[Ffmpeg-devel-irc] ffmpeg-devel.log.20111211

[burek]

[00:07] <ubitux> wow, what's going on ffmpeg-users? oO
[00:08] <ubitux> ("[FFmpeg-user] ffmpeg issue")
[00:11] Action: Compn isnt following ffmpeg-user
[00:11] <ubitux> heh, phil rhodes taking the opportunity to troll again with licensing :)
[00:12] <Compn> goody
[00:29] <_Thomas> what's the best way of integrating ffmpeg and hw codecs?
[00:42] <ubitux> mmh i wonder if copying the a/v defaults and callbacks in lavfi for subtitles is really worth the effort
[00:42] <ubitux> it really sounds overkill in the case of the subtitles&
[01:12] <CIA-41> ffmpeg: 03Kostya Shishkov 07master * r19341c58e0 10ffmpeg/ (7 files in 4 dirs): 
[01:12] <CIA-41> ffmpeg: Dxtory capture format decoder
[01:12] <CIA-41> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[01:12] <CIA-41> ffmpeg: 03Luca Barbato 07master * r0395d37abb 10ffmpeg/doc/faq.texi: 
[01:12] <CIA-41> ffmpeg: doc: remove some stale entries from the faq
[01:12] <CIA-41> ffmpeg: Some entries should be still amended.
[01:12] <CIA-41> ffmpeg: 03Janne Grunau 07master * r2b53e696c8 10ffmpeg/libavcodec/ptx.c: 
[01:12] <CIA-41> ffmpeg: ptx: emit a warning on insufficient picture data
[01:12] <CIA-41> ffmpeg: Return the whole packet as consumed in this case and not the size the
[01:12] <CIA-41> ffmpeg: packet should have had. Move the insufficient data check into the for
[01:12] <CIA-41> ffmpeg: condition to fix a ISO C90 error on bigendian.
[01:13] <CIA-41> ffmpeg: 03Aneesh Dogra 07master * r97334f106c 10ffmpeg/tests/ (9 files in 2 dirs): 
[01:13] <CIA-41> ffmpeg: utvideo: add fate tests covering all codec variants
[01:13] <CIA-41> ffmpeg: Signed-off-by: Janne Grunau <janne-libav at jannau.net>
[01:13] <CIA-41> ffmpeg: 03Justin Ruggles 07master * r28101f6c4e 10ffmpeg/libavformat/flacenc_header.c: 
[01:13] <CIA-41> ffmpeg: flac muxer: fix writing of file header and STREAMINFO header from extradata
[01:13] <CIA-41> ffmpeg: fixes Bug 119
[01:13] <CIA-41> ffmpeg: 03Paul B Mahol 07master * r3c1f60860d 10ffmpeg/libavformat/flacdec.c: 
[01:13] <CIA-41> ffmpeg: flacdec: Support for tracks in cuesheet metadata block
[01:13] <CIA-41> ffmpeg: Signed-off-by: Justin Ruggles <justin.ruggles at gmail.com>
[02:04] <michaelni> hi durandal_1707 
[02:04] <michaelni> vcodec copy with escape130 to avi works now :)
[02:06] <durandal_1707> yeah, i noticed
[02:09] <durandal_1707> is there any way to get full duration of stream when muxing
[02:19] <Daemon404> i think the best you can get without a full pass trhough the stream is an estimate
[02:31] <michaelni> it shouldnt be hard to pass duration over from the demuxer but as Daemon404 said its just an estimate
[02:31] <michaelni> so iam not sure if its usefull for anything on the muxer side
[02:32] <durandal_1707> file format of particular codec set number of sample in header at begining of file
[02:32] <durandal_1707> and seektable depends on number of samples
[02:32] <durandal_1707> reference encoder creates temporary file
[02:32] <durandal_1707> it is just brain dead design
[02:36] <durandal_1707> does anyone have CDXLPlay.exe, elf is no use to me. exe and source seems no more available on net
[02:40] <durandal_1707> argh, there was gscoc task for cdxl demuxer
[02:51] <durandal_1707> heh source is gone
[02:52] <durandal_1707> on gmane but i'm lucky it is archived
[03:03] <Compn> durandal_1707 : do you have address of where it was hosted ?
[03:03] <Compn> i can probably find it for you
[03:03] Action: Compn is a wiz at finding things
[03:06] <Compn> ftp://static.88-198-9-197.clients.your-server.de/gfx/show/cdxlplay-sdl.i386-aros.zip
[03:06] <durandal_1707> well i'm intereted in exe
[03:06] <durandal_1707> but looks like i will need to read pascal
[03:08] <Compn> oh , is there an exe
[03:08] <Compn> i'll looksee
[03:09] <durandal_1707> pascal looks awfull
[03:09] <durandal_1707> i hate basic
[03:12] <Compn> durandal_1707 : http://www.a500.org/downloads/video/cdxlplay-sdl.win32.rar
[03:12] <Compn> there
[03:13] <Compn> there is exe in there this time :)
[03:14] <Compn> http://www.a500.org/downloads/video/index.xhtml
[03:16] <durandal_1707> tnx, got it working with wine
[03:18] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r1f183733e0 10ffmpeg/MAINTAINERS: 
[03:18] <CIA-41> ffmpeg: MAINTAINERS: extend syntax for entries
[03:18] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:18] <CIA-41> ffmpeg: 03Hanspeter Niederstrasser 07master * rbed04e2b03 10ffmpeg/ (Doxyfile doc/platform.texi ffplay.c libavutil/mem.c): 
[03:18] <CIA-41> ffmpeg: Correct use of OSX to Mac OS X and properly format yasm providers on Mac OS X
[03:18] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:55] <durandal_1707> anyone willing to translate some gernam to english?
[03:55] <durandal_1707> s/gernam/german
[04:19] <Daemon404> comments?
[04:19] <Daemon404> (i dont speak german)
[04:21] <durandal_1707> well google did it
[04:25] <Compn> theres a few german speaking people in here
[04:26] Action: Compn is slow
[04:34] <Daemon404> btw could be worse
[04:34] <Daemon404> certain encoders are still written in FORTRAN.
[04:46] <durandal_1707> yes!!!!, got sound working
[04:55] <Compn> durandal_1707 : with ffmpeg ?
[05:18] <durandal_1707> Compn: yup
[05:25] <Daemon404> i have an extra amd geod (lol) and cortex a8 boards lyign around... what's something interesting to do with em?
[06:13] <durandal_1707> can i use av_get_packet in read_packet() for both audio and video at same time?
[06:17] <michaelni> durandal_1707, you can call av_get_packet() multiple times yes
[06:18] <michaelni> Daemon404, ffmpeg fate clients
[06:18] <durandal_1707> and how it will know which is video and which is audio?
[06:19] <michaelni> stream_index
[06:20] <michaelni> AVPacket.stream_index
[06:28] <Daemon404> michaelni, i was vaguely thinking about trying to set up win32/fate
[06:28] <Daemon404> on the geode
[06:28] <Daemon404> er, win32/clang/fate
[06:30] <michaelni> :)))))
[06:30] Action: Daemon404 doesnt see any windows clang on either projects fate sites
[06:38] <durandal_1707> now it would be nice if there is more cdxl samples
[06:39] <durandal_1707> making encoder before decoder is not fun at all
[06:39] Action: michaelni goes to bed
[06:39] <Daemon404> night
[06:40] <Daemon404> durandal_1707, http://wiki.multimedia.cx/index.php?title=CDXL
[06:40] <Daemon404> have you read that?
[06:40] <durandal_1707> i stopped counting when it was reached 11
[06:41] <durandal_1707> there was even new concept for cdxl ...
[06:42] <durandal_1707> i fail to fetch any other sample from aminet
[06:48] <durandal_1707> oh got one more sample that is 2
[06:53] <Daemon404> cdxl seems pretty straightforward except for the HAM stuff
[06:53] <Daemon404> which i couldnt be arsed to read up on
[06:57] <durandal_1707> luckily i have sample which do not use HAM but RGB
[06:58] <durandal_1707> shesh, i dont
[06:58] <Daemon404> lol
[06:58] <durandal_1707> both two samples are encoded with HAM
[06:58] <Daemon404> have fun
[06:59] <Daemon404> thats why i avoided it
[06:59] Action: Daemon404 runs
[07:02] <Daemon404> btw durandal_1707 
[07:02] <Daemon404> are you a gci student?
[07:02] <Daemon404> (sorry to ask. there's a lot around lately.)
[07:07] <durandal_1707> gci students are mostly making code looks good
[07:08] <durandal_1707> i wonder how much points they get for that
[07:09] <durandal_1707> imho they are just trolling libav irc and mailing list
[07:10] <durandal_1707> i think there is HAM6/HAM8 support for IFF
[07:11] <durandal_1707> perhaps i could just use IFF ass video codec in cdxl demuxer ....
[07:15] <durandal_1707> well this will mostly need refactor, to prevent code duplication
[07:15] <durandal_1707> gtg
[09:45] <CIA-41> ffmpeg: 03Reimar Döffinger 07master * r33a43695c6 10ffmpeg/tests/ref/seek/lavf_wtv: 
[09:45] <CIA-41> ffmpeg: Add wtv seek test.
[09:45] <CIA-41> ffmpeg: I am not entirely sure the seek functionality is working correctly,
[09:45] <CIA-41> ffmpeg: there are some strange cases like successful seek but no dts value.
[09:45] <CIA-41> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[09:45] <CIA-41> ffmpeg: 03Reimar Döffinger 07master * r5335bff9ec 10ffmpeg/tests/ (lavf-regression.sh ref/lavf/wtv): 
[09:45] <CIA-41> ffmpeg: Add wtv lavf regression test.
[09:45] <CIA-41> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[09:45] <CIA-41> ffmpeg: 03Reimar Döffinger 07master * r60b252eb57 10ffmpeg/tests/ (lavf-regression.sh ref/lavf/caf): 
[09:45] <CIA-41> ffmpeg: Add caf format regression test.
[09:45] <CIA-41> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[09:45] <CIA-41> ffmpeg: 03Reimar Döffinger 07master * r7c875c2d59 10ffmpeg/tests/ (lavf-regression.sh ref/lavf/sox): 
[09:45] <CIA-41> ffmpeg: Add regression test for "sox" format muxer/demuxer.
[09:45] <CIA-41> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[11:58] <CIA-41> ffmpeg: 03Stefano Sabatini 07master * r6c44ff38d6 10ffmpeg/ (6 files in 3 dirs): lavfi: add cellauto source
[12:15] <j-b> burek: you have 2 tasks to review
[13:26] <ubitux> mmh fate is a bit yellow today :)
[13:54] <michaelni> ubitux, mostly its just that the machines havent yet gotten the utvideo samples
[13:55] <michaelni> there seems to be a bug in wtv too (which is a test added yesterday)
[13:56] <ubitux> ok :)
[14:13] <CIA-41> ffmpeg: 03Clément BSsch 07master * rcc2355817d 10ffmpeg/ (doc/filters.texi libavfilter/vsrc_life.c): life: add mold, mold_color, life_color and death_color options.
[14:13] <CIA-41> ffmpeg: 03Clément BSsch 07master * rc88d6b18ce 10ffmpeg/libavfilter/vsrc_life.c: life: add seed information in filter summary.
[16:24] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r87d781de7a 10ffmpeg/tests/ (lavf-regression.sh ref/lavf/wtv): 
[16:24] <CIA-41> ffmpeg: regressiontest: try to fix wtv
[16:24] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[17:07] <burek> j-b, ok, done.
[17:59] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * ra7aa628e9c 10ffmpeg/tests/ (lavf-regression.sh ref/lavf/wtv ref/seek/lavf_wtv): 
[17:59] <CIA-41> ffmpeg: regressiontest: try mp2 for wtv
[17:59] <CIA-41> ffmpeg: our ac3 decoder outputs different things on different platforms.
[17:59] <CIA-41> ffmpeg: Idea-by: Reimar
[17:59] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:17] <CIA-41> ffmpeg: 03Reimar Döffinger 07master * rc3068d43a8 10ffmpeg/libavformat/nsvdec.c: 
[18:17] <CIA-41> ffmpeg: Support VP8 in NSV.
[18:17] <CIA-41> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[18:29] <Compn> whos putting vp8 in nsv
[18:29] <Compn> whos still using nsv? :D
[18:32] <Daemon404> who EVER used nsv except nullsoft?
[18:33] <Compn> hehe
[18:33] <Compn> i guess winamptv is still around
[19:23] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r616eaea8ad 10ffmpeg/doc/ffmpeg.texi: 
[19:23] <CIA-41> ffmpeg: doc/ffmpeg.texi
[19:23] <CIA-41> ffmpeg: Merge changes done to avconv.texi since the last merge into ffmpeg.texi
[19:23] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[19:23] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r16abd68779 10ffmpeg/ (libavformat/seek-test.c tests/ref/seek/lavf_wtv): 
[19:23] <CIA-41> ffmpeg: seektest: floatfree
[19:23] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[20:44] <durandal_1707> so this cdxl format may have several kind of encoded video
[20:44] <durandal_1707> should i create video codec for each encoded version
[20:44] <durandal_1707> or put all shit into cdxl video decoder?
[20:52] <michaelni> if there are several streams then the demuxer should create several AVStreams
[20:52] <michaelni> havnt looked at the spec ...
[20:53] <durandal_1707> format have two streams per frame header: video and audio
[20:53] <durandal_1707> but video can be encoded in 4 ways
[20:56] <durandal_1707> so i will create CODEC_ID_CDXL_(RGB,HAM,YUV,AVM) and put them in libavcodec/cdxl.c
[21:02] <michaelni> cant say from that if 4 codec_ids or 1 codec_id + something in extradata is better ...
[21:02] <michaelni> AVCodecContext.extradata that is
[21:29] <merbanan> durandal_1707: if they are similar in structure use the same codec id
[21:30] <merbanan> durandal_1707: should be easy to start with one format to begin with
[22:40] <Compn> http://techcrunch.com/2011/12/09/apple-made-a-deal-with-the-devil-no-worse-a-patent-troll/
[22:40] <ubitux> michaelni: n0.9? why not v0.9? :p
[22:41] <Compn> apple going after open source (android) and yet i hear nothing about it.
[22:41] <ubitux> and thx for the release anyway
[22:44] <michaelni> ubitux, libav uses v* tags so that would cause a conflict
[22:44] <ubitux> oh, ok
[23:08] <pasteeater> michaelni: can you also update #ffmpeg topic?
[23:08] <_Thomas> Is there some docs on where in the ffmpeg structure to add support for hardware encoders?  (for various SoCs, for instance)
[23:09] <_Thomas> Some SoCs support complete deocding / encoding of various codecs, for instance
[00:00] --- Mon Dec 12 2011