[Ffmpeg-devel-irc] ffmpeg-devel.log.20140405

burek burek021 at gmail.com
Sun Apr 6 02:05:02 CEST 2014


[00:14] <cone-703> ffmpeg.git 03James Almer 07master:2025d8026fe5: x86/synth_filter: add synth_filter_sse
[00:14] <cone-703> ffmpeg.git 03James Almer 07master:81e02fae6ec1: x86/synth_filter: add synth_filter_avx
[00:14] <cone-703> ffmpeg.git 03James Almer 07master:c74b86699c86: x86/synth_filter: add synth_filter_fma3
[00:14] <cone-703> ffmpeg.git 03Michael Niedermayer 07master:51fd962c0bbf: Merge commit 'c74b86699c86bdf62e8570f41d8a38be5710baa3'
[00:14] <cone-703> ffmpeg.git 03Michael Niedermayer 07master:490d53e335dc: avcodec/x86/dcadsp_init: fix compilation failure without FMA3
[00:22] <cone-703> ffmpeg.git 03Alessandro Ghedini 07master:a49aa440c939: af_volume: implement replaygain pre-amplification
[00:22] <cone-703> ffmpeg.git 03Michael Niedermayer 07master:37af48717067: Merge commit 'a49aa440c939e221194f8d95bf98673f8cf38a06'
[00:43] <cone-703> ffmpeg.git 03Anton Khirnov 07master:9c21c08fa7c5: af_ashowinfo: print a more descriptive name for AV_MATRIX_ENCODING_DOLBY
[00:43] <cone-703> ffmpeg.git 03Michael Niedermayer 07master:b6850e132f36: Merge remote-tracking branch 'qatar/master'
[01:33] <BBB> michaelni: jason checked that a long time ago - the things we really wanted to inline are inlined, afaik, but I don't know details anymore
[01:33] <BBB> michaelni: if you want to inline more and think it gains, go for it
[01:34] <michaelni> no, just wanted to confirm that its ok as is
[01:42] <BBB> I think it is, I mean, stuff might have changed since then, but back then it was ok
[01:42] <BBB> it's a long time ago anyway
[02:04] <Skyler_> I think the inline limits are pretty high nowadays?  I only noticed that happening on older gccs
[02:04] <Skyler_> especially since a lot of the inlining probably doesn't boost code size too much
[02:11] <cone-703> ffmpeg.git 03Michael Niedermayer 07master:a2c6a99ba56a: configure: Fix testing of objective-c headers for qtkit
[02:38] <BBB> Compn: I'm not sure what the point is of asking them anyway
[02:39] <BBB> Compn: it's not exactly alike, but you're kinda asking a copyright violator in china to stop distributing ffmpeg... what's the point? there's no reason for them to stop doing what they're doing, it's perfectly fine... maybe morally rejectable, but who cares about that stuff anyway?
[02:39] <BBB> Compn: if they consider it a great good to cosmetically prettify that code, let them do it
[02:40] <BBB> Compn: if they claim that adds copyright claims, let them think so
[02:41] <Compn> koda asked me to work with the project, cant say i didnt try.
[02:41] <BBB> I prefer to make the end product better
[02:41] <BBB> that's why I work on both hevc as well as vp9, I mean politically that's the silliest thing to do
[02:42] <Compn> thought you were doing it for financial reasons 
[02:43] <Compn> as in, you'd be the person to hire, no matter whos team was hiring. plus they could compete to hire you 
[02:43] <BBB> I can honestly claim that I made $0 on ffvp9, and that I estimate my total income for ffhevc will not be any higher
[02:43] <BBB> actually that's a lie, someone gave me a dinner cheque for ffvp9
[02:43] <Compn> not on those, but on future projects like xevc or xvp9
[02:43] <BBB> but that was more bet-related that anything payment-wise
[02:43] <Compn> lol
[02:44] <Compn> not that i'm dispariging anyone for working for money . i think its good idea 
[02:44] <BBB> I claimed I could make a ffvp9 decoder that'd be faster than ffh264 or ffvp8 for same-quality
[02:44] <BBB> I got laughed at - quite massively
[02:44] <Compn> and i know you've been working hard
[02:44] <BBB> I proved it could be done, and surprise, I got a dinner cheque for that bet
[02:44] <BBB> yay
[02:44] <BBB> anyway, no, I'm not being hired by any multimedia company
[02:45] <BBB> and I don't think I'm really in the market ATM
[02:45] <BBB> I'm ... how do you say that ... hard to get
[02:45] <Compn> i wish you luck! :)
[02:45] <Compn> with whatever you are doing :)
[02:46] <Compn> seems like libavcodec devels are in high demand, i dont see any problems for them
[02:49] <BBB> I don't work in that sector at all right now ;)
[02:50] <BBB> I work here: http://www.twosigma.com/
[02:50] <BBB> also, kids wanna play, brb
[03:31] <cone-703> ffmpeg.git 03Carl Eugen Hoyos 07master:87f29996415a: Force gif aspect ratio multiplication to 64bit.
[03:31] <cone-703> ffmpeg.git 03Carl Eugen Hoyos 07master:9f4ab4e5f9f0: Show stream number when warning about stream without start time.
[03:31] <cone-703> ffmpeg.git 03Carl Eugen Hoyos 07master:90d93eeb386f: Allow .y as extension for image2.
[03:31] <cone-703> ffmpeg.git 03Michael Niedermayer 07master:7a5537fd3fd0: Merge remote-tracking branch 'cehoyos/master'
[03:53] <cone-703> ffmpeg.git 03Thilo Borgmann 07master:8c161d7884f0: lavc/exif: Make EXIF IFD decoding part of private API/ABI.
[03:53] <cone-703> ffmpeg.git 03Thilo Borgmann 07master:2f08dad0286f: lavf/avidec: Read metadata EXIF tags from AVIF tag. Based on patch by Gregory Wolfe (Kodak Alaris) <gregory.wolfe at kodakalaris.com>.
[03:53] <cone-703> ffmpeg.git 03Thilo Borgmann 07master:ec8789ac6353: Reindent after last commit.
[05:47] <cone-703> ffmpeg.git 03James Almer 07master:7d2116dd09d1: x86/synth_filter: compile avx and fma3 functions unconditionally
[10:59] <J_Darnley> um...
[10:59] <J_Darnley> Has the layout of the fate page been deliberately changed?
[11:00] <J_Darnley> Or has something gone seriously wrong?
[11:00] <J_Darnley> Nevermind
[11:20] <superware> can I use the same packet from av_read_frame for two outputs? say avcodec_decode_video2 for display, and av_write_frame for file dumping?
[11:29] <superware> is the righ architecture to use when I have one input and two outputs is to reuse the same AVPacket (av_read_frame)?
[13:48] <cone-668> ffmpeg.git 03Michael Niedermayer 07master:323c049c7e43: configure: more properly disable header when check_header_oc() fails
[13:49] <cone-668> ffmpeg.git 03Michael Niedermayer 07master:3c728ceec66d: configure: use TMPM in check_header_oc
[14:11] <cone-668> ffmpeg.git 03James Almer 07master:a1ac12bddd8d: x86/dcadsp: add ff_dca_lfe_fir0_fma3
[19:05] <cone-788> ffmpeg.git 03Diego Biurrun 07master:3dc6272bed78: Remove a number of unnecessary dsputil.h #includes
[19:05] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:5440151fa42d: Merge commit '3dc6272bed7890a49080e18eacf3c7a4a6594b0d'
[19:10] <cone-788> ffmpeg.git 03Diego Biurrun 07master:b42f49e42f8c: x86: dsputil: Eliminate some unnecessary dsputil_x86.h #includes
[19:10] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:874f27a8f7e3: Merge commit 'b42f49e42f8cde25a788b2d13d03e99ca2956647'
[19:16] <cone-788> ffmpeg.git 03Diego Biurrun 07master:b718b24bbe83: ppc: Drop a bunch of unnecessary dsputil_altivec.h #includes
[19:16] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:865c94bb57db: Merge commit 'b718b24bbe83d3e8787c1dc6449ccbffcd13a4e7'
[19:30] <cone-788> ffmpeg.git 03Diego Biurrun 07master:01c5779f56cf: x86: Drop some unnecessary YASM ifdefs
[19:31] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:ebb21887b853: Merge commit '01c5779f56cf708e6cb88b11cfdc248cae7e2ee8'
[19:36] <cone-788> ffmpeg.git 03Diego Biurrun 07master:c2c5be57494e: x86: h264_qpel: Simplify an #if conditional
[19:36] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:e3c3f277a9db: Merge commit 'c2c5be57494e6117086771bca34c8cd4c72c8e99'
[19:43] <cone-788> ffmpeg.git 03Diego Biurrun 07master:57b5b84e208a: x86: dsputil: Move ff_apply_window_int16_* bits to ac3dsp, where they belong
[19:43] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:e6f69b324e88: Merge commit '57b5b84e208ad61ffdd74ad849bed212deb92bc5'
[19:49] <cone-788> ffmpeg.git 03Diego Biurrun 07master:5f9220437025: dsputil: Move DV-specific ff_zigzag248_direct table to dvdata
[19:49] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:47fc82b5b346: Merge remote-tracking branch 'qatar/master'
[21:44] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:26e70fd5ba50: avcodec/vp8: fix name of copy_luma()
[21:44] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:d167faafe9df: avcodec/wma: use av_freep(), do not leave stale pointers in memory
[21:52] <lukaszmluki> hi, I have question about channel layouts in ffmpeg. For example  AV_CH_LAYOUT_2_1, does ffmpeg always put samples per channel in the same order? In other words, how I know where left channel is placed in stream to not take right channel or LFE instead.
[21:56] <wm4> lukaszmluki: it puts them in the same static order
[21:56] <wm4> lukaszmluki: for each speaker, there's a bit set in the mask
[21:56] <wm4> and the bit index corresponds to the channel number
[21:56] <wm4> err
[21:56] <wm4> not really the bit index
[21:56] <wm4> more like: the Nth set bit has its audio in channel N
[21:58] <lukaszmluki> ok, i expected that, thank for confirm
[22:18] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:a84f9c75bc3f: avformat/omadec: fix probetest failure
[22:25] <Compn> lukaszmluki : theres a file that explains the channel order
[22:25] <Compn> source code
[22:25] <Compn> somewhere :D
[23:13] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:49579c963765: avformat/cinedec: fix sizeof()
[23:13] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:af2e5061bbca: avformat/mov: Fix width/height typo
[23:13] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:47b03415f319: avformat/cinedec: remove redundant zero termination
[23:25] <ubitux> lol koda throwing trolls and asking for everyone to stop contradict him :')
[23:25] <ubitux> wth is wrong with that guy
[23:25] <iive> ?
[23:25] <ubitux> iive: "To be or not to be working with libav" thread
[23:25] <ubitux> Vittorio
[23:26] <ubitux> > But I had really hoped that we could use this thread for something more constructive than "you did this to me", "I don't like you because you smell" and other childish troll attempts.
[23:26] <iive> oh, i though that thread have died.
[23:26] <ubitux> even though he's basically doing that since the beginning
[23:26] <ubitux> :D
[00:00] --- Sun Apr  6 2014


More information about the Ffmpeg-devel-irc mailing list