[00:29] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r0fc684ff7c 10ffmpeg/libavcodec/aacdec.c: [00:29] <CIA-56> ffmpeg: aacdec: pass AVPacket to aac_decode_frame_int() [00:29] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [00:29] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r946ed78f5f 10ffmpeg/ (libavcodec/aacdec.c tests/fate/aac.mak): [00:29] <CIA-56> ffmpeg: aacdec: fix priming/skip for AAC HE/HE2 [00:29] <CIA-56> ffmpeg: There is a remaining error of 2 - 8 samples in some but not all cases, [00:29] <CIA-56> ffmpeg: the source of the error is unknown ATM. [00:29] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [00:29] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r158a70f17a 10ffmpeg/libavformat/mov.c: [00:29] <CIA-56> ffmpeg: mov: fix av_log int type [00:29] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [02:42] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r19008e06c6 10ffmpeg/libavcodec/wmavoice.c: [02:42] <CIA-56> ffmpeg: wmavoice: switch to av_assert [02:42] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [02:42] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rafe6f337ad 10ffmpeg/libswscale/swscale.c: [02:42] <CIA-56> ffmpeg: swscale.c: remove unused variable [02:42] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [05:25] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r19dceef905 10ffmpeg/libavfilter/vf_framestep.c: [05:25] <CIA-56> ffmpeg: vf_framestep: fix uninitilaized variable bug [05:25] <CIA-56> ffmpeg: Found-by: gcc [05:25] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [05:25] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rb3facc4af8 10ffmpeg/libavformat/ac3dec.c: [05:25] <CIA-56> ffmpeg: ac3dec: fix uninitialized var warning [05:25] <CIA-56> ffmpeg: This commit should be purely cosmetic [05:25] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [05:25] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r310fd0d3d4 10ffmpeg/libavfilter/vsrc_testsrc.c: [05:25] <CIA-56> ffmpeg: vsrc_testsrc: fix uninitilaized variable bug [05:25] <CIA-56> ffmpeg: Found-by: gcc [05:25] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [05:25] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r5e363bb9b3 10ffmpeg/libavformat/flvenc.c: [05:25] <CIA-56> ffmpeg: flvenc: kill a uninitialized variable warning. [10:18] <saste> michaelni: when you have some time, could you have a look at the decimate issue? [10:18] <saste> it's in my local repo since March, and I'd like to push it [10:18] <saste> also tell if you're OK with removing mp=smartblur [10:19] <michaelni> saste, "the decimate issue" ? [10:20] <saste> michaelni: yes the bx8 block diff code in decimate seems wrong [10:20] <saste> *8x8 [10:21] <saste> so I had to fix it in order to make it binary-exact with dsputil diff_pixels code [10:22] <saste> and I can't test MMX mp=decimate implementation (is always disabled) [10:24] <michaelni> hmm, yes, seems x is off by 1, y shouldnt matter [10:25] <saste> indeed [10:26] <michaelni> "CpuCaps gCpuCaps; //FIXME initialize this so optims work" [10:26] <michaelni> shouldnt be too hard to initt gCpuCaps from our cpu detect code ... [10:27] <saste> BTW is there some way to set cpu_caps at run time? [10:27] <saste> for example if I want to test C implementation without to recompile everything [10:27] <michaelni> ./ffmpeg -cpuflags ... [10:27] <saste> ok, thanks [11:52] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rd40e2c7a86 10ffmpeg/libavfilter/avfiltergraph.c: [11:52] <CIA-56> ffmpeg: avfiltergraph: kill uninitialized variable warning [11:52] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [11:52] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rb0545800bd 10ffmpeg/libavcodec/wmalosslessdec.c: [11:52] <CIA-56> ffmpeg: wmalosslessdec: remove unused variable [11:52] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [11:53] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r340e8e0b22 10ffmpeg/libavcodec/dnxhdenc.c: [11:53] <CIA-56> ffmpeg: dnxhdenc: fix pointer type warning [11:53] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [13:44] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rf4c020e3ad 10ffmpeg/libswscale/swscale_internal.h: [13:44] <CIA-56> ffmpeg: swscale: fix usePal() [13:44] <CIA-56> ffmpeg: Fixes Ticket1702 [13:44] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [14:39] <Compn> nevcairiel : [aac @ 0126de60]Reserved SBR extensions are not implemented. [14:39] <nevcairiel> reserved sounds wrong [14:41] <Compn> might have just been caused by a seek [14:43] <Compn> because playing it in order doesnt cause that error [17:14] <michaelni> ubitux, is "0830 15:08 Wolfram Gloger (4.0K)  [17:16] <michaelni> or anyone else who knows the affected code ... [17:37] <CIA-56> ffmpeg: 03Mans Rullgard 07master * r8db73c61a7 10ffmpeg/ (5 files in 2 dirs): [17:37] <CIA-56> ffmpeg: build: allow non-standard variations of linker -l/-L flags [17:37] <CIA-56> ffmpeg: This enables replacing the -l and -L flags used to specify the [17:37] <CIA-56> ffmpeg: just-built libraries when linking the tools and shared libs with [17:37] <CIA-56> ffmpeg: non-standard syntaxes. System library flags are already handled [17:37] <CIA-56> ffmpeg: by the filtering mechanism in configure. [17:37] <CIA-56> ffmpeg: Signed-off-by: Mans Rullgard <mans@mansr.com> [17:37] <CIA-56> ffmpeg: 03Anton Khirnov 07master * rbbabeb56fa 10ffmpeg/ (doc/developer.texi libavcodec/avcodec.h): Add reminders to update the codec descriptor list with new codec IDs. [17:37] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * re8e4c8bdde 10ffmpeg/: (log message trimmed) [17:37] <CIA-56> ffmpeg: Merge remote-tracking branch 'qatar/master' [17:37] <CIA-56> ffmpeg: * qatar/master: [17:37] <CIA-56> ffmpeg: build: allow non-standard variations of linker -l/-L flags [17:37] <CIA-56> ffmpeg: Add reminders to update the codec descriptor list with new codec IDs. [18:04] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r0e781d9405 10ffmpeg/libavcodec/utvideoenc.c: [18:04] <CIA-56> ffmpeg: utvideoenc: remove unused variable [18:04] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [18:04] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r370bb893d2 10ffmpeg/libavformat/mp3enc.c: [18:04] <CIA-56> ffmpeg: mp3enc: fix const correctness [18:04] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [18:04] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r6fc892d271 10ffmpeg/libavcodec/libvorbisdec.c: [18:04] <CIA-56> ffmpeg: libvorbisdec: fix const nitpick warning [18:04] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [19:15] <ubitux> michaelni: will look at it in a moment [19:51] <saste> why bugzilla sucks so hard when you want to do a f*****g simple search? [19:52] <ubitux> i never understood how it works [19:53] <ubitux> it just looks like another bloated bug tracker from the 90s [19:54] <Daemon404> still better than clearquest [19:55] <llogan> i hate bugzilla [19:57] <ubitux> Daemon404: the good thing with sucky things is that you can always find worse [19:57] <Daemon404> did someone say bug mailing list? [19:57] <Daemon404> i think someone said bug mailing list. [19:58] <ubitux> :) [19:58] <ubitux> mutt can certainly such better in e-mails than bugzilla in bugs ;) [20:03] <ubitux> s/such/search/ [20:12] <iive> bugzilla is written in perl. [20:21] <saste> iive: don't blame the language for the implementation [20:21] <saste> anyway maybe it is intentionally so hard [20:22] <saste> so that the only users who manages to create a report are skilled or masochistic enough to perform a good report [20:22] <iive> you have to use the right tool for the right job. [20:42] <philipl> ubitux: Do you still need to look at the diff more, or are you happy? (I've fixed the INT_MAX thing locally) [20:49] <ubitux> philipl: what happens in case of a subtitles ` subrip and lace_duration>INT_MAX? [20:49] <ubitux> it looks like none of the duration is set [20:50] <ubitux> saste: is the cflags/ldlibs things fixing anything? [20:50] <philipl> ubitux: Yes, but that's the same as before. [20:50] <philipl> It deliberately doesn't set any duration for non-subrip subtitles. [20:51] <philipl> For ass that's because it inserts inband timing. [20:51] <philipl> For vobsub, I assume its because the inband timing was always there. [20:51] <saste> ubitux: yes, although it is somehow a corner case [20:51] <saste> suppose that you have libs in custom prefix, e.g. /opt/weird/ffmpeg [20:52] <ubitux> philipl: should be ok then [20:52] <philipl> Thanks. [20:52] <saste> *e.g. /weird/path/ [20:53] <saste> then you want to install ffmpeg in /freak/ffmpeg/ [20:53] <saste> when you link, you update your PKG_CONFIG_PATH, so that ffmpeg libs are looked in /freak/ffmpeg, but the cflags/ldflags will point to /weird/path/ [20:53] <saste> with funny results [20:54] <saste> a more practical example, I have libs installed in HOME, but I may want to install ffmpeg in /opt/ffmpeg, and test the examples against that install [20:55] <saste> so I need to override the default cflags/ldflags [20:55] <saste> note that I already patched the build for that much time ago [20:55] <saste> (many build systems don't support that scenario, e.g. libx264) [20:55] <ubitux> ok ok [20:56] <ubitux> well should be ok then [20:58] <saste> i'll push it soon then [21:50] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * re53becc70f 10ffmpeg/doc/examples/Makefile: (log message trimmed) [21:50] <CIA-56> ffmpeg: examples/Makefile: give priority to pkg-config flags [21:50] <CIA-56> ffmpeg: In case CFLAGS/LDLIBS are already defined and conflicting with the [21:50] <CIA-56> ffmpeg: pkg-config flags, give priority to the latter since they are used to [21:50] <CIA-56> ffmpeg: detect compilation flags. [21:50] <CIA-56> ffmpeg: This should fix for example the case where there are many different [21:50] <CIA-56> ffmpeg: instances of a library, CFLAGS=-I/foo/include and pkg-config cflags say [21:50] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * r078e203379 10ffmpeg/doc/examples/scaling_video.c: examples/scaling_video: add file doxy [21:50] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * rcb6defb053 10ffmpeg/doc/examples/scaling_video.c: examples/scaling_video: consistently use printf rather than av_log() [22:05] <ubitux> saste: please add demuxing to .gitignore :) [22:06] <ubitux> saste: maybe we could use av_log() instead of the printf utils in the example? [22:09] <ubitux> ah you just switched from av_log to fprintf heh. [22:21] <saste> yeah we should try to keep some ... consistency between the various examples [22:32] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r438b86fed4 10ffmpeg/libavcodec/faxcompr.c: [22:32] <CIA-56> ffmpeg: faxcompr: fix byte alignment case [22:32] <CIA-56> ffmpeg: Fixes Ticket1653 [22:32] <CIA-56> ffmpeg: Found-by: ami_stuff [22:32] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [22:43] <brimestone> hey guys, i have a piece of footage that is shot in LogC. I'm wondering i can use FFmpeg to add a LUT/Color Profile to get the "normal" looking image in Rec709? [22:46] <Compn> you want to switch the color profile ? [22:46] <Compn> i thought there was a filter to do it in mplayer, which ffmpeg should have now [22:46] <Compn> but i dont remember [22:46] <brimestone> i think that is what i'm talking about [22:47] <Compn> you color profile guys sound like audiophiles :P [22:47] <saste> matrixcolor? [22:47] <brimestone> yes... [22:47] <Compn> saste : http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/2008-August/058341.html [22:47] <Compn> saste : maybe it was never committed... [22:48] <Compn> saste : but theres been at least 100 people asking for this feature, it would make a nice thing if ffmpeg doesnt have it... [22:49] <brimestone> doesnt?? [22:52] <Compn> if you are encoding to x264 , x264 has a colormatrix option [22:52] <Compn> which can specify the thing [22:53] <Compn> dnxhd encoder might as well [22:54] <brimestone> hmmmm [22:55] <Compn> Convert HD YUV BT709 to SD BT601 and vice versa [22:55] <Compn> FFmpeg customized for broadcast and professional usage. [22:55] <Compn> What you can do with FFmbc: [22:55] <Compn> http://code.google.com/p/ffmbc/ [22:56] <Compn> that might be what you want brimestone [22:56] Action: Compn afk [22:56] <Compn> saste : merge ffmbc , gpl ifdef everything! :) [22:57] Action: saste saste wonders why people ask him to do everything [22:58] <Compn> mostly because everyone else avoids me [22:58] <Compn> and michael is too busy ;) [22:58] <ubitux> isn't vf colormatrix what you are looking for? [22:58] <Daemon404> that was ported from avs [22:58] <Daemon404> it isnt unique to ffmbc [22:59] <Compn> maybe it just needs some advertising how to use it [22:59] Action: Compn looks at saste to write docs [22:59] <ubitux> there is some doc for the colormatrix filter [23:00] <Compn> 28.8 colormatrix [23:00] <Compn> The colormatrix filter allows conversion between any of the following color space: BT.709 (bt709), BT.601 (bt601), SMPTE-240M (smpte240m) and FCC (fcc). [23:00] <Compn> The syntax of the parameters is source:destination: colormatrix=bt601:smpte240m [23:00] <Compn> so it is [23:00] <Compn> brimestone : how come you dont read documentation then huh? [23:00] <Compn> :P [23:06] <brimestone> i do? [23:06] <brimestone> i just get confuse with the nomanclature of the Video world... [23:14] <brimestone> that actually does convertion from bt709 to smpte? [23:17] <brimestone> what im trying to do is render a LogC footage from an Arri Alexa camera using a LUT file created by the Cinematographer on the field [23:28] <nyuhu> saste : in your review of the process_command callback for hue youre suggesting to save the old hue values and then reset them ; in itself it is not necessary to save the old values since theyre not directly used in the processing (hue_[sin,cos] attributes are used instead but are computed from the hue). Should we still save them for the consistency ? [23:40] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * r952caf11f2 10ffmpeg/doc/examples/muxing.c: examples/muxing: remove pointless #undef exit [23:40] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * rdb0a0aa243 10ffmpeg/doc/examples/scaling_video.c: examples/scaling_video: add logging in case of image allocation failure [23:43] <saste> nyuhu: whatever is simpler/easier to understand [23:46] <brimestone> anyone here have a clue on how to use -lutyuv? [23:52] <saste> brimestone: -vf lutyuv=... [23:53] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r19d8954966 10ffmpeg/libavformat/utils.c: [23:53] <CIA-56> ffmpeg: find_stream_info: update last_dts more often [23:53] <CIA-56> ffmpeg: This improves dts validity checks and consequently fps detection of files with invalid dts [23:53] <CIA-56> ffmpeg: Fixes Ticket1681 [23:53] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni@gmx.at> [00:00] --- Fri Aug 31 2012