platform/upstream/libav.git
12 years agolavc: convert error_recognition to err_recognition.
Dustin Brody [Wed, 7 Dec 2011 13:51:36 +0000 (08:51 -0500)]
lavc: convert error_recognition to err_recognition.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoavconv: update -map option help text
Janne Grunau [Mon, 12 Dec 2011 11:45:32 +0000 (12:45 +0100)]
avconv: update -map option help text

12 years agox86: Require 7 registers for the cabac asm
Martin Storsjö [Mon, 12 Dec 2011 12:48:53 +0000 (14:48 +0200)]
x86: Require 7 registers for the cabac asm

The change in 599b4c6ef didn't turn out to work properly on
i386 on OS X, where it broke building with PIC enabled.

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agox86: bswap: remove test for bswap instruction
Mans Rullgard [Mon, 12 Dec 2011 01:33:17 +0000 (01:33 +0000)]
x86: bswap: remove test for bswap instruction

Firstly, this test never worked as intended, always reporting
success.  Secondly, bswap is available from 486 onward and can
thus be assumed present.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agobswap: make generic implementation more compiler-friendly
Mans Rullgard [Mon, 12 Dec 2011 00:50:08 +0000 (00:50 +0000)]
bswap: make generic implementation more compiler-friendly

With these changes, gcc 4.5 and later recognise it as a bswap
and use the proper instructions on ARM and x86.  On x86, the
16-bit bswap is recognised from gcc 4.1.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agoh264: remove useless cast
Mans Rullgard [Sun, 11 Dec 2011 23:51:12 +0000 (23:51 +0000)]
h264: remove useless cast

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agoproresdec: fix decode_slice() prototype
Mans Rullgard [Sun, 11 Dec 2011 23:36:05 +0000 (23:36 +0000)]
proresdec: fix decode_slice() prototype

Make the function prototype match the argument of
AVCodecCntext.execute() and remove the cast hiding
this mismatch.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agox86: cabac: replace explicit memory references with "m" operands
Mans Rullgard [Sun, 11 Dec 2011 21:41:59 +0000 (21:41 +0000)]
x86: cabac: replace explicit memory references with "m" operands

This replaces the explicit offset(reg) memory references with
"m" operands for the same locations.  As a result, one fewer
register operand is needed for these inline asm statements.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agoavplay: don't request a stereo downmix
Justin Ruggles [Sat, 10 Dec 2011 21:59:55 +0000 (16:59 -0500)]
avplay: don't request a stereo downmix

Multi-channel output works just fine for me with SDL 1.2.14.

12 years agowmapro: use av_float2int()
Mans Rullgard [Sat, 10 Dec 2011 15:51:25 +0000 (15:51 +0000)]
wmapro: use av_float2int()

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agolavc: avoid invalid memcpy() in avcodec_default_release_buffer()
Mans Rullgard [Mon, 5 Dec 2011 15:38:10 +0000 (15:38 +0000)]
lavc: avoid invalid memcpy() in avcodec_default_release_buffer()

When the buf and last pointers are equal, the FFSWAP() results
in an invalid call to memcpy() with same source and destination
on some targets.  Although assigning a struct to itself is valid
C99, gcc does not check for this before calling memcpy().
See http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32667

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agolavu: replace int/float punning functions
Mans Rullgard [Sun, 27 Nov 2011 14:04:16 +0000 (14:04 +0000)]
lavu: replace int/float punning functions

The existing functions defined in intfloat_readwrite.[ch] are
both slow and incorrect (infinities are not handled).

This introduces a new header with fast, inline conversion
functions using direct union punning assuming an IEEE-754
system, an assumption already made throughout the code.

The one use of Intel/Motorola extended 80-bit format is
replaced by simpler code sufficient under the present
constraints (positive normal values).

The old functions are marked deprecated and retained for
compatibility.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agolavfi: install libavfilter/vsrc_buffer.h
Andrey Utkin [Thu, 8 Dec 2011 19:27:01 +0000 (21:27 +0200)]
lavfi: install libavfilter/vsrc_buffer.h

It declares API necessary for filtering apps.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoRemove extraneous semicolons
Mans Rullgard [Sun, 11 Dec 2011 17:07:04 +0000 (17:07 +0000)]
Remove extraneous semicolons

These semicolons cause invalid empty top-level declarations.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agosdp: Restore the original mp4 format h264 extradata if converted
Martin Storsjö [Fri, 9 Dec 2011 19:00:09 +0000 (21:00 +0200)]
sdp: Restore the original mp4 format h264 extradata if converted

If the sdp is generated before the rtp muxer is initialized
(e.g. as when called from the rtsp muxer), this has to be done,
otherwise the rtp muxer doesn't know that the input really is
in mp4 format.

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agortpenc: Add support for mp4 format h264
Luca Abeni [Fri, 9 Dec 2011 09:02:55 +0000 (11:02 +0200)]
rtpenc: Add support for mp4 format h264

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agortpenc: Simplify code by introducing a separate end pointer
Martin Storsjö [Fri, 9 Dec 2011 18:42:52 +0000 (20:42 +0200)]
rtpenc: Simplify code by introducing a separate end pointer

Also remove redundant parentheses.

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agomovenc: Use the actual converted sample for RTP hinting
Martin Storsjö [Fri, 9 Dec 2011 19:19:57 +0000 (21:19 +0200)]
movenc: Use the actual converted sample for RTP hinting

If an annex b bitstream is muxed into mov, the actual written
sample is reformatted to mp4 syntax before writing.

Currently, the RTP hints that copy data from the normal video
track, where the payload data might be offset compared to the
original sample that the RTP hinting used (when 3 byte
annex b startcodes have been converted into 4 byte mp4 format
startcodes).

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agoFix a bunch of common typos.
Diego Biurrun [Wed, 5 Oct 2011 12:12:42 +0000 (14:12 +0200)]
Fix a bunch of common typos.

12 years agoWarn about avserver being broken.
Anton Khirnov [Sat, 10 Dec 2011 17:30:03 +0000 (18:30 +0100)]
Warn about avserver being broken.

Also remove mentions of it from other avtools' manuals.

12 years agoavconv: drop code for special handling of avserver streams.
Anton Khirnov [Sat, 10 Dec 2011 16:33:52 +0000 (17:33 +0100)]
avconv: drop code for special handling of avserver streams.

It's broken and doesn't work anyway.

This patch means that avconv will ignore encoding options from the ffm
file and will instead use whatever is provided on the commandline as for
normal output.

12 years agorawdec: don't set codec timebase.
Anton Khirnov [Sat, 10 Dec 2011 11:40:05 +0000 (12:40 +0100)]
rawdec: don't set codec timebase.

It's not supposed to be set outside of lavc. Set r_frame_rate and
avg_frame_rate instead.

12 years agolavf doxy: add muxing stuff to lavf_encoding group
Anton Khirnov [Sat, 10 Dec 2011 20:04:30 +0000 (21:04 +0100)]
lavf doxy: add muxing stuff to lavf_encoding group

12 years agolavf doxy: add demuxing stuff to lavf_decoding group
Anton Khirnov [Sat, 10 Dec 2011 20:04:30 +0000 (21:04 +0100)]
lavf doxy: add demuxing stuff to lavf_decoding group

12 years agolavf doxy: expand/reword metadata API doxy.
Anton Khirnov [Sat, 10 Dec 2011 19:39:39 +0000 (20:39 +0100)]
lavf doxy: expand/reword metadata API doxy.

12 years agolavf doxy: add installed headers to groups.
Anton Khirnov [Sat, 10 Dec 2011 19:16:57 +0000 (20:16 +0100)]
lavf doxy: add installed headers to groups.

12 years agolavf doxy: add avio groups into the lavf_io group.
Anton Khirnov [Sat, 10 Dec 2011 19:11:28 +0000 (20:11 +0100)]
lavf doxy: add avio groups into the lavf_io group.

12 years agolavf doxy: rename lavf I/O group to lavf_io.
Anton Khirnov [Sat, 10 Dec 2011 19:09:04 +0000 (20:09 +0100)]
lavf doxy: rename lavf I/O group to lavf_io.

12 years agolavf doxy: add metadata docs to the main lavf group
Anton Khirnov [Sat, 10 Dec 2011 19:06:49 +0000 (20:06 +0100)]
lavf doxy: add metadata docs to the main lavf group

12 years agottadec: check channel count as read from extradata.
Shitiz Garg [Sat, 10 Dec 2011 19:40:57 +0000 (01:10 +0530)]
ttadec: check channel count as read from extradata.

fixes floating-point exception due to channels being set to 0.
fixes Bug 128.

Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
12 years agoAdd CLJR encoding and decoding regression tests
Paul B Mahol [Thu, 8 Dec 2011 23:42:02 +0000 (23:42 +0000)]
Add CLJR encoding and decoding regression tests

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agocljr: remove unused code
Paul B Mahol [Fri, 9 Dec 2011 00:45:27 +0000 (00:45 +0000)]
cljr: remove unused code

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agoflacdec: Support for tracks in cuesheet metadata block
Paul B Mahol [Sat, 10 Dec 2011 04:53:30 +0000 (04:53 +0000)]
flacdec: Support for tracks in cuesheet metadata block

Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
12 years agoptx: fix inverted check for sufficient data
Janne Grunau [Sat, 10 Dec 2011 15:57:32 +0000 (16:57 +0100)]
ptx: fix inverted check for sufficient data

Fix regression introduced in 2b53e69.

12 years agoflac muxer: fix writing of file header and STREAMINFO header from extradata
Justin Ruggles [Sat, 10 Dec 2011 03:51:30 +0000 (22:51 -0500)]
flac muxer: fix writing of file header and STREAMINFO header from extradata

fixes Bug 119

12 years agoptx: emit a warning on insufficient picture data
Janne Grunau [Sat, 10 Dec 2011 11:55:08 +0000 (12:55 +0100)]
ptx: emit a warning on insufficient picture data

Return the whole packet as consumed in this case and not the size the
packet should have had. Move the insufficient data check into the for
condition to fix a ISO C90 error on bigendian.

12 years agoutvideo: add fate tests covering all codec variants
Aneesh Dogra [Fri, 9 Dec 2011 00:31:33 +0000 (00:31 +0000)]
utvideo: add fate tests covering all codec variants

Signed-off-by: Janne Grunau <janne-libav@jannau.net>
12 years agodoc: update to refer to avconv
Luca Barbato [Sat, 10 Dec 2011 00:58:04 +0000 (01:58 +0100)]
doc: update to refer to avconv

12 years agodoc: remove some stale entries from the faq
Luca Barbato [Sat, 10 Dec 2011 00:52:05 +0000 (01:52 +0100)]
doc: remove some stale entries from the faq

Some entries should be still amended.

12 years agoDxtory capture format decoder
Kostya Shishkov [Fri, 9 Dec 2011 10:06:02 +0000 (11:06 +0100)]
Dxtory capture format decoder

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agoisom: sort and pretty-print codec_movaudio_tags[]
Justin Ruggles [Tue, 6 Dec 2011 23:02:14 +0000 (18:02 -0500)]
isom: sort and pretty-print codec_movaudio_tags[]

original order is retained within the same tag or codec id to maintain the
muxing/demuxing priority.

12 years agoisom: remove pointless comments in codec_movaudio_tags[]
Justin Ruggles [Tue, 6 Dec 2011 22:54:07 +0000 (17:54 -0500)]
isom: remove pointless comments in codec_movaudio_tags[]

12 years agoisom: remove commented-out tag for vorbis
Justin Ruggles [Tue, 6 Dec 2011 22:51:54 +0000 (17:51 -0500)]
isom: remove commented-out tag for vorbis

12 years agomovenc: write 'chan' tag for AC-3 in MOV
Justin Ruggles [Tue, 6 Dec 2011 22:35:37 +0000 (17:35 -0500)]
movenc: write 'chan' tag for AC-3 in MOV

12 years agomov: add support for reading and writing the 'chan' tag
Justin Ruggles [Sat, 3 Dec 2011 18:32:45 +0000 (13:32 -0500)]
mov: add support for reading and writing the 'chan' tag

This implements reading the tag in the demuxer and adds support for writing it
in the muxer. Some example channel layout tables for muxing are included for
ac3, aac, and alac, but they are not utilized yet.

12 years agoaudioconvert: add some additional channel and channel layout macros
Justin Ruggles [Tue, 6 Dec 2011 20:53:02 +0000 (15:53 -0500)]
audioconvert: add some additional channel and channel layout macros

12 years agoaudioconvert: change 7.1 "wide" layout to use side surround channels
Justin Ruggles [Tue, 6 Dec 2011 20:38:55 +0000 (15:38 -0500)]
audioconvert: change 7.1 "wide" layout to use side surround channels

12 years agomovenc: simplify handling of pcm vs. adpcm vs. other compressed codecs
Justin Ruggles [Fri, 2 Dec 2011 20:51:52 +0000 (15:51 -0500)]
movenc: simplify handling of pcm vs. adpcm vs. other compressed codecs

Use Sound Sample Description Version 2 for all MOV files.
Updated FATE references accordingly.
Note that ADPCM is treated as compressed audio in version 2.

12 years agodoc: update documentation to use avconv
Luca Barbato [Thu, 8 Dec 2011 16:39:28 +0000 (17:39 +0100)]
doc: update documentation to use avconv

12 years agodoc: update demuxers section
Luca Barbato [Thu, 8 Dec 2011 16:20:17 +0000 (17:20 +0100)]
doc: update demuxers section

Cleanup tags usage and update commandline examples.

12 years agodoc: extend external library coverage
Luca Barbato [Sat, 3 Dec 2011 20:50:19 +0000 (21:50 +0100)]
doc: extend external library coverage

12 years agodoc: split platform specific information
Luca Barbato [Sat, 3 Dec 2011 20:25:04 +0000 (21:25 +0100)]
doc: split platform specific information

Move the platform-specific information in its own file.

12 years agodoc: port the git-howto to texinfo
Luca Barbato [Sat, 3 Dec 2011 17:06:14 +0000 (18:06 +0100)]
doc: port the git-howto to texinfo

12 years agodoc: provide fallback css and customize @float
Luca Barbato [Sat, 3 Dec 2011 22:24:50 +0000 (23:24 +0100)]
doc: provide fallback css and customize @float

Add "note" or "important" class to the generated div.

12 years agodoc: document fate in a texinfo
Luca Barbato [Fri, 2 Dec 2011 00:33:07 +0000 (01:33 +0100)]
doc: document fate in a texinfo

Summarize the information provided in the wiki and the one provided
in fate.txt

12 years agodoxy: change hue value to match our green
Luca Barbato [Thu, 8 Dec 2011 13:29:26 +0000 (14:29 +0100)]
doxy: change hue value to match our green

12 years agocljr: K&R cosmetics
Diego Biurrun [Thu, 8 Dec 2011 14:51:38 +0000 (15:51 +0100)]
cljr: K&R cosmetics

12 years agocljr: return a more sensible value when encountering invalid headers
Diego Biurrun [Thu, 8 Dec 2011 14:53:27 +0000 (15:53 +0100)]
cljr: return a more sensible value when encountering invalid headers

12 years agocljr: drop unnecessary emms_c() calls without MMX code
Diego Biurrun [Thu, 8 Dec 2011 14:38:49 +0000 (15:38 +0100)]
cljr: drop unnecessary emms_c() calls without MMX code

12 years agocljr: remove useless casts
Mans Rullgard [Thu, 8 Dec 2011 14:50:48 +0000 (14:50 +0000)]
cljr: remove useless casts

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agocljr: group encode/decode parts under single ifdefs
Mans Rullgard [Thu, 8 Dec 2011 14:48:12 +0000 (14:48 +0000)]
cljr: group encode/decode parts under single ifdefs

This groups the encode/decode parts under single ifdefs and
eliminates the encode_init() function as it merely calls
common_init().  Also fix whitespace in moved code.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agocljr: remove stray semicolon
Mans Rullgard [Thu, 8 Dec 2011 14:45:36 +0000 (14:45 +0000)]
cljr: remove stray semicolon

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agocljr: add missing return statement in decode_end()
Paul B Mahol [Thu, 8 Dec 2011 12:39:47 +0000 (12:39 +0000)]
cljr: add missing return statement in decode_end()

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agodoc: add pulseaudio to the input list
Luca Barbato [Sat, 3 Dec 2011 20:53:09 +0000 (21:53 +0100)]
doc: add pulseaudio to the input list

12 years agoavconv: remove unsubstantiated comment
Luca Barbato [Thu, 8 Dec 2011 01:17:18 +0000 (02:17 +0100)]
avconv: remove unsubstantiated comment

All the currently supported OS-es manage to use return values
properly.

12 years agoshorten: avoid abort() on unknown audio types
John Brooks [Thu, 8 Dec 2011 09:16:21 +0000 (09:16 +0000)]
shorten: avoid abort() on unknown audio types

Signed-off-by: Janne Grunau <janne-libav@jannau.net>
12 years agocljr: add encoder
Paul B Mahol [Thu, 8 Dec 2011 00:05:23 +0000 (00:05 +0000)]
cljr: add encoder

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agobuild: merge lists of HTML documentation targets
Diego Biurrun [Wed, 7 Dec 2011 20:49:53 +0000 (21:49 +0100)]
build: merge lists of HTML documentation targets

12 years agotests/examples: Mark some variables only used within their files as static.
Diego Biurrun [Wed, 21 Sep 2011 16:44:41 +0000 (18:44 +0200)]
tests/examples: Mark some variables only used within their files as static.

12 years agotests/tools/examples: Replace direct exit() calls by return.
Diego Biurrun [Wed, 21 Sep 2011 16:46:56 +0000 (18:46 +0200)]
tests/tools/examples: Replace direct exit() calls by return.

12 years agox86 cpuid: set vendor union members separately
Sean McGovern [Wed, 7 Dec 2011 21:51:28 +0000 (21:51 +0000)]
x86 cpuid: set vendor union members separately

Solaris Studio (suncc) has difficulty with filling in
members of a union. Instead, let's retrieve and store the
cpuid() results separately. This is still a compiler bug,
however this fix does not cause a regression on other platforms.

Signed-off-by: Janne Grunau <janne-libav@jannau.net>
12 years agocljr: release picture at end of decoding
Paul B. Mahol [Tue, 6 Dec 2011 04:14:55 +0000 (04:14 +0000)]
cljr: release picture at end of decoding

Otherwise after transcoding from cljr we get: Found 1 unreleased buffers!

Signed-off-by: Janne Grunau <janne-libav@jannau.net>
12 years agorv40: NEON optimised rv40 qpel motion compensation
Mans Rullgard [Wed, 7 Dec 2011 22:28:00 +0000 (22:28 +0000)]
rv40: NEON optimised rv40 qpel motion compensation

Based on patch by Janne Grunau.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agodrawtext: remove typo
Luca Barbato [Wed, 7 Dec 2011 19:30:55 +0000 (20:30 +0100)]
drawtext: remove typo

It slipped through the last two iterations.

12 years agopcm-mpeg: implement new audio decoding api
Hendrik Leppkes [Sun, 4 Dec 2011 22:15:51 +0000 (23:15 +0100)]
pcm-mpeg: implement new audio decoding api

Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
12 years agow32thread: port fixes to pthread_cond_broadcast() from x264.
Ronald S. Bultje [Wed, 7 Dec 2011 04:42:55 +0000 (20:42 -0800)]
w32thread: port fixes to pthread_cond_broadcast() from x264.

12 years agodoc: add editor configuration section with Vim and Emacs settings
Diego Biurrun [Mon, 5 Dec 2011 12:18:27 +0000 (13:18 +0100)]
doc: add editor configuration section with Vim and Emacs settings

based on a patch by Victor Vasiliev, vasilvv gmail com

12 years agodxva2.h: include d3d9.h to define LPDIRECT3DSURFACE9
Rafaël Carré [Wed, 7 Dec 2011 15:23:52 +0000 (10:23 -0500)]
dxva2.h: include d3d9.h to define LPDIRECT3DSURFACE9

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
12 years agoavformat/utils: Drop unused goto label.
Diego Biurrun [Wed, 7 Dec 2011 10:30:07 +0000 (11:30 +0100)]
avformat/utils: Drop unused goto label.

libavformat/utils.c:2165:2: warning: label ‘fail’ defined but not used

12 years agodoxygen: Replace '\' by '@' in Doxygen markup tags.
Diego Biurrun [Wed, 7 Dec 2011 12:43:36 +0000 (13:43 +0100)]
doxygen: Replace '\' by '@' in Doxygen markup tags.

12 years agocosmetics: drop some completely pointless parentheses
Diego Biurrun [Wed, 7 Dec 2011 12:48:58 +0000 (13:48 +0100)]
cosmetics: drop some completely pointless parentheses

12 years ago cljr: simplify CLJRContext
Paul B. Mahol [Tue, 6 Dec 2011 17:45:37 +0000 (17:45 +0000)]
 cljr: simplify CLJRContext

There is no need to have delta, offset and gb in CLJRContext.

Signed-off-by: Janne Grunau <janne-libav@jannau.net>
12 years agodrawtext: introduce rand(min, max)
Luca Barbato [Sun, 4 Dec 2011 23:56:21 +0000 (00:56 +0100)]
drawtext: introduce rand(min, max)

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
12 years agodrawtext: introduce explicit draw/hide variable
Luca Barbato [Sun, 4 Dec 2011 20:13:56 +0000 (21:13 +0100)]
drawtext: introduce explicit draw/hide variable

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
12 years agortmp: Use nb_invokes for all invoke commands
Martin Storsjö [Mon, 5 Dec 2011 10:35:06 +0000 (12:35 +0200)]
rtmp: Use nb_invokes for all invoke commands

704af3e29c3ddbc22ac5c8f40e5a0f860d53ac4c broke publishing
of rtmp streams, at least publishing to Wowza servers.

This changes all invoke commands to use nb_invokes.

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agoCode cleanup - mpegvideo.c - 500-1000line
Konstantin Todorov [Mon, 5 Dec 2011 16:40:28 +0000 (18:40 +0200)]
Code cleanup - mpegvideo.c - 500-1000line

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
12 years agorv40: NEON optimised weighted prediction
Janne Grunau [Mon, 5 Dec 2011 21:22:57 +0000 (21:22 +0000)]
rv40: NEON optimised weighted prediction

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agorv40: NEON optimised chroma MC
Janne Grunau [Mon, 5 Dec 2011 21:18:05 +0000 (21:18 +0000)]
rv40: NEON optimised chroma MC

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agoARM: move NEON H264 chroma mc to a separate file
Mans Rullgard [Tue, 6 Dec 2011 12:44:05 +0000 (12:44 +0000)]
ARM: move NEON H264 chroma mc to a separate file

This allows sharing code with the rv40 version of these functions.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agorv34: NEON optimised inverse transform functions
Janne Grunau [Sat, 24 Sep 2011 11:05:55 +0000 (13:05 +0200)]
rv34: NEON optimised inverse transform functions

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agomov: Don't av_malloc(0).
Alex Converse [Mon, 5 Dec 2011 00:38:05 +0000 (16:38 -0800)]
mov: Don't av_malloc(0).

malloc() is allowed to return NULL when zero is the argument. This
causes us to think malloc has failed and return AVERROR(ENOMEM). In
addition OS X malloc() returns an unfreeable non-NULL pointer for size
zero when alignment is greater than 16.

12 years agoavconv: only allocate 1 AVFrame per input stream
Justin Ruggles [Mon, 5 Dec 2011 16:49:38 +0000 (11:49 -0500)]
avconv: only allocate 1 AVFrame per input stream

This avoids allocating/freeing an AVFrame for each packet decode. It also
simplifies error handling in transcode_audio() and transcode_video().

12 years agoavconv: fix memleaks due to not freeing the AVFrame for audio
Justin Ruggles [Mon, 5 Dec 2011 15:36:54 +0000 (10:36 -0500)]
avconv: fix memleaks due to not freeing the AVFrame for audio

12 years agoh264-fate: remove -strict 1 except where necessary (mr4/5-tandberg).
Ronald S. Bultje [Sat, 26 Nov 2011 21:10:17 +0000 (13:10 -0800)]
h264-fate: remove -strict 1 except where necessary (mr4/5-tandberg).

12 years agomisc Doxygen markup improvements
Diego Biurrun [Sun, 30 Oct 2011 18:10:50 +0000 (19:10 +0100)]
misc Doxygen markup improvements

12 years agodoxygen: eliminate Qt-style doxygen syntax
Diego Biurrun [Sun, 30 Oct 2011 17:27:33 +0000 (18:27 +0100)]
doxygen: eliminate Qt-style doxygen syntax

12 years agog722: Add a regression test for muxing/demuxing in wav
Martin Storsjö [Sat, 3 Dec 2011 22:58:31 +0000 (00:58 +0200)]
g722: Add a regression test for muxing/demuxing in wav

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agog722: Change bits per sample to 4
Sjoerd Simons [Thu, 1 Dec 2011 23:08:35 +0000 (01:08 +0200)]
g722: Change bits per sample to 4

Earlier, bits per sample was defined as 8, since
bits_per_coded_sample was used to indicate whether to ignore
the lower bits of the codeword, having values 6, 7 or 8.

g722 encodes 2 samples into one byte codeword, therefore the
bits per sample is 4. By changing this, the generated timestamps
for streams encoded with g722 become correct.

This makes timestamp generation for g722 data correct (both when
encoding and when demuxing from raw g722 files).

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agog722dec: Signal skipping the lower bits via AVOptions instead of bits_per_coded_sample
Martin Storsjö [Sat, 3 Dec 2011 17:54:20 +0000 (19:54 +0200)]
g722dec: Signal skipping the lower bits via AVOptions instead of bits_per_coded_sample

This avoids using bits_per_coded_sample for this information.
bits_per_coded_sample should be 4 for this codec normally,
since two samples are encoded into one 8 bit codeword.

In principle, this might be info that needs to be passed from
a demuxer, and in that case, a private AVOption isn't the best
choice, but no such samples are available at the moment, so
that use case is purely theoretical at the moment.

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agoapi-example: update to use avcodec_decode_audio4()
Justin Ruggles [Fri, 11 Nov 2011 20:44:41 +0000 (15:44 -0500)]
api-example: update to use avcodec_decode_audio4()