* using MOVQ_WONE/MOVQ_BFE as two instruction instead of static memory value access
as its always faster
* PAVGB_MMX macro is using now mm6 -> mm7 is unmodified
* replaced original pixels_xy2_mmx with new faster and equal implementation
* replaced usage of mm7 for other then ZERO contstant in _rnd & _avg file with mm6
Originally committed as revision 632 to svn://svn.ffmpeg.org/ffmpeg/trunk
* calculating average in parallel to use two MMX units
* disabled xy2 put code - it's producing rouding errors
Originally committed as revision 626 to svn://svn.ffmpeg.org/ffmpeg/trunk
the new code
* added PAVGP macros for parallel processing to safe few more cycles on
celerons
Originally committed as revision 625 to svn://svn.ffmpeg.org/ffmpeg/trunk
and being compiled in the same way as _avg.h
* PAVG_MMX macros accept also output parameter
* implemented faster put_pixels_xy2, but it has slightly smaller precission.
But there is not visible difference in the image quality - might be eventualy
easily switched back (#if 0 #endif)- please check
Originally committed as revision 624 to svn://svn.ffmpeg.org/ffmpeg/trunk
rewrote quantizer
fixed bias (+10% compression/quality for h263 like codecs)
qscale=1 support
mpeg1 intra frames looks far less blocky
added codec_id field
Originally committed as revision 423 to svn://svn.ffmpeg.org/ffmpeg/trunk
mpeg4 b-frame enoding support
removed old, out-commented ratecontrol
reuse motion compensation code between encoding & decoding
prefix newly added global functions with ff_ to reduce namespace polution
b-frame ME (unfinished, but working)
added some comments to mpegvideo.h
do MC on encoding only once if possible
bugs? ;)
Originally committed as revision 403 to svn://svn.ffmpeg.org/ffmpeg/trunk
SAD functions rewritten (8x8 support & MMX2 optimizations)
HQ inter/intra decission
msmpeg4 encoding bugfix (MV where too long)
Originally committed as revision 362 to svn://svn.ffmpeg.org/ffmpeg/trunk
sure it works ok. Also it's slow, so use it only when you _really_ need to
measure quality.
- Fix libavcodec Makefile to enable profiling.
Originally committed as revision 314 to svn://svn.ffmpeg.org/ffmpeg/trunk
- introduced MOVQ_WONE, MOVQ_WTWO
* unrolled put_pixels_clamped_mmx
* added .baling where the jump address would be compiled on odd address
* added some testing code to disable all these routines (for time measurement)
Originally committed as revision 308 to svn://svn.ffmpeg.org/ffmpeg/trunk
this includes the range for quantized dct coefficients in dct_quantize() (-1023...1023)
Originally committed as revision 291 to svn://svn.ffmpeg.org/ffmpeg/trunk
mmx & mmx2 quantizer
c dct permutation bugfix
dont copy input on intra only encodings if it can be avoided
dont draw edges on intra only stuff
Originally committed as revision 281 to svn://svn.ffmpeg.org/ffmpeg/trunk
dequantizers skip trailing zeros
msmpeg4 non-intra decoder has its dequantizer "build in" now
Originally committed as revision 260 to svn://svn.ffmpeg.org/ffmpeg/trunk
Zdenek Kabelac
Fabrice Bellard
Michael Niedermayer
Juanjo
Arpi
Nick Kurshev