Commit Graph

8997 Commits

Author SHA1 Message Date
Justin Ruggles
239b961fbf flacdec: Allocate buffers right after STREAMINFO is parsed.
Originally committed as revision 17794 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-04 00:21:46 +00:00
Stefano Sabatini
9d9de444e3 Fix syntax for the libfaad_decoder definition.
Fix compilation (when libfaad is enabled).

Originally committed as revision 17792 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 23:52:31 +00:00
Justin Ruggles
184bc53db4 cosmetics: indentation after last commit
Originally committed as revision 17791 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 23:50:21 +00:00
Justin Ruggles
7d0303587d flacdec: Avoid parsing the STREAMINFO multiple times.
Originally committed as revision 17790 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 23:49:36 +00:00
Michael Niedermayer
6d4f53cbac Remove old scaler.
Originally committed as revision 17786 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 22:51:30 +00:00
Diego Biurrun
59701aeb46 Do not use a macro to generate a single AVCodec declaration.
Originally committed as revision 17773 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 15:48:52 +00:00
Panagiotis Issaris
a0f8005079 When a H.264 stream references a PPS or SPS id which doesn't exist, instead of
just saying that a non-existing id is referenced, show the value of the id.

Originally committed as revision 17771 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 14:32:24 +00:00
Gwenole Beauchesne
0d8ee24c7b Set PixFmtInfo::{x, y}_chroma_shift for VDPAU and VAAPI formats.
Originally committed as revision 17766 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 13:43:24 +00:00
Benoit Fouet
4f46099b90 Use sign_extend().
On behalf of Jai.

Originally committed as revision 17759 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 13:11:10 +00:00
Justin Ruggles
2e78513cee flacdec: Warn about invalid max blocksize and limit the minimum value.
Originally committed as revision 17751 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 05:53:38 +00:00
Alex Converse
09a64ee614 H.264: Simplify decode_residual()
Originally committed as revision 17750 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 05:26:39 +00:00
Justin Ruggles
0c89ca72a5 cosmetics: remove a pointless comment
Originally committed as revision 17749 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 05:25:58 +00:00
Justin Ruggles
95db6659d8 flacdec: Remove unused variable, min_blocksize.
Originally committed as revision 17748 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 05:25:23 +00:00
Justin Ruggles
dde318d5d9 flacdec: Return error when blocksize code of 0 is found. It is a
reserved value per the FLAC format documentation.

Originally committed as revision 17747 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 05:22:44 +00:00
Kostya Shishkov
c978997a88 Enable PB-frames decoding for H.263
Originally committed as revision 17746 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 05:18:39 +00:00
Kostya Shishkov
685502cf3d Add PB-frame decoding support for H.263/i263 (and enable it for i263)
Originally committed as revision 17745 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 05:17:22 +00:00
Justin Ruggles
6053da0182 alacdec: Simplify reading of uncompressed samples by using
get_sbits_long().

Originally committed as revision 17744 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 04:57:02 +00:00
Justin Ruggles
5b37e2fc9a flacdec: Avoid trying to read 33 bits. This occurs when the source
is 32 bits-per-sample and channel decorrelation is used.  Such files
are valid, but not supported currently.

Originally committed as revision 17743 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 04:45:45 +00:00
Justin Ruggles
505cc62f75 flacdec: ALT_BITSTREAM_READER is no longer required.
Originally committed as revision 17742 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 02:53:11 +00:00
Justin Ruggles
9de6e090a7 flacdec: use get_sbits_long() where needed.
Originally committed as revision 17741 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 02:19:01 +00:00
Justin Ruggles
017c0811a1 Add a get_sbits_long() function.
Originally committed as revision 17740 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 02:00:47 +00:00
Måns Rullgård
8858990f07 ALAC: use sign_extend() from mathops.h
Originally committed as revision 17739 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 01:53:54 +00:00
Måns Rullgård
101dfa7d0a Add sign_extend() function to mathops.h
Originally committed as revision 17738 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-03 00:03:55 +00:00
Reimar Döffinger
edd532db64 Simplify show_bits_long and copy the GetBitsContext around only once instead of twice.
Originally committed as revision 17736 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 19:42:31 +00:00
Justin Ruggles
ee4d0322e8 flacdec: Use get/skip_bits_long() for more than 17-bits and
get/skip_bits() for 17-bits or less.

Originally committed as revision 17735 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 17:32:53 +00:00
Kostya Shishkov
b89e77cff7 Factor out block decoding in ff_h263_decode_mb()
Originally committed as revision 17734 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 17:18:49 +00:00
Michael Niedermayer
7115cbda30 Consider all packets in the parser, not just ones with timestamps,
this should greatly simplify handling of packet pos.

Originally committed as revision 17733 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 15:13:58 +00:00
Michael Niedermayer
2cc304215d Favor container packets that end after the first byte of the access
unit in fetch_timestamp(). This should make no difference for valid
streams but may help invalid ones, also its needed for future changes.

Originally committed as revision 17732 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 15:12:26 +00:00
Michael Niedermayer
a4c7a5ea27 Call ff_fetch_timestamp() for mpeg1/2 when a picture start code is found instead
of calling it at the end of a frame with a large negative offset.
This significantly reduces the maximal distance in container packets between
the point where the first byte of the "access unit" was stored and where
we call ff_fetch_timestamp() thus reducing the constraints on our parser.
Also change the parser from next_frame_offset to cur, this is needed
because now the reference is from container packet start instead of
frame start. (i previously misinterpreted this as bug)

Originally committed as revision 17731 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 14:53:18 +00:00
Benoit Fouet
d69da18cb0 Merge three conditions in a single 'if' instead of two.
Originally committed as revision 17728 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 09:22:17 +00:00
Diego Biurrun
12dd57fd29 cosmetics: Reformat long_names so that "DPCM" comes first.
Originally committed as revision 17721 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 06:01:32 +00:00
Diego Biurrun
fd75cedd15 cosmetics: Reformat long_names so that "ADPCM" comes first.
Originally committed as revision 17720 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 05:55:03 +00:00
Diego Biurrun
655463b9eb cosmetics: Reformat long_names so that "PCM" comes first.
Originally committed as revision 17719 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 05:45:44 +00:00
Diego Biurrun
7eb2d654a2 small spelling/grammar fixes
Originally committed as revision 17718 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 05:41:55 +00:00
Diego Biurrun
326b554ccf xvmc works with MPEG-2 as well as MPEG-1, note this in the codec long_name.
Originally committed as revision 17717 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 05:20:00 +00:00
Diego Biurrun
0ffbc258aa Change a bunch of codec long_names to be more consistent and descriptive.
Originally committed as revision 17716 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 05:18:33 +00:00
Justin Ruggles
68dddf53ff Add some missing codecs and update names to match the codec long_name
or vice-versa.

Originally committed as revision 17706 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-02 02:25:31 +00:00
Måns Rullgård
09096c9fad RA144: work around gcc bug/oddity on ARM
This should not make any difference, yet some gcc versions on ARM
produce incorrect output without this fix.

Approved by Vitor.

Originally committed as revision 17698 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-01 23:41:39 +00:00
Måns Rullgård
db794291a2 SVQ3: Fix decoding with A32_BITSTREAM_READER
svq3_decode_slice_header() modifies the buffer used by the bitstream
reader.  Some of the bitstream readers cache a few bytes of data, which
must be flushed after such a modification.  Calling skip_bits_long(gb, 0)
achieves this.

Originally committed as revision 17680 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-01 12:11:07 +00:00
Måns Rullgård
4cd19f6e78 ARM: fix missing MUL16() return type
Originally committed as revision 17679 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-01 12:11:02 +00:00
Vitor Sessak
8d2544bd1b Reindent after last commit
Originally committed as revision 17678 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-01 11:16:57 +00:00
Vitor Sessak
9e27e0d4f8 Do not read uninitialized buffer, no matter if it will be multiplied by
zero later. This should fix some valgrind warnings and hopefully FATE
ra144 test on ARM.

Originally committed as revision 17677 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-01 11:14:21 +00:00
Reimar Döffinger
d819b41ba1 Remove unused variable in mpeg_mc_decode_init
Originally committed as revision 17676 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-01 09:14:20 +00:00
Michael Niedermayer
47cd974ae5 Do not set low_delay if has_b_frames has been set before init()
this will be needed once the parser can figure out has_b_frames
in av_find_stream_info().

Originally committed as revision 17673 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-01 03:13:04 +00:00
Alex Converse
9ef88afac3 AAC: Make the CPE function more consistent with the functions for other channel elements.
Originally committed as revision 17672 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-03-01 00:23:05 +00:00
Michael Niedermayer
efd8c1f64d Adjust time_base exactly when changing ticks_per_frame to 2 this is safer
and for some raw h264 we managed to change the timebase twice before.

Originally committed as revision 17669 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-02-28 20:58:07 +00:00
Måns Rullgård
fe7f149ed8 ARM: fix corner-case overflow in H.264 weighted prediction
Originally committed as revision 17657 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-02-28 13:48:54 +00:00
David S. Miller
da2911d5e8 sub_mb_type[] needs to be 8-byte aligned because it is referenced
as a uint64_t in get_dct8x8_allowed().
Patch by David S. Miller <davem at davemloft dot net>

Originally committed as revision 17655 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-02-28 08:38:33 +00:00
Cédric Schieli
5f59207d94 Fix avpicture_layout to not chroma shift the alpha plane when outputting YUVA420P
Originally committed as revision 17654 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-02-28 07:31:36 +00:00
Ivan Schreter
03831f462e Correctly reset SEI variables.
Patch by Ivan Schreter, schreter gmx net

Originally committed as revision 17651 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-02-27 21:28:41 +00:00