OSDN Git Service

coroid/ffmpeg_saccubus.git
12 years agoWindows Media Image decoder (WMVP/WVP2)
Alberto Delmás [Wed, 17 Aug 2011 12:24:42 +0000 (14:24 +0200)]
Windows Media Image decoder (WMVP/WVP2)

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoac3enc: remove outdated TODO comment for apply_channel_coupling()
Justin Ruggles [Mon, 22 Aug 2011 16:14:57 +0000 (12:14 -0400)]
ac3enc: remove outdated TODO comment for apply_channel_coupling()

12 years agolibx264: only use ABR mode when the user explicitly set bitrate.
Anton Khirnov [Sat, 20 Aug 2011 14:59:47 +0000 (16:59 +0200)]
libx264: only use ABR mode when the user explicitly set bitrate.

12 years agolibx264: use medium preset by default.
Anton Khirnov [Sat, 20 Aug 2011 14:47:36 +0000 (16:47 +0200)]
libx264: use medium preset by default.

12 years agomp2 encoder: make 128k the default bitrate.
Anton Khirnov [Sat, 20 Aug 2011 14:20:27 +0000 (16:20 +0200)]
mp2 encoder: make 128k the default bitrate.

12 years agomovenc: use libx264 by default when possible for mov, mp4 and psp
Anton Khirnov [Sat, 20 Aug 2011 14:14:58 +0000 (16:14 +0200)]
movenc: use libx264 by default when possible for mov, mp4 and psp

12 years agoavienc: saner default audio codec.
Anton Khirnov [Sat, 20 Aug 2011 14:05:17 +0000 (16:05 +0200)]
avienc: saner default audio codec.

libmp3lame if available, ac3 otherwise.

12 years agomatroskaenc: saner default codecs.
Anton Khirnov [Sat, 20 Aug 2011 14:02:22 +0000 (16:02 +0200)]
matroskaenc: saner default codecs.

libvorbis/libx264 for video if available, otherwise ac3/mpeg4.

12 years agoavplay: add examples of how to specify size/pixel format through private options
Stefano Sabatini [Sun, 21 Aug 2011 14:20:37 +0000 (16:20 +0200)]
avplay: add examples of how to specify size/pixel format through private options

Clarify with an example what the user should use instead of the
deprecated (and no longer working) -s/-pix_fmt options.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agolavc: add A|E|D flags to "ac" and "ar" options
Stefano Sabatini [Sun, 17 Jul 2011 13:28:42 +0000 (13:28 +0000)]
lavc: add A|E|D flags to "ac" and "ar" options

Fix setting of the corresponding AVCodecContext fields via
cmdutils.c:filter_codec_opts().
In particular, fix ffplay -ac and -ar options.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoWavPack demuxer: do not rely on index when timestamp is not in indexed range.
Kostya Shishkov [Sat, 20 Aug 2011 11:11:14 +0000 (13:11 +0200)]
WavPack demuxer: do not rely on index when timestamp is not in indexed range.

This fixes the situation when there are not enough entries in the index
(e.g. on initial seek there's only one index entry in the index) and index
search returns just the last known entry. That causes seeking function just to
seek there instead of trying harder to get at the requested position.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
12 years agoWavPack demuxer: store position of the first block in index.
Kostya Shishkov [Sat, 20 Aug 2011 16:14:58 +0000 (18:14 +0200)]
WavPack demuxer: store position of the first block in index.

Currently for multichannel audio position for the last block position is
stored in index (and used for seeking), which is obviously not correct.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
12 years agoWavPack decoder: implement flush function
Kostya Shishkov [Sat, 20 Aug 2011 16:14:44 +0000 (18:14 +0200)]
WavPack decoder: implement flush function

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
12 years agoavconv: Separate initialization from the main transcode loop.
Alex Converse [Fri, 19 Aug 2011 21:45:02 +0000 (14:45 -0700)]
avconv: Separate initialization from the main transcode loop.

12 years agodoc/avconv: replace forgotten av by avconv.
Anton Khirnov [Fri, 19 Aug 2011 15:52:49 +0000 (17:52 +0200)]
doc/avconv: replace forgotten av by avconv.

Signed-off-by: Alex Converse <alex.converse@gmail.com>
12 years agoavconv: reset input_ts_offset between files.
Anton Khirnov [Fri, 19 Aug 2011 15:52:48 +0000 (17:52 +0200)]
avconv: reset input_ts_offset between files.

Signed-off-by: Alex Converse <alex.converse@gmail.com>
12 years agodoc/avconv: extend -ss documentation.
Anton Khirnov [Fri, 19 Aug 2011 15:52:47 +0000 (17:52 +0200)]
doc/avconv: extend -ss documentation.

Signed-off-by: Alex Converse <alex.converse@gmail.com>
12 years agoavconv: call flush_encoders() from transcode() directly.
Anton Khirnov [Fri, 19 Aug 2011 15:52:46 +0000 (17:52 +0200)]
avconv: call flush_encoders() from transcode() directly.

And remove now pointless parameter.

Signed-off-by: Alex Converse <alex.converse@gmail.com>
12 years agoavconv: fix broken indentation.
Anton Khirnov [Fri, 19 Aug 2011 15:52:45 +0000 (17:52 +0200)]
avconv: fix broken indentation.

Signed-off-by: Alex Converse <alex.converse@gmail.com>
12 years agoavconv: rescue poor abused limit_filesize global.
Anton Khirnov [Fri, 19 Aug 2011 15:52:44 +0000 (17:52 +0200)]
avconv: rescue poor abused limit_filesize global.

Keep a per-OutputFile instance of it, thus making -fs work with multiple
output files.

Signed-off-by: Alex Converse <alex.converse@gmail.com>
12 years agoavconv: Set error code before before jumping to fail.
Alex Converse [Fri, 19 Aug 2011 18:26:21 +0000 (11:26 -0700)]
avconv: Set error code before before jumping to fail.

12 years agoh264: fix PCM intra-coded blocks in monochrome case
Jeff Downs [Tue, 5 Jul 2011 17:20:06 +0000 (13:20 -0400)]
h264: fix PCM intra-coded blocks in monochrome case

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agoh264: correct implicit weight table computation for long ref pics
Jeff Downs [Wed, 6 Jul 2011 15:54:36 +0000 (11:54 -0400)]
h264: correct implicit weight table computation for long ref pics

Correct computation of implicit weight tables when referencing pictures
that are marked for long reference.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agoh264: correct the check for invalid long term frame index in MMCO decode
Jeff Downs [Tue, 5 Jul 2011 18:21:54 +0000 (14:21 -0400)]
h264: correct the check for invalid long term frame index in MMCO decode

The current check on MMCO parameters prohibits a "max long term frame index
plus 1" of 16 (frame idx of 15) for the "set max long term frame index" MMCO.
Fix this off-by-one error to allow the full range of legal values.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agoavconv: Fix spelling errors.
Alex Converse [Thu, 18 Aug 2011 19:09:17 +0000 (12:09 -0700)]
avconv: Fix spelling errors.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
12 years agoaac: Only set sample rate and object type from ADTS if output hasn't been configured.
Alex Converse [Tue, 16 Aug 2011 18:04:03 +0000 (11:04 -0700)]
aac: Only set sample rate and object type from ADTS if output hasn't been configured.

Long term it would be nice to support error resilient reconfiguration
but right now setting this every frame does more harm than help.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
12 years agoaac: Set SBR and PS to unsignalled during headerless and ADTS initialization.
Alex Converse [Wed, 10 Aug 2011 23:37:32 +0000 (16:37 -0700)]
aac: Set SBR and PS to unsignalled during headerless and ADTS initialization.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
12 years agoaac: Only output configure if audio was found.
Alex Converse [Tue, 16 Aug 2011 18:03:26 +0000 (11:03 -0700)]
aac: Only output configure if audio was found.

Audio found is not triggered on a CCE because a CCE alone has no output.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
12 years agoavconv: save two levels of indentation in flush_encoders()
Anton Khirnov [Tue, 16 Aug 2011 09:19:20 +0000 (11:19 +0200)]
avconv: save two levels of indentation in flush_encoders()

By replacing
if (foo)
    <do stuff>

with
if (!foo)
    continue;
<do stuff>

12 years agoavconv: factor flushing encoders out of output_packet().
Anton Khirnov [Thu, 18 Aug 2011 06:55:29 +0000 (08:55 +0200)]
avconv: factor flushing encoders out of output_packet().

12 years agoavconv: factor out initializing input streams.
Anton Khirnov [Wed, 6 Jul 2011 06:49:07 +0000 (08:49 +0200)]
avconv: factor out initializing input streams.

12 years agoavconv: remove -intra option.
Anton Khirnov [Wed, 17 Aug 2011 17:26:49 +0000 (19:26 +0200)]
avconv: remove -intra option.

It's equivalent to -g 0.

12 years agoavconv: reset streamid_map between output files.
Anton Khirnov [Wed, 17 Aug 2011 17:11:54 +0000 (19:11 +0200)]
avconv: reset streamid_map between output files.

12 years agoavconv: make timer_start a local var in transcode().
Anton Khirnov [Wed, 17 Aug 2011 16:46:36 +0000 (18:46 +0200)]
avconv: make timer_start a local var in transcode().

12 years agoavconv: cosmetics, move OutputStream.
Anton Khirnov [Wed, 17 Aug 2011 16:40:40 +0000 (18:40 +0200)]
avconv: cosmetics, move OutputStream.

Allows us to get rid of forward InputStream declaration.

12 years agoavconv: remove two unused macros.
Anton Khirnov [Wed, 17 Aug 2011 16:35:43 +0000 (18:35 +0200)]
avconv: remove two unused macros.

12 years agoh264: hide reference frame errors unless requested
Dustin Brody [Wed, 17 Aug 2011 22:08:09 +0000 (18:08 -0400)]
h264: hide reference frame errors unless requested

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
12 years agoswscale: split hScale() function pointer into h[cy]Scale().
Ronald S. Bultje [Wed, 3 Aug 2011 18:25:01 +0000 (11:25 -0700)]
swscale: split hScale() function pointer into h[cy]Scale().

This allows using more specific implementations for chroma/luma, e.g.
we can make assumptions on filterSize being constant, thus avoiding
that test at runtime.

12 years agoMove clipd macros to x86util.asm.
Ronald S. Bultje [Wed, 17 Aug 2011 04:23:53 +0000 (21:23 -0700)]
Move clipd macros to x86util.asm.

This allows sharing them between multiple .asm files.

12 years agoavconv: reindent.
Anton Khirnov [Wed, 17 Aug 2011 10:08:41 +0000 (12:08 +0200)]
avconv: reindent.

12 years agoavconv: rescue poor abused start_time global.
Anton Khirnov [Wed, 17 Aug 2011 09:42:44 +0000 (11:42 +0200)]
avconv: rescue poor abused start_time global.

Keep a per-OutputFile instance of it, thus making -ss work with multiple
output files.

12 years agoavconv: rescue poor abused recording_time global.
Anton Khirnov [Wed, 17 Aug 2011 09:42:44 +0000 (11:42 +0200)]
avconv: rescue poor abused recording_time global.

Keep a per-OutputFile instance of it, thus making -t work with multiple
output files.

12 years agoavconv: merge two loops in output_packet().
Anton Khirnov [Wed, 17 Aug 2011 12:09:22 +0000 (14:09 +0200)]
avconv: merge two loops in output_packet().

12 years agoavconv: fix broken indentation.
Anton Khirnov [Wed, 17 Aug 2011 09:33:30 +0000 (11:33 +0200)]
avconv: fix broken indentation.

12 years agoavconv: get rid of the arbitrary MAX_FILES limit.
Anton Khirnov [Wed, 17 Aug 2011 08:27:17 +0000 (10:27 +0200)]
avconv: get rid of the arbitrary MAX_FILES limit.

12 years agoavconv: get rid of the output_streams_for_file vs. ost_table schizophrenia
Anton Khirnov [Wed, 17 Aug 2011 08:21:37 +0000 (10:21 +0200)]
avconv: get rid of the output_streams_for_file vs. ost_table schizophrenia

Instead store output streams in the same way as input streams.

12 years agoavconv: add a wrapper for output AVFormatContexts and merge output_opts into it
Anton Khirnov [Wed, 17 Aug 2011 07:56:08 +0000 (09:56 +0200)]
avconv: add a wrapper for output AVFormatContexts and merge output_opts into it

12 years agoavconv: make itsscale syntax consistent with other options.
Anton Khirnov [Wed, 17 Aug 2011 07:12:04 +0000 (09:12 +0200)]
avconv: make itsscale syntax consistent with other options.

Move the stream specifier to the option name.

12 years agoavconv: factor out adding input streams.
Anton Khirnov [Tue, 5 Jul 2011 17:38:50 +0000 (19:38 +0200)]
avconv: factor out adding input streams.

12 years agoavconv: Factorize combining auto vsync with format.
Alex Converse [Wed, 17 Aug 2011 02:06:46 +0000 (19:06 -0700)]
avconv: Factorize combining auto vsync with format.

12 years agoavconv: Factorize video resampling.
Alex Converse [Wed, 17 Aug 2011 01:43:27 +0000 (18:43 -0700)]
avconv: Factorize video resampling.

12 years agoavconv: Don't unnecessarily convert ipts to a double.
Alex Converse [Tue, 16 Aug 2011 21:41:16 +0000 (14:41 -0700)]
avconv: Don't unnecessarily convert ipts to a double.

12 years agoffmpeg: remove unsed variable nopts
Diego Biurrun [Wed, 17 Aug 2011 13:35:42 +0000 (15:35 +0200)]
ffmpeg: remove unsed variable nopts

12 years agoRV3/4 parser: remove unused variable 'off'
Kostya Shishkov [Wed, 17 Aug 2011 13:07:49 +0000 (15:07 +0200)]
RV3/4 parser: remove unused variable 'off'

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agoadd XMV demuxer
Sven Hesse [Tue, 16 Aug 2011 19:26:45 +0000 (21:26 +0200)]
add XMV demuxer

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agormdec: parse FPS in RealMedia properly
Kostya Shishkov [Tue, 16 Aug 2011 07:10:57 +0000 (09:10 +0200)]
rmdec: parse FPS in RealMedia properly

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoUse parsers for RealVideo 3/4 to determine correct PTS
Kostya Shishkov [Mon, 15 Aug 2011 10:03:40 +0000 (12:03 +0200)]
Use parsers for RealVideo 3/4 to determine correct PTS

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agormdec: correct DTS calculation in RealMedia container.
Kostya Shishkov [Thu, 11 Aug 2011 12:44:05 +0000 (14:44 +0200)]
rmdec: correct DTS calculation in RealMedia container.

First, container stores only DTS and not PTS as it was believed.
Second, multiple frames in a packet store timestamp instead of position
after the frame length.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agorv10/20: tell decoder to use edge emulation
Kostya Shishkov [Wed, 17 Aug 2011 08:36:33 +0000 (10:36 +0200)]
rv10/20: tell decoder to use edge emulation

This removes out-of-edge motion compensation artifacts (easily spotted green
blocks in avplay, gray blocks in transcoding), for example here:
http://samples.libav.org/samples/real/tv_watching_t1.rm

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agoRevert "avconv: use stream copy by default when possible."
Anton Khirnov [Tue, 16 Aug 2011 18:59:14 +0000 (20:59 +0200)]
Revert "avconv: use stream copy by default when possible."

This reverts commit 4f4f33844a7369e5579912cc02f3c5376b1e6872.

This commit has some ugly corner cases and needs to be discussed
further.

12 years agoavconv: print stream copy information.
Anton Khirnov [Tue, 16 Aug 2011 04:58:52 +0000 (06:58 +0200)]
avconv: print stream copy information.

12 years agoavconv: use stream copy by default when possible.
Anton Khirnov [Sat, 13 Aug 2011 06:37:13 +0000 (08:37 +0200)]
avconv: use stream copy by default when possible.

12 years agomatroskaenc: vertical alignment.
Anton Khirnov [Tue, 16 Aug 2011 04:52:50 +0000 (06:52 +0200)]
matroskaenc: vertical alignment.

12 years agomatroskaenc: implement query_codec()
Anton Khirnov [Thu, 11 Aug 2011 18:35:46 +0000 (20:35 +0200)]
matroskaenc: implement query_codec()

12 years agolavf: add avformat_query_codec().
Anton Khirnov [Thu, 11 Aug 2011 18:34:45 +0000 (20:34 +0200)]
lavf: add avformat_query_codec().

It allows to check if a given codec can be written into a container.

12 years agolavc: add avcodec_get_type() for mapping codec_id -> type.
Anton Khirnov [Thu, 11 Aug 2011 15:45:50 +0000 (17:45 +0200)]
lavc: add avcodec_get_type() for mapping codec_id -> type.

12 years agoflvenc: use int64_t to store offsets
Luca Barbato [Wed, 8 Jun 2011 14:32:07 +0000 (14:32 +0000)]
flvenc: use int64_t to store offsets

Metadata currently is written only at the start of the file in normal
cases, when transcoding from a rtmp source metadata could be
written later and the offset recorded can exceed 32bit.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoavconv: don't segfault on 0 input files.
Anton Khirnov [Tue, 16 Aug 2011 09:32:57 +0000 (11:32 +0200)]
avconv: don't segfault on 0 input files.

12 years agoDo not write ID3v1 tags by default
Diogo Franco [Tue, 16 Aug 2011 09:43:23 +0000 (09:43 +0000)]
Do not write ID3v1 tags by default

ID3v1 are legacy tags with several limitations; furthermore
avconv/ffmpeg writes the tags in UTF-8 which probably has near-0
software support.

Add a -write_id3v1 option to be able to turn it on; disabled by default.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agompegts: log into an AVFormatContext rather than MpegTSContext.
Alex Converse [Tue, 16 Aug 2011 01:14:12 +0000 (18:14 -0700)]
mpegts: log into an AVFormatContext rather than MpegTSContext.

MpegTSContext's AVClass member can be NULL.

12 years agomovenc: change AV_LOG_INFO to AV_LOG_WARNING for some warnings
Anton Khirnov [Thu, 11 Aug 2011 14:33:03 +0000 (16:33 +0200)]
movenc: change AV_LOG_INFO to AV_LOG_WARNING for some warnings

12 years agoavconv: remove a write-only variable
Anton Khirnov [Mon, 15 Aug 2011 16:26:56 +0000 (18:26 +0200)]
avconv: remove a write-only variable

12 years agoavconv: remove pointless parameter from new_*_stream().
Anton Khirnov [Sat, 13 Aug 2011 06:23:44 +0000 (08:23 +0200)]
avconv: remove pointless parameter from new_*_stream().

New streams are always added to the last output file.

12 years agoavconv: cosmetics, move code
Anton Khirnov [Sat, 13 Aug 2011 06:08:13 +0000 (08:08 +0200)]
avconv: cosmetics, move code

Move find_codec_or_die(), choose_codec(), new_output_stream() and
read_avserver_streams() below transcode() where they belong.

12 years agoavconv: cosmetics -- move copy_chapters().
Anton Khirnov [Sat, 13 Aug 2011 05:56:38 +0000 (07:56 +0200)]
avconv: cosmetics -- move copy_chapters().

It's only used in opt_output_file(), so move it right above
opt_output_file().

12 years agoavconv: cosmetics -- move parse_forced_key_frames().
Anton Khirnov [Sat, 13 Aug 2011 05:55:04 +0000 (07:55 +0200)]
avconv: cosmetics -- move parse_forced_key_frames().

parse_forced_key_frames() is only used in new_video_stream(), so move it
right above new_video_stream().

12 years agolavc: add audio flag to the 'b' option, deprecate 'ab'.
Anton Khirnov [Tue, 9 Aug 2011 11:56:36 +0000 (13:56 +0200)]
lavc: add audio flag to the 'b' option, deprecate 'ab'.

Its only reason for its existence was ffmpeg's inability to properly assign
AVOptions to streams. Now this is not a problem anymore, so 'ab' should
go.

12 years agoavconv: rename sameq to same_quant
Anton Khirnov [Tue, 9 Aug 2011 09:32:16 +0000 (11:32 +0200)]
avconv: rename sameq to same_quant

It's often mistakenly used as 'same quality', emphasize that it's not
true in the manual.

12 years agodoc/avconv: add forgotten end of chapter.
Anton Khirnov [Tue, 9 Aug 2011 09:05:04 +0000 (11:05 +0200)]
doc/avconv: add forgotten end of chapter.

12 years agoChangelog: document avconv incompatibilities with ffmpeg.
Anton Khirnov [Fri, 29 Jul 2011 12:17:01 +0000 (14:17 +0200)]
Changelog: document avconv incompatibilities with ffmpeg.

12 years agoavconv: replace -vcodec/-acodec/-scodec with a better system.
Anton Khirnov [Mon, 8 Aug 2011 12:29:37 +0000 (14:29 +0200)]
avconv: replace -vcodec/-acodec/-scodec with a better system.

The new option doesn't depend on its placement wrt -new* options (which
don't exist anymore) and works in a similar way as per-stream AVOptions.

-[vas]codec remain as aliases to -codec:[vas]

12 years agoavconv: remove presets.
Anton Khirnov [Fri, 12 Aug 2011 17:19:19 +0000 (19:19 +0200)]
avconv: remove presets.

Nothing except x264 uses them and native x264 presets should be used
with private options.

12 years agosvq3: propagate codec memory allocation failure in context init
Dustin Brody [Fri, 12 Aug 2011 16:05:46 +0000 (12:05 -0400)]
svq3: propagate codec memory allocation failure in context init

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoFix NASM include directive
Dave Yeo [Sun, 14 Aug 2011 18:46:18 +0000 (11:46 -0700)]
Fix NASM include directive

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
12 years agodsputil_mmx: Honor HAVE_AMD3DNOW
Alex Converse [Sun, 14 Aug 2011 03:02:37 +0000 (20:02 -0700)]
dsputil_mmx: Honor HAVE_AMD3DNOW

12 years agolavf,lavd: remove all usage of AVFormatParameters from demuxers.
Anton Khirnov [Sun, 17 Jul 2011 05:45:33 +0000 (07:45 +0200)]
lavf,lavd: remove all usage of AVFormatParameters from demuxers.

AVFormatParameters are converted into corresponding private options in
av_open_input_file/stream() compat wrappers, so accessing them from
demuxers is redundant.

12 years agojack: add 'channels' private option.
Anton Khirnov [Sun, 17 Jul 2011 06:08:57 +0000 (08:08 +0200)]
jack: add 'channels' private option.

Get rid of AVFormatParameters usage.

12 years agoVC-1: fix reading of custom PAR.
Reimar Döffinger [Sat, 13 Aug 2011 09:58:18 +0000 (11:58 +0200)]
VC-1: fix reading of custom PAR.

Custom PAR num/denum are in 1-256 range.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agoRemove redundant and dubious video codec detection by its extradata
Kostya Shishkov [Sun, 14 Aug 2011 18:02:26 +0000 (20:02 +0200)]
Remove redundant and dubious video codec detection by its extradata

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agompeg12: remove repeat-field code disabled since May 2002
Dustin Brody [Fri, 12 Aug 2011 05:30:20 +0000 (01:30 -0400)]
mpeg12: remove repeat-field code disabled since May 2002

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agopatch checklist: suggest fate instead of regression tests
Michael Niedermayer [Sat, 9 Apr 2011 10:50:03 +0000 (10:50 +0000)]
patch checklist: suggest fate instead of regression tests

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoTurn on resampling on sudden size change instead of bailing out during recode.
Kostya [Sat, 13 Aug 2011 16:44:21 +0000 (16:44 +0000)]
Turn on resampling on sudden size change instead of bailing out during recode.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoavtools: reinitialise filter chain when input video stream changes dimensions
Kostya [Sat, 13 Aug 2011 16:44:04 +0000 (16:44 +0000)]
avtools: reinitialise filter chain when input video stream changes dimensions

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoaacenc: Fix number of coefficients used in a LFE channel.
Nathan Caldwell [Fri, 29 Jul 2011 20:49:04 +0000 (14:49 -0600)]
aacenc: Fix number of coefficients used in a LFE channel.

The spec states:

 *  Only the lowest 12 spectral coefficients of any LFE may be non-zero

We were using the 12 lowest *bands*.

12 years agoaacenc: Fix a segfault with grouped psymodel.
Nathan Caldwell [Sat, 23 Jul 2011 07:19:34 +0000 (01:19 -0600)]
aacenc: Fix a segfault with grouped psymodel.

10l: Forgot about TYPE_LFE being after TYPE_CPE. Which causes a
segfault when encoding 5.1

12 years agoswscale: add dithering to yuv2yuvX_altivec_real
Luca Barbato [Wed, 10 Aug 2011 00:16:26 +0000 (02:16 +0200)]
swscale: add dithering to yuv2yuvX_altivec_real

It just does that part in scalar form, I doubt using a vector store
over 2 array would speed it up particularly.

The function should be written to not use a scratch buffer.

12 years agorv34: free+allocate buffer instead of reallocating it to preserve alignment
Kostya Shishkov [Fri, 12 Aug 2011 19:50:28 +0000 (21:50 +0200)]
rv34: free+allocate buffer instead of reallocating it to preserve alignment

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
12 years agoh264: add missing brackets.
Ronald S. Bultje [Fri, 12 Aug 2011 19:32:31 +0000 (12:32 -0700)]
h264: add missing brackets.

This caused failure of a few fate tests.

12 years agoswscale: use 15-bit intermediates for 9/10-bit scaling.
Ronald S. Bultje [Tue, 2 Aug 2011 22:42:35 +0000 (15:42 -0700)]
swscale: use 15-bit intermediates for 9/10-bit scaling.

12 years agoMove x86util.asm from libavcodec/ to libavutil/.
Ronald S. Bultje [Wed, 3 Aug 2011 16:48:08 +0000 (09:48 -0700)]
Move x86util.asm from libavcodec/ to libavutil/.

This allows using it in swscale also.