[Ffmpeg-devel-irc] ffmpeg-devel.log.20120615

[burek]

[00:09] <saste> ubitux, llogan: what happened with the fancy filtering wiki entry?
[00:09] <durandal_1707> michaelni: png_parser.c and png.c both define same PNG and MNG signatures
[00:10] <ubitux> saste: what do you mean? i don't know what to add :p
[00:11] <saste> ubitux: we had some discussion on #ffmpeg when people proposed various psychedelic effects and i suggested to create an entry in the wiki
[00:11] <ubitux> yes sure the page exists: https://ffmpeg.org/trac/ffmpeg/wiki/FancyFilteringExamples
[00:11] <saste> i thought that was already done but then i couldn't find them
[00:12] <ubitux> there were no other contribs since then afaict
[00:12] <saste> but there is no link from the top level?
[00:12] <ubitux> mmh from the filtering example page maybe
[00:12] <michaelni> durandal_1707, one of the 2 should be removed
[00:13] <ubitux> saste: ah indeed maybe not
[00:13] <ubitux> saste: maybe a reference could be added on https://ffmpeg.org/trac/ffmpeg/wiki/FilteringGuide
[00:13] <llogan> saste: yeah, i failed to unorphan it.
[00:13] <llogan> i'll do that
[00:14] <ubitux> saste: or the front page :)
[00:14] <saste> i have some fancy mirroring graph here :-)
[00:14] <llogan> nice
[00:14] <durandal_1707> michaelni: bikeshed, there is no way to not duplicate signatures inside generated binary code?
[00:15] <ubitux> saste: great :)
[00:15] <ubitux> saste: btw i was thinking of some video filter
[00:15] <llogan> i need to make a libx264 guide too for a lazier way to explain
[00:15] <ubitux> i don't know if i already talked about, but some kind of graphstats filter could be nice
[00:16] <ubitux> so you would be able to graph in real time for instead SAD
[00:16] <ubitux> or various other stuff like that
[00:16] <ubitux> (that makes me think i need to push the ffplay example)
[00:17] <michaelni> durandal_1707, one could replace all "immedeate" in the asm by a pointers that way the literal value would not be duplicated but then there would be duplicated references of equal size ...
[00:18] <ubitux> saste: so, something like: ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]graphstats=sad*3.5,40[filt]; [src][filt]overlay=w'
[00:18] <ubitux> so we will see the video, and next to it SAD would be graphed (and increased in this case), and 40 would be used as a simple threshold line or something
[00:19] <CIA-119> ffmpeg: 03Anton Khirnov 07master * r9034b0ed66 10ffmpeg/avconv.c: avconv: don't try to free threads that were not initialized.
[00:19] <CIA-119> ffmpeg: 03Martin Storsjö 07master * rd1beee0701 10ffmpeg/libavformat/rtpdec_h263_rfc2190.c: 
[00:19] <CIA-119> ffmpeg: rtpdec: Don't require frames to start with a Mode A packet
[00:19] <CIA-119> ffmpeg: While there is no reason for starting a frame with anything else
[00:19] <CIA-119> ffmpeg: than a Mode A packet, some senders seem to consistently use Mode B
[00:19] <CIA-119> ffmpeg: packets for everything. This fixes depacketization of such streams.
[00:19] <CIA-119> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[00:19] <CIA-119> ffmpeg: 03Samuel Pitoiset 07master * r7dc747f50b 10ffmpeg/libavformat/ (rtmppkt.c rtmppkt.h rtmpproto.c): (log message trimmed)
[00:19] <CIA-119> ffmpeg: rtmp: Read and handle incoming packets while writing data
[00:19] <CIA-119> ffmpeg: This makes sure all incoming packets are read and handled (and reacted
[00:19] <CIA-119> ffmpeg: to) while sending an FLV stream over RTMP to a server. If there were
[00:19] <CIA-119> ffmpeg: enough incoming data to fill the TCP buffers, this could potentially
[00:19] <CIA-119> ffmpeg: make things block at unexpected places. For the upcoming RTMPT support,
[00:19] <CIA-119> ffmpeg: we need to consume all incoming data before we can send the next
[00:19] <CIA-119> ffmpeg: 03Diego Biurrun 07master * rd246c18ea6 10ffmpeg/ (avplay.c libavcodec/pthread.c libavdevice/dv1394.h): 
[00:19] <ubitux> oups, maybe not the ',' as sep, ':' will be better :p
[00:19] <ubitux> so graphstats=sad*3.5:40
[00:20] <CIA-119> ffmpeg: 03Luca Barbato 07master * rd2d193c9b6 10ffmpeg/doc/fate.texi: doc: document THREAD_TYPE fate variable
[00:31] <llogan> FilteringGuide has a link to https://github.com/indiefan/FFmpeg-Filters "an updated vf_fade filter"
[00:31] <llogan> anyone know what is "updated" about it? (no, i didn't really look. yes, lazyass.)
[00:32] <durandal_1707> 2 years ago?
[01:45] <CIA-119> ffmpeg: 03Dale Curtis 07master * r71529bd8c5 10ffmpeg/libavformat/matroskadec.c: 
[01:45] <CIA-119> ffmpeg: Fix incorrect unsigned->signed conversion.
[01:45] <CIA-119> ffmpeg: Signed-off-by: Dale Curtis <dalecurtis at chromium.org>
[01:45] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:45] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r88a740afde 10ffmpeg/libavformat/matroskadec.c: 
[01:45] <CIA-119> ffmpeg: matroskadec: change assert to av_assert0()
[01:45] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:45] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r08169fc3d2 10ffmpeg/libavformat/matroskadec.c: 
[01:45] <CIA-119> ffmpeg: matroskadec: move lace_size check up so it catches all code pathes
[01:45] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:45] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r4b7c52346a 10ffmpeg/libavformat/matroskadec.c: 
[01:45] <CIA-119> ffmpeg: matroskadec: change size check in matroska_decode_buffer() to unsigned
[01:45] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:45] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r59c122b3b0 10ffmpeg/libavformat/matroskadec.c: 
[01:45] <CIA-119> ffmpeg: matroskadec: add assert on lack of overflow in pkt_size+offset
[01:45] <CIA-119> ffmpeg: currently a overflow there should be impossible but future changes to
[01:45] <CIA-119> ffmpeg: the code could easily introduce a bug that no longer limits the 2
[01:45] <CIA-119> ffmpeg: values sufficiently so better protect it via av_assert.
[01:45] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[02:55] <CIA-119> ffmpeg: 03Paul B Mahol 07master * ra09ae4ef65 10ffmpeg/libavcodec/ (png.c png.h png_parser.c pngdec.c pngenc.c): 
[02:55] <CIA-119> ffmpeg: Remove duplicated png and mng signatures.
[02:55] <CIA-119> ffmpeg: Reviewed-by: Michael Niedermayer <michaelni at gmx.at>
[02:55] <CIA-119> ffmpeg: Signed-off-by: Paul B Mahol <onemda at gmail.com>
[03:02] <CIA-119> ffmpeg: 03Paul B Mahol 07master * r62c2daab0d 10ffmpeg/configure: 
[03:02] <CIA-119> ffmpeg: configure: do not auto enable mlp decoder when enabling truehd decoder
[03:02] <CIA-119> ffmpeg: It is generally bad idea even if codecs use same code.
[03:02] <CIA-119> ffmpeg: Signed-off-by: Paul B Mahol <onemda at gmail.com>
[03:19] <CIA-119> ffmpeg: 03J. Bohl 07master * r0719e44b58 10ffmpeg/ (libavcodec/motion_est_template.c libavutil/attributes.h): 
[03:19] <CIA-119> ffmpeg: new attribute "extern inline" (fixing linker error with ff_get_mb_score and ff_get_mb_score)
[03:19] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:03] <Compn> oh man
[04:03] <Compn> i saw a bunch of DIVX discs the other day
[04:03] <Compn> for $1 each
[04:03] <Compn> circuit city DIVX format
[04:04] <Compn> still uncracked, last i heard
[04:04] <Compn> i should have bought one, but probably mike already got some
[04:04] <Compn> guess i'll email him and see
[10:34] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * rcd65cd8c5a 10ffmpeg/libavutil/attributes.h: 
[10:34] <CIA-119> ffmpeg: revert "extern inline" usage
[10:34] <CIA-119> ffmpeg: as it breaks ICC:
[10:34] <CIA-119> ffmpeg: libavcodec/libavcodec.a(snowenc.o): In function `encode_q_branch':
[10:34] <CIA-119> ffmpeg: /home/fate/x86_64-linux-gnu-icc-2011.4.191/src/libavcodec/snowenc.c:405: undefined reference to `ff_epzs_motion_search'
[10:34] <CIA-119> ffmpeg: /home/fate/x86_64-linux-gnu-icc-2011.4.191/src/libavcodec/snowenc.c:414: undefined reference to `ff_get_mb_score'
[10:34] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[11:23] <CIA-119> ffmpeg: 03William Yu 07master * rf77a695d2e 10ffmpeg/libavfilter/af_amerge.c: 
[11:23] <CIA-119> ffmpeg: lavfi: af_amerge options array was not ended with NULL
[11:23] <CIA-119> ffmpeg: Add a NULL to prevent undefined behavior
[11:23] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[11:32] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r11153a9371 10ffmpeg/libavformat/ (avio.h aviobuf.c): 
[11:32] <CIA-119> ffmpeg: avio: Collect and print statistics of bytes read & seeks
[11:32] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[15:17] <burek> one noobish question :) why is it easier/more convenient to write an internal encoder (like aac) instead of forking an existing libfaac (or similar) and fixing it to work properly?
[15:18] <iive> licence problems?
[15:19] <burek> oh, ok :) makes sense :)
[15:23] <av500> burek: also source code formatting issues :)
[15:24] <burek> well, if you create a fork, then you can reformat it, right? :)
[15:24] <av500> that would be unethical :)
[15:24] <CIA-119> ffmpeg: 03Martin Sliwka 07master * r9b6f2c7483 10ffmpeg/libavformat/file.c: 
[15:24] <CIA-119> ffmpeg: file: Fix handling of windows named pipes
[15:24] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[15:25] <burek> why? :) really, I don't know if someone would mind, if you fork a project, I mean, for just reformatting the source code.. :)
[15:25] <burek> is that really an issue sometimes?
[17:12] <expresspotato> Hi, does anyone have experience with building FFmpeg for Android?
[18:42] <expresspotato> does anyone know how I can get the average video frame time ?
[18:49] <CIA-119> ffmpeg: 03Clément BSsch 07master * r608bda925a 10ffmpeg/libavformat/ (microdvddec.c srtdec.c): lavf/{srt,microdvd}: correctly raise error on avformat_new_stream() error.
[18:49] <CIA-119> ffmpeg: 03Clément BSsch 07master * r47395a4f52 10ffmpeg/libavcodec/microdvddec.c: lavc/microdvddec: properly reset non persistent color tag.
[18:49] <CIA-119> ffmpeg: 03Clément BSsch 07master * rb678daf05c 10ffmpeg/libavcodec/microdvddec.c: lavc/microdvddec: fix color parsing.
[19:45] <CIA-119> ffmpeg: 03Michael Bradshaw 07master * r67703d64d7 10ffmpeg/libavformat/utils.c: 
[19:45] <CIA-119> ffmpeg: lavf: cosmetic: format else with indentation
[19:45] <CIA-119> ffmpeg: Signed-off-by: Michael Bradshaw <mbradshaw at sorensonmedia.com>
[19:45] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[19:45] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r17bbb818ae 10ffmpeg/libavformat/amr.c: 
[19:45] <CIA-119> ffmpeg: amr: use av_assert()
[19:45] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[19:45] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r47f8303069 10ffmpeg/libavformat/aviobuf.c: 
[19:45] <CIA-119> ffmpeg: lavf/aviobuf: use av_assert()
[19:45] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[19:45] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * rdd2086140c 10ffmpeg/libavformat/nutenc.c: 
[19:45] <CIA-119> ffmpeg: nutenc: use av_assert
[19:45] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[20:39] <CIA-119> ffmpeg: 03Paul B Mahol 07master * r29e9f83192 10ffmpeg/libavcodec/j2kdec.c: 
[20:39] <CIA-119> ffmpeg: j2kdec: check colorspace ncomponents
[20:39] <CIA-119> ffmpeg: This prevents out of array read.
[20:39] <CIA-119> ffmpeg: Signed-off-by: Paul B Mahol <onemda at gmail.com>
[20:46] <CIA-119> ffmpeg: 03Paul B Mahol 07master * rbe32145e9d 10ffmpeg/libavcodec/j2kdec.c: 
[20:46] <CIA-119> ffmpeg: j2kdec: move s->avctx initialization to j2kdec_init()
[20:46] <CIA-119> ffmpeg: Signed-off-by: Paul B Mahol <onemda at gmail.com>
[20:50] <expresspotato> hi, does anyone know how I can get the fps using get_fps?
[21:02] <expresspotato> hmm looks like I'm looking in the wrong place... Does anyone know how I could get the input fps
[21:06] <Compn> huh?
[21:06] <Compn> ffprobe file ?
[21:10] <CIA-119> ffmpeg: 03Janne Grunau 07master * r1cdcf043ca 10ffmpeg/libavformat/segment.c: segment: remove unnecessary <strings.h> include
[21:10] <CIA-119> ffmpeg: 03Janne Grunau 07master * rfbc5740566 10ffmpeg/libavfilter/af_resample.c: af_resample: fix format modifier in debug string for FF_API_SAMPLERATE64
[21:10] <CIA-119> ffmpeg: 03Mans Rullgard 07master * r16b8525963 10ffmpeg/tests/ (3 files in 2 dirs): 
[21:10] <CIA-119> ffmpeg: fate: add snow hpel tests
[21:10] <CIA-119> ffmpeg: Signed-off-by: Janne Grunau <janne-libav at jannau.net>
[21:10] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r90d8506fbf 10ffmpeg/: 
[21:10] <CIA-119> ffmpeg: Merge remote-tracking branch 'qatar/master'
[21:10] <CIA-119> ffmpeg: * qatar/master:
[21:10] <CIA-119> ffmpeg:  af_resample: fix format modifier in debug string for FF_API_SAMPLERATE64
[21:10] <CIA-119> ffmpeg:  segment: remove unnecessary <strings.h> include
[21:10] <CIA-119> ffmpeg:  fate: add snow hpel tests
[21:10] <CIA-119> ffmpeg: Merged-by: Michael Niedermayer <michaelni at gmx.at>
[23:18] <llogan> Zeranoe: someone in #ffmpeg mentioned that your builds were crashing when encoding with libutvideo. i couldn't duplicate it in linux
[23:19] <llogan> (but i used a newet utvideo if that means anything)
[23:19] <llogan> *newer
[23:19] <llogan> this was 6 days ago, IIRC.
[23:44] <ubitux> saste: arg, i forgot to come on the channel :(
[23:45] <saste> you missed nothing particularly spectacular
[23:45] <ubitux> thx for the backlog, i'm reading it :)
[00:00] --- Sat Jun 16 2012