bf29ed5dbc
Cosmetics: 2->4 spaces and some braces
...
Originally committed as revision 5940 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-06 23:15:32 +00:00
fb2d9140e4
Correct MC for B-frames and some improvements (not 100% complete though)
...
Originally committed as revision 5938 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-06 03:44:31 +00:00
78cbfc0c88
Bicubic interpolation requires two additional pixels for block
...
Originally committed as revision 5937 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-06 03:42:39 +00:00
9bd0d79fcb
printf -> av_log, only occurs in DEBUG cases
...
Originally committed as revision 5934 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-04 22:51:20 +00:00
c489d907e1
Yet another typo
...
Originally committed as revision 5933 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-04 19:11:12 +00:00
532877894b
Fix a stupid typo and another error, thanks to Emanuele Giaquinta <exg@gentoo.org> for pointing out the issue and the patch
...
Originally committed as revision 5932 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-04 19:03:01 +00:00
99f649a265
100l: intensity compensation must be also done in 4-MV mode
...
Originally committed as revision 5928 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-04 03:31:01 +00:00
7f624e80ef
standalone snow dsputil init
...
Originally committed as revision 5919 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 16:19:13 +00:00
73e4ff9de9
Almost cosmetic changes in dsputil_init_ppc and vorbis_inverse_coupling_altivec:
...
- whitespace consistency 2->4 [dsputil_init_ppc]
- reorder struct initialization with dsputil_init_altivec [dsputil_init_ppc]
- use casts instead of temp variables [vorbis_inverse_coupling_altivec]
Originally committed as revision 5918 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 16:13:50 +00:00
7c2a270b70
Minor fix
...
Originally committed as revision 5916 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 14:37:58 +00:00
8047fe72a9
avoid alignment hacks, luckly gcc does the right thing on arches different from x86
...
Originally committed as revision 5915 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 13:21:13 +00:00
3b0df5254e
vorbis_inverse_coupling_altivec
...
Originally committed as revision 5914 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 13:19:32 +00:00
90530cec20
Cleanup
...
Originally committed as revision 5913 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 13:00:37 +00:00
cd035a6051
10l, vorbis_inverse_coupling_sse() was really 3dnow
...
Originally committed as revision 5903 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 07:09:29 +00:00
32e8763103
another div -> fastdiv, another 2% faster vorbis.
...
Originally committed as revision 5902 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 06:07:13 +00:00
c198597864
Looks like to be a good time to re-enable WMV3 decoder
...
Originally committed as revision 5901 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 05:15:50 +00:00
48d3fca717
Don't try to decode P-frames from old WMV3 variant until their format is figured
...
Originally committed as revision 5900 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 05:09:36 +00:00
481fb636b1
Some AltiVec optimizations for VC-1
...
Originally committed as revision 5899 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 05:02:31 +00:00
2dac4acfc0
sse & sse2 implementations of vorbis channel coupling.
...
9% faster vorbis (on a K8).
Originally committed as revision 5898 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 03:18:47 +00:00
7bf0049623
int16_t is faster than int_fast16_t for division.
...
2% faster vorbis (on a K8).
Originally committed as revision 5897 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-03 02:18:07 +00:00
27dc20a071
set supported pixel formats correctly
...
Originally committed as revision 5896 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-02 22:48:50 +00:00
832e6e9071
avoid possible segfault situations
...
Originally committed as revision 5895 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-02 21:37:38 +00:00
50b35f7aa6
Cosmetics
...
Originally committed as revision 5890 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-02 11:18:10 +00:00
9c05096967
Forgot to update LIBAVCODEC_VERSION_INT
...
Originally committed as revision 5877 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-08-01 04:00:14 +00:00
6c9ce82f74
remove libtheora wrapper
...
Originally committed as revision 5873 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-31 20:56:15 +00:00
a70c27e813
Fix overflows in bicubic interpolation.
...
Originally committed as revision 5868 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-31 03:32:59 +00:00
8295eb302f
Use bicubic MC (should also remove those ringing artifacts) when needed
...
Originally committed as revision 5864 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-30 03:50:31 +00:00
5ce425f7e6
Use functions from DSPContext in vc1.c
...
Originally committed as revision 5863 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-30 03:47:26 +00:00
64db55aedb
New functions in DSPContext for VC-1 decoding
...
Originally committed as revision 5862 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-30 03:45:02 +00:00
9cf94714d0
Remove spurious references to Ogg Theora encoder, such a thing never existed.
...
Originally committed as revision 5858 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-29 20:47:47 +00:00
595e7bd940
some MMX optimizations for the CAVS decoder
...
Originally committed as revision 5846 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-29 08:45:33 +00:00
09be55df9c
Set MpegEncContext->mspel flag (here it indicates that bicubic MC will be use)
...
Originally committed as revision 5845 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-29 03:11:14 +00:00
ffb9a8b1b6
Support range reduced frames
...
Originally committed as revision 5844 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-29 03:07:30 +00:00
138712fe43
Support grayscale decoding.
...
Originally committed as revision 5843 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-29 03:05:29 +00:00
0154bae137
Add rounding for overlap filters
...
Originally committed as revision 5834 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-28 03:27:37 +00:00
ae08a5631e
10l in r5768 (broke mbaff)
...
Originally committed as revision 5831 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-27 20:37:21 +00:00
1ea501aff0
Remove sws_global_init() from swscaler emulation
...
Originally committed as revision 5830 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-27 12:31:31 +00:00
538389c981
Fix FSF postal address.
...
Originally committed as revision 5829 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-26 01:12:26 +00:00
2793096fe3
Fix a corner case in sws emulation
...
Originally committed as revision 5827 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-25 15:23:44 +00:00
5df6889364
Some B-frames support (parsing and decoding only, no motion compesation is done)
...
Originally committed as revision 5820 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-24 04:24:53 +00:00
c39e3c6fbb
Some fixes for decode_p_mb()
...
Originally committed as revision 5819 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-24 04:20:49 +00:00
9f82cdc7b5
Remove obsolete and misleading warning.
...
Originally committed as revision 5816 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-23 20:45:20 +00:00
bc185f72c0
Do not crash when the first frame is not a keyframe (and thus none of the
...
necessary initialization has been done yet).
Originally committed as revision 5815 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-23 18:22:29 +00:00
83b3df83d3
Correct edge MC for chroma
...
Originally committed as revision 5813 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-23 04:57:36 +00:00
230e9d4ca7
Replace code for clipping MV vectors (which is wrong to use here) with clipping source coords.
...
Originally committed as revision 5811 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-22 03:57:53 +00:00
4b94465927
Don't do non-uniform quantization on zero coeffs.
...
Originally committed as revision 5810 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-22 03:54:00 +00:00
27ed1a0dee
1000l: scale DC predictors instead of predicted value.
...
Originally committed as revision 5808 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-21 04:06:06 +00:00
92ad0d9d3b
Set DC = 0 for skipped MB as it interferes DC prediction
...
Originally committed as revision 5807 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-21 04:03:56 +00:00
b53921aabd
Set correctly quantizer and transform mode when parsing frame header.
...
Originally committed as revision 5806 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-21 03:58:48 +00:00
d9cf32d920
avoid 4 +
...
Originally committed as revision 5799 to svn://svn.ffmpeg.org/ffmpeg/trunk
2006-07-19 23:51:11 +00:00
Luca Barbato
Kostya Shishkov
Mike Melanson
Loren Merritt
Michael Niedermayer
Måns Rullgård
Diego Biurrun
Stefan Gehrer
Luca Abeni
Reimar Döffinger