[Ffmpeg-devel-irc] ffmpeg-devel.log.20150721

burek burek021 at gmail.com
Wed Jul 22 02:05:02 CEST 2015

[00:00:46 CEST] <rcombs> also, I use a simple linear search (and in one case a reverse one) instead of a binary search when looking through the fragment index, but I suppose that could change
[00:06:16 CEST] <baptiste> rcombs, the code should reuse as much as possible as the mfra code
[00:06:33 CEST] <baptiste> I see mfra has an option to specify either dts/pts
[00:07:21 CEST] <baptiste> I never understood why they added sidx instead of reusing mfra
[00:08:48 CEST] <baptiste> I believe you can execute mov_switch_root just before returning from seek successfully
[00:08:55 CEST] <baptiste> that should simplify the function a bit
[00:14:49 CEST] <rcombs> uhhhh
[00:15:59 CEST] <rcombs> I didn't understand either of the things you just said
[00:17:43 CEST] <baptiste> sorry :)
[01:00:10 CEST] <wm4> rcombs: is that search over all samples or just all fragments?
[01:00:46 CEST] <rcombs> wm4: just fragments
[01:00:57 CEST] <wm4> then why care
[01:01:31 CEST] <wm4> there shouldn't be that many fragments right?
[01:01:43 CEST] <wm4> although I have no idea about the magnitude
[01:02:09 CEST] <rcombs> I saw a few hundred in these youtube files
[01:02:23 CEST] <rcombs> 1 per second wouldn't be absurd
[01:02:41 CEST] <rcombs> so I care enough to point out that it might be worth optimizing, but not enough to actually do it
[01:02:43 CEST] <cone-809> ffmpeg 03Vittorio Giovara 07master:5d3addb93794: Add a quality factor packet side data
[01:02:44 CEST] <cone-809> ffmpeg 03Michael Niedermayer 07master:e5bae39f46e5: Merge commit '5d3addb937946eca5391e40b5e6308e74ac6f77b'
[01:02:45 CEST] <cone-809> ffmpeg 03Michael Niedermayer 07master:809780ca425a: avcodec/utils: Do not merge side data for encoders
[01:14:31 CEST] <ZeuZ> Guys, regarding the C API, where can I find an example of something akin to invoking ffmpeg like:  -f image2 -framerate 2 -i /CAMARAS/Camara1/camara1.%d.jpg -y /CAMARAS/Camara1/registro.`echo $clips`.avi  
[01:14:40 CEST] <ZeuZ>  I mean, can I input something akin to camara1.%d.jpg as a filter for each image to be taken? Or should I parse them in order and add them?
[01:27:55 CEST] <cone-809> ffmpeg 03Vittorio Giovara 07master:40cf1bbacc62: Deprecate avctx.coded_frame
[01:27:56 CEST] <cone-809> ffmpeg 03Michael Niedermayer 07master:495eee01238b: Merge commit '40cf1bbacc6220a0aa6bed5c331871d43f9ce370'
[01:47:51 CEST] <cone-809> ffmpeg 03Janne Grunau 07master:f56d8d8dd72b: h264: aarch64: intra prediction optimisations
[01:47:52 CEST] <cone-809> ffmpeg 03Michael Niedermayer 07master:5d5f8b29b4d4: Merge commit 'f56d8d8dd72b1ab52aa814c5a0fccabf8040ef68'
[01:59:44 CEST] <cone-809> ffmpeg 03Janne Grunau 07master:9ed6f9a17cc1: arm: use a local label instead of the function symbol in ff_prefetch_arm
[01:59:45 CEST] <cone-809> ffmpeg 03Michael Niedermayer 07master:eca84946eaa1: Merge commit '9ed6f9a17cc1f7d3699a1223783dadc1ee222069'
[02:01:43 CEST] <cone-809> ffmpeg 03James Almer 07master:651448a8a7bd: avcodec/aacsbr: fix compilation with hardcoded tables
[02:30:46 CEST] <KGB> [13FFV1] 15michaelni pushed 27 new commits to 06master: 02http://git.io/vmxkR
[02:30:46 CEST] <KGB> 13FFV1/06master 14bba8cca 15Ashley Blewer: convert ffv1.lyx to markdown
[02:30:46 CEST] <KGB> 13FFV1/06master 1420f8948 15ablwr: adjusts tables to comply with github-flavored markdown rendering
[02:30:46 CEST] <KGB> 13FFV1/06master 141986fd6 15ablwr: corrects broken tables and moves code block tables to headerless
[02:43:45 CEST] <jamrial> either github's markdown parser sucks, or the resulting ffv1.md file isn't good
[02:45:31 CEST] <rcombs> or both!
[02:45:42 CEST] <KGB> [13FFV1] 15michaelni pushed 2 new commits to 06master: 02http://git.io/vmxtR
[02:45:42 CEST] <KGB> 13FFV1/06master 140a8e9ae 15Michael Niedermayer: Add missing '*'...
[02:45:42 CEST] <KGB> 13FFV1/06master 1427f6cb5 15Michael Niedermayer: update slice count restriction to what is used in the implementation...
[02:45:44 CEST] <rcombs> sturgeon's law says it's both
[03:37:03 CEST] <KGB> [13FFV1] 15dericed opened pull request #11: fix markdown table rendering on github (06master...06markdown-fixes) 02http://git.io/vmxch
[03:50:12 CEST] <cone-809> ffmpeg 03Tom Butterworth 07master:d8488090e507: libavcodec/hap: Name enums
[03:50:14 CEST] <cone-809> ffmpeg 03Tom Butterworth 07master:440c26e9c46b: libavcodec/hap: remove unused struct member
[03:50:14 CEST] <cone-809> ffmpeg 03Tom Butterworth 07master:6b96c70f2d89: avcodec/hapdec: don't log texture format every frame, do it once per decode session
[03:52:14 CEST] <KGB> [13FFV1] 15michaelni closed pull request #11: fix markdown table rendering on github (06master...06markdown-fixes) 02http://git.io/vmxch
[04:45:29 CEST] <KGB> [13FFV1] 15dericed opened pull request #12: Cleaning (06master...06cleaning) 02http://git.io/vmxzN
[05:05:34 CEST] <KGB> [13FFV1] 15dericed opened pull request #13: move code from tables to codeblocks (06master...06tables-to-codeblocks) 02http://git.io/vmxwg
[06:41:27 CEST] <jamrial> commit bc3f19641ce37d14f9cfd7e2693c4833b63fba7d started the filter-channelmap memleaks
[06:41:49 CEST] <jamrial> someone that knows their way around lavfi's filtergraph stuff should check it
[06:47:34 CEST] <rcombs> wait, people know their way around lavfi's filtergraph stuff?
[06:56:45 CEST] <jamrial> haha
[06:58:26 CEST] <durandal_1707> what that commit changed?
[06:59:36 CEST] <rcombs> that can be simplified to "people know their way around software?"
[07:00:00 CEST] <rcombs> because sturgeon's law applies to both software and developers
[07:03:18 CEST] <jamrial> durandal_1707: commit message is "split creating and (re-)configuring complex filtergraphs"
[10:15:46 CEST] <__gb__> hi philipl, that's something that should be possible, yes
[10:18:12 CEST] <KGB> [13FFV1] 15michaelni closed pull request #12: Cleaning (06master...06cleaning) 02http://git.io/vmxzN
[10:41:17 CEST] <cone-626> ffmpeg 03Zhang Rui 07master:cee7acfcfc1b: MAINTAINERS: add myself as a maintainer for async protocol
[14:06:47 CEST] <cone-626> ffmpeg 03Rostislav Pehlivanov 07master:331c1e74941d: aacenc: move the generation of ff_aac_pow34sf_tab[]
[14:53:15 CEST] <D404|Ghetto> interesting
[14:53:18 CEST] <D404|Ghetto> IRT at VDD
[15:13:34 CEST] <cone-626> ffmpeg 03hSÇ 07master:21e2f030bfa1: avcodec: loongson move simple idct functions to a separate file
[15:13:35 CEST] <cone-626> ffmpeg 03hSÇ 07master:b50910e9a9cf: avcodec: loongson optimize xvid idct with mmi
[15:13:36 CEST] <cone-626> ffmpeg 03Tom Butterworth 07master:5d0616ce0e64: avcodec/hapdec: log reason for failure when texture type doesn't match stream
[15:17:14 CEST] <ZeuZ> Hello, I'm looking into programatically convert a series of images into an AVI compressed stream. Can somebody help me find the documentation that pertains to that?
[15:23:46 CEST] <ZeuZ> Took this as basis: https://ffmpeg.org/doxygen/trunk/api-example_8c-source.html -> yet there are some things I've got doubts with like line 250-255
[15:53:31 CEST] <thardin> carl isn't in here?
[15:53:43 CEST] <thardin> anyway, I replied to his patch
[16:02:54 CEST] <TimNich> ULs everywhere, ugh.
[16:03:15 CEST] <D404|Ghetto> UL?
[16:05:12 CEST] <KGB> [13FFV1] 15michaelni pushed 1 new commit to 06master: 02http://git.io/vmjuK
[16:05:12 CEST] <KGB> 13FFV1/06master 142217aaf 15Michael Niedermayer: Suggest a pagebreak before Genral description...
[16:06:12 CEST] <D404|Ghetto> ..wut?
[16:06:24 CEST] <D404|Ghetto> oh. ffv1.
[16:17:53 CEST] <cone-626> ffmpeg 03Zhang Rui 07master:aee909acfb34: fate: add test for async protocol
[16:17:54 CEST] <cone-626> ffmpeg 03Michael Niedermayer 07master:593731efa8ee: tests/checkasm/Makefile: Fix checkasm with SDL
[17:03:46 CEST] <cone-626> ffmpeg 03Luca Barbato 07master:219b39a71a56: parseutil: Use non ambiguous aliases for uhd
[17:03:47 CEST] <cone-626> ffmpeg 03Michael Niedermayer 07master:be0fd074579c: Merge commit '219b39a71a5694b1c14a07b86477f665a5b6849b'
[17:44:40 CEST] <durandal_1707> J_Darnley: how is asm going?
[17:49:56 CEST] <cone-626> ffmpeg 03Michael Niedermayer 07master:f2d7409c7256: avformat/async: Add missing else
[17:49:57 CEST] <cone-626> ffmpeg 03Michael Niedermayer 07master:39867f3e098a: avfilter/af_channelmap: Move ff_add_channel_layout() call to querry_format()
[17:57:14 CEST] <cone-626> ffmpeg 03hSÇ 07master:c303ad85a9e4: avcodec: loongson optimize blockdsp with mmi
[18:04:25 CEST] <cone-626> ffmpeg 03James Almer 07master:3e46c7dbbeb0: avcodec/options-test: don't alloc avctx->coded_frame
[18:04:26 CEST] <cone-626> ffmpeg 03James Almer 07master:d32547a24a3f: avcodec/utils: silence some deprecation warnings
[18:34:53 CEST] <durandal_1707> sidechaincompess is too long, awaiting shorter candidates
[18:40:02 CEST] <cone-626> ffmpeg 03Bernd Bleßmann 07master:99ad832c7ea9: libavfilter/vf_scale: implement process_command
[19:41:18 CEST] <durandal_1707> noone?
[19:44:21 CEST] <Compn> filter ?
[19:44:26 CEST] <Compn> or api ?
[19:44:32 CEST] <Compn> you just want a new name schema?
[19:45:43 CEST] <Compn> ZeuZ : https://trac.ffmpeg.org/wiki/Create%20a%20video%20slideshow%20from%20images is for the user way. you want to make your own program that does it ? confused on what exaclty you want.
[19:51:22 CEST] <durandal_1707> Compn: shorter name for filter...
[19:53:57 CEST] <Compn> whats it do?
[19:54:01 CEST] <Compn> then i can name it :P
[19:55:12 CEST] <Compn> +This filter acts like normal compressor but has the ability to filter detected
[19:55:12 CEST] <Compn> +signal before sending it to the compression stage.
[19:55:14 CEST] <Compn> humm
[19:56:28 CEST] <Compn> whats the second signal ? i mean is it audio background hiss noise ?
[20:02:10 CEST] <cone-626> ffmpeg 03Bernd Bleßmann 07master:d2b78fe6b71f: libavfilter/vf_crop: implement process_command
[20:03:40 CEST] <durandal_1707> Compn: anything what user pick, its meant to provide ducking together with amege and pan filter
[20:05:14 CEST] <durandal_1707> For example if you say something on 2nd input output of the filter will lower volume so what you saying can be heard
[20:35:54 CEST] <cone-626> ffmpeg 03Luca Barbato 07master:977105407cae: hap: Decode using optimal slices sizes
[20:35:55 CEST] <cone-626> ffmpeg 03Michael Niedermayer 07master:2de5737ee21d: Merge commit '977105407cae55876041dddbf4ce0934cdd4cd6c'
[20:37:17 CEST] <Compn> durandal_1707 : i have no idea what to call that then :P
[21:09:07 CEST] <cone-626> ffmpeg 03Luca Barbato 07master:6b2b26e7af3e: dds: Decode using optimal slices sizes
[21:09:08 CEST] <cone-626> ffmpeg 03Michael Niedermayer 07master:8fe00dd18a4a: Merge commit '6b2b26e7af3ede0abfb46eb5725c26d1083f50bc'
[21:28:33 CEST] <cone-626> ffmpeg 03Zhang Rui 07master:5e2098d9064e: avformat/async: fix interrupt_callback usage and return code
[21:28:33 CEST] <cone-626> ffmpeg 03Zhang Rui 07master:830d3a0ebe24: avformat/async: rename async_interrupt_callback to async_check_interrupt
[23:11:32 CEST] <cone-626> ffmpeg 03Vesselin Bontchev 07master:323ec6b5e214: Add support for TEA (Tiny Encryption Algorithm)
[00:00:00 CEST] --- Wed Jul 22 2015

More information about the Ffmpeg-devel-irc mailing list