[Ffmpeg-devel-irc] ffmpeg-devel.log.20150301

[burek]

[00:04] <cone-445> ffmpeg 03Anton Khirnov 07master:d74a8cb7e42f: fmtconvert: drop unused functions
[00:04] <cone-445> ffmpeg 03Michael Niedermayer 07master:5c17377e28e8: Merge commit 'd74a8cb7e42f703be5796eeb485f06af710ae8ca'
[00:15] <cone-445> ffmpeg 03Anton Khirnov 07master:71f1ad37d858: lavc: do not compile fmtconvert unconditionally
[00:15] <cone-445> ffmpeg 03Michael Niedermayer 07master:7fce8c752d87: Merge commit '71f1ad37d858b810b71a4af1c25771beaa50b27b'
[00:26] <cone-445> ffmpeg 03Christian Hujer 07master:802987f8c703: x11grab: Unbreak building
[00:26] <cone-445> ffmpeg 03Andreas Cadhalpun 07master:ea1d0b7ece18: avcodec/utils: use correct printf specifier in ff_set_sar
[00:26] <cone-445> ffmpeg 03Michael Niedermayer 07master:8a776726457a: Merge commit '802987f8c7033ec8b82b35438d3822cf7f761166'
[00:26] <cone-445> ffmpeg 03Michael Niedermayer 07master:2ab3088dda31: Merge commit 'ea1d0b7ece1881c2f1360f8c1e2116791105af21'
[00:46] <cone-445> ffmpeg 03Anton Khirnov 07master:990e4a6639d0: Add a QSV decoding example.
[00:46] <cone-445> ffmpeg 03Michael Niedermayer 07master:e73b3dba3e80: Merge commit '990e4a6639d0714198583c2812b1074b5751898f'
[01:01] <cone-445> ffmpeg 03Martin Storsjö 07master:985639534425: rtpenc_aac: Fix sending fragmented frames
[01:01] <cone-445> ffmpeg 03Martin Storsjö 07master:12b3459979f5: rtpenc_amr: Use s->num_frames instead of s->buf_ptr - s->buf
[01:01] <cone-445> ffmpeg 03Michael Niedermayer 07master:f6f1e1a7a14d: Merge commit '98563953442560dd83aab938f86de3e5a22a891f'
[01:01] <cone-445> ffmpeg 03Michael Niedermayer 07master:6f0a82966d82: Merge commit '12b3459979f5ea6481660cd2c99a0381e2b5ba37'
[01:08] <cone-445> ffmpeg 03Martin Storsjö 07master:0662440b9913: rtpenc_aac: Set a default value for max_frames_per_packet at init
[01:08] <cone-445> ffmpeg 03Martin Storsjö 07master:1fc64e2e0778: rtpenc: Write conditional statements on separate lines
[01:08] <cone-445> ffmpeg 03Michael Niedermayer 07master:ba069e08bdaa: Merge commit '0662440b991361fdb5e732712d997a73e4692e34'
[01:08] <cone-445> ffmpeg 03Michael Niedermayer 07master:cfe33c98b74b: Merge commit '1fc64e2e07787bbca82a72c146588e850e6d098a'
[01:17] <cone-445> ffmpeg 03Martin Storsjö 07master:9c9b0218e85f: rtpenc_aac: Merge a definition with a declaration
[01:17] <cone-445> ffmpeg 03Martin Storsjö 07master:d16c8d28d4e2: rtpenc_aac: Use AV_WB16 instead of manual bitshifts
[01:17] <cone-445> ffmpeg 03Michael Niedermayer 07master:635403e078a0: Merge commit '9c9b0218e85fcd969308632f75af48a4ce229541'
[01:17] <cone-445> ffmpeg 03Michael Niedermayer 07master:41b8d154aa54: Merge commit 'd16c8d28d4e2fca3af1054ffbf635c8cee755fc8'
[01:23] <cone-445> ffmpeg 03Martin Storsjö 07master:7c1e2e646674: rtpenc_xiph: Use AV_WB16 instead of manual bitshifts
[01:23] <cone-445> ffmpeg 03Martin Storsjö 07master:11edeaea3293: rtpenc_xiph: Don't exclude headers from max_payload_size
[01:23] <cone-445> ffmpeg 03Michael Niedermayer 07master:7e5300958885: Merge commit '7c1e2e64667421f931ab48141517f19d309c7eea'
[01:23] <cone-445> ffmpeg 03Michael Niedermayer 07master:9e09931fb6a4: Merge commit '11edeaea3293c41ecf577a330422eabba35f76a2'
[01:31] <cone-445> ffmpeg 03Martin Storsjö 07master:f8c01257f93c: rtpenc: Always do the default initialization regardless of codecs
[01:31] <cone-445> ffmpeg 03Martin Storsjö 07master:d4c7fc02f9f5: rtpenc: Skip redundant initialization
[01:31] <cone-445> ffmpeg 03Michael Niedermayer 07master:a2c4c30b5f61: Merge commit 'f8c01257f93ceda3e03bc4e540a51022d1e2bff2'
[01:31] <cone-445> ffmpeg 03Michael Niedermayer 07master:769b473a4186: Merge commit 'd4c7fc02f9f59e721e76debf4a595df529707545'
[01:42] <cone-445> ffmpeg 03Martin Storsjö 07master:bde2bba45c2f: rtpenc: Restructure if statements in packetizers to simplify adding more conditions
[01:42] <cone-445> ffmpeg 03Martin Storsjö 07master:4f6cd883f06f: rtpenc: Don't set max_frames_per_packet based on the packet frame size or frame rate
[01:42] <cone-445> ffmpeg 03Michael Niedermayer 07master:78c59f3f0c9f: Merge commit 'bde2bba45c2f2df27a8534028bda09a6e7f835e2'
[01:42] <cone-445> ffmpeg 03Michael Niedermayer 07master:5cb6b53050e3: Merge commit '4f6cd883f06f7893a2b60a41e7a4f8ae633dac2f'
[03:13] <cone-445> ffmpeg 03Gilles Chanteperdrix 07master:dcf19008a67f: avfilter/af_volume: fix precision=fixed and volume=0 case
[03:21] <cone-445> ffmpeg 03Michael Niedermayer 07master:9564375deeaf: avutil/ppc/intreadwrite: remove AV_??64 overriding code
[04:12] <cone-445> ffmpeg 03Gilles Chanteperdrix 07master:c95c039ec388: avformat: map T.140 RTP codec to text
[04:12] <cone-445> ffmpeg 03Gilles Chanteperdrix 07master:071f3e1c6e94: Revert "rtsp: Interpret the text media type as AVMEDIA_TYPE_DATA"
[04:12] <cone-445> ffmpeg 03Michael Niedermayer 07master:2b4041661a41: avcodec/mips/Makefile: Build fmtconvert_mips only if CONFIG_FMTCONVERT is set
[15:56] <alimj1> Hello. Good morning.
[15:57] <alimj1> Could you please inform me, what is the current preferred method to use FFMPEG or LibAV in Python?
[15:57] <alimj> Will I be on right track to follow this? http://mikeboers.github.io/PyAV/
[16:02] <alimj> Or maybe a subprocess?
[16:18] <wm4> alimj: the example on the page you linked is already wrong
[16:18] <wm4> it assumes you get exactly one frame per packet, which in this form is wrong
[16:19] <wm4> for example, you typically need to feed the decoder packets for a while at init without getting a frame back
[16:19] <wm4> likewise, you get some frames at the end of the stream without having to feed it packets
[16:27] <alimj> wm4: Thanks for the explanation. I would like to generate a video from sequenced PNGs in Python
[16:28] <alimj> Maybe it would be easier to use a subprocess and simply call avconv externally
[16:28] <wm4> my guess is that calling ffmpeg via subprocess would be the least painful
[16:29] <alimj> So, I will go with the least painful option now. Thanks for the help.
[17:00] <cone-100> ffmpeg 03Michael Niedermayer 07master:e72f8ceabde3: avformat/mpegenc: Ignore max_delay if no other options remain
[17:28] <cone-100> ffmpeg 03Michael Niedermayer 07master:c1e93cdc9efd: avutil/opt: Also test/compare the av_log output i the selftest
[17:38] <cone-100> ffmpeg 03Michael Niedermayer 07master:e9c85ae648b9: avutil/opt: Remove odd trailing whitespace bytes from flags help description
[17:50] <rmklp> is there any tutorial on the AVOption system anyone is aware of? Id like to understand how to add the capability for AVOptions to muxers, encoders etc. and was hoping there was maybe something that I could read in addition to doxy and code examples.  
[17:52] <cone-100> ffmpeg 03Michael Niedermayer 07master:88ec5435bba5: avutil/opt: also test av_opt_show2()
[19:24] <ubitux> thanks michaelni!
[19:25] <ubitux> (#4335)
[21:10] <ubitux> http://fate.ffmpeg.org/report.cgi?time=20150301192738&slot=x86_64-linux-gnu-icc-2011_sp1.13.367
[21:10] <ubitux> is this a bug in icc?
[21:10] <ubitux> other versions don't seem affected
[21:10] <Daemon404> 2011...
[21:11] <ubitux> yeah :)
[21:11] <nevcairiel> icc is generally buggy
[21:11] <nevcairiel> but people keep bothering to workaround its problems anyway
[21:11] <ubitux> btw, shouldn't we have the hevc threading issues fixed before the release?
[21:12] <ubitux> http://fate.ffmpeg.org/report.cgi?time=20150228081657&slot=x86_64-archlinux-gcc-threads
[21:13] <ubitux> and concerning http://fate.ffmpeg.org/report.cgi?time=20150228210155&slot=x86_64-archlinux-gcc-asan i see a lot of "too large alloc"; i wonder if we should do something about it
[21:15] <ubitux> http://fate.ffmpeg.org/report.cgi?time=20150301192220&slot=x86_32-debian-open64-5.0 hehe, this is interesting too...
[21:15] <ubitux> i wonder if i'm doing something really crazy...
[21:17] <ubitux> http://fate.ffmpeg.org/report.cgi?time=20150301201326&slot=powerpc-linux-gnu-gcc-4.3.5 mmh
[21:17] <BtbN> Anyone familiar with the dash muxer? I'm having trouble with it generating the codec string for aac/mp4a when copying an existing aac stream. It puts "mp4a.40" in the codec field, which doesn't seem to be correct. When i re-encode the aac audio, it puts "mp4a.40.2" there, which i'd also expect for the copied stream.
[21:17] <ubitux> all kind of interesting failures
[21:18] <BtbN> The input stream seems to need the aac_adtstoasc bitstream filter. Looking at the dashenc source, it extracts that final .2 from the extradata field, which should be generated by the bsf.
[21:36] <jamrial> http://fate.ffmpeg.org/report.cgi?time=20150301192450&slot=x86_64-msvc12-windows-native same with every msvc slot
[21:38] <jamrial> also http://fate.ffmpeg.org/report.cgi?time=20150301174055&slot=x86-opensolaris-gcc3.4
[22:24] <jamrial> ubitux: that powerpc slot seems to randomly fail a test every other run. hevc, pixfmt, fic-avi, mpeg4, mxpeg, utvideoenc...
[22:25] <ubitux> threading?
[22:27] <jamrial> fate by default uses 1 thread, and there's no mention about using threading in the slot description, so probably not
[23:31] <iive> memory?
[23:32] <ubitux> alignment maybe
[23:35] <iive> arm should also have alignment issues, shouldn't it?
[23:38] <ubitux> maybe it's related to ppc
[23:39] <ubitux> +optimizations
[23:47] <nevcairiel> arm is evil alignment country
[00:00] --- Mon Mar  2 2015