OSDN Git Service

android-x86/external-ffmpeg.git
12 years agolavf doxy: clarify that an AVPacket contains encoded data.
Anton Khirnov [Tue, 3 Jan 2012 06:41:14 +0000 (07:41 +0100)]
lavf doxy: clarify that an AVPacket contains encoded data.

12 years agompegtsenc: allow user triggered PES packet flushing
Jindrich Makovicka [Wed, 21 Mar 2012 11:35:15 +0000 (12:35 +0100)]
mpegtsenc: allow user triggered PES packet flushing

Signed-off-by: Jindrich Makovicka <jindrich.makovicka@nangu.tv>
Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agoAPIchanges: mark the place where 0.7 was cut.
Anton Khirnov [Tue, 20 Mar 2012 06:58:00 +0000 (07:58 +0100)]
APIchanges: mark the place where 0.7 was cut.

12 years agoAPIchanges: mark the place where 0.8 was cut.
Anton Khirnov [Tue, 20 Mar 2012 06:58:00 +0000 (07:58 +0100)]
APIchanges: mark the place where 0.8 was cut.

12 years agoAPIchanges: fill in missing dates and hashes.
Anton Khirnov [Tue, 20 Mar 2012 06:45:45 +0000 (07:45 +0100)]
APIchanges: fill in missing dates and hashes.

12 years agosmacker: convert palette and header reading to bytestream2.
Ronald S. Bultje [Sun, 25 Mar 2012 00:37:43 +0000 (17:37 -0700)]
smacker: convert palette and header reading to bytestream2.

12 years agoalac: convert extradata reading to bytestream2.
Ronald S. Bultje [Sun, 25 Mar 2012 02:10:40 +0000 (19:10 -0700)]
alac: convert extradata reading to bytestream2.

12 years agox86: dsputil: prettyprint gcc inline asm
Diego Biurrun [Fri, 16 Mar 2012 17:42:01 +0000 (18:42 +0100)]
x86: dsputil: prettyprint gcc inline asm

12 years agox86: K&R prettyprinting cosmetics for dsputil_mmx.c
Diego Biurrun [Tue, 6 Mar 2012 12:00:42 +0000 (13:00 +0100)]
x86: K&R prettyprinting cosmetics for dsputil_mmx.c

12 years agox86: conditionally compile H.264 QPEL optimizations
Diego Biurrun [Sun, 18 Dec 2011 13:10:33 +0000 (14:10 +0100)]
x86: conditionally compile H.264 QPEL optimizations

12 years agodsputil_mmx: Surround QPEL macros by "do { } while (0);" blocks.
Diego Biurrun [Thu, 8 Mar 2012 23:59:06 +0000 (00:59 +0100)]
dsputil_mmx: Surround QPEL macros by "do { } while (0);" blocks.

This makes them safe to use in non-fully braced if-blocks and similar.

12 years agoIgnore generated files below doc/.
Diego Biurrun [Fri, 23 Mar 2012 11:20:54 +0000 (12:20 +0100)]
Ignore generated files below doc/.

12 years agodpcm: convert to bytestream2.
Ronald S. Bultje [Mon, 19 Mar 2012 02:33:04 +0000 (19:33 -0700)]
dpcm: convert to bytestream2.

12 years agointerplayvideo: convert to bytestream2.
Ronald S. Bultje [Sun, 25 Mar 2012 00:05:05 +0000 (17:05 -0700)]
interplayvideo: convert to bytestream2.

12 years agomovenc: Merge if statements
Martin Storsjö [Thu, 22 Mar 2012 11:31:31 +0000 (13:31 +0200)]
movenc: Merge if statements

This isn't exactly equivalent with the earlier code for codecs
other than H264 and VC1, but those are two only codecs supported
by this codepath anyway, and it simplifies it a bit.

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agoh264: fix memleak in error path.
Alexander Strange [Sat, 24 Mar 2012 22:34:22 +0000 (15:34 -0700)]
h264: fix memleak in error path.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
12 years agopthread: Immediately release all frames in ff_thread_flush()
Alexander Strange [Sat, 24 Mar 2012 22:25:15 +0000 (18:25 -0400)]
pthread: Immediately release all frames in ff_thread_flush()

Before this, they were only added to the delayed release queue and not
freed until later. This could lead to unnecessary memory use or buffer
exhaustion.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
12 years agoh264: Add check for invalid chroma_format_idc
Alexander Strange [Sat, 24 Mar 2012 21:32:14 +0000 (17:32 -0400)]
h264: Add check for invalid chroma_format_idc

Fixes a crash when FF_DEBUG_PICT_INFO is used.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
12 years agoutvideo: port header reading to bytestream2.
Ronald S. Bultje [Fri, 23 Mar 2012 00:25:22 +0000 (17:25 -0700)]
utvideo: port header reading to bytestream2.

Fixes crash during slice size reading if slice_end goes negative.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agoavc: Add a function for converting mp4 style extradata to annex b
Martin Storsjö [Thu, 22 Mar 2012 11:25:58 +0000 (13:25 +0200)]
avc: Add a function for converting mp4 style extradata to annex b

Make movenc use this function instead of the current custom
conversion function.

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agopthread: free progress if buffer allocation failed.
Ronald S. Bultje [Sat, 24 Mar 2012 19:19:19 +0000 (20:19 +0100)]
pthread: free progress if buffer allocation failed.

Else we run out of progress variables after a few failed buffer
allocations.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agolavc/avconv: support changing frame sizes in codecs with frame mt.
Ronald S. Bultje [Fri, 23 Mar 2012 22:03:03 +0000 (15:03 -0700)]
lavc/avconv: support changing frame sizes in codecs with frame mt.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agolibavformat: Document who sets the AVStream.id field
Martin Storsjö [Thu, 22 Mar 2012 15:05:08 +0000 (17:05 +0200)]
libavformat: Document who sets the AVStream.id field

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agoutvideo: mark output picture as keyframe.
Kostya Shishkov [Sat, 24 Mar 2012 18:54:31 +0000 (19:54 +0100)]
utvideo: mark output picture as keyframe.

Spotted by Антон.

12 years agosunrast: Add support for negative linesize.
Aneesh Dogra [Fri, 23 Mar 2012 17:54:00 +0000 (23:24 +0530)]
sunrast: Add support for negative linesize.

Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
12 years agovp8: fix update_lf_deltas in libavcodec/vp8.c
Janne Salonen [Mon, 19 Mar 2012 07:35:09 +0000 (09:35 +0200)]
vp8: fix update_lf_deltas in libavcodec/vp8.c

lf_delta.ref[i] and lf_delta.mode[i] were incorrectly reset to 0 if
specific delta value was not updated. Fixed to keep the previous value
if flag indicates that element in question is not updated.

Signed-off-by: Janne Salonen <jsalonen@google.com>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
12 years agoralf: read Huffman code lengths without GetBitContext
Kostya Shishkov [Sat, 24 Mar 2012 06:29:51 +0000 (07:29 +0100)]
ralf: read Huffman code lengths without GetBitContext

Those descriptions are stored in nibbles, so they are easy to extract.
And this way we don't need to pad tables for possible bit reader overreads.

12 years agorv34: error out on size changes with frame threading
Janne Grunau [Fri, 23 Mar 2012 21:30:38 +0000 (22:30 +0100)]
rv34: error out on size changes with frame threading

12 years agoaacsbr: Add a debug check to sbr_mapping.
Alex Converse [Wed, 21 Mar 2012 19:20:58 +0000 (12:20 -0700)]
aacsbr: Add a debug check to sbr_mapping.

There have been multiple bugs caused by inconsistencies here.

Based on an idea from Michael Niedermayer.

CC: libav-stable@libav.org
12 years agoaac: Reset some state variables when turning SBR off
Alex Converse [Wed, 21 Mar 2012 19:00:56 +0000 (12:00 -0700)]
aac: Reset some state variables when turning SBR off

This makes sure the reset flag gets set when SBR gets turned back on
and sets control variables for unguided mode back to their defaults.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agoaac: Reset PS parameters on header decode failure.
Alex Converse [Wed, 21 Mar 2012 17:11:02 +0000 (10:11 -0700)]
aac: Reset PS parameters on header decode failure.

If the next header frame codes zero envelopes the previous frame's
values will be used. Consequently the invalid values must be cleared.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agofate: add wmalossless test.
Ronald S. Bultje [Thu, 22 Mar 2012 19:13:24 +0000 (12:13 -0700)]
fate: add wmalossless test.

12 years agoaacsbr: handle m_max values smaller than 4.
Ronald S. Bultje [Thu, 22 Mar 2012 18:50:48 +0000 (11:50 -0700)]
aacsbr: handle m_max values smaller than 4.

Prevents a signflip in the counter, and a subsequent crash because of
overreads/overwrites.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agorv34: Handle only complete frames in frame-mt.
Janne Grunau [Fri, 16 Mar 2012 15:18:57 +0000 (16:18 +0100)]
rv34: Handle only complete frames in frame-mt.

Correct handling of errors to prevent hags or crashes is very complex
otherwise.

The frame initializing is also moved from decode_slice() to
decode_frame() for clarity.

12 years agoMPV: set reference frame pointers to NULL when allocation of dummy pictures fails
Janne Grunau [Tue, 20 Mar 2012 13:44:28 +0000 (14:44 +0100)]
MPV: set reference frame pointers to NULL when allocation of dummy pictures fails

12 years agoconfigure: die if x11grab dependencies are unavailable
Josh Allmann [Wed, 21 Mar 2012 15:01:37 +0000 (08:01 -0700)]
configure: die if x11grab dependencies are unavailable

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agozerocodec: factorize loop
Michael Niedermayer [Thu, 22 Mar 2012 19:50:58 +0000 (15:50 -0400)]
zerocodec: factorize loop

Signed-off-by: Diego Biurrun <diego@biurrun.de>
12 years agoavconv: fix the resampling safety factors for output audio buffer allocation
Justin Ruggles [Wed, 21 Mar 2012 20:56:49 +0000 (16:56 -0400)]
avconv: fix the resampling safety factors for output audio buffer allocation

This matches the output size required for audio_resample()

12 years agoavconv: move audio output buffer allocation to a separate function
Justin Ruggles [Wed, 21 Mar 2012 20:55:30 +0000 (16:55 -0400)]
avconv: move audio output buffer allocation to a separate function

Allows for removing a goto and makes the code easier to follow.

12 years agoavconv: make the async buffer global and free it in exit_program()
Justin Ruggles [Wed, 21 Mar 2012 20:36:23 +0000 (16:36 -0400)]
avconv: make the async buffer global and free it in exit_program()

12 years agoxwma: Validate channels and bits_per_coded_sample.
Alex Converse [Wed, 21 Mar 2012 17:58:07 +0000 (10:58 -0700)]
xwma: Validate channels and bits_per_coded_sample.

This prevents a SIGFPE later on.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agomov: Do not read past the end of the ctts_data table.
Alex Converse [Wed, 21 Mar 2012 18:24:10 +0000 (11:24 -0700)]
mov: Do not read past the end of the ctts_data table.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agomov: Add missing terminator to mov_ch_layout_map_1ch.
Alex Converse [Wed, 21 Mar 2012 16:35:45 +0000 (09:35 -0700)]
mov: Add missing terminator to mov_ch_layout_map_1ch.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: Libav-stable@libav.org
12 years agoasf: reset side data elements on packet copy.
Ronald S. Bultje [Wed, 21 Mar 2012 23:10:37 +0000 (16:10 -0700)]
asf: reset side data elements on packet copy.

Prevents crash (double free) when free()ing the original packet.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agowmavoice: fix stack overread.
Ronald S. Bultje [Wed, 21 Mar 2012 22:47:11 +0000 (15:47 -0700)]
wmavoice: fix stack overread.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agowmalossless: error out if a subframe is not used by any channel.
Ronald S. Bultje [Wed, 21 Mar 2012 22:43:03 +0000 (15:43 -0700)]
wmalossless: error out if a subframe is not used by any channel.

Prevents infinite loop because min_channel_len never increments.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agovqa: check palette chunk size before reading data.
Ronald S. Bultje [Wed, 21 Mar 2012 22:19:31 +0000 (15:19 -0700)]
vqa: check palette chunk size before reading data.

Prevents overreads beyond buffer boundaries.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agowmalossless: reset sample pointer for each subframe.
Ronald S. Bultje [Wed, 21 Mar 2012 22:02:19 +0000 (15:02 -0700)]
wmalossless: reset sample pointer for each subframe.

Prevents overwrites when some subframes only encode some channels.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agowmalossless: error out on invalid values for order.
Ronald S. Bultje [Wed, 21 Mar 2012 19:40:59 +0000 (12:40 -0700)]
wmalossless: error out on invalid values for order.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
12 years agoFATE: Add ZeroCodec test
Derek Buitenhuis [Mon, 19 Mar 2012 22:23:42 +0000 (18:23 -0400)]
FATE: Add ZeroCodec test

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agooggparseogm: fix order of arguments of avpriv_set_pts_info().
Michael Niedermayer [Thu, 8 Mar 2012 20:11:37 +0000 (21:11 +0100)]
oggparseogm: fix order of arguments of avpriv_set_pts_info().

Signed-off-by: Anton Khirnov <anton@khirnov.net>
12 years agopngenc: better upper bound for encoded frame size.
Anton Khirnov [Wed, 21 Mar 2012 08:48:26 +0000 (09:48 +0100)]
pngenc: better upper bound for encoded frame size.

Fixes encoding very large pictures.

Thanks to Костя for providing the formula.

12 years agoaiffdec: set block_duration to 1 for PCM codecs that are supported in AIFF-C
Justin Ruggles [Wed, 21 Mar 2012 18:51:50 +0000 (14:51 -0400)]
aiffdec: set block_duration to 1 for PCM codecs that are supported in AIFF-C

12 years agoaiffdec: factor out handling of integer PCM for AIFF-C and plain AIFF
Justin Ruggles [Wed, 21 Mar 2012 18:48:35 +0000 (14:48 -0400)]
aiffdec: factor out handling of integer PCM for AIFF-C and plain AIFF

12 years agoaiffdec: use av_get_audio_frame_duration() to set block_duration for AIFF-C
Justin Ruggles [Wed, 21 Mar 2012 18:36:33 +0000 (14:36 -0400)]
aiffdec: use av_get_audio_frame_duration() to set block_duration for AIFF-C

12 years agoaiffdec: do not set bit rate if block duration is unknown
Justin Ruggles [Wed, 21 Mar 2012 18:30:33 +0000 (14:30 -0400)]
aiffdec: do not set bit rate if block duration is unknown

CC: libav-stable@libav.org
12 years agowmall: output packet only if we have decoded some samples
Kostya Shishkov [Wed, 21 Mar 2012 18:10:43 +0000 (19:10 +0100)]
wmall: output packet only if we have decoded some samples

Also set CODEC_CAP_DELAY to indicate that decoder may still have some
undecoded data left in internal buffer.

12 years agoadxenc: use AVCodec.encode2()
Justin Ruggles [Wed, 21 Mar 2012 17:58:09 +0000 (13:58 -0400)]
adxenc: use AVCodec.encode2()

12 years agoadxenc: Use the AVFrame in ADXContext for coded_frame
Justin Ruggles [Wed, 21 Mar 2012 17:30:37 +0000 (13:30 -0400)]
adxenc: Use the AVFrame in ADXContext for coded_frame

12 years agoindeo4: fix out-of-bounds function call.
Ronald S. Bultje [Wed, 21 Mar 2012 17:39:10 +0000 (10:39 -0700)]
indeo4: fix out-of-bounds function call.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
Signed-off-by: Kostya Shishkov <kostya.shishkov@gmail.com>
12 years agoconfigure: Restructure help output.
Diego Biurrun [Thu, 8 Mar 2012 23:24:48 +0000 (00:24 +0100)]
configure: Restructure help output.

Break some of the longer sections into smaller sensible pieces;
make some option descriptions and option ordering more consistent.

12 years agoconfigure: Internal-only components should not be command-line selectable.
Diego Biurrun [Thu, 8 Mar 2012 22:53:57 +0000 (23:53 +0100)]
configure: Internal-only components should not be command-line selectable.

12 years agovorbisenc: use AVCodec.encode2()
Justin Ruggles [Wed, 29 Feb 2012 08:16:24 +0000 (03:16 -0500)]
vorbisenc: use AVCodec.encode2()

12 years agolibvorbis: use AVCodec.encode2()
Justin Ruggles [Wed, 29 Feb 2012 06:02:54 +0000 (01:02 -0500)]
libvorbis: use AVCodec.encode2()

12 years agolibopencore-amrnbenc: use AVCodec.encode2()
Justin Ruggles [Tue, 28 Feb 2012 07:51:29 +0000 (02:51 -0500)]
libopencore-amrnbenc: use AVCodec.encode2()

12 years agora144enc: use AVCodec.encode2()
Justin Ruggles [Tue, 28 Feb 2012 06:48:31 +0000 (01:48 -0500)]
ra144enc: use AVCodec.encode2()

12 years agonellymoserenc: use AVCodec.encode2()
Justin Ruggles [Tue, 28 Feb 2012 05:08:11 +0000 (00:08 -0500)]
nellymoserenc: use AVCodec.encode2()

12 years agoroqaudioenc: use AVCodec.encode2()
Justin Ruggles [Tue, 28 Feb 2012 01:53:09 +0000 (20:53 -0500)]
roqaudioenc: use AVCodec.encode2()

The first frame pts must be saved until we have 8 frames since RoQ audio
requires 8 frames in the first packet.

12 years agolibspeex: use AVCodec.encode2()
Justin Ruggles [Tue, 28 Feb 2012 01:01:42 +0000 (20:01 -0500)]
libspeex: use AVCodec.encode2()

12 years agolibvo_amrwbenc: use AVCodec.encode2()
Justin Ruggles [Thu, 23 Feb 2012 22:22:42 +0000 (17:22 -0500)]
libvo_amrwbenc: use AVCodec.encode2()

12 years agolibvo_aacenc: use AVCodec.encode2()
Justin Ruggles [Thu, 23 Feb 2012 06:14:15 +0000 (01:14 -0500)]
libvo_aacenc: use AVCodec.encode2()

12 years agowmaenc: use AVCodec.encode2()
Justin Ruggles [Mon, 20 Feb 2012 17:29:14 +0000 (12:29 -0500)]
wmaenc: use AVCodec.encode2()

12 years agompegaudioenc: use AVCodec.encode2()
Justin Ruggles [Sun, 19 Feb 2012 22:12:48 +0000 (17:12 -0500)]
mpegaudioenc: use AVCodec.encode2()

Update FATE references due to encoder delay.

12 years agolibmp3lame: use AVCodec.encode2()
Justin Ruggles [Sun, 19 Feb 2012 18:16:34 +0000 (13:16 -0500)]
libmp3lame: use AVCodec.encode2()

12 years agolibgsmenc: use AVCodec.encode2()
Justin Ruggles [Thu, 16 Feb 2012 00:20:30 +0000 (19:20 -0500)]
libgsmenc: use AVCodec.encode2()

12 years agolibfaac: use AVCodec.encode2()
Justin Ruggles [Tue, 7 Feb 2012 19:28:50 +0000 (14:28 -0500)]
libfaac: use AVCodec.encode2()

Encoder output is delayed by several frames, so we keep a queue of input
frame timing info to match up with corresponding output packets.

12 years agog726enc: use AVCodec.encode2()
Justin Ruggles [Sat, 4 Feb 2012 16:36:59 +0000 (11:36 -0500)]
g726enc: use AVCodec.encode2()

12 years agog722enc: use AVCodec.encode2()
Justin Ruggles [Sat, 4 Feb 2012 16:26:33 +0000 (11:26 -0500)]
g722enc: use AVCodec.encode2()

FATE reference updated due timestamp rounding because of resampling from
44100 Hz to 16000 Hz in avconv.

12 years agoflacenc: use AVCodec.encode2()
Justin Ruggles [Fri, 3 Feb 2012 23:43:39 +0000 (18:43 -0500)]
flacenc: use AVCodec.encode2()

12 years agoadpcmenc: update to AVCodec.encode2()
Justin Ruggles [Tue, 31 Jan 2012 20:54:45 +0000 (15:54 -0500)]
adpcmenc: update to AVCodec.encode2()

12 years agoac3enc: update to AVCodec.encode2()
Justin Ruggles [Mon, 30 Jan 2012 00:01:10 +0000 (19:01 -0500)]
ac3enc: update to AVCodec.encode2()

Update FATE references due to encoder delay.

12 years agoaacenc: use AVCodec.encode2()
Justin Ruggles [Sat, 28 Jan 2012 17:28:01 +0000 (12:28 -0500)]
aacenc: use AVCodec.encode2()

12 years agoavcodec: add code for a frame queue for use by audio encoders with delay
Justin Ruggles [Thu, 23 Feb 2012 02:52:34 +0000 (21:52 -0500)]
avcodec: add code for a frame queue for use by audio encoders with delay

This simplifies matching of timestamps between input frames and output
packets.

12 years agoavconv: free packet in write_frame() when discarding due to frame number limit
Justin Ruggles [Tue, 20 Mar 2012 19:36:28 +0000 (15:36 -0400)]
avconv: free packet in write_frame() when discarding due to frame number limit

Fixes a memleak when using the -frames option with audio.

12 years agoFATE: use +/- flag option syntax for vp8 emu-edge tests
Justin Ruggles [Tue, 20 Mar 2012 19:10:39 +0000 (15:10 -0400)]
FATE: use +/- flag option syntax for vp8 emu-edge tests

12 years agolavf: make av_interleave_packet_per_dts() private.
Anton Khirnov [Thu, 15 Mar 2012 12:37:43 +0000 (13:37 +0100)]
lavf: make av_interleave_packet_per_dts() private.

There is no reason for it to be public, it's only meant to be used
internally.

12 years agolavf: deprecate av_read_packet().
Anton Khirnov [Sun, 4 Mar 2012 20:41:59 +0000 (21:41 +0100)]
lavf: deprecate av_read_packet().

The caller can achieve the same effect (i.e. getting raw unparsed/mangled
packets) with av_read_frame() and AVFMT_FLAG_NOPARSE |
AVFMT_FLAG_NOFILLIN

12 years agooggdec: output correct timestamps for Vorbis
Justin Ruggles [Fri, 24 Feb 2012 02:40:14 +0000 (21:40 -0500)]
oggdec: output correct timestamps for Vorbis

Takes encoder delay into account by comparing first the coded page
duration with the calculated page duration. Handles last packet duration
if needed, also by comparing coded duration with calculated duration.
Also does better handling of timestamp generation for packets in the
first page for streamed ogg files where the start time is not
necessarily zero.

12 years agoavconv: pass input stream timestamps to audio encoders
Justin Ruggles [Thu, 1 Mar 2012 18:02:49 +0000 (13:02 -0500)]
avconv: pass input stream timestamps to audio encoders

5 FATE test references updated due to using demuxer-generated timestamps that
are either not sample-accurate or are slightly off in the input file.

12 years agolavc: shrink encoded audio packet size after encoding.
Justin Ruggles [Tue, 20 Mar 2012 17:21:52 +0000 (13:21 -0400)]
lavc: shrink encoded audio packet size after encoding.

12 years agoxa: set correct bit rate
Justin Ruggles [Tue, 20 Mar 2012 16:13:15 +0000 (12:13 -0400)]
xa: set correct bit rate

Also fixes stream duration calculation.

12 years agoxa: do not set bit_rate, block_align, or bits_per_coded_sample
Justin Ruggles [Tue, 20 Mar 2012 16:00:00 +0000 (12:00 -0400)]
xa: do not set bit_rate, block_align, or bits_per_coded_sample

The values in the header refer to decoded data, not compressed data.

12 years agoxa: fix end-of-file handling
Justin Ruggles [Tue, 20 Mar 2012 15:53:56 +0000 (11:53 -0400)]
xa: fix end-of-file handling

Do not output an extra packet when out_size is reached.
Also return AVERROR_EOF instead of AVERROR(EIO).

12 years agoxa: fix timestamp calculation
Justin Ruggles [Tue, 20 Mar 2012 15:20:54 +0000 (11:20 -0400)]
xa: fix timestamp calculation

The packet duration is always 28 samples.

12 years agobink: fix typo in FFALIGN() argument
Kostya Shishkov [Tue, 20 Mar 2012 17:43:00 +0000 (18:43 +0100)]
bink: fix typo in FFALIGN() argument

12 years agobink: align plane width to 8 when calculating bundle sizes
Kostya Shishkov [Tue, 20 Mar 2012 06:53:41 +0000 (07:53 +0100)]
bink: align plane width to 8 when calculating bundle sizes

This fixes decoding of Bink files with non-multiple-of-16 width.

12 years agodoc: pass -Idoc texi2html and texi2pod
Mans Rullgard [Tue, 20 Mar 2012 10:54:02 +0000 (10:54 +0000)]
doc: pass -Idoc texi2html and texi2pod

This fixes doc generation in build tree separate from source.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agodoc: texi2pod: add -I flag
Mans Rullgard [Tue, 20 Mar 2012 10:53:06 +0000 (10:53 +0000)]
doc: texi2pod: add -I flag

This allows specifying additional directories to search for
@include files.

Signed-off-by: Mans Rullgard <mans@mansr.com>
12 years agomovenc: Add a min_frag_duration option
Martin Storsjö [Mon, 19 Mar 2012 17:31:15 +0000 (19:31 +0200)]
movenc: Add a min_frag_duration option

The other fragmentation options (frag_duration, frag_size and
frag_keyframe) are combined with OR, cutting fragments at the
first of the conditions being fulfilled.

Signed-off-by: Martin Storsjö <martin@martin.st>
12 years agortsp: Set the default delay to 0.1 s for the RTSP/SDP/RTP demuxers
Martin Storsjö [Mon, 19 Mar 2012 18:40:23 +0000 (20:40 +0200)]
rtsp: Set the default delay to 0.1 s for the RTSP/SDP/RTP demuxers

This enables reordering of UDP packets by default, unless the caller
explicitly sets -max_delay 0.

Signed-off-by: Martin Storsjö <martin@martin.st>