[Ffmpeg-devel-irc] ffmpeg-devel.log.20111127

[burek]

[00:09] <ubitux> the issue with audio visualizations is similar to the issue with subtitles
[00:09] <ubitux> in the first case we want a filter that takes audio in input, and video in output
[00:11] <ubitux> and in the second case we want a filter that takes a subtitle type in input and output video
[00:39] <CIA-36> ffmpeg: 03Anton Khirnov 07master * r488eec1044 10ffmpeg/libavcodec/avcodec.h: lavc: update doxy to use nondeprecated API.
[00:39] <CIA-36> ffmpeg: 03K.Y.H 07master * r51f316a997 10ffmpeg/libavcodec/cook.c: 
[00:39] <CIA-36> ffmpeg: cook: fix apparent typo in extradata parsing
[00:39] <CIA-36> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[00:39] <CIA-36> ffmpeg: 03Carl Eugen Hoyos 07master * r1484b5dec5 10ffmpeg/libavcodec/flicvideo.c: 
[00:39] <CIA-36> ffmpeg: flicvideo: check extradata_size before accessing extradata.
[00:39] <CIA-36> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[00:39] <CIA-36> ffmpeg: 03Reimar Döffinger 07master * r785baa738a 10ffmpeg/libavcodec/nuv.c: 
[00:39] <CIA-36> ffmpeg: nuv: use FFALIGN.
[00:39] <CIA-36> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[00:39] <CIA-36> ffmpeg: 03Reimar Döffinger 07master * rf6afacdb3b 10ffmpeg/libavcodec/nuv.c: 
[00:39] <CIA-36> ffmpeg: nuv: check per-frame header for validity.
[00:39] <CIA-36> ffmpeg: Since it contains dimensions parsing an invalid one has rather
[00:39] <CIA-36> ffmpeg: annoying effects.
[00:39] <CIA-36> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[00:39] <CIA-36> ffmpeg: 03Reimar Döffinger 07master * r7a62ddb689 10ffmpeg/libavcodec/rtjpeg.c: 
[00:39] <CIA-36> ffmpeg: rtjpeg: check get_block return value for error.
[00:39] <CIA-36> ffmpeg: This avoids crashes due to reading out-of-bounds.
[00:39] <CIA-36> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[00:39] <CIA-36> ffmpeg: 03Reimar Döffinger 07master * r7fb55e0b02 10ffmpeg/libavcodec/rtjpeg.c: 
[00:39] <CIA-36> ffmpeg: rtjpeg: simplify get_block() by using get_bits_left.
[00:39] <CIA-36> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[00:39] <CIA-36> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[00:39] <CIA-36> ffmpeg: 03Mans Rullgard 07master * r3bd1162a52 10ffmpeg/libavcodec/gifdec.c: 
[00:39] <CIA-36> ffmpeg: gif: fix invalid signed shifts
[00:39] <CIA-36> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[00:39] <CIA-36> ffmpeg: 03Mans Rullgard 07master * r93c286e54f 10ffmpeg/libavcodec/qtrle.c: 
[00:39] <CIA-36> ffmpeg: qtrle: simplify 32-bit decode using intreadwrite macros
[00:39] <CIA-36> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[00:39] <CIA-36> ffmpeg: 03Mans Rullgard 07master * r644bff6c9b 10ffmpeg/libavcodec/apedec.c: 
[00:39] <CIA-36> ffmpeg: apedec: fix signed integer overflows
[00:39] <CIA-36> (73 lines omitted)
[00:39] <CIA-36> ffmpeg: 03Mans Rullgard 07master * rff6d9cc558 10ffmpeg/libavcodec/snow.c: 
[00:39] <CIA-36> ffmpeg: snow: fix signed overflow in byte to 32-bit replication
[00:39] <CIA-36> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[00:39] <CIA-36> ffmpeg: 03Michael Niedermayer 07master * r5c15b78e4a 10ffmpeg/: (log message trimmed)
[02:16] <CIA-36> ffmpeg: 03Kostya Shishkov 07master * r801393bc96 10ffmpeg/libavcodec/vc1dec.c: 
[02:16] <CIA-36> ffmpeg: vc1: select interlaced scan table by FCM element
[02:16] <CIA-36> ffmpeg: Interlaced videos can contain progressive frames too and now wrong scantable
[02:16] <CIA-36> ffmpeg: is selected for them.
[02:16] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[02:27] <ubitux> buffersink filters are only for debug, right?
[02:29] <iive> have somebody tooked a look of elenril's decode_audio4() stuff. Am I wrong if I think that they would be using data[] pinters to place each channel into separate plane, and thus the new api comes with build-in limitation of 8 channels max?
[02:30] <ubitux> isn't this ruggles stuff?
[02:31] <iive> humm... right, apologies to elenril .
[02:31] <iive> BBB then?
[02:32] <ubitux> Justin Ruggles (ruggles) ` Ronald S. Bultje (BBB)
[02:33] <ubitux> anyway, it looks like indeed to be limited to 8 channels
[02:33] <ubitux> old api: 4, new api: 8, but that sounds weird
[02:33] <michaelni> 8 ch is already today too little :)
[02:33] <michaelni> 7.1 + stereo downmix for example
[02:33] <ubitux> there is "uint8_t **extended_data;"
[02:34] <ubitux> i think it can be extended through this
[02:34] <ubitux> but i didn't read the details
[02:34] <kierank> afaik they'll just change the #define when they want more channels
[02:34] <ubitux> there was some ABI discussion about this in the first patchset
[02:34] <michaelni> they should start at 16 minimum not 8
[02:35] <iive> it doesn't matter with how many they start, they will never be enough.
[02:35] <ubitux> it can be extended
[02:36] <ubitux> i don't think having this value dynamically allocated is a good idea
[02:36] <iive> and yeh, i've started forgetting name<>nicks
[02:36] <michaelni> sure but one doesnt start at a value that is already at the time one puts it there too small
[02:36] <ubitux> but 8 sounds too little somehow
[02:36] <ubitux> i agree with the 16 value :p
[02:36] <ubitux> anything <= 64 sounds reasonable to me :p
[02:36] <ubitux> but > 8
[02:36] <ubitux> :)
[02:36] <iive> how many speakers are in a theater?
[02:38] <iive> oh, justin is the ac3 guy.
[02:39] <iive> anyway, you can easily represent the channels as height, and the number of samples as width.
[02:41] <iive> then depending on the format (planar vs packed) the linesize could be the size of one plane or size of all-channels-sample.
[02:42] <iive> aka planar -> linesize=width*sizeof(sample) ; packed = channels*sizeof(sample).
[03:24] <CIA-36> ffmpeg: 03Peter Ross 07master * r3d977edb04 10ffmpeg/libavcodec/iff.c: 
[03:24] <CIA-36> ffmpeg: HAM6/HAM8 support for IFF ACBM decoder
[03:24] <CIA-36> ffmpeg: Based on patch by ami_stuff
[03:24] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:24] <CIA-36> ffmpeg: 03Peter Ross 07master * rb488679510 10ffmpeg/libavcodec/iff.c: 
[03:24] <CIA-36> ffmpeg: iff: fix invalid reads (ticket 689)
[03:24] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:16] <relaxed> sudo shutdown -h now
[04:16] <relaxed> sorry
[06:12] <CIA-36> ffmpeg: 03Michael Niedermayer 07master * r2bb79b23fe 10ffmpeg/libavcodec/pthread.c: 
[06:12] <CIA-36> ffmpeg: pthread: next try on freeing threads without crashing.
[06:12] <CIA-36> ffmpeg: This should fix mingw
[06:12] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[06:24] <CIA-36> ffmpeg: 03Michael Niedermayer 07master * r47044625ba 10ffmpeg/libavcodec/pthread.c: 
[06:24] <CIA-36> ffmpeg: pthread: check pthread_create() return value.
[06:24] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[15:19] <CIA-36> ffmpeg: 03Michael Niedermayer 07master * r102a2463d3 10ffmpeg/libavformat/img2.c: 
[15:19] <CIA-36> ffmpeg: img2: Allow writing multiple files onto the same output file.
[15:19] <CIA-36> ffmpeg: Fixes Ticket687
[15:19] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[15:56] <CIA-36> ffmpeg: 03Michael Niedermayer 07master * rb7c7eae7d9 10ffmpeg/libavformat/tta.c: 
[15:56] <CIA-36> ffmpeg: tta: better check for totalframes.
[15:56] <CIA-36> ffmpeg: Avoids crash, Fixes Ticket 690
[15:56] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[16:40] <CIA-36> ffmpeg: 03Michael Niedermayer 07master * re99c4bbdf3 10ffmpeg/libavformat/img2.c: 
[16:40] <CIA-36> ffmpeg: img2: update first file only when -updatefirst is specified
[16:40] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:37] <CIA-36> ffmpeg: 03Michael Niedermayer 07master * r6d13499be0 10ffmpeg/ (avconv.c cmdutils.c cmdutils.h ffmpeg.c ffplay.c): 
[21:37] <CIA-36> ffmpeg: cmdutils: pass AVCodec to filter_codec_opts()
[21:37] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:37] <CIA-36> ffmpeg: 03Michael Niedermayer 07master * r1b1223872d 10ffmpeg/libavcodec/ (aaccoder.c aacenc.h): 
[21:37] <CIA-36> ffmpeg: aacenc: add AAC_CODER_NB
[21:37] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:37] <CIA-36> ffmpeg: 03Michael Niedermayer 07master * re64edeed3c 10ffmpeg/libavcodec/ (aacenc.c aacenc.h): 
[21:37] <CIA-36> ffmpeg: aacenc: make the aac coder user choosable.
[21:37] <CIA-36> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:50] <j-b> ANNOUNCE: if you want to add more GCI tasks, this is NOW, or never.
[21:50] <iive> ?
[21:54] <JEEB> Google closing the registration?
[22:05] <michaelni> j-b, how much time is left ?
[22:06] <j-b> tonight
[22:08] <michaelni> :(
[22:08] <michaelni> anyone in here wanting to help mentoring ?
[22:09] <michaelni> ill add a few more tasks but i dont have so much time, its better if i maintain ffmpeg than mentor ...
[22:10] <Compn> you could ask that other project about it :P
[22:10] <Compn> starts with an L 
[22:11] <michaelni> Compn, i dont think they will mentor for ffmpeg
[22:11] <Daemon404> they arecurrently mentoring a crapload of libav stuff for gci
[22:13] <Compn> ah
[22:13] <Compn> the gci website sucks
[22:13] <Compn> only 18 projects listed, dont see ffmpeg or libav
[22:13] <Compn> only videolan
[22:13] <michaelni> Compn, if you want to register i can help you
[22:13] <Compn> no no, too much work
[22:13] <Compn> just trying to look at it
[22:13] <michaelni> yes you need to join videolan
[22:14] <Compn> ah
[22:14] <michaelni> the only way to register is to manually type in some urls that you can find with google on non google pages :)
[22:16] <michaelni> "Find a report a new bug for FFmpeg" is what Code Documentation Outreach User Interface Research Training Translation Quality Assurance ?
[22:16] <Compn> research
[22:23] <michaelni> how many should i create ?
[22:23] <michaelni> ... each task is only good for one bug report
[00:00] --- Mon Nov 28 2011