[Ffmpeg-devel-irc] ffmpeg-devel.log.20120613

burek burek021 at gmail.com
Thu Jun 14 02:05:02 CEST 2012


[00:23] <CIA-119> ffmpeg: 03Paul B Mahol 07master * rc1d62489e1 10ffmpeg/libavcodec/ (Makefile libvorbisdec.c libvorbis.c): 
[00:23] <CIA-119> ffmpeg: libvorbis: split encoder from decoder
[00:23] <CIA-119> ffmpeg: Also fix build dependencies while here.
[00:23] <CIA-119> ffmpeg: Reviewed-by: Michael Niedermayer <michaelni at gmx.at>
[00:23] <CIA-119> ffmpeg: Signed-off-by: Paul B Mahol <onemda at gmail.com>
[00:30] <bcoudurier> <LittleFoot> looks like they got a team of hacky colege nubs working on the new ffmpeg
[00:30] <bcoudurier> lol
[00:36] <Compn> bcoudurier : in reply to what ? :D
[00:36] <Compn> ehe
[00:37] <ubitux> option renames from libav afaict from a quick look
[00:39] <llogan> option changes in general. anything not 4 years old.
[00:40] <iive> bcoudurier: actually SoC is exactly that, but just said with bad tone.
[00:40] <iive> I mean GSoC.
[01:26] <Compn> http://lists.libav.org/pipermail/libav-devel/2012-June/028931.html
[01:35] <iive> Compn: I think that the patch is already committed and merged. So it may be double disabled.
[01:38] <durandal_1707> crap, there is nothing on m95 wiki page which shows video/sound/palette size
[02:01] <burek> is this a bug
[02:01] <burek> http://pastebin.com/Gk9iMqHV
[02:05] <durandal_1707> -probesize ?
[02:06] <burek> is that related to my question?
[02:07] <durandal_1707> yes
[02:08] <Daemon404> i think its copying vp8 to .ts (why....) and just nto setting an id
[02:08] <burek> there are 2 ffmpeg commands, the 1st one remuxes something from webm to mpegts and 2nd command just tries to read that mpegts as input
[02:08] <Daemon404> because there are none
[02:08] <Daemon404> it should refuse to copy it
[02:10] <burek> yes, I guess it should too
[02:10] <burek> rather than to produce an invalid output
[02:54] <burek> I think this is also a bug: http://pastebin.com/YCfKMAcF
[02:54] <burek> durations are not correct
[02:54] <burek> after encoding and when later trying ffmpeg -i input
[02:55] <burek> 00:02:13.06 vs 00:02:14.59
[10:54] <CIA-119> ffmpeg: 03Stefano Sabatini 07master * r134815a030 10ffmpeg/libavfilter/ (audio.c avfilter.c internal.h video.c): 
[10:54] <CIA-119> ffmpeg: lavfi: change name of tracing functions and of the macro which enables them
[10:54] <CIA-119> ffmpeg: The enabling symbol is renamed FF_AVFILTER_TRACE. Avoid conflicts with
[10:54] <CIA-119> ffmpeg: the DEBUG macro, which may be defined for different purposes.
[10:54] <CIA-119> ffmpeg: 03Stefano Sabatini 07master * r465a72b4b4 10ffmpeg/libavfilter/ (avcodec.c buffer.c): 
[10:54] <CIA-119> ffmpeg: lavfi: move definition of avfilter_copy_buf_props() from buffer.c to avcodec.c
[10:54] <CIA-119> ffmpeg: avcodec.c is where it is defined the lavc/lavfi interface code, so seems
[10:54] <CIA-119> ffmpeg: a more adequate location (and doesn't force the inclusion of
[10:54] <CIA-119> ffmpeg: libavcodec/avcodec.h where it is not required).
[11:58] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * rdded4cb20b 10ffmpeg/libavcodec/utils.c: 
[11:58] <CIA-119> ffmpeg: lavc: fix support of unofficial pix_fmt extensions of jpeg
[11:58] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[11:58] <CIA-119> ffmpeg: 03James Zern 07master * rb9e5c5f178 10ffmpeg/doc/encoders.texi: 
[11:58] <CIA-119> ffmpeg: doc/encoders: update libvpxenc mappings
[11:58] <CIA-119> ffmpeg: max-intra-rate / force_key_frames, forgotten in 695e85f, cb7f062.
[11:58] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[13:47] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r5f8f6243ef 10ffmpeg/libswresample/x86/rematrix.asm: 
[13:47] <CIA-119> ffmpeg: swr: fix 10l use of uninitialized data
[13:47] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[13:59] <ubitux> michaelni: are you interested in a avi for which ffmpeg has some cutting issues with codec copy?
[14:02] <michaelni> ubitux, yes sure but atm iam working on random other things so maybe open a ticket
[14:03] <ubitux> ok, will do later
[15:25] <Vardan> hi all
[15:26] <Vardan> people I want to buffer video frames, I mean I want to have a queue and add there video frames, other code will get frames from that queue, how that is possible with ffmpeg?
[15:27] <av500> write your own app
[15:27] <av500> that uses the libraries
[15:28] <Vardan> what libraries? I'm writing my own :)
[15:28] <Vardan> I have tried to use sys/queue.h, but without accuess
[15:28] <Vardan> *success
[15:29] <av500> look at ffplay.c
[15:30] <av500> it uses queues for audio and video packets
[15:30] <Vardan> ok
[16:50] <sandsmark> http://sprunge.us/ZgRY does this patch make sense, or have I misunderstood something?
[16:53] <Vardan> people how to add AVFrame to queue, but not the pointer, be cause I'm using the same pointer for next frame?
[16:57] <av500> av_dup_packet
[16:58] <Vardan> what difference between AVFrame and AVPacket?
[16:58] <sandsmark> bbl
[17:00] <av500> packet decodes into a frame
[17:02] <Vardan> with this: avcodec_decode_video2 ?
[17:05] <Daemon404> ubitux, is show_streams not supposed to show SAR?
[17:47] <ubitux> Daemon404: as sample_aspect_ratio afaik
[17:48] <ubitux> depending on the format you will see either nothing ("N/A") or it won't be displayed
[17:48] <Daemon404> thats bad
[17:48] <ubitux> why?
[17:48] <Daemon404> bevahior should be consistent
[17:48] <Daemon404> SAR should be available for all formats, even if it's 1/1
[17:49] <ubitux> the idea was it's semantically better to do something like x=loadjson(); if x.get('sar'): ...
[17:49] <nevcairiel> isnt it beneficial to know if its 1/1 explicitly, or assumed to be?
[17:49] <ubitux> instead of x['sar'] == 'N/A'
[17:49] <Daemon404> ubitux, the problem is
[17:49] <Daemon404> i have a mov file
[17:49] <Daemon404> with a set sar (4/3)
[17:49] <Daemon404> and it ISNT displaying
[17:49] <Daemon404> in show_streams
[17:50] <Daemon404> ffprobe file.mov shows it in teh streams list (the sar)
[17:50] <Daemon404> but it is not in the json
[17:51] <ubitux> mmh
[17:51] <ubitux> ffprobe use the sample aspect ratio from the codec context
[17:52] <ubitux> while the default "SAR" info (lavf/utils.c) use the AVStream one
[17:52] <Daemon404> shouldnt it be consistent between the two?
[17:52] <Daemon404> ideally
[17:53] <ubitux> i wonder why it is duplicated
[17:53] <ubitux> and which one should be used
[17:53] <Daemon404> because libav* are the pinnacle for good design
[17:53] <Daemon404> of*
[17:53] <ubitux> :)
[17:54] <ubitux> maybe someone else knowing better the internal could comment
[19:00] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r4a6d790a6f 10ffmpeg/libavformat/nutenc.c: 
[19:00] <CIA-119> ffmpeg: nutenc: use av_assert
[19:00] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[19:00] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r53eaca5d6b 10ffmpeg/libavformat/mpegtsenc.c: 
[19:00] <CIA-119> ffmpeg: mpegtsenc: use av_assert
[19:00] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[20:04] <ubitux> Daemon404: i just came accross av_guess_sample_aspect_ration in lavf/utils
[20:05] <ubitux> i wonder if we should use that instead
[20:05] <ubitux> in both class "SAR" report, and ffprobe output formats
[20:08] <Daemon404> im not informed enough to know if it's sane
[20:08] <Daemon404> the 'guess' part makes me nervous
[20:11] <ubitux> it selects between format, stream or frame
[20:12] <ubitux> Daemon404: http://pastie.org/4081498
[20:12] <ubitux> i didn't test it at all
[20:12] <ubitux> but you get the idea
[20:19] <ubitux> Daemon404: if that work for you i can send a patch
[20:21] <Daemon404> ill try it in a bit
[20:30] <ubitux> mpf, lavf/utils is kinda chaotic
[20:30] Action: ubitux has hard time figuring out what's going on with seek stuff
[20:49] <ubitux> oh and there is also a lavf/seek.c
[22:28] <expresspotato> Hello
[22:33] <llogan> expresspotato: Hello
[22:34] <expresspotato> hi llogan
[22:34] <expresspotato> I'm having trouble integrating the latest version of ffmpeg against an old one from about a year and a half ago. It appears get_buffer defined in aviobuf.c has gone away. Do you know which method I should replace it with?
[22:36] <ubitux> (OT: http://savethesounds.info/)
[22:39] <llogan> ubitux: nice. the modem and dot-matrix printer give me fond memories of 1992 and my Packard Bell.
[22:40] <gnafu> ubitux: Hehe, I liked, "Or, if you like industrial music, try turning on all the thumbnails at once!"
[22:43] <expresspotato> I miss windows 95 the most
[22:45] <ubitux> :)
[22:45] <gnafu> For me, I often think how weird it is that my children will never experience a house phone, which was such a part of my childhood.  My wife and I have already moved exclusively to cellphones.
[22:45] <durandal_1707> cellphones are evil
[22:45] <ubitux> 56k and tetris for me.
[22:46] <llogan> i just got my first cellphone. welcome to 1999.
[22:48] <expresspotato> anyways can someone help me with my problem?
[22:49] <durandal_1707> get_buffer is moved out of libavformat
[22:50] <expresspotato> what should I replace it with / or is there another class that contains it?
[22:51] <durandal_1707> i dunno for what it was used ....
[22:51] <durandal_1707> ... in your code ...
[22:53] <expresspotato> you can see this pastbin http://pastebin.com/tbhenvQ2     :)
[22:55] <durandal_1707> why are you integrating 2 different ffmpeg versions?
[22:56] <durandal_1707> anyway perhaps you need avio_read()...
[22:57] <expresspotato> I'm trying to replace the old one with the latest version
[22:57] <expresspotato> there is existing player wrapper code that you see there ( a small portion )
[23:00] <durandal_1707> in e63a362857d9807b23e65872598d782fa53bb6af get_buffer have been renamed to avio_read
[23:02] <CIA-119> ffmpeg: 03Diego Biurrun 07master * raf10feadc2 10ffmpeg/libavcodec/ppc/ (h264_altivec.c h264_template_altivec.c vc1dsp_altivec.c): ppc: Rename H.264 optimization template file for consistency.
[23:02] <CIA-119> ffmpeg: 03Ronald S. Bultje 07master * r404410a53a 10ffmpeg/libswscale/input.c: 
[23:02] <CIA-119> ffmpeg: sws: fix planar RGB input conversions for 9/10/16 bpp.
[23:02] <CIA-119> ffmpeg: Fixes bug 282.
[23:02] <CIA-119> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[23:02] <CIA-119> ffmpeg: 03Martin Storsjö 07master * r0533868642 10ffmpeg/libavformat/rtmpproto.c: 
[23:02] <CIA-119> ffmpeg: rtmp: Tokenize the AMF connection parameters manually instead of using strtok_r
[23:02] <CIA-119> ffmpeg: This fixes builds on platforms without strtok_r (windows).
[23:02] <CIA-119> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[23:02] <CIA-119> ffmpeg: 03Sean McGovern 07master * r8772997d10 10ffmpeg/ (configure libavdevice/jack_audio.c): 
[23:02] <CIA-119> ffmpeg: jack: update to new latency range API as the old one has been deprecated
[23:02] <CIA-119> ffmpeg: Fixes Bugzilla #279.
[23:02] <CIA-119> ffmpeg: Signed-off-by: Diego Biurrun <diego at biurrun.de>
[23:02] <CIA-119> ffmpeg: 03Anton Khirnov 07master * r24b2f4ed21 10ffmpeg/libavfilter/avfiltergraph.c: avfiltergraph: remove a redundant call to avfilter_get_by_name().
[23:02] <CIA-119> ffmpeg: 03Anton Khirnov 07master * r2f296e39a1 10ffmpeg/ (4 files in 2 dirs): lavfi: allow building without swscale.
[23:02] <CIA-119> ffmpeg: 03Anton Khirnov 07master * r9baeff9506 10ffmpeg/ (13 files in 2 dirs): 
[23:02] <CIA-119> ffmpeg: lavfi: replace AVFilterContext.input/output_count with nb_inputs/outputs
[23:03] <CIA-119> ffmpeg: This is more consistent with naming in the rest of Libav.
[23:03] <CIA-119> ffmpeg: 03Anton Khirnov 07master * r9d0bfc5052 10ffmpeg/ (39 files in 2 dirs): 
[23:03] <CIA-119> ffmpeg: lavfi: make AVFilterPad opaque after two major bumps.
[23:03] <CIA-119> ffmpeg: It will allow adding new fields to it without ABI breaks.
[23:03] <CIA-119> ffmpeg: 03Anton Khirnov 07master * r84b9fbe055 10ffmpeg/ (4 files in 2 dirs): 
[23:03] <CIA-119> ffmpeg: lavfi: add avfilter_pad_get_type() and avfilter_pad_get_name().
[23:03] <expresspotato> @durandal_1707 - thank you - looks to be the same functionalityas the exiting file
[23:03] <CIA-119> ffmpeg: 03Anton Khirnov 07master * rd625136877 10ffmpeg/ (8 files in 3 dirs): lavfi: add channelsplit audio filter.
[23:03] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * rc7b9eab2be 10ffmpeg/: (log message trimmed)
[23:03] <CIA-119> ffmpeg: Merge remote-tracking branch 'qatar/master'
[23:03] <CIA-119> ffmpeg: * qatar/master:
[23:03] <CIA-119> ffmpeg:  rtmp: Add a new option 'rtmp_buffer', for setting the client buffer time
[23:03] <CIA-119> ffmpeg:  rtmp: Set the client buffer time to 3s instead of 0.26s
[23:03] <CIA-119> ffmpeg:  rtmp: Handle server bandwidth packets
[23:03] <CIA-119> (27 lines omitted)
[23:08] <expresspotato> can you tell me how you found these changes? I've got a whole bunch more starting with url_buffering_data and url_set_interrupt_cb
[23:24] <durandal_1707> expresspotato: git log
[23:33] <expresspotato> thanks 
[23:38] <expresspotato> seems for me some entires are either missing or not being expressed correctly though the git log. For example: url_set_interrupt_cb was renamed to avio_set_interrupt_cb, but its since been removed and I cannot find the related commit. See this pastebin for proof... http://pastebin.com/cUjJEmV7
[23:54] <durandal_1707> heh
[00:00] --- Thu Jun 14 2012


More information about the Ffmpeg-devel-irc mailing list