <Youka> Isn't there a video format similar simple like wav? Saving raw RGB images with some header data? [22:24] <Mavrik> not really, since it's very inefficient [22:24] <Mavrik> huffyuv is probably the closest [22:24] <Mavrik> or any of other losless codecs [22:24] <Mavrik> plus, most video isn't recorded in RGB [22:26] <Youka> Strange :S thx for these informations [22:34] <defaultro> you guys familiar with twixtor? Can ffmpeg achieve that kind of effect? [22:35] <durandal_1707> twixtor is some kind of interpolated slow motion? [22:36] <defaultro> yup [22:36] <llogan> defaultro: you could look into slowmovideo. it has ffmpeg as a dependency, IIRC. [22:36] <defaultro> really nice very slow motion [22:36] <defaultro> k [22:37] <durandal_1707> there is no such filter in libavfilter, but wrapper or native filter is welcome [22:37] <llogan> and it costs $0 compared to $595. [22:37] <durandal_1707> llogan: where is link with price? [22:38] <llogan> durandal_1707: http://www.revisionfx.com/products/twixtor/pricing/#ae_compatible_4 [22:38] <llogan> defaultro: http://slowmovideo.granjow.net/ [22:39] <defaultro> cool [22:40] <llogan> i've never tried it because i hate cmake. [22:47] <llogan> ioni: see commit 259603b [22:54] <ioni> llogan, yes, i was talking with Carl [00:00] --- Mon Feb 18 2013

1 0

ffmpeg-devel.log.20130216
by burek 17 Feb '13

17 Feb '13

[00:12] <ubitux> saste: mmh i don't remember [00:13] <ubitux> i think i didn't submit because i had some output issues [00:13] <ubitux> i'll have a look in a moment [00:14] <ubitux> we have almost the same code [00:15] <ubitux> i guess it's ok [00:15] <ubitux> huh that was my code maybe [00:16] <ubitux> let me re-bench it locally [00:18] <cone-941> ffmpeg.git 03Stefano Sabatini 07master:fbcc584d3abb: lavfi/unsharp: use named options, and add missing checks on matrix size values [00:18] <cone-941> ffmpeg.git 03Stefano Sabatini 07master:37b42bcb7db9: doc/filters: itemize unsharp examples [00:18] <cone-941> ffmpeg.git 03Stefano Sabatini 07master:2042cd37699f: lavfi/unsharp: extend range for amount values, to reflect the behavior of mp=unsharp [00:18] <cone-941> ffmpeg.git 03Clément BSsch 07master:51ba843f195c: lavfi/kerndeint: use aligned linesizes for the temporary buffer. [00:18] <cone-941> ffmpeg.git 03Stefano Sabatini 07master:f84ef19f3f8b: lavfi/kerndeint: move is_rgb_packed definition to the internal context [00:18] <saste> ubitux, uh I already pushed [00:19] <ubitux> haha it simplifies my test then [00:19] <ubitux> don't remove mp=kerndeint now [00:19] <ubitux> give me a few minutes :P [00:19] <saste> anyway, i think that the benchmarking differences may be due to the fact that i don't enable optims [00:19] <saste> no rush, take your time [00:19] <ubitux> what do you mean with the optim? [00:19] <saste> you can use my benchmarking patch on top of git master [00:20] <saste> --disable-optimizations [00:20] <ubitux> yes i have the same patch locally [00:20] <ubitux> yes but why would that affect the benchmarks diff? [00:21] <saste> yes, indeed it shouldn't [00:26] <ubitux> 68908034 decicycles in kerndeint, 2048 runs, 0 skips [00:26] <ubitux> 68417747 decicycles in mp=kerndeint, 2048 runs, 0 skips [00:27] <ubitux> 68329763 decicycles in mp=kerndeint, 2048 runs, 0 skips [00:28] <ubitux> 69982854 decicycles in kerndeint, 2048 runs, 0 skips [00:28] <ubitux> i'm doing ./ffmpeg -f lavfi -i testsrc=s=720x480 -vf tinterlace,kerndeint -f null - [00:28] <ubitux> (btw, surprise with s=720x479, with both filters) [00:29] <ubitux> saste ^ [00:30] <ubitux> anyway, i don't think the performance matters, but i'm actually very curious about the reason of that difference [00:31] <ubitux> i wasn't able to get it last time [00:34] <cone-941> ffmpeg.git 03James Darnley 07master:724061888be1: lavfi/yadif: remove unnecessary cast [00:34] <cone-941> ffmpeg.git 03James Darnley 07master:87d359ff8f55: lavfi/yadif: support 9, 12, and 14 bits per pixel [00:40] <saste> ubitux: you discovered a bug in testsrc [00:42] <ubitux> oh it's in testsrc? okay [00:42] <ubitux> i didn't dig [00:44] <saste> ubitux, I'm testing with 720x479, and still mp=kerndeint is slower [00:49] <ubitux> so with the command i gave, you get faster results? [00:49] <ubitux> maybe some magic somewhere [00:50] <ubitux> let's not bother too much about it then [00:51] <saste> 166831123 decicycles in mp=kerndeint, 1024 runs, 0 skips [00:51] <saste> 166170803 decicycles in kerndeint, 1024 runs, 0 skips [00:51] <saste> the difference is minimal, but yet it is consistently faster [00:52] <saste> i'll drop mp=kerndeint tomorrow, unless we have more benchmarks which prove the contrary [01:02] <saste> ping on "lavfi/unsharp: add check on matrix x/y size values oddity" [01:02] <saste> so i can finally drop mp=unsharp [03:13] <cone-941> ffmpeg.git 03Marton Balint 07master:b853cfe7eaf1: ffplay: remove refresh thread to achieve more precise frame delay [03:13] <cone-941> ffmpeg.git 03Marton Balint 07master:5534d8f75e27: ffplay: remove remains of video_clock [03:13] <cone-941> ffmpeg.git 03Marton Balint 07master:77ab702a848a: ffplay: only display frames if there are no pending events [03:13] <cone-941> ffmpeg.git 03Michael Niedermayer 07master:b9c544891b05: Merge remote-tracking branch 'cus/stable' [08:55] <cone-323> ffmpeg.git 03Anton Khirnov 07master:ccd70d9c1689: vf_yadif: factorize initializing the filtering callbacks [08:55] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:fab8a89abeae: Merge commit 'ccd70d9c1689990e5aef2de383199bbc7cf60d13' [09:14] <cone-323> ffmpeg.git 03Anton Khirnov 07master:64ed397635ef: vf_yadif: fix out-of line reads [09:14] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:446f7c62a2d1: Merge commit '64ed397635ef2666b0ca0c8d8c60a8bc44581d82' [11:09] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:1e78679768b3: mpegvideo_enc: draw edges on input [12:47] <durandal_1707> michaelni: yadif have unused absrefs variable, probably included with merge [12:49] <durandal_1707> s/included with/left from/ [12:56] <cone-323> ffmpeg.git 03Paul B Mahol 07master:af882e1819ca: lavfi: port MP noise filter [13:05] <durandal_1707> michaelni: why ff_* is global in swr? [13:15] <michaelni> mplayer and our libmpcodecs used some of these i think [13:21] <cone-323> ffmpeg.git 03Anton Khirnov 07master:d2a25c4032ce: get_buffer(): do not initialize the data. [13:21] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:a53b144ec041: avcodec: add avpriv_color_frame() [13:21] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:1c5b8b916d3d: h264: color frames gray [13:21] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:bb29ee62e9ef: Merge commit 'd2a25c4032ce6ceabb0f51b5c1e6ca865395a793' [13:21] <michaelni> durandal_1707, rand()/srand() is not good in a lib it can randomly affect applications that expect a repeatable sequence from rand() [13:24] <michaelni> in vf_noise that is [13:31] <durandal_1707> michaelni: so you wait until i commit this to comment? [13:32] <durandal_1707> also it was already in mp [13:33] <durandal_1707> it can be replaced by some fibonaci stuff saste mentioned [13:41] <michaelni> yes i wait until iam aware of a problem before mentioning, ill try to mention it next time before iam aware of it but iam not sure how to do that yet [13:41] <michaelni> mp is no real lib so it can do things like that [13:42] <durandal_1707> michaelni: i don't follow [13:43] <wm4> uh doesn't lavu have RNGs [13:43] <michaelni> it does [13:43] <michaelni> durandal_1707, i wasnt aware of the problem before noise was commited .... [13:45] <durandal_1707> michaelni: i see srand is used in lavu [13:46] <michaelni> its under #ifdef TEST [13:46] <durandal_1707> michaelni: so what should be used? av_lfg? [13:49] <durandal_1707> michaelni: ok to remove mp=noise it is slower that native one(for me, the SIMD stuff is not enabled at compilation, but this can be added later to native one) [13:50] <michaelni> av_lfg is fine, other generators would be fine too [13:50] <Compn> 26 cycles faster. :) [13:52] <Compn> BBB-work : dxva2 is windows hwaccel , chrome could take advantage of gpu speed up ... [13:55] <durandal_1707> really? [14:11] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:b4bb996d6f41: yadif: remove unused variable [14:11] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:0a707da37b41: cinepack: print an error if cinepak_decode() failed [14:20] <cone-323> ffmpeg.git 03Anton Khirnov 07master:68f930d2188a: h264: simplify calls to ff_er_add_slice(). [14:20] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:131b9c924d6d: Merge commit '68f930d2188aba5b32624887dcbf688c23482834' [14:34] <cone-323> ffmpeg.git 03Anton Khirnov 07master:2491f9ee29ae: h264: remove a pointless if() [14:34] <cone-323> ffmpeg.git 03Anton Khirnov 07master:d9ebb00dcbaa: svq3: remove a pointless if() [14:34] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:8b7568cc3482: Merge commit 'd9ebb00dcbaac3812b8b1fbc3d6e027506c11cbc' [15:58] <durandal_1707> what is max allowed value for color components in YUV 16 bit? [16:08] <cone-323> ffmpeg.git 03Stefano Sabatini 07master:2bac1535db97: Changelog: add missing empty line after the version line [16:24] <saste_> mp=unsharp is faster than unsharp [16:42] <durandal_1707> saste_: really? how much? [16:42] <saste_> durandal_1707, it's due to indirect variable access [16:43] <saste_> fp->x vs. x [16:43] <saste_> since this is in a loop it seems it is measurable [16:44] <saste> the other difference is due to a call to av_clip() [16:44] <saste> which is only relevant with no optims [16:44] <durandal_1707> look if it is different by 100 cycle I do not care [16:44] <saste> so all in all i'm going to push these changes and drop mp=unsharp [16:45] <durandal_1707> what compiler you are using? [16:45] <saste> 68524757 decicycles in unsharp, 4096 runs, 0 skips [16:45] <saste> 68856107 decicycles in mp=unsharp, 4096 runs, 0 skips [16:45] <saste> so with the "optims" unsharp is a bit faster now [16:46] <saste> durandal_1707, gcc [16:46] <durandal_1707> what version? [16:47] <saste> gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3 [16:49] <durandal_1707> saste: it is expected for mp to be slow as it does some extra stuff... [16:49] <durandal_1707> also i think it is still using slices.... [16:50] <saste> durandal_1707, no, i'm just benchmarking the processing code [16:53] <durandal_1707> saste: also lavfi have problems when using mpegts as input in VV->V filter(nut works fine) [16:54] <durandal_1707> i guess it is related to probing [16:54] <saste> which kind of problems? [16:55] <saste> related to framerate or something? [16:55] <saste> there is a lavfi open ticket for that [16:56] <durandal_1707> [buffer @ 0x2940c3a0] Unable to parse option value "0x0" as image size [16:56] <durandal_1707> interestingly it works if input is 2nd [16:57] <durandal_1707> it have 2 video streams, but only one is actually ther [16:57] <durandal_1707> some mpegts bug [17:10] <mateo`> what is the consequences of setting AVFrame.interlaced_frame to 1 for a picture ? does it mean that the picture only contain odd or even fields ? [17:11] <durandal_1707> it means frame is interlaced and not progressive, nothing more - nothing less [17:12] <nevcairiel> an avframe typically always contains both fields weaved together, so yeah, it just indicates that this is the fact [17:13] <mateo`> thx [17:15] <durandal_1707> why would frame contain only odd/even fields? [17:16] <mateo`> it was a dumb question, i just wanted to be sure [17:17] <mateo`> by definition a frame contains both fields [17:18] <Daemon404> nevcairiel, from other channel: i feel like im taking crazy pills [17:22] <Shiz> Daemon404: sounds like guys are having fun [18:59] <cone-323> ffmpeg.git 03Anton Khirnov 07master:54974c62982a: error_resilience: decouple ER from MpegEncContext [18:59] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:1fad547cefb4: Merge commit '54974c62982ae827becdbdb9b620b7ba75d079a0' [19:10] <cone-323> ffmpeg.git 03Anton Khirnov 07master:1d0feb5d1ac0: mpegvideo: split ff_draw_horiz_band(). [19:10] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:ddbbfd95bee4: Merge commit '1d0feb5d1ac04d187b335f0e8d411c9f40b3a885' [22:36] <ubitux> mega accepts bitcoins @_@ [23:33] <cone-323> ffmpeg.git 03Anton Khirnov 07master:2c541554076c: h264: deMpegEncContextize [23:33] <cone-323> ffmpeg.git 03Michael Niedermayer 07master:b7fe35c9e50e: Merge remote-tracking branch 'qatar/master' [00:00] --- Sun Feb 17 2013

1 0

ffmpeg.log.20130216
by burek 17 Feb '13

17 Feb '13

[00:23] <undercash> hello [00:23] <undercash> having a weird issue, i wanted to burn subtitle .srt, so i compiled with libass,. so far so good [00:24] <undercash> but it only display the first sentence then nothing [00:24] <undercash> it s kinda weird.. [00:27] <undercash> http://gyazo.com/db3303aa9e8253459d9e57eba157b106 [00:34] <Kapsel> hi, I'm trying to use ffmpeg to extract video frames from an ARRI Alexa video camera, as PNG's, and then use oiiotool (opencolorio) to apply a color conversion LUT to them, and write it back to an H264. But I see some color shifting, when writing the initial PNG frames, I guess its because ffmpeg reads the data as one color space, and then converts it. any ideas on how to solve rthis? [00:35] <Kapsel> https://sites.google.com/site/openimageio/ this is the tool that I am using [00:36] <Kapsel> works extremely well [00:36] <Kapsel> and converts an absolutely perfect color correct frame, but again, the colors drifts slightly when exporting to png from ffmpeg [00:37] <klaxa> Kapsel: i think you are right, default colorspace for h264 is yuv and not rgb [00:38] <Kapsel> yep, but this happens even when I do my source -> rgb [00:38] <alesan> can h264 support RGB at all? [00:38] <Kapsel> if I compare the PNG's with PNG's taht I export from another program, they differ slightly in colors [00:38] <Kapsel> err, source -> png [00:38] <klaxa> alesan: there is libx264rgb at least... [00:39] <klaxa> >H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 (decoders: h264 h264_vdpau ) (encoders: libx264 libx264rgb ) [00:39] <alesan> hm [00:41] <klaxa> Kapsel: i have not completely understood your work-chain [00:41] <klaxa> so you have a video-file [00:41] <klaxa> then you extract one rgb frame from that [00:41] <klaxa> apply a color filter [00:41] <klaxa> and add it to an h264 stream with ffmpeg [00:42] <klaxa> the resulting stream does not look like the png you exported from your source video? [00:42] <Kapsel> klaxa: sorry. My video file is a ProRes 4444 file, that I extract to PNG's. Then I run them over oiiotool, a program that applies a color LUT, write them to a new file, and then I encode an H264 with ffmpeg. when I do the first source -> png, the colors shifts compared to if I do the same frame with my other software, DaVinci Resolve [00:43] <Kapsel> the DaVinci Resolve extracted PNG is totally accurate, but my ffmpeg PNG is different. [00:43] <Kapsel> I came across this thread, when searching: http://blenderartists.org/forum/showthread.php?209240-Lut-Support-in-Blende… [00:43] <Kapsel> Someone mentions this: [00:43] <Kapsel> ffmpeg -i input -sws_flags lanczos+full_chroma_int+full_chroma_inp+accurate_r nd -f image2 output%05d.png [00:43] <Kapsel> this command extracts to 550d footage to using full range levels 1-254. [00:43] <Kapsel> but I am not sure if those parameters matters in my case. [00:46] <klaxa> so if you are using DaVinci Resolve to extract the frames from the ProRes 4444 file, it's perfect, but if you use ffmpeg, the colors are off? [00:46] <Kapsel> yeah [00:47] <undercash> well sorry, works perfectly.. was just the wrong sub file..... [00:50] <klaxa> Kapsel: can't find a lot in the codecs section of my ffmpeg binary other than 3 encoders and 2 decoders [00:50] <klaxa> i don't know whether the 4444 profile is supported though [00:50] <klaxa> however, http://en.wikipedia.org/wiki/ProRes_422#Open_Source_Projects [00:51] <klaxa> maybe that fork is of interest for you [00:53] <Kapsel> I know of ffmbc, we use it a couple of places. [00:54] <Kapsel> I do think that Prores 4444 decoding is fully supported by ffmpeg, though. [00:55] <klaxa> mmhh... have a look at the output log and see if you can find something like swscaler or something that indicates a colorspace change [00:57] <Kapsel> I have both ffmbc and ffmpeg installed on my laptop, downloading a Prores 4444 file from the office right now, going to do some quick tests. [00:57] <Kapsel> good point, will try that [01:02] <bcoudurier> Kapsel, is your source hd ? [01:02] <Kapsel> bcoudurier: its 2K [01:02] <bcoudurier> is it using bt 709 as colorspace ? [01:03] <Kapsel> no, thats what I'm trying to convert it to. its some logC colorspace right now. [01:03] <Kapsel> afaik. [01:03] <Kapsel> and hey, thank you for your work on ffmpeg and ffmbc for us in broadcast and post [01:04] <Kapsel> http://www.arri.com/camera/digital_cameras/learn/log_c_and_rec_709_video.ht… [01:05] <Kapsel> we shoot in Log C [01:06] <Kapsel> I can send you a .mov if you'd like to see how it looks like. [01:09] <Kapsel> http://dpaste.com/933645/ info on file here [01:14] <Kapsel> http://dpaste.com/933662/ and this is what I get when I export a frame (ignore the errors) [01:16] <Kapsel> interestingly enough, ffmpeg produces a 12.8MB PNG file, ffmbc creates a 3.8MB PNG [01:16] <Kapsel> using the same command, but different bins. [01:18] <Kapsel> full output here: http://dpaste.com/933679/ [01:19] <Kapsel> colors varies slightly, maybe ffmbc does it better for some reason(?) - will try to compare with some material exported from Resolve [01:20] <Kapsel> Stream #0:0(eng): Video: png, rgb48be, 2048x1152 [SAR 1:1 DAR 16:9], q=2-31, 200 kb/s, 90k tbn, 25 tbc [01:20] <Kapsel> Stream #0.0(und): Video: png, rgb24, 2048x1152p [PAR 1:1 DAR 16:9], vbr, lossless, 25.00 fps [01:20] <Kapsel> rgb24 vs rgb28be, maybe thats the answer [01:21] <Kapsel> rgb48be [02:10] <defaultro> other than -vf hue, what other color correction filters we can use in ffmpeg? [02:36] <tds5016> hey. Is it possible to use ffserver to receive an rtp stream, and rebroadcast it as ffmpeg? [02:37] <tds5016> rebroadcast as hls* [07:15] <bigmac> how can i convert a video to a smaller size for streaming over http [07:15] <bigmac> vcodec examples? [08:02] <defaultro> hey folks, I have an existing -vf "mp ......", how do I add hue=h? [08:03] <defaultro> this failed, ffplay 00013.MTS -vf "mp=eq2=1.3:1.1:-.1,hue=h=30:0" [08:11] <defaultro> found it, using comma :) [12:12] <WmA> anyone sees the error in my syntax? [12:12] <WmA> ffmpeg -i "http://192.168.1.2:7522/upnp/channelstream/0.ts" -c:v libx264 -preset slow -crf 19 -x264opts keyint=123:min-keyint=20 -c:a copy output.mp4 [12:12] <WmA> At least one output file must be specified [13:22] <octe_> Trying to follow this guide: http://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20capture%20a%20lightning%20(th… , but it's not working to play back the output from samplicate [13:22] <octe_> http://pastebin.com/LMCgjnmD [13:22] <octe_> that is the output i get [15:33] <Aktau> Hey guys :). Does anyone know what some good settings are to make libvpx encode with variable bitrate. Leaving me to just choose some kind of quality setting? [15:33] <Aktau> I've tried crf but it appears to be ignored [15:33] <Aktau> The ffpresets are cool but they set a bitrate of 2M. Which is for some videos too much and for some perhaps too little [15:34] <Aktau> I'd like to suffer as little quality loss from the source H.264 file as possible [15:34] <Aktau> I'm using the newest ffmpeg freshly compiled [15:34] <Mavrik> sec [15:36] <JEEB> Aktau, -crf is libx264-only [15:37] <Aktau> JEEB: I've heard many conflicting things. This is from the WebM project itself: http://wiki.webmproject.org/ffmpeg [15:37] <Mavrik> Aktau: -crf should work [15:37] <Aktau> JEEB: it actually lists a crf [15:37] <Mavrik> Aktau: if you have new enough ffmpeg [15:37] <JEEB> o_O [15:37] <Mavrik> I see that in older versions only bitrate CBR mode was available [15:37] <JEEB> uhh [15:38] <Aktau> Mavrik: ffmpeg 1.1.2 [15:38] <Aktau> I compiled it yesterday via homebrew for mac, works like a charm [15:38] <Aktau> Just looking for the right flags [15:38] <JEEB> I'd be surprised if there was a -crf option for libvpx :s [15:38] <JEEB> ask on #vp8 [15:38] <JEEB> as far as I know, crf has only been in libx264 [15:39] <Mavrik> JEEB: I'm looking at the source. [15:39] <Aktau> Yea that's what many people on the web say as well [15:39] <Aktau> Been looking for almost 2 days how to do this [15:39] <Mavrik> JEEB: "-crf" parameter maps to libvpx's "constant quality" option [15:39] <JEEB> lol [15:39] <Mavrik> http://ffmpeg.org/doxygen/1.0/libvpxenc_8c-source.html [15:39] <Mavrik> line 271. [15:39] <JEEB> that is not crf >_< [15:39] <JEEB> IIRC the constant quality in libvpx is something quite different [15:39] <JEEB> at least looking at their documentation [15:39] <Aktau> ffmpeg -i ironman3-biggame_h1080p.mov -codec:v libvpx -threads 8 -vpre 1080p-aktauvar -crf 11 -codec:a libvorbis ironman3-biggame_h1080p-var.webm [15:39] <Aktau> Is my current commandline [15:40] <Aktau> This creates a really small file, source is 64MB, output is about 4 [15:40] <Aktau> Quality is really bad so I assume it does 200kb/s as it shows after entering the command [15:40] <Mavrik> JEEB: terminology aside, "-crf" just switches libvpx to "CQ" mode and passes the number as CQ factor :) [15:40] <JEEB> yes [15:41] <Mavrik> which is what Aktau wants I guess [15:41] <JEEB> I understand that [15:41] <JEEB> no [15:41] <JEEB> he wants x264's crf [15:41] <JEEB> something like that [15:41] <JEEB> while libvpx's "constant quality" as far as I know [15:41] <JEEB> is something else [15:41] <JEEB> unless they've recently implemented something better [15:41] <JEEB> http://blog.webmproject.org/2011/03/vp8-constrained-quality-cq-encoding.html [15:41] <Aktau> I guess I want something to regulate quality without fixing the bitrate [15:42] <Mavrik> JEEB: ffs, their documentation calls "VPX_CQ" the "constant quality" parameter [15:42] <Mavrik> and in other docs they call it "constrained quality" [15:42] <Mavrik> -_- [15:42] <JEEB> yes, their documentation sucks on it and is not definitive [15:43] <JEEB> and since I don't really care for vp8, I've never really derped at the #vp8 folk about whether or not this is something similar to x264's crf [15:43] <JEEB> putting it under the -crf option in ffmpeg is another piece of craziness as well [15:43] <Aktau> That, it sure is [15:43] <Aktau> I'll inquire there as well [15:44] <Mavrik> slides from senior engineer: "This is not a constant quality mode; but can be made to [15:44] <Mavrik> behave like one (almost). [15:44] <Mavrik> " [15:44] <Mavrik> they need better docs -_- [15:45] <JEEB> I hope they'll make their docs and the encoder better for when they start finishing up VP9 :P [15:46] <Mavrik> considering how they rushed it to Chromium... I doubt it :P [16:30] <defaultro> is it possible to reduce green color? [16:30] <saste_> defaultro, lutrgb [16:31] <defaultro> awesome, I'll search for it [16:37] <defaultro> what's default value of rgb, is it 1? [16:37] <defaultro> i got it working btw :) just need to know the default values [16:42] <saste_> there is no default value for a pixel color [16:42] <saste_> defaultro, what you mean? [16:49] <defaultro> just got back [16:49] <defaultro> the default value that will not change rgb [16:49] <defaultro> i'll try 0 [16:50] <defaultro> 0 made video all black :) [16:50] <durandal_1707> defaultro: there is expression which is evaluated [16:51] <durandal_1707> using constant values is not much of use [16:51] <octe> did anyone see my question? [16:51] <defaultro> k [16:51] <octe> Trying to follow this guide: http://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20capture%20a%20lightning%20(th… , but it's not working to play back the output from samplicate [16:51] <octe> http://pastebin.com/LMCgjnmD that's the output i get [16:51] <octe> maybe i need to specify more parameters explicitly? [16:53] <durandal_1707> octe: you sure it is mpegts stream? [16:53] <octe> that is the format specified in the streamer [16:54] <durandal_1707> does it work if you do not force it? [16:55] <octe> udp://127.0.0.1:5678: Invalid data found when processing input [16:57] <durandal_1707> increse probe/analyize duration [16:57] <octe> durandal_1707, trying it without samplicate in the middle doesnt seem to work either [16:58] <octe> ah [16:58] <durandal_1707> octe: if it fails to auto detect it usually forcing it will not help at all [17:00] <octe> durandal_1707, just following the tutorial from the wiki.. [17:00] <octe> what's a good value to try for probesize / analyzeduration? [17:00] <durandal_1707> big one [17:21] <rindolf> Hi all. [17:23] <rindolf> How do I record the screen and a mono sound track from my microphone at the same time on Linux using ffmpeg-1.1.2-1.mga3.tainted on Mageia Linux 3/Cauldron? [17:23] <rindolf> Something like a screencast. [17:52] <octe> rindolf, something like this works for me: ffmpeg -f x11grab -s 1920x1200 -r 15 -i :0.0 -f pulse -ac 2 -i default -vcodec libx264 -tune zerolatency -vb 5000k -pix_fmt yuv420p -s 1920x1200 -acodec libmp3lame -ab 96k -ar 22050 -threads 0 -f flv out.flv [17:52] <rindolf> octe: thanks. [17:53] <Fjorgynn> why 420p? [17:53] <Fjorgynn> why not 720P? [17:53] <Fjorgynn> ah, I read wrong [17:53] <Mavrik> evenin' [17:54] <rindolf> octe: http://paste.debian.net/234689/ - I'm getting this. [17:55] <rindolf> Mavrik: evening. [17:56] <Fjorgynn> rindolf: http://www.youtube.com/watch?v=eFM9ZI4BY0w [17:56] <Fjorgynn> maybe? [17:56] <Fjorgynn> and why flv, use x264 imo [17:56] <rindolf> Fjorgynn: OK. [17:57] <rindolf> Fjorgynn: windows? [17:57] <rindolf> Fjorgynn: I'm on Linux. [17:57] <rindolf> Fjorgynn: Give me freedom or give me death. [17:57] <Mavrik> ... [17:59] <rindolf> Fjorgynn: someone I know become much more cynical/etc. after he switched from using Debian Linux to Windows (for Photoshop and for video stuff). [17:59] <Mavrik> dude [17:59] <rindolf> Mavrik: what? [17:59] <Mavrik> please leave OS evangelism out of this channel. [17:59] <rindolf> Mavrik: OK. [18:00] <rindolf> Mavrik: I want to record a screencast on Linux using ffmpeg. [18:00] <octe> Fjorgynn, flv is just a container isnt it? [18:00] <Mavrik> anyway, what seems to be the problem? [18:00] <octe> it's still x264 [18:00] <rindolf> Not on Windows, not on whatever. [18:00] <Mavrik> mhm [18:00] <octe> rindolf, did you change the screen size parameter? [18:00] <Mavrik> rindolf: and I see you're using x11grab [18:00] <octe> to match [18:00] <rindolf> Mavrik: http://paste.debian.net/234692/ [18:00] <rindolf> octe: no, I did not. [18:00] <octe> rindolf, is your screen smaller than 1920x1200? [18:00] <octe> then that is probably the issue. [18:00] <Mavrik> rindolf: can you pastebin the actual command as well? [18:01] <Mavrik> I see only a bash script [18:01] <rindolf> Ah, it's x1080 [18:01] <Mavrik> oh, it's in the log [18:01] <rindolf> Mavrik: it's what octe gave. [18:01] <Mavrik> yup, your resolution is probably wrong [18:02] <Fjorgynn> rindolf: I know that hak5 have a tutorial [18:03] <rindolf> OK, now I'm getting no sound. [18:04] <rindolf> Fjorgynn: will you forget about hak5? http://www.shlomifish.org/philosophy/computers/web/use-qmail-instead/ ? [18:04] <rindolf> Mavrik: OK, resolution is fine now and the video was recorded but no sound. [18:04] <rindolf> Let me paste the new script. [18:04] <rindolf> Mavrik, octe : thanks. [18:04] <Mavrik> rindolf: hmm, do you have pulseaudio server on your machine? [18:04] <Mavrik> IIRC Debian doesn't install it by deafult [18:04] <Mavrik> *default [18:05] <octe> yeah, that command uses pulseaudio [18:05] <octe> change the -f pulse to -f alsa to use alsa i think? [18:05] <rindolf> Mavrik: http://paste.debian.net/234694/ - see this. [18:05] <Mavrik> -f alsa -i hw0,0 I think [18:05] <Mavrik> or is it hw0:0 [18:05] <Mavrik> not sure what alsa device aliases are [18:05] <octe> if you're using pulse check what input you get in the pulseaudio volume control [18:06] <rindolf> Mavrik: I'm using ALSA. [18:06] <rindolf> Mavrik: there's "default". [18:07] <octe> do you have audio recording working in general with alsa? [18:07] <rindolf> octe: yes, works fine in Audacity. [18:17] <rindolf> http://paste.debian.net/234696/ - this is what I get - script + output - in the console. [21:57] <defaultro> anyone alive :) what's the best compression setting so that when we upload to youtube, it won't affect the image quality too much [22:00] <klaxa> defaultro: youtube will fuck up your quality anyways, don't bother too much [22:00] <defaultro> k [22:01] <rindolf> defaultro: maybe try using an .mp4 file. [22:01] <defaultro> i am on mp4 [22:05] <Xgates> hi guys [22:05] <divVerent> defaultro: just use high quality settings [22:05] <Xgates> if I use this to trim a video the audio is messed up; [22:05] <Xgates> ffmpeg -ss 00:00:00 -i 1.avi -c:v copy -c:a copy -t 00:03:13 output.avi [22:05] <divVerent> so the quality loss will be mostly on youtube's side [22:06] <divVerent> then your encoding process just won't matter, quality loss wise [22:06] <divVerent> that's as good as you can get then [22:06] <Xgates> then I found this for mencoder and it works; [22:06] <Xgates> mencoder -ss 00:00:00 -endpos 00:03:13 -oac copy -ovc copy 1.avi -o trimmed.avi [22:06] <Xgates> so not sure if mencoder works better for simple video trimming... [22:08] <defaultro> i read somewhere that if our encoding is really high, the loss will be little [22:08] <Xgates> it's mpeg4 & mp3 and if I leave out the -c:a copy it works fine, not sure why it won't do the copy and work... [22:09] <Xgates> so is there one command that will work on any codec for trimming? [22:10] <Xgates> without transcoding, just copying... [22:12] <divVerent> defaultro: "little", no [22:12] <divVerent> but the higher your quality settings are [22:12] <defaultro> k [22:12] <divVerent> the less your own quality loss affects the result [22:12] <divVerent> so it then is only youtube's quality loss, which you can't control [22:12] <defaultro> yup [22:13] <divVerent> I don't know which crf to use, but if you want to find out... [22:13] <defaultro> what would be the highest quality setting? q=15? [22:13] <defaultro> i mean crt [22:13] <divVerent> upload a HQ video to youtube, with a very low CRF [22:13] <divVerent> then download it from youtube [22:13] <divVerent> and see what crf the file has [22:13] <defaultro> ah [22:13] <divVerent> but, yes, crf 15 sounds safe to me [22:14] <defaultro> can i find out what crf was use in my encoding? [22:14] <divVerent> yes [22:14] <divVerent> run "strings" on the video file [22:14] <defaultro> ok [22:14] <divVerent> then you get a long string from x264 [22:14] <divVerent> which contains all x264 settings [22:14] <defaultro> k [22:14] <divVerent> not sure if youtube strips that, though [22:14] <divVerent> if they DO strip it, go by file size [22:15] <divVerent> make your upload twice the size download from youtube will yield, then you also are pretty safe [22:15] <divVerent> or just use paranoidly high quality settings, like crf 15 :P [22:15] <defaultro> ah, this one used 23 for crf. That's why isn't that good [22:15] <divVerent> and PCM or REALLY high bitrate MP3 or AAC for audio [22:15] <defaultro> i'm fine with audio [22:15] <divVerent> crf 23 is default [22:15] <defaultro> got it [22:23] <defaultro> this is failing [22:23] <defaultro> ffmpeg -y -i $2 -crf $1 -vcodec libx264 -s 1920x1080 -r 30000/1001 -vf 'lutrgb=r=1.1*val:g=.9*val:b=1*val, hue=0:1.5, mp=eq2=1:1.5:0' -acodec libfaac -ac 2 -vpre slow -threads 0 00013.mp4 [22:23] <defaultro> /usr/local/share/ffmpeg/libx264-slow.ffpreset: Invalid option or argument: 'directpred=3 [22:23] <JEEB> don't use preset files [22:23] <defaultro> what should I change? [22:23] <JEEB> use -preset and the actual internal x264 presets [22:24] <JEEB> http://mewiki.project357.com/wiki/X264_Settings#preset [22:24] <defaultro> did you mean -vpre slow? [22:24] <JEEB> ^ list of presets [22:24] <JEEB> yes [22:24] <JEEB> also I am not fully sure why you are setting the frame rate [22:24] <JEEB> esp. after -i [22:25] <defaultro> it's because video is at 60p [22:25] <defaultro> oh, where should I place it? [22:25] <JEEB> ok, so you are converting frame rate [22:25] <JEEB> then it's ok [22:25] <defaultro> k [22:25] <JEEB> well, -r after -i converts frame rate [22:25] <JEEB> -r before -i sets a frame rate (in certain cases) [22:25] <JEEB> and usually people don't have to set -r [22:26] <JEEB> because the input time stamps get used [22:26] <defaultro> so mine is fine? [22:26] <JEEB> yes [22:26] <defaultro> k [22:26] <defaultro> ffmpeg is running now :) I removed -vpre and changed it to -preset slow [22:26] <defaultro> Thanks a lot [22:36] <defaultro> hmmm. video became blank gray [22:36] <defaultro> i'm rerunning it. Command is ffmpeg -y -i 00013.MTS -crf 15 -vcodec libx264 -s 1920x1080 -r 30000/1001 -vf 'lutrgb=r=1.1*val:g=.9*val:b=1*val, hue=0:1.5, mp=eq2=1:1.5:0' -acodec libfaac -ac 2 -preset slow -threads 0 00013.mp4 [22:42] <defaultro> it's still color gray :( what could be the culprit parameter? [22:46] <Mavrik> um [22:46] <Mavrik> maybe the video filter where you change colors? [22:47] <defaultro> it works i'm sure because i used it last night but without the other parameters [22:48] <defaultro> i took out -r and -s [22:48] <defaultro> let's set if it's going to work this time [22:50] <defaultro> still blank gray :( ffmpeg -y -i 00013.MTS -crf 23 -vcodec libx264 -vf 'lutrgb=r=1.1*val:g=.9*val:b=1*val, hue=0:1.5, mp=eq2=1:1.5:0' -acodec libfaac -ac 2 -preset slow -threads 0 00013.mp4 [22:50] <defaultro> this is what I used and it works, ffmpeg -i 00013.MTS -vf 'lutrgb=r=1.1*val:g=.9*val:b=1*val, hue=0:1.3' 00013.mp4 [22:51] <defaultro> but crf on that is 23 which is default [22:51] <defaultro> maybe -preset slow is affecting it [22:52] <JEEB> no, no it shouldn't [22:52] <defaultro> not sure why when using the long command, output becomes gray [22:52] <JEEB> you are most probably doing something wrong, as simple as that [22:52] <JEEB> check your filter chain and so forth [22:52] <JEEB> have fun [22:52] <defaultro> it works if using the short command I showed above [22:58] <defaultro> i took out -vf and video is now being displayed [22:58] <defaultro> where should -vf be placed? [22:58] <JEEB> the placement of -vf itself is OK [22:58] <JEEB> I'm pretty sure you have problems with the filter chain [22:59] <defaultro> k [22:59] <defaultro> i took out two from the filter chain and left lutrgb [23:00] <defaultro> video is back to blank gray again with filter lutrgb [23:01] <defaultro> now trying filter hue [23:01] <defaultro> it works [23:02] <defaultro> trying with both hue and mp [23:02] <defaultro> works again. so lutrgb is the culprit [23:13] <defaultro> played with lutrgb values and it wouldn't cooperate :( [23:13] <defaultro> maybe lutrgb is not compatible with libx264 [23:13] <JEEB> try converting to rgb before it and back to yuv420p after it [23:14] <defaultro> how do I do that? [23:14] <defaultro> the reason I am using lutrgb is to just reduce the green channel [23:15] <JEEB> look at the listing of video filters, there should be one to change colorspaces around [23:15] <defaultro> k [23:16] <JEEB> because either you are failing at life regarding the usage of that filter, or you were and are using a diff. colorspace between those two command lines, and in the latter case you could just manually convert before and after the filter [23:17] <defaultro> k [23:23] <saste> defaultro, looks like you didn't read lutrgb docs [23:23] <defaultro> why? [23:28] <defaultro> saste, can you please tell me where the lutrgb doc is? [23:28] <defaultro> is this it, http://multivac.fatburen.org/localdoc/ffmpeg/ffmpeg.html#lut_002c-lutrgb_00… [23:30] <Xgates> I edited a video from my digital camera with ffmpeg and now I want to trim off the end with this command but the audio doesn't work; [23:30] <Xgates> ffmpeg -i 6.avi -vcodec copy -mp3 copy -ss 00:00:00 -t 00:03:13 trimmed.avi [23:30] <Xgates> is there a one size fits all command for when audio is remuxed how to do this, no matter what audio format? [23:31] <Xgates> errr my bad no -mp3 copy but -acodec copy doesn't work [23:31] <Xgates> this is the output; http://pastebin.ca/2314537 [23:31] <klaxa> that is weird because -c:a copy should work especially with mp3 [23:31] <Xgates> yeah it's not [23:31] <Xgates> hmm [23:32] <klaxa> you might also want to update your ffmpeg [23:32] <Xgates> I compiled ffmpeg-0.11.2 yeah thinking maybe to get the latest 1x the only reason I didn't is because I mainly use WinFF to convert videos and working good with 0.11x [23:33] Action: Xgates checks [23:34] <saste> defaultro, lutrgb=r=1.1*val:g=.9*val:b=1*val [23:34] <saste> this looks mostly right, but 1.1*val is going to generate out-of-range values [23:34] <defaultro> that's what I used [23:34] <saste> which are clipped to 255 IIRC [23:35] <defaultro> it works when I don't add the other parameters [23:35] <defaultro> so float is not allowed? [23:35] <defaultro> but i've been trying non-float, it still generates gray video [23:36] <saste> defaultro, complete command [23:36] <defaultro> one sec, it's running [23:37] <defaultro> for my recent test, I took out lutrgb and used lutyuv, it showed the video [23:37] <defaultro> this one generated grey video, ffmpeg -y -i 00013.MTS -crf 23 -vcodec libx264 -vf lutrgb="r=1*val" -acodec libfaac -ac 2 -preset ultrafast -threads 0 00013.mp4 [23:38] <Xgates> ok compile time for 1.1x [23:39] <saste> defaultro, what about the original source? [23:39] <defaultro> what about the original source? [23:39] <saste> also "lutrgb=r=1*val" is not changing the video [23:40] <defaultro> based from my earlier test, it will increase red channel [23:40] <defaultro> what is your question about original source? [23:41] <saste> defaultro, no, r=1*val is just setting red to the original value, while the other components are kept to the same value [23:41] <saste> that is it is not changing the input at all [23:41] <defaultro> got it [23:41] <defaultro> now I remember why I did 1.1 [23:41] <saste> also if you have gray input, you get gray output [23:41] <defaultro> my original source is full color [23:42] <defaultro> but it has some slight green cast [23:42] <saste> but then i don't know what you're trying to achieve [23:42] <defaultro> saste, check this out, http://www.youtube.com/watch?v=twwWOsWUSIk&list=UUrcz0Vxiv4ZOKFEDScLw5UQ&in… [23:43] <defaultro> that's the one I created last night using ffmpeg. However, the command was very simple. crf was set to default of 23 [23:43] <defaultro> now, I am retrying with more ffmpeg parameters but when I added the same -vf filters, it turned color gray [23:43] <saste> defaultro, what about lutrgb=g=0.5*val [23:44] <saste> but in that case you are pixel intensity [23:44] <defaultro> ok [23:44] <saste> you are *decreasing* [23:44] <defaultro> fyi, this is what I used last night to create that youtube video - ffmpeg -i 00013.MTS -vf 'lutrgb=r=1.1*val:g=.9*val:b=1*val, hue=0:1.3, mp=eq2=1:1.5:0' 00013.mp4 [23:45] <defaultro> now, when I added that same filter to my longer command line, it generated grey video [23:46] <defaultro> this one generated grey video [23:46] <defaultro> ffmpeg -y -i 00013.MTS -crf 15 -vcodec libx264 -s 1920x1080 -r 30000/1001 -vf 'lutrgb=r=1.1*val:g=.9*val:b=1*val, hue=0:1.5, mp=eq2=1:1.5:0' -acodec libfaac -ac 2 -preset slow -threads 0 00013.mp4 [23:46] <saste> not here [23:47] <defaultro> what do you mean "not here"? [23:47] <relaxed> defaultro: you may want to build ffmpeg with libfdk (--enable-libfdk-aac), as it's supposed to be a much better aac encoder. [23:47] <saste> defaultro, that I tried that filterchain and I get no grey video [23:48] <defaultro> ok saste. I don't get grey video if I use the shorter command line [23:48] <defaultro> i get grey video once I use the longer command [23:48] <defaultro> relaxed, I'll retry it later [23:48] <defaultro> i would like to make the filter work [23:49] <defaultro> do you think i have a bad ffmpeg version? [23:50] <defaultro> ffmpeg version N-48389-g341e40f [23:50] <Xgates> ok done, I gues 10 mins. on an i3 not bad :) [23:56] <relaxed> defaultro: latest from git is usually recommended [23:56] <defaultro> k [23:57] <defaultro> when specifying multiple filters, how do you combine them? Do we use comma, space or semicolon? [23:58] <defaultro> i tried two -vf, the latter is only one that gets executed [00:00] --- Sun Feb 17 2013

1 0

ffmpeg-devel.log.20130215
by burek 16 Feb '13

16 Feb '13

[00:16] <cone-893> ffmpeg.git 03Vignesh Venkatasubramanian 07master:003be0a9c33e: Removing network.h from matroskadec.c [00:16] <cone-893> ffmpeg.git 03rogerdpack 07master:12c71f648cbc: dshow: Fix MSVC support, remove av_export, which was apparently unneeded anyway. [00:27] <cone-893> ffmpeg.git 03Michael Niedermayer 07master:8102a097a5ce: doc/APIchanges: List merge commit hashes and version numbers [00:29] <cone-893> ffmpeg.git 03Stefano Sabatini 07master:772b949d8eea: examples/scaling_video: fix typo [01:06] <Paranoialmaniac> michaelni: fiel atom is only in mov. at least, fiel atom is not defined in 14496-12(iso base media)/14 (mp4)/15 (avc), even if fiel atom is present in a non-mov file, reader shall ignore it [01:09] <cone-893> ffmpeg.git 03Stefano Sabatini 07master:7ac3ccc5f238: lavfi/unsharp: use the same macros used in the original MP filter [01:59] <michaelni> Paranoialmaniac, thanks! [02:07] <Paranoialmaniac> undefined atoms under a certain brand shall be ignored by reader. they are just junk under that brand [04:03] <cone-893> ffmpeg.git 03Michael Niedermayer 07master:4a6fa7963bd2: lavf: dont try to find fps for attached pictures [08:40] <cone-466> ffmpeg.git 03Cyrille Faucheux 07master:75758f84def2: build: fixes a "can't cd to..." issue when installing shared libraries. [09:07] Action: ubitux still trying to figure out what Nicolas really has in mind about the sub_charenc mode thing [12:16] <cone-241> ffmpeg.git 03Diego Biurrun 07master:ab441e20ffdf: avutil: Move emms code to x86-specific header [12:16] <cone-241> ffmpeg.git 03Diego Biurrun 07master:759a3a217783: configure: Move MinGW CPPFLAGS setting to libc section, where it belongs [12:16] <cone-241> ffmpeg.git 03Diego Biurrun 07master:4db96649ca70: avutil: Ensure that emms_c is always defined, even on non-x86 [12:16] <cone-241> ffmpeg.git 03Michael Niedermayer 07master:61fbb4cd57d7: Merge commit '4db96649ca700db563d9da4ebe70bf9fc4c7a6ba' [12:27] <cone-241> ffmpeg.git 03Diego Biurrun 07master:49fe280753e0: h264idct: Replace duplicate scan8 table by appropriate #include [12:27] <cone-241> ffmpeg.git 03Michael Niedermayer 07master:048ecbd3f833: Merge commit '49fe280753e0f167ac3d9f227f0c0f7744501fc1' [12:34] <cone-241> ffmpeg.git 03Diego Biurrun 07master:3594554a064d: sparc: dsputil: Simplify high_bit_depth checks [12:34] <cone-241> ffmpeg.git 03Michael Niedermayer 07master:f98598942f7f: Merge remote-tracking branch 'qatar/master' [13:24] <Shiz> >archlinux [13:24] <Shiz> no wonder it failed [13:31] <kode54> hah [13:53] <ubitux> Shiz: libav moved some stuff again, no wonder it failed [16:02] <Shiz> ubitux: you're not libav though, are you :-) [16:15] <ubitux> kierank: give me an hour [16:15] <kierank> no problem [16:16] <ubitux> sorry for the delay, i'm trying to catch up with all i missed this month :) [16:16] <ubitux> kierank: btw, you're using that filter? [16:16] <Compn> Shiz : ffmpeg merges from libav, including their bugs :D [16:16] <kierank> i am investigating using it [16:16] <kierank> ubitux: the biggest issue of course is getting the amount to normalise by to volume [16:17] <kierank> but i think that'll be doable in the latest libav merge [16:17] <ubitux> can't you just use af volume as a second pass? [16:17] <ubitux> doing a simple substraction [16:17] <kierank> i am filtering in realtime [16:17] <Shiz> Compn: I'm aware, it's just kind of hypocitic to bitch about libav if you merge their changes without apparently checking if it works for your own repo :) [16:18] <ubitux> Shiz: this fate instance has for very particular purpose to test such thing [16:19] <Shiz> ic [16:19] <kierank> ubitux: do you know where it says what length you are meant to use per "time slice" of audio in the documents [16:19] <Compn> libav has their own fate system too, http://fate.libav.org [16:19] <ubitux> Shiz: bitching about the distro wasn't innocent either, it's a bit hypocrit to critic such behaviour [16:19] <Compn> oh burn [16:19] <ubitux> kierank: the 100ms and 3s? [16:20] <ubitux> erm 400ms [16:20] <kierank> are those segments the amounts of time that you then pass to the volume filter? [16:20] <kierank> and normalise separately [16:20] <ubitux> i don't know about the real time normalization [16:21] <ubitux> i've really spent only 2 weeks on this :D [16:21] <kierank> that's the part i don't quite understand about r128 [16:21] <kierank> r128 or atsc a/85 is going to be mandatory in some countries for realtime encoding [16:21] <ubitux> i'd say adjust volume in 400ms windows [16:22] <ubitux> based on the overall loudness [16:22] <kierank> yeah [16:22] <ubitux> (integrated) [16:23] <ubitux> or maybe using the other loudness [16:23] <ubitux> to get something smoother [16:26] <mateo`> hi guys ! [16:27] <kierank> hello [16:30] <mateo`> is there a way to force the number of threads of a decoder from a demuxer ? [16:30] <mateo`> i'm trying to solve this ticket https://ffmpeg.org/trac/ffmpeg/ticket/1102 [16:32] <mateo`> the problem i'm facing is that i need to reconstruct a full frame from 2 jpeg2000 frame [16:32] <mateo`> i have a current wip here https://github.com/mbouron/FFmpeg/commits/mxfdec_j2ki but it removes threading support from the libopenjpeg wrapper [16:33] <mateo`> (and i'm not familiar at all with the threading api) [16:38] <mateo`> any hints would be appreciated ;) [16:40] <cone-462> ffmpeg.git 03David A. Sedacca 07master:de21e6736e36: lavfi/ebur128: fix channel weights [16:40] <cone-462> ffmpeg.git 03sedacca(a)comcast.net 07master:b64de24fd7a5: lavfi/ebur128: advance pointer to samples [16:41] <ubitux> erm, i should have checked the authorship [16:41] <TimNich> :P [16:42] <kierank> thanks [16:48] <ubitux> 16:16:59 <+kierank> ubitux: the biggest issue of course is getting the amount to normalise by to volume [16:48] <ubitux> 16:17:12 <+kierank> but i think that'll be doable in the latest libav merge [16:48] <ubitux> what are you refering to? [16:48] <ubitux> lavfi already has metadata injection [16:48] <kierank> elenril said it may be possible with per frame metadata [16:48] <ubitux> we actually can already do that given relatively little effort [16:48] <ubitux> we already have them in buffer ref in lavfi [16:48] <ubitux> so it's the sasme [16:48] <ubitux> same* [16:49] <kierank> well yes but that's going to become AVBuffer or something [16:49] <kierank> to be unified [16:49] <ubitux> currently it's copied at the AVFrame at the end [16:49] <ubitux> the feature you need is already available [16:49] <kierank> ok [16:50] <ubitux> it's already used in two filters iirc [16:50] <ubitux> silencedetect and scenedetect [19:06] <cone-462> ffmpeg.git 03Stefano Sabatini 07master:0018221c03c2: doc: fix reference to ffmpeg-bitstream-filters.html page [19:17] <cone-462> ffmpeg.git 03Hendrik Leppkes 07master:5ad43af9a62c: lavfi/kerndeint: use av_pix_fmt_desc_get instead of directly accessing the table [19:53] <saste> ubitux: i'm benchmarking kerndeint vs. mp=kerndeint and noticing no serious difference (even without the linesize "optimizations") [19:53] <saste> how was you testing it? [19:54] <saste> also, do you have numbers to share? [21:39] <BBB-work> who is "Owner fate" on fate.ffmpeg.org? [21:40] <BBB-work> (I'm looking for the owner of "http://fate.ffmpeg.org/report.cgi?time=20130215202302&slot=x86_64-msvc10-wi…") [21:41] <llogan> damnit roger [21:43] <BBB-work> nevcairiel, Daemon404: is that one of you? [21:44] <nevcairiel> thats mine [21:44] <nevcairiel> why? [21:44] <BBB-work> \o/ [21:44] <BBB-work> does that machine have avx? [21:44] <nevcairiel> yes [21:44] <BBB-work> uhm... [21:44] <BBB-work> ok [21:44] <BBB-work> so why does it pass? [21:45] <BBB-work> I've been finding 2 avx-specific win64 bugs already [21:51] <nevcairiel> i havent seen any issues recently [21:54] <BBB-work> so one issue (patch coming sometime soon) is that h264 loopfilter uses redzone [21:54] <BBB-work> win64 doesn't support that [22:01] <Compn> BBB-work : any chance you can review the dxva2 patch on the list ? [22:01] <Compn> everyone seems to ignore it :\ [22:03] <nevcairiel> because we hate it [23:02] <BBB-work> Compn, dxva2? I don't even know what that is [23:14] <BBB-work> (if you're asking me because it's windows; I have absolutely no love for windows, and don't ever use it; I keep msvs working for chrome :) ) [23:39] <cone-941> ffmpeg.git 03Michael Niedermayer 07master:5e947aeb5945: sws/x86: improve rounding for yuv2yuvX [23:49] <saste> ubitux: ping kerndeint [00:00] --- Sat Feb 16 2013

1 0

ffmpeg.log.20130215
by burek 16 Feb '13

16 Feb '13

[00:10] <tds5016> so if I disable-all, what all would I need to enable for ffmpeg -i fs3 -c:v copy -f mpegts stream3.ts to work? [00:10] <tds5016> ./configure --disable-yasm --disable-all seems not to work :-( [00:12] <tds5016> correction: it works, but it doesn't output an ffmpeg binary anywhere :-( [00:19] <ubitux> --disable-all is what it means& [00:24] <tds5016> I'm now getitng osmething produced, but I'm missing something to make that command work... [00:24] <tds5016> I keep getting invalid input. [00:25] <ubitux> --enable-demuxer=mpegts --enable-protocol=file and stuff like that i guess [00:26] <ubitux> (assuming you disabled everything) [00:26] <ubitux> you'll likely need a few ones like this [00:39] <stephanedev> hi everybody. i would like to stream a live video over rtp without any noticeable latency on a LAN. right now i have a latency ~1s. [00:40] <stephanedev> i am already using some codec options to reduce latency [00:41] <stephanedev> namely (("quality", "realtime"), ("lag-in-frames", "0")) for vp8 and ("tune", "zerolatency") for h264 [00:42] <stephanedev> this reduced the latency from a couple of seconds to around one second [00:42] <stephanedev> even when using the loopback interface this latency happens [00:44] <stephanedev> when streaming directly the output of the vp8 encoder without using rtp (and using my own code instead of letting ffmpeg lib handling the network bits), there is no noticeable latency [00:45] <stephanedev> do you know of any other option in the ffmpeg C API to reduce this latency? [00:46] <stephanedev> is it due to some internal buffering in the muxer or the demuxer? [03:47] <lobotomius> hi ffmpeg! [03:47] <lobotomius> how do i use this unsharpen filter [03:47] <lobotomius> what are the params to sharpen the clip [03:47] <lobotomius> ? tia [06:33] <bigmac> i have a question about a movie collection and is there a way to convert them all to a smaller size streamable on my droid phone... [06:33] <bigmac> maybe one command will work for all different formats, sizes and sizes? [06:34] <bigmac> [700mb.mkv, 3gig.avi, 5gig.flv] [06:34] <bigmac> those would be example source files [06:36] <bigmac> src= mp4 avi mkv [06:36] <bigmac> output= 300mb.flv [06:36] <bigmac> or something [06:38] <bigmac> or 300mb.swf [06:46] <sacarasc> bigmac: Input matters relatively little. [06:46] <sacarasc> Except for length, if you're wanting to have a specific size. [06:49] <bigmac> i can batch script a directory but ultimatly i want to use a single ffmpeg command to get decent quility video i want to embed but download speed is my main concern [06:50] <bigmac> maybe a example command if a embeddable output video? [06:52] <bigmac> can i specify output file size? with width and hight? [06:57] <bigmac> ffmpeg -i INPUT.mp4 -vcodec copy -acodec copy OUTPUT.flv [06:58] <bigmac> thats same quality, im not sure how to loose file size [07:01] <bigmac> -fs 3000 [07:20] <bigmac> i dont know what im doing lol [07:21] <bigmac> how to reduce the output video size for better streaming [10:20] <llogan> bigmac: see https://ffmpeg.org/trac/ffmpeg/wiki/x264EncodingGuide [10:20] <llogan> two-pass if you're targeting specific output file size [11:06] <barnseenio> relaxed: about? [11:09] <relaxed> barnseenio: yes [11:10] <barnseenio> relaxed: pm? [16:17] <captainloc> Hi guys. Im broadcasting a live stream from one place to another and then i'm capturing this stream, cutting it into chunk. It's a mpeg-ts stream with H264 and AAC. I'm saving this stream locally, cutting it based upon a guide. It can appens that my live stream stop and restart but i keep recording and appending in the same file. Then i transcode this file to MP4. But randomly i get problems with Video : audio seems correct and plays but video stop and i [16:22] <haole> hey there... I'm working in a digital TV application in which I need to send a private binary file over a MPEG stream... I need to develop a software to "break" my binary as MPEG sections with a possibly private type/descriptor/whatever... can ffmpeg help me achieving this? [18:13] <alesan> hi [18:13] <alesan> I am recording from three video sources at the same time and audio. So far I have captured the output of each camera, encoded in a different video file, and the audio in a .ogg [18:14] <alesan> but it's a nightmare to synchronize everything [18:14] <alesan> is there a way to record and sync right away multiple video sources? and maybe one or more audio sources too [18:16] <klaxa> maybe synchronize recording devices clocks and add timestamps? :X [18:17] <klaxa> *devices' [18:21] <alesan> klaxa, the device is one and I am reading out three USB webcams [18:22] <klaxa> are you using one ffmpeg instance or are you spawning multiple ones? [18:22] <alesan> adding timestamps... but shouldn't the three videos be in the same file at that point? otherwise how can the timestamps relsato correctly to each other? [18:22] <alesan> right now it's three different processes, each one records on a separate file completely indipendently [18:23] <klaxa> you can put multiple video streams into one file and extract them for later processing [18:23] <alesan> if they are in the same file, I do not want to extract them later :) [18:23] <alesan> how do I put more than one video stream in the same file? [18:23] <klaxa> they would be in the same file [18:24] <klaxa> the matroska container for example supports multiple video streams [18:24] <klaxa> you could extract them for post-editing [18:24] <alesan> I know mkv supports more tracks, that would be awesome but... how to do that? I have searched for the past two days [18:25] <alesan> I see no example at all, only joining several video files that already exist with mkvmerge [18:26] <klaxa> ffmpeg -i <input 0> -i <input 1> -i <input 2> -map 0 -map 1 -map 2 everything_in_one_file.mkv [18:26] <klaxa> i *think* that should work [18:26] <klaxa> i never tried it myself [18:27] <alesan> -map ? let me see what it does [18:28] <klaxa> https://www.ffmpeg.org/ffmpeg.html#Main-options [19:32] <Aktau> Hey guys [19:33] <Aktau> I'm trying to convert an apple h.264 file (1080p) to a decent WebM file (libvpx and libvorbis) [19:33] <Aktau> I can do that just fine with the libvpx-1080p.ffpreset [19:33] <Aktau> But that sets the nitrate fixed [19:33] <Aktau> In this case to 2M [19:34] <Aktau> It looks nice and all but I would like to equal the nitrate of the original, which was a bit more than 16M [19:34] <Aktau> Is there a nice way to tell ffmpeg to take the same nitrate? [19:34] <Aktau> s/nitrate/bitrate/ [19:46] <Aktau> I guess I would like to specify quality and let the bitrate be determined by libvpx [19:46] <Aktau> 't would be nice if anyone with experience with vp8 could enlighten me :) (using 1.1.2 on mac btw, freshly compiled) [20:23] <john_____> hi, I'm using ffmpeg 0.11.2 with android 2.3.3, When I play h264 videos, there's this hiccup effect. With 0.10.8, everything was fine... Any idea ? [20:24] <bahar> john_____ what do you mean by hiccup event? [20:25] <john_____> bahar: the video plays normally, then suddenly, goes back few frames. Then proceeds forward, but jumps back few frame. and so on [20:35] <llogan> john_____: does it occur with ffmpeg from git head? [20:38] <john_____> llogan: tried to compile and run 1.0 but didn't succeed. We sticked to 0.x But to fix the deadlock in libestagefright, we need to use 0.11.1 or 0.11.2 [20:40] <llogan> development is active, and testing git head is always the first step in case the issue has already been resolved [20:44] <john_____> llogan: ok, will try. Is it working with android 2.3.3 ? Does somebody has a recipe to build libstagefright with 1.x ? [20:48] <peper03> Is there any way to get av_read_frame to return MPEG2 packets with 0x1bf (PRIVATE_STREAM_2) as start code? I need them in sync with the normal packets to make improvements to DVD playback but mpegps_read_pes_header() actively filters them out. [20:55] <rcombs> hey, all [20:56] <rcombs> I'm trying to use FileSalvage and dd to recover a file [20:56] <rcombs> it's an MOV [20:56] <rcombs> I've been alternating between "invalid data" and "no MOOV atom found" [20:57] <rcombs> in ffplay, which of those would come first? [20:57] <rcombs> e.g. if I get "no MOOV atom", could it indicate that I'm not extracting enough data? [21:00] <llogan> probably, especially if the moov atom is near the end of the file [21:00] <rcombs> if the file was written by a DSLR, I'd expect the MOOV to be at the end, yeah? [21:03] <llogan> i'm not sure. it's not something i'm really that familiar with [21:04] <llogan> but an uneducated guess would be "yes". [21:11] <rcombs> OK, which would indicate I've got the right start to the file: invalid data, or no MOOV? [21:22] <qubitnik> hello! [21:22] <Fjorgynn> cake? [21:23] <qubitnik> thx! [21:27] <qubitnik> I have an ffmpeg question (suprise, suprise), its about reusing outputs: I want one Single Input to be transcoded in multiple bitrates+resoultions (e.g. 720,480,240) but also pushed to other servers (e.g. rtmp), one quality should be saved to file, and the 3 qualities should be segmented for hls [21:29] <qubitnik> I know how to do all the single Steps, but my Question is: What is the preferred way to provide each ffmpeg process with the already encoded streams, avoiding reencoding, and beeing able to use the streams any time not just at initial start [21:34] <Mavrik_> qubitnik: ffmpeg really doesn't support this use case well [21:34] <Mavrik_> (or at all) at the moment [21:34] <qubitnik> hmm i made some experiments with a rtmp server in the middle [21:34] <Mavrik_> qubitnik: if you have different resultions and bitrates you WILL have to reencode for each output [21:34] <qubitnik> but i am not happy with that [21:37] <qubitnik> yes i know, that is not the problem, but is there a way other ffmpeg processes can grab these outputs and use it again just copying audio+video and only doing remux for pushing,recording,segmenting? [21:37] <llogan> maybe the tee muxer will be useful to you but i don't know if it will work with segment [21:38] <Mavrik_> hmm [21:38] <qubitnik> @llogan: yes, but it is a bit too static for me, using an rtmp server in the middle works, but adds another component+latency [21:39] <Mavrik_> well another rather inefficient option is to just send UDP packets over localloop [21:39] <Mavrik_> have one ffmpeg encode to MPEG2-TS and send to udp://127.0.0.1:<port> [21:39] <Mavrik_> and other ffmpegs read from it and remux [21:39] <Mavrik_> it'll add latency of course [21:40] <Mavrik_> but for anything better you'll just have to pay for a "real" solution like Wowza :) [21:40] <qubitnik> i also experimented with udp://127.0.0.1:5000 Streams and 1 Output is using them, and a second output is a copy to another port [21:41] <qubitnik> i did so because this udp is point-2-point only one ffmpeg instance could use it, multicast is not an option in my scenario [21:41] <Mavrik_> well as I said [21:41] <Mavrik_> Wowza licenses aren't so expensive for pro usage [21:42] <qubitnik> well, if you need one wowza license for each client, it is... [21:43] <qubitnik> what i want to do tooks too much performance, its 1, maximum 2 liveshows per server [21:43] <qubitnik> and they have a per server license [21:43] <Mavrik> I had to develop my own solution based on libav* libs to do just what you want to [21:44] <Mavrik-> everything else is just sub-par :\ [21:45] <qubitnik> yes using the libav* apis is the right way to go, but i am not good in c [21:45] <qubitnik> more into java [21:45] <Mavrik-> mhm, it's a problem then [21:45] <qubitnik> yes [21:45] <qubitnik> and libav* is not so well documented :) [21:45] <Mavrik-> true ^^ [21:47] <qubitnik> what did you wrote with libav? [22:01] <peper03> Is there a better place to ask about the PRIVATE_STREAM_2 issue? I tried on the libav-user mailing list several days ago and got no response there either. Just don't want to waste my (or anyone else's) time asking in the wrong place. [22:02] <Mavrik_> well, you can ask [22:02] <Mavrik_> but I doubt there's much people here that would know more than libav-user list [22:07] <peper03> Ok. Doesn't look hopeful, then. I'm guessing there must be a reason for explicitly filtering those packets out but I'm blown if I can understand why. Thanks anyway. [22:47] <neXyon> hi [22:48] <neXyon> what are the standard codecs (video, audio) for avi files? I know the container supports many codecs, but which ones are used and best supported from devices? [22:49] <neXyon> mpeg4 + mp3? that's what ffmpeg uses by default [22:51] <Mavrik_> something like that [22:51] <Mavrik_> neXyon: what device are you targeting? [22:51] <Mavrik_> usually choosing anything else than avi is a wise idea :) [22:52] <neXyon> Mavrik_: yeah, the problem is, I don't know, that's why I'm gonna take my video in different formats and resolutions with me xD [22:52] <Mavrik_> hmm [22:52] <Mavrik_> mp4 with H.264/AAC is much more widely supported than anything in .avi really [22:54] <neXyon> Mavrik_: yeah, that's the original format anyway [23:15] <hi117> how do i extract a text stream from a video file? [23:19] <hi117> er actually its the skeleton stream from a ogg [23:23] <neXyon> ffmpeg -i short.mp4 -sameq -s 1280x720 short_hd.mp4 => gives an error, is that size not supported? o_O [23:24] <Mavrik_> 1.) never use sameq [23:24] <Mavrik_> 2.) what error? [23:24] <neXyon> Mavrik_: what else? xD [23:24] <neXyon> Error while opening encoder for output stream #0:1 - maybe incorrect parameters such as bit_rate, rate, width or height [23:25] <Mavrik_> look higher [23:29] <neXyon> http://pastie.org/private/tuzgssgnsb8gc34qrypmg [23:29] <neXyon> no clue why it writes none @ audio, and why not use sameq? [23:30] <JEEB> because sameq is "copy quants from source", and hasn't really worked even as such for years [23:30] <sacarasc> -sameq uses the same quantizer as the source, not the same quality. [23:30] <JEEB> yup [23:30] <JEEB> also the reason why your line didn't work is written on line 31 [23:30] <neXyon> ah ok xD [23:32] <neXyon> and how can I tell it to use the same quality? [23:32] <JEEB> long story short, you can't [23:32] <JEEB> you don't have thousands of chinese kids working in your computer [23:32] <JEEB> the closest one has to such is libx264's -crf option [23:32] <JEEB> which is the closest to "constant quality" [23:32] <JEEB> default is 23 [23:33] <JEEB> higher is more compression, possibly worse quality, lower is less compression and possibly better quality [23:33] <JEEB> the idea is to find the highest crf value that still looks good [23:33] <neXyon> ah ok :) thanks [23:33] <JEEB> which is the maximum amount of compression you can get while the video still looks good [23:34] <JEEB> -c:v libx264 -crf 23 -preset medium <- these would be libx264's defaults [23:34] <JEEB> preset is speed vs compression, while crf controls quality level [23:34] <JEEB> http://mewiki.project357.com/wiki/X264_Settings#preset <- all the presets listed [23:34] <JEEB> so yeah, find the highest crf value that still looks good [23:35] <JEEB> and then pick the slowest preset that is still fast enough for you [23:35] <JEEB> and that's it :) [23:35] <neXyon> k thanks :D [23:35] <JEEB> you can use -ss and -t to cut 1-2 minute parts from the middle of a film or something [23:36] <JEEB> for testing :) [23:36] <JEEB> so you don't have to encode the whole thing [23:40] <neXyon> JEEB: thanks, but I don't have enough time to do testing xD [23:40] <neXyon> JEEB: have a very soon deadline :D [00:00] --- Sat Feb 16 2013

1 0

ffmpeg-devel.log.20130214
by burek 15 Feb '13

15 Feb '13

[00:02] <someone-noone> Hey! If I want to reorder packets by pts (decoder is not ffmpeg), how can I know when to "flush" those packets? When new key-frame arrives or when dts==pts? Or may be some other algorithm? [00:04] <nevcairiel> for mpeg2 and vc1 its pretty easy, you basically delay the last I/P frame until the next one arrives, for H.264 thats quite a bit more complex [00:06] <someone-noone> nevcairiel, the first codec I'm working on is exactly h264 [00:06] <nevcairiel> well good luck with that [00:06] <nevcairiel> i wouldn't know how to do it :D [00:06] <someone-noone> :D [00:06] <nevcairiel> you could just store the packets and then simply sort by time [00:07] <nevcairiel> since pts is out of order, you can just re-sort them to give you a linear time [00:07] <nevcairiel> but its meh :) [00:07] <nevcairiel> anyway, sleeping time [00:09] <someone-noone> nevcairiel, pts is in milliseconds time. So& it's not so obvious. However, now I do following: while (pts!=dts || list.size() < 10) wait() [00:09] <someone-noone> But it's not very accurate :) [00:10] <someone-noone> okay, what is a "good time" for this channel? I mean, it's night here too, but I don't know when is it better time to ask [00:11] <someone-noone> not || but && [00:14] <michaelni> someone-noone, if you want to know how h264 pixture reordering works, the authorative source is the itu h264 spec [00:14] <michaelni> its not a nice easy to read doc though [00:15] <someone-noone> anyway thanks :_ [00:43] <cone-580> ffmpeg.git 03Ronald S. Bultje 07master:faf8eca08dcb: h264: remove clear_blocks call in threading init. [00:43] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:701e9b82547c: h264: Use mb itself as memcpy anchor and assert the other anchors position [01:00] <cone-580> ffmpeg.git 03Vignesh Venkatasubramanian 07master:30c5c45b123c: Adding support for parsing BlockAdditional [01:01] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:a08ebf04b1b0: matroskadec: simplify additional_id writing code [03:35] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:9f16cb9e50a5: doc/APIchanges: fix odd .01 versions [06:58] <wm4> so, how am I supposed to use libavformat in presence of sparse video packets (lots of sound packets, but video packets come only after a number of seconds, like mjpeg "slide shows" in mp4 with audio) - I think this is an API weakness, or am I overlooking something? [06:58] <wm4> trying to find the next video packets ends in being flooded with audio packets [06:58] <wm4> ffplay has the same problem [06:59] <wm4> and in presence of the cover art hack, you'd end up demuxing the whole audio file just to find out that there are no more video packets [10:47] <wm4> ffmpeg still fails to inject cover art images when trying to seek before start of the file [10:47] <wm4> or, at least my code just Works on libav [10:48] <wm4> haven't managed to reproduce it with 'ffmpeg' [10:49] <durandal_1707> how you recognise cover art on libav? [10:49] <wm4> I don't think I'm doing anything special here [10:52] <durandal_1707> so what not special are you doing? [10:53] <durandal_1707> do you know why it happens? [10:53] <wm4> not really [10:53] <durandal_1707> also what container? [10:53] <wm4> mp3 [10:54] <wm4> also converting mp3s with cover art to mkv produces strange files [10:54] <wm4> maybe ffmpeg should repeat the picture in the video stream [10:55] <durandal_1707> what "strange files" means? [10:55] <durandal_1707> doesn't ffmpeg disposition for covers and libav doesn't? [10:56] <durandal_1707> lets try again: [10:56] <durandal_1707> ffmpeg have disposition for covers and libav does not? [10:57] <wm4> they both have it set [10:57] <wm4> the AV_DISPOSITION_ATTACHED_PIC flag on the stream [10:58] <wm4> how can I seek to negative timestamps with ffplay? [10:58] <durandal_1707> ok, can we get back with strange mkv files? [10:58] <durandal_1707> wm4: with left/right cursor/arrow keys? [10:59] <wm4> the "strange" thing about the produced mkv file is that it 1. starts at 10 seconds timestamp (?), and 2. nothing can seek in it (ffplay just jumps to the start) [11:00] <durandal_1707> and resulted video stream (covert arts) is not marked as cover art? [11:00] <nevcairiel> many things try to seek to a video frame [11:00] <nevcairiel> idealy, such an attached pic should not result in a video stream in the mkv [11:00] <nevcairiel> but also an attached pic [11:00] <wm4> durandal_1707: no, the produced mkv picture is a "real" video stream [11:00] <durandal_1707> that is bug, please report it [11:01] <durandal_1707> nevcairiel: you are talking to the void [11:01] <wm4> wut [11:06] <wm4> durandal_1707: can I make ffprobe to seek? [11:12] <durandal_1707> wm4: seek in what sense? [11:12] <wm4> durandal_1707: call av[format]_seek_ [11:13] <wm4> before dumping packets with -show_packets [11:15] <durandal_1707> wm4: don't think so, feature request should be reported in usual place [11:16] <durandal_1707> wm4: why you did not tell in bug report [11:17] <durandal_1707> #2270 that disposition flag is not set? [11:17] <wm4> isn't that obvious from the output [11:18] <wm4> isn't that why you ask everyone to post the full output [11:18] <durandal_1707> not for cehoyos, instead you mentioned metadata and attachments which confused him even more [11:19] <nevcairiel> attachments are just metadata [11:19] <nevcairiel> its all in id3 tags for example in mp3 [11:19] <nevcairiel> the disposition flag probably is set, ffmpeg output just doesnt show it, btw [11:19] <wm4> oops [11:20] <durandal_1707> why would muxer create dispostion stream in h264? [11:20] <wm4> it really doesn't [11:20] <nevcairiel> the muxer just creates a video stream [11:20] <nevcairiel> it ignores the disposition, apparently [11:22] <durandal_1707> this doesn't happen with avconv? [11:22] <cone-893> ffmpeg.git 03Diego Biurrun 07master:c6507946d428: dsputil: Move STRIDE_ALIGN macro to the only place it is used [11:22] <cone-893> ffmpeg.git 03Derek Buitenhuis 07master:130cefc9dced: doc/platform: Fix 10l typo [11:22] <cone-893> ffmpeg.git 03Michael Niedermayer 07master:8bbb487e443a: Merge remote-tracking branch 'qatar/master' [11:22] <wm4> durandal_1707: didn't check [11:23] <wm4> durandal_1707: yes avconv does the same [11:24] Action: durandal_1707 is not motiviated to fix their mess rigth now [11:27] <durandal_1707> wm4: about seek, does ffplay show same issue? [11:31] <wm4> durandal_1707: no, ffplay doesn't mind if there's no video packet after a seek, I think [11:33] <durandal_1707> whatever export attachment as stream does not work when seeking [11:40] <durandal_1707> michaelni: michaelni what is supposed output when qp filter is applied before pp one? [11:43] <michaelni> pp filter applies postprocessing depending on qp values, qp filter can change qp, higher qp stronger postprocessing [11:48] <durandal_1707> michaelni: it doesn't work because qcale<->qp_table are not copied? [11:49] <michaelni> yes, dunno if thats the only reason [11:52] <saste> durandal_1707, can you send an updated noise? [11:53] <saste> if you see no review from me within today, then feel free to push it [12:00] <wm4> durandal_1707: I made a craptacular test program: http://pastebin.com/QCFPXdUU [12:01] <wm4> durandal_1707: when I use this on a mp3 with cover art, I never see a packet from stream 1 [12:01] <wm4> durandal_1707: but when changing the negative seek target pts to something positive, it shows up [12:04] <durandal_1707> saste: i want to remove mp=softskip and mp=dsize (and mp=qp, non functional but will port it) [12:05] <durandal_1707> and maybe also ilpack [12:07] <durandal_1707> michaelni: libavcodec/h264.c:1257:30: warning: comparison of distinct pointer types ('CABACContext *' [12:07] <durandal_1707> (aka 'struct CABACContext *') and 'int16_t (*)[512]') [12:07] <durandal_1707> av_assert0(&h->cabac == &h->mb_padding + 1); [12:07] <durandal_1707> ~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~ [12:07] <durandal_1707> ./libavutil/avassert.h:38:11: note: expanded from macro 'av_assert0' [12:07] <durandal_1707> if (!(cond)) { \ [12:07] <durandal_1707> ^ [12:09] <nevcairiel> its meant to compare those pointers [12:18] <cone-893> ffmpeg.git 03Diego Biurrun 07release/1.1:077beee4653f: x86: ac3: Fix HAVE_MMXEXT condition to only refer to external assembly [12:18] <cone-893> ffmpeg.git 03Tim Walker 07release/1.1:5393a5600ddb: mlpdec: set the channel layout. [12:18] <cone-893> ffmpeg.git 03Tim Walker 07release/1.1:59f22ef91a1e: mlpdec: TrueHD: use Libav channel order. [12:18] <cone-893> ffmpeg.git 03Tim Walker 07release/1.1:5af78cc98d80: mlp: store the channel layout for each substream. [12:18] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:e1a86b1433f1: mlpdec: dont leave a invalid huff_lsb in the context. [12:18] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:e67491a2a49a: Merge commit '99ccd2ba10eac2b282c272ad9e75f082123c765a' [12:18] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:1d20d975aa89: Merge commit '3ffcccb4fbaae4d5ad775506f1f2761f2029affa' [12:18] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:9e3e11a348e0: Merge commit '1fd2deedcc6400e08b31566a547a5fac3b38cefb' [12:19] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:6baaaa0174bf: Merge commit '5af78cc98d807f3b43510410dad46e1840c5c99f' into release/1.1 [12:19] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:358e4081edb8: mlp: fix channel order. [12:24] <saste> durandal_1707, i'm not against removing mp filters if useless, but i prefer if it is michael to give the ok, since that's mostly his code [12:24] <saste> ubitux, still willing to optimize kerndeint? [12:25] <nevcairiel> try to figure out why it fails in msvc-dll builds while you're in there :D [12:28] <durandal_1707> michaelni: your opinion on mp=dsize and mp=softskip removal? [12:29] <durandal_1707> saste: michaelni did not write those two filters [12:29] <michaelni> i leave it to saste todecide [12:29] <durandal_1707> maybe thay where useful with mplayer/mencoder but thay are not with ffmpeg [12:30] <durandal_1707> i tried my best to prove they do anything and i failed [12:30] <durandal_1707> unlike mp=qp, which is useless but port could be useful [12:33] <durandal_1707> also my goal is to write more generic filters - they are way more useful [12:35] <cone-893> ffmpeg.git 03Michael Karcher 07release/1.1:901682ff78da: atrac3: use correct loop variable in add_tonal_components() [12:35] <cone-893> ffmpeg.git 03Anton Khirnov 07release/1.1:e0e425042193: dnxhdenc: fix invalid reads in dnxhd_mb_var_thread(). [12:35] <cone-893> ffmpeg.git 03Anton Khirnov 07release/1.1:e835ce83e2ac: vf_delogo: fix an uninitialized read. [12:35] <cone-893> ffmpeg.git 03Anton Khirnov 07release/1.1:5bee21d724dc: vf_delogo: fix copying the input frame. [12:36] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:5a3c8f95d5bf: Merge commit '5bee21d724dc47d115faae3f5065a6db74e1594a' into release/1.1 [12:41] <wm4> hm, so under some condition, libav actually fails to inject a video packet too... actually the seek can sometimes fail [12:43] <durandal_1707> saste: sent new noise [12:46] <saste> durandal_1707, thanks for the noise :) [12:46] <durandal_1707> what is different between unsharp and mp=unsharp? [12:46] <saste> durandal_1707, interface [12:46] <saste> also there is a possible out-of-buffer read in unsharp [12:47] <saste> give me some time and i'll send you a link to the thread [12:47] <durandal_1707> saste: so no reason to keep mp=unsharp? [12:47] <saste> i'd prefer to keep it until unsharp is fixed [12:47] <durandal_1707> what needs fixing? [12:47] <saste> for theoretical security reasons people may prefer to use mp=unsharp [12:48] <durandal_1707> nonsense, it is nowhere documented [12:48] <saste> durandal_1707, i forgot the details, but it is all in the thread [12:50] <durandal_1707> no, i just found thread that you found that output differs [12:51] <durandal_1707> which is WTF, filter should not be commited if it have obvious bug [12:51] <durandal_1707> like earwax [13:00] <durandal_1707> unsharp have strange macro UPSHIFT [13:00] <durandal_1707> *SHIFTUP [13:01] <cone-893> ffmpeg.git 03Anton Khirnov 07release/1.1:00bf66785f7d: fraps: fix off-by one bug for version 1. [13:01] <cone-893> ffmpeg.git 03Anton Khirnov 07release/1.1:7e35c50b81d9: yop: check that extradata is large enough. [13:01] <cone-893> ffmpeg.git 03Anton Khirnov 07release/1.1:108ca6fad1e0: yop: check for input overreads. [13:01] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:81bcf9454e19: Merge commit '108ca6fad1e0e9af8d6337f908bfd23807b7fbd6' into release/1.1 [13:03] <durandal_1707> ok i removing softskip and dsize [13:11] <durandal_1707> what is happening with mplayerhq.hu? I cant read its manual [13:17] <michaelni> elaborate what where doesnt work ? [13:17] <michaelni> manual i mean ? [13:17] <durandal_1707> native unsharp indeed gives different output, maybe because its syntax is broken [13:18] <durandal_1707> michaelni: not acessible, stalled [13:22] <michaelni> i dont know what you mean by "not acessible, stalled" it doesnt timeout it doesnt give http errors [13:23] <michaelni> also ive no time for guessing games [13:23] <michaelni> which URL does what ? [13:25] <durandal_1707> itsn't ffmpeg so why would it be so imporant? [13:25] <durandal_1707> no it works [13:26] <durandal_1707> *now [13:26] <michaelni> if theres a problem with the server it may be important [13:26] <michaelni> also mplayerhq.hu uses a mirror network [13:27] <michaelni> www1.mplayerhq.hu www2, ... [13:30] <iive> meaning you may have been routed to some of the mirrors that is having problems. [13:30] <durandal_1707> lol, unsharp can do some unintentional effect [13:31] <cone-893> ffmpeg.git 03Anton Khirnov 07release/1.1:1f8bf163e4b4: aasc: fix output for msrle compression. [13:31] <cone-893> ffmpeg.git 03Kostya Shishkov 07release/1.1:d0249f1c2e55: qtrle: fix the topmost line for 1bit [13:31] <cone-893> ffmpeg.git 03Anton Khirnov 07release/1.1:b7765d00f911: msrledec: check bounds before constructing a possibly invalid pointer, [13:31] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:7f8846405e8a: Merge commit 'b7765d00f911fe0f8fcda21b93a540f27d2ba2f5' into release/1.1 [13:32] <durandal_1707> saste: actually unsharp gives same output as mp=unsharp [13:32] <saste> durandal_1707, yes [13:33] <saste> durandal_1707, but it accepts param values for which it will read oob values [13:33] <durandal_1707> so what should be done? change syntax? [13:33] <saste> the port was ill-done [13:33] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/0.11:470ee0c660ed: h264_refs: Print default in case we are missing a reference. [13:34] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/0.11:ebe645f02b3f: h264: Reset last_pocs in case of reference or frame number inconsistencies [13:34] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.0:9c659b3a678f: h264_refs: Print default in case we are missing a reference. [13:34] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.0:169d84934423: h264: Reset last_pocs in case of reference or frame number inconsistencies [13:34] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:2ac6b573a408: h264: Reset last_pocs in case of reference or frame number inconsistencies [13:34] <saste> since the unsharp syntax is different from mp=unsharp [13:34] <saste> and it accepts different ranges [13:35] <durandal_1707> it can also crash [13:35] <durandal_1707> i want to fix obvious nonsense/useless nonfeatures/misfeatures/bugs/and atc [13:35] <durandal_1707> so what solution you prefer? [13:36] <durandal_1707> keep syntax and add missing checks? [13:45] <durandal_1707> config_props can be called several times in row without uninit? [13:55] <cone-893> ffmpeg.git 03Kostya Shishkov 07release/1.1:5479e08cc44a: xxan: properly handle odd heights. [13:55] <cone-893> ffmpeg.git 03Derek Buitenhuis 07release/1.1:4eede1fca24a: doc/platform: Fix 10l typo [13:55] <cone-893> ffmpeg.git 03Martin Storsjö 07release/1.1:5310da7e83ec: arm: Fall back to runtime cpu feature detection via /proc/cpuinfo [13:55] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:7d3e21762317: Merge remote-tracking branch 'qatar/release/9' into release/1.1 [14:17] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/1.1:71fee2ab1e77: sws: dont write out of array on bigendian [14:19] <cone-893> ffmpeg.git 03Xi Wang 07release/0.8:e163d884ef6c: rtmp: fix multiple broken overflow checks [14:19] <cone-893> ffmpeg.git 03Xi Wang 07release/0.8:b59ee5dcf119: rtmp: fix buffer overflows in ff_amf_tag_contents() [14:19] <cone-893> ffmpeg.git 03Janne Grunau 07release/0.8:801eff785aa1: rv34: error out on size changes with frame threading [14:19] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/0.8:03ddc260668b: indeo5dec: Make sure we have had a valid gop header. [14:19] <cone-893> ffmpeg.git 03Anton Khirnov 07release/0.8:604d72aa0d05: dfa: improve boundary checks in decode_dds1() [14:19] <cone-893> ffmpeg.git 03Anton Khirnov 07release/0.8:440e98574bde: indeo4/5: check empty tile size in decode_mb_info(). [14:19] <cone-893> ffmpeg.git 03Anton Khirnov 07release/0.8:301761792a69: mpeg12: do not decode extradata more than once. [14:19] <cone-893> ffmpeg.git 03Reinhard Tartler 07release/0.8:db5b454c3d20: Update changelog for 0.7.7 release [14:19] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/0.8:e4831bb9a678: huffyuvdec: Check init_vlc() return codes. [14:19] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/0.8:4f91c4564493: huffyuvdec: Skip len==0 cases [14:19] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/0.8:acada70ffbd7: Merge remote-tracking branch 'qatar/release/0.7' into release/0.8 [14:23] <durandal_1707> last time to say no to dsize & softskip removal [14:26] <durandal_1707> cehoyos, bug is not wish [14:46] <durandal_1707> saste: http://ffmpeg.org/pipermail/ffmpeg-devel/2012-February/120640.html <- why this was never applie? [14:49] <cone-893> ffmpeg.git 03Paul B Mahol 07master:968f8acec882: lavfi: remove dsize libmpcodecs wrapper [14:49] <cone-893> ffmpeg.git 03Paul B Mahol 07master:41ae43cade37: lavfi: remove softskip libmpcodecs wrapper [14:56] <saste> durandal_1707, i was not anymore sure it was a good idea [14:56] <durandal_1707> why? [14:56] <durandal_1707> please explain [14:57] <saste> maybe we should keep separate options [14:57] <saste> and use shorthands to keep syntax compatibility [14:58] <saste> having an ad-hoc specification language for the params is a bit inconsistent [14:59] <durandal_1707> saste: so the only reason why you prefer to keep mp=unsharp is because native one crash? [14:59] <saste> especially now that we have shorthands and options [14:59] <saste> does it crash? [14:59] <durandal_1707> yes [15:00] <saste> durandal_1707, as i explained, i was not confortable with the other patch which was limiting the ranges [15:00] <saste> because i couldn't tell why those ranges were choosen [15:00] <saste> and i'm sorry that i've being busy but i'd still would like to update my unsharp patchset [15:02] <durandal_1707> sorry but too much time passed [15:03] <durandal_1707> mp=unsharp clip them somehow [15:04] <ubitux> saste: not soon [15:04] <ubitux> got some other priorities somehow now [15:12] <durandal_1707> saste: mp=unsharp set blocks size to odd values between [3, 63] [15:12] <saste> durandal_1707, missing micro bumps [15:13] <saste> durandal_1707, give me some time to work on unsharp, now i gonna leave [15:14] <cone-893> ffmpeg.git 03Carl Eugen Hoyos 07master:7d0e3b197c81: Write the fiel atom to mov files independently of the used video coded. [15:14] <cone-893> ffmpeg.git 03Carl Eugen Hoyos 07release/0.10:7a21b089c275: sws: dont write out of array on bigendian [15:14] <cone-893> ffmpeg.git 03Carl Eugen Hoyos 07release/0.11:2f98537ea074: sws: dont write out of array on bigendian [15:14] <cone-893> ffmpeg.git 03Carl Eugen Hoyos 07release/1.0:140ca8f6c035: sws: dont write out of array on bigendian [15:19] <durandal_1707> this is unacceptable [15:30] <cone-893> ffmpeg.git 03Kostya Shishkov 07release/0.10:a94f789c334c: indeo3: initialise pixel planes on allocation [15:30] <cone-893> ffmpeg.git 03Luca Barbato 07release/0.10:1076ea8115ad: mp3: exit on parsing error in mp_decode_frame [15:30] <cone-893> ffmpeg.git 03Anton Khirnov 07release/0.10:a4c9260e6914: pthread: set the frame properties from the thread context, not user. [15:30] <cone-893> ffmpeg.git 03Xi Wang 07release/0.10:165f783235a0: rtpenc: fix overflow checking in avc_mp4_find_startcode() [15:30] <cone-893> ffmpeg.git 03Xi Wang 07release/0.10:69b3fedc09d9: rtmp: fix multiple broken overflow checks [15:30] <cone-893> ffmpeg.git 03Xi Wang 07release/0.10:ef953f760ef1: rtmp: fix buffer overflows in ff_amf_tag_contents() [15:30] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/0.10:ba4b57e8024a: huffyuvdec: Check init_vlc() return codes. [15:30] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/0.10:b07c79125270: huffyuvdec: Skip len==0 cases [15:30] <cone-893> ffmpeg.git 03Michael Niedermayer 07release/0.10:5b7f7f3809a8: Merge remote-tracking branch 'qatar/release/0.8' into release/0.10 [15:53] <cone-893> ffmpeg.git 03Carl Eugen Hoyos 07release/0.10:83446017128b: Write the fiel atom to mov files independently of the used video coded. [15:53] <cone-893> ffmpeg.git 03Carl Eugen Hoyos 07release/0.11:ec18baadfa17: Write the fiel atom to mov files independently of the used video coded. [15:53] <cone-893> ffmpeg.git 03Carl Eugen Hoyos 07release/1.0:397e769f74dd: Write the fiel atom to mov files independently of the used video coded. [15:53] <cone-893> ffmpeg.git 03Carl Eugen Hoyos 07release/1.1:057051b84879: Write the fiel atom to mov files independently of the used video coded. [16:19] <michaelni> Paranoialmaniac, is there a fiel atom in mp4 like in mov ? [16:25] <TimNich> I don'y have the mp4 spec to hand but Quicktime Pro writes them to .mp4 files [16:25] <JEEB> not like QuickTime really cares about the specs [16:26] <michaelni> well iam asking because carls commit makes ffmpeg write them into mp4 too now [16:26] <nevcairiel> i probably have the iso spec somewhere, but i hate reading it :D [16:26] <JEEB> the ISO base media format spec is freely available [16:26] <JEEB> the ones that are not are the short MP4 file format spec [16:26] <JEEB> and the -15 AVC-in-"mp4" spec [16:27] <JEEB> (AVC file format) [16:27] <TimNich> Yes, but I tend to have qtff to hand at all times, but not the part 10 specs [16:27] <nevcairiel> -12 should be the isom [16:28] <JEEB> oh right, 12 [16:29] <nevcairiel> where in the hierarchy is fiel? [16:29] <JEEB> http://www.iso.org/iso/home/store/catalogue_tc/catalogue_detail.htm?csnumbe… <- click the ITTF link [16:29] <JEEB> for a free copy of -12 [16:39] <cone-893> ffmpeg.git 03Nicolas George 07master:bb3303b94c1c: lavfi/vf_ass: ignore subtitles decoding errors. [16:39] <cone-893> ffmpeg.git 03Nicolas George 07master:c557a5b08efe: lavfi/vf_ass: reindent after last commit. [16:44] <michaelni> TimNich, QT pro writes fiel also into h264 in mp4 files ? [16:47] <TimNich> not for H264, but for some other codecs..\ [16:50] <JEEB> also stuff becomes even more fun because "mp4" is not a single entity [16:50] <JEEB> given the written brand [16:51] <TimNich> ISO/IEC 14496-12:2012(E) doesn't specifically mention the 'fiel' atom, but that doesn't necessarily mean its disallowed.. [16:51] <JEEB> various brands have various feature sets [16:51] <JEEB> f.ex. QT does have some brands of its own [16:51] <JEEB> which then follow qtff [16:51] <JEEB> just like DTS and Dolby have their own, too [16:51] <JEEB> "oh ze joy" [16:52] Action: michaelni fails to find a single mp4 with fiel except what ffmpeg generates now [16:54] <TimNich> If I put proves in mp4 with QTPro I get a fiel atom [16:55] <JEEB> check the brands written [16:55] <TimNich> s/proves/prores/ [16:57] <TimNich> QTPro brand QT compatible QT&.. [16:57] <nevcairiel> why would something like h264 get a fiel atom anyway? [16:58] <TimNich> quite, why should it need it? but then why should Prores need it. they all have the infer in the stream parameters anyway... [16:58] <nevcairiel> i have no idea about prores, just know that h264 has it :D [16:59] <michaelni> so write fiel if mov || prores ? [16:59] <TimNich> nevcairiel: btw the file is in trak->mdia->stbl->stsd [16:59] <nevcairiel> there are a lot more formats that will probably get it [17:00] <TimNich> I think the matrix is more complicated than that [17:01] <michaelni> what about write if !h264 && !mpeg4 [17:02] <michaelni> mpeg4-asp video that is [17:04] <michaelni> id like to avoid it for at least things like ipod and other hw player specific files [17:05] <michaelni> i doubt it will do any good to them ... [17:06] <TimNich> The problem I see is that for .mov it is only mandated for raw video *but* included in other formats that shouldn't need it (like Prores) while similar formats, like DNX don't use it. The big question is does its presence do any harm? It should be silently ignored by applications that do not want or expect it, but can we garuantee that? [17:07] <nevcairiel> sadly plenty applications arent smart enough to simply ignore unknown data, but tend to explode [17:09] <michaelni> also theres the case that what fiel says differs from what the codec stream says where it can store it like h264 [17:10] <TimNich> I think thats important and afaik untested, which is why I wanted to do extensive testing on DX before being completely happy. [17:12] <TimNich> If the section of mov.c with all those 'if else' was turned into a set of 'case' sections it would be easier to fine tune, and only add in codecs as needed. [17:22] <durandal_1707> nobody working on bitexact dirac? [17:24] <kierank> dirac is largely dead [17:24] <durandal_1707> as codec? [17:25] <kierank> yes [17:25] <kierank> jpeg2k is all the rage [17:26] <TimNich> this week at least.... [17:28] <durandal_1707> someone from libav is working on fully working native decoder, i think i had cotact with repo but i forgot where it was [17:28] <michaelni> TimNich, dnxhd has fiel or no fiel ? [17:30] Action: Compn tried to tell people about reinventing dirac a) before it was finished and b) before there were even samples / codec was used [17:30] <Compn> no one listens to me :P [17:30] <TimNich> dnxhd has no fiel in my "genuine" samples [17:30] <michaelni> ok, thx [17:31] <TimNich> got to go now back tomorrow [17:33] <durandal_1707> vf_transpose have interesting way to detect number of data[planes] [17:50] <cone-893> ffmpeg.git 03Michael Niedermayer 07master:405cc0d90520: movenc: hotfix, dont store fiel for h264 / mpeg4-asp / dnxhd [17:50] <cone-893> ffmpeg.git 03James Almer 07master:8c95d177dfa9: fate: Add encrypted tta stream test [19:11] <durandal_1707> i cant use ffplay with -filter_complex? [19:25] <llogan> durandal_1707: "Set noise seed for pixel component or all of them" <- this phrase doesn't make sense to me. what is "them"? [19:27] <durandal_1707> all pixel components [19:32] <llogan> "Set noise seed for specific pixel component of all pixel components" or something would be more clear to lusers like me [19:32] <llogan> s/of/or [19:32] <iive> how about "set noise seed for one or all pixel components" ? [19:34] <llogan> "set noise seed for all-one, all-one, all-one: all-one or none" (can you guess the reference?) [19:38] <iive> nope [19:46] <durandal_1707> i think i will add float and double pixel format [19:52] <llogan> iive: some weird-ass hippy soap in the US. Dr. Bronners. [19:54] <iive> oh, i guess they haven't ported it to EU market yet. [19:56] <ubitux> durandal_1707: i've been away for a while and saw a lot of activity from you, are you waiting for some particular review? [20:11] <durandal_1707> ubitux: yes, unsharp, kerndeint and tinterlace removal from mp [20:17] <durandal_1707> ubitux: there is also trivial overlay patch that adds yuv444p but it somehow is broken with ffplay, because it inject strange conversion in middle of chain, breaking chroma values [20:18] <durandal_1707> dunno if that is lavfi bug [20:22] <ubitux> michaelni: do you expect someone to check all the hash or you're waiting for more general reviews? [20:22] <ubitux> durandal_1707: about unsharp and tinterlace, dunno [20:22] <ubitux> about kerndeint i had a relative performance concern, which i wasn't really able to understand [20:23] <ubitux> i have no idea when i'll have time to look again at it [20:23] <ubitux> i'm not sure to be the best person for the yuv444p thing; i think i've actually already expected a similar problem with ffplay [20:28] <michaelni> ubitux, dunno, i had no specific expectattions [20:28] <ubitux> sorry i'm not motivated to check every single hash :)) [20:29] <michaelni> :) [20:29] <michaelni> they should be ok [20:29] <michaelni> version numbers are trickier to lookup so higher chance that theres a wrong one in it [20:30] <michaelni> i mean version numbers are not unique before we added .100 [20:30] <michaelni> also there are typos in the file [20:36] <cone-893> ffmpeg.git 03Michael Niedermayer 07master:ad6802f975a9: apichanges: fix date [20:36] <cone-893> ffmpeg.git 03Michael Niedermayer 07master:33d6330652c0: apichanges: Use , instead of / to seperate multiple hashes [20:55] <durandal_1707> ubitux: also you may review noise [21:42] <cone-893> ffmpeg.git 03Michael Niedermayer 07master:2f3bc5122822: apichanges: fix 2 wrong hashes [21:43] <cone-893> ffmpeg.git 03Vignesh Venkatasubramanian 07master:ce6a8e5947ca: Adding AlphaMode element to Matroska Parser [21:49] <ubitux> durandal_1707: ok, i'm having a look [22:16] <ubitux> michaelni: i won't make any more review for APIChanges, so you can consider it a LGTM from me [22:51] <durandal_1707> ubitux: filters that use filter_complex need PRESERVE? [22:53] <ubitux> filters that use filter_complex? [22:53] <ubitux> what does that mean? [22:54] <ubitux> i'm not familiar with the preserve thing btw [22:54] <durandal_1707> filter that operates on > 1 buffers [22:54] <durandal_1707> one need to queue such buffs [22:55] <durandal_1707> i encountered problem with my filter if i use 2 images as inputs i will get nothing as output [22:56] <durandal_1707> may not be related to it at all [22:56] <ubitux> that's what i was going to say [22:58] <ubitux> * Filters that intend to keep a reference after the filtering process [22:58] <ubitux> is finished (after filter_frame returns) must have the PRESERVE [22:58] <ubitux> permission on it and remove the WRITE permission if they create a new [22:58] <ubitux> reference to give it away. [22:58] <ubitux> next point might be related too [23:02] <ubitux> divVerent: ping [23:11] <cone-893> ffmpeg.git 03Clément BSsch 07master:cf8dec7d6421: lavfi/showspectrum: pretty-align constants. [23:11] <cone-893> ffmpeg.git 03Clément BSsch 07master:022437518086: lavfi/showspectrum: simplify intensity_color_table declaration. [23:47] <cone-893> ffmpeg.git 03Clément BSsch 07master:35a995f4518a: doc/resampler: fix two typo. [23:47] <cone-893> ffmpeg.git 03Clément BSsch 07master:d5ce725cb32b: Fix a few "its" vs "it's" typo. [23:59] <cone-893> ffmpeg.git 03Michael Niedermayer 07master:c3fb20bab4f0: ffmpeg: Check for parameter changes at the output of the audio filter graph [00:00] --- Fri Feb 15 2013

1 0

ffmpeg.log.20130214
by burek 15 Feb '13

15 Feb '13

[00:32] <meelu> is there a list of codecs ffmpeg supports, where is allows libmp3 conversion [00:32] <meelu> from video [00:32] <meelu> i mean is there a list of video codecs it supports where it can convert to mp3 [00:33] <llogan> i don't quite understand your question, but did you see "ffmpeg -codecs"? [00:33] <meelu> tried ffmpeg -formats, is this it [00:34] <meelu> i want to know if i can use "-y -acodec libmp3lame -ab " on all llogan [00:34] <llogan> you want to know what formats support mp3 audio? [00:35] <llogan> i meant "containers", not "formats" [00:35] <meelu> i want to know which type of videos i can run this command on [00:35] <meelu> %ffmpeg% -i %tempFile% -vol %volume% -y -acodec libmp3lame -ab %quality%k %newFile% [00:35] <llogan> you can run that on any input file [00:35] <meelu> ah great [00:35] <meelu> so i can do that with anything [00:36] <llogan> although consider using -q:a instead of -ab [00:36] <meelu> im going to make something using php to verify a file is supported when a user uploads it [00:36] <meelu> why? [00:36] <llogan> https://ffmpeg.org/trac/ffmpeg/wiki/Encoding%20VBR%20%28Variable%20Bit%20Ra… [00:42] <tseth> hi. can someone help me with this error? [00:42] <tseth> h264_mp4toannexb failed for stream 0, codec copy: Invalid argume [00:42] <tseth> argument* [00:42] <tseth> ffmpeg -i fs0 -vcodec copy -acodec copy -vbsf h264_mp4toannexb -f mpegts ts_stream0.ts [00:42] <tseth> that is my command... fs0 is a rawdump of an h264 stream nals [00:50] <tseth> better question... does anoyne know how to convert raw h.264 over to mpeg2-ts [00:50] <tseth> ? [00:52] <cps> Hi, I'm working on the open source animation program Pencil2D. I'm trying to get file export working using ffmpeg. [00:53] <cps> As a test, I'm trying to encode a vector of RGB image data (derived from a png). [00:54] <cps> I have a couple of questions about the decoding_encoding.c example. [00:54] <cps> Firstly, is this the right place to ask? [00:55] <klaxa> cps: if you are using ffmpeg and not the fork (avconv) then yeah probably [00:56] <cps> Great! [00:58] <cps> The example uses "AV_PIX_FMT_YUV420P". What I'm not clear about is how to convert and then write the image data to the AVFrame [01:00] <cps> Am I right in thinking that I need to convert the RGB data to YCbCr for each pixel, and then average the YCbCr data for each quad? [01:06] <tseth> so I basically realized what's happening. I'm trying to put an mpeg2-ts container around an h264 payload... when I run this command: ffmpeg -i fs0 -acodec copy -vbsf h264_mp4toannexb -f mpegts stream0.ts the 10 second stream pops out a 00:00:0.4 seconds long. [01:06] <tseth> any idea what I'm doing wrong? [01:16] <tseth> can someone please help me out? I'm getting incredibly frustrated? [01:18] <tseth> llogan: http://pastie.org/6158686 [01:18] <tseth> the video clip is 10 seconds long. [01:19] <tseth> it's a raw h264 dump, all I'm trying to do, is to convert it to an mpeg2-ts format that safari will play nice with. [01:19] <tseth> if anyone has any idea how to help, it would really mean the world to me. [01:21] <llogan> ask in #libav. you're not using ffmpeg from FFmpeg. [01:21] <llogan> we do not support third party tools here. [01:22] <tseth> llogan: am I able to do this with ffmpeg? [01:23] <llogan> you're not using ffmpeg. [01:23] <tseth> damn... I just used the apt-get for ffmpeg.... [01:23] <tseth> how can I get ffmpeg? [01:23] <llogan> compile or use a static build [01:23] <llogan> the "ffmpeg" is a result of others misrepresenting the name ffmpeg. [01:24] <llogan> *the "ffmpeg" in the ubuntu repo [01:24] <tseth> does ubuntu not have an ffmpeg. [01:24] <tseth> and is libav essentially the same thing? (honest question... I have no idea.) [01:24] <llogan> no. it's a fork. [01:24] <llogan> see http://stackoverflow.com/a/9477756/1109017 [01:24] <llogan> and http://blog.pkh.me/p/13-the-ffmpeg-libav-situation.html [01:25] <llogan> ubuntu no longer provides ffmpeg from FFmpeg, but a fake "bizarro" version from libav. [01:25] <tseth> why would they do that? [01:26] <tseth> Would the real ffmpeg solve the issue I'm bumping into here? [01:27] <llogan> because the debian/ubuntu maintainer is a member of the fork and they want users to switch to libav (while allowing users to bleieve they are using ffmpeg from FFmpeg). [01:27] <llogan> i'm not sure if ffmpeg will help, but we simply dont support fake ffmpeg here. [01:28] <tseth> If this were a real version of ffmpeg, what would you suggest to do this? [01:29] <tseth> libav looks dead, and I'm frustrated, and you're the only one who seemed to be willing to help :-( [01:29] <tseth> #libav* [01:29] <llogan> i suggest trying it with a static build from the link i provided. it will be simple to try. [01:29] <llogan> i've never been in that channel. i don't know if they often try to help users. [01:30] <tseth> will downloading that ffmpeg break the current one? [01:31] <klaxa> tseth: no, you can put it in any directory you want and execute it from there [01:31] <llogan> no. you simply download the archive, extract the archive, and run the binary [01:31] <klaxa> no installation needed whatsoever [01:34] <llogan> do you mean to be re-encoding the video? also your input has no audio, so -acodec copy is superfluous. [01:35] <tseth> oh wow. looks like this version worked. [01:41] <tseth> llogan: you are correct. [01:42] <tseth> any idea why what I'm doing wouldn't work with http live streaming? it half worked :-) [01:42] <klaxa> that is often the case :) [01:42] <tseth> it starts the video, plays through my section, then craps out. [01:42] <tseth> :-) [01:42] <llogan> it's what I call YALB...yet another libav bug [01:43] <llogan> "it" as in safari? [01:48] <tseth> ah, gotchya. [01:48] <tseth> hahaha. [01:48] <tseth> yeah, I think I tracked the issue down :-) [01:48] <tseth> one last question... what is the licensing on ffmpeg? [01:48] <tseth> (hoping for mit or berkeley, expecting gpl v2) [01:53] <klaxa> i think most is lgpl [01:53] <klaxa> http://ffmpeg.org/legal.html [02:06] <skered-> I have an mp3 file that only has right audio, can ffmpeg make an audio file that has the right channel for both left/right (or a mono file using the right side)? [02:08] <tseth> lgpl = more restrictive, no? [02:21] <base698> Hi All! I have this command: ffmpeg -i pipe:0 -acodec libvorbis test.ogg and it works great. Just wanted to know if you can do something like this: ffmpeg -i pipe:0 -acodec libvorbis pipe:1 > somefile.ogg [02:21] <base698> I had to rebuild ffmpeg for ogg support and in doing that it changed the man page, which did have a huge section about pipes [02:25] <klaxa> base698: you probably have to specify a container format with -f [02:25] <klaxa> for example -f ogg [02:25] <llogan> base698: man pages have been segmented. see the end of "man ffmpeg". [02:25] <llogan> such as: man ffmpeg-protocols [02:25] <base698> holy shit success! [02:26] <base698> and thanks even more for the man info [02:26] <base698> that's the man page I remember [02:47] <iive> tseth: lgpl is less strict than gpl. [03:10] <defaultro> so happy, -vf hue=30 fixed the greencast from my videos :) [03:16] <tseth> new question... [03:16] <tseth> is it possible to use ffmpeg to resize a video? [03:17] <klaxa> yes [03:17] <klaxa> see -s [03:17] <tseth> what would the options be for doing such said black magic? [03:17] <klaxa> i think it should also be possible to use -vf scale [03:17] <tseth> so if I wanted to reencode to 720p, how would I do that? [03:18] <klaxa> ffmpeg -i <input> <input options> -s 1280x720 output.container [03:18] <klaxa> i think that should work [03:18] <klaxa> i think you can put -s 1280x720 even before <input options> but not before -i <input> [03:19] <tseth> hmm, looks like this should already be done :-/. [03:19] <tseth> for some reason the ts I cvreate doesn't work on the ipad. [03:19] <klaxa> it's also rather output options really [03:19] <klaxa> not input options, silly me [03:19] <tseth> it does work on the safari desktop thoguh. [03:19] <tseth> for live streaming. [03:19] <tseth> I know this isn't your area, but.... [03:20] <tseth> ffmpeg -i fs0 -vbsf h264_mp4toannexb stream0.ts am I missing some options to make this a valid http live stream? [03:20] <tseth> for the ipad* [03:20] <klaxa> heh, no idea [03:20] <klaxa> i have no ipad and everything i stream via http is done with matroska [03:21] <llogan> -vf scale is more flexible than -s and has the advantage of working easier with other filters in a filterchain [03:21] <llogan> such as -vf scale=1280:-1 if you want it to automatically figure out the height to preserve aspect [03:22] <tseth> gotchya. [03:22] <tseth> hmm, that's really weird. [03:22] <tseth> haha. [03:24] <llogan> i'm assuming your command is re-encoding H.264 to mpeg2 video, so -vbsf h264_mp4toannexb is not needed if that is what you want [03:25] <tseth> llogan: I don't think I want to be reencoding either. I think what I want is just to reencapsulate? [03:26] <tseth> unless I'm completely misunderstanding. [03:26] <llogan> add "-codec copy" as an output option with "-bsf:v h264_mp4toannexb" [03:30] <tseth> this may sound like a really strange question... [03:30] <tseth> but does ffmpeg use cron at all? [03:35] <llogan> you can invoke just about anything with cron [03:38] <tseth> you're right. ffmpeg is the only thing I was running. I ended up with about 100 copies of it. [03:38] <tseth> haha. [03:38] <tseth> there's something screwed with that server.... [03:39] <tseth> Failed to open bitstream filter h264_mp4toannexb for stream 0 with codec copy: Invalid argument [03:39] <tseth> llogan: I'm getting that :-) [03:40] <tseth> http://pastie.org/6159610 [03:41] <tseth> llogan: that's the output :-) [03:51] <llogan> tds5016: i don't know why it's doing that and I can't duplicate it [03:51] <tds5016> any ideas llogan ? [03:51] <tds5016> no worries. Sorry. :-/. I always get lucky with this. [03:52] <tds5016> This problem is so stupidly obnoxious... haha. [03:53] <llogan> ask on ffmpeg-user mailing list if you don't figure it out. make sure to include your complete ffmpeg command and the complete console output (you can omit the repeated "failed" messages but leave at least one) [03:54] <tds5016> could I be missing timecode info? [03:57] <tds5016> Stream format for the stream segmenter must be MPEG elementary audio and video streams, wrapped in an MPEG-2 transport stream, and using the following encoding. [03:57] <tds5016> hmmmm..... [03:57] <tds5016> how do I check if I've already got an elementary video stream? [04:00] <tds5016> sorry if that's a stupid question... I'm just triyng ot figure all of this out :-/. [04:14] <tds5016> is there a differennce between baseline and elementary? [04:18] <relaxed> elementary stream means a raw h264 file. (not in a container) [04:18] <tds5016> I thinkt hat's what I have :-/. how could I tell? [04:19] <relaxed> run ffmpeg -i file [04:19] <relaxed> look at input format [04:20] <relaxed> For example: Input #0, h264, from 'fs0': [04:20] <tds5016> http://pastie.org/6159849 [04:20] <tds5016> also, btw...... pastie.com and pastie.org... NOT the same thing ;-) [04:20] <relaxed> so, yes, your input is a h264 elementary stream [04:21] <tds5016> then I have no idea why in the hell mobile safari is choking on it :- [04:21] <tds5016> :-( [04:21] <tds5016> http://pastie.org/6159854 [04:22] <tds5016> is that the right output for what should be an mpeg2-ts with a payload of an elementary h264 stream? [04:22] <relaxed> your video is now mpeg2, is that what you want? [04:22] <relaxed> no [04:22] <tds5016> I want h264 wrapped in mpeg2-ts [04:22] <tds5016> relaxed: can you help me figure out how I'd do that? [04:23] <relaxed> "ffmpeg -i fs0 -c:v copy -f mpegts output.ts" [04:23] <tds5016> http://pastie.org/6159861 [04:24] <relaxed> NO, run what I just gave you. [04:24] <relaxed> -vbsf h264_mp4toannexb is only needed when demuxing h264 from the mp4 container. [04:25] <relaxed> or stream copying [04:25] <tds5016> hmm, that looked like it broke :-/. [04:25] <tds5016> it only plays for a portion of a second, then breaks. [04:26] <relaxed> pastie.org or it didn't happen [04:26] <tds5016> Failed to open bitstream filter h264_mp4toannexb for stream 0 with codec copy: Invalid argument [04:26] <tds5016> I keep getting that :-( [04:26] <tds5016> sad face. [04:26] <tds5016> haha. [04:26] <tds5016> when I add that section back in. [04:27] <relaxed> I need pastebins and not a personal recount from you. [04:28] <tds5016> http://pastie.org/6159883 [04:28] <tds5016> that's without the mp4toannexb [04:29] <tds5016> for some reason when I get hte file, it only plays a portion of a second of it, then it stops playing. [04:30] <tds5016> oh, it looks like it's a problem with the video tag auto playing.... [04:30] <tds5016> it looks like it gets what it can from http, then just says "screw it, I'm at the end" then dies if it doesn't have the whole file. [04:31] <defaultro> does ffmpeg support denoise? [04:32] <tds5016> relaxed: do you have any idea how I'd be able to fix that? or that's a web problem? [04:35] <tds5016> hmmm.... magically started working. [04:35] <tds5016> you guys rock! [04:35] <tds5016> thank you so much! [04:35] <tds5016> well, partially worked... [04:36] <tds5016> still breaks on ipad in the video tag [04:53] <tds5016> thanks for all of the help guys. you are phenomenal people, all of you! [05:31] <relaxed> defaultro: ffmpeg -filters | grep -i denoise [05:38] <defaultro> thanks relaxed! [05:39] <defaultro> hqdn3d V->V Apply a High Quality 3D Denoiser. [08:13] <willbradley> hey is ffmpeg good for extracting h264 data from inside a large binary? [08:13] <willbradley> google is failing me [08:49] <creep> At the end of the month, Jaimie Alexander was injured on the London film set, after she slipped while walking in the rain. [08:49] <creep> breaking news haha [08:50] <creep> willbradley<< ffmpeg plays it, you do "extracting" with dd for example [08:52] <robotics> Hi everyone! Sorry for interrupting you, but, does anybody know, why while converting avi to mpeg2, one second of the end of video is lost? [08:52] <robotics> *why when... [08:53] <creep> robotics<< hi, no problem, i'll put the chips down while writing then [08:54] <robotics> creep: i'm sorry, i don't know english very well. Could you explain in detail [08:56] <creep> robotics<< you know, when you eat you hold food in your hand you would want to eat and take it to your mouth usually [08:57] <robotics> ok, don't waste your time. [11:10] <jeje34> hello to all :-) [11:21] <jeje34> I have a question about av_init_packet and av_new_packet. In my code, when I have a H264 frame to decode with FFMPEG 1.1.1, I use av_init_packet and av_free_packet. But, I see av_new_packet seems be better because it can realloc if necessary with FF_INPUT_BUFFER_PADDING_SIZE and set the av_destruct pointer. With av_init_packet, I call av_init_packet, I make a temporary buffer sizing it to my [11:21] <jeje34> (H264 frame size+FF_INPUT_BUFFER_PADDING_SIZE), memset it to 0, copy my H264 frame in this one and so affect the AVPacket.data to this buffer and the size to the H264 frmae size. When finish decoding, I use av_free_packet to free the memory. I try to use av_new_packet because I don't have to create my temporary buffer, but I have an error on the av_free_packet (in reality in av-destruct-packet) [11:23] <jeje34> between the av_new_packet and av_free_packet I call the avcodec_decode_video2 on this packet, so the size of this packet after decoding is set to 0 [11:55] <jeje34> another question I have is how to use the av_lock_mgr because in liv AVCodec 5.2 (I know it's old) I had a function called av_lockmgr_addref and I use it doing if (av_lockmgr_addref()==1)av_lockmgr_register(&ff_lockmgr); But now, it seems there's only the function av_lockmgr_register. Do I need to call it at each thread using ffmpeg? [12:50] <Bor0> how can I correctly increase the audio volume of a video that I'm displaying with SDL? I tried something like for (int i=0;i<len1;i++) {... *p *= (is->volume / 100.0); } but increasing amplitude like this will generate noise at a point (and above it) [12:50] <Bor0> are there any native functions for this? [12:54] <Mavrik> there's an audio filter for that [12:54] <Mavrik> btw, volume is always adjusted on logarithmic scale [12:55] <Bor0> can you point me to some examples? can libavfilter do this? I'd rather not do it manually [12:55] <durandal_1707> see volume filter [12:58] <Bor0> ok, and regarding the logarithmic scale thing, if I were to do it manually (out of curiousity), then I'd multiply by log(x), instead of x? 0 < x < 100 [13:00] <Mavrik> http://dsp.stackexchange.com/questions/2990/how-to-change-volume-of-a-pcm-1… [13:00] <Mavrik> this should help a little ;) [13:19] <Bor0> excellent, thanks. [13:27] <funyun_> hi. if i want to encode 2 videos into 1, how would i go about doing that? ffmpeg -i vid1 -i vid2 ? [13:37] <durandal_1707> burek: the nick arg works for !pb only. why? [13:46] <funyun_> saste: here's what i understood from that but it's not working. any idea why? i did this, ffmpeg -i vid1.wmv -i vid2.wmv -filter_complex '[0:0] [0:1] [1:0] [1:1] concat=n=2:v=1:a=1 [v] [a]' -map '[v]' -map '[a]' -vcodec libx264 -preset veryfast -tune film -x264opts level=41:deblock=-2,-2 -crf 18 test.mp4 [13:46] <funyun_> the error i get is Stream specifier ':0' in filtergraph description [0:0] [0:1] [1:0] [1:1] concat=n=2:v=1:a=1 [v] [a] matches no streams. [13:51] <funyun_> anyone know what i'm doing wrong? [14:14] <saste> funyun_, you're missing to link the graph outputs to the output file [14:14] <saste> -map [a] -map [v] -codec:v libx264 ... [14:15] <saste> uhm not really [14:17] <funyun_> saste: http://pastie.org/6163183 [14:21] <saste> funyun_, can't see nothing wrong in your command, and it works here [14:22] <saste> you might try to update your ffmpeg [14:22] <funyun_> alright, i'll try [14:52] <funyun_> saste: now i get "Unrecognized option 'preset'." [16:30] <diroots> hi there [16:31] <jeje34> Hi, is it better to in order:1- avcodec_all_context3(NULL) 2- avcodec_find_decoder(CODEC_IH_H264) to finally 3- avcodec_open2( CodecCtx, Codec, NULL); or is better to directly do 1- Codec = avcodec_find_decoder(CODEC_ID_H264) , 2- avcodec_alloc_context3(Codec) and so 3- avcodec_open2( CodecCtx, Codec, NULL); because the 2 methods works but which is the best??? [16:31] <diroots> I have a source (H.264) video with a Stream #0.2 : Data: unknown (tmcd), do you know wtf this is? [16:32] <diroots> and when I re-encode it to x264, I loose this stream #0.2 is there a way to keep it within the output video ? [18:10] <taylanub> Is $CC being ignored by ffmpeg's build ? I see a line in config.log, but then it doesn't seem to use it. [18:45] <ubitux> taylanub: yes iirc it's ignored [18:45] <ubitux> i sent a patch a long while ago for this [18:45] <ubitux> but it was somehow rejected for a reason like& [18:46] <ubitux> "ppl tend to put crap in it" [18:47] <taylanub> ... [18:48] <ubitux> actually it was at the time of the fork [18:48] <ubitux> http://lists.libav.org/pipermail/libav-devel/2011-April/001454.html the thread died after this [18:49] <relaxed> taylanub: can't you use --cc= [18:52] <taylanub> relaxed: I can, but of course I'd prefer it to just use my global $CC. [19:17] <relaxed> taylanub: -cc=$CC done [19:20] <taylanub> Still not as good as: <nothing> done :P [19:46] <ubitux> relaxed: that's especially troublesome when you are using an automated build of ffmpeg, like let's say, in ffmpeg [19:47] <ubitux> in mplayer* [21:24] <john_____> hi, does anyone has some knowledge about libstagefright.cpp ? [21:50] <durandal_1707> john_____: what is issue? [22:09] <funyun_> hi. i'm trying to encode 2 videos into 1 video. but i get this error http://pastebin.com/1r2tumWx , can anyone help me? [22:15] <funyun_> anyone have any idea? [22:23] <funyun_> JEEB: are you familiar with concat in ffmpeg? [22:25] <durandal_1707> funyun_: concat what? [22:25] <durandal_1707> protocol?filter?demuxer? [22:25] <ubitux> seems to be the filter here [22:31] <funyun_> durandal_1707: yes, filter [22:31] <funyun_> durandal_1707: here's my problem http://pastebin.com/1r2tumWx [22:51] <llogan> ubitux: looks like ffmpeg in Extra now has your config requests. [22:52] <llogan> s/extra/testing [22:53] <ubitux> llogan: oh? all of them? [22:55] <ubitux> funyun: try to put the audio streams after the video ones maybe [23:15] <tds5016> hi. I was wondering if anyone could send a list of dependencies for ffmpeg? [23:16] <llogan> ubitux: i think so [23:17] <ubitux> nice ok [23:17] <llogan> tds5016: depends on what you want [23:18] <tds5016> ffmpeg -i fs3 -c:v copy -f mpegts stream3.ts [23:18] <tds5016> enough to do that :-P [23:20] <tds5016> how would I track down the dependencies on that? [23:20] <tds5016> thanks for the help man. llogan you are awesome. [23:26] <ubitux> tds5016: gcc [23:26] <ubitux> i guess that's all [23:26] <tds5016> really? no deps aside from that for what I'm doing there? [23:26] <ubitux> for remuxer i don't think so [23:26] <ubitux> remuxing* [23:27] <ubitux> yasm is generally recommended for a ffmpeg build though [23:27] <tds5016> yasm? [23:27] <ubitux> assembler [23:27] <ubitux> for the optims [23:27] <ubitux> (mostly in codecs) [23:27] <tds5016> I'll be crosscompiling using gcc. [23:28] <tds5016> I don't think I can use yasm for this. [23:28] <tds5016> what would the configure options be to allow what I have there to work? [23:28] <ubitux> none [23:29] <ubitux> you can disable a lot of things though [23:29] <ubitux> like --disable-everything and just enable the few flags you need [23:29] <ubitux> s/flags/things/ [23:31] <tds5016> awesome. so just configure and it'll install? I don't need to disable/enable anything and it'll all work? [23:31] <tds5016> correction: SHOULD all work? [23:35] <ubitux> yes [23:36] <ubitux> note: you don't need to make install to test, make is enough, you can ./ffmpeg [23:37] <tds5016> thanks! [23:37] <tds5016> you rock :-) [00:00] --- Fri Feb 15 2013

1 0

ffmpeg-devel.log.20130213
by burek 14 Feb '13

14 Feb '13

[00:31] <j-b> Compn: indeed [01:16] <cone-102> ffmpeg.git 03Michael Niedermayer 07master:4f1279154ee9: shorten: dont leave invalid channel counts in the context. [04:36] <cone-884> ffmpeg.git 03Daniel Verkamp 07master:2acd5fb55325: doc/demuxers: quote command line arguments with @code [04:52] <Compn> nevcairiel : highgod is going to post his 3rd patch ,adding opencl support to scale filter [04:53] <Compn> one of these days i'm just going to commit his stuff :P [06:45] <Compn> https://github.com/nblock/pkgbuild/blob/master/otrdecoder/PKGBUILD [06:45] <Compn> what this otrkey thing is [06:46] <Compn> will have to check it later, maybe some germans can shine a light [10:01] <nevcairiel> Compn: why would i care about opencl [10:01] <nevcairiel> there is also general uglyness in the patches, anyone can review them :p [11:32] <cone-580> ffmpeg.git 03Justin Ruggles 07master:a3de4010c2d0: 8svx: decode directly to the user-provided AVFrame [11:32] <cone-580> ffmpeg.git 03Justin Ruggles 07master:ffd2123095bd: aac: decode directly to the user-provided AVFrame [11:32] <cone-580> ffmpeg.git 03Justin Ruggles 07master:55d2e12aefa2: ac3: decode directly to the user-provided AVFrame [11:32] <cone-580> ffmpeg.git 03Justin Ruggles 07master:e57daa876bf0: adpcm: decode directly to the user-provided AVFrame [11:32] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:4789955ec426: Merge commit 'e57daa876bf0cf50782550e366e589441cd8c2bd' [11:44] <cone-580> ffmpeg.git 03Justin Ruggles 07master:0588935e64fe: adx: decode directly to the user-provided AVFrame [11:44] <cone-580> ffmpeg.git 03Justin Ruggles 07master:bae4f47938bb: adxenc: alloc/free coded_frame instead of keeping it in the ADXContext [11:44] <cone-580> ffmpeg.git 03Justin Ruggles 07master:5cd597f22ff5: alac: decode directly to the user-provided AVFrame [11:44] <cone-580> ffmpeg.git 03Justin Ruggles 07master:9b0b355e9731: als: decode directly to the user-provided AVFrame [11:44] <cone-580> ffmpeg.git 03Justin Ruggles 07master:e3db34291f44: amrnb: decode directly to the user-provided AVFrame [11:44] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:d13f434dbb55: Merge commit 'e3db34291f4401a16f6ac92721617a9f33cd4c31' [11:51] <cone-580> ffmpeg.git 03Justin Ruggles 07master:da28bb3f4a7c: amrwb: decode directly to the user-provided AVFrame [11:51] <cone-580> ffmpeg.git 03Justin Ruggles 07master:5932e2d7d2c0: ape: decode directly to the user-provided AVFrame [11:51] <cone-580> ffmpeg.git 03Justin Ruggles 07master:0ded61b57a3d: atrac1: decode directly to the user-provided AVFrame [11:51] <cone-580> ffmpeg.git 03Justin Ruggles 07master:9a75ace2b2cb: atrac3: decode directly to the user-provided AVFrame [11:51] <cone-580> ffmpeg.git 03Justin Ruggles 07master:5cc0bd2cb47c: binkaudio: decode directly to the user-provided AVFrame [11:51] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:f03cdbd0454f: Merge commit '5cc0bd2cb47cbb1040f2bb0ded8d72a442c79b20' [11:58] <cone-580> ffmpeg.git 03Justin Ruggles 07master:a6bb39add273: pcm: decode directly to the user-provided AVFrame [11:58] <cone-580> ffmpeg.git 03Justin Ruggles 07master:e42e5a89d623: bmvaudio: decode directly to the user-provided AVFrame [11:58] <cone-580> ffmpeg.git 03Justin Ruggles 07master:cddf8998f1af: comfortnoise: decode directly to the user-provided AVFrame [11:58] <cone-580> ffmpeg.git 03Justin Ruggles 07master:7b78321597a9: cook: decode directly to the user-provided AVFrame [11:58] <cone-580> ffmpeg.git 03Justin Ruggles 07master:182821cff43f: dca: decode directly to the user-provided AVFrame [11:58] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:0a5138695aa4: Merge commit '182821cff43f5f977004d105b86c47ceb20d00d6' [12:03] <cone-580> ffmpeg.git 03Justin Ruggles 07master:97c7bdc6d417: cinaudio: decode directly to the user-provided AVFrame [12:03] <cone-580> ffmpeg.git 03Justin Ruggles 07master:b8e9c99ef19c: flac: decode directly to the user-provided AVFrame [12:03] <cone-580> ffmpeg.git 03Justin Ruggles 07master:036e9b045f14: g722: decode directly to the user-provided AVFrame [12:03] <cone-580> ffmpeg.git 03Justin Ruggles 07master:7e52fd6bca83: g723.1: decode directly to the user-provided AVFrame [12:03] <cone-580> ffmpeg.git 03Justin Ruggles 07master:cb7b47a61dba: g726: decode directly to the user-provided AVFrame [12:03] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:d88e674a15dc: Merge commit 'cb7b47a61dba0b9329ecede5dd3211dc0662dc05' [12:12] <cone-580> ffmpeg.git 03Justin Ruggles 07master:6fdfdb23d36f: gsm: decode directly to the user-provided AVFrame [12:12] <cone-580> ffmpeg.git 03Justin Ruggles 07master:9b28e583572c: imc/iac: decode directly to the user-provided AVFrame [12:12] <cone-580> ffmpeg.git 03Justin Ruggles 07master:b8788671288e: dpcm: decode directly to the user-provided AVFrame [12:12] <cone-580> ffmpeg.git 03Justin Ruggles 07master:a8ea936a0a00: libilbc: decode directly to the user-provided AVFrame [12:12] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:5459cf411412: Merge commit 'a8ea936a0a00570f61a16a588821b52f6a3115c2' [12:18] <cone-580> ffmpeg.git 03Justin Ruggles 07master:4f69612d3e61: libgsm: decode directly to the user-provided AVFrame [12:18] <cone-580> ffmpeg.git 03Justin Ruggles 07master:0cd08367dd6a: libopencore-amr: decode directly to the user-provided AVFrame [12:18] <cone-580> ffmpeg.git 03Justin Ruggles 07master:19b2cb268f5e: libopus: decode directly to the user-provided AVFrame [12:18] <cone-580> ffmpeg.git 03Justin Ruggles 07master:bed957bb1116: libspeex: decode directly to the user-provided AVFrame [12:18] <cone-580> ffmpeg.git 03Justin Ruggles 07master:86bfcfcf2364: mace: decode directly to the user-provided AVFrame [12:18] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:afe30fe06008: Merge commit '86bfcfcf2364bc837b7bb582c66a8a15a332414f' [12:27] <cone-580> ffmpeg.git 03Justin Ruggles 07master:dc33fbbfd0f4: mlp/truehd: decode directly to the user-provided AVFrame [12:27] <cone-580> ffmpeg.git 03Justin Ruggles 07master:2c785e250a1f: mpegaudio: decode directly to the user-provided AVFrame [12:27] <cone-580> ffmpeg.git 03Justin Ruggles 07master:3a23752c5a06: mpc7/8: decode directly to the user-provided AVFrame [12:27] <cone-580> ffmpeg.git 03Justin Ruggles 07master:0fe4056f953a: nellymoser: decode directly to the user-provided AVFrame [12:27] <cone-580> ffmpeg.git 03Justin Ruggles 07master:0905c96390b0: pcm-bluray: decode directly to the user-provided AVFrame [12:27] <cone-580> ffmpeg.git 03Justin Ruggles 07master:1b9b6d6e5ea5: qcelp: decode directly to the user-provided AVFrame [12:27] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:65da7007043c: Merge commit '1b9b6d6e5ea556b6d307f9d473f54f6406fdc3c8' [12:32] <cone-580> ffmpeg.git 03Justin Ruggles 07master:e01e1a4673a5: qdm2: decode directly to the user-provided AVFrame [12:32] <cone-580> ffmpeg.git 03Justin Ruggles 07master:cbeb3ed38d9d: ralf: decode directly to the user-provided AVFrame [12:32] <cone-580> ffmpeg.git 03Justin Ruggles 07master:f7e8c87c02dc: ra144: decode directly to the user-provided AVFrame [12:32] <cone-580> ffmpeg.git 03Justin Ruggles 07master:79fb2a1f1725: ra288: decode directly to the user-provided AVFrame [12:32] <cone-580> ffmpeg.git 03Justin Ruggles 07master:5d5c248c3df3: s302m: decode directly to the user-provided AVFrame [12:32] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:08059f6150fc: Merge commit '5d5c248c3df30fa91a8dde639618c985b9a11c53' [12:40] <cone-580> ffmpeg.git 03Justin Ruggles 07master:fed74c0ae497: shorten: decode directly to the user-provided AVFrame [12:40] <cone-580> ffmpeg.git 03Justin Ruggles 07master:09d6831f4901: sipr: decode directly to the user-provided AVFrame [12:40] <cone-580> ffmpeg.git 03Justin Ruggles 07master:903b62cc0b88: smackaud: decode directly to the user-provided AVFrame [12:40] <cone-580> ffmpeg.git 03Justin Ruggles 07master:4a2b26fc1b1a: tak: decode directly to the user-provided AVFrame [12:40] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:2becf21d9fd8: Merge commit '4a2b26fc1b1ad123eba473a20e270f2b0ba92bca' [12:47] <cone-580> ffmpeg.git 03Justin Ruggles 07master:3997fef95262: truespeech: decode directly to the user-provided AVFrame [12:47] <cone-580> ffmpeg.git 03Justin Ruggles 07master:ad2104ba449a: tta: decode directly to the user-provided AVFrame [12:47] <cone-580> ffmpeg.git 03Justin Ruggles 07master:3b7d43383f5e: twinvq: decode directly to the user-provided AVFrame [12:47] <cone-580> ffmpeg.git 03Justin Ruggles 07master:f80f8dd4c25a: vmdaudio: decode directly to the user-provided AVFrame [12:47] <cone-580> ffmpeg.git 03Justin Ruggles 07master:ee6ca11b6575: vorbis: decode directly to the user-provided AVFrame [12:48] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:dca6fb08a71b: Merge commit 'ee6ca11b657515ad736ec0d2b8635e098d0a2680' [12:53] <cone-580> ffmpeg.git 03Justin Ruggles 07master:9873d71f316c: wavpack: decode directly to the user-provided AVFrame [12:53] <cone-580> ffmpeg.git 03Justin Ruggles 07master:f4a283eec49b: wmapro: decode directly to the user-provided AVFrame [12:53] <cone-580> ffmpeg.git 03Justin Ruggles 07master:c815ca36412e: wma: decode directly to the user-provided AVFrame [12:53] <cone-580> ffmpeg.git 03Justin Ruggles 07master:205a95f7b517: wmaenc: alloc/free coded_frame instead of keeping it in the WMACodecContext [12:53] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:0dff771f31cf: Merge commit '205a95f7b5178362874bc1e65eae9866723491c1' [12:58] <cone-580> ffmpeg.git 03Justin Ruggles 07master:5a7288822f24: wmavoice: decode directly to the user-provided AVFrame [12:58] <cone-580> ffmpeg.git 03Justin Ruggles 07master:8ae50d87e771: ws-snd1: decode directly to the user-provided AVFrame [12:58] <cone-580> ffmpeg.git 03Justin Ruggles 07master:646831e697a5: lavr: print out the mix matrix in ff_audio_mix_set_matrix() [12:58] <cone-580> ffmpeg.git 03Justin Ruggles 07master:1647da89dd8a: lavr: make sure that the mix function is reset even if no mixing will be done [12:58] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:91043de8256f: Merge commit '1647da89dd8ac09a55c111589f7a30d7e6b87d90' [13:05] <cone-580> ffmpeg.git 03Justin Ruggles 07master:a3735bb92a6b: lavr: cosmetics: reindent [13:05] <cone-580> ffmpeg.git 03Justin Ruggles 07master:600b4c973fb9: lavr: fix matrix reduction for upmixing in certain cases [13:05] <cone-580> ffmpeg.git 03Justin Ruggles 07master:157542ebc15d: lavr: fix mixing matrix reduction when normalization is disabled [13:05] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:845fa2f5c964: Merge remote-tracking branch 'qatar/master' [13:36] <cone-580> ffmpeg.git 03Paul B Mahol 07master:9145818ecd51: vima: decode directly to the user-provided AVFrame [13:36] <cone-580> ffmpeg.git 03Paul B Mahol 07master:2cced2a85442: paf: decode directly to the user-provided AVFrame [13:36] <cone-580> ffmpeg.git 03Paul B Mahol 07master:8f7c7ff2d605: evrcdec: decode directly to the user-provided AVFrame [13:38] <cone-580> ffmpeg.git 03Anton Khirnov 07release/0.5:0b25c3b67cef: mpeg12: do not decode extradata more than once. [13:38] <cone-580> ffmpeg.git 03Michael Niedermayer 07release/0.5:deaaacbc3f8d: Merge remote-tracking branch 'qatar/release/0.5' into release/0.5 [14:26] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:b2e57eb5a3cb: mjpegdec: pass nb_components into ljpeg_decode_yuv_scan [15:04] <cone-580> ffmpeg.git 03Nicolas George 07master:6d680d20ad8b: lavf/concatdec: add safe option. [15:04] <cone-580> ffmpeg.git 03Nicolas George 07master:db9dc5154d1a: lavf/concatdec: add signature directive. [15:04] <cone-580> ffmpeg.git 03Nicolas George 07master:b68dd8a1614b: lavf/concatdec: allow probing. [15:13] <cone-580> ffmpeg.git 03Nicolas George 07master:a17ececcc792: libcelt: decode directly to the user-provided AVFrame. [15:32] <durandal_1707> michaelni: libvorbis decoder is broken [15:34] <kierank> is ubitux back [15:43] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:fe84577273c9: libvorbisdec: set sample type [16:07] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:324d96644b57: libvorbisdec: Remove AVFrame from context [16:22] Action: ubitux is back kierank [16:24] <j-b> ubitux: we went and met your former Bosses this morning [16:25] <ubitux> "we"? [16:25] <j-b> we! [16:25] <ubitux> wepee~ [16:48] <durandal_1707> there are not deps for filters in fate? [16:51] <durandal_1707> can something be done about FATE_AVCONV nonsense in makefiles? [17:00] <michaelni> it could be renamed trivially but it would add conflicts on merges [17:07] Action: durandal_1707 waiting for s/FATE/LATE [17:09] <ubitux> Libav Automated Mess Environment [17:09] <ubitux> you can also have this variant [17:11] <av500> +1 [17:17] <durandal_1707> now i have big pix fmts table, and it is big and ugly [17:37] <durandal_1707> why all image formats and even raw cant be displayed when only part of data is present? [17:40] <cone-580> ffmpeg.git 03Paul B Mahol 07master:cf73e0bff3a2: lavu/intreadwrite: add AV_[RW][BLN]48 [17:40] <cone-580> ffmpeg.git 03Paul B Mahol 07master:34126cd090c3: lavfi/transpose: support more pix fmts [18:14] <durandal_1707> michaelni: what QP changer filter does? [18:16] <durandal_1707> lol, it changes some quantization tables.. [18:26] <michaelni> changes Quantization Parameter [18:26] <durandal_1707> -> does nothing in lavfi [18:28] <durandal_1707> why there are 3 blur filters - they could be merged into one [18:35] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:21cd905cd44a: swr: check channel layouts before using them. [18:47] <cone-580> ffmpeg.git 03Daniel Verkamp 07master:5dd78ef34e20: WAV muxer: reindent [18:49] <durandal_1707> burek: you still want to write playlist(m3u) demuxer? [20:19] <cone-580> ffmpeg.git 03Clément BSsch 07master:15e7533ca1a1: doc/dev: clarify public symbols prefixes. [21:32] <ubitux> don't worry about this one [21:32] <ubitux> most likely because of the upgrade in the middle of the tests [21:43] <kierank> ubitux: were there any more problems with those r128 patches? [21:43] <ubitux> i'll test them and apply, but i wanted a comment from the author about the back speakers masks [22:15] <tg2> when was version syntax changed from git-2012-11-08-aeb4d27 to N-49695-g4e2c636 ? [22:17] <llogan> tg2: i don't know if it was changed. [22:17] <tg2> what does the N-49695-g4e2c636 signify [22:18] <tg2> build # - commit? [22:18] <tg2> or something [22:18] <llogan> git-2012-11-08-aeb4d27 probably means you're using a shallow clone [22:18] <llogan> "git clone --depth 1 ..." [22:18] <ubitux> 49695th commit, hash=4e2c636? [22:19] <tg2> yeah trying ot figure out what tha tsecond hash is [22:19] <llogan> i'm guessing you're using ffmpeg-git from arch or followed the ubuntu compile guide [22:20] <tg2> used git clone [22:20] <tg2> then compiled [22:21] <tg2> anyway, have a weird bug i'm seeing betweeen two versions, both compiled within a day of eachother, one does it one doesnt with teh same file same command [22:21] <llogan> ah, yes, in version.sh: Shallow Git clones (--depth) do not have the N tag [22:21] <tg2> [mpeg4 @ 0x6be18a0] Invalid and inefficient vfw-avi packed B frames detected [22:21] <tg2> on one [22:21] <tg2> ah ok llogan [22:21] <tg2> that makes sense [22:23] <tg2> http://pastie.org/6157210 [22:24] <tg2> with cmd: http://pastie.org/6157216 [22:55] <BBB-work> someone broke msvc dll builds by marking ff_codec_bmp_tags[] with av_export [22:55] <BBB-work> can whoever did that fix it? [22:55] <BBB-work> https://github.com/libav/c99-to-c89/issues/5 [23:09] <beastd> BBB-work: ^^^^ i think it could have been him. need to check. [23:11] <michaelni> aa715cdc <-- [23:21] <beastd> BBB-work: A generic question? Do you have a strategy handle shared data symbols between DLLs? Or you completely avoid it? [23:25] <llogan> It's GSoC time again. [23:34] <cone-580> ffmpeg.git 03Michael Niedermayer 07master:c230af9bccc3: h264: Reset last_pocs in case of reference or frame number inconsistencies [23:48] <nevcairiel> beastd: av_export can in general share data symbols, however its a bit of a pain. You cant use them as static initializers anymore (which causes the current breakage), and stuff like that [23:48] <nevcairiel> additionally, a ff_* symbol should not be exported. :p [23:50] <nevcairiel> it may be easier in many cases to add a getter functionf or such tables [23:50] <BBB-work> that doesn't work for static users like this [23:50] <BBB-work> generally, avoid it [23:51] <BBB-work> if at all possible [23:51] <beastd> nevcairiel: well, i said generally (not this instance) [23:51] <beastd> BBB-work, nevcairiel: but thanks for your answers [23:52] <nevcairiel> its generally a painful thing to do, so don't :) [23:52] Action: beastd also tends to think that avoiding it is the probably the best strategy [23:53] <beastd> so if it was roger (dshow guy) who broke this someone should send a mail (i do not think he is here on irc) [23:59] <beastd> Need to get some sleep now. [23:59] <beastd> bye... [00:00] --- Thu Feb 14 2013

1 0

ffmpeg.log.20130213
by burek 14 Feb '13

14 Feb '13

[00:21] <fatpony> nice tip llogan thanks [00:21] <tds5016> hi. [00:21] <tds5016> Is it possible to use ffmpeg to convert from h264 stream over to an mpegts for http live streaming? [00:26] <tds5016> I've basically got an h264 file, and I'm trying to convert it over to an http live streaming format to share out. [01:47] <jankarlitos> Hello. Is anyone capable of get the streams out of this shoutcast URL?: http://johnny.serverroom.us:6336 [01:49] <klaxa> um... [01:49] <klaxa> that's pretty much the stream already [01:50] <klaxa> wget http://johnny.serverroom.us:6336/ -O stream_dump.mp3 [01:51] <jankarlitos> klaxa, yes that's the only way i could get it. But i need to transcode it with ffmpeg [01:51] <klaxa> ffmpeg -i http://johnny.serverroom.us:6336/ -c:a libvorbis stream_dump_encoded.ogg [01:51] <klaxa> or whatever [01:52] <jankarlitos> It doesn't work [01:53] <klaxa> mmhh.. it appears to hang that's weird [01:54] <klaxa> this shouldn't be a problem, for the time being you could create a pipe with wget [01:54] <klaxa> wget http://johnny.serverroom.us:6336/ | ffmpeg -i - -c:a libvorbis stream_dump.ogg [01:54] <klaxa> argh [01:55] <klaxa> wget http://johnny.serverroom.us:6336/ -O - | ffmpeg -i - -c:a libvorbis stream_dump.ogg [01:55] <klaxa> like that [01:56] <jankarlitos> Already try that. The result: "Cannot write to - (Broken pipe)" [01:56] <klaxa> what is your command line? [01:58] <jankarlitos> The same as yours: wget http://johnny.serverroom.us:6336/ -O - | ffmpeg -i - -acodec libvorbis out.ogg [01:59] <klaxa> mmhh that sounds like a restriction set by your system [02:01] <klaxa> hmm maybe you can use named pipes but that would get ugly [02:14] <jShaf> how do I skip the beginning to the very first keyframe for video INPUT? [02:14] <jShaf> like -ss <time of the first keyframe> [05:19] <happarappa> ffmpeg.exe -i test.mp4 -sameq test.avi <-- This results in an .avi of poored quality than the input .mp4. What the hell? Just when I thought I had ffmpeg figured out, it goes and does this to me. [05:20] <klaxa> sameq just means same quantizer not same quality [05:20] <klaxa> i think it should be avoided if possible [05:23] <happarappa> Same quantizer? [05:24] <happarappa> Never even heard of any such thing as "quantizer". [05:24] <Plorkyeran> then you definitely don't want to be using sameq [05:25] <fenduru> Anyone know a way to force an encoder to NOT add padding? [05:26] <fenduru> ffprobe reports a .wav as 0.97 seconds long, but after encoding to aac it is 10.0, and after encoding to mp3 it is 10.03. These are audio segments so playing them in sequence is causing gaps [05:27] <happarappa> F... F... S. :| [05:34] <relaxed> it was removed from ffmpeg because it most cases it didn't even work. [05:37] <happarappa> What was? [05:43] <relaxed> -sameq [05:45] <happarappa> relaxed: Huh? It's not removed? [05:50] <relaxed> You must be using an older version. [05:52] <relaxed> happarappa: use '-q:v 3' instead for good quality [06:05] <happarappa> Not using an old version... [06:05] <happarappa> I want the same exact quality. [06:14] <relaxed> then use -c:v copy, otherwise it's impossible [06:17] <Plorkyeran> or a lossless codec [06:50] <happarappa> :| [07:52] <asher^> hi guys. does ffmpeg have an option similar to -ss where you can seek to half way through a video rather than a specific time? [07:54] <klaxa> yes [07:54] <klaxa> asher^: see: http://ffmpeg.org/trac/ffmpeg/wiki/Seeking%20with%20FFmpeg [07:54] <asher^> thanks [07:56] <asher^> klaxa, i cant see it there. i should be more specific, id like to know if theres an option where i dont need to know the duration in advance [07:57] <klaxa> um... [07:57] <klaxa> i don't think so? [07:57] <klaxa> for what usecase would you need that? [07:57] <asher^> im just taking screen grabs of videos programatically where i dont know the length in advance. i thought id just grab them from the middle [07:58] <asher^> some are 5 mins, some an hour [07:58] <klaxa> so you want to specify like... -ss 50% ? [07:59] <asher^> yeah something like that [08:00] <klaxa> hmm... well you could certainly script it with the help of ffprobe [08:00] <asher^> yeah i guess ill have to do something like that [08:00] <asher^> thanks [08:01] <asher^> hmm, looks like something called ffmpegthumbnailer might do it [08:01] <asher^> although that looks like linux too [08:38] <spaam> klaxa: nice page. [08:39] <klaxa> i know, i didn't write it though [08:39] <klaxa> but it's rather useful, i didn't know ffmpeg could do fast-seeking with keyframes [08:39] <spaam> it was burek and some other guy D: [09:20] <happarappa> Why is ffmpeg the most confusing nonsense ever made? [09:21] <happarappa> I just want to convert from one file format to another with the same amount of quality/untouched. [09:21] <klaxa> happarappa: because it's the leading FOSS in video encoding [09:21] <klaxa> that's not possible unless you use lossless codecs [09:21] <klaxa> or copy the codec [09:21] <klaxa> this is a general truth and has nothing to do with ffmpeg [09:22] <klaxa> oh ffmpeg also does video filtering [09:22] <klaxa> same stuff with audio :V [09:23] <klaxa> i mean come on, sure it's not easy to use, but it's still one of the greatest open source projects [09:25] <happarappa> :( [09:25] <happarappa> Yes, it is amazing... if it were just nicer to use. [09:25] <happarappa> Supports an insane amount of formats. [09:26] <klaxa> actually i find it rather easy to use [09:26] <klaxa> everything is well documented [09:26] <happarappa> I get a crappy video result from my conversion. [09:26] <happarappa> I want it to look just like the original. [09:26] <klaxa> <klaxa> that's not possible unless you use lossless codecs [09:26] <klaxa> <klaxa> or copy the codec [09:26] <happarappa> ... What? [09:27] <happarappa> Of course it's possible. [09:27] <happarappa> It's damaging the video. [09:27] <klaxa> you cannot achieve the exact s-- [09:27] <JEEB> a result close to the original is possible [09:27] <JEEB> not the exact same [09:27] <happarappa> It's MUCH worse. [09:27] <happarappa> ffmpeg.exe -i test.mp4 -sameq test.avi [09:27] <JEEB> pastebin your command line and terminal output [09:27] <JEEB> ugh [09:27] <JEEB> sameq is not what you think [09:27] <JEEB> and was actually broken in most cases [09:27] <happarappa> Whatever it means, the result is not what I want... [09:28] <JEEB> it is not same quality but "same quantizer", which doesn't even make sense between two different formats [09:28] <JEEB> also I think it was removen later [09:28] <JEEB> as an option [09:28] <JEEB> because people kept using it even if it was broken [09:28] <JEEB> so what do you want your output to be? [09:29] <JEEB> for what are you encoding? [09:29] <happarappa> Visually impossible to tell apart. [09:29] <JEEB> no, i mean [09:29] <happarappa> For what am I encoding? [09:29] <JEEB> avi is a container [09:29] <JEEB> what do you want into it? [09:29] <JEEB> just like mp4, avi is a "box" [09:29] <JEEB> you can put stuff into the "box" [09:29] <happarappa> I don't care. When files are jungled through the black box that is "ffmpeg", they work in various software such as video editors. [09:30] <JEEB> :| [09:30] <happarappa> Prior to being put into the black box, they are not accepted. [09:30] <happarappa> (Renaming the files does not work.) [09:30] <JEEB> well, naturally [09:30] <JEEB> if the editor doesn't support the container format [09:30] <JEEB> or the video inside it, then naturally [09:30] <JEEB> for video editors I recommend utvideo [09:30] <JEEB> because you can get decoder components for every OS for it [09:31] <JEEB> Windows, OS X, Linux [09:31] <JEEB> ffmpeg -i input -c:v utvideo -c:a pcm_s16le out.avi [09:32] <klaxa> heh promoting utvideo again :) [09:32] <aji> i wonder what percentage of this channel's messages are ffmpeg command lines [09:32] <klaxa> (btw, is it multi-threaded yet?) [09:32] <JEEB> the decoder is, the encoder is harder to MT [09:32] <JEEB> at least with slices [09:32] <JEEB> frame-based MT would be easier [09:33] <JEEB> happarappa, also it kind of would help me if you would tell which OS and editor you're going to use [09:33] <JEEB> and I asked for the terminal output because I kind of could have seen the version of ffmpeg you have :V [09:34] <JEEB> pretty much everything I ask for is for a reason [09:35] <JEEB> klaxa, also why would I not promote it? It has components for every OS after all [09:35] <JEEB> I really don't see a reason not to use it if you're doing video editing :) [09:35] <klaxa> i wasn't complaining, just noting :) [09:36] <jeje34> Hi to all ;-) [09:36] <JEEB> Daemon404 actually multithreaded the prediction per-slice but that ended up being bottlenecked by the huffman afterwards [09:36] <JEEB> so it seems like the easiest way to make the libavcodec utvideo encoder faster is to just frame thread it :< [09:37] <klaxa> hmm okay i don't understand this well enough yet to comment on it :P [09:41] <jeje34> I already have integrated FFMPEGs Library to decode H264 video streaming from IP Cameras and display them in windows of a Windows OS. But now, I'm trying to decrease the CPU usage when I have several decompressions at the same time. So I compile FFMPEG (version release 1.1.1) with --enable-dxva2 --enable-decoder=h264_dxva2 --enable-hwaccel=h264_dxva2 [09:43] <jeje34> After do this, I try to modify my code to use AVHWAccel. but it's really hard to find documentation or sample code source to use it...hte only thing I found is I have to use av_hwaccel_next to find the good one (hwaccel->id == CODEC_ID_H264) and register it by av_register_hwaccel but It seems it isn't enougth to use the hardware decompression [09:43] <JEEB> yes [09:43] <jeje34> if someone can help me, thanks [09:44] <JEEB> hwaccel has you do work too, because libavcodec can have no idea on how you want to use DXVA [09:44] <JEEB> https://lists.ffmpeg.org/pipermail/ffmpeg-user/2012-May/006600.html [09:44] <JEEB> take a look at this post [09:45] Action: JEEB goes back to doing laundry [09:48] <jeje34> thanks fot this reply! for the point 3- Create a dxva_context structure and initialize it's fields with the above objects, is it a context like an AVCodecContext? And how initialize the fields? [09:50] <jeje34> if someone has already use AVHWAccel in his code, it will be very helpfull [10:09] <JEEB> jeje34, there is an app mentioned in the post, so if you want you can look at that as an example [10:13] <happarappa> JEEB: I don't see what my OS or my video editor has to do with this. [10:15] <JEEB> happarappa, because if you are using a packaged ffmpeg it can be old, if you are using linux video editors they can be built with old libavcodec [10:15] <JEEB> those naturally limit the alternatives [10:15] <jeje34> JEEB> you telling about "a useful tool for diagnosing DXVA2 is DXVAChecker" [10:15] <JEEB> jeje34, no [10:15] <JEEB> far before that [10:15] <jeje34> ok the msdn link [10:16] <JEEB> well, that is related yes, but you know -- the reference to VLC [10:16] <JEEB> because DXVA is a MS thing, you will find all the docs related to it from MSDN :P [10:17] <jeje34> but do you think it really decrease my CPU time when I decode several streams at the same time (because it need me to make a lot of changes in my implementation...) [10:17] <happarappa> JEEB: None of that made any sense to me. Sorry. [10:17] <happarappa> "a packaged ffmpeg"? [10:17] <JEEB> package management [10:18] <JEEB> on linux most distributions have a package management system [10:18] <JEEB> where most people get their software [10:18] <JEEB> on OS X you have macports and homebrew for similar usage [10:18] <happarappa> Windows here. [10:18] <JEEB> ok [10:18] <JEEB> then the case B) of you having an editor that can't use the VFW component is unrelated [10:19] <JEEB> then Ut Video should work just fine and be lossless as long as you install the windows codec for it (DirectShow/VFW) [10:19] <JEEB> http://umezawa.dyndns.info/wordpress/?p=3655 [10:19] <JEEB> of course your windows build of ffmpeg has to be new enough to support ut video [10:19] <JEEB> but yes, that line I gave you up there should work then [10:20] <JEEB> <JEEB> ffmpeg -i input -c:v utvideo -c:a pcm_s16le out.avi [10:20] <JEEB> it will be big, but it will also be lossless as well as intra so it will be quick to seek through [10:20] <JEEB> when editing [10:21] <JEEB> (lossless does not mean uncompressed, it is compressed) [10:21] <JEEB> also if you have too old of an ffmpeg, then you can grab a newer one from zeranoe http://ffmpeg.zeranoe.com/builds/ [10:29] <jeje34> I also have another question about my FFMpeg use. I compile it with --enable-w32threads (because I'm a window user). But when I try to set a thread_count in my AVCodecContext (like 3), I always have a log error The maximum value for lowres supported by the decoder is 0. when calling avcodec_open2 [10:29] <jeje34> So I can't use FFMPEG Multithreading to decode H264? [10:30] <JEEB> no idea if you're just using libavcodec wrong [10:30] <JEEB> because there has been H.264 multithreaded frame decoding since summer 2011 [10:31] <JEEB> sliced decoding is a more derpy thing, and only works with slices [10:31] <JEEB> sliced threading [10:34] <jeje34> ok JEEEB, I set my ffmpeg code part here: http://pastebin.com/sNpcpHjf [10:36] <jeje34> but I want to decompress H264 frames coming from a camera. So in my code I recompose one entire frame before passing it to FFMPEG so I think it will be more FF_THREAD_FRAME I need to use [10:36] <JEEB> I have no idea, I have only coded within libavcodec, not used it myself :P Unfortunately you'll have to have someone else look at that code, or look at other examples using libavcodec's threading (various things use it) [10:36] <jeje34> ok thanks [10:43] <jeje34> another one question: do you think in my case, using a CodecCtx->thread_count could decrease my CPU usage to decode my H264 stream? [10:44] <jeje34> when I have several streams to decode at the same time [10:46] <JEEB> depending on how many things you have going at the same time, you can make them derp less with each other by limiting the overall amount of threads, the notion of "could decrease my CPU usage" is kind of derp -- if you are bound by, say, a single thread it will use less CPU but be slower, while more threads mean more power can be used for one stream. In case of you doing multiple parallel jobs of course, you might want to limit the [10:46] <JEEB> amount of threads depending on the amount of cores in the system and the amount of streams you're doing, but that really won't "decrease your CPU usage", it will just make those separate processes not poke each other as much [11:00] <jeje34> ok JEEB, thanks for all your answers. If I don't set a thread_count but have compiling FFMPEG with --enable-w32threads, what happen with FFMPEG? because when calling avcodec_alloc_context3, it initialze the AVCodecContext with avcodec_get_context_defaults3, so setting the threadcount to 0 [11:02] <JEEB> that should be the default now > --enable-w32threads [11:02] <JEEB> on windows [11:02] <JEEB> and it just enables threading [11:02] <JEEB> I think it depends on the default in the decoder? No idea, as I've said I've never used the API [11:02] <JEEB> so I don't know the defaults and how it is going to function [11:07] <jeje34> ok, I'm going to seek in the FFMPEG source code to understand a little bit more what happen [11:14] <fatpony> i can't seem to crop one pixel, is that normal? can i only crop to even values? [11:15] <JEEB> if your source has chroma subsampling (4:2:2 or 4:2:0 YCbCr [colloquially called "YUV"]), then yes [11:16] <JEEB> because with 4:2:0 you have one value for a 2x2 area, and 4:2:2 has one value for a 2x1 area [11:17] <fatpony> ah so i guess that's what yuv420p stands for [11:17] <JEEB> yup [11:18] <fatpony> it's strange because i can clearly see that a black border beginning at an odd height [11:18] <fatpony> s/that// [11:19] <JEEB> luma (grayscale image) is full resolution, but the chroma planes (color information) are subsampled [11:19] <JEEB> so yes, I guess having the black area start at an odd height is possible I guess [11:21] <fatpony> ah i see, thanks! [13:49] <jShaf> how do i trim the beginning up to first keyframe on the video input? [13:59] <xroberx> hi [13:59] <xroberx> If I only want to do pixel format conversion, which is the right functions ? sws_scale() ? [14:00] <durandal_1707> yes [14:00] <xroberx> I do not want to perform scaling though [14:01] <xroberx> durandal_1707: so... I guess I'll have to set the source width/height equal to the destination width/height, right ? [14:01] <durandal_1707> same function can do most of unscaled conversions too [14:01] <durandal_1707> depends on input->output formats [14:01] <durandal_1707> and obviously same dimensions [14:02] <weecka> Hello guys, maybe you can help me. I am receiving lots of these warnings: max resync size reached, could not find sync byte.78 bitrate=N/A. Whole output is here http://pastebin.com/Ke9wse1a. What does this means? [14:03] <durandal_1707> weecka: how you create file? what can play it? and soo on .. [14:05] <weecka> Honestly to say, I'm trying to convert flv to mpegts, which then could be played using apple streaming protocol. [14:05] <xroberx> durandal_1707: do you know if there is any ARM optimized version ? By looking in the libswscale folder I can see there is only optimizations for x86, ppc, sparc and bfin. [14:06] <weecka> but I'm not sure, what I'm doing wrong. [14:06] <durandal_1707> xroberx: as you noticed, there is noone [14:07] <xroberx> ok [14:08] <durandal_1707> weecka: the mpegts muxer may be just buggy, so what is your command that do flv->mpegts? [14:09] <weecka> I've it did like this: ffmpeg -y -i mindmovie-5173.flv -f mpegts -acodec libmp3lame -ar 48000 -ab 64k -s 320x480 -vcodec libx264 -b 512k -maxrate 512k -bufsize 512k -qcomp 0.6 -qmin 30 -qmax 51 -qdiff 4 -level 30 -aspect 320:480 -g 30 -async 2 output-iphone-big.ts [14:14] <weecka> Also, I've tried to do flv->m4v and then to segment it, but the segment durations somehow are then way of target. [14:41] <xroberx> hi [14:43] <weecka> durandal_1707: any thoughts? [14:43] <xroberx> durandal_1707: I've found this ARM optimized yuv2rgb conversion library : http://wss.co.uk/pinknoise/yuv2rgb/ The license is BSD. Now the question is: would it be possible to integrate it in ffmpeg's libswscale ? [14:45] <durandal_1707> xroberx: what kind of integration? [14:46] <durandal_1707> writing patches that add arm optimization to swscale is certainly possible and likely to be accepted and applied [14:47] <xroberx> durandal_1707: ok, good to know, I'll contact the author then, but considering it's BSD licensed I guess it can be done [14:47] <durandal_1707> if you can't code you find someonce who can [14:47] <creep> h [14:47] <durandal_1707> xroberx: why would you contact author? [14:48] <xroberx> durandal_1707: because of this sentence in his webpage: "If you do use use this code as part of a piece of software (or hardware), please let me know, purely for my own interest." [14:48] <durandal_1707> it's BSD license which means you can do almost anything with it (except claiming you write that code ...) [14:48] <durandal_1707> xroberx: good, but first code must be used ... [14:48] <durandal_1707> currently is not [14:49] <xroberx> yes [14:52] <xroberx> durandal_1707: do you know happen to know who is the person to contact if I wanted to send some patches to integrate the library ? [14:53] <JEEB> ffmpeg-devel mailing list for patches [14:53] <jarno> Hi Jeeb [14:53] <jarno> ...and everyone else... [14:54] <xroberx> JEEB: ok, thanks [14:55] <jarno> ...I have a question related to encoding and mux(ing)...if I have understood it correctly, the steps should be 1) find encoder 2) allocate context and 3) create output stream, right? [14:56] <jarno> ...so the avcodec_alloc_context3() sets default values as per the chosen encoder, right? [14:57] <jarno> ...and then these can be changed as needed? [15:34] <Macey> hi all whats the output format for just streaming a TS file to UDP? [15:36] <Mavrik> mpegts? [15:36] <Macey> if i do -f mpegts it just recreates the container [15:38] <Mavrik> mhm [15:38] <Mavrik> what do you want to do exactly? [15:41] <Macey> i have a fully formed TS file (NIT/SDT/PMTs e.t.c) that i want to push over UDP [15:41] <Macey> no transcoding nothing, just a straight tx [15:41] <Mavrik> ffmpeg isn't the right tool for the job then [15:41] <Macey> ah ok [15:41] <Macey> vlc? [15:41] <Mavrik> why don't you use something made to just dump data to network like netcat? [15:42] <Macey> needs to read the PCR and tx it at a rate [15:42] <Mavrik> both vlc and ffmpeg are meant to actually change streams, not be dumb network pipes [15:42] <Mavrik> VLC might know that, but I'm not really sure if it can do what you want withou modifying the stream [15:42] <Macey> ok thanks [15:50] <jeje34> Hello, just to know ... are functions av_lockmgr_addref and av_lockmgr_register are deprecated? [15:55] <jeje34> I can find av_lockmgr_register but not av_lockmgr_addref in avcodec.h (using last FFMPEG release 1.1.1) [16:36] <catalinb> Hi, I'm trying to build ffmpeg on windows using ./chromium/scripts/build_ffmpeg.sh [16:36] <catalinb> I'm calling the script from a mingw shell [16:37] <catalinb> I encounter the error: 'c99wrap is unable to create an executable file.' [16:37] <catalinb> I have a mingw installation and gcc is in path [16:38] <catalinb> How do I specify the right compiler? [17:48] <jeje34> hi to all [17:49] <jeje34> I have some question about using FFMPEG to decode h264 videos streaming from IPCamera. My code works (I can decode) but I have some questions about the initialization of the AVCodecContext fields [17:50] <jeje34> there are lot of them set - decoding: Set by user [17:51] <jeje34> I can really understand what are the best values for flags and flags2 of the AVCodecContext structure [17:54] <Mavrik> jeje34: usually most fields don't have to be set [17:54] <Mavrik> unless you have a very-nonstandard use case [17:55] <Mavrik> jeje34: those "set by user" don't always mean MUST be set by user ;) [17:55] <jeje34> I have a standard use case (I hope all manufacturer of IP camera have a good implementation of H264 video stream) [17:55] <jeje34> ;-) [17:56] <Mavrik> ^^ [17:56] <jeje34> but when I see some flags like CODEC_FLAG2_FAST [17:57] <jeje34> I'm not very sure if it's better to set [17:58] <Mavrik> jeje34: usually, leaving default settings is prefered [17:58] <Mavrik> seems like CODEC_FLAG2_FAST gains some speed at cost of decoded image quality [18:56] <Guest17651> hi, I'm trying to export VOB videos to ogg (theora/vorbis) for the web, using kdenlive. The videos are about 45 minutes each. Do you have good settings to recommend? [19:55] <someone-noone> if dts==pts does it mean that frame is a key-frame? [19:56] <someone-noone> or there is some cases when it's not true [19:56] <someone-noone> ? [19:58] <JEEB> no, it just means that it's not a b-frame [19:59] <someone-noone> Okay, thanks. That was a silly question. But I have another one, probalby less silly :) [20:00] <someone-noone> If I want to reorder packets in pts way and I want to flush packets when they are ordered. If pts==dts is a good event for that? [20:00] <someone-noone> If I want to reorder packets in pts order* [20:01] <someone-noone> Or better to look-up for key-frames for this event? [20:06] <Paranoialmaniac> JEEB: {k? it is true if pts always equals dts. but, if not, a frame with pts==dts is an un-referenced b-frame [22:15] <mordonez1> Hu guys, how can I put an image in a video in a specified time and position? [22:25] <llogan> the "specified time" is the tricky part. i don't think overlay filter has a temporal option [22:26] <llogan> of course you could section the video into parts, add the image overlay (you want an overlay, right), and then concat the sections together. [22:26] <mordonez1> yeah, I want to put an image inside a square on the video [22:26] <mordonez1> the square only appear on second 10 to 12 [22:27] <llogan> you can do that with the drawbox and overlay filters [22:28] <llogan> but i don't think either have "make filter work from 10 to 12 seconds" type of option [22:28] <mordonez1> you have any example or something that I can use to start with it? [22:38] <llogan> mordonez1: ffmpeg -i video.mp4 -i image.png -filter_complex drawbox,overlay output [22:38] <llogan> http://ffmpeg.org/ffmpeg-filters.html#drawbox [22:38] <llogan> http://ffmpeg.org/ffmpeg-filters.html#overlay-1 [23:40] <mordonez1> Thaks llogan [23:56] <someone-noone> Hey! If I want to reorder packets by pts (decoder is not ffmpeg), how can I know when to "flush" those packets? When new key-frame arrives or when dts==pts? Or may be some other algorithm? [00:00] --- Thu Feb 14 2013

1 0