Commit Graph

282 Commits

Author SHA1 Message Date
Reimar Döffinger
14b8607065 Add support for hard-coded MDCT-related ff_sine_windows tables.
Originally committed as revision 21108 to svn://svn.ffmpeg.org/ffmpeg/trunk
2010-01-09 13:28:04 +00:00
Loren Merritt
b1159ad928 refactor and optimize scalarproduct
29-105% faster apply_filter, 6-90% faster ape decoding on core2
(Any x86 other than core2 probably gets much less, since this is mostly due to ssse3 cachesplit avoidance and I haven't written the full gamut of other cachesplit modes.)
9-123% faster ape decoding on G4.

Originally committed as revision 20739 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-12-05 15:09:10 +00:00
Reimar Döffinger
1ffc6e8327 Add ff_init_ff_cos_tabs function and use it in rdft.c to ensure that the
necessary ff_cos_tabs tables are initialized.
Fixes issue 1507 (QDM2 broken since r20237 without hardcoded tables).

Originally committed as revision 20464 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-11-05 18:29:06 +00:00
Reimar Döffinger
223217746c Pad ff_cos_tabs and ff_sin_tabs so that index n points to the table for n bits.
While this "wastes" up to 2x32 bytes it makes the code slightly simpler and
less confusing.

Originally committed as revision 20449 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-11-04 17:30:23 +00:00
Loren Merritt
e17ccf60fe huffyuv: add some const qualifiers
Originally committed as revision 20290 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-10-18 20:47:25 +00:00
Loren Merritt
2f77923d72 simd add_hfyu_left_prediction
2.2x faster than C on conroe, 3.6x on penryn.
4-6% faster huffyuv decoding if using left or plane mode and yuv

Originally committed as revision 20287 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-10-18 20:10:10 +00:00
Justin Ruggles
fde82ca7e4 Move autocorrelation function from flacenc.c to lpc.c. Also rename the
corresponding dsputil functions and remove their dependency on the FLAC
encoder.
Fixes Issue1486.

Originally committed as revision 20266 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-10-17 21:00:39 +00:00
Alexander Strange
2d4bbdecee Huffyuv: Add missing const to src pointers in dsputil functions.
Originally committed as revision 20259 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-10-16 23:04:41 +00:00
Reimar Döffinger
75df2edbb9 Add support for hardcoded ff_sin_* tables.
Originally committed as revision 20244 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-10-15 18:04:55 +00:00
Reimar Döffinger
4ee726b670 Move/add COSTABLE/SINTABLE macros to dsputil to add extern definitions
for ff_cos_* and ff_sin_* without introducing too much code duplication.

Originally committed as revision 20243 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-10-15 17:55:51 +00:00
Reimar Döffinger
aafd659518 Do not initialize ff_cos_* tables again in rdft_init, they are already
initialized by ff_fft_init and using different code can result in slightly
different values, in addition it crashes when the tables are hardcoded.
On amd64 this slightly changes qdm2 output.

Originally committed as revision 20237 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-10-15 11:50:32 +00:00
Reimar Döffinger
16aec74761 Use hardcoded instead of runtime-calculated ff_cos_* tables if
--enable-hardcoded-tables was used.
Due to the size, the code for the tables is generated at compile time.

Originally committed as revision 20232 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-10-14 21:41:24 +00:00
Nathan Caldwell
73c6f59830 Move HuffYUV left prediction to dsputil.
Patch by Nathan Caldwell, saintdev gmail

Originally committed as revision 20179 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-10-06 16:06:15 +00:00
Måns Rullgård
952e872198 Drop unused args from vector_fmul_add_add, simpify code, and rename
The src3 and step arguments to vector_fmul_add_add() are always zero
and one, respectively.  This removes these arguments from the function,
simplifies the code accordingly, and renames the function to better
match the new operation.

Originally committed as revision 20061 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-27 16:51:54 +00:00
Måns Rullgård
53b572118e Add some dsputil functions useful for AAC decoder
Originally committed as revision 19955 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-22 00:48:41 +00:00
Måns Rullgård
94274b82f6 Allow arch-specific mdct code to request interleaving of cos/sin tables
Originally committed as revision 19939 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-21 02:56:06 +00:00
Måns Rullgård
01b2214758 Merge FFTContext and MDCTContext
Originally committed as revision 19931 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-20 17:30:20 +00:00
Måns Rullgård
68336ea8d4 10l: fix ff_fft_init_arm() prototype
Originally committed as revision 19865 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-15 23:14:55 +00:00
Måns Rullgård
f486321395 Move per-arch fft init bits into the corresponding subdirs
Originally committed as revision 19864 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-15 21:14:14 +00:00
Vitor Sessak
6776061b04 Add two more sizes to ff_sine_windows[] and also pad it with NULLs so
that FF_ELEMS(ff_sine_windows[x]) == 1 << x.

Fix issue 1384.

Originally committed as revision 19862 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-15 16:18:16 +00:00
Måns Rullgård
edc0f5dcba ARM: NEON optimised MDCT
Originally committed as revision 19819 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-11 02:24:22 +00:00
Måns Rullgård
46c32e2654 Prepare for optimised forward MDCT implementations
This adds a function pointer for forward MDCT to FFTContext and
initialises it with the existing C function.  ff_calc_mdct() is
changed to an inline function calling the selected version as
done for other fft/mdct functions.

Originally committed as revision 19818 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-11 02:24:19 +00:00
Måns Rullgård
648d792042 ARM: NEON optimised FFT and MDCT
Vorbis and AC3 ~3x faster.

Parts by Naotoshi Nojiri, naonoj gmail

Originally committed as revision 19806 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-10 08:50:03 +00:00
Reimar Döffinger
c55e5b5fee ff_sine_windows table contains only constant data, too.
Originally committed as revision 19785 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-06 08:56:10 +00:00
Reimar Döffinger
faca56191c The ff_cos_tabs table itself is constant, too, so mark it as such.
Originally committed as revision 19784 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-09-06 08:53:14 +00:00
Måns Rullgård
e78516c6d8 One declaration of mm_flags is enough
Originally committed as revision 19740 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-08-29 16:55:55 +00:00
Måns Rullgård
4e36a5b46f Move declarations of some mmx functions to dsputil_mmx.h
Originally committed as revision 19739 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-08-29 16:55:50 +00:00
Måns Rullgård
5b1b014761 Move DECLARE_ALIGNED_8 definition next to DECLARE_ALIGNED_16
Originally committed as revision 19738 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-08-29 16:55:46 +00:00
Måns Rullgård
c4d6d318cf Make DECLARE_ALIGNED_8 align to 8 bytes, never to 16
The DECLARE_ALIGNED_8 macro is defined to align to 16 bytes instead
the 8 suggested by the name on some CPUs.  None of the uses of this
macro ever need 16-byte alignment, cases which once did having been
changed to always specify 16 bytes explicitly.

Originally committed as revision 19737 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-08-28 23:39:02 +00:00
Vitor Sessak
50e23ae9d3 Mark parameter src of vector_clipf() as const
Originally committed as revision 19729 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-08-27 15:38:59 +00:00
Vitor Sessak
0a68cd876e SSE optimized vector_clipf(). 10% faster TwinVQ decoding.
Originally committed as revision 19728 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-08-27 14:49:36 +00:00
Jason Garrett-Glaser
4f717c69ed idct_dc for VC-1/WMV3 decoder; ~11% faster decoding overall.
Includes mmx2 asm for the various functions.
Note that the actual idct still does not have an x86 SIMD implemtation.
For wmv3 files using regular idct, the decoder just falls back to simple_idct,
since simple_idct_dc doesn't exist (yet).

Originally committed as revision 19204 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-06-16 09:00:55 +00:00
Siarhei Siamashka
7d485f165f Support for getting (i)MDCT output multiplied by a constant scaling factor.
Scaling (i)MDCT output has no runtime overhead and can be used to improve
performance of audio codecs. All the changes are only needed in
'ff_mdct_init' function and slow down initialization a bit.

Originally committed as revision 18855 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-05-16 14:17:08 +00:00
Ramiro Polla
13bd2044d3 mlp: Simplify adressing of state and coeffs arrays for both filters by making
the arrays sequential.

Originally committed as revision 18841 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-05-15 15:34:22 +00:00
Ramiro Polla
bf4f19dc65 mlpdec: Move MLP's filter_channel() to dsputils.
Originally committed as revision 18721 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-04-30 12:30:05 +00:00
David Conrad
3992526b3c Split VC1 loop filter into separate functions for h/v and size
Originally committed as revision 18521 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-04-15 06:22:17 +00:00
David Conrad
0e58865d6e Move VC1 loop filter to DSPContext
Originally committed as revision 18520 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-04-15 06:20:22 +00:00
David Conrad
8013da7364 VC1: add and use avg_no_rnd chroma MC functions
Originally committed as revision 18518 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-04-14 23:56:10 +00:00
David Conrad
c374691b28 Rename put_no_rnd_h264_chroma* to reflect its usage in VC1 only
Originally committed as revision 18517 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-04-14 23:55:39 +00:00
David Conrad
6cecd63005 VC1: Do qpel when needed for both MVs in a B frame
Originally committed as revision 18511 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-04-14 19:26:33 +00:00
Alex Converse
184fcc60b7 Mark src constant in copy_block*().
Originally committed as revision 18506 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-04-14 04:07:35 +00:00
Roman Shaposhnik
3899eb2f65 Making the arrays accomodate an extra intra 8x8 cmp function
Originally committed as revision 17446 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-02-19 00:28:08 +00:00
Aurelien Jacobs
26f3ae828b move vp6_filter_diag4() to a new vp6dsp.c file and use it throught dsputil
Originally committed as revision 17111 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-02-09 23:37:31 +00:00
Loren Merritt
3daa434a40 ff_add_hfyu_median_prediction_mmx2
overall ffvhuff decoding speedup: 28% on core2, 25% on k8.

Originally committed as revision 17059 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-02-08 17:45:30 +00:00
Diego Biurrun
bad5537e2c Use full internal pathname in doxygen @file directives.
Otherwise doxygen complains about ambiguous filenames when files exist
under the same name in different subdirectories.

Originally committed as revision 16912 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-02-01 02:00:19 +00:00
Alex Converse
6860254044 Add the rdft family of transforms (fft/ifft of an all real sequence) to dsputil.
Originally committed as revision 16864 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-01-30 20:15:48 +00:00
Alex Converse
8f05c995fd Clean up FFT related comments.
Originally committed as revision 16721 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-01-22 19:46:23 +00:00
Diego Biurrun
406792e7b0 cosmetics: Remove pointless period after copyright statement non-sentences.
Originally committed as revision 16684 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-01-19 15:46:40 +00:00
Aurelien Jacobs
b250f9c66d Change semantic of CONFIG_*, HAVE_* and ARCH_*.
They are now always defined to either 0 or 1.

Originally committed as revision 16590 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-01-13 23:44:16 +00:00
Diego Biurrun
6a5d31ac25 Fix build: Add intreadwrite.h and bswap.h #includes where necessary.
Originally committed as revision 16556 to svn://svn.ffmpeg.org/ffmpeg/trunk
2009-01-11 22:19:48 +00:00