1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-02 03:06:28 +02:00
Commit Graph

690 Commits

Author SHA1 Message Date
Michael Niedermayer
17592475b3 alternative bitstream writer (disabled by default, uncomment #define ALT_BISTREAM_WRITER in common.h if u want to try it)
Originally committed as revision 295 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-02-12 15:43:16 +00:00
Juanjo
e03c341ef3 - More work on preliminary bit rate control, just to be able to get an
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
2002-02-10 06:10:50 +00:00
Juanjo
37fbfd0a91 - Now the ME is done for the entire picture when enconding, the
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
2002-02-10 01:56:50 +00:00
Juanjo
81401c1fbc - Fixes on RTP and GOB headers for H.263.
- 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
2002-02-09 16:59:56 +00:00
Arpi
3cb32e3d1f PATCH by Rik Snel <rsnel@cube.dyndns.org>
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
2002-02-09 01:25:06 +00:00
Arpi
80e103d04c PATCH by Rik Snel <rsnel@cube.dyndns.org>
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
2002-02-09 01:23:41 +00:00
Michael Niedermayer
0b2540a0c1 avoid copying input when encoding non intra stuff too
Originally committed as revision 288 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-02-07 23:48:24 +00:00
Michael Niedermayer
2f349de286 (commit by michael)
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
2002-01-27 13:30:18 +00:00
Michael Niedermayer
d2b3c3d7bb (commit by michael)
dc coeff isnt dequantized (bug reported by falk hueffner)

Originally committed as revision 278 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-24 15:25:11 +00:00
Michael Niedermayer
4e3269e87f (commit by michael)
quantizer range bugfix (mpeg1 & mpeg4 encoding)

Originally committed as revision 277 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-24 14:34:03 +00:00
Michael Niedermayer
ae40484c1c (commit by michael)
bye bye weird al rounding bug ;)

Originally committed as revision 268 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-15 22:22:41 +00:00
Michael Niedermayer
5975626d6f (commit by michael)
emms

Originally committed as revision 267 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-15 05:02:26 +00:00
Michael Niedermayer
3d9fccbf0f (commit by michael)
draw_edges_mmx()

Originally committed as revision 266 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-15 04:07:29 +00:00
Michael Niedermayer
badaf88eeb (commit by michael)
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
2002-01-13 04:59:37 +00:00
Arpi
eede215681 rv10 sig11 fix by Michael Niedermayer <michaelni@gmx.at>
Originally committed as revision 258 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-12 14:36:40 +00:00
Michael Niedermayer
5b3438c630 different fix, s->mbintra_table used only if h263_pred set. - patch by Michael Niedermayer <michaelni@gmx.at>
Originally committed as revision 257 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-11 22:07:12 +00:00
Vladimir Dergachev
01dbbd0a9a This fixes segfaults because of uninitialized s->mbintra_table variable - patch by Vladimir Dergachev <volodya@mindspring.com>
Originally committed as revision 256 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-11 21:32:34 +00:00
Arpi
477c35a9f0 mpeg4 encoder fix by Michael Niedermayer <michaelni@gmx.at>
Originally committed as revision 254 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-11 21:23:48 +00:00
Michael Niedermayer
bff6ecaa9c AC table reset (memset) optimization - patch by Michael Niedermayer <michaelni@gmx.at>
Originally committed as revision 251 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-10 00:53:21 +00:00
Michael Niedermayer
ef093645ef fix different UMV handling for mpeg4 vs. h263 - patch by Michael Niedermayer <michaelni@gmx.at>
Originally committed as revision 249 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-10 00:46:40 +00:00
Stephen Davies
fb4a9613ee qscale estimate fix, diff extended to 64bit wide. patch by Stephen Davies <steve@daviesfam.org>
Originally committed as revision 247 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-01-08 00:28:55 +00:00
Arpi
3a842ea5d1 informal msg
Originally committed as revision 238 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-12-26 14:02:08 +00:00
Arpi
393fe8dd8e unrestricted MC fixed - thanks to Michael Niedermayer for idea
Originally committed as revision 236 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-12-16 12:44:34 +00:00
Juanjo
644d98a4f2 - Preliminary RTP friendly mode for H.263.
- GOB headers for H.263 coding on RTP mode.
- Improved GOB header detection for H.263 decoder.

Originally committed as revision 222 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-11-19 02:13:14 +00:00
Arpi
79e7b30557 exporting qscale data for postprocessing (for MPlayer)
Originally committed as revision 164 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-10-16 23:10:38 +00:00
Juanjo
544286b3d3 Moved some H.263+ variables to MpegEncContext to be thread-safe.
Increase video_buffer on ffmpeg to avoid buffer overrun on big pictures.


Originally committed as revision 114 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-31 06:21:27 +00:00
Juanjo
6dbd39fe70 Added support to Unrestricted Motion Vectors (UMV)
on H.263+ encoding/decoding. According to Annex D.
Now MV on H.263+ are 32x32 long.


Originally committed as revision 113 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-28 23:32:34 +00:00
Fabrice Bellard
58f26ba916 changed opendivx to mpeg4
Originally committed as revision 102 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-15 13:12:23 +00:00
Fabrice Bellard
1a56543279 win32 fixes
Originally committed as revision 84 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-13 21:48:05 +00:00
Fabrice Bellard
519c2b6d11 moved matrix init away from MPV_common_init()
Originally committed as revision 77 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-12 00:52:37 +00:00
Fabrice Bellard
514547517b fixed mpeg2 interlaced DCT
Originally committed as revision 74 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-11 22:07:21 +00:00
Fabrice Bellard
425dddb7f1 fixed mpeg1 first block bug (pb with black picture optimisation for B frames)
Originally committed as revision 59 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-08 21:57:07 +00:00
Fabrice Bellard
4f1c7e3c92 fixed block permutation in encoder (not optimal - should move it in forward DCT code)
Originally committed as revision 51 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-07 22:48:42 +00:00
Arpi
4af7bcc185 MMX/MMXEXT iDCT support, using external functions currently defined in libmpeg2
Gives average 13-20% mpeg decoding speedup on x86 systems.


Originally committed as revision 30 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-03 18:33:03 +00:00
Arpi
2d6d0c1d66 - removed #include ../config.h, because it's included from common.h
which is included from avcodec.h which is included from those files...


Originally committed as revision 29 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-02 18:46:26 +00:00
Nick Kurshev
54329dd5a5 Adding fastmemcpy stuff to speedup mplayer project
Originally committed as revision 27 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-08-02 08:29:38 +00:00
Fabrice Bellard
21af69f787 use block[] in structure to have it aligned on 8 bytes for mmx optimizations - dct_unquantize is always a function pointer - added specialized dct_unquantize_h263
Originally committed as revision 22 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-07-30 23:26:26 +00:00
Nick Kurshev
a9b3f63084 Sync with mplayer's stuff
Originally committed as revision 14 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-07-30 09:04:34 +00:00
Fabrice Bellard
3bb4e23a8a added skip macroblock optimization (big perf win on black regions for example)
Originally committed as revision 13 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-07-24 20:43:41 +00:00
Fabrice Bellard
de6d9b6404 Initial revision
Originally committed as revision 5 to svn://svn.ffmpeg.org/ffmpeg/trunk
2001-07-22 14:18:56 +00:00