[Ffmpeg-devel-irc] ffmpeg-devel.log.20141019

[burek]

[00:50] <kierank> ffmpeg is on tv
[00:51] <amalia_> kierank Which TV is that ?
[00:52] <kierank> bbc four, "the code"
[01:02] <amalia_> kierank Thanks for your tips on the BBC  I have just downloaded the ffmpeg code from github. How do I compile it ?
[01:06] <amalia_> Compn J_Darnley How do I compile the ffmpeg code on my machine ?
[01:06] <kierank> https://trac.ffmpeg.org/wiki/CompilationGuide/Generic
[01:06] <amalia_> Thanks kierank
[01:09] <pross> kierank: that is an australian show \o/
[01:09] <kierank> yup
[01:09] <kierank> https://fbcdn-sphotos-d-a.akamaihd.net/hphotos-ak-xpf1/v/t1.0-9/1794677_585293414909614_7587187744299178872_n.jpg?oh=6b96d83ec423ec9224c71b1995bfcece&oe=54AF208E&__gda__=1421473615_2d6436ecc0035fb4b1bd1cc54527e905
[01:10] <kierank> and there are some historgrams
[01:10] <kierank> which could be the waveform filter
[01:13] <iive> kierank: ffplay when playing audio only shows histograms
[01:15] <J_Darnley> kierank: neat
[01:16] <J_Darnley> I wonder what decoder(s) are disabled in that configure line.
[04:18] <cone-741> ffmpeg.git 03Michael Niedermayer 07master:993a2487de7d: avcodec/nellymoserenc: use av_freep(), do not leave stale pointers in memory
[04:18] <cone-741> ffmpeg.git 03Michael Niedermayer 07master:c5e295cb9d57: avcodec/mpegaudiodec_template: use av_freep(), do not leave stale pointers in memory
[04:18] <cone-741> ffmpeg.git 03Michael Niedermayer 07master:765abb3b76a1: avcodec/libvpxenc: use av_freep(), do not leave stale pointers in memory
[04:28] <cone-741> ffmpeg.git 03Michael Niedermayer 07master:e23af4662dc6: avcodec/utils: Display pixel format information as none ... if unknown instead of omitting it
[05:05] <cone-741> ffmpeg.git 03Michael Niedermayer 07master:5c8eb16769e5: avcodec: add codec_whitelist
[05:05] <cone-741> ffmpeg.git 03Michael Niedermayer 07master:057ea2a9823a: avformat: Add format_whitelist
[05:16] <cone-741> ffmpeg.git 03Simon Thelen 07master:69071e8e9c04: doc/filters: fix localtime drawtext example.
[05:51] <jamrial> wait so, the symmetric-key block ciphers OPW project had three objectives, implementing CAST-128, Twofish and Camellia. This means the student went and did a third of the project as the qualification task?
[05:55] <wm4> jamrial: I guess the project objectives could be raised on demand
[06:02] <jamrial> maybe aes-ni as kierank requested on trac. though the project requisites never stated asm skills, so probably not.
[15:08] <amalia> I am new to github - Created a folder https://github.com/Ngassa/myFFmpeg for a copy of the ffmpeg code but it's empty. What do I do ? 
[15:09] <amalia> I am an OPW student new to github
[15:09] <wm4> amalia: you don't need to use github, but if you want to, then just clone the official ffmpeg repo on github
[15:10] <amalia> Okay. Just what command do I use ?
[15:11] <wm4> every repo has a "fork" button
[15:17] <Compn> wm4 : she clicked that , presented with 3 options, 'quick setup' 'create new repo on command line' 'push code from computer'
[15:17] <Compn> i guessed 'create new repo on the command line' , i havent done github before :P
[15:58] <cone-35> ffmpeg.git 03Michael Niedermayer 07master:a71a5c911eb0: avformat/riff: Add UGY2
[15:58] <cone-35> ffmpeg.git 03Kevin Mitchell 07master:9d51bad6255e: avfilter/vf_idet: add counts to frame metadata
[17:35] <Compn> michaelni : seems rtmpdump-web is not updating http://rtmpdump.mplayerhq.hu
[17:36] <kierank> jamrial: yes but she seems quite competent
[17:54] <michaelni> Compn, what exactly is where not updating ? what file or content ?
[18:19] <cone-35> ffmpeg.git 03Nicholas Robbins 07master:22cfa1f759ac: libavfi: added option to vf_perspective to specify transformation by giving destinations of corners of source.
[18:25] <Compn> michaelni : erm, updating the index.htm is not being updated on the website
[18:26] <Compn> index.html
[18:26] <Compn> changed the mplayer link from a git repo to mplayerhq.hu (bottom of page)
[18:26] <Compn> and it hasnt updated
[18:33] <Compn> michaelni : nevermind. looks like its working somehow
[18:33] <Compn> i must have done something wrong before , sorry for the noise
[18:39] <wm4> so why is libutvideo being kept around
[18:40] <wm4> didn't someone say the lib is unmaintained and abandoned
[18:43] <Compn> wm4 : michael offered to maintain it
[18:44] <Compn> or do you mean the libutvideo itself and not our wrapper ?
[18:44] <wm4> itself
[18:44] <Compn> i was unclear which part was unmaintained, good question
[18:44] <JEEB> it seems like someone is still kind of poking it from the release tarballs of Ut Video and kind of keeping it compiling
[18:44] <JEEB> it lacks the asm and the new features of Ut Video (10bit etc)
[18:45] <JEEB> and I think no-one is making even sure that its output matches the VFW one
[19:05] <cone-35> ffmpeg.git 03Kacper MichajBow 07master:f22cf88fd317: rtmpproto: Don't mistake app for playpath.
[20:17] <cone-35> ffmpeg.git 03Michael Niedermayer 07master:f63d38138716: ffprobe: Add support for color_trc/color_pri/chroma_loc
[20:39] <amalia> In which directory is the ALS decoder code found ?
[20:40] <JEEB> it's not a container but an actual a/v/s/whatever format so it is found under libavcodec
[20:40] <JEEB> containers and otherwise things to parse to get to actual a/v/s/whatever streams = libavformat (lavf)
[20:41] <JEEB> actual a/v/s/whatever formats/streams and their decoding/encoding = libavcodec (lavc)
[20:45] <amalia> What about the RLS-LMS encoder ?
[20:46] <JEEB> unless it's a container/format to be parsed or muxed (often called an 'encoder' in lavc/lavf speak, even though it's not really encoding), it's in lavc
[20:46] <JEEB> also git grep is your friend
[21:02] <wm4> amalia: all decoders and encoders are in the libavcodec sub-directory, but yeah, grepping helps
[21:02] <wm4> it will often happen that you have to track down something by reading code, grepping the sources, etc.
[21:02] <amalia> Thanks very very much  wm4
[21:03] <JEEB> wm4, saying that can get kind of hard because lavf loves calling muxers "encoders" :P
[21:04] <amalia> There are quite a number of files in that library
[21:04] <JEEB> yes, which is why you use tools like git grep to quickly find where exactly that thing is, or you just search within the file names something similar :P
[21:05] <wm4> true, libavformat has "enc" and "dec" suffixes on muxer and demuxer implementation files
[21:15] <Compn> would  you rather mu and dem ? :P
[21:15] <amalia> ?
[21:15] <JEEB> _mux _demux
[21:16] <Compn> amalia : just JEEB and i discussing the color of the bikeshed
[21:16] <JEEB> yeh
[21:16] <Compn> nothing important
[21:16] <wm4> libavdevice is the most dumb here
[21:16] <wm4> so reading from an alsa device is done by alsa-audio-dec.c
[21:17] <amalia> JEEb So an examplery git grep command would be ?
[21:18] Last message repeated 1 time(s).
[21:18] <JEEB> git grep -i "RMS"
[21:18] <amalia> That's much better
[21:18] <JEEB> related documentation: http://git-scm.com/docs/git-grep
[21:18] <amalia> Thank you
[22:07] <ubitux> ok, now i need an avi sample, with ssa in it, but with multiple Dialogue at the same time
[22:08] <nevcairiel> didnt my sample do that on the credits stream?
[22:10] <ubitux> nevcairiel: ah? let me check...
[22:16] <ubitux> nevcairiel: your sample only has 5 events
[22:16] <ubitux> but actually, it seems to use the ass demuxer itself so&
[22:16] <ubitux> it won't be a problem actually
[22:17] <ubitux> ah there are other streams indeed
[22:17] <nevcairiel> the credits stream only has like one bunch of events right at the start, the other ones have normal subtitle dialog
[22:18] <ubitux> anyway, all of this is really evil
[22:19] <ubitux> i'm going to drop the ssa packets support since it won't be necessary in this case actually
[22:19] <ubitux> and since nothing in ffmpeg outputs these&
[22:19] <ubitux> or maybe nut
[22:20] <ubitux> but can anyone find me a sample for that? :P
[22:34] <cone-35> ffmpeg.git 03Andrey Utkin 07master:65639e6bef80: avutil/error: Introduce new error codes for 4XX and 5XX replies from remote servers
[23:07] <ubitux> lol @ canon storing jpeg thumbnail, into the jpeg itself
[23:07] <ubitux> (in some custom exif tag)
[23:09] <wm4> why is that surprising
[23:10] <wm4> it only gets dumb if the thumbnail has exif and thumbnail too
[23:10] <J_Darnley> Does ffmpeg read that instead of the actual image?
[23:10] <ubitux> no
[23:11] <J_Darnley> Oh I'm glad that ffmpeg isn't that silly yet.
[23:11] <ubitux> and not the jpeg thumb doesn't have exif with a thumb
[23:11] <ubitux> no*
[23:26] <ubitux> i really won't be easy to test if i'm not breaking the ssa decoder, so does anyone object if i drop it?
[23:26] <ubitux> (i guess no one care but...)
[23:29] <cone-35> ffmpeg.git 03Clément BSsch 07master:3d258696057d: avcodec/ass: assume raw=0 in ff_ass_add_rect_bprint
[23:29] <cone-35> ffmpeg.git 03Clément BSsch 07master:c669af0f248e: avcodec/ass: add ASS trailing \r\n in one single place
[00:00] --- Mon Oct 20 2014