[Ffmpeg-devel-irc] ffmpeg-devel.log.20121201

burek burek021 at gmail.com
Sun Dec 2 02:05:02 CET 2012


[00:04] <ubitux> haha :D
[00:04] <ubitux> yeah
[00:41] <cone-663> ffmpeg.git 03Michael Niedermayer 07a7ee6281f7ef: qdm2: check array index before use, fix out of array accesses
[00:41] <cone-663> ffmpeg.git 03Michael Niedermayer 07b90e795f737c: check std tag size before reading.
[01:00] <ubitux> i begin to have quite a huge pile of patches ready to go
[01:01] <ubitux> i guess i'll start pushing sunday
[04:09] <cone-440> ffmpeg.git 03Michael Niedermayer 07f0695b09dd47: pcmdec: check codec_id
[04:09] <cone-440> ffmpeg.git 03Michael Niedermayer 07822d10e7c52e: swscale: fix 12 and 14 bit YUV planar output with scaling.
[07:31] <pross-au> cd
[07:32] <pross-au> aaaaaah
[07:43] <pross-au> If anyone is awake.. Who or how can i request files be moved from /incoming into the public facing /samples directory?
[07:57] <Compn> pross-au : request here or on list
[07:57] <Compn> carl or i do most of the moving
[07:58] <Compn> michael and reimar both have access as well
[07:58] <Compn> but usually i'm not doing much , feel free to ask me irc or email
[07:58] Action: Compn sleeps now tho
[07:58] <Compn> night
[08:44] <burek> ERROR: libopencore_amrnb not found
[08:44] <burek> libopencore-amrnb.a
[08:44] <burek> note the underscore.. :)
[08:45] <burek> wait, ignore me
[08:45] <nevcairiel> works for me
[08:45] <burek> :D
[08:56] <burek> it works, i made a mistake and built 64bit lib instead of 32bit :)
[11:28] <cone-504> ffmpeg.git 03Paul B Mahol 07b912895d94fd: mvi: check av_mallocz() result
[11:28] <cone-504> ffmpeg.git 03Paul B Mahol 073b56ed3478f9: motionpixels: use meaningful error code
[11:28] <cone-504> ffmpeg.git 03Paul B Mahol 0724e5cb4836e4: xldec: return meaningful error code
[11:28] <cone-504> ffmpeg.git 03Paul B Mahol 0751bcf8ac8542: xldec: K&R formatting cosmetics
[11:28] <cone-504> ffmpeg.git 03Paul B Mahol 07679c7578a667: wnv1: return more meaningful error codes
[11:28] <cone-504> ffmpeg.git 03Paul B Mahol 07e4af8ed8c7e3: ws-snd1: return meaningful error code
[11:28] <cone-504> ffmpeg.git 03Paul B Mahol 076b11ae7112e4: shorten: return more meaningful error codes
[11:53] <burek> can I request an update of http://www.ffmpeg.org/download.html
[11:53] <burek> to mark one static build as build for 3.x+ kernels and the other for 2.6.x+ kernels
[11:53] <burek> people think it's the same thing (a mirror)
[11:57] <cone-504> ffmpeg.git 03Paul B Mahol 072152b60f39d5: gifdec: use av_fast_malloc()
[12:20] <nevcairiel> burek: the second one is also only 64-bit
[12:24] <burek> that should be mentioned too
[12:42] <michaelni> burek, if something on the webpage isnt correct, send a patch, pull request or similar
[12:42] <durandal_1707> burek: don't ask to ask
[12:43] <burek> but but..
[12:43] <burek> it's just 20 letters :)
[12:46] <cone-504> ffmpeg.git 03Paul B Mahol 079d5cfce85540: gifdec: rudimentary seeking
[13:00] <cone-504> ffmpeg.git 03Michael Niedermayer 0713d351bd45d2: tests/ref/seek/lavf_gif: update for 9d5cfce85540: gifdec: rudimentary seeking
[13:06] <durandal_1707> why seeking test rerun synth one each time, and even if I'm only interested in single demuxer ...
[13:07] <durandal_1707> ... it creates files for all of them
[13:09] <cone-504> ffmpeg.git 03Stefano Sabatini 0712cd992d67d9: doc/ffmpeg-filters: add short description
[13:11] <saste> keeping our private copy of texi2pod is silly
[13:11] <saste> should I try to convince gcc guys to create a separate project for it?
[13:13] <saste> another option would be to switch to a more sensible toolchain, e.g. pandoc
[13:18] <cone-504> ffmpeg.git 03Martin Storsjö 07b70abd5b6143: avutil: Include io.h with a separate condition from windows console functions
[13:18] <cone-504> ffmpeg.git 03Martin Storsjö 077e6a11bcf717: avutil: Use a configure check to enable windows console functions
[13:18] <cone-504> ffmpeg.git 03Martin Storsjö 073bbe63d1a458: configure: Use headers in the check for _beginthreadex for w32threads
[13:18] <cone-504> ffmpeg.git 03Michael Niedermayer 075473f6258c63: Merge remote-tracking branch 'qatar/master'
[16:06] <cone-504> ffmpeg.git 03Stefano Sabatini 0713c39e967677: lavfi/geq: add T variable and example
[16:38] <wm4> has anyone made measurements yet whether the removal of the slice path in libavfilter made things slower?
[17:16] <cone-504> ffmpeg.git 03Michael Niedermayer 07b5cedf8b66c7: fate/mp3: increase fuzz for NEON, due to different rounding from lrintf()
[17:18] <wm4> does avio have something that identifies a stream as network/slow? or must I use AVIO_SEEKABLE_NORMAL to guess
[17:56] <burek> can anyone say something meaningful about this question: http://ffmpeg.gusari.org/viewtopic.php?f=16&t=748
[17:56] <burek> i generally understand what he asked, but have no idea what to answer
[18:01] <juanmabc> isn't the jerky usually not a high bitrate setting? the other would be a 24 fps to ~28 frames without padding but pitching the frames & audio
[18:01] <juanmabc> i.e. a little fast forward encoded
[18:04] <burek> I've bought some more time asking for output log
[18:05] <burek> i'm not sure, but it seems he just needs -r as an output option, right?
[18:08] <juanmabc> if it does padding a setpts=0.5*PTS seems to do fast motion
[18:36] <Compn> burek : ask him to look at the frames and see if there is a duplicate frame ever 4 frames or so
[18:36] <Compn> i forgot what needs to be done but its either ivtc or pullup/softskip from mplayer/mencoder
[18:36] <juanmabc> also i would not wonder if its player can't handle the stream issuing jerkiness like choppy playback
[18:36] <Compn> yeah
[18:37] <Compn> ask him also if the stream moves backwards , like a frame 3 frames behind pops up
[18:37] <Compn> out of order frame problem, disable b-frames
[18:38] <Compn> the frameorder is something like 12345267893
[18:38] <Compn> so previous frames stuck in there :\
[18:49] <ubitux> saste: won't pandoc require to change the markup of all the docs?
[18:49] <saste> ubitux: no
[18:49] <saste> it is just a generic converter
[18:49] <saste> think at it like the ffmpeg of docs ;-)
[18:49] <ubitux> !
[18:50] <ubitux> then it's certainly awesome
[18:50] <ubitux> mmh it's not packaged on my distro
[18:51] <ubitux> :(
[18:51] <ubitux> shouldn't be hard to add though
[18:51] <saste> the problem is that it currently *doesn't* support texi
[18:58] <ubitux> patch welcome!
[19:11] <kierank> ubitux: did your r128 filter pass the testsuite?
[19:11] <kierank> the ebu testsuite that is
[19:11] <kierank> also you should submit ffmpeg to http://tech.ebu.ch/loudness/loudness-faq-general
[19:14] <ubitux> i've tested it against the samples, and it seemed to work, i didn't check very carefully thought, might be worth having a closer look
[19:18] <ubitux> yeah i thought about submitting it, but i forgot and was too lazy
[19:34] <cone-504> ffmpeg.git 03Stefano Sabatini 07c9ff32215b43: lavu/opt: allow to set sample and pixel format with av_opt_set_int()
[20:13] <wm4> is the "http_proxy" environment variable the only way to make libavformat use a proxy?
[20:15] <wm4> ah, there's httpproxy too
[20:21] <cone-504> ffmpeg.git 03Michael Niedermayer 077389bb12e6b3: svq1dec: update w/h only if the header is successfully parsed.
[20:21] <cone-504> ffmpeg.git 03Michael Niedermayer 07873049e6d854: mxfdec: check index_tables before dereferencing in close.
[20:44] <divVerent> ubitux: can you explain this: https://gist.github.com/4184471
[20:49] <divVerent> ubitux: was tracking down encoding bitrate differences between mpv and ffmpeg
[20:49] <divVerent> and it turned out to be different sws flags
[20:49] <divVerent> and while tracking it down, also noticed that a lavfilter "scale" invoked by -vf gets different default flags from a "scale" invoked by -filter_complex
[20:49] <divVerent> does this even make any sense? how can this even happen at all?
[21:17] <cone-504> ffmpeg.git 03Carl Eugen Hoyos 075442083b1c54: Support Sorenson Spark in f4v files streamed by Flash Media Server.
[21:58] <ubitux> divVerent: i saw various patches for setting the sws flags in the filtergraph
[21:58] <ubitux> i think it's a bit tricky to transmit them sometimes
[21:59] <ubitux> saste or michaelni are more likely to be able to answer that&
[22:06] <cone-504> ffmpeg.git 03Paul B Mahol 0756e149fda991: PVF demuxer
[22:20] <wm4> hm, this error threw me into an endless loop when seeking in a network stream: [http @ 0x8a06c60]HTTP error 416 Requested range not satisfiable
[22:21] <wm4> can't seem to reproduce it
[22:31] <cone-504> ffmpeg.git 03Michael Niedermayer 073b2cd83a829e: dcadec: check lfe field
[22:31] <cone-504> ffmpeg.git 03Michael Niedermayer 07aae478036223: vmnc: Check for integer overflow
[23:18] <wm4> ubitux: so what's your reason again not to rename the mplayer functions in libavfilter to make libavfilter usable from mplayer?
[23:20] <michaelni> wm4, a patch renaming them is welcome
[23:32] <Compn> wm4 : are you saying --enable-lavfi is broken ?
[23:33] Action: Compn hasnt tried
[23:34] <wm4> Compn: mplayer-svn handles this differently: it messes directly with the ffmpeg build system and forces --disable-filter=mp
[23:35] <wm4> is there any specific prefix for internal symbols?
[23:41] <Compn> oh
[23:41] <Compn> ff_
[23:41] <Compn> for ffmpeg
[23:42] <Compn> could use av_ too for libav :P
[23:44] <wm4> but didn't ff_ used to be for public symbols too
[23:44] <Compn> oh, maybe i'm just making shit up :D
[23:44] <Compn> better ask someone smart
[23:47] <michaelni> ff_ is fine
[00:00] --- Sun Dec  2 2012


More information about the Ffmpeg-devel-irc mailing list