[Ffmpeg-devel-irc] ffmpeg-devel.log.20120120

burek burek021 at gmail.com
Sat Jan 21 02:05:03 CET 2012

[00:02] <ubitux> Compn: isn't it cigaes?
[00:02] <ubitux> i don't remember seeing him on irc
[01:22] <CIA-31> ffmpeg: 03Clément BSsch 07master * r9d35fa43ee 10ffmpeg/libavfilter/ (af_pan.c vf_fade.c vf_yadif.c vsrc_color.c vsrc_mandelbrot.c): lavfi: add missing periods in descriptions.
[02:25] <CIA-31> ffmpeg: 03Diego Biurrun 07master * r45f2908d02 10ffmpeg/ (avconv.c avplay.c cmdutils.c cmdutils.h ffmpeg.c): cosmetics: Rename ffsink to avsink.
[02:25] <CIA-31> ffmpeg: 03Diego Biurrun 07master * r395e6402fe 10ffmpeg/tests/ (42 files in 2 dirs): fate: Give some tests more sensible names.
[02:25] <CIA-31> ffmpeg: 03Diego Biurrun 07master * re5903e2972 10ffmpeg/tests/fate/ (22 files): fate: add some shorthands to run groups of tests
[02:25] <CIA-31> ffmpeg: 03Michael Niedermayer 07master * r48706f41e1 10ffmpeg/: (log message trimmed)
[02:25] <CIA-31> ffmpeg: Merge remote-tracking branch 'qatar/master'
[02:25] <CIA-31> ffmpeg: * qatar/master:
[02:25] <CIA-31> ffmpeg:  mpeg12: check for available bits to avoid an infinite loop
[02:25] <CIA-31> ffmpeg:  fate: add some shorthands to run groups of tests
[02:25] <CIA-31> ffmpeg:  fate: Give some tests more sensible names.
[02:25] <CIA-31> ffmpeg:  cosmetics: Rename ffsink to avsink.
[02:25] <CIA-31> ffmpeg: 03JindYich Makovi
ka 07master * ra85ce653fb 10ffmpeg/libavcodec/mpeg12.c: 
[02:25] <CIA-31> ffmpeg: mpeg12: check for available bits to avoid an infinite loop
[02:25] <CIA-31> ffmpeg: Signed-off-by: Jindrich Makovicka <makovick at gmail.com>
[02:25] <CIA-31> ffmpeg: Signed-off-by: Diego Biurrun <diego at biurrun.de>
[02:29] <ubitux> michaelni: sorry for the merge conflict with the tests/fate/*.mak
[02:29] <ubitux> michaelni: i didn't expect Diego to do exactly what i did, but putting the rename before&
[02:32] <ubitux> i guess those reindent, cosmetics and rename are the most painful things to merge... :p
[06:35] Action: Compn wonders why ronald emails one person about doing a joint venture with the two projects for soc instead of just making a post on -devel
[06:36] Action: Compn wonders to himself :D
[06:36] <ohsix> he used to be a cool guy
[13:21] <Cervajz> Hello
[13:21] <Cervajz> How can I tell "avformat_find_stream_info" to do not probe private streams in TS, please?
[13:21] <Cervajz> This function starts to read whole TS file when it reaches my private stream
[13:21] <Cervajz> Then it takes very long time
[13:22] <Cervajz> The private streams are marked as "6"
[13:38] <CIA-31> ffmpeg: 03Stefano Sabatini 07master * r72092188ef 10ffmpeg/ffmpeg.c: ffmpeg: remove unused and commented variable OutputStream.sync_ipts
[13:38] <CIA-31> ffmpeg: 03Stefano Sabatini 07master * reef4b70450 10ffmpeg/ffprobe.c: ffprobe: reindent after previous commit
[13:38] <CIA-31> ffmpeg: 03Stefano Sabatini 07master * r4601ad769d 10ffmpeg/ffprobe.c: 
[13:38] <CIA-31> ffmpeg: ffprobe: fix value_string() prefix printing for values with negative logarithm
[13:38] <CIA-31> ffmpeg: The index for the binary_unit_prefixes array is expected to be positive,
[13:38] <CIA-31> ffmpeg: so avoid to use negative indexes for accessing it.
[13:38] <CIA-31> ffmpeg: 03Stefano Sabatini 07master * r58b10b4c7e 10ffmpeg/ffprobe.c: ffprobe: factorize common code in value_string()
[13:38] <CIA-31> ffmpeg: 03Stefano Sabatini 07master * r23934e5f41 10ffmpeg/ffprobe.c: ffprobe: increase precision for the shown float values
[15:24] <ubitux> saste: "Note tha a ratio"  i think you missed a letter here
[15:25] <saste> yes, the doxy contains even more errors
[20:37] <CIA-31> ffmpeg: 03Michael Niedermayer 07master * r2d9e2c6464 10ffmpeg/libavutil/opt.c: 
[20:37] <CIA-31> ffmpeg: opt: Fix "libavutil/opt.c:811:20: warning: initialization discards const qualifier from pointer target type [enabled by default]"
[20:37] <CIA-31> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:15] <mcbaine1>  Hi there ::: I'm really excited about going to FOSDEM  ... I put an advert in the national newspaper to see if anyone else is going and to get in touch etc... Anybody available for abit of travelling / FLOSS advice here atmo.   ??
[21:28] <merbanan> mcbaine1: a few libav people are going
[21:29] <mcbaine1> merbanan, All I hear is that it's the latest and greatest in EUROPE                                   :)
[21:30] Action: gnafu puts his hand up to shield his eyes from the sun as he peers off into the distance, looking at a smiling face some miles away.
[22:32] <ubitux> michaelni: i have a lot of "Truncating packet of size XXX" XXX being 1 or a value between 400 and 1000+ when decoding MP3 since you added the av_log in ee181f84a31d5ba3efd28bd635f249cc5240f5a8
[22:32] <ubitux> is it a normal behaviour?
[22:33] <ubitux> for instance, here i'm decoding a "large" amount of mp3, and got that kind of output: http://pastie.org/3221336
[22:34] <ubitux> i also wonder if the max_analyze_duration warning is really necessary
[22:35] <ohsix> if the analysis fails it's good to know when it finally gives up ;]
[22:35] <ubitux> these are normal mp3 files, generated with ffmpeg
[23:03] <michaelni> ubitux, the mp3 demuxer reads in fixed size amounts and only afterwards the parser splits into mp3 frames
[23:03] <michaelni> so the last read will get truncated
[23:03] <michaelni> its no error
[23:03] <michaelni> somehow we need to suppress the message ...
[23:05] <cbsrobot> michaelni: in sws: yuv2planeX_10_c_template
[23:05] <cbsrobot> there's: int shift = 11 + 16 - output_bits;
[23:05] <cbsrobot> wheres do these numers come from ?
[23:05] <cbsrobot> *numbers ?
[23:06] <michaelni> ubitux, AVIOContext->maxsize= -1 somewhere in mp3dec.c might work
[23:06] <cbsrobot> I'm looking whats needed to add yuv4XXp12*
[23:07] <ubitux> michaelni: ok i'ill look into it, thank you
[23:07] <michaelni> cbsrobot, just use trial and error with the numbers :)
[23:07] <ubitux> but i won't have time until sunday
[23:08] <michaelni> ubitux, well its fairly harmless spam so ill let you fix it when you have time ...
[23:10] <ubitux> a lot of people are wondering what's wrong with the max_analyze_duration warning, and it's only a warning; here the spam is an error, so users might think something is wrong
[23:11] <ohsix> oh, i didn't look at the log, i thought it was just the one :D
[23:19] <mcbaine1> http://goo.gl/6knuF
[23:23] <gnafu> http://longurl.org/expand?url=http%3A%2F%2Fgoo.gl%2F6knuF
[23:28] <durandal_1707> wow hacking base64 got quite popular
[23:29] <CIA-31> ffmpeg: 03Reimar Döffinger 07master * r686ba5f4ea 10ffmpeg/libavutil/base64.c: 
[23:29] <CIA-31> ffmpeg: Minor optimization of base64 decode.
[23:29] <CIA-31> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[23:29] <CIA-31> ffmpeg: 03Reimar Döffinger 07master * r6af4c9814a 10ffmpeg/libavutil/base64.c: 
[23:29] <CIA-31> ffmpeg: Optimize output buffer size check in base64 decode.
[23:29] <CIA-31> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[23:29] <CIA-31> ffmpeg: 03Reimar Döffinger 07master * rd60d718c49 10ffmpeg/libavutil/base64.c: 
[23:29] <CIA-31> ffmpeg: Slightly optimize base64 encode.
[23:29] <CIA-31> ffmpeg: Move handling of last byte outside of innermost loop.
[23:29] <CIA-31> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[23:29] <durandal_1707> is there variant for yasm yet?
[00:00] --- Sat Jan 21 2012

More information about the Ffmpeg-devel-irc mailing list