[Ffmpeg-devel-irc] ffmpeg-devel.log.20120704

burek burek021 at gmail.com
Thu Jul 5 02:05:04 CEST 2012


[00:05] <CIA-41> ffmpeg: 03Stefano Sabatini 07master * raf392efe51 10ffmpeg/libavformat/segment.c: 
[00:05] <CIA-41> ffmpeg: lavf/segment: rename SegmentContext.pb to list_pb
[00:05] <CIA-41> ffmpeg: The new name is more auto-explicative, as the pb is used for the segment
[00:05] <CIA-41> ffmpeg: list file.
[00:16] <durandal_1707> michaelni: how to prevent overflow in tables allocation for paf? limiting nb_frames and frame_blocks to INT_MAX/sizeof(uint32_t) ?
[00:18] <michaelni> without rechecking the code, yes
[00:36] <CIA-41> ffmpeg: 03Mans Rullgard 07master * r4719ea7e1e 10ffmpeg/libavcodec/flacdec.c: 
[00:36] <CIA-41> ffmpeg: flacdec: remove redundant setting of avctx->sample_fmt
[00:36] <CIA-41> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[00:36] <CIA-41> ffmpeg: 03Martin Storsjö 07master * r4d8516fdb1 10ffmpeg/libavcodec/ (snow.c snowdec.c snowenc.c): 
[00:36] <CIA-41> ffmpeg: snow: Check mallocs at init
[00:36] <CIA-41> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[00:36] <CIA-41> ffmpeg: 03Mans Rullgard 07master * rfb93e61e2b 10ffmpeg/libavfilter/x86/ (gradfun.c yadif.c): 
[00:36] <CIA-41> ffmpeg: x86: lavfi: fix gradfun/yadif build with mmx/sse disabled
[00:36] <CIA-41> ffmpeg: These functions are defined conditionally so any uses need to have
[00:36] <CIA-41> ffmpeg: preprocessor guards.
[00:36] <CIA-41> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[00:36] <CIA-41> ffmpeg: 03Luca Barbato 07master * r906f9dce85 10ffmpeg/avplay.c: 
[00:36] <CIA-41> ffmpeg: avplay: fix write on freed memory for rawvideo
[00:36] <CIA-41> ffmpeg: Do not assume avpacket and the decoded frames are independent.
[00:36] <CIA-41> ffmpeg: To be absolutely sure and not sprinkle av_free_packet around the code
[00:36] <CIA-41> ffmpeg: the call had been placed before getting the frame and on the error path.
[00:36] <CIA-41> ffmpeg: 03Luca Barbato 07master * r09a445ce34 10ffmpeg/libavformat/flvdec.c: 
[01:12] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rf91b7d8030 10ffmpeg/cmdutils.c: 
[01:12] <CIA-41> ffmpeg: mdutils: print error message on error get_buffer related errors
[01:12] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:12] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rf339ebc1ff 10ffmpeg/libavcodec/utils.c: 
[01:12] <CIA-41> ffmpeg: lavc/utils: print error message on get_buffer related errors
[01:12] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:12] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r93d672967d 10ffmpeg/libavcodec/utils.c: 
[01:12] <CIA-41> ffmpeg: video_get_buffer: return ENOMEM instead of -1 on malloc failure
[01:12] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:12] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r5eb4af6c59 10ffmpeg/libavcodec/ (snow.c snowdec.c): 
[01:12] <CIA-41> ffmpeg: snow: move init code that depends on picture paramaters to after these parameters are known.
[01:12] <CIA-41> ffmpeg: This should fix debug 2048 amongth other things
[01:12] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:20] <CIA-41> ffmpeg: 03Stefano Sabatini 07master * r776a4e42a7 10ffmpeg/doc/muxers.texi: 
[01:20] <CIA-41> ffmpeg: doc/muxers/segment: apply minor documentation updates and adjustements
[01:20] <CIA-41> ffmpeg: Update documentation to match code behavior after previous changes, also
[01:20] <CIA-41> ffmpeg: reword the examples section.
[02:12] <Compn> whoaaa
[02:12] <Compn> ami_stuff figured out yuy8 :)
[02:12] <Compn> er yuv8
[03:37] <burek> is -af fully integrated now in ffmpeg or -f lavfi is to be used still?
[03:45] <durandal_1707> burek: AFAIK -af works with ffmpeg but not with ffplay
[03:45] <burek> I see, good enough :) thanks :)
[04:07] <burek> ok, I moved tutorials from the forum to the wiki: https://ffmpeg.org/trac/ffmpeg/wiki
[04:07] <burek> feel free to edit it if you like
[04:08] <burek> or dislike :)
[04:17] <CIA-41> ffmpeg: 03Piotr Bandurski 07master * ra54bc1ebb8 10ffmpeg/libavformat/riff.c: 
[04:17] <CIA-41> ffmpeg: riff: add "YUV8" FourCC
[04:17] <CIA-41> ffmpeg: sample:
[04:17] <CIA-41> ffmpeg: http://samples.mplayerhq.hu/V-codecs/yuv8/MAILTEST.AVI
[04:17] <CIA-41> ffmpeg: Reviewed-by: compn <tempn at twmi.rr.com>
[04:17] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:17] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rdf3a96c497 10ffmpeg/ffmpeg.c: 
[04:17] <CIA-41> ffmpeg: ffmpeg: Add sanity check on frame duplication
[04:17] <CIA-41> ffmpeg: This prevents long loops and OOM in case of damaged timestamps in some
[04:17] <CIA-41> ffmpeg: cases
[04:17] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[07:50] <jesk> any suggestion how to fix "Invalid MPEG-4 Descriptor"? Any tool?
[07:50] <jesk> "Invalid MPEG-4 Descriptor" is echoed from MP4box
[08:36] <jesk> VLC does not support the audio or video format "undf". VLC does not support the audio or video format "undf". 
[08:36] <jesk> whe
[08:37] <jesk> gnah sorry, multicast DoSed my wifi :)
[08:37] <jesk> and that without any benefit, stream not working
[08:38] <jesk> got anyone some clue how to fix that?
[08:47] <CIA-41> ffmpeg: 03Carl Eugen Hoyos 07master * rab2356cb79 10ffmpeg/configure: Let the ffmpeg executable select the abuffersink filter.
[09:09] <ubitux> + at item Amazing Studio Packed Animation File        @tab   @tab X
[09:09] <ubitux> +    @tab Multimedia format used in game Heart Of Darkness.
[09:09] <ubitux> :o
[09:09] <ubitux> HoD e
[11:37] <ubitux> i was wondering, maybe i should disable the memory poisoning for valgrind, so use of uninitialized data are detected&
[11:53] <michaelni> ubitux, yes
[11:53] <ubitux> i added --disable-memory-poisoning for the full one
[11:53] <ubitux> should be done in the next run
[11:58] <michaelni> thx
[14:01] <ubitux> is Georg Lippitsch on IRC?
[14:02] <burek> does anyone understand this: http://ffmpeg.gusari.org/viewtopic.php?f=12&t=617
[14:02] <burek> is it a feature suggestion or an issue troubleshooting or what?
[14:03] <ubitux> wtf is this 30px <pre> height? :D
[14:03] <burek> where? :)
[14:04] <ubitux> the code samples
[14:05] <ubitux> it's awful to scroll them
[14:05] <ubitux> i don't understand what the guy want
[14:05] <burek> it's the default, I dunno :)
[14:05] <burek> I just don't know if he's asking for help or showcasing something :D
[14:07] <burek> btw, ubitux, try at: georg.lippitsch at gmx.at
[14:07] <ubitux> burek: i know the mail, i wanted to know if he was here :)
[14:39] <varaderoguy> [with applogies to those in #ffmpeg channel]
[14:40] <varaderoguy> I would like to what the RED Camera support is like at the moment
[14:40] <varaderoguy> I can see from a nightly build from a week ago, that decoder was supported, but to what degree?
[14:40] <varaderoguy> Is the encrypted RED Camera material supported yet?
[14:41] <kierank> probably no
[14:42] <varaderoguy> +kierank: is that a definate no, or just a not quite sure at the moment....
[14:43] <kierank> afaik there's no decryption in ffmpeg apart from one special case
[14:43] <kierank> so no
[14:44] <varaderoguy> +kierank: was this because of the encpytion or because RED [the company] got arsy [lawyers etc] 
[14:44] <kierank> it's because nobody implemented it
[14:45] <kierank> it is probably crackable if someone had the time
[14:45] <varaderoguy> I noticed that RED have offered a developer code if needed
[14:46] <varaderoguy> not sure if that would be a glob or actual code
[14:46] <kierank> iirc they encrypted it because of ffmpeg
[14:47] <Tjoppen> hooray for vendor lock-in
[14:50] <varaderoguy> Stupid Q: Has anybody actually - formally - asked RED to see whether they would be willing to contribute code?
[14:50] <kierank> the r3d sdk says you can't release source code apparently
[14:50] <kierank> well red offered me a job once
[14:50] <kierank> and i hate their bs so i said no
[14:51] <Tjoppen> I thought red was an underdog, what with 4k cameras being way expensive before they came around
[14:51] <Tjoppen> making their acting seem strange
[14:51] <varaderoguy> okay - for the sake of 'getting something working' - even if there was a precomiled glob, would that get around the issue?
[14:52] <varaderoguy> ...and hooks were written into ffmpeg to get the glob to work???
[14:53] <kierank> well it wouldn't be allowed in ffmpeg
[14:53] <varaderoguy> ...and that's because of GPL
[14:53] <varaderoguy> :-(
[14:53] <Tjoppen> you could roll your own, and put it up on github or something
[14:53] <Tjoppen> the magic of DVCS
[14:54] <kierank> you wouldn't be allowed to distribute the source
[14:54] <kierank> because of the red sdk
[14:54] <varaderoguy> +Kierank: So you've actually taken a look at the source
[14:54] <kierank> no, i just know the red sdk says you can't distribute source code
[14:55] <Tjoppen> yes, but you can still provide the wrapper
[14:55] <Tjoppen> for people with the code
[14:55] <kierank> apparently the licence prohibits that as well
[14:55] <kierank> i.e you can't reference functions in the header
[14:55] <varaderoguy> ????
[14:55] <Tjoppen> it can't. see the recent oracle vs google case over the java api
[14:56] <kierank> well that's beyond my pay grade, whether what you've signed is legal or not
[14:56] <varaderoguy> The joke seems to be that their hardware card looks like an nVidia with an encryption chip on it...
[14:57] <Tjoppen> or you just publish it anonymously. whatever
[14:57] <varaderoguy> I'm just not feeling rich enough
[14:57] <varaderoguy> What I was thinking was this:
[14:58] <varaderoguy> if I could get RED to offer up the SDK with the proviso that ffmpeg could hook into it with headers, would you be game to programme that?
[14:59] <kierank> well you can hire someone to do it
[14:59] <leandrosansilva> Hello to all. Does avfilter_copy_buf_props replace  avfilter_fill_frame_from_video_buffer_ref or I need to execute some code before?
[15:00] <varaderoguy> I'll make some phonecalls
[15:00] <leandrosansilva> because I was using  avfilter_fill_frame_from_video_buffer_ref and my code was working properly, but when I changed to avfilter_copy_buf_props, this function always return error
[15:00] <iive> varaderoguy: what is this red camera?
[15:00] <iive> i assume it is not cheap web cam.
[15:01] <Tjoppen> iive: www.red.com
[15:01] <varaderoguy> @iive: www.red.com
[15:01] <kierank> varaderoguy: would be better to hire someone to reverse engineer the encryption of course
[15:01] <iive> please, don't send me in the forest to look up for a single tree.
[15:02] <Tjoppen> iive: 4k camera, format is bayer sensor data encoded as jpeg2000
[15:02] <Tjoppen> or used to be at least
[15:02] <varaderoguy> Kierank: of course, I agree.....depends on whether my company will sponser such a job....
[15:02] <varaderoguy> Main Concept might take it up....
[15:03] <iive> is it photo or video/movie camera? why does it need special codec support?
[15:03] <ubitux> leandrosansilva: you can just rename the call
[15:03] <ubitux> leandrosansilva: look at the function in lavfi/avcodec.c, it just forwards the call
[15:03] <varaderoguy> iive: its a top end 2k/4k video camera - which purpose is to try and replace film
[15:03] <leandrosansilva> Now it returns the error code "Invalid argument"
[15:04] <ubitux> leandrosansilva: then something is wrong in what you pass to the API :P
[15:05] <leandrosansilva> ubitux, yes, may be
[15:06] <ubitux> leandrosansilva: check the ref->type, and check if dst is not NULL
[15:09] <DEgITx> AVCodecContext have any field with header size?
[15:14] <leandrosansilva> ubitex, strange... should I manually set the ref->type?
[15:14] <leandrosansilva> I'm filling it with av_buffersink_read()
[15:15] <ubitux> you mean av_buffersink_get_buffer_ref()?
[15:16] <ubitux> check doc/examples for filtering example
[15:18] <leandrosansilva> hum... ops
[15:18] <ubitux> (you should not set ref->type manually no)
[15:20] <leandrosansilva> yes, I realized I was using the wrong function
[15:20] <leandrosansilva> thx a lot
[15:21] <ubitux> np
[15:23] <leandrosansilva> ubitux, have you already used the scale filter?
[15:23] <ubitux> ask your question :p
[15:28] <leandrosansilva> ok, it's a bit dificult without showing my code, but I'm using the scale filter based in the example in doc/examples, but when I try to scale a image (I'm using a jpeg file) to a smaller size, the final image is wrong.
[15:29] <leandrosansilva> the original image is: http://imageshack.us/photo/my-images/515/imagem1lq.jpg/
[15:29] <ubitux> wonderful image
[15:29] <leandrosansilva> and the scaled (to half) is http://imageshack.us/photo/my-images/225/outn.jpg/
[15:29] <leandrosansilva> the image is 1920x1080
[15:30] <leandrosansilva> and the filter I'm using is scale=1920/2:1080/2
[15:30] <ubitux> then show the code
[15:30] <leandrosansilva> probably it's some problem with my code, because when I use ffmpeg or ffplay it works fine
[15:31] <ubitux> it's likely you write the frame based on the dimension of the source
[15:31] <leandrosansilva> in fact the code is a "spagetti", because it's mixed with other (c++) code
[15:31] <leandrosansilva> so I'll try
[15:31] <leandrosansilva> hum...
[15:31] <leandrosansilva> you may be right
[15:31] <leandrosansilva> let me check here
[15:53] <DEgITx> guys, so what about my question... is there any field in AVCodecContext with video file header size? Now i'm getting it from packet position after av_read_frame() ... thats not very good...
[15:59] <CIA-41> ffmpeg: 03Nicolas George 07master * raa39516a04 10ffmpeg/libavutil/ (error.c error.h): 
[15:59] <CIA-41> ffmpeg: error: define AVERROR_BUFFER_TOO_SMALL.
[15:59] <CIA-41> ffmpeg: Some API require a buffer and return a specific error
[15:59] <CIA-41> ffmpeg: if it is too small. This error code helps to map codes.
[15:59] <CIA-41> ffmpeg: 03Nicolas George 07master * r7beeea8f17 10ffmpeg/libavutil/ (error.c error.h version.h): 
[15:59] <CIA-41> ffmpeg: error: define AVERROR_EXTERNAL.
[15:59] <CIA-41> ffmpeg: This code is intended for errors in external libraries
[15:59] <CIA-41> ffmpeg: when no corresponding error code can be found.
[15:59] <CIA-41> ffmpeg: AVERROR_UNKNOWN is too vague for that and looks like AVERROR_BUG.
[16:18] <burek> docs bug: option "overrun_nonfatal" for udp protocol is not documented, but found in: ./libavformat/udp.c:262: *         'overrun_nonfatal=1': survive in case of circular buffer overrun
[16:19] <burek> http://ffmpeg.org/ffmpeg.html#udp doesn't mention it
[16:21] <durandal_1707> burek: make a bug report or even better make a patch
[16:23] <burek> well, again.. if I take time to do all that, someone will have to review it anyway and again we spend double amount of time on that..
[16:23] <burek> I'll create a ticket
[16:24] <durandal_1707> burek: huh, i disagree (currently not much effort is spent on documenting anything)
[16:25] <burek> you probably weren't present several days ago when wiki discussion was on here :)
[16:25] <burek> the problem is that I have to learn how to properly use git to be able to revert things if I scr.. something up and stuff
[16:25] <burek> anyway, it's easier to create a ticket hoping someone will eventually fix it
[16:26] <durandal_1707> i lost hope long ago....
[16:27] <burek> hope for what? :)
[16:27] <durandal_1707> better tomorrow
[16:27] <burek> :)
[16:28] <burek> shortly, my belief is that we should find a suitable way for devels and ordinary users to edit documentation easily
[16:28] <burek> that way nobody has to learn texinfo file format and git usage
[16:28] <burek> to be able to contribute
[16:30] <CIA-41> ffmpeg: 03Nicolas George 07master * r2fc354f90d 10ffmpeg/ (10 files in 3 dirs): (log message trimmed)
[16:30] <CIA-41> ffmpeg: ffmpeg: rework checks for the -t option.
[16:30] <CIA-41> ffmpeg: This commit is based on libav's implementation and
[16:30] <CIA-41> ffmpeg: makes sure to compare output timestamps together.
[16:30] <CIA-41> ffmpeg: It also reduces the differences with avconv.
[16:30] <CIA-41> ffmpeg: The changes to the test reference files are caused
[16:30] <CIA-41> ffmpeg: by an additional packet at the end, the timestamp
[16:41] <durandal_1707> michaelni: what about log colors for other stuff like: sws, swr, protocols and similar?
[16:51] <michaelni> we need more colors ...
[16:52] <durandal_1707> michaelni: why you did not removed av_find_stream_info? libav did.
[16:53] <ubitux> it was put back in 8a07429318acfbf497981a0ba1f68e2966ce3fdf
[16:54] <ubitux> removed to early maybe?
[16:56] <durandal_1707> perhaps because av_find_stream_info is still mentioned in headers
[16:58] <burek> sorry to interrupt, but is there any filter that can delay both audio/video streams (like -itsoffset that delays video only)
[16:58] <burek> or any other way.. a guy has 2 inputs (webcams) which he can't sync properly (2 a/v streams)
[17:10] <durandal_1707> michaelni: more colors are really trivial to add but some terminals just lie how many colors they support and may cause colored stuff to blink instead...
[17:21] <JEEB> burek, I really recommend you link mewiki generally for stuff like x264's presets :) That dump is fine, too, but the wiki is usually more newbie-friendly
[17:21] <durandal_1707> michaelni: lavu currently uses first 8 bits as background and next 8 bits as foreground
[17:21] <burek> ok
[17:22] <durandal_1707> michaelni: s/8/4
[17:22] <JEEB> burek, http://mewiki.project357.com/wiki/X264_Settings#preset with or without the #preset depending on what exactly is needed
[17:23] <durandal_1707> michaelni: so type for color from uint8_t should be changed to uint16_t
[18:51] <durandal_1707> michaelni: i have working 256 color support, just need to tweak colors...
[18:53] <michaelni> \o/
[18:58] <durandal_1707> michaelni: pick some colors 12bit rgb colors 6x6x6 bits per component
[19:09] <CIA-41> ffmpeg: 03Nicolas George 07master * r8b421fad24 10ffmpeg/libavcodec/dpx.c: 
[19:09] <CIA-41> ffmpeg: dpx: sanitize aspect ratio.
[19:09] <CIA-41> ffmpeg: Fixes some samples where the ratio is declared as -1/-1.
[19:24] <ubitux> durandal_1707: nice the PAF support :)
[19:25] <ubitux> durandal_1707: btw, PAF_AUDIO belongs before Q*
[19:26] <durandal_1707> michaelni: 256 color patch on ml
[19:28] <durandal_1707> ubitux: what?
[19:28] <ubitux> libavcodec/allcodecs.c
[19:28] <ubitux> you put the P* entry after the Q*
[19:28] <ubitux> just an alphabetical nit
[19:29] <durandal_1707> aha....
[19:32] <ubitux> durandal_1707: do you know a community based on extracting stuff like that from video games?
[19:33] <ubitux> most of the time the media stuff and basically all the content is packed in some random proprietary format
[19:33] <ubitux> s/format/archive/
[19:48] <durandal_1707> ubitux: http://wiki.xentax.com/
[19:49] <ubitux> oh nice
[19:49] <ubitux> thx
[19:51] <durandal_1707> well extracting big paf should not be hard
[19:51] <ubitux> sure, i was just curious
[21:09] <CIA-41> ffmpeg: 03Martin Storsjö 07master * r1be8c90847 10ffmpeg/tools/ismindex.c: 
[21:09] <CIA-41> ffmpeg: ismindex: Properly report errors reading the MFRA atom
[21:09] <CIA-41> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[21:09] <CIA-41> ffmpeg: 03Martin Storsjö 07master * r30327865f3 10ffmpeg/tools/ismindex.c: (log message trimmed)
[21:09] <CIA-41> ffmpeg: ismindex: Verify that all bitrate variants match
[21:09] <CIA-41> ffmpeg: In Smooth Streaming, the fragments are addressed by time, and
[21:09] <CIA-41> ffmpeg: the manifest only stores one list of time offests for all streams,
[21:09] <CIA-41> ffmpeg: so all streams need to have identical fragment offsets. Warn if
[21:09] <CIA-41> ffmpeg: this isn't the case, so that the user can fix the files instead of
[21:09] <CIA-41> ffmpeg: getting failures at runtime when the fragments can't be found.
[21:09] <CIA-41> ffmpeg: 03Mans Rullgard 07master * r983fa8bc46 10ffmpeg/tests/ (Makefile fate/flac.mak): 
[21:09] <CIA-41> ffmpeg: fate: add flac encode/decode tests with various options
[21:09] <CIA-41> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[21:09] <CIA-41> ffmpeg: 03Mans Rullgard 07master * r360aa2465a 10ffmpeg/libavcodec/flacenc.c: 
[21:09] <CIA-41> ffmpeg: flacenc: add option for forcing stereo decorrelation mode
[21:09] <CIA-41> ffmpeg: This is mainly useful for testing.
[21:09] <CIA-41> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[21:10] <CIA-41> ffmpeg: 03Mans Rullgard 07master * r0da301e105 10ffmpeg/libavcodec/flacdec.c: 
[21:10] <CIA-41> ffmpeg: flacdec: remove curr_bps from FLACContext
[21:10] <CIA-41> ffmpeg: This value does not need to be persistent across calls.
[21:10] <CIA-41> ffmpeg: 03Mans Rullgard 07master * rd155b60fc8 10ffmpeg/libavcodec/ (flac.c flac.h flacenc.c): flac: make FLAC_CHMODE_* constants consecutive
[21:10] <CIA-41> ffmpeg: 03Mans Rullgard 07master * r93e7ef9a24 10ffmpeg/libavcodec/flacdec.c: 
[21:10] <CIA-41> ffmpeg: flacdec: allocate sample buffers with av_malloc
[21:10] <CIA-41> ffmpeg: The buffers are only allocated once, although it can happen from
[21:10] <CIA-41> ffmpeg: any of a few different places, so there is no need to use realloc.
[21:10] <CIA-41> ffmpeg: Using av_malloc() ensures they are aligned suitably for SIMD
[21:10] <CIA-41> ffmpeg: optimisations.
[21:10] <CIA-41> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[21:10] <CIA-41> ffmpeg: 03Mans Rullgard 07master * r87466f8124 10ffmpeg/libavcodec/flacdec.c: flacdec: factor out code setting avctx->sample_fmt
[21:10] <CIA-41> ffmpeg: 03Anton Khirnov 07master * rbf2a7f9161 10ffmpeg/tests/ (Makefile fate/cover-art.mak): FATE: add a test for itunes cover art.
[21:11] <CIA-41> ffmpeg: 03Mans Rullgard 07master * r25accf93ad 10ffmpeg/libavcodec/ (flacdec.c flacdsp.c flacdsp.h): 
[21:11] <CIA-41> ffmpeg: flacdec: move lpc filter to flacdsp
[21:11] <CIA-41> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[21:11] <CIA-41> ffmpeg: 03Anton Khirnov 07master * r1c840fa227 10ffmpeg/tests/fate/cover-art.mak: FATE: add a test for WMA cover art.
[21:11] <CIA-41> ffmpeg: 03Anton Khirnov 07master * r8059d3b7d4 10ffmpeg/tests/fate/cover-art.mak: FATE: add a test for apetag cover art
[21:11] <CIA-41> ffmpeg: 03Mans Rullgard 07master * r4a8528349f 10ffmpeg/libavcodec/ (Makefile flacdec.c flacdsp.c flacdsp.h flacdsp_template.c): 
[21:11] <CIA-41> ffmpeg: flacdec: split off channel decorrelation as flacdsp
[21:11] <CIA-41> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[21:11] <CIA-41> ffmpeg: 03Martin Storsjö 07master * r296d0da8bd 10ffmpeg/avplay.c: (log message trimmed)
[21:11] <CIA-41> ffmpeg: avplay: Add an option for not limiting the input buffer size
[21:11] <CIA-41> ffmpeg: For reading from normal files on disk, the queue limits for
[21:11] <CIA-41> ffmpeg: demuxed data work fine, but for reading data from realtime
[21:11] <CIA-41> ffmpeg: streams, they mean we're not reading from the input stream
[21:11] <CIA-41> ffmpeg: at all once the queue limit has been reached. For TCP streams,
[21:11] <CIA-41> ffmpeg: this means that writing to the socket from the peer side blocks
[21:11] <CIA-41> ffmpeg: 03Samuel Pitoiset 07master * rbc3dbcc8e5 10ffmpeg/ (6 files in 3 dirs): 
[21:11] <CIA-41> ffmpeg: Add Blowfish encryption support in libavutil
[21:11] <CIA-41> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[21:11] <CIA-41> ffmpeg: 03Ronald S. Bultje 07master * r1ce1e6360d 10ffmpeg/libavformat/mov.c: 
[21:11] <CIA-41> ffmpeg: mov: use forward declaration of a function instead of a table.
[21:11] <CIA-41> ffmpeg: MSVC gives a compile error on the tentative definition of
[21:11] <CIA-41> ffmpeg: mov_default_parse_table[].
[21:11] <CIA-41> ffmpeg: 03Diego Biurrun 07master * r2047e40e6e 10ffmpeg/ (7 files in 7 dirs): Clarify Doxygen comment for FF_API_* #defines.
[21:11] <CIA-41> ffmpeg: 03Ronald S. Bultje 07master * r723b266d72 10ffmpeg/libavcodec/x86/dsputilenc_mmx.c: dsputilenc: group yasm and inline asm function pointer assignment.
[21:11] <CIA-41> ffmpeg: 03Mans Rullgard 07master * r7c29377b70 10ffmpeg/ (5 files in 4 dirs): 
[21:11] <CIA-41> ffmpeg: lavfi: reclassify showfiltfmts as a TESTPROG
[21:11] <CIA-41> ffmpeg: This tool uses lavfi internal symbols not accessible in shared
[21:11] <CIA-41> ffmpeg: libraries. TESTPROGS are linked statically to allow them use of
[21:11] <CIA-41> ffmpeg: library internals not normally exported.
[21:11] <CIA-41> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[21:11] <CIA-41> ffmpeg: 03Diego Biurrun 07master * rfa2d747352 10ffmpeg/tools/graph2dot.c: 
[21:11] <CIA-41> ffmpeg: graph2dot: fix printf format specifier
[21:11] <CIA-41> ffmpeg: tools/graph2dot.c:85: warning: format %lld expects type long long int, but argument 4 has type int
[21:11] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r039e9fe01c 10ffmpeg/: (log message trimmed)
[21:12] <CIA-41> ffmpeg: Merge remote-tracking branch 'qatar/master'
[21:12] <CIA-41> (33 lines omitted)
[23:45] <michaelni> ubitux, do you want to review "movtextdec: MPEG4 Part 17 Timed Text Decoder." ?
[23:45] <ubitux> i am ATM
[23:46] <michaelni> ok thx
[23:48] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rfa36f33422 10ffmpeg/ (12 files in 4 dirs): 
[23:48] <CIA-41> ffmpeg: sws: support 12&14 bit planar colorspaces
[23:48] <CIA-41> ffmpeg: Reviewed-by: Paul B Mahol <onemda at gmail.com>
[23:48] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[23:48] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r1f29504204 10ffmpeg/libavutil/blowfish.h: 
[23:48] <CIA-41> ffmpeg: blowfish.h: include author attribution
[23:48] <CIA-41> ffmpeg: Reviewed-by: Clément BSsch <ubitux at gmail.com>
[23:48] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[23:48] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r0713c38ea3 10ffmpeg/libavutil/xtea.h: 
[23:48] <CIA-41> ffmpeg: xtea.h: include author attribution
[23:48] <CIA-41> ffmpeg: Reviewed-by: Clément BSsch <ubitux at gmail.com>
[23:48] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:00] --- Thu Jul  5 2012


More information about the Ffmpeg-devel-irc mailing list