[Ffmpeg-devel-irc] ffmpeg-devel.log.20150109

[burek]

[00:00] <compn> tell that to cygwin1.dll :P
[00:01] <compn> >also 20 other cygwin dlls
[00:07] <nevcairiel> this is why noone uses cygwin for shit anymore
[00:07] <aetasx> exactly
[00:08] <nevcairiel> and most people link the mingw extra-crt staticly
[00:08] <nevcairiel> although i've seen that not be the case, and thats ugly as hell <.<
[00:10] <aetasx> vlc ships with libgcc dll
[00:10] <aetasx> at least their nightlies do
[00:11] <j-b> releases don't
[00:11] <j-b> of course
[00:12] <aetasx> dunno, its been awhile since I've used one.  there was some feature or something I upgraded to 2.2 to get but cant recall what
[00:13] <aetasx> oh right, it couldnt play x265 within mp4s
[00:15] <aetasx> that was before I realized it was too damn slow to encode to justify using it
[00:15] <j-b> it got better, no?
[00:16] <aetasx> I cant recall but I remember doing medium encodes and it only going at something like under 60fps
[00:17] <aetasx> for video thats 320x240
[00:19] <aetasx> x264 could hit the same size, same quality much faster
[00:19] <nevcairiel> x265 is still under fast development, so any opinions formed half a year ago may already be quite invalid :d
[00:20] <aetasx> this was months ago
[00:20] <aetasx> but I still consider it in active development so its not like I can expect the same speed results
[00:20] <nevcairiel> half a year is "6 months", so it would qualify!
[00:21] <aetasx> would 2 months still qualify? :p
[00:21] <nevcairiel> dunno, possibly, i havent been checking on its progress that much over the last time
[00:22] <aetasx> x264 has been out for quite awhile so its not really unexpected that it would be much slower
[00:23] <aetasx> it may just end up being for more professionally done video than for the random stuff I need
[00:23] <jamrial> the bulk of optimizations on x265 has been fox avx2, so unless you have a haswell you're out of luck
[00:23] <aetasx> nah, nehalem
[00:50] <cone-173> ffmpeg.git 03Carl Eugen Hoyos 07master:cd3405282c70: Skip Exif data when auto-detecting jpeg images.
[04:04] <ac_slater_> sigh. Hey all. My libavformat shared library output is missing `ff_isom_write_hvcc`. Oh, ffmpeg 2.5
[04:08] <ac_slater_> I've tried enabling all of the HEVC features but no go. I'll paste a config.log
[04:12] <ac_slater_> config.log ... http://sprunge.us/FhIF
[04:26] <jamrial> as you were told last night that function is not related to either hevc decoder or hevc de/muxer. it's only needed for mov, matroska, webm and rtp muxers
[04:27] <ac_slater_> jamrial: I lost connection for a bit last night so I must have missed it
[04:28] <ac_slater_> jamrial: sorry mate. 
[04:28] <ac_slater_> jamrial: so I narrowed it down to one VERY specific thing. If you have `--enable-protocol=rtp` but not a muxer or demuxer for RTP, this will happen
[04:28] <ac_slater_> is that intended?
[04:29] <jamrial> look for CONFIG_RTP_MUXER in libavformat/Makefile and add hevc.o at the end of the list of objects it should compile
[04:30] <jamrial> unlike in git master, that's not in ffmpeg 2.5
[04:30] <jamrial> it's probably the source of your problem
[04:33] <jamrial> see if it builds after that. if it does then fccfc22d1 should be backported
[04:35] <ac_slater_> jamrial: awesome ill report back
[04:40] <ac_slater_> jamrial: hmm well that didnt do it. I'll find the difference with this. But for now (since Im in a rush) adding matroska mux/demux fixes the build for me
[04:45] <jamrial> eh, that commit should be backported anyway
[04:45] <jamrial> michaelni_: ok with the above?
[04:46] <michaelni_> jamrial, fccfc22d1, sure
[04:47] <jamrial> ok
[04:51] <cone-161> ffmpeg.git 03Martin Storsjö 07release/2.5:93cbdcd4d30c: libavformat: Build hevc.o when building the RTP muxer
[05:05] <cone-161> ffmpeg.git 03Peter Ross 07master:624384503643: avformat/aiffdec: improve readability
[05:05] <cone-161> ffmpeg.git 03Michael Niedermayer 07master:d1c6b7b2d1ef: configure: fix convoluted shlib code
[12:59] <cone-671> ffmpeg.git 03wm4 07master:b88e80589bd1: vp9: fix parser return values in error case
[14:09] <cone-671> ffmpeg.git 03Michael Niedermayer 07master:2959a612326c: avformat/utils: Use avio_closep() to avoid leaving stale pointers in memory
[14:09] <cone-671> ffmpeg.git 03Michael Niedermayer 07master:643dee846add: avformat/sapenc: Use avio_closep() to avoid leaving stale pointers in memory
[14:09] <cone-671> ffmpeg.git 03Michael Niedermayer 07master:ef3a6a48b56d: avformat/tee: Simplify code by using avio_closep()
[15:21] <anshul_mahe> is there any way in command line to disable bug of perticular codec
[15:24] <anshul_mahe> Its hard to find my codc(what I implementing) related debug message when other codecs overflow my terminal 
[15:28] <wm4> well, the log callback has a context
[15:28] <wm4> you can get the AVCodecContext from that I think
[15:29] <wm4> but you might miss messages, because av_log is often called with no context
[15:29] <wm4> or with a different context, even if it should be related to the codec you're debugging
[15:31] <compn> anshul_mahe : you can force your codec ? -vcodec codecname
[15:31] <compn> or not sure what you want
[16:07] <anshul__> wm4: I am greping my codec context name, but this thing is taking long time, I was looking for some better trick
[16:09] <anshul__> its not that I am missing them on my screen, I redirect and take them in file, but this long procedure is slowing me down
[16:55] <cone-671> ffmpeg.git 03Michael Niedermayer 07master:323355413dbe: tools/uncoded_frame: Use avio_closep() to avoid leaving stale pointers in memory
[16:55] <cone-671> ffmpeg.git 03Michael Niedermayer 07master:04091552616e: ffmpeg: remove unneeded NULL pointer check
[16:55] <cone-671> ffmpeg.git 03Michael Niedermayer 07master:b1d223301f03: avformat/hdsenc: Remove redundant NULL pointer checks
[17:18] <cone-671> ffmpeg.git 03Anton Khirnov 07release/2.4:883795fb0fac: h264: restore a block mistakenly removed in e10fd08a
[17:18] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.4:33c8bb2c6a87: Merge commit '883795fb0face00a07349fbd32f4775431a9f30d' into release/2.4
[17:32] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:bcd7f35717db: avformat/segment: Use av_freep() avoid leaving stale pointers in memory
[17:32] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:68c63470895d: avformat/cdxl: Fix integer overflow of image_size
[17:32] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:eac0451e4738: avformat/flvdec: do not inject dts=0 metadata packets which failed to be parsed into a new data stream
[17:32] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:5bb31e856d7c: avformat/flvdec: Increase string array size
[17:32] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:252ba4a9259f: avfilter/vf_sab: fix filtering tiny images
[17:32] <cone-671> ffmpeg.git 03wm4 07release/2.5:e0a12b3dc3a2: avcodec/dvdsubdec: fix out of bounds accesses
[17:32] <cone-671> ffmpeg.git 03Dale Curtis 07release/2.5:022bfd3dd47c: mov: Avoid overflow with mov_metadata_raw()
[17:32] <cone-671> ffmpeg.git 03Dale Curtis 07release/2.5:9143ab0e5a75: mov: Fix overflow and error handling in read_tfra().
[17:33] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:25312a427bda: avformat/mov: fix integer overflow in mov_read_udta_string()
[17:33] <cone-671> ffmpeg.git 03Dale Curtis 07release/2.5:0ce35b8ce8b9: mov: Fix negative size calculation in mov_read_default().
[17:33] <cone-671> ffmpeg.git 03Paul B Mahol 07release/2.5:c7a2ac6b6bd5: lavfi: check av_strdup() return value
[17:33] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:0d3a07852c3c: avformat/mov: use av_freep() to avoid leaving stale pointers in memory
[17:33] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:2791eba1d78a: avformat/matroskadec: Use av_freep() to avoid leaving stale pointers in memory
[17:33] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:918ed73b70a9: avformat/utils: Clear pointer in ff_alloc_extradata() to avoid leaving a stale pointer in memory
[17:33] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:b3c082412cda: cmdutils: Use 64bit for file size/offset related variable in cmdutils_read_file()
[17:33] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:d0041dc8c4d7: avformat/mov: Fix mixed declaration and statement warning
[17:33] <cone-671> ffmpeg.git 03James Almer 07release/2.5:b834dc14da08: configure: bump year
[17:33] <cone-671> ffmpeg.git 03wm4 07release/2.5:3d71024f8ab3: avcodec/dvdsubdec: error on bitmaps with size 0
[17:33] <cone-671> ffmpeg.git 03wm4 07release/2.5:b895e29941f0: avcodec/dvdsubdec: fix accessing dangling pointers
[17:33] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:f9bbc26e691f: ffmpeg: Clear error message array at init.
[17:33] <cone-671> ffmpeg.git 03wm4 07release/2.5:3032291b3a88: vp9: fix parser return values in error case
[17:35] <cone-671> ffmpeg.git 03Arwa Arif 07master:a299cd5ab356: lavfi: port mp=pp7 to libavfilter
[18:19] <cone-671> ffmpeg.git 03Michael Niedermayer 07release/2.5:a7dd37169c9a: Update for 2.5.3
[20:46] <cone-671> ffmpeg.git 03Michael Niedermayer 07master:727cb2bda9df: avcodec/ass: Use av_realloc_array()
[20:46] <cone-671> ffmpeg.git 03Michael Niedermayer 07master:aefb99f2c055: avcodec/ass_split: use *alloc_array()
[21:25] <cone-671> ffmpeg.git 03Anton Khirnov 07master:7e4e010b80e7: mpeg4audio: check the init_get_bits() return value
[21:25] <cone-671> ffmpeg.git 03Michael Niedermayer 07master:1cf0f2798596: Merge commit '7e4e010b80e76862e83afbd41c25d50e72f0b44c'
[21:36] <cone-671> ffmpeg.git 03Anton Khirnov 07master:51da7d02748c: matroskaenc: refuse to write AAC without valid extradata
[21:36] <cone-671> ffmpeg.git 03Michael Niedermayer 07master:f95cd5a235ba: Merge commit '51da7d02748cc54b7d009115e76efa940b99a8ef'
[22:52] <compn> whoa
[22:52] <compn> intel and amd both contributing to ffmpeg at the same time :D
[22:53] <compn> plus mips before. how many other cpu manf have submitted patches? :)
[00:00] --- Sat Jan 10 2015