[Ffmpeg-devel-irc] ffmpeg-devel.log.20120827

[burek]

[00:05] <kierank> maister: what are you trying to do with swscale
[00:05] <Daemon404> kierank, rgb scaling
[00:05] <kierank> good luck with that
[00:05] <Daemon404> currently it converts to yuv (with subsampling) and back
[00:05] <Daemon404> yeah gl;hf
[00:05] <kierank> yes
[00:06] <Daemon404> i just use avisynth for that
[00:06] <Daemon404> btw kierank i tested msvc-built libavcodec's avc decoding yesterday against gcc-built
[00:07] <kierank> i saw
[00:07] <Daemon404> with no qpel or cabac asm
[00:41] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r851bac4b06 10ffmpeg/libavfilter/vf_hqdn3d.c: 
[00:41] <CIA-56> ffmpeg: vf_hqdn3d: avoid declaring variable in for()
[00:41] <CIA-56> ffmpeg: This should fix compilation with gcc 2.95 and possibly others
[00:41] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:42] <Daemon404> >supporting gcc 2.95
[00:42] <Daemon404> oh boy 
[00:43] <ubitux> :))
[00:43] <ubitux> consistency Daemon404, consistency ;)
[00:57] <Daemon404> gnafu, fyi, no other errors with pngs
[01:39] <CIA-56> ffmpeg: 03jamal 07master * r95d1af1188 10ffmpeg/Makefile: 
[01:39] <CIA-56> ffmpeg: build: Fix compilation when configuring with a prog suffix
[01:39] <CIA-56> ffmpeg: Fixes ticket #1676
[01:39] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:55] <CIA-56> ffmpeg: 03Marton Balint 07master * ra366beaf6a 10ffmpeg/ffplay.c: 
[01:55] <CIA-56> ffmpeg: ffplay: fix odd YUV width by cropping the video
[01:55] <CIA-56> ffmpeg: Should fix issue #1322.
[01:55] <CIA-56> ffmpeg: Signed-off-by: Marton Balint <cus at passwd.hu>
[01:55] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:01] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rb9b97c2fb9 10ffmpeg/libavformat/aiffdec.c: 
[03:01] <CIA-56> ffmpeg: aiffdec: fix bits per sample
[03:01] <CIA-56> ffmpeg: Fixes Ticket1660
[03:01] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:53] <gnafu> Daemon404: Thanks for checking that.  I appreciate it, and I know it's appreciated by others.
[05:25] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * ra503afb11f 10ffmpeg/libavformat/mpc.c: 
[05:25] <CIA-56> ffmpeg: mpc: clip seeking timestamp in valid range
[05:25] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[05:25] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r4e8b206092 10ffmpeg/libavformat/mpc.c: 
[05:25] <CIA-56> ffmpeg: mpc: fix seeking to the end
[05:25] <CIA-56> ffmpeg: Fixes Ticket1689
[05:25] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[09:15] <CIA-56> ffmpeg: 03Carl Eugen Hoyos 07master * r73b40dafc8 10ffmpeg/libavformat/nutdec.c: Use correct format identifier for uint64_t.
[10:11] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * rc0bca6425d 10ffmpeg/doc/examples/ (Makefile decoding_encoding.c): 
[10:11] <CIA-56> ffmpeg: examples/decoding_encoding: store temporary files in current dir
[10:11] <CIA-56> ffmpeg: Do not clutter the temporary directory with files, also "/tmp" is not
[10:11] <CIA-56> ffmpeg: always available, e.g. in Windows.
[10:11] <CIA-56> ffmpeg: Also add the clean-test Makefile rule, which will clean the generated
[10:11] <CIA-56> ffmpeg: files.
[10:11] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * r3b0e2763f4 10ffmpeg/doc/examples/Makefile: 
[10:11] <CIA-56> ffmpeg: examples/Makefile: apply misc minor fixes to the RM rule
[10:11] <CIA-56> ffmpeg: Use $(RM) in place of rm, drop useless "-r" for removing files.
[10:11] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * r60924dfe55 10ffmpeg/doc/examples/decoding_encoding.c: 
[10:11] <CIA-56> ffmpeg: examples/decoding_encoding: make operation logic more self-evident / predictable
[10:11] <CIA-56> ffmpeg: Add a switch which controls which operation to perform. Useful to test
[10:11] <CIA-56> ffmpeg: only a single feature. Also the new behavior is more self-consistent.
[12:12] <saste> why dsputil_init is deprecated?
[12:30] <saste> I'm getting undefined reference to `ff_MPV_common_init_x86'
[12:31] <saste> apparently caused by dc402854271ef3d4eaaa3c53fa85237e996a06be
[12:31] <saste> that is because HAVE_MMX != ARCH_X86
[12:32] <saste> and I'm not enabling optimizations
[13:09] <saste> michaelni: ffmpeg.org is down
[13:30] <Compn> cant connect to the box either, seems down down
[13:30] <Compn> my tracert is strange :D
[13:32] <Compn> i get two hops and then all timeout
[13:48] <Compn> it seems back up now
[13:48] <Compn> maybe not
[14:00] <Compn>  13:51:37 up 13 min,  2 users,  load average: 0.43, 0.37, 0.23
[14:00] <Compn> fresh boot
[14:09] <Compn> saste : its back up now, down due to power out
[14:09] <saste> Compn: i see
[14:11] <ubitux> saste: calling ff_dsputil will cause problem
[14:12] <ubitux> saste: see 44be67595db74340495eca4534faf56d54e2a028
[14:13] <saste> ubitux: again, we should export that functionality aka make it public
[14:13] <saste> now accessing it from lavfi is hackish at best
[14:14] <ubitux> sure, just saying it will cause problem as is
[14:20] <ubitux> saste: btw, is the ///< [...] /// [...] doxy really working?
[14:20] <ubitux> i think you should use /**< ... */ instead
[14:20] <ubitux> but maybe i'm wrong
[14:20] <saste> ubitux: ah yes
[14:21] <saste> ubitux: I don't care anyway
[14:21] <ubitux> :)
[14:50] <durandal_1707> ubitux: ping
[14:52] <saste> ubitux: http://www.stack.nl/~dimitri/doxygen/docblocks.html
[14:53] <saste> the ///< thing is supported
[14:55] <xxthink> @Compn:?
[14:56] <xxthink> are you online?
[14:58] <ubitux> durandal_1707: pong
[14:58] <ubitux> saste: ok :)
[15:00] <durandal_1707> ubitux: i was just wondering to you have link to failing exr?
[15:00] <ubitux> huh, i don't think so, why?
[15:02] <durandal_1707> nvm, I will wait Daemon404 to open bug report
[15:09] <ubitux> saste: 'amovie=april.flac,showspectrum,pad=iw*2 [a]; amovie=niea7-op.mp3,showspectrum [b]; [a][b] overlay=w'
[15:09] <ubitux> this works
[15:09] <durandal_1707> michaelni: AFAIK that aiff demuxer fix is inconsistent with aiff muxer
[15:09] <ubitux> but if i use a amovie='http\://radio-web-radio' it doesn't work as expected
[15:10] <ubitux> ah it seems to work with showwaves
[15:11] <ubitux> but not when there is one below another one
[16:38] <CIA-56> ffmpeg: 03Diego Biurrun 07master * r3f02c533f3 10ffmpeg/libavcodec/x86/Makefile: build: fft: x86: Drop unused YASM-OBJS-FFT- variable
[16:38] <CIA-56> ffmpeg: 03Martin Storsjö 07master * r125c6c0752 10ffmpeg/libavformat/rtsp.c: 
[16:38] <CIA-56> ffmpeg: rtsp: Free the rtpdec context properly
[16:38] <CIA-56> ffmpeg: The condition for calling the rtpdec cleanup was broken in
[16:38] <CIA-56> ffmpeg: df8cf076c86.
[16:38] <CIA-56> ffmpeg: This fixes a memory leak.
[16:38] <CIA-56> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[16:38] <CIA-56> ffmpeg: 03Luca Barbato 07master * re0f7a9f6e7 10ffmpeg/libavcodec/huffyuv.c: huffyuv: update to current coding style
[16:38] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r4abb88d7e8 10ffmpeg/: (log message trimmed)
[16:38] <CIA-56> ffmpeg: Merge remote-tracking branch 'qatar/master'
[16:38] <CIA-56> ffmpeg: * qatar/master:
[16:38] <CIA-56> ffmpeg:  vf_hqdn3d: Don't declare the loop variable within the for loop
[16:38] <CIA-56> ffmpeg:  huffyuv: update to current coding style
[16:38] <CIA-56> ffmpeg:  huffman: update to current coding style
[16:38] <CIA-56> ffmpeg:  rtsp: Free the rtpdec context properly
[16:38] <CIA-56> ffmpeg: 03Martin Storsjö 07master * r2bd67175c7 10ffmpeg/libavfilter/vf_hqdn3d.c: 
[16:38] <CIA-56> ffmpeg: vf_hqdn3d: Don't declare the loop variable within the for loop
[16:38] <CIA-56> ffmpeg: This C99 feature is generally not used in the libav codebase,
[16:39] <CIA-56> ffmpeg: since it breaks building with some fringe compilers (GCC 2.95,
[16:39] <CIA-56> ffmpeg: MSVC).
[16:39] <CIA-56> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[16:39] <CIA-56> ffmpeg: 03Luca Barbato 07master * rf76e47877c 10ffmpeg/libavcodec/huffman.c: huffman: update to current coding style
[16:58] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r99f5090714 10ffmpeg/libavformat/aiffdec.c: 
[16:58] <CIA-56> ffmpeg: aiffdec: try #2 to fix bps
[16:58] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[17:45] <Compn> j-b : shot5.avi in incoming is verient sn40 codec , works in mplayer with binary codec.
[17:45] <Compn> streams.videolan.org/incoming/Shot%205.avi
[17:46] <Compn> i guess the .txt explains as much
[17:54] Action: Compn wishes he cound find all of the non-working codec bugs in vlc trac easier
[17:57] <ubitux> nyuhu, saste "The code got relicensed with the original author permission."
[17:57] <ubitux> how is that relicensed?
[17:57] <ubitux> it still looks GPL to me
[17:57] <ubitux> (smartblur)
[18:04] <nyuhu> ubitux : yes that's true& I am a bit confused with the licenses so I didn't know if that counted as a relicense
[18:12] <nyuhu> I let it under GPL since michaelni told me that he prefers to wait for companies which needs them under LGPL, though it is true especially for complex filters (which is not the case here so I don't know)
[18:25] <ubitux> then it might be better to just remove the relicense sentence, and add a gpl dependency in the configure
[18:25] <ubitux> (smartblur_filter_deps="gpl")
[18:27] <tiborfabian> hi, i'm new here. i hope that this is the right place to ask my question.
[18:30] <tiborfabian> can i normalize the audio track of a video file using ffmpeg itself? 
[18:30] <ubitux> yes, but this sounds like a user question; so maybe #ffmpeg
[18:31] <kierank> there are tools out there that do normalisation such as ebur128
[18:34] <ubitux> ./ffplay -f lavfi 'amovie=niea7-op.mp3,asplit=3[out1][a][b]; [a]showspectrum=s=640x240,pad=iw:ih*2[spectrum]; [b]showwaves=s=640x240[waves]; [spectrum][waves]overlay=0:h [out0]'
[18:34] <ubitux> e
[18:37] <Compn> ehe
[18:38] <Compn> michaelni : geocodec bastardized h264 > http://samples.ffmpeg.org/V-codecs/g264.avi
[18:38] <Compn> ffmpeg reports [h264 @ 0126de60]no frame!
[18:39] <Compn> well mplayer reports that, of course :)
[18:39] <Compn> since i dont feel like recompiling ffmpeg with G264 in riff.c ...
[18:46] <ubitux> https://ffmpeg.org/trac/ffmpeg/wiki/FancyFilteringExamples#showwavesandshowspectrum  maybe the pic is too huge?
[18:47] <CIA-56> ffmpeg: 03Georg Lippitsch 07master * r5cc5d9d5f7 10ffmpeg/ (libavcodec/dpx.c tests/ref/lavf/dpx): 
[18:47] <CIA-56> ffmpeg: dpx: 10 and 12 bit decoding
[18:47] <CIA-56> ffmpeg: Rewrite 10 bit dpx decoder to decode into GBRP10 color space
[18:47] <CIA-56> ffmpeg: instead of converting to RGB48.
[18:47] <CIA-56> ffmpeg: Add 12 bit decoder to decode into GBRP12 color space.
[18:47] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[19:01] <CIA-56> ffmpeg: 03Reimar Döffinger 07master * r237f53ce85 10ffmpeg/libavcodec/cllc.c: (log message trimmed)
[19:01] <CIA-56> ffmpeg: cllc: simplify/fix swapped data buffer allocation.
[19:01] <CIA-56> ffmpeg: Using the malloc variant avoids pointless memcpy on size
[19:01] <CIA-56> ffmpeg: increase and simplifies handling allocation failure.
[19:01] <CIA-56> ffmpeg: Also change code to ensure that allocation, bswap and bitstream
[19:01] <CIA-56> ffmpeg: reader all use the same size, even when the packet size is odd
[19:01] <CIA-56> ffmpeg: for example.
[19:07] <CIA-56> ffmpeg: 03Kari Lentz 07master * rbe862c0dac 10ffmpeg/libavformat/movenc.c: 
[19:07] <CIA-56> ffmpeg: movenc: support for itunes tempo tag
[19:07] <CIA-56> ffmpeg: Implements support for tmpo atom for mp4 files, typically used to store BPM. -metadata "tmpo=127" as a command line option will record 127 as the BPM in the meta data.
[19:07] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[19:09] <ubitux> nyuhu: btw, you added the swscale dependency in smartblur in the Makefile, but you need to add it as well in the configure
[19:09] <ubitux> so, something like smartblur_filter_deps="gpl swscale" maybe
[19:42] <ubitux> http://www.notcot.com/archives/2012/08/realitat---microsonic-landscap.php
[20:43] <nyuhu> ubitux : ok got it
[20:47] <ubitux> nyuhu: afaict adding the dep in the makefile is to make sure it will have the correct link parameters, and in the configure it will make sure the deps are build (or that your filter won't if the dep isn't satisfied)
[20:48] <nyuhu> I see
[20:50] <Daemon404> "nyuhu" reminds me of banjo kazooie...
[20:51] <nyuhu> never played it :o
[20:54] <ubitux> reminds me of a certain loli cry...
[20:54] <ubitux> (http://lolicri.es/#elfen-lied-lucy)
[20:54] <JEEB> ah, elfen lied
[20:54] <nyuhu> \o/
[20:55] <Daemon404> it scares me that that url is real.
[20:55] <ubitux> :)
[23:39] <ubitux> saste: i think $(RM) is supposed to be "rm -f"
[23:39] <saste> ubitux: is it?
[23:40] <saste> I checked our Makefile code before to post the patch, but maybe i'm wrong
[23:40] <ubitux> well at least it's set to this here
[23:40] <ubitux> not that it matters much
[23:41] <ubitux> also, i wonder if the copyright boilerplate is appropriate for example
[23:42] <ubitux> it's kind of restrictive to request API user to put that copyright in their code
[23:42] <ubitux> for *the examples*
[23:43] <ubitux> though, we are doing it in all of them
[23:48] <saste> ubitux: that license is one of the most permissive
[23:48] <ubitux> yep sure, i'm talking about the Copyright line
[23:49] <ubitux> it's mandatory to keep it for any code based on
[23:49] <saste> also the code is not meant to be used as is, but as a template for other code or to be read as example reference
[23:50] <saste> The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software
[23:51] <saste> this affects only the part which is "copied"
[23:51] <saste> but ianal, and i don't care that much
[23:51] <ubitux> yes sure as i said, it really doesn't matter much
[23:51] <ubitux> 'just wondering :)
[23:54] <ubitux> saste: btw, i added a showspectrum/showwaves example in the fancy filter wiki page
[23:55] <ubitux> not sure if that's what you had in mind the other day
[23:55] <ubitux> feel free to change it :)
[00:00] --- Tue Aug 28 2012