platform/upstream/libav.git
16 years agoDo not lose the '\n'.
Michael Niedermayer [Fri, 26 Sep 2008 03:24:05 +0000 (03:24 +0000)]
Do not lose the '\n'.

Originally committed as revision 15424 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agosync to latest mingw api, remove unneeded struct definitions
Baptiste Coudurier [Fri, 26 Sep 2008 03:20:35 +0000 (03:20 +0000)]
sync to latest mingw api, remove unneeded struct definitions

Originally committed as revision 15423 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years ago2 in 1, an ASS and SSA demuxer.
Michael Niedermayer [Fri, 26 Sep 2008 02:19:21 +0000 (02:19 +0000)]
2 in 1, an ASS and SSA demuxer.
Tested with -scodec copy on all ASS & SSA samples that i found on mphq (exactly 2).

Originally committed as revision 15422 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoinclude sys/select.h, fix freebsd 5.5 compilation
Baptiste Coudurier [Fri, 26 Sep 2008 02:13:38 +0000 (02:13 +0000)]
include sys/select.h, fix freebsd 5.5 compilation

Originally committed as revision 15421 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoonly include sys/select.h if present, fix mingw compilation
Baptiste Coudurier [Fri, 26 Sep 2008 02:12:37 +0000 (02:12 +0000)]
only include sys/select.h if present, fix mingw compilation

Originally committed as revision 15420 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoPrevent image2 from complaining about non monotone timestamps as
Michael Niedermayer [Fri, 26 Sep 2008 02:07:56 +0000 (02:07 +0000)]
Prevent image2 from complaining about non monotone timestamps as
there are no timestamps in "raw" image2.

Originally committed as revision 15419 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAvoid out of picture accesses when CODEC_FLAG_EMU_EDGE is set.
Michael Niedermayer [Thu, 25 Sep 2008 21:44:07 +0000 (21:44 +0000)]
Avoid out of picture accesses when CODEC_FLAG_EMU_EDGE is set.
Fixes issue427.

Originally committed as revision 15418 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoDo instead of divising a float by a constant, multiply by its inverse
Vitor Sessak [Thu, 25 Sep 2008 20:03:57 +0000 (20:03 +0000)]
Do instead of divising a float by a constant, multiply by its inverse

Originally committed as revision 15417 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoEnsure that no non-existing frame is used.
Michael Niedermayer [Thu, 25 Sep 2008 19:25:28 +0000 (19:25 +0000)]
Ensure that no non-existing frame is used.
Fixes issue624.

Originally committed as revision 15416 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoImplement av_nearer_q() and av_find_nearest_q_idx() functions.
Stefano Sabatini [Thu, 25 Sep 2008 19:23:13 +0000 (19:23 +0000)]
Implement av_nearer_q() and av_find_nearest_q_idx() functions.

Originally committed as revision 15415 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoDo not free sps & pps on width/height changes.
Michael Niedermayer [Thu, 25 Sep 2008 19:01:39 +0000 (19:01 +0000)]
Do not free sps & pps on width/height changes.
Fixes issue557.

Originally committed as revision 15414 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoDo not use the generic "alloc missing references" code for h.264 as it does
Michael Niedermayer [Thu, 25 Sep 2008 14:34:14 +0000 (14:34 +0000)]
Do not use the generic "alloc missing references" code for h.264 as it does
not work correctly in that case.
Fixes issue652.

Originally committed as revision 15413 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoInitialize next/last_picture* as the generic code does not always
Michael Niedermayer [Thu, 25 Sep 2008 14:29:35 +0000 (14:29 +0000)]
Initialize next/last_picture* as the generic code does not always
set them to sane values.

Originally committed as revision 15412 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoPrint which slices are IDR not just the slice type.
Michael Niedermayer [Thu, 25 Sep 2008 12:04:25 +0000 (12:04 +0000)]
Print which slices are IDR not just the slice type.

Originally committed as revision 15411 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix CAVLC+8x8+MBAFF loop filter.
Michael Niedermayer [Thu, 25 Sep 2008 00:41:49 +0000 (00:41 +0000)]
Fix CAVLC+8x8+MBAFF loop filter.
Fixes FREXT02_JVC_C.

Originally committed as revision 15410 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agocosmetics, remove GXF: in log messages since it is present in context
Baptiste Coudurier [Wed, 24 Sep 2008 23:40:55 +0000 (23:40 +0000)]
cosmetics, remove GXF: in log messages since it is present in context

Originally committed as revision 15409 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoskip invalid audio samples in gxf packets
Baptiste Coudurier [Wed, 24 Sep 2008 23:31:04 +0000 (23:31 +0000)]
skip invalid audio samples in gxf packets

Originally committed as revision 15408 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoField -> MBAFF direct mode can use data from fields that are not
Michael Niedermayer [Wed, 24 Sep 2008 23:24:24 +0000 (23:24 +0000)]
Field -> MBAFF direct mode can use data from fields that are not
in the reference list. Follow the spec and no comment on the sanity of this
design ...
Fixes HPCAMAPALQ_BRCM_B

Originally committed as revision 15407 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoinclude sys/select.h to get select, according to posix 2001,
Baptiste Coudurier [Wed, 24 Sep 2008 22:11:53 +0000 (22:11 +0000)]
include sys/select.h to get select, according to posix 2001,
fix compilation on freebsd 5.5

Originally committed as revision 15406 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoinclude sys/select.h instead of unistd.h to get select,
Baptiste Coudurier [Wed, 24 Sep 2008 22:08:50 +0000 (22:08 +0000)]
include sys/select.h instead of unistd.h to get select,
according to posix 2001, fix compilation on freebsd 5.5

Originally committed as revision 15405 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMove constant multiplication out of the loop
Vitor Sessak [Wed, 24 Sep 2008 20:31:24 +0000 (20:31 +0000)]
Move constant multiplication out of the loop

Originally committed as revision 15404 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCosmetics: move convolve() together with the other DSP functions
Vitor Sessak [Wed, 24 Sep 2008 20:05:22 +0000 (20:05 +0000)]
Cosmetics: move convolve() together with the other DSP functions

Originally committed as revision 15403 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSimplify: move division by constant off the loop
Vitor Sessak [Wed, 24 Sep 2008 20:03:58 +0000 (20:03 +0000)]
Simplify: move division by constant off the loop

Originally committed as revision 15402 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agostop parsing if tag size is wrongly < 8 to avoid infinite loop
Baptiste Coudurier [Wed, 24 Sep 2008 18:55:00 +0000 (18:55 +0000)]
stop parsing if tag size is wrongly < 8 to avoid infinite loop

Originally committed as revision 15401 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAvoid segfault due to not set AVCodec.
Michael Niedermayer [Wed, 24 Sep 2008 17:46:23 +0000 (17:46 +0000)]
Avoid segfault due to not set AVCodec.

Originally committed as revision 15400 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agomerge amr wb case with amr nb and qcelp
Baptiste Coudurier [Wed, 24 Sep 2008 17:17:11 +0000 (17:17 +0000)]
merge amr wb case with amr nb and qcelp

Originally committed as revision 15399 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoonly write trailer when state says so, should fix issue #646
Baptiste Coudurier [Wed, 24 Sep 2008 17:11:46 +0000 (17:11 +0000)]
only write trailer when state says so, should fix issue #646

Originally committed as revision 15398 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agodo not override sample rate for amr nor qcelp, should fix some qcelp files at 16000hz
Baptiste Coudurier [Wed, 24 Sep 2008 17:09:59 +0000 (17:09 +0000)]
do not override sample rate for amr nor qcelp, should fix some qcelp files at 16000hz

Originally committed as revision 15397 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCorrect wrong lower limit and condition used in APE decoder
Kostya Shishkov [Wed, 24 Sep 2008 12:45:28 +0000 (12:45 +0000)]
Correct wrong lower limit and condition used in APE decoder

Originally committed as revision 15396 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoASF stands for "advanced stream format" not "adaptive ...", changing it to "ASF".
Michael Niedermayer [Wed, 24 Sep 2008 12:03:25 +0000 (12:03 +0000)]
ASF stands for "advanced stream format" not "adaptive ...", changing it to "ASF".

Originally committed as revision 15395 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMerge the 2 lpc loops.
Michael Niedermayer [Tue, 23 Sep 2008 23:16:48 +0000 (23:16 +0000)]
Merge the 2 lpc loops.
This changes the output by (stddev:    0.21 PSNR:109.51, file:ddinterview28.ra),
the changes are due to float rounding inaccuracies and do not happen if doubles
are used.

Originally committed as revision 15394 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFactorize lpc code slightly.
Michael Niedermayer [Tue, 23 Sep 2008 22:49:22 +0000 (22:49 +0000)]
Factorize lpc code slightly.

Originally committed as revision 15393 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAdd MIPS to list of architectures requiring PIC flags.
Diego Biurrun [Tue, 23 Sep 2008 11:10:38 +0000 (11:10 +0000)]
Add MIPS to list of architectures requiring PIC flags.
Prior to gcc version 4.3 PIC was the default, now it needs to be enabled
explicitly. Passing the flag to older gcc versions does not hurt.

Originally committed as revision 15392 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCosmetics: reindent after last commit
Kostya Shishkov [Tue, 23 Sep 2008 08:45:46 +0000 (08:45 +0000)]
Cosmetics: reindent after last commit

Originally committed as revision 15391 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAdd RLE4 and RLE8 decoding support for BMP
Kostya Shishkov [Tue, 23 Sep 2008 08:45:12 +0000 (08:45 +0000)]
Add RLE4 and RLE8 decoding support for BMP

Originally committed as revision 15390 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAdd support for 1-bit, 4-bit, 8-bit and some 16-bit raw BMP
Kostya Shishkov [Tue, 23 Sep 2008 08:38:32 +0000 (08:38 +0000)]
Add support for 1-bit, 4-bit, 8-bit and some 16-bit raw BMP

Originally committed as revision 15389 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCalculate line size correctly for bit depths < 8
Kostya Shishkov [Tue, 23 Sep 2008 08:32:18 +0000 (08:32 +0000)]
Calculate line size correctly for bit depths < 8

Originally committed as revision 15388 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAdd myself to BMP format maintainers (with permission from Mans on IRC)
Kostya Shishkov [Tue, 23 Sep 2008 08:26:59 +0000 (08:26 +0000)]
Add myself to BMP format maintainers (with permission from Mans on IRC)

Originally committed as revision 15387 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAdd support for RVTR fourCC
Vitor Sessak [Tue, 23 Sep 2008 06:24:00 +0000 (06:24 +0000)]
Add support for RVTR fourCC

Originally committed as revision 15386 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoA few empty lines for readability.
Michael Niedermayer [Mon, 22 Sep 2008 20:46:52 +0000 (20:46 +0000)]
A few empty lines for readability.

Originally committed as revision 15385 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAnother try to fix temporal direct mode references.
Michael Niedermayer [Mon, 22 Sep 2008 20:43:35 +0000 (20:43 +0000)]
Another try to fix temporal direct mode references.
Fixes at least
CAPAMA3_Sand_F.264
CVMAPAQP3_Sony_E.jsv

Originally committed as revision 15384 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agospelling cosmetics
Diego Biurrun [Mon, 22 Sep 2008 05:40:52 +0000 (05:40 +0000)]
spelling cosmetics

Originally committed as revision 15383 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAltiVec detection support for OpenBSD, patch by Brad, brad comstyle com.
Brad [Mon, 22 Sep 2008 05:38:50 +0000 (05:38 +0000)]
AltiVec detection support for OpenBSD, patch by Brad, brad comstyle com.

Originally committed as revision 15382 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix all the recent problems Justin had run into with the AC3 parser,
Michael Niedermayer [Mon, 22 Sep 2008 03:00:26 +0000 (03:00 +0000)]
Fix all the recent problems Justin had run into with the AC3 parser,
including issue 459.

Originally committed as revision 15381 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix ref_shift so that it is correct for more/all? MBAFF/PAFF mixes.
Michael Niedermayer [Mon, 22 Sep 2008 02:32:22 +0000 (02:32 +0000)]
Fix ref_shift so that it is correct for more/all? MBAFF/PAFF mixes.
Fixes at least MR9_BT_B.

Originally committed as revision 15380 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoindent
Michael Niedermayer [Sun, 21 Sep 2008 23:51:36 +0000 (23:51 +0000)]
indent

Originally committed as revision 15379 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCorrect comment in the direct mode code.
Michael Niedermayer [Sun, 21 Sep 2008 23:49:53 +0000 (23:49 +0000)]
Correct comment in the direct mode code.
(note, yes this is unrelated to the previous simplification, the
 code always behaved like it is documented now.)

Originally committed as revision 15378 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSimplify initialization of a few direct mode related variables.
Michael Niedermayer [Sun, 21 Sep 2008 23:46:59 +0000 (23:46 +0000)]
Simplify initialization of a few direct mode related variables.

Originally committed as revision 15377 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoV4L2 support for OpenBSD, patch by Brad, brad comstyle com
Brad [Sun, 21 Sep 2008 22:47:52 +0000 (22:47 +0000)]
V4L2 support for OpenBSD, patch by Brad, brad comstyle com

Originally committed as revision 15376 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake MBAFF temporal direct mode closer to the spec.
Michael Niedermayer [Sun, 21 Sep 2008 02:39:09 +0000 (02:39 +0000)]
Make MBAFF temporal direct mode closer to the spec.
Fixes at least:
CAMANL2_TOSHIBA_B
CVMANL2_TOSHIBA_B
camp_mot_mbaff0_full

Originally committed as revision 15375 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMove direct_dist_scale_factor() and direct_ref_list_init() after fill_mbaff_ref_list()
Michael Niedermayer [Sun, 21 Sep 2008 02:23:51 +0000 (02:23 +0000)]
Move direct_dist_scale_factor() and direct_ref_list_init() after fill_mbaff_ref_list()
as future changes will need the mbaff lists inited.

Originally committed as revision 15374 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSet poc for field references in MBAFF.
Michael Niedermayer [Sun, 21 Sep 2008 02:19:45 +0000 (02:19 +0000)]
Set poc for field references in MBAFF.

Originally committed as revision 15373 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRename check_func2 to check_func_headers.
Diego Biurrun [Sat, 20 Sep 2008 16:17:27 +0000 (16:17 +0000)]
Rename check_func2 to check_func_headers.

Originally committed as revision 15372 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoLog an error message in case of invalid number of channels.
Stefano Sabatini [Sat, 20 Sep 2008 11:13:14 +0000 (11:13 +0000)]
Log an error message in case of invalid number of channels.

Originally committed as revision 15371 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRemove fixme that is based on the assumption of !direct_8x8_inference && MBAFF.
Michael Niedermayer [Fri, 19 Sep 2008 18:53:31 +0000 (18:53 +0000)]
Remove fixme that is based on the assumption of !direct_8x8_inference && MBAFF.

Originally committed as revision 15370 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRemove check and error message for "MBAFF + !direct_8x8_inference is not implemented"
Michael Niedermayer [Fri, 19 Sep 2008 18:49:17 +0000 (18:49 +0000)]
Remove check and error message for "MBAFF + !direct_8x8_inference is not implemented"
because this mode does not exist, H.264-2007 says "When frame_mbs_only_flag is
equal to 0, direct_8x8_inference_flag shall be equal to 1."

Originally committed as revision 15369 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAvoid undefined behavior for removing elements that were not in the tree.
Michael Niedermayer [Fri, 19 Sep 2008 12:41:12 +0000 (12:41 +0000)]
Avoid undefined behavior for removing elements that were not in the tree.

Originally committed as revision 15368 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoEnsure that muxer and encoder aspect ratios match. So no
Michael Niedermayer [Fri, 19 Sep 2008 12:37:38 +0000 (12:37 +0000)]
Ensure that muxer and encoder aspect ratios match. So no
invalid files are generated.

Originally committed as revision 15367 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSupport comments in preset files.
Michael Niedermayer [Thu, 18 Sep 2008 21:07:11 +0000 (21:07 +0000)]
Support comments in preset files.

Originally committed as revision 15366 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake pps/sps scaling_matrix initialization closer to the spec.
Michael Niedermayer [Thu, 18 Sep 2008 16:36:19 +0000 (16:36 +0000)]
Make pps/sps scaling_matrix initialization closer to the spec.
This does not affect any of the conformance streams.

Originally committed as revision 15365 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoZero 'idx' for each iteration of the coupling gain loop and as it is only used
Robert Swain [Thu, 18 Sep 2008 16:02:51 +0000 (16:02 +0000)]
Zero 'idx' for each iteration of the coupling gain loop and as it is only used
within this loop, we can move its declaration there too. This fixes bitstream
desync when decoding streams containing CCEs.

Based on a patch by Alex Converse (alex converse gmail com)

Originally committed as revision 15364 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCosmetics: missing brackets
Robert Swain [Thu, 18 Sep 2008 15:48:37 +0000 (15:48 +0000)]
Cosmetics: missing brackets

Originally committed as revision 15363 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCast the return value of get_bits() to signed integer such that the subtraction
Alex Converse [Thu, 18 Sep 2008 15:35:08 +0000 (15:35 +0000)]
Cast the return value of get_bits() to signed integer such that the subtraction
is conducted as a signed calculation.

Patch by Alex Converse (alex converse gmail com)

Originally committed as revision 15362 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoEnsure that extradata is not parsed twice.
Michael Niedermayer [Thu, 18 Sep 2008 15:13:15 +0000 (15:13 +0000)]
Ensure that extradata is not parsed twice.
Fixes FRExt1_Panasonic.

Originally committed as revision 15361 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoadd newline at the end of message
Kostya Shishkov [Thu, 18 Sep 2008 14:08:44 +0000 (14:08 +0000)]
add newline at the end of message

Originally committed as revision 15360 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoZMBV encoder forgot to fill coded frame pointer
Kostya Shishkov [Thu, 18 Sep 2008 13:17:44 +0000 (13:17 +0000)]
ZMBV encoder forgot to fill coded frame pointer

Originally committed as revision 15359 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agosmall wording fixes
Diego Biurrun [Thu, 18 Sep 2008 08:10:12 +0000 (08:10 +0000)]
small wording fixes

Originally committed as revision 15358 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAdd fourcc for MS RLE4
Kostya Shishkov [Thu, 18 Sep 2008 05:25:29 +0000 (05:25 +0000)]
Add fourcc for MS RLE4

Originally committed as revision 15357 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFactorize out code used for MS RLE format decoding in different decoders.
Kostya Shishkov [Thu, 18 Sep 2008 05:20:54 +0000 (05:20 +0000)]
Factorize out code used for MS RLE format decoding in different decoders.

Originally committed as revision 15356 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRemove duplication of TRANSPOSE8 macro, also removing some redundant load/stores
David Conrad [Thu, 18 Sep 2008 03:08:33 +0000 (03:08 +0000)]
Remove duplication of TRANSPOSE8 macro, also removing some redundant load/stores

Originally committed as revision 15355 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake a separate macro for writing an entire block from registers to memory
David Conrad [Thu, 18 Sep 2008 03:08:13 +0000 (03:08 +0000)]
Make a separate macro for writing an entire block from registers to memory

Originally committed as revision 15354 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFactorize SSE2_(Row|Column)_IDCT into one macro
David Conrad [Thu, 18 Sep 2008 03:07:41 +0000 (03:07 +0000)]
Factorize SSE2_(Row|Column)_IDCT into one macro

Originally committed as revision 15353 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoUniform *ToY and *ToUV function signatures
Luca Barbato [Thu, 18 Sep 2008 00:26:37 +0000 (00:26 +0000)]
Uniform *ToY and *ToUV function signatures

Originally committed as revision 27636 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agoSplit mono2Y in monowhite and monoblack
Luca Barbato [Thu, 18 Sep 2008 00:24:32 +0000 (00:24 +0000)]
Split mono2Y in monowhite and monoblack

Originally committed as revision 27635 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agoFactorize unit32_t* casts for palette pointer
Luca Barbato [Thu, 18 Sep 2008 00:22:53 +0000 (00:22 +0000)]
Factorize unit32_t* casts for palette pointer

Originally committed as revision 27634 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agorevert r14983, value is not sample aspect ratio, it is display aspect ratio
Baptiste Coudurier [Wed, 17 Sep 2008 23:14:48 +0000 (23:14 +0000)]
revert r14983, value is not sample aspect ratio, it is display aspect ratio

Originally committed as revision 15352 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCosmetics: remove excessive blank lines
David Conrad [Wed, 17 Sep 2008 20:52:54 +0000 (20:52 +0000)]
Cosmetics: remove excessive blank lines

Originally committed as revision 15351 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoEnsure MMX/SSE2 VP3 IDCT selection isn't disabled when only Theora is enabled
David Conrad [Wed, 17 Sep 2008 19:49:31 +0000 (19:49 +0000)]
Ensure MMX/SSE2 VP3 IDCT selection isn't disabled when only Theora is enabled

Originally committed as revision 15350 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCosmetics: reindent
David Conrad [Wed, 17 Sep 2008 19:38:44 +0000 (19:38 +0000)]
Cosmetics: reindent

Originally committed as revision 15349 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agolibx264 ffpresets:
Robert Swain [Wed, 17 Sep 2008 19:31:39 +0000 (19:31 +0000)]
libx264 ffpresets:
 * default - x264 CLI default values
 * fastfirstpass - disables options unnecessary for first pass
 * normal - Dark Shikari’s Q55 preset
 * hq - Dark Shikari’s Q75 preset
 * max - all options maxed out (at time of writing)

Originally committed as revision 15348 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRewrite SSE2 VP3 IDCT in inline asm
David Conrad [Wed, 17 Sep 2008 19:30:51 +0000 (19:30 +0000)]
Rewrite SSE2 VP3 IDCT in inline asm

Originally committed as revision 15347 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoRewrite MMX VP3 IDCT in inline asm
David Conrad [Wed, 17 Sep 2008 19:30:33 +0000 (19:30 +0000)]
Rewrite MMX VP3 IDCT in inline asm

Originally committed as revision 15346 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMMX/SSE2 VP3 IDCT are bitexact now that the dequantization matrices are permutated...
David Conrad [Wed, 17 Sep 2008 19:30:03 +0000 (19:30 +0000)]
MMX/SSE2 VP3 IDCT are bitexact now that the dequantization matrices are permutated correctly

Originally committed as revision 15345 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoTake idct permutation into account when initializing dequantization matrices
David Conrad [Wed, 17 Sep 2008 19:29:40 +0000 (19:29 +0000)]
Take idct permutation into account when initializing dequantization matrices

Originally committed as revision 15344 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoUse SAMPLE_FMT_FLT instead of doing the float->int conversion in the
Vitor Sessak [Wed, 17 Sep 2008 19:26:37 +0000 (19:26 +0000)]
Use SAMPLE_FMT_FLT instead of doing the float->int conversion in the
decoder. This changes the output very little and any difference should
be inaudible.

Originally committed as revision 15343 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoSecond-stage gain and Moving Average Predictor codebooks.
Vladimir Voroshilov [Wed, 17 Sep 2008 06:13:07 +0000 (06:13 +0000)]
Second-stage gain and Moving Average Predictor codebooks.

Originally committed as revision 15342 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAnother set of approved changes for G.729: includes, formats structure,
Vladimir Voroshilov [Wed, 17 Sep 2008 06:05:12 +0000 (06:05 +0000)]
Another set of approved changes for G.729: includes, formats structure,
several parts of main decoding routine and initialization code.

Originally committed as revision 15341 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoValidate pulse position and error out if an invalid position is encountered.
Alex Converse [Tue, 16 Sep 2008 15:59:43 +0000 (15:59 +0000)]
Validate pulse position and error out if an invalid position is encountered.

Patch by Alex Converse (alex converse gmail com)

Originally committed as revision 15340 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix segfault with rgb24 and full_internal_chrome due to non existing alpha
Michael Niedermayer [Tue, 16 Sep 2008 03:49:54 +0000 (03:49 +0000)]
Fix segfault with rgb24 and full_internal_chrome due to non existing alpha
byte being written after the array.

Originally committed as revision 27628 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agoOnly print "invalid new backstep" when it is really invalid.
Michael Niedermayer [Tue, 16 Sep 2008 01:49:32 +0000 (01:49 +0000)]
Only print "invalid new backstep" when it is really invalid.
Fixes issue541.

Originally committed as revision 15339 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix regression test failure caused by golomb limit not being considered
Michael Niedermayer [Tue, 16 Sep 2008 01:24:12 +0000 (01:24 +0000)]
Fix regression test failure caused by golomb limit not being considered
in r15334.

Originally committed as revision 15338 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agosync to latest x264 git, field was renamed
Baptiste Coudurier [Tue, 16 Sep 2008 01:13:55 +0000 (01:13 +0000)]
sync to latest x264 git, field was renamed

Originally committed as revision 15337 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoMake decoder more robust by default against broken encoders.
Michael Niedermayer [Tue, 16 Sep 2008 00:31:32 +0000 (00:31 +0000)]
Make decoder more robust by default against broken encoders.
Fixes issue540.

Originally committed as revision 15336 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoFix generated md5, it was wrong for some input lengths.
Michael Niedermayer [Mon, 15 Sep 2008 22:10:28 +0000 (22:10 +0000)]
Fix generated md5, it was wrong for some input lengths.
Fix issue634.

Originally committed as revision 15335 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoCorrect the threshold of get_ur_golomb_jpegls() at which the optimized
Michael Niedermayer [Mon, 15 Sep 2008 20:30:22 +0000 (20:30 +0000)]
Correct the threshold of get_ur_golomb_jpegls() at which the optimized
case is used.
Fix issue245.

Originally committed as revision 15334 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agoAvoid using floating point for calculating filter coefficients.
Michael Niedermayer [Mon, 15 Sep 2008 03:33:09 +0000 (03:33 +0000)]
Avoid using floating point for calculating filter coefficients.

Originally committed as revision 27612 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

16 years agowording/spelling
Diego Biurrun [Sun, 14 Sep 2008 22:46:38 +0000 (22:46 +0000)]
wording/spelling

Originally committed as revision 15333 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agocosmetics: fix indentation after r15321
Justin Ruggles [Sun, 14 Sep 2008 22:27:47 +0000 (22:27 +0000)]
cosmetics: fix indentation after r15321

Originally committed as revision 15332 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agowrite actual min and max frame size to FLAC header. update regression test checksum.
Justin Ruggles [Sun, 14 Sep 2008 22:25:50 +0000 (22:25 +0000)]
write actual min and max frame size to FLAC header. update regression test checksum.

Originally committed as revision 15331 to svn://svn.ffmpeg.org/ffmpeg/trunk

16 years agocosmetics: Fix indentation.
Diego Biurrun [Sun, 14 Sep 2008 22:20:14 +0000 (22:20 +0000)]
cosmetics: Fix indentation.

Originally committed as revision 15330 to svn://svn.ffmpeg.org/ffmpeg/trunk