Ffmpeg-devel-irc

Download

ffmpeg-devel-irc@ffmpeg.org

August 2013

1 participants
54 discussions

ffmpeg-devel.log.20130830
by burek 31 Aug '13

31 Aug '13

[00:45] <durandal11707> michaelni: how complicated would be adding packed yuva444? (YUVAYUVAYUVA....) [00:46] <ubitux> llogan: fun idea, but don't ask me to do it [00:47] <ubitux> durandal11707: you can hardsub bitmap subs, there is hack in ffmpeg to inject bitmaps into frames for overlay iirc [00:47] <ubitux> it's even documented [00:47] <durandal11707> ubitux: nicolas already told me hours ago [00:48] <ubitux> time is a bitch [00:53] <durandal11707> haha [01:20] <BBB> ubitux: is uvpred done? [01:20] <ubitux> wasn't it already? [01:21] <BBB> you said you'd make a small adjustment (remove duplicae off_u/v and stride_u/v [01:21] <BBB> ) [01:21] <BBB> otherwise it's done yes [01:34] <ubitux> fuck [01:34] <ubitux> my [01:34] <ubitux> isp [01:34] <ubitux> BBB: what message from me did you get? [01:38] <ubitux> BBB: no changes from last time; the only differing thing is that i pass 2x the offset (and you said linesize[1]==linesize[2]) [01:38] <ubitux> but im not sure about those assumptions [01:38] <ubitux> i dont think that hurts currently [01:38] <michaelni> durandal11707, dunno, see comits that added other packed yuv formats [01:39] <durandal11707> michaelni: considering how planar is hard, packed is harder [01:47] <BBB> ubitux: yes, that, and off_uv being passed twice [01:48] <BBB> ubitux: off_u and off_v are always the same (and they are; y/uv can be different, but u and v are always the same) [01:48] <BBB> ubitux: but we can ignore if you feel strongly and then just take as-is so we can finish off (i.e. I'll enable inter frame fate tests and we can start working on simd/mt) [01:49] <BBB> ubitux: because I think we're finished feature-wise now [01:58] <durandal11707> saste: pushed that probe thing? [01:58] <BBB> member:ubitux: yes, that, and off_uv being passed twice [01:58] <BBB> member:ubitux: off_u and off_v are always the same (and they are; y/uv can be different, but u and v are always the same) [01:58] <BBB> member:ubitux: but we can ignore if you feel strongly and then just take as-is so we can finish off (i.e. I'll enable inter frame fate tests and we can start working on simd/mt) [01:58] <BBB> ubitux: because I think we're finished feature-wise now [01:58] <BBB> ubitux: c/p from a few minutes ago [01:59] <ubitux> i'm ok for off_[uv] [01:59] <BBB> ok, so let's leave the stride in-place, I'll cherry-pick and then enable fate tests [01:59] <BBB> ffvp9 done \o/ [02:01] <ubitux> not really for ref linesizes [02:01] <ubitux> will fix in a moment, if my isp allows me [02:01] <ubitux> BBB: done, pushed [02:01] <ubitux> BBB: btw, did you see the decode error with the sample i pasted? [02:03] <ubitux> done, and broken \o/ [02:03] <ubitux> BBB: try ./ffplay -noframedrop -ss 29 ~/out.webm [02:04] <BBB> omg [02:04] <ubitux> also, i don't know if it's a problem with the encode [02:05] <ubitux> but there are some kind of freezes [02:05] <BBB> can vpxdec decode it? [02:05] <ubitux> i didn't look at all :) [02:05] <ubitux> there are various issues with that sample [02:05] <ubitux> but it might be related to libvpx as well [02:05] <BBB> it seems to decode fine to me [02:06] <cone-912> ffmpeg.git 03Michael Niedermayer 07master:259292f9d484: avcodec/mpegvideo: Dont incorrectly warn about missing keyframes [02:06] <BBB> (w/o your patch) [02:06] <ubitux> huh? [02:08] <ubitux> that reminds me i need to open a ticket for the ffplay lag after a seek [02:09] <BBB> yes libvpx and ffmpeg produce same output to me [02:09] <BBB> ffplay has issues, tha may be a problem [02:09] <BBB> but ffmpeg works [02:09] <BBB> no idea what that means or anything [02:09] <ubitux> in ffplay between 30-40 sec i get a buggy output [02:10] <ubitux> crazy blocks all over [02:10] <ubitux> like a complete mess [02:10] <durandal11707> what if you use non-ffplay ? [02:11] <ubitux> it's the same [02:11] <ubitux> http://ubitux.fr/pub/pics/_ffpv9-ffplay-etv.jpg [02:11] <ubitux> reproducible with ffmpeg [02:13] <ubitux> ./ffmpeg -ss 29 -i ~/out.webm -y out.y4m && ./ffplay out.y4m [02:17] <ubitux> the decode with libvpx is ok [02:33] <ubitux> BBB: cant reproduce? [02:53] <BBB> ubitux: checking [02:53] <BBB> oh yes I can [02:54] <BBB> ok looks like an actual bug [02:54] <BBB> that's cool [02:54] <durandal11707> *not [02:55] <durandal11707> why whenever i type make: doc/fate.txt is 'rebuild'? [02:56] <ubitux> BBB: it happens again at the end btw [02:56] <ubitux> BBB: what do you want me to work on tomorrow? [02:56] <ubitux> "cleanups"? [02:56] <BBB> simd, mt, other optimizations? [02:56] <BBB> cleanups is fine also [02:56] <ubitux> well, it's the end of the month [02:57] <ubitux> i'm starting my new job on monday [02:57] <durandal11707> but is pure c bitexact? [02:57] <BBB> durandal11707: there's some small bugs, I'm working them out, but the dsp code is bitexact yes [02:57] <ubitux> i dont mind focusing on one or two mt function though [02:57] <BBB> mt isn't function, mt is the whole thing [02:57] <BBB> frame-mt, that is [02:57] <ubitux> i meant simd sorry [02:57] <BBB> if you look at ffvp8, it's quite trivial to see how to do it [02:57] <BBB> ok [02:57] <ubitux> brain is off [02:57] <BBB> :) [02:58] <ubitux> fuck it's 3 am [02:58] <ubitux> i need my 12 hours sleep [02:58] <BBB> ok [02:58] <BBB> gnite [02:58] <ubitux> 'nite [03:40] <durandal11707> filter that takes 699kb in source code [03:43] <durandal11707> and its nearest neighbor 2-4 scaler [05:29] <cone-253> ffmpeg.git 03Michael Niedermayer 07master:20b965a1a43a: avcodec/ffv1dec: check global header version [05:29] <cone-253> ffmpeg.git 03Michael Niedermayer 07master:547d690d6760: ffv1dec: check that global parameters dont change in version 0/1 [05:39] <BBB> ubitux: pushed your commit (sorry took a while) [07:00] <BBB> ubitux: and yes I can reproduce the issue now, will fix (fixed another minor issue also, causing some artifacts around the borders) [09:35] <kurosu_> BBB: how is that vp9 decoder going? feature complete, or you are avoiding implementing some kind of unusual tool? (ie do you know what is the spec coverage) [09:37] <BBB> kurosu_: mostly there, small bugs here and there, some related to odd implementation details in libvpx that I obviously have to follow [09:38] <BBB> kurosu_: I'm avoiding resolution changing support ATM, maybe will do that later, but for now that's not a priority [09:38] <BBB> kurosu_: then again I'm not 100% sure if that's part of profile 0 or not :) [09:38] <BBB> I believe it is [09:38] <BBB> kurosu_: and obviously no simd/mt/anything yet, so it's somewhat slow [09:39] <kurosu_> I didn't find the openhevc implementation terribly sexy; hopefully ffvp9 has a better start in that domain [09:40] <BBB> what's openhevc? [09:40] <BBB> is that "ffhevc"? [09:40] <kurosu_> resolution changing? ie, you code some frames at eg half resolution? reminds me of mpeg4 reduced resolution, although I'm not sure where it belongs [09:40] <kurosu_> BBB: yes [09:40] <BBB> https://github.com/rbultje/ffmpeg/tree/vp9 [09:40] <BBB> ubitux is also working on it [09:41] <kurosu_> afaik, they developped in on their own inside of that French stuff then are now trying to push it on libav [09:41] <BBB> I thought smarter wrote it? [09:43] <kurosu_> wasn't that only the start of it ? actually I'm maybe imagining things and I don't actually know the connection [09:43] <kurosu_> afaik there's a French consortium trying to promote hevc through an "open" development, [09:43] <kurosu_> https://github.com/OpenHEVC [09:44] <BBB> they probably work together [09:44] <kurosu_> mostly a French university doing that work iirc [09:44] <BBB> I don't know [09:44] <kurosu_> I haven't quite followed either [09:45] <kurosu_> probably you can ask smarter I guess :D [09:45] <kurosu_> (if you wanted to know, not that I'm asking that you ask) [11:25] <cone-303> ffmpeg.git 03Michael Niedermayer 07release/2.0:c7ee4bc016e5: ffv1dec: check that global parameters dont change in version 0/1 [11:40] <cone-303> ffmpeg.git 03Timothy Gu 07master:40b8350b57ad: doc/encoders: reformat libmp3lame doc [11:40] <cone-303> ffmpeg.git 03Timothy Gu 07master:e45e72f5f89e: doc/encoders: reformat and add some clarification in libtwolame doc [11:57] <cone-303> ffmpeg.git 03Diego Biurrun 07master:a6b650118543: ppc: cosmetics: Consistently format CPU flag detection invocations [11:57] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:09c94b57ca2c: Merge commit 'a6b650118543e1580e872896d8976042b7c32d01' [11:57] <cone-303> ffmpeg.git 03Sean McGovern 07master:01a82f1dc544: ppc: don't return a value from a function declared void [12:01] <smarter> kurosu_: so, https://github.com/OpenHEVC/libav is the latest version of the hevc decoder, based on what I did during the gsoc last year, improved by me and others (mostly people from IETR-INSA) [12:06] <cone-303> ffmpeg.git 03Diego Biurrun 07master:79aec43ce813: x86: Add and use more convenience macros to check CPU extension availability [12:06] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:f0a35623826e: Merge commit '79aec43ce813a3e270743ca64fa3f31fa43df80b' [12:48] <cone-303> ffmpeg.git 03Diego Biurrun 07master:6369ba3c9cc7: x86: avcodec: Use convenience macros to check for CPU flags [12:48] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:8be0e2bd43d9: Merge commit '6369ba3c9cc74becfaad2a8882dff3dd3e7ae3c0' [12:48] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:c1913064e38c: avcodec/x86/vp8dsp: Fix cpu flag checks so they work [12:48] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:7fb758cd8ed0: avcodec/x86/lpc: Fix cpu flag checks so they work [12:54] <BBB> ubitux: fixed [12:57] <ubitux> cool :) [12:58] <cone-303> ffmpeg.git 03Diego Biurrun 07master:e998b56362c7: x86: avcodec: Consistently structure CPU extension initialization [12:58] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:62a6052974d8: Merge commit 'e998b56362c711701b3daa34e7b956e7126336f4' [12:59] <durandal_1707> michaelni: why you merged "convenience macros" if they do not work? [13:05] <michaelni> durandal_1707, why do you assume that they do not work ? [13:06] <durandal_1707> because i read 7fb758cd8ed08e4a37f10e25003953d13c68b8cd commit log [13:08] <michaelni> "avcodec/x86/lpc: Fix cpu flag checks so they work" Fixes the cpu flag checks in x86/lpc, how does that imply that the macors defined outside x86/lpc and used in several other places, dont work ? [13:09] <durandal_1707> it doesn't [13:09] <durandal_1707> but why it got broken... [13:10] <durandal_1707> they appars to work in some but not all cases... [13:21] <cone-303> ffmpeg.git 03Sean McGovern 07master:f1f728cbe4e8: ppc: don't return a value from a function declared void [13:21] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:05507348afa1: Merge remote-tracking branch 'qatar/master' [13:29] <durandal_1707> ubitux: were you serious about porting hqx filters? This code takes more than 10k lines of code. [13:30] <ubitux> i'm pretty sure it can be refactored in a few lines [13:31] <ubitux> maybe with a lut or two [13:31] <ubitux> someone raised a better one btw [13:31] <ubitux> a shader based, but cant remember the name [13:31] <ubitux> iirc another 2 letters filter unwebsearchable [13:31] <durandal_1707> shader one could only be faster [13:32] <ubitux> the algo was different [13:32] <durandal_1707> also there is algo that probably generated this nonsense, but where it is ... [13:35] <durandal_1707> about refactoring: switch takes most of lines, but i doubt i can refactor this... [13:35] <ubitux> it's a long work [13:36] <ubitux> durandal_1707: "xBR is better" [13:36] <durandal_1707> xBR ? [13:37] <durandal_1707> i think right solution is self generating code [13:37] <ubitux> like 5xBR [13:37] <durandal_1707> i got what 5x means... [13:38] <ubitux> https://github.com/libretro/common-shaders/tree/master/xbr [13:38] <ubitux> btw... https://github.com/libretro/common-shaders/blob/master/hqx/hq4x.cg [13:43] <durandal_1707> ok that is much less lines, but how(if possible at all) can i convert this to c? [13:46] <durandal_1707> well i dunno what half4 means [13:47] <durandal_1707> perhaps half is 16 bit float... [13:47] <nevcairiel> half4 is a 16-bit float, and a vector of 4 [13:47] <durandal_1707> nice [13:48] <nevcairiel> half4x4 would be a matrix [13:51] <durandal_1707> if i do this in pure c it would be extremly slow? [14:15] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:b05cd1ea7e45: ffv1dec: Check bits_per_raw_sample and colorspace for equality in ver 0/1 headers [14:33] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:4f5454d20130: avcodec/mpegvideo: reduce log level for messages about allocating frames. [14:45] <cone-303> ffmpeg.git 03Paul B Mahol 07master:48cd1037f661: cmdutils: silence warning about incompatible pointer types [15:04] <durandal_1707> michaelni: if you not gonna push license patch, i will [15:10] <michaelni> if someone pushes it, please ommit: [15:10] <michaelni> > - * Copyright (c) 2007 The Libav Project [15:10] <michaelni> > + * Copyright (c) 2007 The FFmpeg Project [15:10] <durandal_1707> ? [15:10] <michaelni> libavformat/network.c [15:11] <durandal_1707> why? [15:11] <michaelni> because you are not the copyright holder of the file ? [15:13] <durandal_1707> but for other files in that patch its also correct [15:13] <durandal_1707> and for others you changed when merging [15:14] <durandal_1707> so i'm very confused [15:16] <durandal_1707> or this was joke... [16:33] <cone-303> ffmpeg.git 03Carl Eugen Hoyos 07master:8fe1fb41ac28: Fix compilation with --disable-mmx. [16:54] <BBB> ubitux: ok all fate tests (without emu-edge) pass now [16:55] <BBB> ubitux: with emu-edge still some issues, these are relatively easy to fix, will do that later [16:55] <BBB> ubitux: simd/mt time now [17:16] <durandal_1707> michaelni: so you changed mind about license patch? [17:17] <michaelni> ? [17:18] <durandal_1707> read log [17:18] <durandal_1707> michaelni: stop playing games with me [17:27] <iive> michaelni: about networks.c, I don't think that libav project existed at 2007... [17:29] <michaelni> iive, you are probably correct, but it doesnt really make a difference does it ? i mean except for the entertainment value the contradiction has [17:30] <michaelni> durandal_1707, i read the log and i dont really know what you talk about. If its about the patch, i dint say that i will apply it nor did i say i wont, iam nt stoping you from applying it nor anyone else [17:33] <durandal_1707> well there is always revert [17:33] <iive> if I understand correctly, all "This file is part" are OK, as they are not part of the copyright notice. [17:34] <iive> i mean, this file is part of ffmpeg and libav... so it is true either way. [17:35] <durandal_1707> what about: "the ffmpeg project" vs "The FFmpeg Project" [17:35] Action: durandal_1707 have nothing better to do [17:37] <iive> as for network.c it should probably contain 2 copyright notices, 2007-2013 FFMpeg and whatever the first and last modification 2011-2013 libav. [17:37] <durandal_1707> its not FFMpeg but FFmpeg [17:38] <iive> of course it is [17:38] <durandal_1707> and really changing that each time someone refactor code is pita [17:39] <iive> Then 2007 FFmpeg, 2011 Libav ? [17:39] <durandal_1707> no way, i will commit what I have [17:40] <iive> having wrong copyright notice is indeed high amusement value, but in legal trouble it could turn into nightmare. [17:40] <iive> then... 2013 FFmpeg, 2013 Libav ? i guess you have merged stuff in it this year. [17:41] <iive> in/if [17:41] <durandal_1707> so i need to update this for every file in repo (except ffmpeg only code) [17:48] <iive> most of the files seem to have the copyright assigned to the person who wrote them. not to a project [17:49] <durandal_1707> yes, this patch just changes This file is part of .... [17:49] <durandal_1707> ... is free software; ... distributed in hope, etc... [17:50] <iive> yes, so no issue. as the discussion also seems to agree. [17:50] <iive> michael said he won't commit it, but he won't oppose it. If i got it right. and today i don't get a lot of things right. [17:51] <durandal_1707> you are not alone [17:55] <cone-303> ffmpeg.git 03Thilo Borgmann 07master:d814a839ac11: Reinstate proper FFmpeg license for all files. [18:51] <durandal_1707> michaelni: i guess adding support for yuva/gbr(a) to snow is now trivial? [19:50] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:7b47d7f75e6f: avcodec/pngdec: Fix padded alloc code with threads [19:50] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:60fed98e6347: avcodec/pngdec: fix last_row_size type [20:24] <cone-303> ffmpeg.git 03Paul B Mahol 07master:ea3ce0085921: wnv1: remove unused avctx from codec private context [20:24] <cone-303> ffmpeg.git 03Paul B Mahol 07master:057dce5f21cd: kgv1dec: make decoder independent of sizeof(AVFrame) [20:24] <cone-303> ffmpeg.git 03Paul B Mahol 07master:c04268447627: kgv1dec: remove unused avctx from codec private context [21:43] <durandal_1707> why this again [21:45] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:5c504e4df7b4: vformat/subtitles: check av_copy_packets return code [21:45] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:6e1b1a27a403: avcodec/avpacket: Use av_free_packet() in error cleanups [22:49] <cone-303> ffmpeg.git 03Lukasz Marek 07master:0b46d6f3efa7: lavu/bprint: add append buffer function [23:23] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:86736f59d6a5: avcodec/pngdsp: fix (un)signed type in end comparission [23:42] <cone-303> ffmpeg.git 03Michael Niedermayer 07release/0.11:2d945ac68f7a: avcodec/pngdsp: fix (un)signed type in end comparission [23:42] <cone-303> ffmpeg.git 03Michael Niedermayer 07release/1.0:5bd2b24db399: avcodec/pngdsp: fix (un)signed type in end comparission [23:42] <cone-303> ffmpeg.git 03Michael Niedermayer 07release/1.1:a2e7fd406c5b: avcodec/pngdsp: fix (un)signed type in end comparission [23:42] <cone-303> ffmpeg.git 03Michael Niedermayer 07release/1.2:ddce97c7b0e4: avcodec/pngdsp: fix (un)signed type in end comparission [23:42] <cone-303> ffmpeg.git 03Michael Niedermayer 07release/2.0:a4522ae516b5: avcodec/pngdsp: fix (un)signed type in end comparission [23:47] <cone-303> ffmpeg.git 03Michael Niedermayer 07master:454a11a1c9c6: avcodec/dsputil: fix signedness in sizeof() comparissions [00:00] --- Sat Aug 31 2013

1 0

ffmpeg.log.20130830
by burek 31 Aug '13

31 Aug '13

[00:05] <llogan> magnulu: you could have just put it in ~/bin [00:06] <llogan> and not messed with system stuff [00:21] <klaxa> well if he wants to have it available for all users, putting it in /usr/local/bin should be the right thing, no? [00:24] <llogan> it's one option. could use checkinstall or something if you want it in package management system [01:19] <CentRookie> hi there XD [01:19] <CentRookie> do you know if there is an option that prevents ffmpeg from overscaling? [01:19] <CentRookie> interpolating [01:21] <durandal11707> can you explain it in more detail what you are trying to solve? [01:21] <CentRookie> im running a script that converts videos but some have a lower resolution than the set resolution [01:22] <CentRookie> was wondering if there is a switch to prevent overscaling [01:22] <CentRookie> underscaling is fine [01:22] <CentRookie> *upscaling [01:22] <klaxa> i for one don't know of any, you could script it though [01:22] <CentRookie> well yeah, scripting is always a resolution [01:50] <CentRookie> oh, found a way [01:51] <CentRookie> -vf scale="min(640\,iw):trunc(ow/a/2)*2" [01:55] <CentRookie> now how to get it to work with complex filter [01:57] <durandal11707> what complex filter? you confused me again [02:02] <CentRookie> the command works with simple filter [02:02] <CentRookie> but i use filter_complex [10:09] <kaictl> Okay, so I've been using ffmpeg to record my desktop lately, and I was wondering if there was a way to only capture a single window, and not just by size (in case it resizes or moves. the output size is held at 852x480). [10:15] <kaictl> (using linux, not windows. that probably matters.) [11:07] <relaxed> kaictl: you expect it to track a window? no [11:08] <kaictl> relaxed: meh. not a dealbreaker. is there any way to manipulate ffmpeg while it is running? pause the video or audio input/output? [11:10] <relaxed> I don't think so [11:11] <relaxed> there's ctrl+z [11:12] <relaxed> then `fg` to start it again [11:14] <kaictl> true, but i'd like to use this for streaming and have some way to just pause the video output without stopping it, just streaming the last image if possible. [11:16] <kaictl> might be time to just grab a separate monitor for this kind of stuff. [11:56] <CentRookie> nice [11:57] <CentRookie> is there a way to post a certain useful command line and put it as an example on the wiki page? [12:34] <relaxed> example? [12:36] <zap0> sample? [12:43] <CentRookie> a line for auto resizing without upscale in filter_complex [12:43] <CentRookie> 844 is the destined output width, but if the source's width is smaller, it will keep the original width [12:43] <CentRookie> ffmpeg -i /test/test.mp4 -filter_complex scale='min(844\,iw):(trunc(ow/a/2))*2' output.mp4 -y [12:44] <CentRookie> hehe, there is a frowny on my screen, but i hope you see the code correctly [12:45] <CentRookie> it is such an important function, but there is not a single site that covers this line in complex filter [12:45] <CentRookie> only in -vf [12:45] <zap0> write a blog. describe it, give keywords. google will index it if it's quite unique. [12:46] <CentRookie> i just think it might be referenced somewhere ppl will actually read [12:48] <relaxed> shouldn't you scale up and pad to maintain the aspect ratio? [12:50] <CentRookie> it does by a [12:59] <kode101> so i have an mp4 and a wav file, and they're roughly the same len [12:59] <kode101> the mp4 has an audio track i no longer need [12:59] <kode101> do i convert the wav to aac then mux, or do i do something else [12:59] <kode101> goal is the mp4 with the new wav as its audio track [13:55] <Mavrik> kode101, you can just take both as input, encode audio to aac and mux it with a single command [13:55] <Mavrik> use "map" to properly map streams [13:55] <kode101> nice [13:56] <kode101> what the command [13:57] <Mavrik> I just told you. [13:57] <Mavrik> use "-i" twice and use map to map streams to output [13:58] <Mavrik> video + new audio [13:58] <kode101> ffmpeg -i haha [13:58] <kode101> silly [13:58] <kode101> i thought mux was a thing [13:58] <kode101> its not, its just there [13:58] <kode101> what so do i need to map? [13:59] <kode101> i tried [13:59] <kode101> ffmpeg -i input.mp4 -i input.wav -vcodec copy -acodec aac output.mp4 [14:00] <Mavrik> kode101, yes, you need to use map to tell ffmpeg to use video track from first output and audio track from the second [14:00] <Mavrik> and to ignore the audio track from first input :) [14:00] <relaxed> kode101: man ffmpeg|less +/^'STREAM SELECTION' [14:00] <kode101> ive removed the audio from the file [14:00] <kode101> so can skip mapping here no? [14:00] <kode101> i have a video with no audio and audio with no video so ffmpeg just works it out [14:01] <Mavrik> probably yeah [14:01] <Mavrik> check the output, the mapping is always written out at the start [14:02] <relaxed> learning would save you headaches later [14:02] <kode101> tru [14:02] <kode101> thanks yo [14:02] <relaxed> read about -map [16:34] <XwZ> hi, I'm using ffmpeg to grab images from an rtsp camera with limited slots. When I try to disconnect from the camera, the TEARSDOWN message is sent to the camera but the slot is not freed [16:35] <XwZ> i'm using avformat_open_input and avformat_close_input to open and close the stream [16:36] <XwZ> do you have an idea of the problem ? if I use avplay or ffplay it works fine (frustrating) [19:26] <DasMoeh> hey, is there a function in libav to skip all frames till next keyframe? and how many frames are normally between to keyframes? can i get this information somehow? [19:47] <vilalian> DasMoeh: Check out the av_seek_frame/av_seek_file functions. How many frames and whether it is consistent between depends on what codec you are in, but look at gop_size and keyint_min variables on your AVCodecContext. [19:49] <DasMoeh> ok thanks, when i use av_seek_frame with the timestamp from my actual frame i get the next keyframe? [19:53] <vilalian> DasMoeh: I've seen it vary in behavior a bit on different codecs, but it usually (should be always?) returns the first keyframe before the requested time, so you need some idea of where you're trying to jump to. [19:53] <vilalian> DasMoeh: if you're just trying to get to the next keyframe I would just read frames from where I am until I got there. [19:54] <DasMoeh> That's what i'm doing now. But i'll safe the time needed for decoding each frame. [19:55] <DasMoeh> *want safe [20:05] <DasMoeh> mh, gop_size = 12, keyint_min = 25... but description says keyint_min is the minimum GOP size? [20:38] <Mavrik> DasMoeh, slightly late but here is goes [20:39] <Mavrik> first the bad news: there is no reason for the keyframes to be equally distributed [20:39] <Mavrik> in more advanced formats they certanly aren't [20:45] <DasMoeh> Mavrik: thx [20:45] <Mavrik> also, to decode only keyframes [20:45] <Mavrik> run the av_read_packet loop and check for the keyframe flag on the packet [20:49] <vilalian> I'm reading a video stream and encoding to disk as h264. Reading that file in and writing it out again with identical codec parameters (verified that libx264 is logging that the bitrate requested is the same) results in a file that is much higher bitrate. Original (and correct) ~600k bitrate file is rewritten at ~1500k. ffprobe shows no difference in the files other than the frame sizes (no extra streams or anything). I'm a bit stumped [20:49] <vilalian> to what could be causing the size inflation. [21:05] <DasMoeh> <Mavrik> run the av_read_packet loop and check for the keyframe flag on the packet <-- thanks, that's a good idea. Didn't found that flag befor. [21:05] <Mavrik> yeah, there's a AV_PKT_FLAG_KEY or something like that [21:06] <DasMoeh> yes now i know what i'm looking for and found the flag [00:00] --- Sat Aug 31 2013

1 0

ffmpeg-devel.log.20130829
by burek 30 Aug '13

30 Aug '13

[20:17] <burek> durandal_1707, i wasnt aware it was offline :/ [20:18] <llogan> what happened to it? [20:18] <burek> im looking into logs now [20:20] <burek> a ddos it seems [20:20] <llogan> SEA! by allah's beard! [20:22] <durandal_1707> burek: by who? [20:24] <burek> who cares :) [20:25] <durandal_1707> i do [20:25] <burek> no harm done :) [20:25] <burek> except for few days of outage :) [20:26] <durandal_1707> who did it? [20:26] <burek> internet gremlins :) [20:27] <llogan> how can you tell anyway with botnet rentals, windows zombies, etc? [20:27] <durandal_1707> no if you tell, i will give it back [20:27] <burek> im so desperate now :) [20:28] <burek> why does it matter anyway :) [20:28] <durandal_1707> what if it happens again? [20:28] <burek> the world will stop [20:28] <durandal_1707> yes, i can not see when someone adds spam [20:29] <durandal_1707> and refreshing trac page is waste of my time [20:29] <burek> use rss feed? [20:29] <durandal_1707> perhaps you should add something that will detect ddos attacks and ignore such messages? [20:30] <burek> i will some time in the future [20:31] <llogan> durandal_1707: i had the assumption that you didn't like fflogger [20:32] <durandal_1707> assumptions are bad [20:32] <llogan> some comment(s) in the past that I can't remember or maybe I am imagining about log spamming channel [20:34] <burek> llogan, btw, thanks for your help on the forum [20:35] <burek> i was really busy lately and couldnt afford to visit forum more often [20:35] <burek> you really helped a lot there [20:37] <llogan> burek: no problem. it's part of my daily morning procrastination routine. [20:37] <burek> i was thinking about migrating the forum to the Q&A type of a web site [20:37] <burek> which might be more efficient i guess [20:37] <burek> but im not sure if its worth the effort [20:38] <llogan> i think the forum is fine as is. users can go to superuser if they want the other format. there are some informed answerers there [20:40] <burek> ok [20:55] <llogan> ubitux: what do you think about webvtt output corresponding with select filter scene? users could use it for "thumbnail previews" with supported players [22:58] <cone-912> ffmpeg.git 03Michael Niedermayer 07master:811d58e08386: avcodec/utils: support non edge emu for grayscale [22:58] <cone-912> ffmpeg.git 03Michael Niedermayer 07master:ec120efaa90a: doc/snow: add gray/alpha/gbr [22:58] <cone-912> ffmpeg.git 03Michael Niedermayer 07master:c4224fff1b0a: avcodec/snow: gray support [23:18] <durandal11707> michaelni: check return value of av_frame_alloc? [23:30] <cone-912> ffmpeg.git 03Michael Niedermayer 07master:24b4e6c373f8: snow: Check av_frame_alloc() failures [00:00] --- Fri Aug 30 2013

1 0

ffmpeg.log.20130829
by burek 30 Aug '13

30 Aug '13

[21:07] <hans_henrik> any1 have any idea what a "GARMIN BITMAP" is, or how to convert a jpg to 1? :p [21:08] <hans_henrik> (viewing it with a hex editor suggests its called GARMIN BITMAP") [21:08] <llogan> i do not, but please share a sample. [21:09] <kode101> what should i use to concat a bunch of ts files? [21:10] <hans_henrik> http://4chanx.org/hans/temp/truck_blue.srf [21:11] <hans_henrik> kode101, idk, but best guess, -i concat:"input1.ts|input2.ts|input3.ts" [21:12] <improviser> porn? [21:12] <hans_henrik> improviser, yes, blue truck porn [21:13] <hans_henrik> actually i think its this "truck" in this image http://www.gpsmagazine.com/assets/review-nuvi465t/trucking13.jpg [21:16] <llogan> do you know of anything that can decode the file? [21:17] <llogan> othan than some expensive garmin shit [21:18] <hans_henrik> llogan, no not really :/ [21:22] <hans_henrik> ohhhhh, i found it, http://techmods.net/nuvi/ , has srf2png and png2srf ^^ | ping llogan [21:22] <hans_henrik> and a nice "SRF file format details" [21:22] <hans_henrik> so som1 has taken the time to figure out the format, most of it, anyway ^^ [21:24] <llogan> that's a good start i guess. also "garmin garage" site has a bunch more srt files if you want more [21:26] <hans_henrik> llogan, garmin garage? where is that? [21:30] <llogan> hans_henrik: http://www8.garmin.com/vehicles/product.html?vName=Homer [21:30] <kode101> what formats should i mux audio and video into for further processing? [21:33] <llogan> what kind of processing? [21:35] <kode101> python [21:35] <kode101> im stretching the audio and video with different algos [21:35] <kode101> then remuxing [21:36] <kode101> so i think i need to turn an mp4 into a wav and as pure a video file as i can [21:36] <durandal_1707> kode101: use concat demuxer [22:51] <magnulu> hi.. I am on squeeze-amd64 - anyone got experience on building ffmpeg-2.0.1 from source? [22:51] <magnulu> OR can provide me with an installable .deb? :-) [22:51] <klaxa> you can just grab a static binary [22:51] <klaxa> or read the ubuntu compilation guide (it also works for debian presumably) [22:52] <magnulu> ok.. I'll give the first option a try, thanks a lot.. [22:54] <magnulu> would it be a good idea to remove the old version with apt-get before using the new one? [22:57] <klaxa> hmm... not if you have software that depends on the shared libraries [22:57] <klaxa> if you use the static build, you shouldn't worry about deinstalling stuff anyway [22:58] <magnulu> just replace the binaries in /usr/bin then? and copy the presets to /usr/share/ffmpeg i presume? [22:59] <klaxa> oh god no, if anything put it in /usr/local/bin [23:00] <klaxa> putting things into /usr/bin is... well you shouldn't do it, because it might mess with some automatic installation stuff [23:00] <magnulu> ok.. [23:21] <magnulu> I ended up uninstalling the old version of ffmpeg (nothing else was removed) and placed the binaries in /usr/local/bin and the presets in /usr/local/share/ffmpeg.. testing with the mmac script now [23:21] <magnulu> thanks for your help, highly appreciated! [00:00] --- Fri Aug 30 2013

1 0

ffmpeg-devel.log.20130824
by burek 25 Aug '13

25 Aug '13

[00:02] <cone-662> ffmpeg.git 03Paul B Mahol 07master:2a7545951926: pngdec: do not release buffer on failure instead report full progress [00:21] <wm4> lol that youtube patch [00:22] <wm4> converted from a perl script [01:11] <cone-662> ffmpeg.git 03Michael Niedermayer 07master:3941a4f5c2a4: snowenc: change a bunch of assert() to av_assert() [01:11] <cone-662> ffmpeg.git 03Michael Niedermayer 07master:5cc8b816875d: mpeg4videodec: fix GEOV/GEOX fliping [01:22] <durandal_1707> oh why lavc needs to do flipping? [01:22] <cone-662> ffmpeg.git 03Paul B Mahol 07master:83b915d495ea: truemotion1: use av_freep() [01:22] <cone-662> ffmpeg.git 03Paul B Mahol 07master:b8ff4f5ea3d3: truemotion1: check av_fast_malloc() return value [01:26] Action: durandal_1707 wonders why vp3.c does not have K&R commit [01:27] <ubitux> better delete it [01:27] <ubitux> "it's too ugly" [01:27] <durandal_1707> why, only theora part looks ugly [01:27] <ubitux> dunno [01:28] <ubitux> havent opened the file [02:27] <BBB> ubitux: uses two predictors instead of one (like B-frame prediction in classic MPEG-like codecs; in vp9, this is a per-partition/per-block decision) [02:27] <BBB> ubitux: and the references to predict from are stored in b->ref[0] and b->ref[1] [03:12] <ubitux> BBB: you mean if b->comp is set, it means picking from 2 different references instead of one? [03:12] <BBB> yes [03:13] <BBB> we have dsp->mc[6-log2sz][filter][0=put,1=avg][xfilter][yfilter]() [03:13] <BBB> 0=put is first reference, 1=avg is second reference [03:14] <ubitux> if there is a second one it's doing an average between the two i guess? [03:14] <ubitux> (50-50?) [03:14] <BBB> yes [03:15] <BBB> think of it as two separate predictor filters (with separate MVs), and then averaging between the two as (a+b+1)>>1 [03:15] <BBB> in simd, it's faster to do one store, then a load, avg, store, rather than two stores, two loads, avg, store [03:15] <BBB> so that's why it's implemented like this [03:15] <BBB> but it does the same thing [03:38] <ubitux> BBB: btw, mmh, there is no 10-bit with vp9? [03:38] <ubitux> (yeah totally unrelated i know) [03:39] <BBB> no [03:40] <ubitux> do you know the reasons? [03:45] <BBB> I think it's planned for a future extension profile or something along those lines [03:46] <ubitux> ah, cool, ok [03:46] <BBB> just like the current profile is 420 only and 3-plane only (yuv), but the codebase already supports 444 and alpha, which are also planned for stabilization in a future profile [03:47] <ubitux> yeah, ok [03:48] <ubitux> michaelni: please use FF_CEIL_RSHIFT :( [03:49] <ubitux> (in reference to 5cc8b8168) [03:50] <michaelni> ubitux, you could write a cronjob that does a checkout, replaces them and commits it and sends me a pull req :) [03:50] <ubitux> yeah that sounds like the most optimal way :D [03:53] <cone-662> ffmpeg.git 03Michael Niedermayer 07master:9a271a9368ea: jpeg2000: check log2_cblk dimensions [04:15] <cone-662> ffmpeg.git 03Michael Niedermayer 07master:b99d3613cfdb: avcodec/h263dec: use FF_CEIL_RSHIFT() [04:25] <BBB> ubitux: also I think I may soon have all keyframes decoding correctly in terms of giving same output as libvpx [04:29] <kierank> 02:38:53 <"ubitux> BBB: btw, mmh, there is no 10-bit with vp9? --> not much point [06:50] <BBB> ubitux: ok, all non-emuedge fate tests I added pass now; still a few emuedge bugs remaining, will fix those next [06:50] <BBB> ubitux: any progress on your end? [11:21] <cone-129> ffmpeg.git 03Diego Biurrun 07master:8506ff97c9ea: vp56: Mark VP6-only optimizations as such. [11:21] <cone-129> ffmpeg.git 03Michael Niedermayer 07master:f9418d156fe6: Merge commit '8506ff97c9ea4a1f52983497ecf8d4ef193403a9' [11:27] <BBB> what was the method again to use valgrind for all fate tests? [11:30] <cone-129> ffmpeg.git 03Diego Biurrun 07master:84784c297fe6: libfdk-aacdec: formatting cosmetics [11:30] <cone-129> ffmpeg.git 03Michael Niedermayer 07master:edf6fb64e07c: Merge commit '84784c297fe6a6e538a7e111dcdbd8b893c2d275' [11:40] <michaelni> BBB, --valgrind=valgrind [11:43] <cone-129> ffmpeg.git 03Diego Biurrun 07master:f407856968dc: arm: h264chroma: Do not compile h264_chroma_mc* dependent on h264 decoder [11:43] <cone-129> ffmpeg.git 03Michael Niedermayer 07master:6067186f3a76: Merge remote-tracking branch 'qatar/master' [14:50] <wm4> ubitux: the webvtt patch has bogus OOM handling; it should do av_free_packet(pkt) instead of av_free(pkt), right? [14:51] <wm4> or, that should be additional [14:52] <wm4> it also doesn't check av_malloc return values [15:50] Action: BBB pokes ubitux [16:09] <ubitux> BBB: yeah i think i start to get it, but it takes me some time [16:09] <BBB> ok [16:09] <ubitux> wm4: no idea, will look later eventually [16:09] <ubitux> BBB: why the 4 inter pred modes start at 10? [16:10] <BBB> so they can be used in the same context as intra modes (of which there are 10) [16:10] <BBB> if we can somehow separate that, then I'm fine with removing that [16:12] <ubitux> ah ok; it would have been more obvious if that enum was in vp9.h below its brother (and eventually using N_INTRA_PRED_MODES as start value) [16:13] <ubitux> i see 16 intra pred modes though [16:13] <ubitux> (maybe that was supposed to be 0x10?) [16:19] <ubitux> BBB: am i missing sth obvious? [16:23] <ubitux> on a side note, it would be nice to have some prefix for inter & intra pred mode macro [16:27] <BBB> ubitux: 10 are coded, the rest are inferred [16:27] <BBB> naming is just to be consistent with h264pred.h and vp8.h [16:28] <BBB> eventually I'd like to merge these kind of things [16:28] <BBB> ubitux: basically there's 10 modes [16:28] <BBB> ubitux: but some modes mean something different depending on which edges are available [16:28] <BBB> ubitux: e.g. DC on top row means something else than DC in the middle of the image [16:29] <ubitux> oh, ok [16:29] <BBB> so for code simlicity (simd) reasons, we do that branch in the calling code, and thus add extra intra modes which we use if the coded mode is dc but not all edges are available [16:29] <BBB> so that's dc_top, dc_left, dc_128 [16:29] <BBB> then dc_127/dc_129 are just variations thereof (vp9 defines dc with no edges as 128, but the top edge is 127 and the left edge is 129 for all other intra pred modes [16:30] <BBB> vp8 did that too [16:30] <ubitux> ok [16:30] <BBB> so b->mode is only 0-9 [16:30] <BBB> but the intra pred simd functions can be called with 0-14 [16:38] <BBB> my todo list is shrinking quite quickly now [16:38] <BBB> nice [16:40] <ubitux> BBB: i don't get the difference between NEAREST and NEAR mv [16:40] <ubitux> is it related to the reference used? [16:41] <ubitux> zeromv is suppose that's the latest one (no further sub division), and newmv is the opposite (a subdivision, so in case of 8x8 the latest possible 4 partitions?) [16:41] <BBB> it's what motion vector you use [16:41] <ubitux> s/is supposed/i suppose/ [16:41] <BBB> ZEROMV means the motion vector is y=0, x=0 [16:41] <BBB> NEWMV means we code an actual motion vector difference in the bitstream using NEARESTMV as a reference [16:42] <BBB> NEARESTMV is whatever is the first motion vector returned from find_ref_mvs() [16:42] <BBB> NEARMV is the second motion vector returned from find_ref_mvs() [16:43] <ubitux> ah. [16:43] <BBB> the first/second motion vector simply are unique MVs found in surrounding blocks [16:43] <BBB> first the left neihgbour, then the top, then the topleft, then those further away [16:43] <BBB> see the table in find_ref_mvs() [16:43] <BBB> we first try to find one that uses the same reference as us [16:44] <BBB> if that didn't work, we look for ones that use a different reference [16:44] <BBB> larger blocks use a larger neighbourhood search radius, so to say, so they more sparsely look in blocks that are located further away [16:45] <BBB> then once you know your motion vector, the rest is straightforward [16:46] Action: BBB goes to bed [16:47] <ubitux> ok thx, 'night :) [19:06] <cone-549> ffmpeg.git 03Michael Niedermayer 07master:7495186fd49f: avcodec/h263dec: fix aspect of lead h263 EHC [20:00] <cone-549> ffmpeg.git 03Michael Niedermayer 07master:5c6a58746b41: ffplay: make next_nb_channels[] static const [22:13] <ubitux> BBB: ping [23:13] <cone-549> ffmpeg.git 03Michael Niedermayer 07master:f55a7ba0376d: avcodec/ituh263dec: detect and warn about RTP [23:28] <cone-549> ffmpeg.git 03Michael Niedermayer 07master:88909beca39e: avcodec/movenc: move chapter_properties under the #if of the code that uses it [23:31] <ubitux> michaelni: avformat* [23:36] <michaelni> ^^;; [00:00] --- Sun Aug 25 2013

1 0

ffmpeg.log.20130824
by burek 25 Aug '13

25 Aug '13

[01:31] <CentRookie> hello all :) [01:32] <CentRookie> Does somebody of your knowledge encoders tell me what it is that I have to change in order to fix MB-tree frametype 0 doesn't match actual frametype 2 [01:33] <CentRookie> I try to run 1pass with ultrafast but obviously it doesnt work well with my 2nd pass settings, but first pass works well with veryfast [01:39] <CentRookie> hm, i guess it cant be fixed [01:54] <CentRookie> so silent [01:54] <CentRookie> maybe they are all ghosts [01:54] <CentRookie> does somebody have experience with multi file 2 pass encoding? [02:10] <mjiig> when i try to make a timelapse video using the command "ffmpeg -y -r 5.0 -f image2 -i /home/angus/glapse/%09d.jpg /home/angus/glapse/timelapse.mp4.avi" i get a video that's much much whiter than the original images [02:10] <mjiig> is there anyway i can avoid this? [02:19] <CentRookie> -chromaoffset -2 [02:22] <mjiig> that doesn't change anything [02:22] <CentRookie> :( [02:23] <CentRookie> i doubt it actually parses the images like videos, since you have no interframes [02:23] <CentRookie> so there isnt much you can do with filters [02:26] <mjiig> ugh, just got a friend to look at it and it's fine for him, so i think it's probably the player that's screwing not ffmpeg [02:27] <mjiig> sorry for the time waster [02:41] <rafael2k> people, I'm with trouble with decoding HD h.264 using crystalhd hw accel. w/ gst-crystal all works fine, any recent issue related to crystalhd h.264 dec? [02:41] <rafael2k> btw, I'm using bcm 70012 hw decoder [05:59] <elkng> how much RAM is enough for ffmpeg if you encode video ? if there is 8GB RAM will ffmpeg use it all ? what if there is 32GB RAM ? [06:00] <relaxed> elkng: you can use `top` to view its memory usage. [06:02] <relaxed> I doubt it would use that much memory unless your frames were *HUGE* [06:02] <relaxed> like xbox [06:02] <klaxa> lol [06:05] <elkng> I'm converting 1920x800 to 650x350 and top shows it uses "81MB RSS" so its enough about 128MB ? [06:06] <elkng> seem like the only thing that benefit from huge amount of RAM is kernel compilation [06:06] <klaxa> virtualization :x [06:06] <klaxa> that uses quite some ram too [06:06] <elkng> relaxed: "like xbox", what do you mean by that ? [06:07] <elkng> klaxa: blender [06:07] <klaxa> that too [06:07] <klaxa> also: http://knowyourmeme.com/memes/huge-like-xbox-hueg-like-xbox [06:07] <relaxed> "huge like xbox" was a meme [06:07] <klaxa> re: xbox hueg [06:08] <elkng> or gimp could consume much RAM [06:08] <elkng> or converters like imagemagic [06:09] <elkng> actually I remember run imagemagic to convert some image, it was about 10000x5000 or so on a machine with 1GB RAM and it was killed by system because of "out of memory" issues, and I wasn't been able to convert those image, even with 1GB RAM [06:11] <elkng> so seems like all ffmpeg need RAM for is for one current frame ? so it covert video frame by frame and the most needed amount of RAM for ffmpeg procees depends on size of one single frame ? [06:12] <klaxa> yeah imagemagic eats ram like pacman eats pills [06:12] <klaxa> *imagemagick [06:12] <klaxa> i think it also depends on the codec? [06:12] <elkng> I though there is some sort of optimization for ffmpeg, some kind of cashing or so, the more RAM the faster procees [06:12] <klaxa> x264 needs a lot of ram for m-trees i think although i have no idea how they work exactly [06:12] <klaxa> but i think they optimize P and B frames? [06:13] <klaxa> hmm... encoding one frame will probably take significantly longer than reading a frame from the disc [06:13] <klaxa> *disk [06:13] <elkng> I'm converting now from "1920x800 x264" to "650x350 mpeg4" [06:15] <elkng> and it eats 81MB RAM [06:16] <klaxa> you could put the input file to /tmp if it is mounted as a ramfs and output to /tmp too [06:16] <klaxa> that would speed up reading the file from the filesystem [06:16] <klaxa> since it is actually in ram already [06:16] <klaxa> but i don't think it will speed up the process really UNLESS, your bottleneck is your disk io [06:16] <klaxa> which i highly doubt [07:14] <elkng> I have atom 1.6 [07:15] <klaxa> optimizing on the ram-side is probably misplaced then :P [09:14] <liquidmetal> I have a general codec question [09:14] <liquidmetal> http://developer.android.com/reference/android/media/MediaCodecInfo.CodecCa… [09:14] <liquidmetal> This 'color format' says it's packed AND semi-planar. [09:14] <liquidmetal> how is that possible [09:14] <liquidmetal> ? [11:27] <JEEB> liquidmetal, I wonder if that's something like NV12 [11:28] <JEEB> (not related to nvidia even with that name) [11:28] <liquidmetal> JEEB, I understand planar formats [11:28] <liquidmetal> I understand packed formats [11:28] <liquidmetal> aren't they mutually exclusive? [11:29] <JEEB> well, if you have one plane as planar and two of the other planes as packed together [11:29] <liquidmetal> (okay, I kinda understand those two terms) [11:29] <JEEB> although to be honest that thing doesn't document what it is so lol [11:30] <JEEB> liquidmetal, and yes [11:30] <JEEB> COLOR_TI_FormatYUV420PackedSemiPlanar is NV12 [11:30] <JEEB> from a random piece of code I found on the internet [11:30] <liquidmetal> ah! [11:30] <liquidmetal> So packed semi-planar means, one part is planar and the other is packed [11:31] <JEEB> somehing like that, the name in this case is rather ambiguous [11:31] <liquidmetal> Got it! [11:31] <liquidmetal> Now I need to figure out how to decode NV12 on android [11:43] <fschuetz> Why is there no avformat_close_output method? Am I missing something? [11:49] <fschuetz> I could really use some help with converting an audio file to another format. I have checked the examples and documentation, however each example seems to take a different approach to tasks like opening/closing files and writing to them. [11:50] <fschuetz> There also seems to be a general absence of symmetry in ffmpeg. Is there a particular reason, why opening and writing to files are so very different for input and output? [12:34] <fschuetz> A great start would be, if someone could explain to me, what open_audio in http://ffmpeg.org/doxygen/trunk/doc_2examples_2muxing_8c-example.html does [12:35] <fschuetz> I think everything it computes should be determined by output format and codec. [13:13] <liquidmetal> Looking at http://developer.android.com/reference/android/media/MediaCodecInfo.CodecCa… I found that there are two different color formats mentioned there: [13:13] <liquidmetal> COLOR_FormatYUV422SemiPlanar [13:13] <liquidmetal> COLOR_FormatYUV422PackedSemiPlanar [13:14] <liquidmetal> This makes me wonder - what's the difference between these two formats? [13:17] <blez> hello [13:17] <blez> is there a version of ffmpeg that supports hardware acceleration [13:18] <fschuetz> compile it [13:21] <Mavrik> blez, what is "hardware acceleration" for you? [13:21] <Mavrik> liquidmetal, I think line alignment [13:21] <blez> CUDA support for example? [13:21] <Mavrik> liquidmetal, but I haven't used it enough to be sure [13:21] <Mavrik> blez, for which part of transcoding process? [13:22] <liquidmetal> Mavrik, I'm sure the naming wouldn't be android specific - so there must be some documentation about this. [13:22] <Mavrik> anyway, no, there are no CUDA encoders in ffmpeg at the moment since they're silly [13:22] <liquidmetal> Any clue where I can find more about this? [13:22] <Mavrik> liquidmetal, why are you sure naming isn't android specific? [13:22] <Mavrik> liquidmetal, and you will find that info in SoC documentation [13:22] <Mavrik> so look at qualcomm, nvidia [13:23] <liquidmetal> SoC documentation? [13:23] <liquidmetal> Mavrik, sure because of faith in Android :) Which means I could be wrong [13:24] <liquidmetal> But I just want to know the difference between 'packed semiplanar' and 'semiplanar' [13:25] <blez> what about OpenCL? [13:31] <Mavrik> liquidmetal, and I'm telling you you need to consult documentation of a SoC that produces that kind of images [13:31] <Mavrik> since in Android doc it's not clearly specified [13:32] <Mavrik> liquidmetal, also, if you love youself just a little [13:32] <Mavrik> you'll forget about MediaCodec API for a few versions of Android more at least [13:32] <Mavrik> it's a horrible horrible mess [13:33] <liquidmetal> Mavrik, they now have tests for those APIs - so they'll stay consistent with newer version of android [13:34] <liquidmetal> Just that the % of devices they work on is limited - I don't care about that just yet [13:34] <Mavrik> liquidmetal, the problem is that pixel formats aren't defined [13:34] <Mavrik> and the format in which the devices expect frames varies wildly without a reliable way to check [13:35] <Mavrik> anyway, as I said, I think the non-packed YUV:4:2:2 has to be line aligned on 16-byte mark, but I'm not 100% sure [13:39] <Mavrik> liquidmetal, doh sorry, I'm talking crazy talk [13:39] <Mavrik> liquidmetal, "packed" formats have luma and chroma channels interleaved [13:39] <liquidmetal> Mavrik, how's that different from semiplanar? [13:40] <Mavrik> good question actually [13:41] <Mavrik> that's also what doesn't make sense to me [14:18] <CentRookie> hi [14:18] <CentRookie> im trying to run multiple 2 pass encodings simultanously [14:19] <CentRookie> but the mbtrees keep overwriting each other [14:19] <CentRookie> is there a way to define the mbtree names? [14:21] <relaxed> -passlogfile [14:22] <CentRookie> it also applies to mbtree? [14:22] <CentRookie> and i only need that for first pass encoding right [14:23] <relaxed> yes and no [14:23] <CentRookie> thanks :) [14:23] <relaxed> the second pass needs the name too [14:23] <CentRookie> hmm [14:23] <CentRookie> do i assign it as input somehow? [14:24] <CentRookie> or do i say for 2nd pass -passlogfile video02 [14:24] <relaxed> the same name for both [14:24] <relaxed> passes [14:24] <CentRookie> ok [14:24] <CentRookie> helpful like always :) [14:25] <CentRookie> wish there was a parameter in ffmpeg to auto remove logtrees after last pass [14:27] <relaxed> it's trivial to script [14:28] <CentRookie> yup [14:28] <CentRookie> still it is so useful [14:28] <CentRookie> if there was such an option i mean [15:36] <fschuetz> i am trying to convert audio. decode works and got packet, however avcodec_encode_audio2 crashes. Is there something I need to do with the decoded frame, before I can hand it to avcodec_encode_audio2? Code is here: http://pastebin.com/UkQywDHF [15:51] <fscz> in the examples/muxing.c write_audio_frame there is a resampling step. Is resampling necessary, if you get the frame from avcodec_decode_audio4? [15:59] <fscz> Would be smarter, if this step was integrated in the respective encode/decode functions [16:08] <hackeron> hey, I'm trying to record audio from alsa with: ffmpeg -f alsa -ac 1 -i hw:1,0 -dn -vn -codec:a libfdk_aac -flags +qscale -ar 44100 -y test.mkv -- but I am getting: "[alsa @ 0x2333bc0] cannot set sample format 0x10000 2 (Invalid argument)" -- the sample format used by arecord is "S32_LE" but I can't see it in the list of -sample_fmts -- any ideas? [16:10] <klaxa> just try s32p maybe? :x [16:11] <klaxa> and s32 [16:11] <klaxa> one of them might sound correct [16:12] <hackeron> klaxa: I'm not sure where to put it, but I tried: ffmpeg -f alsa -ac 1 -sample_fmt s32 -i hw:1,0 -sample_fmt s32 -dn -vn -codec:a libfdk_aac -flags +qscale -ar 44100 -y test.mkv -- and it throws the same: "[alsa @ 0xe74cc0] cannot set sample format 0x10000 2 (Invalid argument)" [16:12] <hackeron> am I putting it in the wrong place? [16:12] <klaxa> no i think that place is right [16:12] <klaxa> wait [16:12] <hackeron> no matter what sample format I use, it always says: "cannot set sample format 0x10000 2" [16:12] <klaxa> remove the second one [16:12] <klaxa> hmm yeah weird [16:13] <hackeron> same thing with: ffmpeg -f alsa -ac 1 -sample_fmt s32 -i hw:1,0 -dn -vn -codec:a libfdk_aac -flags +qscale -ar 44100 -y test.mkv [16:13] <hackeron> ffmpeg -f alsa -ac 1 -sample_fmt dblp -i hw:1,0 -dn -vn -codec:a libfdk_aac -flags +qscale -ar 44100 -y test.mkv --- also returns: "cannot set sample format 0x10000 2 (Invalid argument)" :/ [16:14] <klaxa> can you try: ffmpeg -f alsa -i hw:1,0 test.wav ? [16:15] <klaxa> and if that works add more and more arguments to the command line [16:15] <hackeron> yep: [alsa @ 0xd738c0] cannot set sample format 0x10000 2 (Invalid argument) [16:15] <hackeron> hw:1,0: Input/output error [16:15] <klaxa> so [16:15] <klaxa> input output error sounds quite suspicioius [16:15] <klaxa> second hardware soundcard, first device? [16:16] <hackeron> this is the format from arecord: http://pastie.org/8265575 [16:16] <hackeron> the output I mean [16:16] <hackeron> so the hardware is fine :/ [16:17] <hackeron> and the resulting wav file plays just fine [16:17] <klaxa> hmm hmm [16:20] <hackeron> klaxa: the full ffmpeg debug output: http://pastie.org/8265584 [16:21] <klaxa> hmm dunno wait for someone who knows about this to show up :/ [16:23] <hackeron> also, check this out: http://pastie.org/8265588 -- so no matter what sample format I try, it always says: "cannot set sample format 0x10000 2" - hmmm :/ [16:44] <hackeron> for a workaround I'm recording through dsnoop which is working :) [16:45] <hackeron> but the audio input has 12 channels - how do I record just channel 6 for instance? -- I'm trying ffmpeg -f alsa -ac 12 -i plug:capt -ar 44100 -map_channel 0.0.5 -y test.wav -- but I'm getting silence (probably channel 1) [16:45] <hackeron> any ideas? [16:55] <zap0> 0.0.5 ? [16:55] <hackeron> is that wrong? [16:56] <hackeron> I tried this: ffmpeg -f alsa -ac 12 -i plug:capt -ar 44100 -map_channel 0.0.0 -y test0.wav -map_channel 0.0.1 -y test1.wav -map_channel 0.0.2 -y test2.wav -map_channel 0.0.3 -y test3.wav -map_channel 0.0.4 -y test4.wav -map_channel 0.0.5 -y test5.wav -map_channel 0.0.6 -y test6.wav -map_channel 0.0.7 -y test7.wav -map_channel 0.0.8 -y test8.wav -map_channel 0.0.9 -y test9.wav -map_channel 0.0.10 -y test10.wav -map_channel 0.0.11 -y test11.wav [16:56] <hackeron> according to the manual, it should record each channel to a separate file [16:57] <hackeron> but every one of the files is just complete silence :/ -- if I do just ffmpeg -f alsa -ac 12 -i plug:capt -ar 44100 -y test.wav -- I get a 12 channel wav file and there's sound on 8 of the 12 channels [16:57] <hackeron> (only 8 channels have microphones so this is expected) [16:59] <hackeron> the example given is ffmpeg -i INPUT -map_channel 0.0.0 OUTPUT_CH0 -map_channel 0.0.1 OUTPUT_CH1 -- which is what I did above :/ - any ideas why I am getting silence in all the output files? [16:59] <zap0> hackeron, http://ffmpeg.org/ffmpeg.html then search for "map_channal' has some examples [17:00] <hackeron> zap0: that's what I'm doing [17:00] <hackeron> the example is: "ffmpeg -i INPUT -map_channel 0.0.0 OUTPUT_CH0 -map_channel 0.0.1 OUTPUT_CH1" -- what I'm doing is: "ffmpeg -f alsa -ac 12 -i plug:capt -ar 44100 -map_channel 0.0.0 -y test0.wav -map_channel 0.0.1 -y test1.wav -map_channel 0.0.2 -y test2.wav -map_channel 0.0.3 -y test3.wav -map_channel 0.0.4 -y test4.wav -map_channel 0.0.5 -y test5.wav -map_channel 0.0.6 [17:00] <hackeron> -y test6.wav -map_channel 0.0.7 -y test7.wav -map_channel 0.0.8 -y test8.wav -map_channel 0.0.9 -y test9.wav -map_channel 0.0.10 -y test10.wav -map_channel 0.0.11 -y test11.wav" [17:00] <hackeron> (for 12 channels instead of 2) [17:00] <hackeron> and getting silence on all channels [17:01] <zap0> if file 0, stream 0 really the input? [17:01] <hackeron> yes: Input #0, alsa, from 'plug:capt': [17:01] <hackeron> Duration: N/A, start: 1377355941.479040, bitrate: 9216 kb/s [17:01] <hackeron> Stream #0:0: Audio: pcm_s16le, 48000 Hz, 12 channels, s16, 9216 kb/s [17:02] <hackeron> am I misunderstanding something? [17:03] <hackeron> if I try to use any other file or stream input, other than 0, it says: "mapchan: invalid input file stream index #0.1" [17:03] <hackeron> and if I use a channel higher than 11, it says "mapchan: invalid audio channel #0.0.12" - so I'm using the right parameters it seems [17:03] <zap0> why haven't you told in the format.. s16le [17:03] <zap0> it/ [17:04] <hackeron> zap0: I don't need to, dsnoop does in .asoundrc - also ffmpeg -f alsa -ac 12 -i plug:capt -ar 44100 -y test.wav -- records all 12 channels just fine [17:05] <zap0> if you say so. [17:07] <hackeron> zap0: ok, I changed the command to: ffmpeg -f alsa -ac 12 -sample_fmt s16 -i plug:capt -ar 44100 -map_channel 0.0.0 -y test0.wav -map_channel 0.0.1 -y test1.wav -map_channel 0.0.2 -y test2.wav -map_channel 0.0.3 -y test3.wav -map_channel 0.0.4 -y test4.wav -map_channel 0.0.5 -y test5.wav -map_channel 0.0.6 -y test6.wav -map_channel 0.0.7 -y test7.wav -map_channel 0.0.8 -y test8.wav -map_channel 0.0.9 -y test9.wav -map_channel 0.0.10 -y ... [17:07] <hackeron> ... test10.wav -map_channel 0.0.11 -y test11.wav -- no change, silence in all output files [17:07] <hackeron> any other ideas? [17:07] <zap0> are they the file length you expected? [17:08] <zap0> also, im not sure s16 is valid.. how does it know endianess? [17:08] <hackeron> yep, all 7 seconds - if I add -t 5 to the beginning, all are 5 seconds [17:09] <hackeron> the only ones available are: s8 s16 s32 flt dbl u8p s16p s32p fltp dblp -- which one is the correct one? [17:11] <zap0> don't know... i don't use -sample_fmt i use something else [17:11] <hackeron> what do you use? [17:11] <zap0> i'm trying to find it (hence the delay) [17:11] <zap0> -f s16le [17:13] <hackeron> ok, current command is: ffmpeg -f alsa -ac 12 -i plug:capt -f s16le -ar 44100 -map_channel 0.0.0 -y test0.wav -map_channel 0.0.1 -y test1.wav -map_channel 0.0.2 -y test2.wav -map_channel 0.0.3 -y test3.wav -map_channel 0.0.4 -y test4.wav -map_channel 0.0.5 -y test5.wav -map_channel 0.0.6 -y test6.wav -map_channel 0.0.7 -y test7.wav -map_channel 0.0.8 -y test8.wav -map_channel 0.0.9 -y test9.wav -map_channel 0.0.10 -y test10.wav -map_channel ... [17:13] <hackeron> ... 0.0.11 -y test11.wav [17:13] <hackeron> silence in all output wav files :/ [17:13] <hackeron> input is: Stream #0:0: Audio: pcm_s16le, 48000 Hz, 12 channels, s16, 9216 kb/s - so that looks right [17:14] <zap0> have you given the output file format? [17:14] <hackeron> if I output to a single output.wav file, it outputs all 12 channels correctly [17:14] <hackeron> no, I haven't [17:15] <hackeron> doesn't seem like I need to? < # file test1.wav [17:15] <hackeron> test1.wav: RIFF (little-endian) data, WAVE audio, Microsoft PCM, 16 bit, mono 48000 Hz [17:16] <zap0> hwo do you know it's silence? [17:16] <hackeron> I opened it in audacity, there is absolutely no signal [17:17] <zap0> if you open it in notepad... can you literally see the file is just a bunch of NUL chars ? (of whatever value is audio-silence) ? [17:17] <zap0> or whatever/ [17:17] <hackeron> yes: ^@^@^@^@^@^@^@^@^@^@^@^@ [17:17] <hackeron> just a bunch of that, nothing else [17:17] <zap0> do you have a sample of this 12chan file i could download and try ? [17:18] <hackeron> sure [17:18] <hackeron> one sec [17:18] <zap0> k [17:18] <zap0> back in 3 [17:20] <hackeron> recorded with: ffmpeg -t 10 -f alsa -ac 12 -i plug:capt -y test.wav -- http://itstar.co.uk/test.wav [17:21] <zap0> .ogx file? it's downloading..... slowly.... ETA 14mins [17:22] <hackeron> so as you can see, all 12 channels are there and there is signal on channels 3,4,5,6,7,8 and 11 and 12 is pink noise [17:22] <hackeron> .ogx? -- it's test.wav: RIFF (little-endian) data, WAVE audio, Microsoft PCM, 16 bit, 12 channels 48000 Hz [17:23] <hackeron> so now I need to figure out why -map_channel isn't working to split the channels to separate output files (or to pick just a single channel) [17:24] <zap0> yep.. test.ogx and VLC plays it (although i only have stereo speakers), but it's info says 12 chns [17:24] <hackeron> if you open it with audacity, it will show a waveform for each separate channel [17:24] <hackeron> and allow you to solo each channel [17:26] <hackeron> so any ideas how to record just 1 channel out of the 12? [17:26] <zap0> the levels appear to be VERY low.. i 'see' 3,4,5,6,7,8,11,12 the others appear to be silent. [17:26] <hackeron> say channel 4 [17:26] <hackeron> yes, the others are silent [17:29] <hackeron> zap0: so any ideas how to record just channel 4 for instance? [17:31] <zap0> trying a few things [17:31] <hackeron> thanks :) [17:31] <mecil9> hi guys [17:31] <mecil9> anybody can help me? [17:32] <mecil9> i has been git ffmpeg to my computer [17:32] <mecil9> make all [17:32] <mecil9> error:libmp3lame >=3.98.3 not found [17:33] <mecil9> what's means [17:34] <hackeron> mecil9: http://bit.ly/1dCQNEK [17:36] <mecil9> i can't open the pages [17:38] <zap0> hackeron.. ffmpeg says during file creation.. "-map_channel is forwarded to lavfi similarly to -af pan=0x4:c0=c6." [17:39] <zap0> hackeron, then pan says: "This syntax is deprecated. Use '|' to separate the list items" [17:39] <hackeron> zap0: yeh, I saw that -- is ffmpeg doing it wrong? [17:39] <Hfuy> Codec question: if I present an h.264 codec (not necessarily the one in ffmpeg) with a black-and-white image, is it likely to be able to use the fact that there's no energy in the U and V channels to improve compression? [17:39] <zap0> hackeron, maybe.. try using | instead [17:39] <Hfuy> I understand that h.264 specifies the decoder, not the encoder, so presumably various codecs could handle this differently. [17:41] <hackeron> zap0: no difference at all [17:42] <hackeron> zap0: I'm doing: for i in 0 1 2 3 4 5 6 7 8 9 10 11; do ffmpeg -i test.wav -t 5 -dn -vn -codec:a libfdk_aac -flags +qscale -ar 44100 -af "pan=0x4|c0=c${i}" -y ch${i}.aac; done --- no deprecetaion warnings anymore, silence in output files (3KB per file, empty waveform, etc) [17:42] <hackeron> zap0: are you able to record just 1 channel from that test recording? [17:44] <zap0> i have a ffmpeg command line that produces are correct file (although it is still full of silence). [17:44] <zap0> it makes a .wav in the right size, header etc.. just full of NUL chars :( [17:45] <hackeron> yeh, that's what I'm getting :/ [17:45] <hackeron> all examples I can find say something like: ffmpeg -i stereo.wav -map_channel 0.0.1 right_mono.wav -- the 1 in 0.0.1 being the channel id -- but the output file is silence regardless of channel id :/ [17:45] <zap0> is this a one-off problem? can you not just use Audacity? [17:45] <hackeron> no, it isn't, it's for recording live audio [17:46] <zap0> ok [17:47] <zap0> the other day i write a WAV reader/writer in C. i'd have just used my own code by now :) [17:48] <hackeron> lol, can it record from just 1 specified channel from an alsa source with 12 channels? [17:48] <Hfuy> I once wrote a WAV reader in Javascript, in windows scripting host. Complicated it ain't. [17:48] <hackeron> and create segment files? [17:48] <zap0> hackeron: maybe use -af pan ? ? http://superuser.com/questions/601972/ffmpeg-isolate-one-audio-channel [17:49] <hackeron> zap0: that's the page I have open on the screen - no change [17:49] <hackeron> zap0: output file is silence :/ [17:49] <hackeron> it looks like ffmpeg is broken when dealing with 12 channel inputs? [17:50] <Hfuy> I haven't heard the whole conversation - are you trying to process a 12 channel wav? [17:50] <zap0> Hfuy, i'm currently trying to build a simple waveform generator/player in JS, to run on mobile phone browsers! [17:50] <Hfuy> zap0: The problem is not so much the language itself, it's the facilities provided by the environment. The trick for doing it in windows was simply how to get a binary stream into a series of numbers. Figure that out and it's trivial. [17:51] <hackeron> Hfuy: yep, here's a wav file, it is 12 channels: http://itstar.co.uk/test.wav -- I need to be able to get just 1 channel out of it with ffmpeg. All the documented methods including -af pan and -map_channel are not working and producing an empty output file. For example: for i in 0 1 2 3 4 5 6 7 8 9 10 11; do ffmpeg -i test.wav -codec:a libfdk_aac -flags +qscale -ar 44100 -af "pan=1|c0=c${i}" -y ch${i}.aac; done [17:51] <Hfuy> I'm actually lying anyway - I did it in windows script host, but there was some aspect or other of binary handling that the JScript interpreter's tendency to make everything into ASCII was screwing up. I had to use a tiny bit of VBscript to get around it. But it worked. [17:52] <Hfuy> Mmm, I wouldn't be surprised if any $SOFTWARE had a problem reading multichannel audio, they often do. But equally it isn't a complex format. [17:52] <zap0> i feel so dirty, even just hearing about the use of VB script ;) [17:52] <Hfuy> zap0: Imagine how dirty I felt writing it. But it was only a couple of lines. [17:53] <Hfuy> hackeron: Sorry, you're massively exceeding my experience with ffmpeg. [17:54] <Hfuy> I don't even know if your commandlines have the correct intent, let alone their likely performance. [17:54] <Hfuy> I have to ask, though: where the hell did you get a 12-channel wave from?! [17:56] <Hfuy> zap0: http://pastebin.com/YCpVZw7S [17:57] Action: zap0 reads [17:58] <hackeron> ok, filed a bug report: http://trac.ffmpeg.org/ticket/2899 [17:59] <zap0> lol.. i just tried ffmpeg 2008 .. "unrecognized option '-map_channel'" [17:59] <Hfuy> zap0: If you can spot the big ugly cheat, which isn't in that file, you can have one of my doughnuts. [18:00] <hackeron> zap0: yeh, it was added in 2011 I believe? [18:00] <zap0> on a diet [18:00] <zap0> hackeron, good idea. [18:00] <zap0> : re bug report.. good idea,. [18:00] <hackeron> thanks :) [18:01] <zap0> hackeron, it's not likely to get looked at unless a 12 channel input file is available... so add a link in your bug report. or at least email so you can be contacted [18:01] <hackeron> zap0: the first line is a link [18:02] <Hfuy> I would expect it's somewhat unlikely to be looked at anyway. [18:02] <zap0> if a dev has an interest in 12 chn audio... it might get a lot of attention [18:02] <Hfuy> And the likelihood of that is... [18:02] <hackeron> Hfuy: actually every bug I filed to ffmpeg has generally been looked at very quickly [18:03] <Hfuy> Hey, did I figure out how to parse wave files without needing my evil vbscript hack? I think I did! [18:03] Action: Hfuy is a genius [18:04] <zap0> hackeron, success!! [18:04] <braincracker> my genius mice's buttons died early ;/ [18:04] <hackeron> zap0: yeh??? [18:04] <zap0> hackeron.. oh oh oh... O M G... w000tttt!!!! /me runs about the room naked! [18:04] <hackeron> zap0: I'm going to join you, what is it??? [18:04] <zap0> hackeron.. used an older ffmpeg.. the file is non-full-of-NULs [18:05] <zap0> lemme listen.. back in 3 [18:05] <hackeron> zap0: lol! - ok, now to figure out what someone broke and how and get a developer to revert it [18:05] <braincracker> zap0 <= tin-foil-hat is a must! [18:06] <braincracker> Hfuy <= in C? for loop? [18:06] <zap0> hackeron, i selected channel 11,, it sounds a bit like white noise heard thru a toilet roll stuck to ones ear. [18:06] <Hfuy> braincracker: Sorry? [18:06] <Hfuy> hackeron: where did this 12 channel audio come from, out of interest? [18:06] <hackeron> zap0: yep, channels 11 and 12 are pink/white noise -- try something like channels 3 to 8 [18:06] <braincracker> [180330] <Hfuy> Hey, did I figure out how to parse wave files without needing my evil vbscript hack? I think I did! [18:06] <hackeron> Hfuy: M-Aaudio lt1010 sound card [18:06] <braincracker> don't you just hate vb* ? [18:07] <Hfuy> Oh I do. [18:07] <hackeron> Hfuy: it has 8 analog inputs and 4 digital inputs [18:07] <Hfuy> I did it in Javascript. [18:07] <Hfuy> I can't quite figure out how, but apparently I did. [18:08] <braincracker> okey, most ms things work like this. [18:09] <zap0> hackeron, channel 3 sounded like pink noise too. channel 8 sounds very quite.. just turned up it sounds a bit like shitty audio chips on motherboards that produce digital noise into their crappy pre-amps [18:09] <zap0> is there a specific channel that has something very distringuishable ? [18:09] <hackeron> zap0: yeh, it is recording sound in a few rooms that are empty right now - but at least it is working :D [18:09] <zap0> yes! [18:09] <hackeron> zap0: well, you should be able to hear rain on 7 and 8 I believe? [18:10] <hackeron> zap0: ok, so any ideas what revision broke it and what specific code? [18:10] <braincracker> nobody knows how, it just works [18:10] <hackeron> also, can you add a comment what version works for you? [18:10] <Hfuy> I'm not sure if it's really an MS thing or a JS thing. [18:10] <Hfuy> The situation is that the file reader you get in JScript expects to work on text files, and munges character values above 128 in certain circumstances. [18:10] <zap0> hackeron, ffmpeg.exe --version ffmpeg version N-35295-gb55dd10, built on Nov 30 2011 00:52:52 with gcc 4.6.2 [18:10] <braincracker> ddos logs will be forwarded to authorities [18:11] <zap0> Hfuy, surely it has a binary mode!? [18:12] <hackeron> zap0: it's so annoying :( - I need the latest version of ffmpeg because it has all the -segment beauty but -map_channel is broken, grrr [18:12] <zap0> hackeron, at least we identified it's a bug.. and not a missing feature.. so it should be fixable... quickly-ish [18:12] <Hfuy> zap0: Well, to be completely fair, the function is called openTextFile() [18:13] <hackeron> zap0: yeh, I've added a comment: "Also note that ffmpeg version: N-35295-gb55dd10, built on Nov 30 2011 00:52:52 with gcc 4.6.2 works fine, but latest trunk is broken. [18:13] <hackeron> " [18:13] <zap0> hackeron, let me find a mid point-date-wise.. see if thats broken too.. maybe something near... oct/nov 2012 [18:17] <zap0> hackeron, ffmpeg-20121003-git-df82454 throws an error, and writes a zero sized output file. [18:18] <Hfuy> zap0: Aha. You have to do some chicanery with translating unicode numbers to their ASCII equivalents. [18:18] <hackeron> zap0: aha, so somewhere between 201111 and 20121003, lol? [18:18] <Hfuy> If you read a "text" file byte >127 then do charCodeAt() on it, you'll get a unicode number if it's >127. [18:19] <zap0> hackeron, indeed!! the last verison i just quoted has some git reference... perhaps that is valuable to someone [18:19] <Hfuy> Which is why I got Visual Studio and started using C# instead :) [18:21] <hackeron> zap0: hmm, someone replied - have a look at the ticket [18:22] <Hfuy> I don't get that. How is there supposed to be a "known" channel layout for 12 arbitrary inputs? [18:23] <zap0> cause 5.1 and 7.1 has known layouts... 12 is not a "standard" [18:23] <Hfuy> Pan filter just takes numeric input, though, doesn't it? [18:24] <Hfuy> (and in fact, in may situations, the channel layout of streams known to contain 5.1 and 7.1 tracks is really not very consistent!) [18:25] <hackeron> zap0: hmm, how do I specify the aformat=channel_layouts=0xFFF? [18:25] <zap0> hackeron, i don't know. im still staring at it trying to comprehend that too [18:26] <hackeron> heh [18:27] <Hfuy> In movie postproduction, multichannel surround is almost always ferried around as a set of single channel files. [18:27] <Hfuy> For this exact reason. [18:27] <Hfuy> Few things support multichannel files, and even fewer support them properly. [18:28] <Hfuy> Really you need something like -af "assign_channels=l,r,c,ls,rs,lfe" [18:29] <hackeron> Hfuy: I want to record 1 channel, in mono, a channel number I specify from an input with 12 channels - there is no left/right/whatever - every channel is a different room [18:29] <Hfuy> Oh I completely understand. But if it's going to insist on somehow knowing what the channels represent, there ought to be a way to assign them labels. [18:30] <Hfuy> But I agree there seems to be no reason why that should be necessary simply to split out one of the channels. [18:30] <zap0> RIFF format provides packets for data [18:30] <zap0> someone needs to set (YET ANOTHER) standard ;) [18:31] Action: Hfuy cries [18:32] <Hfuy> The reason I wrote that wave parser was so as to have the ability to read and write "broadcast wave" extensions, with timecode etc. [18:32] <zap0> i hereby declare a new RIFF packet called 'channel layout', containing a list like 1=23ºW @ 206.4mm from center. 2=... [18:32] <Hfuy> Gathering example files from field audio recorders, I immediately discovered a collection of RIFF chunks I'd never heard of before. [18:33] <Hfuy> There were n of these proprietary chunks, where n is in fact slightly larger than the number of recorder manufacturers involved. [18:33] <zap0> Hfuy, lol.. the reason i'm write audio in JS is for a timecode generator! [18:33] <Hfuy> Writing a slate app? [18:33] <zap0> more or less! [18:33] <Hfuy> I would counsel against it :/ [18:34] <zap0> why? [18:34] <Hfuy> We tried two different ipad slate apps against a real Ambient clockit slate. [18:35] <Hfuy> The problem I think is that the accuracy of the slate apps is dependent on the clock accuracy of the audio hardware in the ipad. [18:35] <Hfuy> And it isn't good enough. [18:35] <Hfuy> It lost whole frames an hour, which is way not good enough. [18:35] <zap0> are you implying the accuracy/drift is an issue? [18:35] <Hfuy> There are circumstances where you could make it work, but it isn't good enough to jam sync then walk away. [18:35] <zap0> yes, i guess you are! [18:36] <Hfuy> Depends what you're doing I guess. [18:36] <zap0> i am quite aware of the haphazard timing of these android/ipad consumer hardware. [18:36] <Hfuy> If you want to let it listen to incoming SMPTE timecode using the audio input, and just display what you're getting, fine. [18:36] <hackeron> zap0: any luck? -- I tried ffmpeg -v debug -i test.wav -filter:a aformat=channel_layouts=0xFFF -af "pan=0x4|c0=c4" -y ch4.wav -- still getting silence in the output [18:37] <Hfuy> Equally if it's going to be a timecode master and you're going to record its output onto a spare audio track for later syncing, probably fine. [18:38] <Hfuy> Perhaps you could put some sort of calibration term into the software but I'm not sure how much of the drift we saw is down to interrupts and so on.\ [18:39] <zap0> how often do you record a single take that goes for over an hour ? [18:40] <Hfuy> Not often. But that's not the factor, if you want to jam sync it. [18:40] <Hfuy> The issue is has it been RUNNING for an hour since it was last synced. [18:42] <zap0> this is for some simple stuff anyway... if i wanted something i'd have to rely on, then i'd use this microcontroller to do it, and run it off a real-tme-clock module thingy. [18:43] <Hfuy> I think really this is a microcontroller project. [18:43] <Hfuy> I've been pondering doing just that for ages, but it's a lot easier if you can simply ensure the uC is accurately clocked, as opposed to trying to refer your code to an external RTC. [18:43] <Hfuy> And I think you can do that. [18:43] <zap0> i'm using a uC for displaying milliseconds anyway... for the high-speed cameras [18:44] <Hfuy> Ooh, high-speed cameras [18:44] Action: Hfuy rubs his hands [18:44] <Hfuy> It's not as if SMPTE code is complicated, anyway. [18:45] <zap0> i've seen some people trying to use POV displays for high-speed. [18:45] <Hfuy> I just wish it encoded frame rate. [18:45] <zap0> there are some empty blocks in SMPTE you can write your own data into [18:45] <zap0> although not all hardware likes it when you do [18:45] <Hfuy> Only a couple of bits. [18:45] <Hfuy> Although that'd be enough to indicate whether we were at a fractional frame rate or not. [18:45] <Hfuy> But, as you say... [18:46] <zap0> how many frame rate changes per second do you need ? just write 1 bps of your frame-rate-info stream, until its done! [18:46] <Hfuy> Heh. [18:47] <Hfuy> Really the issue when I was doing it was simply being able to tell the difference between, say, 29.97 and 30. [18:47] <Hfuy> Which I found was fine, even from analogue tape. So it really isn't a huge deal. [18:48] <zap0> lol.. NTSC [18:49] <hackeron> zap0: this seems to work! < ffmpeg -v debug -i test.wav -filter:a "aformat=channel_layouts=0xFFF,pan=0x4|c0=c4" -y ch4.wav [18:50] <Hfuy> zap0: Tell me about it. I live in the UK, where we can count to 25. IN WHOLE NUMBERS. [18:51] <zap0> lol... .au PAL too ;) [18:51] Action: Hfuy waves a very small union jack [18:51] <sacarasc> So, you're out to sea, Hfuy? [18:51] <Hfuy> Oh dear. A heraldic pedant. [18:51] <Hfuy> OK, OK. Union FLAG. [18:51] <sacarasc> \o/ [18:52] <zap0> hackeron, well done! [18:52] <Hfuy> Nobody in the UK understands the difference, or has any idea what you're on about when you talk about the "union flag." And if you use the term in international company, they tend to think of the American civil war. [18:52] <mecil9> hackeron [18:52] <Hfuy> So yes, I tend to use the more common term. But be happy; I know how not to indicate I'm in distress when flying said flag. [18:53] <zap0> Hfuy, i have no idea... but i'm going to guess you are talking about the Cross-of_... overlayed on the cross-of-.... that makes up the multiple colours? [18:54] <Hfuy> The George Cross is a red cross on white. George, by one mythology, is the patron saint of England. [18:54] <hackeron> mecil9: yes? [18:54] <Hfuy> The Cross of St. Andrew is a diagonal white cross on blue. By said mythology, Andrew is patron saint of Scotland. [18:55] <mecil9> hackeron new error played [18:55] <mecil9> libx264 must be >=0.118 [18:55] <mecil9> i git x264 from git.videolan.org [18:56] <mecil9> make it [18:56] <Hfuy> Aaaand the cross of St. Patrick is a diagonal red cross on white. But we made that bit smaller. Because frankly, who cares bout the Irish. :) [18:56] <Hfuy> And mainly because it was selected more or less at random from the flags of the great houses of Ireland. [18:56] <Hfuy> Infodump ends. [18:58] <zap0> hackeron, i've noted your example in my 'notebook of ffmpeg tricks' which grows by the day! [19:00] <hackeron> zap0: haha, cna I see that notebook? [19:01] <fscz> I've decoded a frame, using avcodec_decode_audio4, however when I put this frame into avcodec_encode_audio2 it crashes on me. [19:02] <fscz> I've debugged it all the way to where the segault is thrown, which is at samplefmt.c/av_samples_copy [19:02] <fscz> the problem is, that the source parameter of the function is messed up and therefore the memcpy call in that function fails [19:04] <fscz> I am not sure about this, but i think the problem might be in avcodec_encode_audio2, as it enters the branch where pad_last_frame is called. [19:04] <fscz> this happens, even though it is the very first frame that was decoded. [19:10] <hackeron> zap0: ok, here's my command to record audio :D < ffmpeg -loglevel info -f alsa -ac 12 -i plug:capt -map 0 -analyzeduration 0 -dn -vn -codec:a libfdk_aac -flags +qscale -global_quality 1 -afterburner 1 -f segment -segment_time 60 -segment_wrap 10 -segment_list_flags live -segment_list_size 10 -reset_timestamps 1 -segment_list 'test.csv' -ar 44100 -filter:a "aformat=channel_layouts=0xFFF,pan=1|c0=c4" -y test_%02d.mkv [19:10] <hackeron> simples :P [19:16] <Hfuy> Does ffmpeg have -filter:v "camerawork=good" [19:16] <Hfuy> and if not why not [19:18] <fscz> Am I missing something obvious, or is nobody here who has the answer? [19:55] <shahinkhan> Hi, can anybody help me with this? : how can I have less delay when I use ffplay via RTSP [21:31] <taladan> hey folks, quick q - is there a way to use ffmpeg to capture from two different windows at once without having to have those windows overlayed on the desktop? [21:32] <taladan> like, for instance if I was streaming out to a source and wanted to overlay a webcam video in the lower right quadrant of the screen, could I have the webcam on a seperate portion of my desktop without having to have it positioned over the lower right quadrant of the main window I'm streaming? [21:59] <fscz> decoding audio file, I got following error: [vorbis @ 0x2226e0] Not a Vorbis I audio packet. Error decoding frame: Invalid data found when processing input [21:59] <fscz> is this recoverable [22:00] <fscz> maybe with skipping? [00:00] --- Sun Aug 25 2013

1 0

ffmpeg-devel.log.20130823
by burek 24 Aug '13

24 Aug '13

[00:38] <cone-826> ffmpeg.git 03Michael Niedermayer 07master:c443689afbb3: avformat/movenc: use av_freep() instead of av_free() except for local variables before return [00:47] <durandal_1707> michaelni: so you are now reporting CVE candidates? [01:05] <michaelni> durandal_1707, you make it sound like theres something new about that [01:07] <cone-826> ffmpeg.git 03Michael Niedermayer 07master:8bb11c3ca77b: avcodec/jpeg2000dec: Check cdx/y values more carefully [01:08] <durandal_1707> well, i get impression others reported it in past [01:09] <beastd> michaelni: You have some security mechanism in place for creation of the web pages? [01:09] <beastd> remote: ok/Makefile Makefile differ: char 169, line 4 [01:10] <beastd> I pushed archive page as requested by ubitux and durandal_1707 . So I had to add a new source to the Makefile. [01:11] <ubitux> thx :) [01:11] <michaelni> beastd, yes [01:12] <beastd> michaelni: what is the procedure? should i change the ref file on the web server and rebuild? [01:12] <michaelni> i think so [01:13] <beastd> ubitux: np. i do not think creating a news archive is related to adding a new entry. but you got me with hat one :) [01:13] <ubitux> :) [01:14] <michaelni> durandal_1707, true, most where, not all though [01:15] <durandal_1707> where are news older than 2007? [01:16] <beastd> dunno [01:16] <ubitux> in the git log maybe [01:17] <durandal_1707> oldest commits is from 2011 [01:17] <durandal_1707> so i think i should use other tools... [01:18] <ubitux> archive.org ? [01:18] <durandal_1707> no, TARDIS [01:19] <ubitux> wow it was in php [01:19] <ubitux> http://web.archive.org/web/20020922011917/http://ffmpeg.sourceforge.net/ [01:19] <ubitux> (later) [01:20] <durandal_1707> omg, what i missed [01:20] <ubitux> and it was written "FFMpeg" [01:20] <ubitux> heresy! [01:21] <ubitux> http://web.archive.org/web/20050301010331/http://ffmpeg.sourceforge.net/ind… [01:21] <beastd> ok ppl, archive page is online. will go to sleep now. [01:21] <beastd> see you [01:22] <ubitux> http://web.archive.org/web/20070116171005/http://ffmpeg.mplayerhq.hu/ moar news [01:22] <ubitux> durandal_1707: i guess this one provides you the missing bits: http://web.archive.org/web/20081119203954/http://ffmpeg.org/ [01:27] <durandal_1707> perhaps missing archive entries should be added back? [03:19] <cone-826> ffmpeg.git 03Carl Eugen Hoyos 07master:2baa12f1d194: Fix dependencies for h263 vaapi decoder. [04:57] <cone-826> ffmpeg.git 03Michael Niedermayer 07master:16a0d75c769a: avcodec/mjpegdec: fix overread in find_marker() [07:26] <BBB> and back from vacation [07:26] <BBB> ubitux: any news? :) [09:12] <ubitux> BBB: well i was mostly trying to get a better picture of the whole thing so i wasn't very^Wat all productive [09:14] <ubitux> i'll have a few questions in a few hours [09:36] <BBB> ubitux: booh! well anyway [09:36] <BBB> ubitux: yes ask qs [09:58] <BBB> ubitux: or at least let me know how far you are, we can split work for particular tasks if you want [09:58] <BBB> e.g. I can do sub8x8 bitstream parsing and confirm at least all that works [09:58] <BBB> ubitux: or something along those lines [10:42] Action: BBB gives ubitux a big fat poke [10:43] <ubitux> BBB: i was on sub8x8, which was as you said similar to the VP8_SPLITMVMODE thing [10:44] <BBB> bitstream parsing? or reconstruction? [10:44] <ubitux> but i dont know how i'm suppose to test this [10:44] <BBB> (or both?) [10:44] <ubitux> it was the reconstruction; but i guess i have the bitstream parsing to do too (first), right? [10:44] <BBB> yes [10:44] <BBB> otherwise reconstruction will never trigger, I suppose [10:44] <ubitux> i saw that code triggered once, but probably a bug [10:45] <ubitux> anyway, where is that supposed to be done? [10:45] <BBB> if (b->bl == BL_8X8 && b->bp == PARTITION_NONE) { [10:45] <BBB> // sub8x8 mode/mv coding [10:45] <BBB> // inter mode ctx = inter_mode_ctx_lut[a_mode][l_mode]; [10:45] <BBB> printf("Inter sub8x8 mode/mv coding not yet done\n"); [10:45] <BBB> return -1; [10:45] <ubitux> (sorry i'm a bit lost at the overall codec) [10:45] <BBB> line 1455 [10:45] <ubitux> ok [10:45] <BBB> you basically do the inter equivalent of the intra coding in line 1080-1120 [10:46] <BBB> that is, parse one inter mode (see line 1421-1426), then parse one mv (if newmv), then parse the second inter mode, next mv (if newmv), etc. [10:47] <BBB> either twice (for PARTITION_H/V) or 4x (for PARTITION_SPLIT) [10:47] <BBB> you also need a small extension to find_ref_mvs() to be sub8x8 compatible [10:47] <BBB> I can point out what's missing in our code or the relevant code in libvpx so you can figure it out yourself [10:48] <ubitux> that would help [10:48] <ubitux> (in libvpx) [10:48] <ubitux> (i need a reference somehow...) [10:49] <BBB> vp9/common/vp9_mvref_common.c [10:50] <BBB> see calls to get_sub_block_mv() [10:50] <ubitux> i see, thx [10:50] <BBB> block index is 0 (topleft), 1 (topright), 2 (bottomleft) or 3 (bottomright 4x4 subblock in a 8x8 parent block) [10:50] <BBB> if the blocks are 4x8, only 0 and 1 exist [10:50] <BBB> if the blocks are 8x4, only 0 and 2 exist [10:50] <ubitux> btw, i looked a bit at the vp9 ml; there is not much info [10:51] <ubitux> i found some kind of xml draft for overall design, but not sure if accurate and really up-to-date [10:51] <BBB> the direct neighbours are filled in vp9_append_sub8x8_mvs_for_idx() in vp9/common/vp9_findnearmv.c [10:51] <ubitux> ok [10:51] <BBB> basically what it does is look for direct left and/or direct above neighbour (if we're not subblock 0) [10:52] <BBB> so for block 1, the direct left would be block 0 [10:52] <BBB> for block 2, the direct above would be 0 [10:52] <BBB> for block 3, the direct left is 2 and above is 1 [10:52] <BBB> then after that it calls find_ref_mvs (ffvp9) / vp9_find_mv_refs_idx (libvpx) to fill in the more distant mv references [10:53] <BBB> for sub8x8, it is subblock aware for the direct left/above neighbours [10:53] <BBB> that is, for the left 8x8 neighbour, it uses the edge block closest to our subblock [10:53] <BBB> for 0/1, that is 1 (left), for 2/3, that is 3 (left); for 0/2, that is 2 (above) and the 1/3, that is 3 (above) [10:53] <ubitux> it's still using mv from a maximum of 3 ref frames? (from a maximum of a pool of 8 frames?) [10:53] <BBB> of course all of this matters only if these blocks use sub9x8x8 coding [10:53] <BBB> yes [10:54] <ubitux> why limit the limit to 3 ref frames when there is 8 frames cached? [10:54] <BBB> no idea [10:54] <ubitux> ok [10:54] <BBB> :-p [10:54] <ubitux> :) [10:55] <ubitux> ok well, i'll try to make a report of my progress in ~12 hours [10:55] <ubitux> i have to go right now [10:55] <ubitux> thx for the help [10:56] <BBB> bye [10:56] <ubitux> there is two random trivial commits in my branch if you're interested [10:56] <BBB> what do I do now? :) [10:56] <BBB> ok I'll check [10:56] <ubitux> haha [10:56] <ubitux> what do you do? i dunno; is my work that blocking for your progress? [11:02] <BBB> ubitux: well a little right? I can't get reconstruction of the first inter frame correct until sub8x8 is done, and I can't test things like bw adaptivity until after first inter frame is correct [11:02] <ubitux> erm [11:03] <ubitux> you can give me something else if that matters for your [11:03] <ubitux> -r [11:03] <ubitux> i'm just curious about how it works, but as i know near to nothing about the codec, i guess anything else will be fine [11:04] <BBB> no it's fine, I'm gonna do some intra pred simd meanwhile [11:04] <BBB> I'll hope you finish it this weekend or so [11:05] <BBB> and your patches (plus minor one by me) pushed [11:05] <ubitux> i'll do my best, but gotta go for real now [11:05] <ubitux> ok :) [11:14] <BBB> I know what I'll do, I'll add some fate tests for the vp90-2-* files [11:14] <BBB> (first frame only ofcourse) [11:45] <zidanne> How do we write S16P to fifo to be used in audio encoding for mp3lame? like this? for(int i=0;i<plane_count;i++) av_fifo_generic_write(fifo, (uint8_t *)data[i], length, NULL ); [12:28] <cone-662> ffmpeg.git 03Stefano Sabatini 07master:0be3be901196: lavf/tee: copy metadata to output chained muxers [12:28] <cone-662> ffmpeg.git 03Paul B Mahol 07master:8fbf940e1673: lavfi/tile: do not leak input frame [12:44] <cone-662> ffmpeg.git 03Paul B Mahol 07master:dd1d29bd5f31: pngdec: use av_fast_padded_malloc(z) [12:52] <durandal11707> is there filter that would repeat each frame several times? [12:54] <durandal11707> i don't think so [12:55] <iive> mp=harddup ? [12:55] <durandal11707> that was removed [12:56] <durandal11707> and it did nothing [12:56] <Daemon404> durandal11707, vf_fps? [12:56] <Daemon404> just use an integer multiple of the current framerate [12:57] Action: BBB kicks lazy Daemon404 [12:57] <durandal11707> is it supposed to work with overlay? [12:57] <durandal11707> when you want to loop single main frame over and over again [12:58] <Daemon404> oh i unno [12:58] <Daemon404> ive always used images as input [12:58] <Daemon404> not videos [12:59] <durandal11707> yes, but if you loop image it will decode over and over again [12:59] <Daemon404> ... lol [13:03] <durandal11707> i get 19 vs 28 here [13:03] <durandal11707> fps [13:10] <durandal11707> hmm, this doesn't work: ffmpeg -i bg.png -f lavfi -i testsrc -filter_complex '[0]fps=999[fps],[fps][1]overlay' [13:19] <durandal11707> because you can't seek in lavfi there is no loop filter [13:33] <cone-662> ffmpeg.git 03John Stebbins 07master:1f70a5ad284b: mov: use tkhd enabled flag to set the default track [13:33] <cone-662> ffmpeg.git 03Michael Niedermayer 07master:c6f4a3a70837: Merge commit '1f70a5ad284b33e8b3e2b40a5cb33055419781b7' [14:07] <zidanne> I am using http://pastebin.com/466tUJyd to convert S16 to S16P so I can feed it to avcodec_encode_audio for libmp3lame codec. However, encoded file is just a looping noise like a very fast shooting machine gun. Is the algorithm wrong? [14:11] <Skyler_> I'm not sure, but maybe check to be sure that linesize is in units of int16_t, rather than units of bytes? [14:12] <Skyler_> I'm not sure which it is, but if it's in bytes and you're acting on it as if it's in units of samples, the code might be wrong [14:12] <Daemon404> why not just call libavresample [14:12] <Daemon404> the lazy way~ [14:17] <cone-662> ffmpeg.git 03John Stebbins 07master:30ce289074e8: movenc: Make tkhd "enabled" flag QuickTime compatible [14:17] <cone-662> ffmpeg.git 03Michael Niedermayer 07master:800ea20cadce: Merge remote-tracking branch 'qatar/master' [14:20] <michaelni> Daemon404, any reason why you suggest avr over libswresample ? [14:22] <Daemon404> no [14:22] <Daemon404> its easier to type [14:28] <zidanne> I have tried using swr_convert but it didn't work for S16 to S16P [14:30] <zidanne> (however it worked for FLTP to S16, which was valid for another scenario) [14:35] <michaelni> zidanne, swr certainly works for s16->s16p, its a common convertion, not something obscure that isnt tested [14:36] <michaelni> zidanne, if it doesnt work for you please provide a reproduceable testcase and ill take a look [14:45] <zidanne> it crashes on "conv_AV_SAMPLE_FMT_S16_to_AV_SAMPLE_FMT_S16" [14:46] <michaelni> zidanne, how can this be reproduced ? [14:49] <zidanne> I decode an mmsh stream (asf) and I convert it into AV_SAMPLE_FMT_S16 so I can feed the audio system. At the same time, I want to have it in AV_SAMPLE_FMT_S16P too (so I can encode it into mp3). So, I use swr_convert for the "second" time but this time with the output of the previous swr_convert. So, now it converts from S16 to S16P instead of FLTP to S16. [14:51] <michaelni> zidanne, can you provide a reproduceable testcase, a small C program that causes this crash ? [14:52] <michaelni> swr doesnt crash when used by other applications [14:53] <michaelni> or if it does crash for you with ffmpeg and some command line, please provide this command line [14:57] <durandal11707> so apropos repeat filter: to be or not to be, if loop ever came out it can be deprecated [15:13] <durandal11707> or why seeking in lavfi is so hard? [15:13] <zidanne> michaelni: my bad.. I was not creating an array and assigning the output of the previous swr_convert to it's [0]. Now it didn't crash. [15:14] <durandal11707> wasn't there example in source tree [15:14] <durandal11707> so you could look at it? [15:15] <durandal11707> perhaps its not commented at all and thats why so many people are lost [15:15] <zidanne> But i still get the same audio noise. So I must be using the AVAudioFifo incorrectly.. [15:16] <durandal11707> when swr converts to s16p it returns array with pointers to array [15:17] <durandal11707> anyway why there is not function that takes AVFrame? [15:18] <zidanne> durandal11707: I was using swr_convert to convert S16 to S16P. As swr_convert always wants an array, that was my problem [15:19] <wm4> durandal11707: yeah, that'd be very useful [15:19] <wm4> durandal11707: write it [15:19] <durandal11707> wm4: i have paypal [15:20] <durandal11707> zidanne: so what you use now? link... [15:24] <Daemon404> [14:13] <@durandal11707> or why seeking in lavfi is so hard? [15:24] <Daemon404> because seekign with lavf is hard [15:24] <Daemon404> or not possible [15:24] <Daemon404> without an index. [15:24] <Daemon404> which we do not implement. [15:29] <durandal11707> let filter just call avformat_seek_frame [15:34] <wm4> durandal11707: the main problem perhaps is that you'd have to figure out the seek target PTS [15:34] <wm4> which would require information traveling backwards through the filter chain, or so [15:34] <wm4> OTOH, you could just add a seek command to the movie src [15:55] <zidanne> I think, we could have a "ffmpeg tips" page. Where we can add notes related to common mistakes, etc.. (: [16:00] <michaelni> zidanne, you can create a page for that on the ffmpeg wiki if you like [16:13] <michaelni> link to wiki: https://ffmpeg.org/trac/ffmpeg [16:14] <michaelni> shorter link : https://trac.ffmpeg.org/ [18:13] <cone-662> ffmpeg.git 03Michael Niedermayer 07master:912ce9dd2080: jpeg2000: fix dereferencing invalid pointers [18:13] <cone-662> ffmpeg.git 03Michael Niedermayer 07master:09927f3eaa93: jpeg2000: zero reslevel array on allocation [18:13] <cone-662> ffmpeg.git 03Michael Niedermayer 07master:9e477a377033: jpeg2000: fix null pointer dereference in case of malloc failure [20:15] <cone-662> ffmpeg.git 03Michael Niedermayer 07master:aadfadd784bb: avformat/redspark: check coef_off [20:46] <durandal11707> michaelni: so just report progress and do not care what is put into frame? [20:47] <michaelni> filling the otherwise uninitialized parts by copying fro the previous frame would be better [20:48] <durandal_1707> well its hard to find what is initialized and what is not [20:49] <durandal_1707> some files have multiple idats and only last one may fail [20:49] <michaelni> then just setting progress should be fine for now if it fixes the crashes and deadlocks [20:50] <durandal_1707> i can't reproduce either [20:53] <durandal_1707> so if you can test it to find out no crashes/locks happen it would be great [20:53] <ubitux> BBB: what does the b->comp bit mean? [21:12] <zidanne> How can this be possible? (: When I av_free "resampledOut" after "memcpy(ffData->data,(const uint8_t *)resampledOut,ffData->size);" I get "malloc: *** error for object 0xf5fbf5fb: pointer being freed was not allocated" [21:13] <zidanne> if I comment out free, everything works perfectly (: [21:13] <ubitux> maybe because you changed your pointer [21:14] <zidanne> The only thing is, I give the pointer to swr_convert with &resampledOut. [21:26] <zidanne> the problem was av_freep. I changed it to av_free and now it works [21:41] <cone-662> ffmpeg.git 03Alexander Strasser 07master:b329ff3d43ca: MAINTAINERS: Add my GPG fingerprint [23:09] <durandal_1707> michaelni: so you tried it? [23:14] <michaelni> durandal_1707, i can if you push the code somewhere [23:14] <michaelni> i cant test without having the code [23:15] <durandal_1707> but i think one line change is less work.... [23:16] <michaelni> do you want your code tested or do you want me to redo the debuging and fix it myself ? [23:18] <michaelni> i cant test your code without seeing your code of course [23:20] <michaelni> you dont have to push it if you dont want but if you want me to test your code you have to provide your code by some means [23:20] <durandal_1707> ok, stop typing you are wasting time [23:22] <durandal_1707> you can see "code" in png branch [23:30] <zidanne> lol (: [23:31] <michaelni> durandal_1707, no crashes or deadlocks with that file anymore and fate passed [23:42] <durandal_1707> michaelni: but asan warnings? [23:43] <michaelni> asan showed nothing [23:43] <michaelni> neither valgrind but that didnt show anything before either [23:45] <Daemon404> 36 [23:47] <durandal_1707> huh, youtube protocol....... [23:50] <ubitux> ._. [23:50] <ubitux> why didn't he asked before :( [23:50] <ubitux> we have libquvi support... :( [23:51] <durandal_1707> but this one is LGPL [23:51] <ubitux> durandal_1707: it's going to require regular updates [23:52] <ubitux> like maybe every month [23:52] <durandal_1707> than that is not really protocol... [00:00] --- Sat Aug 24 2013

1 0

ffmpeg.log.20130823
by burek 24 Aug '13

24 Aug '13

[00:00] <axorb> IamTrying: ffmpeg -i Supernova1.mov -vcodec copy -acodec copy 1.mpg [00:00] <Mavrik> JennieL, weeeelll&.. yeah. It's also quite buggy. Lemme read your post again :) [00:00] <JennieL> Mavrik: Thanks. I am NOT making any promises that any of the info there is helpful! ;-) [00:00] <Mavrik> JennieL, ok, bad news is& you can't really do what you want in the way you have your system set up [00:00] <IamTrying> axorb, [mpeg @ 0x17315a0] buffer underflow i=0 bufi=234980 size=238677 [mpeg @ 0x17315a0] packet too large, ignoring buffer limits to mux it [00:01] <IamTrying> [mpeg @ 0x17315a0] VBV buffer size not set, muxing may fail [00:01] <JennieL> Mavrik: Which part? Just want to trasncode & stream the mpegts to my TV. [00:01] <JennieL> Or even VLC, at this point [00:02] <JennieL> and, of course, have the controls work [00:02] <Mavrik> JennieL, since you are sending output from ffmpeg directly to your TV and ffmpeg doesn't have any controls to move around the stream after it started processing (it's a tool for transcoding files primarly, not streaming :/) [00:02] <axorb> rtmp would be good if your TV supports it [00:02] <IamTrying> JennieL, Mavrik, is right, VLC and all of them are good for nothing, Gstreamer fights with VLC and VLC fights with Gstreamer, at the end of the day both are good for nothing [00:02] <JennieL> Mavrik: ffplay does ... [00:02] <Mavrik> ffplay is a simple example of a player :) [00:03] <JennieL> Ok, getting confused. Is it not possible? [00:03] <Mavrik> you need a DLNA server which can restart transcoding on another point when you want to seek. [00:03] <Mavrik> a DLNA server that understands that request and will properly restart encoding [00:03] <JennieL> It's already in there. Mediatomb ... [00:04] <JennieL> @ http://ffmpeg.org/pipermail/ffmpeg-user/2013-August/016996.html -- "I setup a "Mediatomb" DLNA server to get the videos onto my TVs." [00:04] <IamTrying> axorb, dont know but this did not worked ffmpeg -i Supernova1.mov -vcodec copy -acodec copy 1.mpg [00:05] <Mavrik> JennieL, what you're setting up is ffmpeg with command to transcode WHOLE file from BEGINNING without any timing parameters always [00:06] <Mavrik> JennieL, it also says in MediaTomb FAQ that they do not support seeking on transcoded streams [00:06] <JennieL> Mavrik: Is that a limitation of Mediatomb? or DLNA? The mediatomb folks told me this works. [00:07] <Mavrik> mediatomb [00:07] <axorb> MediaTomb should be doing the transcoding anyway [00:07] <Mavrik> seeking probably works if you don't have to transcode. [00:07] <JennieL> axorb: Mediatomb uses an ffmpeg-based script ... [00:07] <Mavrik> but your ffmpeg command certanly won't support seeking [00:07] <JennieL> Mavrik: Do you offhand know of a more reliable DLNA that *would* do this? [00:07] <Mavrik> I think PS3 Media Server supports seeking while remuxing [00:08] <JennieL> For Linux that'd be the UMS? [00:08] <Mavrik> I used it for a bit awhile ago, wouldn't know how compatible is it with other DLNA clients [00:08] <axorb> IamTrying: ffmpeg -i Supernova1.mov -vcodec libx264 -acodec aac -strict experimental 1.mpg [00:08] <axorb> then figure out what the -sameq option is now [00:08] <axorb> but you need to specify the codecs, and this will transcode instead of remux (which is ideal) [00:08] <durandal_1707> -sameq option never was useful for you ...... [00:09] <axorb> okay, don't use -same1 [00:09] <axorb> sameq [00:09] <IamTrying> axorb, OK but is not this all is a hard way? its like 2013 can it not be like $ ffmpeg -multi-files-from 1 2 3 4 5 -to one.mov ? [00:10] <durandal_1707> IamTrying: no [00:10] <axorb> you're concatting right? [00:10] <durandal_1707> why you are ignoring info people already gave you? [00:10] <axorb> https://trac.ffmpeg.org/wiki/How%20to%20concatenate%20(join,%20merge)%20med… [00:10] <axorb> ffmpeg -f concat -i <(for f in ./*.wav; do echo "file '$f'"; done) -c copy output.wav [00:10] <JennieL> Mavrik: Sigh. UMS forums seem to blame ffpmeg (http://www.universalmediaserver.com/forum/viewtopic.php?f=9&t=672&p=3903&hi…) That's assuming I've found the right topic. [00:10] <axorb> there's a loop for you [00:11] <axorb> and ffmpeg can seek, it's just hard to get it right [00:11] <Mavrik> JennieL, ffmpeg can't seek once transcoding has started, you have to restart it form another point [00:11] <axorb> ^ you need to cancel the current job [00:11] <axorb> and use -ss [00:12] <axorb> VLC can seek while transcoding though if you use an rtmp output [00:13] <JennieL> axorb: Hm. So no way to 'connect' that cancel_job+seek_to_new_location to the buttons on the remote? [00:14] <IamTrying> OK - axorb thank you , it works great. [00:14] <JennieL> This stuff really is SO far over my head ... [00:14] <axorb> DLNA would be the one making the seek request [00:14] <axorb> and the server would have to translate that into canceling the job and starting a new one [00:14] <IamTrying> durandal_1707, thank you too, its working [00:14] <JennieL> axorb: So, back to a DLNA server that actually works ... i.e. !Mediatomb [00:15] <Mavrik> JennieL, as I said, try out PS3 Media Server& I've had some success with it on Xbox and it supported seeking then [00:15] <axorb> never used DLNA, but it definitely is possible [00:15] <axorb> I just implemented it actually [00:15] <axorb> except with HLS instead of DLNA [00:16] <durandal_1707> hmm why they claim ffmpeg does not support internal subtitles? [00:16] <axorb> durandal_1707: http://trac.ffmpeg.org/ticket/2067 [00:16] <axorb> maybe? [00:17] <JennieL> Mavrik: PMS is on linux? (I'm looking now ...) [00:17] <Mavrik> think so. [00:17] <JennieL> Mavrik: I must be blind ... where on the MediaTomb FAQ did you find that stmt? [00:18] <JennieL> I'm poking around here, http://mediatomb.cc/dokuwiki/faq:faq, and am missing it [00:18] <Mavrik> 4.2.1: http://mediatomb.cc/pages/transcoding#id2892931 [00:18] <Mavrik> anyway, sleep, g'nite [00:18] <IamTrying> durandal_1707, funny. i worked in one company and had a nice experience where i was working there i had few people and my boss included. 1) when ever i write application using java/python and release it (i always put a README file) 2) none of them use that README file and instead of learning they asks me can you not add this feature and that, which is already given to them 3) LOL - you just reminded me that one by saying that NO LOL [00:20] <JennieL> Mavrik: thanks [00:27] <JennieL> Off to try UMS. TA! [00:31] <durandal_1707> IamTrying: the command you mentioned would when used with ffmpeg mux all stream from all input files into output file [00:37] <IamTrying> durandal_1707, yes temporary its working great thank you!! appreciate it [00:38] <durandal_1707> huh, what is working great? [05:47] <Codewaffle> I'm using ffmpeg to realtime/live transcode video files and stream them straight to an rtmp server (ffmpeg -re -i blah.avi -f flv rtmp://localhost/live/thing) - is there any way to seek/pause/etc when this is going on? [10:42] <inqb> hello [10:43] <inqb> can anyone please help me? [10:46] <spaam> only if you tell us about what you want help with . if not we wont help you. [10:46] <inqb> :) [10:46] <inqb> i'm trying to understand libav libraries [10:46] <inqb> why do avformat_write_header() emits warning: "Codec for stream 0 does not use global headers but container format requires global headers" [10:47] <Mavrik> well, because it's probably true :) [10:48] <inqb> :) [10:48] <Mavrik> your video codec didn't create global header info and the container you're muxing into requires global metadata [10:49] <Mavrik> inqb, which codec and format are we talking about here? [10:49] <inqb> h264/mpeg-ts [10:50] <inqb> h264 stream is demuxed from mov container [10:56] <Mavrik> hmmm, that's wierd [10:56] <Mavrik> TS shouldn't require global headres [10:56] <Mavrik> MOV does however [10:57] <Mavrik> if I recall correctly you're supposed to send that through h264_mp4toannexb bitstreamfilter to H.264 stream gets PPS/SPS packets injected periodically [10:57] <inqb> yes, you're right... [10:57] <inqb> i was trying to write to .mp4 [11:18] <fschuetz> Hello everyone. Trying to compile a slim ffmpeg for android that should only do audio encoding. Trouble is, I seem to need avfilter.a file, which is 27 MB in size. Also I noticed that a lot of code is build, that is disabled in my configure script. Is this intended? [11:19] <Mavrik> fschuetz, nope [11:20] <Mavrik> fschuetz, I suggest you configure it with --disable-everything and then enable things you need [11:20] <Mavrik> you seem to be compiling all filters [11:21] <shlomo> How can I read a video file and get a raw array of pixels for each frame? I couldn't find a simple example [11:22] <Mavrik> shlomo, there's a decode_example.c in doc/examples I think [11:23] <Mavrik> shlomo, you need to demux file, decode video stream and you'll get raw video frames from decoder [11:25] <shlomo> Mavrik, Does doc/examples/demuxing.c do it? [11:25] <Mavrik> demuxing does just demuxing. [11:27] <Mavrik> shlomo, filtering_video.c does it [11:27] <Mavrik> you just don't need the filters bit [11:31] <inqb> how do i fix "H.264 bitstream malformed, no startcode found, use the h264_mp4toannexb bitstream filter (-bsf h264_mp4toannexb)" error? [11:31] <fschuetz> ok. found my problem. thanks [11:31] <inqb> i've tried AVBitStreamFilterContext *filter = av_bitstream_filter_init("h264_mp4toannexb"); [11:31] <inqb> but av_bitstream_filter_filter(filter, c, NULL, &packet->data, &packet->size, packet->data, packet->size, 0); always return 0; [11:32] <shlomo> Mavrik, instead of lines 219-234, just use frame? https://github.com/FFmpeg/FFmpeg/blob/master/doc/examples/filtering_video.c… [11:32] <Mavrik> shlomo, yeah, you'll have raw frame data in frame->data[] array [11:33] <Mavrik> shlomo, remember, you'll probably get a YUV420 format, not RGB :) [11:34] <Mavrik> you can use swscale library to convert formats and resize frames if you have to :) [11:34] <Mavrik> inqb, hmm, I know there were some wierd things about that when used in code [11:34] <Mavrik> I had to inject PSS/SPS packets manually [11:38] <inqb> do you know of any tutorial how to do that? [11:38] <inqb> i can see, that someone had the same problem: http://ffmpeg.org/pipermail/libav-user/2012-March/001484.html [11:38] <Mavrik> no, not really [11:41] <Mavrik> inqb, reading the source of mpegtsenc.c of ffmpeg could give you some idea why does it require that [11:41] <Mavrik> my usecase was different from yours [11:45] <inqb> Mavrik: thx for hinting me about PSS/SPS [11:45] <Mavrik> mhm [11:46] <inqb> i found this post: http://aviadr1.blogspot.com/2010/05/h264-extradata-partially-explained-for.… [11:46] <inqb> and i think this is what i need [11:46] <Mavrik> yeah, that's it [11:46] <Mavrik> you have to create AnnexB format for MPEG-TS [11:46] <inqb> yup [11:46] <inqb> and the input is from .mov file [11:46] <Mavrik> the thing I can't really tell you right now is where will you find SPS/PSS data in your input [11:47] <Mavrik> x264 encoder dups them into it's priv_data field in AVCodecContext [11:47] <Mavrik> you'll probably find them somewhere in AVFormatContext of mov [11:47] <Mavrik> then I just prepended every keyframe with those packets [11:47] <Mavrik> and everything went well [11:47] <inqb> ok [11:58] <inqb> ok, then why [filter = av_bitstream_filter_init("h264_mp4toannexb");] doesn't work? [11:59] <Mavrik> it expects some fields formatted in certain way same with bitstream [11:59] <Mavrik> check the source [11:59] <inqb> maybe i need to pass packet->side_data instead of packet->data to the av_bitstream_filter_filter()? [12:02] <Mavrik> no, because it manipulates the data to inject packets [12:28] <d-fens_> hi, regarding http://ffmpeg.org/pipermail/ffmpeg-user/2013-August/017037.html, can somebody help me out there how to do that with the filter_complex in a fast way? [12:29] <d-fens_> switching the inputs puts the background over the sequence :( [12:39] <inqb> finally some progress [12:39] <inqb> h264_mp4toannexb works fine [12:40] <inqb> i had to set AVCodecContext.extradata [12:40] <inqb> but... [12:41] <inqb> the resulting video plays for 14sec instead of 10min, display only keyframes (i think) [13:06] <inqb> and i found out the reason: pts/dts values weren't rescaled to output stream's time_base [13:36] <fschuetz> Hello everyone. I've built a ffmpeg and would like to use it to convert audio files from different formats to vorbis. Any advice, where I should start? I've checked the doc at http://ffmpeg.org/doxygen/trunk/ but can't really find the entry point. [13:45] <inqb> fschuetz: try http://www.inb.uni-luebeck.de/~boehme/using_libavcodec.html [13:46] <fschuetz> thanks [13:46] <JEEB> ugh [13:46] <JEEB> that's rather old [13:46] <JEEB> at least looking at the changelog ending in 2009 [13:46] <JEEB> so I recommend looking at the samples in the source directory [13:46] <JEEB> as well as keeping some tabs of the ffmpeg doxygen from the main site open [14:02] <fschuetz> ok [14:03] <Mr_E-> https://ffmpeg.org/trac/ffmpeg/ticket/2748 I can confirm this one aswell, It makes a lot of problems not only with hw decoders, but also software like udpxy. [14:09] <fschuetz> avformat_open_input returns an error code. How can I check, what it means? [14:16] <JEEB> Mr_E-, you might want to look into OBE if you want better mpeg-ts muxing [14:22] <Mr_E-> The muxing is fine, it just needs to output 1 size packets only, and not use it as a max value :) [14:32] <Mavrik> JEEB, OBE_ [14:32] <Mavrik> ? [14:43] <elkng> I compilled ffmpeg from source and I used options for gcc: "-O2 -march=i486 -mtune=i686" is it alright to use those options ? or its better to use "-O0" or "-O1" ? [14:48] <JEEBsv> Mavrik: Open Broadcast Encoder [14:48] <JEEBsv> by kierank, uses his libmpegts muxer [14:48] <JEEBsv> which is much stricter than libavformat's as far as I know [14:49] <Mavrik> oh that looks very nice [16:17] <kotfare> hi there ... any quick command to transcode a DBT-S2 , 10mbps from satellite to UDP multicast ? [16:17] <kotfare> i found a lot of commands but most of them doesnt has a smooth streaming on the output [16:17] <kotfare> DVB-S2 * [17:42] <Mista_D> cant pass a vf with ' from a varaible to FFmpeg. http://pastebin.ca/2436477 [18:03] <Codewaffle> any way to render embedded (ssa, stream 0.3 - not already hardsubs) subtitles in one pass? Currently I'm extracting the subtitles with ffmpeg out to SSA and then feeding it back in to a second call and it's a little annoying :/ [18:09] <Mista_D> Got it. vf='-filter:v foo:bar/' works. [18:16] <fschuetz> Is there something similar to avformat_open_input to write a file [18:31] <Mista_D> Image overlay to fade out after 10 sec... Any advise? [18:33] <durandal11707> Mista_D: you need to be much more specifici [18:35] <fschuetz> I am using the Android NDK to convert audio files from different formats to a user supplied output format. [18:37] <fschuetz> I managed to open an input file, find the format and stream and decode a packet to frame [18:38] <Mista_D> durandal11707: -filter_complex 'overlay=10:main_h-overlay_h-10:enable=lte(t,10)' -- expected it to last 5 sec, but the command is refused. [18:38] <fschuetz> So right now I would like to encode the those decoded frames to packets (outputformat) and write them to a file [18:39] <fschuetz> the snippet I am working right now http://pastebin.com/F1BvprFT [18:40] <Mavrik> uh& that's rather& wrong. [18:40] <fschuetz> ok. what do I have to do? [18:47] <durandal11707> Mista_D: how is it refused? [18:48] <Mista_D> durandal11707: [overlay @ 0x15231a20] Option 'enable' not found [18:49] <durandal11707> Mista_D: your ffmpeg is too old [18:50] <Mista_D> durandal11707: new one fails too with " Missing ')' or too many args in 'lte(t' " [18:51] <fschuetz> Can someone give me some directions on encoding/decoding audio please? [18:53] <durandal11707> Mista_D: that is because you do not escape it correctly [18:54] <Mavrik> fschuetz, it's obvious you haven't even read the small documentation about the methods you're using [18:54] <Mavrik> you're ignoring the return types [18:54] <Mavrik> it seems you haven't even checked the examples in doc folder of ffmpeg [18:58] <fschuetz> I do read the doc. However I find it somewhat confusing. I've also tried to find source that does, what I want, as obviously ffmpeg must have it already. Just could not find it. [18:59] <Mista_D> durandal11707: here's command, all escaped well. http://pastebin.ca/2436494 [19:00] <durandal11707> Mista_D: you did not escaped ',' [19:03] <Mista_D> durandal11707: Nice. Thanks. Any way to make it fade at the end? [19:03] <durandal11707> what you need to fade main or overlay or both? [19:08] <Mista_D> durandal11707: Overlay fadeout after 10 sec with complete fade out by 12 sec. [19:09] <durandal11707> then you use fade filter on it, i do not see how is that complicated [19:14] <fschuetz> what's the difference between avcodec_free_frame and av_frame_free [19:26] <CentRookie> hi, I'd like to hear your opinion. when it comes to speed x quality, what do you think brings the best buck per cpu encoding time, ref 8 instead of 4 or bstratagy2 instead of 1 ? [19:27] <CentRookie> to enable both would cost too much in encoding tme [19:35] <cbreak> more reference frames help with noisy footage, like images from a snow storm, rain, filmed content [19:36] <CentRookie> whats the diminishing rate of return treshold? [19:36] <CentRookie> ref 12 ? [19:37] <cbreak> depends. [19:37] <cbreak> a few years ago, x264 was able to print encoding statistics [19:38] <cbreak> you should be able to see the histogram of reference frames and iframe distances and so on [19:44] <CentRookie> well [19:44] <CentRookie> thats not really helpful to a beginner like me [19:47] <Mavrik> why are you even dealing with this then? [19:47] <Mavrik> what kind of use-case prevents you from using presets? [19:48] <CentRookie> why do you guys keep pushing presets? [19:48] <CentRookie> i dont care about presets [19:49] <Mavrik> then learn how x264 works. [19:50] <Mavrik> presets are there for people that don't want to learn how x264 works and what the parameters mean [19:50] <Mavrik> if you don't want to use presets then learn to read the x264 encoding statistics so you can optimize beyond those presets for your use-case [19:54] <cbreak> CentRookie: as beginner, you should just start out with one of the presets, and then extend them as you go and learn how encoding works. [19:54] <cbreak> CentRookie: all parameters are tradeoffs. Sometimes between compression and time, quality and space, or even different kinds of quality. [19:55] <cbreak> and most of them hugely depend on the specifics of the content [19:55] <cbreak> there wouldn't really be a point in options that are not a tradeoff :D [19:56] <CentRookie> well it takes 2 min to read into 1 option, there are roughly 60 or more options, so it would take 120 min just to read them, I have spent already several weeks on this, so I kinda know what they mean. [19:56] <CentRookie> still reading a histogram would go into the matter in deeper [19:56] <CentRookie> not speaking about finding out which of the options are depreciated [19:57] <cbreak> 2 mins? that's quick. [19:57] <CentRookie> as the documentation isn't really up to date concerning ffmpeg 2 [19:57] <CentRookie> 2 mins to get a general understanding [19:57] <cbreak> ffmpeg doesn't really matter [19:57] <cbreak> if you use x264, then it does all the encoding [19:57] <cbreak> all parameters are from it [19:57] <CentRookie> well ffmpeg and x264 documentation should be similiar [19:58] <CentRookie> cant use one without the other except you go with a whole other encoder [20:01] <durandal11707> what documentation? [20:05] <cbreak> CentRookie: I used to use x264 without ffmpeg (directly) [20:06] <cbreak> x264 as encoder, mk4tool for muxing, something else for aac, and mplayer for decoding the input (which in turn, of course, used ffmpeg :) [20:22] <CentRookie> mm yes [20:23] <CentRookie> actually is there a way to automate tests? [20:24] <CentRookie> and to output side by side comparison images [20:24] <CentRookie> dunno if somebody has already thought about that kind of feature [20:24] <CentRookie> im running tests and am creating tens of video files and ahve to test them by hand and the naked eye [20:25] <CentRookie> cbreak, i have a question regarding the first few seconds [20:25] <CentRookie> in all my videos there seem to be an initial phase where image quality is really bad, like the first 3 seconds [20:26] <CentRookie> i have not been able to figure out what causes it [20:26] <CentRookie> might be movflag faststart? [20:26] <cbreak> as far as I know, h264 can't handle fades well. maybe. [20:27] <cbreak> the fast start I know is about the location of the index [20:27] <CentRookie> well that index takes a certain amount of space right [20:28] <cbreak> you'll have it anyway [20:28] <CentRookie> and if bitrate control tries to compensate it, it will affect video quality [20:28] <cbreak> only if you're dumb enough to use bandwidth limiting with two pass or so :) [20:28] <CentRookie> nope, all 1 pass [20:28] <CentRookie> but what do you mean with bandwidth limiting [20:28] <cbreak> with CRF it should try to get consistent quality [20:29] <cbreak> CentRookie: I mean a bitrate mode as opposed to a quality mode like crf [20:29] <CentRookie> ah i see [20:29] <CentRookie> but still dont really understand, isnt rc=2pass often used? [20:30] <cbreak> only if you don't care as much about quality and encode speed [20:30] <cbreak> and instead value exact file sizes most. [20:30] <file> what's this about my size? [20:30] <CentRookie> well its a trade off [20:31] <CentRookie> i try to encode 1 hour worth of video into 120-130mb [20:31] <CentRookie> so am working at quite the border [20:31] <cbreak> then you'd use 2 pass rate limiting [20:31] <CentRookie> no, i use 1 pass [20:31] <CentRookie> havent thought too much abot 2 pass [20:32] <CentRookie> tkes too much encoding time [20:32] <cbreak> one pass will give you uneven quality [20:32] <CentRookie> i know [20:32] <CentRookie> that's the trade you pay [20:32] <CentRookie> im not encoding 1 super important videos [20:32] <CentRookie> but lots of videos [20:32] <CentRookie> thousands [20:32] <cbreak> then I'd use crf [20:32] <CentRookie> yep [20:32] <CentRookie> im using crf 24 [20:33] <cbreak> you can't decide how big the movies will be, but at least they all get the same quality [20:33] <CentRookie> but it tends to blow up file size [20:33] <cbreak> then use a bigger number [20:34] <CentRookie> say, if i do 2 pass but lower ref 8 to 4 and bstratgy 2 to 1, and very fast first pass, then would I get better quality with the lower encoding settings? [20:34] <cbreak> no. [20:35] <CentRookie> in what way no [20:35] <cbreak> 2 pass vs 1 pass doesn't increase quality. [20:35] <CentRookie> the quality should be more even [20:35] <cbreak> it just makes sure it is all more or less the same quality [20:35] <cbreak> less reference frames hurt quality [20:36] <CentRookie> well when we talk about low bitrate encoding, more even quality means higher overall quality [20:36] <cbreak> crf is usually superior to any number of passes in quality [20:36] <cbreak> for the same bitrate [20:36] <cbreak> (and same other settings) [20:36] <CentRookie> depends [20:36] <CentRookie> usually i can get superior compression with 2pass [20:37] <cbreak> nope. [20:37] <cbreak> with the same bitrate, you obviously get the exact same compression. [20:37] <CentRookie> much better quality on smaller bitrate [20:37] <cbreak> and since crf makes more even quality than any number of passes [20:37] <cbreak> you get better quality [20:38] <CentRookie> might be, but it doesnt control bitrate well [20:38] <cbreak> and since you only use one pass worth of encoding time, you could even throw in more expensive options [20:38] <cbreak> it doesn't control bitrate at all [20:38] <CentRookie> lets say i do 1 pass encoding for 1 hour video, for hd i would need at least 1200kbps video rate for decent quality, if i do 2 pass, i can reach the same quality with 600kbps [20:39] <CentRookie> crf is decent enough, but the bitrate control is nowhere as good as in 2pass [20:39] <CentRookie> i get much smaller files with 2pass at same quality [20:39] <cbreak> no [20:39] <CentRookie> well thats my experience [20:40] <CentRookie> from testing hundreds of encodings [20:48] <CentRookie> well i guess it can depend on input and bitrate you are working on [21:47] <hackeron> hey, I have a sound card with 1 audio input that provides 12 channels. How do I tell ffmpeg I want to just channel number 6 only? [21:48] <hackeron> (mono) [22:12] <CentRookie> hmmrrgrr [22:13] <CentRookie> i cant fix this noise effect no matter how high i set the bitrate [22:14] <CentRookie> has anyone experienced that after encoding, the first 3-4 seconds are really ugly and bad in quality, until image quality stabilizes? [22:15] <CentRookie> I thought it might be because of faststart, but it wasnt [22:47] <cbreak> is it bad in the input too? [22:47] <cbreak> is it bad with crf too? [22:49] <Fieldy> i'm recording a full screen game to .mkv with the intention of it being high quality and fast to record, and then later i'll convert it to something smaller. is there something faster than .mkv I could use? disk space isn't an issue nor is speed (it's a high end SSD) [22:51] <sacarasc> .mkv is a container only. [22:51] <sacarasc> The container part of encoding is pretty damn quick. [22:51] <Fieldy> okay so if I am losing frames it's a computational issue and I need to make the area i'm recording smaller? [22:51] <CentRookie> no, input is very good [22:51] <CentRookie> sorry, for late reply, was coding [22:52] <sacarasc> Fieldy: Something like -c:v libx264 -preset:v veryfast -crf 0 will do a lossless dump of it. [22:52] <sacarasc> Sorry [22:52] <sacarasc> *ultrafast [22:52] <sacarasc> I think. [22:52] <CentRookie> crf with max bitrate is bad, yes, without max it is ridiculous huge [22:53] <Fieldy> what does -crf do? it is not in the manpage I have [22:53] <JEEB> it is a x264-specific option [22:53] <JEEB> sets the constant rate factor [22:53] <JEEB> closest thing we have to "constant quality" in the video encoding :P [22:53] <JEEB> CentRookie, how are you using crf + maxrate? [22:53] <JEEB> it sounds like you're doing something wrong [22:54] <Fieldy> probably. the command i'm doing is: ffmpeg -f alsa -ac 2 -i pulse -f x11grab -r 25 -s 1920x1026 -i :0.0+0,54+nomouse -acodec pcm_s16le -vcodec libx264 -preset ultrafast -crf 0 -threads 0 ~/temp/video.mkv [22:54] <JEEB> crf + maxrate (_and_ bufsize) should be rather good [22:54] <Fieldy> to be honest there's so many examples out there in so many places, i end up with a mashup which could easily be wrong [22:54] <JEEB> at least better than similar 1pass bit rate-based [22:54] <JEEB> Fieldy, video-wise that looks fine [22:54] <JEEB> lossless H.264 [22:55] <Fieldy> the issue is that it starts out at 25fps but drops to 14. on playback, it appears to play twice the actual speed [22:55] <JEEB> crf 0 with a 8bit x264 is lossless, with >8bit (9,10bit) x264 the lossless value is somewhere around negative [23:00] <Fieldy> so if i can keep that at 25, i'll be golden. [23:09] <CentRookie> how I am using maxrate and crf? together ^^ [23:09] <CentRookie> it works on certain videos [23:09] <CentRookie> but i have this problem even without crf [23:09] <CentRookie> i used to do ratetol [23:09] <sacarasc> Fieldy: What does your CPU usage look like when you're doing it? [23:10] <Fieldy> sacarasc: not sure, closed everything down, spent way too much time on this today. but probably fairly high, what i'm capturing is very cpu intensive and multithreaded. it's a pretty strong system with lots of ram and a fast SSD, but it can only do so much. thanks for your input [23:11] <sacarasc> CPU is more likely to be the bottleneck over anything else. [23:11] <Fieldy> probably wouldn't have to mess with all of this if the linux version of the flight sim had functioning video capture in it (the win version does, grumble) [23:51] <CentRookie> has anyone experienced "Incomplete MB-tree stats file." while encoding multipass ? [00:00] --- Sat Aug 24 2013

1 0

ffmpeg-devel.log.20130822
by burek 23 Aug '13

23 Aug '13

[01:33] <cone-349> ffmpeg.git 03Michael Niedermayer 07master:3819db745da2: avcodec/rpza: Perform pointer advance and checks before using the pointers [09:53] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:eadb3ad7580a: lavf/tee: initialize ret in parse_bsfs() [11:22] <zidanne> Range header in http.c triggers a bug on some icy (shout cast) servers. [11:22] <zidanne> When I comment out Accept and Range headers in http.c, it connects to problematic servers successfully. [11:35] <wm4> zidanne: what bugs? [11:35] <zidanne> ok, you can try it: [11:42] <zidanne> http://pastebin.com/V9vnXRFN [11:43] <wm4> zidanne: what, both Accept and Range? [11:44] <zidanne> I didn't try commenting out just 1 yet. (I commented out both headers till now). Probably Range is the responsible one. I am trying it now. [11:47] <zidanne> ok [11:47] <zidanne> I tried it an the problem is only at "Range:" header [11:47] <zidanne> Server does not like "Range: bytes=0-\r\n" [11:48] <wm4> it shouldn't be needed (if offset is 0), but according to the code comment, it's to detect seekability with some broken servers [11:49] <wm4> but if this makes some other servers not work at all, this probably shouldn't be done [11:49] <wm4> zidanne: what exactly is happening? here, it doesn't close the connection, just gets... stuck [11:49] <wm4> maybe a workaround is possible [11:49] <zidanne> I tried wireshark [11:49] <zidanne> server stops replying [11:50] <zidanne> and ffmpeg hangs at poll() [11:50] <zidanne> ffmpeg hangs on waiting for data.. it never comes from the server.. [11:50] Action: Daemon404 likes seeing giant lists of avfilter deprecation warnings [11:53] <wm4> zidanne: see commit f0bb88e2bc8b20e0181 [11:54] <wm4> of course it doesn't say details [11:54] <wm4> because, as we all know, commit message bytes are scarce and valuable [11:55] <wm4> zidanne: looks like you could set the seekable option to 0 [11:56] <wm4> I'd rather have it work with all servers by default, and require setting seekable to 1 for servers where we misdetected seekability [12:03] <zidanne> what did you change? I couldn't find the change [12:06] <wm4> try ffplay http://95.211.60.38:6004 -seekable 0 [12:07] <zidanne> so, no change in the code, we have to specifically add seekable 0 to options. [12:08] <zidanne> Maybe seekable could be 0 as default. You may not want to seek by default but play by default (for remote streams). [12:09] <zidanne> (but I understand that this can be a breaking change...) [12:12] <wm4> no, as I explained above, if the offset is 0, the Range header is sent only to determine seekability with some broken (?) servers [12:12] <wm4> but this in turn breaks connecting to your icy server [12:13] <wm4> maybe the problem can be worked around by changing the order of sent headers? after all, the icy server (or http.c) doesn't disconnect, just gets stuck somehow [12:16] <zidanne> if changing the order changes things (It shouldn't according to http protocol definition), it means that the developer of icy server did not implemented http protocol correctly. [12:18] <wm4> a correct implemented http protocol doesn't just freeze after sending headers either [12:20] <zidanne> I tried to contact the developer but no response yet& I will add seekable 0 to options parameter of my avformat_open_input [12:20] <zidanne> thanks, this at least solves the problem in my scenario where I really don't need seekability in my specific situation. [12:54] <durandal_1707> ffmpeg still picks for filter gray format instead rgb when input is pal8 [12:57] <durandal_1707> and it looks like if filter supports gbrp chain will still pick yuv444p [13:02] <cone-987> ffmpeg.git 03Paul B Mahol 07master:139a98be8e2c: lavfi/gradfun: support gbrp [13:08] <cone-987> ffmpeg.git 03Diego Biurrun 07master:0b45269c2d73: x86: h264_idct: Remove incorrect comment [13:08] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:503aec142526: Merge commit '0b45269c2d732d15afa2de9c475d85fcf5561ac4' [13:11] <durandal_1707> git br [13:14] <cone-987> ffmpeg.git 03Rafaël Carré 07master:4622f11f9c83: w32pthread: help compiler figure out undeeded code [13:14] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:221a99aae767: Merge commit '4622f11f9c83db8a2e08408c71ff901826ca652c' [13:42] <cone-987> ffmpeg.git 03Clément BSsch 07master:f8ef91ff3d6b: movenc: add faststart option for web streaming [13:42] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:d68adbd666cb: Merge commit 'f8ef91ff3d6bb83d601d816ef9368f911021c64b' [14:03] <cone-987> ffmpeg.git 03John Stebbins 07master:fe5d5a8ffcaf: movenc: Make chapter track QuickTime compatible [14:03] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:606a30c5a115: Merge commit 'fe5d5a8ffcafdc14c0d26eaea6464c89e120cc9e' [14:21] <cone-987> ffmpeg.git 03Clément BSsch 07master:60198742ff85: movenc: fix detection of 64bit offset requirement [14:21] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:25e4ec6aa14d: Merge commit '60198742ff851f11a3757c713fc75a9c19b88566' [14:28] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:7af7b45c3802: lavf/image2: extend start_number range to accept zero [14:34] <cone-987> ffmpeg.git 03Diego Biurrun 07master:e7b31844f68e: x86: Split DCT and FFT initialization into separate files [14:34] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:f903b4266381: Merge remote-tracking branch 'qatar/master' [14:35] <michaelni> ubitux, mateo` , if you have time please check the past few days changes introduced to mov*c by the merges [14:38] <ubitux> i'm slightly concerned about the co64_required() simplification [14:38] <ubitux> that might be correct but.. [14:38] <michaelni> ubitux, dont hesitate to undo/revert it if you think theres an issue [14:38] <ubitux> i need to look a bit more closely [14:42] <ubitux> my main concern being that i'm not convinced about the tracks being ordered properly in every situation [14:43] <ubitux> but well it's been a long time i haven't look at this stuff [15:08] <mateo`> michaelni: i'll take a look at it this week-end (no time until then) [15:13] <mateo`> ubitux: they picked up your patch ? [15:14] <ubitux> they cherry-picked the faststart patchset, but did a few changes [15:15] <mateo`> have they mentionned why ? [15:15] <ubitux> because it makes the code is simpler, there is a discussion on their ml [15:15] <ubitux> but i didn't read it closely [15:15] <mateo`> ok good [15:18] <durandal_1707> michaelni: why 2850 is not closed? [15:22] <durandal_1707> what about all this license violation bugs? [15:37] <durandal_1707> why 1900 is not fixed? [16:46] <cone-987> ffmpeg.git 03Paul B Mahol 07master:6e643239d995: pngdec: frame multithreading support [17:08] <cone-987> ffmpeg.git 03Paul B Mahol 07master:b1e276f8df25: lavfi/hue: allow changing brightness [00:00] --- Fri Aug 23 2013

1 0

ffmpeg.log.20130822
by burek 23 Aug '13

23 Aug '13

[00:22] <jangle_> I"m attempting to make my own h264 encoded video, and ship it out over rtsp using live555. I'm having problems, and I'm at the point where I believe that I need to make sure I'm using the h264 encoder correctly, so I want to take the encoded data and put it in a matroska file. Since I'm generating the source video myself, I can't use a cli transform or capture program to make the mkv, I need to make it programmatically. Does the ffmpeg project [00:22] <jangle_> provide this capability, or am I confused? [00:23] <sacarasc> IYou should be able to do that, but I don't know how, personally. [00:24] <sacarasc> And make sure you're using the libx264 encoder, as the h264 codec in ffmpeg is just a decoder. [00:31] <jangle_> thanks. I am using x264 [00:31] <jangle_> I am modifying a program that currently works, that encodes using xvid [00:32] <jangle_> but when I put h264 data in the same place, I can crash vlc :-) [00:32] <jangle_> things happen that shouldn't, and the live555 folks think the already working program is faulty, so I need to confirm the functionality at every step, at least for my own sanity [00:41] <cbreak> jangle_: libavformat [00:42] <cbreak> works together with libavcodec if you wan [00:44] <jangle_> I crash in libavcodec when using vlc, so I've been trying to link vlc with a dev build of libavcoded so I can debug with gdb, but thats proving to be harder for me than I thought it would [00:45] <jangle_> does that sound sane? [01:13] <jangle_> http://pastebin.com/ePAEb6WV [01:14] <jangle_> need some illumination. I am setting up my x264 encoder to operate on rgb data. but apparently, my encoder isn't emmitting this specification, so I've explicitly specified to ffmpeg what the pixel format is, and I am presented with the pasted error [01:16] <jangle_> I'm pretty sure I've even linked ffmpeg against the same libx264 binary that I'm using for encoding&. [01:25] <durandal_1707> jangle_: use libx264rgb encoder to encode rgb data as rgb data and not yuv444p [01:26] <durandal_1707> encoder was split because it caused other users to encoder rgb while in fact they wanted yuv [01:27] <durandal_1707> so current inconsistency is because of inifinite human stupidity [01:27] <jangle_> I'm not sure I understand you. in my encoder setup, I specify the input pixel format to be rgb? [01:27] <jangle_> you're saying there's a fork of libx264 that I should use for this purpose instead? [01:29] <klaxa> i don't think it's a fork [01:31] <jangle_> so i would look for this in the ffmpeg project? [01:45] <durandal_1707> jangle_: there are 2 encoders [01:46] <durandal_1707> libx264 accepts only yuv and libx264rgb accepts only rgb [01:46] <durandal_1707> is it now more clear? [01:51] <danharibo> hi [01:51] <danharibo> is it possible to use ffmpeg to stream (with as little CPU utilization as possible) to a different machine running ffmpeg which is streaming to some online service? [02:21] <axorb> RTMP, but VLC might be better at that [02:26] <jangle_> durandal_1707: no, but I figured out that I specify to ffmpeg to use a different encoder (-vcodec libx264rgb), which abstracts the use of libx264 to accept rgb input, not that I had to add a different block of library code to the build of ffmpeg I was using [02:26] <jangle_> durandal_1707: and while I don't get an error about being unable to determine the input color format, my output file is still black [02:30] <jangle_> is there an easy way to specify that no transcoding should be done, that ffmpeg should just stuff the input into the specified container file without modification? [02:32] <klaxa> jangle_: -c:v copy or with the legacy argument -vcodec copy [02:32] <klaxa> using -c:v is recommended though afaik [02:34] <jangle_> thanks, so [02:34] <jangle_> that looks like, -i infile -pix_fmt rgb24, -vcodec -c:v libx264rgb? [02:37] <jangle_> or rather -c:v libx264rgb instead of -vcodec [02:37] <jangle_> so instead of a black screen I get a green one now [02:37] <jangle_> and the file encoder format according to vlc is still wrong, but at least it is different than what it used to be [02:38] <jangle_> ahhh..... [02:39] <jangle_> avcodec in vlc is complaining of an invalid avcodec, gbrp, [02:39] <jangle_> ffmpeg sees that as the input codec pixel format... [02:40] <jangle_> thank you all, this has been illuminating [03:07] <durandal_1707> jangle_: if you see black or green in your player that means your play does not support such profile [03:07] <durandal_1707> *player [03:07] <jangle_> ok [03:07] <jangle_> thanks, tahts good to know [03:09] <durandal_1707> if vlc crash that means vlc is buggy, perhaps you use very old version [03:10] <jangle_> I'm using git trunk, and have reported the crash, which lands in avcoded [03:10] <durandal_1707> and if you want to support many players than you should only use yuv420p format with h264 [03:10] <durandal_1707> git trunk of vlc? [03:11] <jangle_> I"ve been trying to debug my encoder using this crash, thinking they were related [03:11] <jangle_> thanks. I was hoping to do a drop-in replacement and see if avoiding the colorspace conversion would clear up a problem I was having [03:11] <jangle_> yes [03:11] <jangle_> git trunk of vlc. [03:12] <jangle_> that was two weeks ago, so I haven't tried it lately, and I know what setting causes the crash so I stopped using it [03:12] <jangle_> its related to the use of cabac, which I don't yet fully understand, but [03:12] <durandal_1707> could see if output was correct by playing it with mplayer/ffplay [03:13] <durandal_1707> but perhaps you mean you modified x264 encoder? [03:14] <durandal_1707> did you reported crash on vlc bug tracker? [03:15] <durandal_1707> note i consider crash as serious bugs, so that's why i need clarification it does not happen with ffmpeg itself [03:17] <jangle_> I reported the crash on the vlc bugtracker [03:17] <jangle_> I haven't modified the x264 library, just trying to use it [03:18] <jangle_> so [03:18] <jangle_> try it with ffplay or mplayer? [03:19] <jangle_> there is a closed bug of mine I submitted, where it crashes when you play a file I made [03:19] <jangle_> you can try playing that file yourself [03:19] <jangle_> that bug was closed because vlc thinks they've fixed it for 2.1.0-pre, but I'm getting trunk now and I'm going to try it again. [03:29] <jangle_> I still get the crash, unless I"m not updating correctly, or i'm misunderstanding the reason they closed the bug, that it "will be fixed when a feature targeted for 2.1.0 is complete" [03:33] <jangle_> ok, I'm fetching ffplay [03:45] <durandal_1707> jangle_: you have link to file/bug? [03:46] <jangle_> sure one moment [03:46] <jangle_> https://trac.videolan.org/vlc/ticket/9123 [03:47] <jangle_> and the other 2 bug reports I filed are about the same problem, and have crash dumps in instances I was reading from a stream instead of a file [04:04] <jangle_> how do I build ffplay? [05:29] <jangle_> the suspect file plays in ffplay [05:29] <jangle_> and correctly too, or mostly [05:32] <jangle_> cant play my newer files yet though [10:46] <shurnor> why if i --disable-shared and --enable-config, 'file ffmpeg' keeps telling that is 'dynamically linked (uses shared libs)' ? [10:46] <shurnor> why if i --disable-shared and --enable-*static... [10:48] <shurnor> maybe are missing some --extra-libs and/or --extra-cflags, according to http://sopues.blogspot.com.es/2007/02/how-to-compile-ffmpeg-statically.html [11:07] <Mavrik> shurnor, what does ldd say? [11:12] <ealdeguer> hi everyone [11:12] <ealdeguer> My goal is to split a XDCAM or a H264 video, frame-accurately, with ffmpeg. I guess that the problem comes from its long GOP structure, but I'm looking for a way to split the video without re-encoding it. I apply an offset to encode only a specific section of the video (let say from the 10th second to the end of the media) Any ideas ? [11:22] <shurnor> i love you [11:24] <shurnor> says not a dynamic executable (but is targeted to arm built from x86_64) [11:25] <shurnor> am making now a try in 5 minutes with the x86_64 targeted build [11:45] <Mavrik> ealdeguer, well, you will have to split the video on IDR frames, there's no way around it if you don't want to reencode [11:52] <xlinkz0> how do i change the timebase? [11:56] <ealdeguer> Mavrik, thank you it makes sense. I saw on google a tutorial with a YUV reencoding, what is your feeling about that choice ? [11:57] <Mavrik> I don't know what exactly is your use case or your goals [12:11] <xlinkz0> JEEB: do you know by any chance how to modify the time_base ? [12:32] <shurnor> am almost done preparing both static and shared versions for a ldd... [12:37] <shurnor> http://ideone.com/3Ch8aX [12:37] <shurnor> so my answer would be 'static compile yields yet a dynamic binary'? [12:42] <shurnor> (but dependant on seven shared objects instedof fourteen) [13:20] <xlinkz0> could someone please explain how to use the -time_base option? [13:21] <xlinkz0> wherever i put it i get errors [13:21] <xlinkz0> i tried putting it between the input and output file, i get [13:21] <xlinkz0> Codec AVOption time_base () specified for output file #0 (1.mp4) is not an encoding option. [13:21] <xlinkz0> before the input file doesn't work either, idk what to do [13:32] <inqb> hello [14:14] <shurnor> i finally got the 100% static binaries appending -static to extra c cxx and ld flags [14:16] <shurnor> but i dont really know in which of both c cxx and/or ld extra flags is that mandatory ·_· [15:42] <mkozjak> hmm [15:43] <mkozjak> been thinking how to use https://trac.ffmpeg.org/wiki/Create%20a%20mosaic%20out%20of%20several%20inp… to create a video with multichannel output video, but i don't know where to place amerge [15:43] <mkozjak> can someone help to get me started? [15:44] <CentRookie> multi channel or multi stream audio? [15:45] <mkozjak> multistream audio :) [15:46] <mkozjak> so i need a clip with one video stream and 4 stream audio, let's say [15:46] <mkozjak> i'll give it 4 different videos, each of them having one audio and one video stream [15:48] <CentRookie> i see [15:48] <CentRookie> and you want to go with mkv? [15:49] <CentRookie> does your original source have only 1 stream? [15:49] <mkozjak> i could go with mp4 video container, and aac, mp3, or ac3 audio codec [15:50] <mkozjak> CentRookie: my source has two streams, one for audio and one for video [15:50] <mkozjak> possibly it'll have one for subtitles, but i can ignore that [15:50] <CentRookie> no i meant audio stream [15:50] <CentRookie> if the original has already 4 audio streams [15:50] <CentRookie> you can simply copy stream it [15:50] <CentRookie> dont have to reencode [15:51] <CentRookie> if you have 4 files with each 1 audio stream, then its probably best to extract the 4 audio streams first [15:51] <mkozjak> yeah, the latter [15:51] <mkozjak> 4 files with each 1 audio stream [15:51] <CentRookie> as audio1.m4a audio2.m4a [15:51] <CentRookie> yeah [15:51] <CentRookie> good [15:51] <CentRookie> you use map [15:53] <mkozjak> ok [15:54] <CentRookie> ffmpeg -i video.avi -i video1audiosource.avi -i video2audiosource.avi -i video3audiosource.avi -i video3audiosource.avi -i video4audiosource.avi -map 0:0 -map 1:1 -map 2:1 -map 3:1 -map 4:1 -"now come the video codec, bitrate, audio codec, bitrate parameters" finaloutput.mp4 [15:55] <CentRookie> map options work like this "-map fileID:streamID" [15:56] <CentRookie> where 0:0 is like : "Take the first file ( id=0 ) and first stream (video stream id=0) [15:56] <CentRookie> so it is taking the video steam of the first -i inputfile.avi [15:57] <CentRookie> after that it takes the audio streams (id :1) [15:57] <CentRookie> of file 0, 1,2,3 [16:00] <mkozjak> video.avi and video1audiosource.avi are files of similar properties? or video1audiosource.avi only having an audio source? [16:01] <CentRookie> it doenst matter if one of those has multiple streams [16:01] <CentRookie> you can select the file with -map fileID and the stream with -map fileID:streamID [16:01] <mkozjak> ok, thank you, i will go try it now :) [16:02] <CentRookie> IDs are numbered from 0 to n-1 [16:02] <CentRookie> so 0 for first, 3 for forth file [16:03] <mkozjak> ok [16:06] <CentRookie> oh you can add different audio properties like aac and mp3 streams too, but just try it with one audio first [16:08] <mkozjak> CentRookie: can i use '-map' after -filter_complex ? [16:08] <mkozjak> oh, yes, seems like i can [16:10] <mkozjak> now it would be nice to name that audio channels so they have same pids all the time [16:15] <inqb> hello [16:15] <inqb> can someone help me with libavformat? [16:16] <inqb> i'm trying to do some demuxing/muxing without transcoding [16:17] <CentRookie> what do you want to do [16:18] <inqb> i'm trying to demux h264 video stream from big_buck_bunny_480p_h264.mov [16:18] <inqb> and then mux it into some other format (like avi) [16:18] <CentRookie> what did you try until now [16:19] <inqb> getting stream frames with av_read_frame [16:19] <inqb> and then writing them with av_interleaved_write_frame [16:19] <CentRookie> dont you just want to change container? [16:20] <inqb> yes, i do [16:20] <CentRookie> from mov to .avi? [16:20] <inqb> yup [16:20] <CentRookie> then just do ffmpeg -i bunny.mov -c:v copy -c:a copy bunny.avi [16:21] <inqb> i need to do this in C [16:21] <inqb> this is just a first step of POC of bigger project [16:23] <CentRookie> in C? you want to write a program huh [16:23] <CentRookie> well thats the command line [16:23] <CentRookie> you just need to pass that as string to ffmpeg [16:26] <inqb> ok, my goal is to perform live-stream transcoding with NVENC [16:26] <inqb> using FFMPEG as demuxer/muxer [16:30] <inqb> i've managed to demux mpeg4 stream from an avi file and then mux it into new avi container [16:31] <inqb> so this would indicate that the code is good so far [16:32] <inqb> what's special about h264/mov stream/container combination, that the stream can't be just copied to another container? [16:33] <CentRookie> nothing is special [16:33] <CentRookie> you can put h264 stream into any container [16:33] <inqb> the mov file contains tmcd stream, but pts/dts values of h264 looks fine [16:33] <CentRookie> the question is if it is supported by the devices that should read it [16:34] <CentRookie> are you using jwplayer or something to read the stream? [16:35] <inqb> vlc [16:35] <inqb> and when i put the stream to an avi container it says the file lenght is 39min [16:35] <inqb> when i put it into mp4 it says 1:14 [16:35] <mkozjak> is there a compiled ffmpeg 2.x for ubuntu precise? [16:37] <inqb> ffmpeg version git-2013-08-22-3819db7 Copyright (c) 2000-2013 the FFmpeg developers [16:38] <inqb> mkozjak: you can compile it yourself: http://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide [16:38] <mkozjak> getting libx264.c:562: undefined reference to `x264_encoder_open_135' [16:38] <mkozjak> so that's why i'm asking :D [16:38] <mkozjak> i have x264 built and installed [16:39] <inqb> i've followed the instruction today and it worked... [16:39] <inqb> how did you install x264? [16:39] <mkozjak> inqb: from here: https://gist.github.com/faleev/3435377 [16:40] <mkozjak> inqb: probably LD_LIBRARY_PATH is not correct or something [16:40] <mkozjak> i had to link /usr/lib/x86_64-linux-gnu/libx264.so.120 to /usr/lib/libx264.so for ffmpeg to configure correctly [18:18] <Guest21326> why do I see a blue icon called ffdshow in taskbar when converting/editing a video? [18:18] <durandal_1707> maybe because you installed ffdshow [18:50] <jtriley> anyone know if there's a guide on transcoding for adobe fms 3.5 with ffmpeg? cant seem to find any details but cloudfront rtmp streaming of ffmpeg transcoded mp4s seems to have huge audio sync issues [18:50] <jtriley> progressive download of same ffmpeg transcoded mp4s works fine.. [21:16] <shurnor> -static only seemed to be mandatory in --extra-ldflags [21:37] <Case> FFmpeg fails to decode Vorbis file posted here: http://www.hydrogenaudio.org/forums/index.php?showtopic=102350&hl= [21:37] <Case> the file has nonsensical granulepos info but even after those are fixed by stream rewrite decoding fails [21:51] <viric> how can I add a silent audio track to a file? [21:52] <viric> (for concat purposes) [21:52] <viric> I've a v+a file, then v only, then v+a again. [21:52] <durandal_1707> by encoding silence [21:53] <zap0> with the right params you should be able to use /dev/zero as a src. [21:53] <viric> -f lavfi -i aevalsrc=0 can do it, I saw in stackoverflow [21:56] <viric> auhm What is: [21:56] <viric> [Parsed_overlay_4 @ 0x97fd880] Buffer queue overflow, dropping. [21:56] <viric> ? [21:56] <viric> (I'm reencoding a video from a file) [21:57] <durandal_1707> overlay filter tropped some frame because sources have different pts [21:58] <durandal_1707> *dropped [21:58] <viric> pts? [21:58] <durandal_1707> filter use pts to sync both inputs [21:58] <durandal_1707> packet timestamp [21:59] <viric> ah. [21:59] <viric> mh right [22:00] <durandal_1707> so if that is not wanted, use setpts filter.... [22:01] <viric> I'll read about it [22:01] <viric> I don't know much what is that of pts :) [22:03] <viric> setpts to what streams? [22:03] <viric> I've multiple '-i' [22:07] <durandal_1707> the one you use with overlay [22:09] <viric> hm can I know if it's audio requiring that, or video? [22:09] <viric> simply to the video overlay doesn't solve it :( [22:09] <durandal_1707> overlay filter buffers video frames [22:10] <viric> but I think the trouble is with concat [22:10] <durandal_1707> iirc I already have discussion here about concatenation of silence audio [22:13] <viric> sorry, I don't see how that is related. [22:14] <viric> does any line with any loglevel report about pts? [22:14] <viric> two files have tb:1/90000, one tb:1/1000 [22:14] <viric> can it be that? [22:14] <durandal_1707> showinfo shows it among others [22:14] <durandal_1707> that is time base [22:15] <viric> here is the ffmpeg output: http://sprunge.us/HOPV [22:16] <viric> showinfo... let me check. [22:20] <viric> a mistery for me. [22:21] <viric> durandal_1707: sure isn't the timebase thing? [22:24] <durandal_1707> viric: pts uses time base [22:25] <durandal_1707> so both should be same.... [22:26] <viric> I tried adding setpts everywhere, and no success. [22:26] <viric> well, I added 'setpts' without parameters. Which clearly shows that I don't know what I am doing [22:29] <beastd> viric: maybe try to find the overlay examples in the ffmpeg docs. i think it mentions setpts too. [22:30] <viric> ok [22:32] <viric> hm maybe it has to do with concat + later overlay [22:32] <durandal_1707> viric: there is documentation, setpts without any arguments does nothing [22:32] <viric> I'm trying: [22:32] <viric> movie=portada.mov:loop=1,setpts=PTS-STARTPTS [p]; [22:33] <viric> [0:0] yadif,scale=960:540,setpts=PTS-STARTPTS [deint0]; [22:33] <viric> [1:0] showinfo,yadif,scale=960:540,setpts=PTS-STARTPTS [deint1]; [22:33] <viric> [deint0] [0:1] [deint1] [1:1] concat=n=2:v=1:a=1 [outv] [outa]; [22:33] <viric> [outv][p] overlay [outv3]' [22:33] <viric> (I've two -i, for [0] and [1]) [22:33] <durandal_1707> if tb is different you should change it with settb [22:34] <durandal_1707> chaning only pts while tb is different may not work [22:34] <durandal_1707> but you first need to be sure what you are doing [22:35] <viric> so the point is that 'pts' should be equal for every input stream to overlay, right? (checking with showinfo) [22:35] <durandal_1707> i do not know... [22:36] <durandal_1707> if you want to overlay 1-1 frame from 2 sources than yes, otherwise not [22:37] <viric> I have the same fps, so I think I want 1-1 frame. [22:40] <viric> I guess I have to learn about how video works. [22:45] <viric> I think I got pts equal... [22:45] <viric> show input at both overlay inputs: [22:45] <viric> [Parsed_showinfo_2 @ 0x81e8ee0] n:2 pts:14400 pts_time:0.08 [22:45] <viric> [Parsed_showinfo_6 @ 0x81ea620] n:2 pts:14400 pts_time:0.08 [22:50] <viric> mh no idea. [22:50] <viric> showinfo shows the same for both, now, at every frame. Same pts, same pts_time, same n, ... [22:51] <viric> this is my showinfo at overlay inputs:http://sprunge.us/DcDT [22:52] <khali> I have a .avi file of which ffprobe says: 640x360 [SAR 2:1 DAR 32:9], SAR 1:1 DAR 16:9 [22:53] <durandal_1707> so? [22:53] <khali> the SAR 1:1 is correct and the 2:1 is not... can anyone suggest an ffmpeg command that would remove the wrong SAR 2:1? [22:54] <durandal_1707> one is from container and other from bitstream [22:54] <durandal_1707> so if i guessed correct order, remuxing should fix it [22:54] <khali> durandal_1707: which is whic, BTW? [22:57] <viric> argh [22:58] <viric> durandal_1707: do you know what was it? I had *two* output files by error [22:59] <viric> ffmpeg -i .... file.webm file.wbm [22:59] <llogan> what do you mean by, "I had *two* output files by error"? [23:00] <llogan> what is a .wbm file? [23:00] <viric> in my command line, I didn't notice I wrote *twice* the output file [23:00] <durandal_1707> viric: same thing is encoded into 2 files [23:00] <viric> file.webm file.webm [23:01] <viric> If I type a single file.webm as output, I don't have any buffer overflow [23:06] <viric> uf. I was almost learning all the pts thing, had all matched... and yet didn't work. Buffer overflow. Good that I noticed the double filename. [23:06] <llogan> becase the two output files are different. one is using default settings since it has no output options applied to it [23:06] <llogan> ...i assume. i have no idea what your command is [23:12] <khali> durandal_1707: I ended up reencoding the file to mp4... avi sucks anyway [23:13] <durandal_1707> khali: you mean remuxing [23:13] <durandal_1707> transcoding decrease quality for lossy codecs [23:14] <khali> durandal_1707: no, I had the original source still available, so I encoded it to x264 in mp4 container [23:14] <khali> smaller and nice it is [23:14] <khali> thanks for your support BTW [23:14] <khali> now is bed time, see you [23:32] <IamTrying> How can i join 3 mov files to all-in-one.mov file? this is not working: $ mencoder -oac copy -ovc copy -idx -o output.mov Supernova1.mov Supernova2.mov Supernova3.mov [23:33] <IamTrying> No audio [23:33] <durandal_1707> IamTrying: this is not mencoder support channel [23:34] <IamTrying> durandal_1707, OK how do you do it with FFmpeg? [23:34] <IamTrying> durandal_1707, just need 3 mov file to 1 mov file with video/audio [23:34] <durandal_1707> if there all have same codecs/fps/timebase/format/order/etc with concat demuxer [23:35] <IamTrying> durandal_1707, like this ffmpeg -i "concat:input1.mpg|input2.mpg|input3.mpg" -c copy output.mpg ? [23:36] <durandal_1707> nope, that is concat protocol, same as cat [23:37] <JennieL> Hello. I have a question about which decoder vcodec to use. For a src file that mediainfo tells me Video:Codec ID: --> "V_MPEG4/ISO/AVC", would I use vcodec== h264, mpeg4, msmpeg4v1, msmpeg4v2 or msmpeg4v3? I think I'd use h264, but they seem to overlap. [23:38] <JennieL> In case it matters, the stream output format I want is "mpegts" [23:38] <llogan> IamTrying: http://ffmpeg.org/faq.html#How-can-I-join-video-files_003f [23:38] <llogan> https://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20concatenate%20%28join%2C%20m… [23:39] <durandal_1707> JennieL: libx264 [23:39] <durandal_1707> for encoder, for decoder h264 [23:40] <axorb> JennieL: why not use ffprobe instead of mediainfo? [23:41] <JennieL> durandal_1707: Ok, thanks [23:41] <JennieL> axorb: Well, for starters, I hadn't FOUND ffprobe yet ;-p [23:41] <axorb> :P [23:41] <JennieL> I have now ... [23:41] <axorb> ffprobe <input_file> -show_streams [23:42] <IamTrying> llogan, if i use this then i lose -sameq ? http://stackoverflow.com/a/7333453/285594 [23:42] <JennieL> axorb: much better, thx! [23:42] <IamTrying> durandal_1707, if i convert mov to mpg and cat mpg to ffmpeg i lose -sameq of mov ? [23:43] <llogan> sameq no longer exists [23:43] <axorb> JennieL: and if you have an updated version of ffmpeg, -print_format json [23:43] <llogan> http://stackoverflow.com/a/7333453/285594 [23:43] <llogan> http://ffmpeg.org/faq.html#Why-was-the-ffmpeg-_002dsameq-option-removed_003… [23:43] <IamTrying> llogan, will it keep video/audio quality ??? and same .mov format? [23:43] <llogan> https://trac.ffmpeg.org/wiki/Option%20%27-sameq%27%20does%20NOT%20mean%20%2… [23:45] <IamTrying> llogan, so use -qscale 0 ? [23:45] <IamTrying> ffmpeg -i 1.mov -qscale 0 1.mpg ? [23:45] <JennieL> axorb: Thanks -- those'll be helpful I hope. I've been having a real problem trying to get my stream output responsding to controls, and to troubleshoot I've been trying to make sure everything is "right" step by step. HAving good/easy info output helps. [23:46] <IamTrying> cat 1.mpg | ffmpeg -f mpeg -i - -qscale 0 -vcodec mov output.mov ? [23:46] <IamTrying> VAlid? ffmpeg -i 1.mov -qscale 0 1.mpg; ffmpeg -i 2.mov -qscale 0 2.mpg; cat 1.mpg | ffmpeg -f mpeg -i - -qscale 0 -vcodec mov output.mov ? [23:47] <IamTrying> s/cat 1.mpg/cat 1.mpg 2.mpg [23:47] <llogan> IamTrying: have you read any of the links i gave you? [23:47] <IamTrying> VALID or INVALID ? ffmpeg -i 1.mov -qscale 0 1.mpg; ffmpeg -i 2.mov -qscale 0 2.mpg; cat 1.mpg 2.mpg | ffmpeg -f mpeg -i - -qscale 0 -vcodec mov output.mov ? [23:47] <IamTrying> llogan, yes, and its too much to read i just need 3 files in 1 [23:48] <IamTrying> Can i use not use this to keep same quality for video codec/audio codec? $ ffmpeg -i 1.mov -qscale 0 1.mpg; ffmpeg -i 2.mov -qscale 0 2.mpg; cat 1.mpg 2.mpg | ffmpeg -f mpeg -i - -qscale 0 -vcodec mov output.mov ? [23:49] <IamTrying> -vcodec for mov is mov or something else? [23:50] <axorb> libx264 or something [23:50] <axorb> mov is just a container [23:51] <IamTrying> while playing with VLC axorb , showing video : H264-MPEG4 AVC, audio: A52 Audio (aka AC3) (a52) [23:51] <axorb> oh, you can just use copy [23:51] <axorb> -vcodec copy -acodec copy [23:51] <IamTrying> axorb, libx264 is open source broken h264 no? [23:51] <axorb> but that won't rencode [23:51] <IamTrying> ok copy is good axorb [23:52] <axorb> it'll literally use the same data as the input file, just wrap it in a mpg instead of a mov [23:52] <llogan> help vampire [23:52] <axorb> and yes, x264 is open source h264 [23:53] <axorb> ** [23:53] <IamTrying> OK - axorb then better never ever use it, cause it wasted my 2 year not giving stable quality [23:53] <axorb> well [23:53] <axorb> I'm using x264 in production and it's perfectly fine [23:54] <IamTrying> axorb, x264 with java crash daily like 10 times in my production release [23:54] <IamTrying> axorb, finally i quit by hating x264 and java [23:54] <axorb> not sure how java factored into things, but ffmpeg using x264 works completely fine [23:54] <JennieL> Anyone here by any chance know some ffpmeg magic for getting an output stream's (h264 .mkv -> mpegts conversion) action-controls to work? Like Pause, Skip, FastForward, etc? [23:55] <axorb> JennieL: you probably want HLS for that [23:55] <axorb> it's Apple's streaming protocol that splits a video into 10 second long ts files [23:55] <Mavrik> uh [23:55] <Mavrik> what's HLS have to do with that_ [23:55] <Mavrik> ? [23:56] <Mavrik> JennieL, that's player dependent, what is your exact use-case? [23:56] <axorb> Mavrik: I've had difficulty with ts files [23:56] <axorb> VLC barely plays them [23:56] <axorb> at least ones produced by ffmpeg [23:56] <JennieL> Mavrik: I've been having a nice discussion all by myself about it: http://ffmpeg.org/pipermail/ffmpeg-user/2013-August/016996.html [23:57] <JennieL> Mavrik: Neither a Panasonic BluRay standalone player or VLC will allow be to control transcoded streams. [23:57] <JennieL> Native mkvs are fine. [23:57] <JennieL> That's "allow me" [23:57] <Mavrik> hmm, I see [23:57] <axorb> does ts have global headers? [23:57] <JennieL> axorb: HLS? New one on me. Googling. [23:57] <Mavrik> axorb, TS is a streaming format [23:57] <Mavrik> so no. [23:57] <axorb> so yeah, you won't get stuff like seek support [23:58] <Mavrik> and TS can be trickplayed and rewinded as well. [23:58] <Mavrik> axorb, yes you will. [23:58] <Mavrik> stop saying stuff without support, c'mon :P [23:58] <axorb> but the player would have to decode everything right? [23:58] <Mavrik> nop [23:58] <axorb> hrm, okay [23:58] <IamTrying> axorb, Failed to set value '0' for option 'qscale' when trying $ ffmpeg -i Supernova1.mov -qscale 0 1.mpg [23:58] <Mavrik> it's a streaming format, you can start playing it anywhere :) [23:59] <Mavrik> JennieL, anyway, that seems like a player limitation with TS container [23:59] <axorb> Mavrik: but you won't know where to seek for a timestamp [23:59] <axorb> I guess you can guess based on the bitrate [23:59] <JennieL> Mavrik: I though VLC was the swiss-army-knife of players. [23:59] <Mavrik> axorb, either the timestamp table is inserted periodically or you guess yeah :) [00:00] --- Fri Aug 23 2013

1 0