[Ffmpeg-devel-irc] ffmpeg-devel.log.20130614

[burek]

[00:03] <durandal_1707> michaelni: i think its resolved now
[00:05] <michaelni> durandal_1707, thanks
[00:16] <wm4> michaelni: could you reproduce the problem with the video I posted? is it a known problem? shall I create a ticket?
[00:26] <michaelni> wm4, reproduced, i dont think its a known issue
[00:26] <michaelni> opening a ticket cant hurt
[00:28] <ubitux> i can't figure out where the qp table is set when using -f lavfi -i testsrc
[00:28] <ubitux> any idea?
[00:29] <ubitux> oh my bad
[00:29] <ubitux> it's not
[00:47] <cone-185> ffmpeg.git 03Michael Niedermayer 07master:129edcb50fac: jpeg2000dec: dont set ncomponents to an invalid value
[00:54] <ubitux> microchip_: do you have a blocky sample i could use to do some tests?
[01:28] <cone-185> ffmpeg.git 03Clément BSsch 07master:a2c547ffecb1: lavfi: add spp filter.
[01:28] <cone-185> ffmpeg.git 03Clément BSsch 07master:852f74bd859d: lavfi/mp: remove mp=spp.
[02:16] <cone-185> ffmpeg.git 03Michael Niedermayer 07master:e850a064cee3: jpeg2000dec: Check compno in get_coc()
[04:05] <cone-185> ffmpeg.git 03Jean Delvare 07master:cef42ded8dd0: drawbox: Respect thickness parameter
[04:05] <cone-185> ffmpeg.git 03Timothy Gu 07master:6fe419bf7314: doc/general: Make the license status of the Android libraries clearer
[06:58] <cone-185> ffmpeg.git 03Clément BSsch 07master:147adf2c4fc0: lavfi/spp: fix description.
[10:45] <khali> ubitux: if I send a patch set which modifies the same filter more than once, should I increase LIBAVFILTER_VERSION_MICRO each time, or only once?
[10:47] <ubitux> you can only bump at the end
[10:47] <ubitux> i guess :)
[10:47] <ubitux> it's not really important for micro bump anyway
[10:48] <khali> ubitux: well you asked me to increase it in my future patches, didn't you?
[10:48] <ubitux> i just asked for the last one since it was changing the defaults and behaviour, but that's not important
[10:57] <khali> ubitux: OK, I don't care myself, so I'll just do as told
[10:58] <ubitux> your patch was applied without the bump, but it's my fault for not mentioning it on the ml anyway; dont worry about that
[11:43] <LordDoskias> hello
[11:43] <LordDoskias> seems, like a wrong channel :D 
[11:49] <khali> ubitux: I'll send more patches soon so the bump will happen anyway
[11:49] <bor0> is there a way to get progress in % using FFmpeg? for instance, it is easy to calculate progress when transcoding video from one type to another, but consider the case where a user inputs a command which generates 3 thumbnails from a given video. how could I calculate the current progress over total progress in that case?
[12:59] <cone-378> ffmpeg.git 03Hendrik Leppkes 07master:59d3c24e28ba: vc1_parser: fix parsing of the frame headers in interlaced streams
[13:42] <cone-378> ffmpeg.git 03Peter Ross 07master:d891d354c551: mpegts: Enable muxing of SMPTE KLV metadata
[14:48] <damir__> hello. I am using mpeg2video encoder. when I set 'context->scenechange_threshold = MAX_MIN', encoder still inserts keyframes at scene changes
[14:48] <damir__> can anyone clarify?
[14:48] <damir__> i want to disable the scene change detection completely
[14:49] <damir__> typo... is set to INT_MAX
[14:52] <av500> maybe it does not work
[14:53] <damir__> is there any other way to disable scene change detection (or to limit keyframe frequency)?
[14:53] <damir__> i want keyframes exactly every 12 frames, no more no less
[14:55] <michaelni> damir__, can this be reproduced with command line ffmpeg ?
[14:56] <damir__> let me try
[14:57] <damir__> the problem is I-frame insertion causes 'closed gops' that don't start with I-frame
[14:57] <damir__> because it offsets everything
[14:58] <michaelni> Hi BuxiNess 
[14:59] <j-b> Salut Nicolas!
[14:59] <michaelni> do you have a moment to review my patch on ffmpeg-devel ? i didnt double check it against the spec so it needs a review
[15:10] <BuxiNess> Hi all
[15:10] <BuxiNess> michaelni, not yet
[15:10] <BuxiNess> I'll do that begining next week
[15:17] <michaelni> BuxiNess, ok
[15:28] <cone-378> ffmpeg.git 03Michael Niedermayer 07master:7fff3df6b277: avcodec_encode_audio2: fix memleak on last padded frame
[19:06] <durandal_1707> what i need to bring with me for this secret meeting?
[19:06] <ubitux> cookies
[19:10] <cone-39> ffmpeg.git 03Paul B Mahol 07master:a75d2a6505bf: apetag: do not require seekable output
[19:25] <ubitux> so, since it seems pp7, fspp and uspp are still wanted
[19:25] <ubitux> who's going to port them?
[19:26] <ubitux> we have 13 filters left to port
[19:26] <ubitux> assuming eq/eq2 can be merged into one, there is a total of 12 filters remaining
[19:31] <wm4> ubitux: I'd say delete all these pp filters and point the ones who want it to mplayer
[19:31] <ubitux> wm4: fspp seems to be wanted by at least two persons (see ml)
[19:31] <durandal_1707> just delete all libmpcodecs and point ones who want them to PATCHWELCOME
[19:31] <ubitux> uspp seems to have better results than spp for low qp
[19:32] <ubitux> and pp7 has better results according to microchip_ 
[19:32] <ubitux> i personally don't care much, but there are some demands
[19:33] <ubitux> wm4: see http://ffmpeg.org/pipermail/ffmpeg-devel/2013-June/144729.html
[19:33] <wm4> then ask them to port it
[19:34] <ubitux> that won't happen
[19:34] <ubitux> unfortunately.
[19:34] <wm4> sigh
[19:34] <wm4> well, not my mess fortunately
[19:34] <ubitux> but yeah i'll ask for
[20:17] <cone-39> ffmpeg.git 03Michael Niedermayer 07master:5deb96c5641c: avcodec/jpeg2000dec: Check Psot
[20:17] <cone-39> ffmpeg.git 03Michael Niedermayer 07master:69e4d8e6a4ce: avcodec/jpeg2000dec: check len before parsing header
[21:26] <cone-39> ffmpeg.git 03Michael Niedermayer 07master:3c4c5ca1c194: avcodec/g2meet: Check monochrome cursor width
[00:00] --- Sat Jun 15 2013