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


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@293 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

- 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.


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@292 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

this includes the range for quantized dct coefficients in dct_quantize()  (-1023...1023)


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@291 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

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)


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@290 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@289 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@288 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@287 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@286 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@285 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@284 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

avoiding useage of ebx


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@283 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

* using ALPHA code optimalization for non ARCH_I386


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@282 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

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


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@281 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@280 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

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


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@279 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

dc coeff isnt dequantized (bug reported by falk hueffner)


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@278 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

quantizer range bugfix (mpeg1 & mpeg4 encoding)


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@277 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@276 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@275 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@274 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@273 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@272 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@271 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@270 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

faster simple idct in MMX


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@269 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

bye bye weird al rounding bug ;)


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@268 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

emms


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@267 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

draw_edges_mmx()


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@266 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

faster simple_idct in C


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@265 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

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


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@264 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

memset() in MMX


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@263 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

print "using pp filters ..." if -v


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@262 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

new msmpeg4 dequantizer did dequantize too early


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@261 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

dequantizers skip trailing zeros
msmpeg4 non-intra decoder has its dequantizer "build in" now


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@260 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

fixed a bug in get_vlc() when using the ALT_BITSTREAM_READER (rv10 & mpeg12 decoders where not working)
some optimizations
 replaced a if() with a >>
 changed get_bits1() a bit so it hopefully is faster on p4


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@259 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@258 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

different fix, s->mbintra_table used only if h263_pred set. - patch by Michael Niedermayer <michaelni@gmx.at>


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@257 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

This fixes segfaults because of uninitialized s->mbintra_table variable - patch by Vladimir Dergachev <volodya@mindspring.com>


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@256 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

s->c_dc_scale was 7 if s->qscale==2 but should be 8 (the bug is visible in deep red areas in high bitrate clips) - patch by Michael Niedermayer <michaelni@gmx.at>


git-svn-id: file:///var/local/repositories/ffmpeg/trunk@255 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@254 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b