virtualenv/FFmpeg
1
0
Fork 0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-12 19:18:44 +02:00
Code Issues Releases Activity
32,604 Commits 37 Branches 400 Tags 463 MiB
791de61bbb
Commit Graph

15511 Commits

Author SHA1 Message Date
Anton Khirnov
d55fa6f9cb lavc: factorize setting got_packet_ptr in avcodec_encode_video2() 2012-02-25 07:48:46 +01:00
Anton Khirnov
52953d61ca lavc: signal no output when a NULL frame is passed to audio encoder without delay 2012-02-25 07:48:36 +01:00
Paul B Mahol
99cff417f3 xwdenc: fix monow encoding 
Signed-off-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
 2012-02-25 07:45:06 +01:00
Paul B Mahol
904817cb9f cdxl: fix ham6/8 on big endian 
Signed-off-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
 2012-02-25 00:18:14 -05:00
Justin Ruggles
19adb0bc2d vorbisdec: add a flush() function 
clear MDCT overlap buffer and reset previous window mode when seeking
 2012-02-24 17:37:25 -05:00
Ronald S. Bultje
830f70442a fraps: release reference buffer on pix_fmt change. 
Prevents crash when trying to copy from a non-existing plane in e.g.
a RGB32 reference image to a YUV420P target image

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-24 14:28:15 -08:00
Ronald S. Bultje
87840eeb71 avcodec: disallow reget_buffer() if pix_fmt changed. 2012-02-24 14:28:15 -08:00
Ronald S. Bultje
33cd32b389 kgv1: use avctx->get/release_buffer(). 
Also fixes crashes on corrupt bitstreams.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-24 14:28:15 -08:00
Dale Curtis
b716542691 mpegaudiodec: Prevent premature clipping of mp3 input buffer. 
Instead of clipping extrasize based on EXTRABYTES, clip based on the
amount of buffer actually left. Without this fix, there are warbles
and other distortions in the test case below.

http://kevincennis.com/mix/assets/sounds/1901_voxfx.mp3
 2012-02-24 13:47:04 -05:00
Ronald S. Bultje
be129271ea lcl: error out if uncompressed input buffer is smaller than framesize. 
This prevents crashes when trying to read beyond the end of the buffer
while decoding frame data.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-24 07:28:57 -08:00
Ronald S. Bultje
ab492ca2ab mjpeg: abort decoding if packet is too large. 
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-24 07:28:43 -08:00
Ronald S. Bultje
46b3fbc30b golomb: use HAVE_BITS_REMAINING() macro to prevent infloop on EOF. 
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-24 07:28:33 -08:00
Ronald S. Bultje
b44b41633f get_bits: add HAVE_BITS_REMAINING macro. 2012-02-24 07:28:14 -08:00
Alex Converse
447363870f tiff: Prevent overreads in the type_sizes array. 
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-23 16:23:51 -08:00
Alex Converse
e32548d133 tiff: Make the TIFF_LONG and TIFF_SHORT types unsigned. 
TIFF v6.0 (unimplemented) adds signed equivalents.
 2012-02-23 16:23:51 -08:00
Christophe GISQUET
2784d18791 SBR DSP x86: implement SSE sbr_hf_g_filt 
Unrolling the main loop to process, instead of 4 elements:
- 8: minor gain of 2 cycles (not worth the extra object size)
- 2: loss of 8 cycles.

Assigning STEP to a register is a loss. Output address (Y) is almost always
unaligned.

Timings:
- C (32/64 bits): 117/109 cycles
- SSE: 57 cycles

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
 2012-02-23 15:50:09 -08:00
Christophe GISQUET
34454c761f SBR DSP x86: implement SSE sbr_sum_square_sse 
The 32bits targets have been compiled with -mfpmath=sse for proper reference.
sbr_sum_square C  /32bits: 82c (unrolled)/102c
               C  /64bits: 69c (unrolled)/82c
               SSE/32bits: 42c
               SSE/64bits: 31c

Use of SSE4.1 dpps to perform the final sum is slower.
Not unrolling to perform 8 operations in a loop yields 10 more cycles.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
 2012-02-23 15:50:06 -08:00
Christophe GISQUET
2e74a5abc2 SBR DSP: use intptr_t for the ixh parameter. 
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
 2012-02-23 15:48:40 -08:00
Derek Buitenhuis
480b133e6f wavpack: Don't shift minclip/maxclip 
Since we are clipping before we shift the values to
16 or 32 bits, we should not shift the min/max clip
values to compensate.

Fixes 8 and 24 bit lossy decoding.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
 2012-02-23 20:43:39 +01:00
Anton Khirnov
ff311c0918 lavc: drop encode() support for video. 
All video encoders have been converted to encode2(), all new encoders
should also use only encode2().
 2012-02-23 20:26:26 +01:00
Anton Khirnov
171273fec8 snowenc: switch to encode2(). 2012-02-23 20:26:14 +01:00
Anton Khirnov
8d3348c2c4 snowenc: don't abuse input picture for storing information. 2012-02-23 20:26:04 +01:00
Anton Khirnov
7340008f18 a64multienc: switch to encode2(). 
We have no decoder, so cannot test if the output is decodable.
 2012-02-23 19:56:55 +01:00
Anton Khirnov
cfc6ab3199 a64multienc: don't write into output buffer when there's no output. 2012-02-23 19:56:45 +01:00
Anton Khirnov
4da6d194e5 libxvid: switch to encode2(). 2012-02-23 19:55:23 +01:00
Anton Khirnov
760b004086 tiffenc: switch to encode2(). 2012-02-23 19:53:20 +01:00
Anton Khirnov
62acb4937e tiffenc: properly forward error codes in encode_frame(). 2012-02-23 19:52:08 +01:00
Anton Khirnov
2257f66ec5 lavc: drop libdirac encoder. 
Libschroedinger is the preferred way to encode dirac video now, it
produces better output and has a nicer API.
 2012-02-23 19:50:46 +01:00
Anton Khirnov
577fed3b7c gifenc: switch to encode2(). 2012-02-23 19:50:46 +01:00
Anton Khirnov
d8f3365fb0 libvpxenc: switch to encode2(). 2012-02-23 19:50:46 +01:00
Anton Khirnov
0ecf54f9dc flashsvenc: switch to encode2(). 2012-02-23 19:50:46 +01:00
Ronald S. Bultje
01cb62aba2 lcl: don't overwrite input memory. 
If the PNG filter is enabled, a PNG-style filter will run over the
input buffer, writing into the buffer. Therefore, if no zlib compression
was used, ensure that we copy into a temporary buffer, otherwise we
overwrite user-provided input data.
 2012-02-23 10:31:46 -08:00
Anton Khirnov
84db922bf9 libxavs: switch to encode2(). 2012-02-23 19:25:50 +01:00
Anton Khirnov
d3b577f00d libxavs: add an AVClass at the beginning of XavsContext. 2012-02-23 19:25:50 +01:00
Anton Khirnov
a5f848c86d libxavs: split extradata writing out of encode_nals(). 
This is done in preparation for the following patch implementing
encode2().

This commit is analogous to 05d699222d for
libx264.
 2012-02-23 19:25:50 +01:00
Anton Khirnov
8e8124e173 ljpegenc: switch to encode2(). 2012-02-23 19:25:50 +01:00
Ronald S. Bultje
2b83e8b700 truemotion2: error out if the huffman tree has no nodes. 
This prevents crashers and errors further down when reading nodes in the
empty tree.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-23 07:46:25 -08:00
Anton Khirnov
3c0ed7d1a8 asv1enc: switch to encode2(). 2012-02-23 14:11:08 +01:00
Anton Khirnov
df53a47317 libschroedingerenc: switch to encode2(). 2012-02-23 13:54:34 +01:00
Anton Khirnov
278d88689b ffv1enc: switch to encode2(). 2012-02-23 13:51:29 +01:00
Anton Khirnov
f7fa73ac91 libtheoraenc: switch to encode2(). 2012-02-23 13:48:34 +01:00
Anton Khirnov
6d9c27dc85 jpeglsenc: switch to encode2(). 2012-02-23 11:12:39 +01:00
Anton Khirnov
148fc99506 lclenc: switch to encode2(). 2012-02-23 11:11:42 +01:00
Anton Khirnov
21b46747ad qtrleenc: switch to encode2(). 2012-02-23 11:08:50 +01:00
Justin Ruggles
62d5f9e5ca flacdec: set channel_layout based on channel count 
Channel layouts are specified in the FLAC format description at
http://flac.sourceforge.net/format.html

fixes Bug 209
 2012-02-22 20:14:00 -05:00
Justin Ruggles
3798205a77 mov: set channel layout for AC-3 streams based on the 'dac3' atom info 
fixes Bug 225
 2012-02-22 20:07:02 -05:00
Alex Converse
1cd9a6154b aac: fix infinite loop on end-of-frame with sequence of 1-bits. 
Based-on-work-by: Ronald S. Bultje <rsbultje@gmail.com>
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-22 11:23:43 -08:00
Tim Walker
6bbaf6e0dd mlpdec_parser: fix a few channel layouts. 
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
 2012-02-22 10:50:22 -05:00
Christophe GISQUET
b5696ff2b8 prores: use natural integer type for the codebook index 
The operations that use it require it to be promoted to a larger (natural)
type and thus perform sign extension on it.

While an optimal compiler may account for this, gcc 4.6 (for x86 Windows)
fails. Using the natural integer type provides a 2% speedup for Win64
and 1% for Win32.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
 2012-02-22 09:21:39 +01:00
Anton Khirnov
5d42ac7ffb mpegvideo_enc: only allocate output packet when we know there will be output 
Fixes a memleak.
 2012-02-21 22:47:01 +01:00
Aneesh Dogra
79c7064c5f sunrast: Add a sample request for RMP_RAW colormap. 
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
 2012-02-21 16:00:06 -05:00
Justin Ruggles
70749c483c avcodec: do not override pts or duration from the audio encoder 
This allows encoders to set pts and/or duration even if they do not use
CODEC_CAP_DELAY.
 2012-02-21 15:18:51 -05:00
Alex Converse
b0f29db5c2 Mark mutable static data const where appropriate. 2012-02-21 09:47:07 -08:00
Janne Grunau
6a4cf065c7 dvbsubdec: avoid undefined signed left shift in RGBA macro 2012-02-21 16:01:31 +01:00
Paul B Mahol
ca085e667b mlpdec: use av_log_ask_for_sample() 
Signed-off-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Janne Grunau <janne-libav@jannau.net>
 2012-02-21 16:01:31 +01:00
Paul B Mahol
08630d25a2 png: make .long_name more descriptive 
Signed-off-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Diego Biurrun <diego@biurrun.de>
 2012-02-21 15:47:05 +01:00
Ronald S. Bultje
3ab9a2a557 rv34: change most "int stride" into "ptrdiff_t stride". 
This prevents having to sign-extend on 64-bit systems with 32-bit ints,
such as x86-64. Also fixes crashes on systems where we don't do it and
arguments are not in registers, such as Win64 for all weight functions.
 2012-02-20 14:58:25 -08:00
Justin Ruggles
770a5c6d02 adpcmenc: Use correct frame_size for Yamaha ADPCM. 
Output packet size should match avctx->block_align. The target output packet
size is 1024 bytes.
Before:
mono   - 1024 samples -> 512 bytes
stereo - 2048 samples -> 2048 bytes
After:
mono   - 2048 samples -> 1024 bytes
stereo - 1024 samples -> 1024 bytes
 2012-02-20 15:52:32 -05:00
Justin Ruggles
91a28b0e8e avcodec: add ff_samples_to_time_base() convenience function to internal.h 2012-02-20 15:20:17 -05:00
Justin Ruggles
41ac9bb253 adx parser: set duration 2012-02-20 15:08:41 -05:00
Justin Ruggles
b3a4c7e0f1 mlp parser: set duration instead of frame_size 2012-02-20 15:08:41 -05:00
Justin Ruggles
2460b168b4 gsm parser: set duration 2012-02-20 15:08:41 -05:00
Justin Ruggles
7575ffac8a mpegaudio parser: set duration instead of frame_size 2012-02-20 15:08:40 -05:00
Justin Ruggles
16e54ac725 (e)ac3 parser: set duration instead of frame_size 2012-02-20 15:08:40 -05:00
Justin Ruggles
c7f3f1c91e flac parser: set duration instead of frame_size 2012-02-20 15:08:40 -05:00
Justin Ruggles
e9cda85351 avcodec: add duration field to AVCodecParserContext 
This will allow parsers to export the duration of the current frame being
output, if known, instead of using AVCodecContext.frame_size.
 2012-02-20 15:08:40 -05:00
Paul B Mahol
0996f406c4 pnmdec: remove useless .pix_fmts 
Signed-off-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
 2012-02-20 13:46:57 -05:00
Justin Ruggles
e00959a9b1 libmp3lame: support float and s32 sample formats 2012-02-20 12:32:31 -05:00
Justin Ruggles
e232225276 libmp3lame: renaming, rearrangement, alignment, and comments 2012-02-20 12:32:30 -05:00
Justin Ruggles
232e16dd02 libmp3lame: use the LAME default bit rate 
Also, only set bit rate for CBR.
 2012-02-20 12:32:30 -05:00
Justin Ruggles
1f516c0451 libmp3lame: use avpriv_mpegaudio_decode_header() for output frame parsing 2012-02-20 12:32:30 -05:00
Justin Ruggles
e3d2c89e9d libmp3lame: cosmetics: remove some pointless comments 2012-02-20 12:32:30 -05:00
Justin Ruggles
bf909fc456 libmp3lame: convert some debugging code to av_dlog() 
also remove unneeded commented-out full frame data debugging
 2012-02-20 12:32:30 -05:00
Justin Ruggles
469d2a8e8e libmp3lame: remove outdated comment. 
We now require at least libmp3lame 3.98.3.
lame_encode_buffer_interleaved() still doesn't work for mono, but it does not
"die"; it just expects a stereo interleaved buffer.
 2012-02-20 12:32:30 -05:00
Justin Ruggles
35cfd7d09c libmp3lame: do not set coded_frame->key_frame. 
it is already set in avcodec_alloc_frame()
 2012-02-20 12:32:30 -05:00
Justin Ruggles
8dad25ebf7 libmp3lame: improve error handling in MP3lame_encode_init() 2012-02-20 12:32:30 -05:00
Justin Ruggles
310c372e12 libmp3lame: remove unneeded 'stereo' field from Mp3AudioContext 2012-02-20 12:32:30 -05:00
Anton Khirnov
55fd7da107 pcxenc: switch to encode2(). 2012-02-20 07:50:44 +01:00
Anton Khirnov
fb11e22fda sgienc: switch to encode2(). 2012-02-20 07:50:44 +01:00
Anton Khirnov
61eaf45c99 targaenc: switch to encode2(). 2012-02-20 07:50:44 +01:00
Anton Khirnov
1e742ea3c8 targaenc: don't modify input frame. 
Set the pict_type/key_frame properties on the coded picture.
 2012-02-20 07:50:44 +01:00
Anton Khirnov
3227770092 roqvideoenc: switch to encode2(). 2012-02-20 07:50:44 +01:00
Anton Khirnov
4fd7cfef0b roqvideoenc: add const qualifier to the input frame. 2012-02-20 07:50:44 +01:00
Anton Khirnov
1ea5755046 pnmenc: switch to encode2(). 2012-02-20 07:50:44 +01:00
Anton Khirnov
bc9c70e5a3 huffyuv: switch to encode2(). 2012-02-20 07:50:44 +01:00
Anton Khirnov
2abee9be82 v410enc: switch to encode2(). 2012-02-20 07:50:44 +01:00
Anton Khirnov
11505f39e1 zmbvenc: switch to encode2(). 2012-02-20 07:50:44 +01:00
Anton Khirnov
8c8c7b5e37 zmbvenc: move header writing to the end of encode_frame(). 
This makes switching to encode2() simpler, because it allows us to know
exactly how large should the output buffer be before we start writing
into it.
 2012-02-20 07:50:44 +01:00
Ronald S. Bultje
8fb26950ed h264: don't use redzone in loopfilter on win64. 
Red zone usage is not allowed in the Win64 ABI.
 2012-02-19 15:31:03 -08:00
Luca Barbato
aac63cef20 examples: unbreak compilation 
Update api so it will compile again.
 2012-02-19 19:10:28 +01:00
Paul B Mahol
e04ca1d4ce ttadec: cosmetics: reindent 
Signed-off-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
 2012-02-19 13:04:10 -05:00
Aneesh Dogra
566df2eea2 sunrast: use RLE trigger macro inplace of the hard coded value. 
Signed-off-by: Anton Khirnov <anton@khirnov.net>
 2012-02-19 15:55:57 +01:00
Aneesh Dogra
1a58daed0f sunrastenc: set keyframe flag for the output packet. 
Signed-off-by: Anton Khirnov <anton@khirnov.net>
 2012-02-19 15:54:27 +01:00
Anton Khirnov
445a7d48b1 mpegvideo_enc: switch to encode2(). 2012-02-19 08:26:30 +01:00
Anton Khirnov
5455384219 mpegvideo_enc: force encoding delay of at least 1 frame when low_delay=0 
This allows the following commit to extrapolate better dts for the first
frame. Pts difference between the first two frames is reused as the
difference between pts and dts of the first frame.
 2012-02-19 08:22:48 +01:00
Kostya Shishkov
6d702dc072 proresenc: force bitrate not to exceed given limit 
Apple ProRes Format Specifications mentions target data size for every frame,
so make sure frame meets it. This also allows encoder to demand much smaller
packet sizes for output.
 2012-02-18 18:34:01 +01:00
Ronald S. Bultje
c742ab4e81 vc1parse: call vc1_init_common(). 
The parser uses VLC tables initialized in vc1_common_init(), therefore
we should call this function on parser init also.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-18 09:16:39 -08:00
Ronald S. Bultje
9d3050d3e9 wma: don't return 0 on invalid packets. 
Return 0 means "please return the same data again", i.e. it causes an
infinite loop. Instead, return an error.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-18 09:13:38 -08:00
Ronald S. Bultje
74699ac8c8 mjpegb: don't return 0 at the end of frame decoding. 
Return 0 indicates "please return the same data again", i.e. it causes
an infinite loop. Instead, return that we consumed the buffer if we
finished decoding succesfully, or return an error if an error occurred.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
 2012-02-18 08:41:45 -08:00
Ronald S. Bultje
b1af4e9c27 vp8dsp: split long line. 
Signed-off-by: Martin Storsjö <martin@martin.st>
 2012-02-18 02:32:09 +02:00