Originally committed as revision 322 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 321 to svn://svn.ffmpeg.org/ffmpeg/trunk

q-pel mc support
neither is totally bugfree yet though :(

Originally committed as revision 320 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 318 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 317 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

Originally committed as revision 313 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 312 to svn://svn.ffmpeg.org/ffmpeg/trunk

- H.263+ should be able to decode streams with AIC now :)

Originally committed as revision 311 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 310 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

* ignoring error from msmpeg4_decode_block

Originally committed as revision 307 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 306 to svn://svn.ffmpeg.org/ffmpeg/trunk

* simplified indexing

Originally committed as revision 305 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 304 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 303 to svn://svn.ffmpeg.org/ffmpeg/trunk

bug found by Lennert Buytenhek <buytenh@gnu.org>

Originally committed as revision 302 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 301 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 300 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 299 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 298 to svn://svn.ffmpeg.org/ffmpeg/trunk

(makeing some constant arrays global, merging put_bits(), ...)

Originally committed as revision 297 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 296 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 295 to svn://svn.ffmpeg.org/ffmpeg/trunk

average variance for picture's MBs so we can adjust qscale on the MB layer.

Originally committed as revision 294 to svn://svn.ffmpeg.org/ffmpeg/trunk

DCT/Quantization is done after we have calculated all the MV of the picture.
- This is the preamble for a better bit rate control.

Originally committed as revision 293 to svn://svn.ffmpeg.org/ffmpeg/trunk

- RTP callback system implemented for H.263/H.263+.
- Bug fix on DC coefficients of H.263.
- H.263 decoder now returns PTS on picture number.

Originally committed as revision 292 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

this patch enhances the jpeg header writer. It can be asked to omit
quantisation and huffman tables and it can write different horizontal and
vertical sampling factors. (the last thing is useless for libavcodec
itself (because libavcodec only handles YUV420P at ecoder level), but the
values are initialized so that operation of libavcodec is not impaired)

Originally committed as revision 290 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 289 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 288 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 287 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 286 to svn://svn.ffmpeg.org/ffmpeg/trunk

avoiding useage of ebx

Originally committed as revision 283 to svn://svn.ffmpeg.org/ffmpeg/trunk

* using ALPHA code optimalization for non ARCH_I386

Originally committed as revision 282 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

Originally committed as revision 280 to svn://svn.ffmpeg.org/ffmpeg/trunk

motion-estimation on width/height not divisable through 16 files bugfix

Originally committed as revision 279 to svn://svn.ffmpeg.org/ffmpeg/trunk

dc coeff isnt dequantized (bug reported by falk hueffner)

Originally committed as revision 278 to svn://svn.ffmpeg.org/ffmpeg/trunk

quantizer range bugfix (mpeg1 & mpeg4 encoding)

Originally committed as revision 277 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 276 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 274 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 273 to svn://svn.ffmpeg.org/ffmpeg/trunk

faster simple idct in MMX

Originally committed as revision 269 to svn://svn.ffmpeg.org/ffmpeg/trunk

bye bye weird al rounding bug ;)

Originally committed as revision 268 to svn://svn.ffmpeg.org/ffmpeg/trunk

emms

Originally committed as revision 267 to svn://svn.ffmpeg.org/ffmpeg/trunk

draw_edges_mmx()

Originally committed as revision 266 to svn://svn.ffmpeg.org/ffmpeg/trunk

faster simple_idct in C

Originally committed as revision 265 to svn://svn.ffmpeg.org/ffmpeg/trunk

16-bit divide instead of 32-bit on x86 in msmpeg_pred_dc()

Originally committed as revision 264 to svn://svn.ffmpeg.org/ffmpeg/trunk

memset() in MMX

Originally committed as revision 263 to svn://svn.ffmpeg.org/ffmpeg/trunk