[svn] / branches / release-1_0-branch / xvidcore / ChangeLog Repository:
ViewVC logotype

Annotation of /branches/release-1_0-branch/xvidcore/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1541 - (view) (download)

1 : edgomez 1382 # Ed.Gomez: This ChangeLog is generated from a personal tree maintained
2 :     # under the arch revision control tool. That's why dates may be skewed. I
3 :     # also removed all my email adresses from the output because they are not
4 :     # relevant.
5 :    
6 : edgomez 1482
7 : edgomez 1541 #########################################################################
8 :     # 1.0.2 (Bitstream Version 36)
9 :     #########################################################################
10 :    
11 :     2004-08-22 13:08:44 GMT patch-53
12 :    
13 :     Summary:
14 :     Thread safety problem in idct C version
15 :     Revision:
16 :     xvidcore--stable--1.0--patch-53
17 :    
18 :     From ed.gomez:
19 :     * Fixed a thread safety problem in C version of the
20 :     idct function. Added some comments on some static
21 :     data not marked as RO.
22 :    
23 :     modified files:
24 :     src/bitstream/mbcoding.c src/dct/idct.c
25 :    
26 :    
27 :     2004-08-21 11:45:55 GMT patch-52
28 :    
29 :     Summary:
30 :     Stupid typo+error in fdct_xxx_skal macro generator.
31 :     Revision:
32 :     xvidcore--stable--1.0--patch-52
33 :    
34 :     From Nicolas Boulay:
35 :     * Found a typo mistake (ecx->eax) and an error in the same line
36 :     But as we're lucky, the unrolled version was bugfree, and that
37 :     is that one which is used.
38 :    
39 :     modified files:
40 :     src/dct/x86_asm/fdct_mmx_skal.asm
41 :    
42 :    
43 :     2004-07-26 20:21:24 GMT patch-51
44 :    
45 :     Summary:
46 :     ChangeLog Update
47 :     Revision:
48 :     xvidcore--stable--1.0--patch-51
49 :    
50 :     ChangeLog Update
51 :    
52 :    
53 :     modified files:
54 :     ChangeLog
55 :    
56 :    
57 : edgomez 1525 2004-07-24 11:33:57 GMT patch-50
58 :    
59 :     Summary:
60 :     BVOP direct/interpolated ref block rounding fix.
61 :     Revision:
62 :     xvidcore--stable--1.0--patch-50
63 :    
64 :     From ed.gomez:
65 :     * BVOP direct/interpolated ref block rounding fix.
66 :     It's been using rounding=1 for averaging stage since ever.
67 :     The standard says it's rounding=0.
68 :    
69 :     See standard clause 7.6.9.4 for explicit code and Section 6.3.5
70 :     that says "rounding=0" in bframes as they don't set the
71 :     vop_rounding_type in VOP header. Both sections match, xvid was
72 :     wrong
73 :    
74 :     modified files:
75 :     src/decoder.c
76 :    
77 :    
78 :     2004-07-23 20:37:09 GMT patch-49
79 :    
80 :     Summary:
81 :     Removed data qualifer in .rodata
82 :     Revision:
83 :     xvidcore--stable--1.0--patch-49
84 :    
85 :     From ed.gomez:
86 :     * long standing warning by yasm... data isn't a keyword for
87 :     .(ro)data sections.
88 :    
89 :     modified files:
90 :     src/bitstream/x86_asm/cbp_mmx.asm
91 :     src/bitstream/x86_asm/cbp_sse2.asm
92 :     src/dct/x86_asm/fdct_mmx_ffmpeg.asm
93 :     src/dct/x86_asm/fdct_mmx_skal.asm
94 :     src/dct/x86_asm/fdct_sse2_skal.asm
95 :     src/dct/x86_asm/idct_3dne.asm src/dct/x86_asm/idct_mmx.asm
96 :     src/dct/x86_asm/idct_sse2_dmitry.asm
97 :     src/dct/x86_asm/simple_idct_mmx.asm
98 :     src/image/x86_asm/colorspace_rgb_mmx.asm
99 :     src/image/x86_asm/colorspace_yuyv_mmx.asm
100 :     src/image/x86_asm/interpolate8x8_3dn.asm
101 :     src/image/x86_asm/interpolate8x8_3dne.asm
102 :     src/image/x86_asm/interpolate8x8_mmx.asm
103 :     src/image/x86_asm/interpolate8x8_xmm.asm
104 :     src/image/x86_asm/qpel_mmx.asm
105 :     src/image/x86_asm/reduced_mmx.asm
106 :     src/motion/x86_asm/sad_3dn.asm src/motion/x86_asm/sad_3dne.asm
107 :     src/motion/x86_asm/sad_mmx.asm src/motion/x86_asm/sad_sse2.asm
108 :     src/motion/x86_asm/sad_xmm.asm
109 :     src/quant/x86_asm/quantize_h263_3dne.asm
110 :     src/quant/x86_asm/quantize_h263_mmx.asm
111 :     src/quant/x86_asm/quantize_mpeg_mmx.asm
112 :     src/quant/x86_asm/quantize_mpeg_xmm.asm
113 :     src/utils/x86_asm/cpuid.asm
114 :     src/utils/x86_asm/interlacing_mmx.asm
115 :     src/utils/x86_asm/mem_transfer_3dne.asm
116 :     src/utils/x86_asm/mem_transfer_mmx.asm
117 :    
118 :    
119 :     2004-07-19 18:45:14 GMT patch-48
120 :    
121 :     Summary:
122 :     Complete previous xvid_decraw patch
123 :     Revision:
124 :     xvidcore--stable--1.0--patch-48
125 :    
126 :     Complete previous xvid_decraw patch
127 :    
128 :    
129 :     modified files:
130 :     examples/xvid_decraw.c
131 :    
132 :    
133 :     2004-07-18 11:58:48 GMT patch-47
134 :    
135 :     Summary:
136 :     ISO C99'ism fix
137 :     Revision:
138 :     xvidcore--stable--1.0--patch-47
139 :    
140 :     ISO C99'ism fix
141 :    
142 :    
143 :     modified files:
144 :     src/encoder.c
145 :    
146 :    
147 :     2004-07-17 11:32:42 GMT patch-46
148 :    
149 :     Summary:
150 :     Make sure time incr is never larger than 16bit.
151 :     Revision:
152 :     xvidcore--stable--1.0--patch-46
153 :    
154 :     From ed.gomez:
155 :     * Keep both fbase and fincr under 16bit limit.
156 :    
157 :     modified files:
158 :     src/encoder.c
159 :    
160 :    
161 :     2004-07-17 10:00:42 GMT patch-45
162 :    
163 :     Summary:
164 :     Future version interoperability
165 :     Revision:
166 :     xvidcore--stable--1.0--patch-45
167 :    
168 :     From ed.gomez:
169 :     * Zeroing the structures is the best way not to pass
170 :     wrong data when dealing with slightly new XviD (like
171 :     head). If we don't zero here, then brightness is not
172 :     initialized in HEAD leading to crash.
173 :    
174 :     modified files:
175 :     examples/xvid_decraw.c
176 :    
177 :    
178 :     2004-07-10 17:30:40 GMT patch-44
179 :    
180 :     Summary:
181 :     ChangeLog update
182 :     Revision:
183 :     xvidcore--stable--1.0--patch-44
184 :    
185 :     ChangeLog update
186 :    
187 :    
188 :     modified files:
189 :     ChangeLog
190 :    
191 :    
192 : edgomez 1482 2004-07-10 17:27:06 GMT patch-43
193 :    
194 :     Summary:
195 :     Small mem leak in vfw.
196 :     Revision:
197 :     xvidcore--stable--1.0--patch-43
198 :    
199 :     From sysKin:
200 :     * Free zones.
201 :    
202 :     modified files:
203 :     vfw/src/codec.c
204 :    
205 :    
206 :     2004-07-10 16:55:53 GMT patch-42
207 :    
208 :     Summary:
209 :     Fix wrong matrix reading logic.
210 :     Revision:
211 :     xvidcore--stable--1.0--patch-42
212 :    
213 :     From sysKin:
214 :     * Fix a problem reading quantization matrix.
215 :    
216 :     modified files:
217 :     src/bitstream/bitstream.c
218 :    
219 :    
220 :     2004-06-26 21:26:35 GMT patch-41
221 :    
222 :     Summary:
223 :     low delay guessing (il)logic fix.
224 :     Revision:
225 :     xvidcore--stable--1.0--patch-41
226 :    
227 :     From sysKin:
228 :     * bframes were decoded wrong when trying to guess low_delay flag
229 :     as specified in the standard when vol_control_parameters aren't
230 :     available.
231 :    
232 :     modified files:
233 :     src/decoder.c
234 :    
235 :    
236 :     2004-06-13 19:15:05 GMT patch-40
237 :    
238 :     Summary:
239 :     Small memory error in ia32 cpuid function.
240 :     Revision:
241 :     xvidcore--stable--1.0--patch-40
242 :    
243 :     From ed.gomez:
244 :     * Valgrind detected a write to suspicious stack space. To avoid
245 :     any false reporting, added an explicit stack space allocation.
246 :    
247 :     modified files:
248 :     src/utils/x86_asm/cpuid.asm
249 :    
250 : edgomez 1449 #########################################################################
251 : edgomez 1468 # 1.0.1 (Bitstream Version 35)
252 :     #########################################################################
253 :    
254 : edgomez 1482 2004-06-05 22:55:56 GMT patch-39
255 :    
256 :     Summary:
257 :     Marking 1.0.1 release
258 :     Revision:
259 :     xvidcore--stable--1.0--patch-39
260 :    
261 :     Marking 1.0.1 release
262 :    
263 :    
264 :     modified files:
265 :     ChangeLog TODO build/generic/configure.in src/xvid.c
266 :     src/xvid.h
267 :    
268 :    
269 : edgomez 1468 2004-06-02 20:58:38 GMT patch-38
270 :    
271 :     Summary:
272 :     DC clipping bug for real
273 :     Revision:
274 :     xvidcore--stable--1.0--patch-38
275 :    
276 :     From ed.gomez:
277 :     * patch-25 was supposed to fix a DC clipping bug. However i
278 :     added the additional clipping code in the wrong place. But
279 :     at least, my fix didn't cause any trouble, it was just noop.
280 :     This patch should really fix this very "unlikely bug" (i just
281 :     want to remind the reader that this bug isn't easy to trigger,
282 :     and eg: my test sequences don't trigger it at all)
283 :    
284 :     BS version incremented: 35
285 :    
286 :     modified files:
287 :     src/decoder.c src/motion/estimation_rd_based.c
288 :     src/prediction/mbprediction.c src/prediction/mbprediction.h
289 :     src/xvid.h
290 :    
291 :    
292 :     2004-05-31 21:11:49 GMT patch-37
293 :    
294 :     Summary:
295 :     time fixes to decoder.
296 :     Revision:
297 :     xvidcore--stable--1.0--patch-37
298 :    
299 :     From ed.gomez:
300 :     * timestamps were badly computed by teh decoder in some corner cases
301 :     (1fps). This bug revealed that, timestamps were indeed wrong as
302 :     expected, but that bvop blocks in direct mode (vectors interpolated)
303 :     were somewhat compensated with wrong vectors in these same corner
304 :     cases.
305 :    
306 :     modified files:
307 :     src/bitstream/bitstream.c src/decoder.c src/decoder.h
308 :    
309 :    
310 :     2004-05-30 09:36:13 GMT patch-36
311 :    
312 :     Summary:
313 :     Wrong license header.
314 :     Revision:
315 :     xvidcore--stable--1.0--patch-36
316 :    
317 :     From ed.gomez:
318 :     * Pascal did agree a plain GPL migration long ago, but this file
319 :     remained GPL+location restriction.
320 :    
321 :     modified files:
322 :     src/image/reduced.c
323 :    
324 :    
325 :     2004-05-29 09:02:25 GMT patch-35
326 :    
327 :     Summary:
328 :     More missing va_end() calls.
329 :     Revision:
330 :     xvidcore--stable--1.0--patch-35
331 :    
332 :     From pete:
333 :     * portab.h is plenty of missing calls to va_end().
334 :    
335 :     modified files:
336 :     src/portab.h
337 :    
338 :    
339 :     2004-05-28 21:28:21 GMT patch-34
340 :    
341 :     Summary:
342 :     FPS=1 problem in decoder.
343 :     Revision:
344 :     xvidcore--stable--1.0--patch-34
345 :    
346 :     From ed.gomez:
347 :     * patch-24 did fix bad behavior in encoder, so at least, compliant
348 :     streams were generated but the decoder was still doing the maths
349 :     a wrong way. Apply same logic to decoder.
350 :    
351 :     Thanks to the patch-24 bug reporter for this followup.
352 :    
353 :     modified files:
354 :     src/bitstream/bitstream.c
355 :    
356 :    
357 :     2004-05-27 20:04:01 GMT patch-33
358 :    
359 :     Summary:
360 :     Nasty typo in pvop vector lambdas.
361 :     Revision:
362 :     xvidcore--stable--1.0--patch-33
363 :    
364 :     From sysKin:
365 :     * s/+/*/ in the lambda value array for vectors in the pvop
366 :     estimation module.
367 :    
368 :     modified files:
369 :     src/motion/estimation_pvop.c
370 :    
371 :    
372 :     2004-05-26 13:23:38 GMT patch-32
373 :    
374 :     Summary:
375 :     Bits/Bytes confusion in the VFW frontend.
376 :     Revision:
377 :     xvidcore--stable--1.0--patch-32
378 :    
379 :     From sysKin:
380 :     * confusion between the kilo, in kilobits (1000) and
381 :     the kilo in kilobytes (1024, should be named KiB anyway)
382 :     * biSizeImage is in bytes, not bits according to the Win32
383 :     API.
384 :    
385 :     modified files:
386 :     vfw/src/codec.c vfw/src/config.c
387 :    
388 :    
389 :     2004-05-26 09:28:31 GMT patch-31
390 :    
391 :     Summary:
392 :     Close variable argument list.
393 :     Revision:
394 :     xvidcore--stable--1.0--patch-31
395 :    
396 :     From ed.gomez:
397 :     * Close the variable argument list as specified by the ANSI C
398 :     standard. Reported by Carsten <list4darav at gmx dot de> on
399 :     xvid-devel.
400 :    
401 :     modified files:
402 :     src/image/font.c
403 :    
404 :    
405 :     2004-05-26 09:00:26 GMT patch-30
406 :    
407 :     Summary:
408 :     ICM compatibility for VFW
409 :     Revision:
410 :     xvidcore--stable--1.0--patch-30
411 :    
412 :     From sysKin:
413 :     * Makes the VFW frontend compatible with ICM applications
414 :     (Ooo, MS Office... etc). Reported on IRC.
415 :    
416 :     modified files:
417 :     vfw/src/config.c vfw/src/driverproc.c
418 :    
419 :    
420 :     2004-05-26 08:58:56 GMT patch-29
421 :    
422 :     Summary:
423 :     Small trellis bug
424 :     Revision:
425 :     xvidcore--stable--1.0--patch-29
426 :    
427 :     From sysKin:
428 :     * Last coeff wasn't summed.
429 :    
430 :     Reported by Jean Marc.
431 :    
432 :     modified files:
433 :     src/utils/mbtransquant.c
434 :    
435 :    
436 :     2004-05-26 08:46:45 GMT patch-28
437 :    
438 :     Summary:
439 :     Small bug in bframe ME.
440 :     Revision:
441 :     xvidcore--stable--1.0--patch-28
442 :    
443 :     From sysKin:
444 :     * Small bug in bframe ME.
445 :    
446 :     modified files:
447 :     src/motion/estimation_bvop.c
448 :    
449 :     #########################################################################
450 : edgomez 1449 # 1.0.0 final (Bitstream Version 34)
451 :     #########################################################################
452 : edgomez 1382
453 : edgomez 1468 2004-05-08 22:26:06 GMT patch-27
454 :    
455 :     Summary:
456 :     Marking 1.0.0 final
457 :     Revision:
458 :     xvidcore--stable--1.0--patch-27
459 :    
460 :     From ed.gomez:
461 :     * Marking 1.0.0 final \o/
462 :    
463 :     modified files:
464 :     ChangeLog build/generic/configure.in src/xvid.h
465 :    
466 :    
467 : edgomez 1449 2004-05-06 17:56:52 GMT patch-26
468 :    
469 :     Summary:
470 :     Small mismatch in hint<->widget in VFW
471 :     Revision:
472 :     xvidcore--stable--1.0--patch-26
473 :    
474 :     From sysKin:
475 :     * Small mismatch in hint<->widget.
476 :    
477 :     modified files:
478 :     vfw/src/resource.rc
479 :    
480 :    
481 :     2004-05-02 22:40:50 GMT patch-25
482 :    
483 :     Summary:
484 :     DC prediction fix.
485 :     Revision:
486 :     xvidcore--stable--1.0--patch-25
487 :    
488 :     From ed.gomez:
489 :     * DC predictors weren't clipped to the [-2048, 2047] range.
490 :    
491 :     BS version increased to 33
492 :    
493 :     Thanks to jnorish on our forums to point out the problem.
494 :    
495 :     modified files:
496 :     src/bitstream/bitstream.c src/decoder.c
497 :     src/motion/estimation_rd_based.c src/prediction/mbprediction.c
498 :     src/prediction/mbprediction.h src/xvid.h
499 :    
500 :    
501 :     2004-05-02 10:30:29 GMT patch-24
502 :    
503 :     Summary:
504 :     Possible VOL header corruption.
505 :     Revision:
506 :     xvidcore--stable--1.0--patch-24
507 :    
508 :     From ed.gomez:
509 :     * The VOL header could be corrupted when passing fincr=fbase=1 which
510 :     happens for fps=1 sequences.
511 :    
512 :     BS version bumped up to 32
513 :    
514 :     Original report:
515 :     http://www.xvid.org/modules.php?op=modload&name=phpBB2&file=viewtopic&t=2026&highlight=
516 :    
517 :     modified files:
518 :     src/bitstream/bitstream.c src/xvid.h
519 :    
520 :    
521 :     2004-04-30 23:10:19 GMT patch-23
522 :    
523 :     Summary:
524 :     Some very light Unix build system changes
525 :     Revision:
526 :     xvidcore--stable--1.0--patch-23
527 :    
528 :     To prepare testing framework merging.
529 :    
530 :     From ed.gomez:
531 :     * Some typos
532 :     * Copyright updates (it's 2004 since a few months ;-)
533 :     * Added some checking to bootstrap.sh
534 :     * Added m4 AC_PREREQ macro to configure.in
535 :    
536 :     modified files:
537 :     build/generic/Makefile build/generic/bootstrap.sh
538 :     build/generic/configure.in
539 :    
540 :    
541 :     2004-04-20 19:40:29 GMT patch-22
542 :    
543 :     Summary:
544 :     Small visual fix.
545 :     Revision:
546 :     xvidcore--stable--1.0--patch-22
547 :    
548 :     From sysKin:
549 :     * Small visual fix
550 :    
551 :     modified files:
552 :     vfw/src/config.c
553 :    
554 :    
555 :     2004-04-20 19:38:24 GMT patch-21
556 :    
557 :     Summary:
558 :     Fix crash in decoder for non IFrame 1st frame.
559 :     Revision:
560 :     xvidcore--stable--1.0--patch-21
561 :    
562 :     From sysKin:
563 :     * Fixed the crash caused by non IFrame 1st frame.
564 :    
565 :     modified files:
566 :     src/decoder.c
567 :    
568 :    
569 :     2004-04-18 16:21:50 GMT patch-20
570 :    
571 :     Summary:
572 :     Typo
573 :     Revision:
574 :     xvidcore--stable--1.0--patch-20
575 :    
576 :     Typo
577 :    
578 :    
579 :     modified files:
580 :     vfw/src/resource.rc
581 :    
582 :    
583 :     2004-04-17 17:04:20 GMT patch-19
584 :    
585 :     Summary:
586 :     vfw opens audio file in shared access mode
587 :     Revision:
588 :     xvidcore--stable--1.0--patch-19
589 :    
590 :     vfw opens audio file in shared access mode
591 :    
592 :    
593 :     modified files:
594 :     vfw/src/config.c
595 :    
596 :    
597 :     2004-04-15 22:39:12 GMT patch-18
598 :    
599 :     Summary:
600 :     Tiny xvid_decraw cleaning
601 :     Revision:
602 :     xvidcore--stable--1.0--patch-18
603 :    
604 :     Tiny xvid_decraw cleaning
605 :    
606 :    
607 :     modified files:
608 :     examples/xvid_decraw.c
609 :    
610 :    
611 :     2004-04-15 19:14:31 GMT patch-17
612 :    
613 :     Summary:
614 :     Tiny minor fixes for msvc.
615 :     Revision:
616 :     xvidcore--stable--1.0--patch-17
617 :    
618 :     From pete:
619 :     * Missing arch endianness define in project files.
620 :     * Add a textual warning about win32 console EOF misreading.
621 :     * Prevent a SIGFPE when no frames were decoded.
622 :    
623 :     modified files:
624 :     build/win32/xvid_decraw.dsp build/win32/xvid_encraw.dsp
625 :     examples/xvid_decraw.c
626 :    
627 :    
628 :     2004-04-14 22:41:07 GMT patch-16
629 :    
630 :     Summary:
631 :     Fixed missing 1st frame in dshow output.
632 :     Revision:
633 :     xvidcore--stable--1.0--patch-16
634 :    
635 :     From sysKin:
636 :     * decoder flags were overwritten, this was preventing from outputing
637 :     the first frame immediatly.
638 :    
639 :     modified files:
640 :     dshow/src/CXvidDecoder.cpp
641 :    
642 :    
643 :     2004-04-14 22:39:17 GMT patch-15
644 :    
645 :     Summary:
646 :     Ressource leaking in dshow.
647 :     Revision:
648 :     xvidcore--stable--1.0--patch-15
649 :    
650 :     From sysKin:
651 :     * Same kind of ressource leaking as in vfw. Same cure.
652 :    
653 :     modified files:
654 :     dshow/src/CXvidDecoder.cpp dshow/src/CXvidDecoder.h
655 :    
656 :    
657 :     2004-04-14 19:45:23 GMT patch-14
658 :    
659 :     Summary:
660 :     Fixed small bug in trellis code.
661 :     Revision:
662 :     xvidcore--stable--1.0--patch-14
663 :    
664 :     From ed.gomez (spotted by jean marc):
665 :     * Trellis optimization was computing the sum |coeffs| wrongly
666 :     because the Compute_sum function wasn't doing zigzag scanning
667 :     and stopped at last non zero coeffs in linear scanning...
668 :     The trivial fix could have been to pass the zigzag to the compute
669 :     sum function so it could have done its job right.
670 :     But... Why computing stuff that is already known in the upper layer ?
671 :     I removed the compute sum function, and just pass the original sum
672 :     value in trellis function parameters, just in case we have to return
673 :     it again because trellis failed optimizing the block coeffs.
674 :    
675 :     This fix the bug and saves cycles :-) (but should not be noticeable)
676 :    
677 :     modified files:
678 :     src/utils/mbtransquant.c src/xvid.h
679 :    
680 :    
681 :     2004-04-12 12:06:12 GMT patch-13
682 :    
683 :     Summary:
684 :     Don't do SAD and RD based searches for qp.
685 :     Revision:
686 :     xvidcore--stable--1.0--patch-13
687 :    
688 :     From sysKin:
689 :     * MakeGoodFlags function wasn't disabling SAD based search when
690 :     using RD. This was causing slowdown for no gain at all. This
691 :     patch should speed up encoding in qp mode.
692 :    
693 :    
694 :     modified files:
695 :     src/motion/estimation_pvop.c
696 :    
697 :    
698 :     2004-04-12 12:01:19 GMT patch-12
699 :    
700 :     Summary:
701 :     3dnow functions proper separation.
702 :     Revision:
703 :     xvidcore--stable--1.0--patch-12
704 :    
705 :     From pete (thx to a forum report):
706 :     * Separate correctly pure 3dnow functions and 3dnow+mmxext functions.
707 :     This fix "Illegal instruction" crash on old k6-2 CPUs.
708 :    
709 :     modified files:
710 :     src/xvid.c
711 :    
712 :    
713 :     2004-04-12 11:57:20 GMT patch-11
714 :    
715 :     Summary:
716 :     Better MV clipping code.
717 :     Revision:
718 :     xvidcore--stable--1.0--patch-11
719 :    
720 :     From sysKin:
721 :     * Better MV clipping. DivX 5 generates out of range vectors and clipping
722 :     them directly borks the decoding of other MVs for which out of range
723 :     MVs were predictors. So it's just better to clip them for the block
724 :     decoding and keep an unclipped version for predictions.
725 :    
726 :     modified files:
727 :     src/decoder.c
728 :    
729 :    
730 :     2004-04-08 20:34:54 GMT patch-10
731 :    
732 :     Summary:
733 :     PGM support back in xvid_decraw.
734 :     Revision:
735 :     xvidcore--stable--1.0--patch-10
736 :    
737 :     From ed.gomez:
738 :     * pgm/pnm format added back to xvid_decraw
739 :     + pgm/pnm formats are now default for yv12/i420/rgb24 pixel format.
740 :     + tga is default for rgb16/32
741 :     + use option -f to choose tga, or pnm/pgm
742 :    
743 :     modified files:
744 :     examples/xvid_decraw.c
745 :    
746 :    
747 :     2004-04-07 22:30:15 GMT patch-9
748 :    
749 :     Summary:
750 :     3DNow Ext functions use MMXEXT opcodes.
751 :     Revision:
752 :     xvidcore--stable--1.0--patch-9
753 :    
754 :     From Soltius (XviD Forum):
755 :     * Most of 3dnow extension functions do use MMXEXT opcodes, so
756 :     classify these functions as 3dnowext+mmxext. Avoids K6-2
757 :     boxes to crash with an invalid instruction error reported
758 :     by the host OS.
759 :    
760 :     PS: original bug report
761 :     http://www.xvid.org/modules.php?op=modload&name=phpBB2&file=viewtopic&t=1656
762 :    
763 :     modified files:
764 :     src/xvid.c
765 :    
766 :    
767 :     2004-04-07 22:01:54 GMT patch-8
768 :    
769 :     Summary:
770 :     RGB 16bit output fix.
771 :     Revision:
772 :     xvidcore--stable--1.0--patch-8
773 :    
774 :     From ScarletteTout (XviD Forum):
775 :     * Fix RGB 16bit output in C functions.
776 :    
777 :     From ed.gomez:
778 :     * Replaced PGM output by TGA output so it's easy to implement
779 :     RGB 16/24/32 and greyscale bitmaps support in a single format.
780 :     (pgm could have supported RGB 24 and Greyscale only)
781 :     * Added colorspace choice to xvid_decraw
782 :     Use option -c csp, where csp is either rgb16, rgb24, rgb32, yv12 or i420
783 :     Defaults to i420.
784 :    
785 :     PS: original bug report
786 :     http://www.xvid.org/modules.php?op=modload&name=phpBB2&file=viewtopic&t=1960&highlight=
787 :    
788 :     modified files:
789 :     examples/xvid_decraw.c src/image/colorspace.c
790 :    
791 :    
792 :     #########################################################################
793 :     # 1.0.0 RC4 (Bitstream Version 30)
794 :     #########################################################################
795 :    
796 :     2004-04-04 20:21:38 GMT patch-7
797 :    
798 :     Summary:
799 :     DShow widget hiding.
800 :     Revision:
801 :     xvidcore--stable--1.0--patch-7
802 :    
803 :     From Michael:
804 :     * No need to keep widget visibles if they won't be in 1.0.0.
805 :    
806 :     modified files:
807 :     dshow/src/xvid.ax.rc
808 :    
809 :    
810 :     2004-04-04 20:17:52 GMT patch-6
811 :    
812 :     Summary:
813 :     Compiler quirk in portab.h
814 :     Revision:
815 :     xvidcore--stable--1.0--patch-6
816 :    
817 :     From Michael:
818 :     * The VC.NET workaround was causing trouble. Inversed the test.
819 :    
820 :     modified files:
821 :     src/portab.h
822 :    
823 :    
824 :     2004-04-04 14:19:10 GMT patch-5
825 :    
826 :     Summary:
827 :     Marking RC4
828 :     Revision:
829 :     xvidcore--stable--1.0--patch-5
830 :    
831 :     Marking RC4
832 :    
833 :    
834 :     modified files:
835 :     ChangeLog build/generic/configure.in src/xvid.h
836 :    
837 :    
838 : edgomez 1408 2004-04-04 14:07:00 GMT patch-4
839 :    
840 :     Summary:
841 :     Frame dropping disabling for bframes.
842 :     Revision:
843 :     xvidcore--stable--1.0--patch-4
844 :    
845 :     From sysKin & Pete:
846 :     * Disable frame dropping with bframes enabled. These two options
847 :     do not play fine together.
848 :    
849 :     modified files:
850 :     src/encoder.c
851 :    
852 :    
853 :     2004-04-04 14:05:50 GMT patch-3
854 :    
855 :     Summary:
856 :     Dead code removal.
857 :     Revision:
858 :     xvidcore--stable--1.0--patch-3
859 :    
860 :     From sysKin:
861 :     * FrameCodeP was always called with contanst parameters. Removed
862 :     these parameters and associated dead code.
863 :    
864 :     modified files:
865 :     src/encoder.c
866 :    
867 :    
868 :     2004-04-04 14:03:42 GMT patch-2
869 :    
870 :     Summary:
871 :     Typo in ME fast comparison.
872 :     Revision:
873 :     xvidcore--stable--1.0--patch-2
874 :    
875 :     From sysKin:
876 :     * Small typo in Fast ME code.
877 :    
878 :     modified files:
879 :     src/motion/estimation_common.c
880 :    
881 :    
882 :     2004-04-02 23:58:19 GMT patch-1
883 :    
884 :     Summary:
885 :     VFW Resource leak fix (try #2)
886 :     Revision:
887 :     xvidcore--stable--1.0--patch-1
888 :    
889 :     From Suiryc on IRC:
890 :     * both encoder and decoder ending functions were calling
891 :     the dll freeing code. This was an error as the first function
892 :     called would unbind core function for the second called one.
893 :     Thus xvidcore could not release buffers.
894 :    
895 :     modified files:
896 :     vfw/src/codec.c vfw/src/driverproc.c
897 :    
898 :    
899 :     2004-04-02 20:33:02 GMT base-0
900 :    
901 :     Summary:
902 :     tag of ed.gomez@free.fr--2004-1/xvidcore--devapi4--1.0--patch-53
903 :     Revision:
904 :     xvidcore--stable--1.0--base-0
905 :    
906 :     (automatically generated log message)
907 :    
908 :    
909 :     2004-03-31 19:32:47 GMT patch-53
910 :    
911 :     Summary:
912 :     Ressources leaking in VFW.
913 :     Revision:
914 :     xvidcore--devapi4--1.0--patch-53
915 :    
916 :     From sysKin:
917 :     * Storing ressources in global vars is making multithreaded/instanced
918 :     apps leaking lot of memory. Moved these vars to codec struct.
919 :    
920 :     Thanks to dalox <fdalleau at hymatom d0t fr> to spot and fix the bug.
921 :    
922 :     modified files:
923 :     vfw/src/codec.c vfw/src/codec.h vfw/src/driverproc.c
924 :    
925 :    
926 :     2004-03-31 19:28:51 GMT patch-52
927 :    
928 :     Summary:
929 :     Fix to bad NVOP+bframe interaction.
930 :     Revision:
931 :     xvidcore--devapi4--1.0--patch-52
932 :    
933 :     From syskin:
934 :     * When generating a NVOP, it interferes with bframe flushing and
935 :     packing.
936 :    
937 :     modified files:
938 :     src/encoder.c
939 :    
940 :    
941 :     2004-03-31 19:24:42 GMT patch-51
942 :    
943 :     Summary:
944 :     Fix GMC 2 warp point.
945 :     Revision:
946 :     xvidcore--devapi4--1.0--patch-51
947 :    
948 :     From Skal:
949 :     * Fix 2 warp points GMC.
950 :    
951 :     modified files:
952 :     src/motion/gmc.c
953 :    
954 :    
955 :     2004-03-31 19:18:46 GMT patch-50
956 :    
957 :     Summary:
958 :     Added intra quant testing.
959 :     Revision:
960 :     xvidcore--devapi4--1.0--patch-50
961 :    
962 :     From Skal:
963 :     * Added intra quant testing to xvid_bench
964 :    
965 :     modified files:
966 :     examples/xvid_bench.c
967 :    
968 :    
969 :     2004-03-31 19:07:55 GMT patch-49
970 :    
971 :     Summary:
972 :     input width/height check
973 :     Revision:
974 :     xvidcore--devapi4--1.0--patch-49
975 :    
976 :     From Skal:
977 :     * Input width and height aren't checked and this can cause crash.
978 :    
979 :     modified files:
980 :     src/encoder.c
981 :    
982 :    
983 :     2004-03-28 01:02:21 GMT patch-48
984 :    
985 :     Summary:
986 :     $ CVS expansion removed
987 :     Revision:
988 :     xvidcore--devapi4--1.0--patch-48
989 :    
990 :     $ CVS expansion removed
991 :    
992 :    
993 :     modified files:
994 :     dshow/src/Configure.cpp dshow/src/config.c
995 :     examples/xvid_encraw.c
996 :    
997 :    
998 :     2004-03-28 00:45:23 GMT patch-47
999 :    
1000 :     Summary:
1001 :     Allow bigger frames
1002 :     Revision:
1003 :     xvidcore--devapi4--1.0--patch-47
1004 :    
1005 :     Allow bigger frames
1006 :    
1007 :    
1008 :     modified files:
1009 :     examples/xvid_encraw.c
1010 :    
1011 :    
1012 :     2004-03-28 00:41:54 GMT patch-46
1013 :    
1014 :     Summary:
1015 :     Fix in postproc header
1016 :     Revision:
1017 :     xvidcore--devapi4--1.0--patch-46
1018 :    
1019 :     Fix in postproc header
1020 :    
1021 :    
1022 :     modified files:
1023 :     src/image/postprocessing.h
1024 :    
1025 :    
1026 :     2004-03-28 00:33:02 GMT patch-45
1027 :    
1028 :     Summary:
1029 :     Fixed xvidvfw build on real mingw+msys systems
1030 :     Revision:
1031 :     xvidcore--devapi4--1.0--patch-45
1032 :    
1033 :     Fixed xvidvfw build on real mingw+msys systems
1034 :    
1035 :    
1036 :     modified files:
1037 :     vfw/bin/Makefile
1038 :    
1039 :    
1040 :     2004-03-15 21:48:48 GMT patch-44
1041 :    
1042 :     Summary:
1043 :     VFW updates.
1044 :     Revision:
1045 :     xvidcore--devapi4--1.0--patch-44
1046 :    
1047 :     VFW updates.
1048 :    
1049 :     modified files:
1050 :     vfw/src/config.c vfw/src/resource.rc
1051 :    
1052 :    
1053 :     2004-03-15 21:44:17 GMT patch-43
1054 :    
1055 :     Summary:
1056 :     GMC bugfix.
1057 :     Revision:
1058 :     xvidcore--devapi4--1.0--patch-43
1059 :    
1060 :     From skal:
1061 :     * GMC bugfixes.
1062 :    
1063 :     modified files:
1064 :     src/motion/gmc.c
1065 :    
1066 :    
1067 :     2004-03-15 21:41:18 GMT patch-42
1068 :    
1069 :     Summary:
1070 :     Buffer overrun fix in post proc.
1071 :     Revision:
1072 :     xvidcore--devapi4--1.0--patch-42
1073 :    
1074 :     From sysKin:
1075 :     * Fixed buffer overrun in postproc code.
1076 :    
1077 :     modified files:
1078 :     src/image/postprocessing.c src/image/postprocessing.h
1079 :    
1080 :    
1081 :     2004-03-15 21:33:22 GMT patch-41
1082 :    
1083 :     Summary:
1084 :     Forgotten files.
1085 :     Revision:
1086 :     xvidcore--devapi4--1.0--patch-41
1087 :    
1088 :     Me:
1089 :     * Sorry i missed these file additions. Important for the packages
1090 :     as i build them from the tla archive.
1091 :    
1092 :     new files:
1093 :     dshow/src/.arch-ids/Configure.cpp.id
1094 :     dshow/src/.arch-ids/config.c.id
1095 :     dshow/src/.arch-ids/config.h.id dshow/src/.arch-ids/debug.h.id
1096 :     dshow/src/Configure.cpp dshow/src/config.c dshow/src/config.h
1097 :     dshow/src/debug.h
1098 :    
1099 :    
1100 :     2004-03-03 21:01:09 GMT patch-40
1101 :    
1102 :     Summary:
1103 :     MV clipping in decoder.
1104 :     Revision:
1105 :     xvidcore--devapi4--1.0--patch-40
1106 :    
1107 :     From sysKin:
1108 :     * clip MVs to valid ranges.
1109 :    
1110 :     modified files:
1111 :     src/decoder.c src/decoder.h
1112 :    
1113 :    
1114 :     #########################################################################
1115 : edgomez 1449 # 1.0.0 RC3 (Bistream Version 29)
1116 : edgomez 1408 #########################################################################
1117 :    
1118 :     2004-02-29 13:17:10 GMT patch-39
1119 :    
1120 :     Summary:
1121 :     Marking 1.0.0 RC3
1122 :     Revision:
1123 :     xvidcore--devapi4--1.0--patch-39
1124 :    
1125 :     * Marking RC3
1126 :    
1127 :     modified files:
1128 :     ChangeLog build/generic/configure.in src/xvid.h
1129 :    
1130 :    
1131 : edgomez 1382 2004-02-29 13:08:38 GMT patch-38
1132 :    
1133 :     Summary:
1134 :     Win32 project fixes for xvid_encraw and xvid_decraw.
1135 :     Revision:
1136 :     xvidcore--devapi4--1.0--patch-38
1137 :    
1138 :     From sysKin:
1139 :     * xvid_dec/encraw were linking against libxvidcore.lib. Now they link
1140 :     against xvidcore.dll.a
1141 :    
1142 :     NB: xvid_bench is left as is though it doesn't link. This program
1143 :     requires access to internal functions which aren't available
1144 :     through the dll link lib. No good solution exists for MSVC to
1145 :     build both a dll+its link lib+static lib.
1146 :    
1147 :     modified files:
1148 :     build/win32/xvid_decraw.dsp build/win32/xvid_encraw.dsp
1149 :    
1150 :    
1151 :     2004-02-29 12:56:36 GMT patch-37
1152 :    
1153 :     Summary:
1154 :     DShow updates
1155 :     Revision:
1156 :     xvidcore--devapi4--1.0--patch-37
1157 :    
1158 :     Bunch of DShow updates
1159 :    
1160 :     modified files:
1161 :     dshow/src/CXvidDecoder.cpp dshow/src/resource.h
1162 :     dshow/src/xvid.ax.rc
1163 :    
1164 :    
1165 :     2004-02-29 12:55:41 GMT patch-36
1166 :    
1167 :     Summary:
1168 :     VFW updates.
1169 :     Revision:
1170 :     xvidcore--devapi4--1.0--patch-36
1171 :    
1172 :     * Bunch of VFW frontends updates.
1173 :    
1174 :     modified files:
1175 :     vfw/src/codec.c vfw/src/codec.h vfw/src/config.c
1176 :     vfw/src/resource.rc vfw/src/status.c vfw/vfw.dsp
1177 :    
1178 :    
1179 :     2004-02-29 12:49:24 GMT patch-35
1180 :    
1181 :     Summary:
1182 :     Artefact workaround in bframes.
1183 :     Revision:
1184 :     xvidcore--devapi4--1.0--patch-35
1185 :    
1186 :     From sysKin:
1187 :     * Workaround for some artefacts appearing in bframes.
1188 :    
1189 :     modified files:
1190 :     src/motion/estimation_bvop.c
1191 :    
1192 :    
1193 :     2004-02-29 12:46:13 GMT patch-34
1194 :    
1195 :     Summary:
1196 :     Decoder buffer overflow fix.
1197 :     Revision:
1198 :     xvidcore--devapi4--1.0--patch-34
1199 :    
1200 :     From sysKin:
1201 :     * Avoids buffer overflow when reading the last align byte. Our
1202 :     bitstream do 4 byte reading because of some platform contraints
1203 :     (ARM) and can cause buffer overflow reads.
1204 :    
1205 :     modified files:
1206 :     src/decoder.c
1207 :    
1208 :    
1209 :     2004-02-29 11:53:47 GMT patch-33
1210 :    
1211 :     Summary:
1212 :     Compatibility decoding for old bitstreams.
1213 :     Revision:
1214 :     xvidcore--devapi4--1.0--patch-33
1215 :    
1216 :     From syskin:
1217 :     * old core versions used in dev-api-3 distributed by nearly all win32 bin
1218 :     builders used to have a edging bug. So when this information is known,
1219 :     workaround the bug.
1220 :    
1221 :     modified files:
1222 :     src/decoder.c src/encoder.c src/image/image.c
1223 :     src/image/image.h
1224 :    
1225 :     #########################################################################
1226 : edgomez 1449 # 1.0.0 RC2 (Bistream Version 28)
1227 : edgomez 1382 #########################################################################
1228 :    
1229 :     2004-02-08 01:06:40 GMT patch-32
1230 :    
1231 :     Summary:
1232 :     Marking RC2
1233 :     Revision:
1234 :     xvidcore--devapi4--1.0--patch-32
1235 :    
1236 :     Marking RC2
1237 :    
1238 :    
1239 :     modified files:
1240 :     ChangeLog build/generic/configure.in
1241 :    
1242 :    
1243 :     2004-02-07 13:54:24 GMT patch-31
1244 :    
1245 :     Summary:
1246 :     Win32 project outputs dll lib for linking.
1247 :     Revision:
1248 :     xvidcore--devapi4--1.0--patch-31
1249 :    
1250 :     From pete:
1251 :     * output a lib to link against the dll (xvidcore.dll.a).
1252 :    
1253 :     modified files:
1254 :     build/win32/libxvidcore.dsp
1255 :    
1256 :    
1257 :     2004-02-07 13:51:01 GMT patch-30
1258 :    
1259 :     Summary:
1260 :     DShow update.
1261 :     Revision:
1262 :     xvidcore--devapi4--1.0--patch-30
1263 :    
1264 :     From pete:
1265 :     * cmd line driving
1266 :    
1267 :     From sysKin(?):
1268 :     * bugfixes related to video flipping
1269 :     * bugfix for the 'crash at the end" bug
1270 :    
1271 :     modified files:
1272 :     TODO dshow/dshow.dsp dshow/src/CAbout.cpp dshow/src/CAbout.h
1273 :     dshow/src/CXvidDecoder.cpp dshow/src/CXvidDecoder.h
1274 :     dshow/src/xvid.ax.def
1275 :    
1276 :    
1277 :     2004-02-07 13:47:45 GMT patch-29
1278 :    
1279 :     Summary:
1280 :     VFW updates
1281 :     Revision:
1282 :     xvidcore--devapi4--1.0--patch-29
1283 :    
1284 :     Sorry feeling lazy about splitting this patch...
1285 :    
1286 :     From peter:
1287 :     * bitrate calculator
1288 :    
1289 :     From sysKin:
1290 :     * WMP9 bugfix
1291 :    
1292 :     modified files:
1293 :     TODO vfw/src/codec.c vfw/src/config.c vfw/src/config.h
1294 :     vfw/src/resource.h vfw/src/resource.rc
1295 :    
1296 :    
1297 :     2004-02-07 13:43:26 GMT patch-28
1298 :    
1299 :     Summary:
1300 :     Implicit overflow tuning for 2nd pass.
1301 :     Revision:
1302 :     xvidcore--devapi4--1.0--patch-28
1303 :    
1304 :     From sysKin:
1305 :     * When doing a bigger 2nd pass, the overflow loop must be more
1306 :     aggressive else no bonus bits are reinjected. So we can auto
1307 :     tune the overlow values in that case.
1308 :    
1309 :     modified files:
1310 :     src/plugins/plugin_2pass2.c
1311 :    
1312 :    
1313 :     2004-02-07 13:38:33 GMT patch-27
1314 :    
1315 :     Summary:
1316 :     GMC+interlaced bugfix in decoder.
1317 :     Revision:
1318 :     xvidcore--devapi4--1.0--patch-27
1319 :    
1320 :     From sysKin:
1321 :     * GMC+interlaced bugfix in decoder.
1322 :    
1323 :     modified files:
1324 :     src/decoder.c
1325 :    
1326 :    
1327 :     2004-02-07 13:35:16 GMT patch-26
1328 :    
1329 :     Summary:
1330 :     Reverted patch-23
1331 :     Revision:
1332 :     xvidcore--devapi4--1.0--patch-26
1333 :    
1334 :     From christoph:
1335 :     * reverted patch-23, old code was right.
1336 :     * Important typo for the YVYU csp (passing the y plane instead of u).
1337 :    
1338 :     modified files:
1339 :     src/image/image.c src/xvid.h
1340 :    
1341 :    
1342 :     2004-01-31 11:20:36 GMT patch-25
1343 :    
1344 :     Summary:
1345 :     DShow support for more mpeg4 fourccs.
1346 :     Revision:
1347 :     xvidcore--devapi4--1.0--patch-25
1348 :    
1349 :     From sysKin:
1350 :     * Added support for the MP4V fourcc.
1351 :    
1352 :     modified files:
1353 :     dshow/src/CAbout.cpp dshow/src/CAbout.h
1354 :     dshow/src/CXvidDecoder.cpp dshow/src/CXvidDecoder.h
1355 :     dshow/src/resource.h dshow/src/xvid.ax.rc
1356 :    
1357 :    
1358 :     2004-01-31 11:12:38 GMT patch-24
1359 :    
1360 :     Summary:
1361 :     DivX decoder compatibility
1362 :     Revision:
1363 :     xvidcore--devapi4--1.0--patch-24
1364 :    
1365 :     From sysKin:
1366 :     * DivX decoder compatibility improved for packed bitstreams.
1367 :     It should now detect them and play them fine.
1368 :    
1369 :     modified files:
1370 :     src/bitstream/bitstream.c
1371 :    
1372 :    
1373 :     2004-01-31 11:10:26 GMT patch-23
1374 :    
1375 :     Summary:
1376 :     YV12/I420 confusion fixed.
1377 :     Revision:
1378 :     xvidcore--devapi4--1.0--patch-23
1379 :    
1380 :     From christoph:
1381 :     * I420/YV12 were swapped since ... ages.
1382 :     * CSP_USER renamed to CSP_PLANAR
1383 :    
1384 :     modified files:
1385 :     src/encoder.c src/image/image.c src/xvid.h
1386 :    
1387 :    
1388 :     2004-01-31 10:53:20 GMT patch-22
1389 :    
1390 :     Summary:
1391 :     Arch separation for mem transfer functions
1392 :     Revision:
1393 :     xvidcore--devapi4--1.0--patch-22
1394 :    
1395 :     Arch separation for mem transfer functions
1396 :    
1397 :    
1398 :     modified files:
1399 :     src/utils/mem_transfer.h
1400 :    
1401 :    
1402 :     2004-01-27 14:47:08 GMT patch-21
1403 :    
1404 :     Summary:
1405 :     Write to registry Flip video flag in dshow
1406 :     Revision:
1407 :     xvidcore--devapi4--1.0--patch-21
1408 :    
1409 :     From sysKin:
1410 :     * The flip video flag is now saved in registry.
1411 :     * Changed internal flags name convention (use n prefix for all now)
1412 :    
1413 :     modified files:
1414 :     dshow/src/CAbout.cpp dshow/src/CAbout.h
1415 :     dshow/src/CXvidDecoder.cpp
1416 :    
1417 :    
1418 :     2004-01-27 14:42:52 GMT patch-20
1419 :    
1420 :     Summary:
1421 :     Added bitrate calc to VFW
1422 :     Revision:
1423 :     xvidcore--devapi4--1.0--patch-20
1424 :    
1425 :     From Pete:
1426 :     * Added bitrate calculator.
1427 :     * Changed up a few function calls to static type.
1428 :    
1429 :     modified files:
1430 :     vfw/src/codec.c vfw/src/codec.h vfw/src/config.c
1431 :     vfw/src/config.h vfw/src/resource.h vfw/src/resource.rc
1432 :     vfw/vfw.dsp
1433 :    
1434 :    
1435 :     2004-01-27 14:29:49 GMT patch-19
1436 :    
1437 :     Summary:
1438 :     Bugfix in decoder
1439 :     Revision:
1440 :     xvidcore--devapi4--1.0--patch-19
1441 :    
1442 :     From sysKin:
1443 :     * when stats are not used, don't write to the stats pointer.
1444 :    
1445 :     modified files:
1446 :     src/decoder.c
1447 :    
1448 :     #########################################################################
1449 : edgomez 1449 # 1.0.0 RC1 (Bistream Version 26)
1450 : edgomez 1382 #########################################################################
1451 :    
1452 :     2004-01-25 16:01:06 GMT patch-18
1453 :    
1454 :     Summary:
1455 :     Marking RC1
1456 :     Revision:
1457 :     xvidcore--devapi4--1.0--patch-18
1458 :    
1459 :     Marking RC1
1460 :    
1461 :     modified files:
1462 :     ChangeLog build/generic/configure.in src/xvid.h
1463 :    
1464 :    
1465 :     2004-01-25 15:37:57 GMT patch-17
1466 :    
1467 :     Summary:
1468 :     VFW update (again²)
1469 :     Revision:
1470 :     xvidcore--devapi4--1.0--patch-17
1471 :    
1472 :     From sysKin:
1473 :     * Added Constant Quant encoding.
1474 :    
1475 :     modified files:
1476 :     vfw/src/codec.c vfw/src/config.c vfw/src/config.h
1477 :     vfw/src/resource.rc
1478 :    
1479 :    
1480 :     2004-01-25 15:35:38 GMT patch-16
1481 :    
1482 :     Summary:
1483 :     Missing MB quants for PP.
1484 :     Revision:
1485 :     xvidcore--devapi4--1.0--patch-16
1486 :    
1487 :     From sysKin:
1488 :     * MB Quants are used by the PP code, so don't forget to update
1489 :     them even if the block is skipped or not coded.
1490 :    
1491 :     modified files:
1492 :     src/decoder.c
1493 :    
1494 :    
1495 :     2004-01-23 13:25:52 GMT patch-15
1496 :    
1497 :     Summary:
1498 :     VFW update (again)
1499 :     Revision:
1500 :     xvidcore--devapi4--1.0--patch-15
1501 :    
1502 :     From sysKin:
1503 :     * Status window updates.
1504 :     * Big resource.h cleanup, it seems msvc isn't able to do it automatically.
1505 :    
1506 :     modified files:
1507 :     vfw/src/resource.h vfw/src/resource.rc vfw/src/status.c
1508 :    
1509 :    
1510 :     2004-01-23 11:17:20 GMT patch-14
1511 :    
1512 :     Summary:
1513 :     VFW gcc warnings
1514 :     Revision:
1515 :     xvidcore--devapi4--1.0--patch-14
1516 :    
1517 :     VFW gcc warnings
1518 :    
1519 :    
1520 :     modified files:
1521 :     vfw/src/codec.c vfw/src/config.c vfw/src/config.h
1522 :    
1523 :    
1524 :     2004-01-22 20:54:53 GMT patch-13
1525 :    
1526 :     Summary:
1527 :     DShow updates.
1528 :     Revision:
1529 :     xvidcore--devapi4--1.0--patch-13
1530 :    
1531 :     From sysKin:
1532 :     * Fixed registry params type. Bool cannot be used or something weird
1533 :     happens when writing to registry
1534 :     * Defaults set to what the Reset widget sets.
1535 :    
1536 :     modified files:
1537 :     dshow/src/CAbout.h dshow/src/CXvidDecoder.cpp
1538 :    
1539 :    
1540 :     2004-01-22 20:35:27 GMT patch-12
1541 :    
1542 :     Summary:
1543 :     VFW updates.
1544 :     Revision:
1545 :     xvidcore--devapi4--1.0--patch-12
1546 :    
1547 :     From sysKin:
1548 :     * GUI improvements.
1549 :     * Stats fixing.
1550 :     * Automatic config clear upon installation.
1551 :     * Added postprocessing options in there too.
1552 :    
1553 :     modified files:
1554 :     vfw/bin/xvid.inf vfw/src/codec.c vfw/src/codec.h
1555 :     vfw/src/config.c vfw/src/config.h vfw/src/resource.h
1556 :     vfw/src/resource.rc vfw/src/status.c vfw/src/status.h
1557 :    
1558 :    
1559 :     2004-01-22 20:28:54 GMT patch-11
1560 :    
1561 :     Summary:
1562 :     Minor updates to text files
1563 :     Revision:
1564 :     xvidcore--devapi4--1.0--patch-11
1565 :    
1566 :     Minor updates to text files
1567 :    
1568 :    
1569 :     modified files:
1570 :     AUTHORS TODO
1571 :    
1572 :    
1573 :     2004-01-22 20:27:10 GMT patch-10
1574 :    
1575 :     Summary:
1576 :     2pass plugin changes.
1577 :     Revision:
1578 :     xvidcore--devapi4--1.0--patch-10
1579 :    
1580 :     From sysKin:
1581 :     * Disabled QPel during first pass as well.
1582 :     * Fix a mistaking condition when enabling largers 2nd passes.
1583 :    
1584 :     From ed.gomez:
1585 :     * Fix the fix logic. The previous fix enclosed a condition it should
1586 :     not have touched. So i removed the mistaking condition, which was
1587 :     wrong anyway as stated in the comment, and got back the sane condition
1588 :     test.
1589 :    
1590 :     modified files:
1591 :     src/plugins/plugin_2pass1.c src/plugins/plugin_2pass2.c
1592 :    
1593 :    
1594 :     2004-01-17 13:03:11 GMT patch-9
1595 :    
1596 :     Summary:
1597 :     Unitialized pointers during plugin creation.
1598 :     Revision:
1599 :     xvidcore--devapi4--1.0--patch-9
1600 :    
1601 :     From sysKin:
1602 :     - plugins which do not require private data were leaving the param2
1603 :     unitialized. Just init it to NULL. This bug wasn't causing any
1604 :     trouble anyway...
1605 :    
1606 :     modified files:
1607 :     src/plugins/plugin_dump.c src/plugins/plugin_psnr.c
1608 :     vfw/src/codec.c
1609 :    
1610 :    
1611 :     2004-01-17 01:09:01 GMT patch-8
1612 :    
1613 :     Summary:
1614 :     DShow forwwards AR information.
1615 :     Revision:
1616 :     xvidcore--devapi4--1.0--patch-8
1617 :    
1618 :     From syskin(?):
1619 :     - forward AR information to DShow framework.
1620 :     - grayed some widgets.
1621 :    
1622 :     modified files:
1623 :     dshow/src/CXvidDecoder.cpp dshow/src/CXvidDecoder.h
1624 :     dshow/src/xvid.ax.rc
1625 :    
1626 :    
1627 :     2004-01-07 13:50:29 GMT patch-7
1628 :    
1629 :     Summary:
1630 :     Scaled zones fix.
1631 :     Revision:
1632 :     xvidcore--devapi4--1.0--patch-7
1633 :    
1634 :     From sysKin (ideas from Koepi iirc):
1635 :     * Fix scaled zones computing prescaled data and so on instead of
1636 :     guessing everything with a global zone weight factor.
1637 :    
1638 :     modified files:
1639 :     src/plugins/plugin_2pass2.c
1640 :    
1641 :    
1642 :     2004-01-06 01:06:39 GMT patch-6
1643 :    
1644 :     Summary:
1645 :     Tab->Spaces in header only
1646 :     Revision:
1647 :     xvidcore--devapi4--1.0--patch-6
1648 :    
1649 :     Tab->Spaces in header only
1650 :    
1651 :    
1652 :     modified files:
1653 :     src/xvid.h
1654 :    
1655 :    
1656 :     2004-01-04 18:35:35 GMT patch-5
1657 :    
1658 :     Summary:
1659 :     Typo fixed
1660 :     Revision:
1661 :     xvidcore--devapi4--1.0--patch-5
1662 :    
1663 :     Typo fixed
1664 :    
1665 :    
1666 :     modified files:
1667 :     build/generic/configure.in
1668 :    
1669 :    
1670 :     2004-01-04 13:40:51 GMT patch-4
1671 :    
1672 :     Summary:
1673 :     VFW safer code.
1674 :     Revision:
1675 :     xvidcore--devapi4--1.0--patch-4
1676 :    
1677 :     From sysKin:
1678 :     * Protects some parts of the code depending on a previous
1679 :     xvidcore opening. Avoids resources leaking.
1680 :    
1681 :     modified files:
1682 :     vfw/src/codec.c vfw/src/driverproc.c
1683 :    
1684 :    
1685 :     2004-01-04 13:33:28 GMT patch-3
1686 :    
1687 :     Summary:
1688 :     Fixes VC debug target name
1689 :     Revision:
1690 :     xvidcore--devapi4--1.0--patch-3
1691 :    
1692 :     Fixes VC debug target name
1693 :    
1694 :    
1695 :     modified files:
1696 :     vfw/vfw.dsp
1697 :    
1698 :    
1699 :     2004-01-02 23:10:56 GMT patch-2
1700 :    
1701 :     Summary:
1702 :     Win32 linking policy revised.
1703 :     Revision:
1704 :     xvidcore--devapi4--1.0--patch-2
1705 :    
1706 :     Finnaly Win32 linking policy is to separate all XviD components:
1707 :     - xvidcore.dll exports XviD API
1708 :     - xvidvfw.dll links against xvidcore DLL
1709 :     - xviddshow.dll links against xvidcore DLL
1710 :    
1711 :     From sysKin:
1712 :     * Changed DShow linking policy in VS project file.
1713 :     * Changed VFW linking policy in VS project file.
1714 :     * Added runtime xvidcore.dll loading in DShow and VFW.
1715 :     * Installs xvidcore.dll along side with xvidvfw.dll.
1716 :    
1717 :     From ed.gomez:
1718 :     * Changed libxvidcore.dll mingw32/cygwin target name to
1719 :     xvidcore.dll in the configure script.
1720 :     * Changed xvid.dll VFW target name to xvidvfw.dll in the
1721 :     generic Makefile.
1722 :    
1723 :     PS: unlike CVS, i reverted back to MS build tools in VS project files
1724 :    
1725 :     modified files:
1726 :     build/generic/configure.in build/win32/libxvidcore.dsp
1727 :     dshow/dshow.dsp dshow/src/CXvidDecoder.cpp
1728 :     dshow/src/CXvidDecoder.h vfw/bin/Makefile vfw/bin/sources.inc
1729 :     vfw/bin/xvid.inf vfw/src/codec.c vfw/src/codec.h
1730 :     vfw/src/config.c vfw/src/config.h vfw/vfw.dsp
1731 :    
1732 :    
1733 :     2004-01-02 22:02:07 GMT patch-1
1734 :    
1735 :     Summary:
1736 :     Better seeking in dshow
1737 :     Revision:
1738 :     xvidcore--devapi4--1.0--patch-1
1739 :    
1740 :     From Michael:
1741 :     * Better DShow seeking.
1742 :    
1743 :     modified files:
1744 :     dshow/src/CXvidDecoder.cpp
1745 :    
1746 :    
1747 :     2004-01-02 12:28:39 GMT base-0
1748 :    
1749 :     Summary:
1750 :     tag of ed.gomez@free.fr--2003-1/xvidcore--devapi4--1.0--patch-162
1751 :     Revision:
1752 :     xvidcore--devapi4--1.0--base-0
1753 :    
1754 :     Archive cycling...
1755 :    
1756 :     #########################################################################
1757 : edgomez 1449 # 1.0.0 beta3 (Bistream Version 25)
1758 : edgomez 1382 #########################################################################
1759 :    
1760 :     2003-12-26 22:21:35 GMT patch-162
1761 :    
1762 :     Summary:
1763 :     Marking 1.0.0 beta3
1764 :     Revision:
1765 :     xvidcore--devapi4--1.0--patch-162
1766 :    
1767 :     Marking beta3
1768 :    
1769 :     modified files:
1770 :     ChangeLog TODO build/generic/configure.in src/xvid.h
1771 :    
1772 :    
1773 :     2003-12-25 20:57:52 GMT patch-161
1774 :    
1775 :     Summary:
1776 :     Thread safe PP.
1777 :     Revision:
1778 :     xvidcore--devapi4--1.0--patch-161
1779 :    
1780 :     From Michael:
1781 :     * Thread safe PP, context is now stored in DECODER struct.
1782 :    
1783 :     modified files:
1784 :     src/decoder.c src/decoder.h src/image/postprocessing.c
1785 :     src/image/postprocessing.h
1786 :    
1787 :    
1788 :     2003-12-25 20:49:36 GMT patch-160
1789 :    
1790 :     Summary:
1791 :     Added Turbo option to VFW GUI
1792 :     Revision:
1793 :     xvidcore--devapi4--1.0--patch-160
1794 :    
1795 :     From Michael:
1796 :     * Added turbo mode that enables all fast ME flags.
1797 :    
1798 :     modified files:
1799 :     vfw/src/codec.c vfw/src/config.c vfw/src/config.h
1800 :     vfw/src/resource.h vfw/src/resource.rc
1801 :    
1802 :    
1803 :     2003-12-25 20:46:16 GMT patch-159
1804 :    
1805 :     Summary:
1806 :     Cartoon mode usage written to stream.
1807 :     Revision:
1808 :     xvidcore--devapi4--1.0--patch-159
1809 :    
1810 :     From Michael:
1811 :     * @encoding: write cartoon usage in the version user data.
1812 :     * @decoding: detect cartoon flag appended to version string
1813 :     and force FILM PP disabling.
1814 :    
1815 :     modified files:
1816 :     src/bitstream/bitstream.c src/bitstream/bitstream.h
1817 :     src/decoder.c src/decoder.h src/encoder.c
1818 :    
1819 :    
1820 :     2003-12-21 13:34:03 GMT patch-158
1821 :    
1822 :     Summary:
1823 :     Removed unused var in VFW GUI
1824 :     Revision:
1825 :     xvidcore--devapi4--1.0--patch-158
1826 :    
1827 :     Removed unused var in VFW GUI
1828 :    
1829 :    
1830 :     modified files:
1831 :     vfw/src/config.c
1832 :    
1833 :    
1834 :     2003-12-21 13:32:52 GMT patch-157
1835 :    
1836 :     Summary:
1837 :     Two pass small update
1838 :     Revision:
1839 :     xvidcore--devapi4--1.0--patch-157
1840 :    
1841 :     from syskin:
1842 :     * allow second pass to be bigger than 1st one
1843 :     (not tested, the quant mapping formula may not be adapted for this
1844 :     usage, so take this change as experimental, and prefer doing
1845 :     second pass still smaller than 1st one)
1846 :     * let ivops benefit from positive overflow.
1847 :    
1848 :     from me:
1849 :     * set frame type in quant zones (was a buglet)
1850 :    
1851 :     modified files:
1852 :     src/plugins/plugin_2pass2.c
1853 :    
1854 :    
1855 :     2003-12-20 22:28:07 GMT patch-156
1856 :    
1857 :     Summary:
1858 :     New VFW defaults
1859 :     Revision:
1860 :     xvidcore--devapi4--1.0--patch-156
1861 :    
1862 :     From michael:
1863 :     * new VFW default values
1864 :    
1865 :     modified files:
1866 :     vfw/src/config.c
1867 :    
1868 :    
1869 :     2003-12-20 22:12:38 GMT patch-155
1870 :    
1871 :     Summary:
1872 :     Added ARGB colorspace.
1873 :     Revision:
1874 :     xvidcore--devapi4--1.0--patch-155
1875 :    
1876 :     Christoph Nägeli naegelic(at)ee{dot}ethz<dot>ch asked me to add C
1877 :     support for ARGB colorspace. He provided the encoding part, and i
1878 :     extended the original patch in order to have full support for ARGB
1879 :     both for encoding and decoding (though it's C only, read *slow*).
1880 :    
1881 :     modified files:
1882 :     src/image/colorspace.c src/image/colorspace.h
1883 :     src/image/image.c src/xvid.c src/xvid.h
1884 :    
1885 :    
1886 :     2003-12-20 21:29:37 GMT patch-154
1887 :    
1888 :     Summary:
1889 :     Added 2pass1 comment about fast 1st pass.
1890 :     Revision:
1891 :     xvidcore--devapi4--1.0--patch-154
1892 :    
1893 :     Just added a comment on fast 1st pass, so it explains why
1894 :     we do it that way and why some things are left aside.
1895 :    
1896 :     modified files:
1897 :     src/plugins/plugin_2pass1.c
1898 :    
1899 :    
1900 :     2003-12-20 20:03:51 GMT patch-153
1901 :    
1902 :     Summary:
1903 :     Win32 VC6 wrong libc linking.
1904 :     Revision:
1905 :     xvidcore--devapi4--1.0--patch-153
1906 :    
1907 :     From sysKin:
1908 :     * Changed single thread libc linking to multithreaded version.
1909 :    
1910 :     modified files:
1911 :     vfw/vfw.dsp
1912 :    
1913 :    
1914 :     2003-12-20 15:28:53 GMT patch-152
1915 :    
1916 :     Summary:
1917 :     VOL flags updating -- take #2
1918 :     Revision:
1919 :     xvidcore--devapi4--1.0--patch-152
1920 :    
1921 :     From sysKin:
1922 :     * VOL flags updates fix take #2
1923 :    
1924 :     modified files:
1925 :     src/encoder.c
1926 :    
1927 :    
1928 :     2003-12-20 15:10:30 GMT patch-151
1929 :    
1930 :     Summary:
1931 :     Fast ME tunings.
1932 :     Revision:
1933 :     xvidcore--devapi4--1.0--patch-151
1934 :    
1935 :     From michael:
1936 :     * fast refinement for 8x8 blocks
1937 :     * more reliable behavior for all fast ME decisions
1938 :    
1939 :     modified files:
1940 :     src/motion/estimation_bvop.c src/motion/estimation_pvop.c
1941 :     src/xvid.h
1942 :    
1943 :    
1944 :     2003-12-20 14:59:58 GMT patch-150
1945 :    
1946 :     Summary:
1947 :     VFW AR revamping -- take #2
1948 :     Revision:
1949 :     xvidcore--devapi4--1.0--patch-150
1950 :    
1951 :     From sysKin:
1952 :     * more AR revamping
1953 :    
1954 :     modified files:
1955 :     vfw/src/codec.c vfw/src/config.c vfw/src/config.h
1956 :     vfw/src/resource.h vfw/src/resource.rc
1957 :    
1958 :    
1959 :     2003-12-20 14:57:40 GMT patch-149
1960 :    
1961 :     Summary:
1962 :     Unbuffured IO for 1st pass stat files
1963 :     Revision:
1964 :     xvidcore--devapi4--1.0--patch-149
1965 :    
1966 :     Unbuffured IO for 1st pass stat files
1967 :    
1968 :    
1969 :     modified files:
1970 :     src/plugins/plugin_2pass1.c
1971 :    
1972 :    
1973 :     2003-12-18 17:44:07 GMT patch-148
1974 :    
1975 :     Summary:
1976 :     Forgotten bit of patch-141
1977 :     Revision:
1978 :     xvidcore--devapi4--1.0--patch-148
1979 :    
1980 :     Damn i forgot to merge the 1st pass changes... so lame, i tested the
1981 :     original patch but not the merged one.
1982 :    
1983 :     modified files:
1984 :     src/plugins/plugin_2pass1.c
1985 :    
1986 :    
1987 :     2003-12-18 14:45:39 GMT patch-147
1988 :    
1989 :     Summary:
1990 :     More postprocessing.
1991 :     Revision:
1992 :     xvidcore--devapi4--1.0--patch-147
1993 :    
1994 :     From michael:
1995 :     * added film noise effect.
1996 :     * moved postproc initialization to decoder initialization.
1997 :     * added support for this postproc filter into DShow.
1998 :    
1999 :     modified files:
2000 :     dshow/src/CAbout.cpp dshow/src/CAbout.h
2001 :     dshow/src/CXvidDecoder.cpp dshow/src/resource.h
2002 :     dshow/src/xvid.ax.rc src/decoder.c src/image/postprocessing.c
2003 :     src/image/postprocessing.h src/xvid.c src/xvid.h
2004 :    
2005 :    
2006 :     2003-12-18 14:38:19 GMT patch-146
2007 :    
2008 :     Summary:
2009 :     BFrames ME speed up flags.
2010 :     Revision:
2011 :     xvidcore--devapi4--1.0--patch-146
2012 :    
2013 :     From michael:
2014 :     * Added 3 ME flags to skip some bvop ME steps and thus speed up
2015 :     ME for bvops (at the expense of quality loss).
2016 :     - skip delta search
2017 :     - fast interpolate mode
2018 :     - early stop
2019 :    
2020 :     modified files:
2021 :     src/motion/estimation.h src/motion/estimation_bvop.c
2022 :     src/motion/estimation_common.c src/motion/estimation_pvop.c
2023 :     src/xvid.h
2024 :    
2025 :    
2026 :     2003-12-18 14:11:01 GMT patch-145
2027 :    
2028 :     Summary:
2029 :     VFW defaults changed
2030 :     Revision:
2031 :     xvidcore--devapi4--1.0--patch-145
2032 :    
2033 :     From michael:
2034 :     * disabled DXN profiles. Better not say we are compatible to avoid
2035 :     problems with DXN. Mostly because of VBV lacking.
2036 :     * Changed overflow defaults to 5/5/5 with new 2pass code.
2037 :    
2038 :     modified files:
2039 :     vfw/src/config.c
2040 :    
2041 :    
2042 :     2003-12-17 16:45:59 GMT patch-144
2043 :    
2044 :     Summary:
2045 :     Forgotten bit for AR support in VFW
2046 :     Revision:
2047 :     xvidcore--devapi4--1.0--patch-144
2048 :    
2049 :     Koepi might have forgotten to send me this change:
2050 :     * added resource id to the resource header
2051 :    
2052 :     modified files:
2053 :     vfw/src/resource.h
2054 :    
2055 :    
2056 :     2003-12-17 15:11:37 GMT patch-143
2057 :    
2058 :     Summary:
2059 :     Lower starting quantizer for CBR encoding
2060 :     Revision:
2061 :     xvidcore--devapi4--1.0--patch-143
2062 :    
2063 :     From christoph:
2064 :     * Lower starting quantizer for CBR encodings.
2065 :    
2066 :     modified files:
2067 :     src/plugins/plugin_single.c
2068 :    
2069 :    
2070 :     2003-12-17 15:04:33 GMT patch-142
2071 :    
2072 :     Summary:
2073 :     Small xvid_encraw updates
2074 :     Revision:
2075 :     xvidcore--devapi4--1.0--patch-142
2076 :    
2077 :     From christoph:
2078 :     * set upper frame size limit to 4096 pixels
2079 :     * GME refinement flag set where it belongs to.
2080 :    
2081 :     modified files:
2082 :     examples/xvid_encraw.c
2083 :    
2084 :    
2085 :     2003-12-17 15:01:52 GMT patch-141
2086 :    
2087 :     Summary:
2088 :     Two pass update.
2089 :     Revision:
2090 :     xvidcore--devapi4--1.0--patch-141
2091 :    
2092 :     This patch improves the two pass code, quantizer distribution is
2093 :     smoother and results seem to be better.
2094 :    
2095 :     * Two pass now scales only a specific part of the frame length.
2096 :     This required changing the stats file format (added a header+MV
2097 :     length field) and the xvid_plg_data_t structure (binary compatible).
2098 :     * Overflow improvement and degradation set to 10% instead of 60%
2099 :     Assymetric values may help...
2100 :     * Some cleanup work done on the encoder part of the API header.
2101 :    
2102 :     NB: plg data struct will be cleaned up before 1.0 so if you rely on it
2103 :     please read the header file to know which part will disappear.
2104 :    
2105 :     modified files:
2106 :     src/encoder.c src/plugins/plugin_2pass1.c
2107 :     src/plugins/plugin_2pass2.c src/xvid.h
2108 :    
2109 :    
2110 :     2003-12-17 13:53:34 GMT patch-140
2111 :    
2112 :     Summary:
2113 :     VOL flags fix.
2114 :     Revision:
2115 :     xvidcore--devapi4--1.0--patch-140
2116 :    
2117 :     From sysKin:
2118 :     * VOL flags updated as they should
2119 :    
2120 :     modified files:
2121 :     src/encoder.c
2122 :    
2123 :    
2124 :     2003-12-17 11:07:15 GMT patch-139
2125 :    
2126 :     Summary:
2127 :     VFW GUI Update.
2128 :     Revision:
2129 :     xvidcore--devapi4--1.0--patch-139
2130 :    
2131 :     From Koepi:
2132 :     * typo in GMC description
2133 :     * Added AR widget (experimental)
2134 :    
2135 :     modified files:
2136 :     vfw/src/codec.c vfw/src/config.c vfw/src/config.h
2137 :     vfw/src/resource.rc
2138 :    
2139 :    
2140 :     2003-12-14 12:44:36 GMT patch-138
2141 :    
2142 :     Summary:
2143 :     Fix PP crash
2144 :     Revision:
2145 :     xvidcore--devapi4--1.0--patch-138
2146 :    
2147 :     Fix PP crash
2148 :    
2149 :    
2150 :     modified files:
2151 :     src/decoder.c
2152 :    
2153 :    
2154 :     2003-12-14 12:43:21 GMT patch-137
2155 :    
2156 :     Summary:
2157 :     Texture bit counting for bframes
2158 :     Revision:
2159 :     xvidcore--devapi4--1.0--patch-137
2160 :    
2161 :     Texture bit counting for bframes
2162 :    
2163 :    
2164 :     modified files:
2165 :     src/bitstream/mbcoding.c
2166 :    
2167 :    
2168 :     2003-12-12 23:58:18 GMT patch-136
2169 :    
2170 :     Summary:
2171 :     SSE2 code enabled.
2172 :     Revision:
2173 :     xvidcore--devapi4--1.0--patch-136
2174 :    
2175 :     * SSE2 code enabled by default (only sane ones, _no_ idct)
2176 :     * Aligned data in xvid_bench to avoid crashes with SSE2 code
2177 :     because of unaligned read accesses.
2178 :    
2179 :     modified files:
2180 :     examples/xvid_bench.c src/xvid.c
2181 :    
2182 :    
2183 :     2003-12-12 22:50:33 GMT patch-135
2184 :    
2185 :     Summary:
2186 :     DShow update.
2187 :     Revision:
2188 :     xvidcore--devapi4--1.0--patch-135
2189 :    
2190 :     From michael (from nic):
2191 :     * Dshow updates (colorspace etc...)
2192 :     * Deblocking option.
2193 :    
2194 :     modified files:
2195 :     dshow/src/CAbout.cpp dshow/src/CAbout.h
2196 :     dshow/src/CXvidDecoder.cpp dshow/src/CXvidDecoder.h
2197 :     dshow/src/resource.h dshow/src/xvid.ax.rc
2198 :    
2199 :    
2200 :     2003-12-12 14:18:13 GMT patch-134
2201 :    
2202 :     Summary:
2203 :     Added missing postproc init
2204 :     Revision:
2205 :     xvidcore--devapi4--1.0--patch-134
2206 :    
2207 :     Added missing postproc init
2208 :    
2209 :    
2210 :     modified files:
2211 :     src/xvid.c
2212 :    
2213 :    
2214 :     2003-12-12 14:16:29 GMT patch-133
2215 :    
2216 :     Summary:
2217 :     YUV space clarifications and fix.
2218 :     Revision:
2219 :     xvidcore--devapi4--1.0--patch-133
2220 :    
2221 :     From christoph:
2222 :     * Put some comments
2223 :     * fixed UV swapping in USER case.
2224 :    
2225 :     modified files:
2226 :     src/image/image.c
2227 :    
2228 :    
2229 :     2003-12-12 14:01:52 GMT patch-132
2230 :    
2231 :     Summary:
2232 :     Default compilation flags change.
2233 :     Revision:
2234 :     xvidcore--devapi4--1.0--patch-132
2235 :    
2236 :     from christoph:
2237 :     - -fgcse was causing trouble on Suse9 gcc
2238 :     - -01->-02
2239 :    
2240 :     modified files:
2241 :     build/generic/configure.in
2242 :    
2243 :    
2244 :     2003-12-12 13:43:35 GMT patch-131
2245 :    
2246 :     Summary:
2247 :     Pixel Aspect Ratio support improvement.
2248 :     Revision:
2249 :     xvidcore--devapi4--1.0--patch-131
2250 :    
2251 :     * 1:1 VGA is default now (old behavior)
2252 :     * When passed EXT PAR type, we now (try to) sanityse the
2253 :     par_width/par_height value:
2254 :     - make it positive
2255 :     - using 0 (typical memset reseting) defaults to 1
2256 :     - simplify the PAR (using gcd)
2257 :     - then range it in [1..255] (can be lossy)
2258 :     * Specify valid range in API header
2259 :    
2260 :     Bitstream version bumped to 24.
2261 :    
2262 :     PS: this patch supersedes sysKin's one in CVS.
2263 :    
2264 :     modified files:
2265 :     src/encoder.c src/xvid.h
2266 :    
2267 :    
2268 :     2003-12-10 22:57:50 GMT patch-130
2269 :    
2270 :     Summary:
2271 :     xvid_decraw cmdline changes.
2272 :     Revision:
2273 :     xvidcore--devapi4--1.0--patch-130
2274 :    
2275 :     * -nframes -> -frames (why did i put a 'n' there ?)
2276 :     * -save changed its meaning, it now controls per frame Elementary
2277 :     Stream saving.
2278 :     * -o string is now independent from -save.
2279 :    
2280 :     So now it's possible to save both a ES file per frame + a ES file for the
2281 :     sequence.
2282 :    
2283 :     modified files:
2284 :     examples/xvid_encraw.c
2285 :    
2286 :    
2287 :     2003-12-10 15:08:20 GMT patch-129
2288 :    
2289 :     Summary:
2290 :     Decoder bugfixes.
2291 :     Revision:
2292 :     xvidcore--devapi4--1.0--patch-129
2293 :    
2294 :     From syskin:
2295 :     * bvops MBs were going banana from time to time because they were
2296 :     referencing wrong future ref MBs.
2297 :     * decoder now informs the client app about bvop lag, returning
2298 :     XVID_TYPE_NOTHING, up to the client app to display (or not)
2299 :     the bvop lag frame (black with error message).
2300 :    
2301 :     Fixing previous patch a bit:
2302 : edgomez 1541 * added $Id: ChangeLog,v 1.2.2.6 2004-08-29 10:56:56 edgomez Exp $ fields
2303 : edgomez 1382 * Fixed copyright
2304 :    
2305 :     modified files:
2306 :     dshow/src/CXvidDecoder.cpp src/decoder.c
2307 :     src/image/postprocessing.c src/image/postprocessing.h
2308 :    
2309 :    
2310 :     2003-12-10 14:53:58 GMT patch-128
2311 :    
2312 :     Summary:
2313 :     Deblocking code.
2314 :     Revision:
2315 :     xvidcore--devapi4--1.0--patch-128
2316 :    
2317 :     Patch from michael:
2318 :     * added deblocking code
2319 :    
2320 :     Merge work:
2321 :     * Added postprocessing.[ch] to project files
2322 :     * added #include "image/postprocessing.h" directive in decoder.c
2323 :     * new lines missing (gcc is so pedantic)
2324 :    
2325 :     NB: slice rendering + postprocessing is impossible. Slice rendering is
2326 :     somewhat abandoned.
2327 :    
2328 :     new files:
2329 :     src/image/.arch-ids/postprocessing.c.id
2330 :     src/image/.arch-ids/postprocessing.h.id
2331 :     src/image/postprocessing.c src/image/postprocessing.h
2332 :    
2333 :     modified files:
2334 :     build/generic/sources.inc build/win32/libxvidcore.dsp
2335 :     src/decoder.c src/xvid.h
2336 :    
2337 :    
2338 :     2003-12-08 18:33:26 GMT patch-127
2339 :    
2340 :     Summary:
2341 :     Don't read out of bounds
2342 :     Revision:
2343 :     xvidcore--devapi4--1.0--patch-127
2344 :    
2345 :     Don't read out of bounds
2346 :    
2347 :    
2348 :     modified files:
2349 :     src/plugins/plugin_2pass2.c
2350 :    
2351 :    
2352 :     2003-12-08 18:31:41 GMT patch-126
2353 :    
2354 :     Summary:
2355 :     Macroblock structure cleanup
2356 :     Revision:
2357 :     xvidcore--devapi4--1.0--patch-126
2358 :    
2359 :     Macroblock structure cleanup
2360 :    
2361 :    
2362 :     modified files:
2363 :     src/global.h
2364 :    
2365 :    
2366 :     2003-12-07 15:09:41 GMT patch-125
2367 :    
2368 :     Summary:
2369 :     Small fixes.
2370 :     Revision:
2371 :     xvidcore--devapi4--1.0--patch-125
2372 :    
2373 :     From gruel:
2374 :     * xvid.h: Minor color space correction.
2375 :    
2376 :     From sysKin:
2377 :     * codec.c: Zones fix
2378 :    
2379 :     modified files:
2380 :     src/xvid.h vfw/src/codec.c
2381 :    
2382 :    
2383 :     2003-12-07 14:57:14 GMT patch-124
2384 :    
2385 :     Summary:
2386 :     HUGE file handling in twopass.
2387 :     Revision:
2388 :     xvidcore--devapi4--1.0--patch-124
2389 :    
2390 :     Because of a lacking cast, two pass did not handle well some very
2391 :     large target size (bitrate mode is not affected).
2392 :    
2393 :     It should now be safe specifying target sizes up to 2^31kB which
2394 :     represents 2TB. Someone using XviD in studios ? ;-)
2395 :    
2396 :     modified files:
2397 :     src/plugins/plugin_2pass2.c
2398 :    
2399 :     #########################################################################
2400 : edgomez 1449 # 1.0.0 beta2 (Bitstream Version 23)
2401 : edgomez 1382 #########################################################################
2402 :    
2403 :     2003-12-05 14:43:53 GMT patch-123
2404 :    
2405 :     Summary:
2406 :     Marking 1.0.0 Beta2
2407 :     Revision:
2408 :     xvidcore--devapi4--1.0--patch-123
2409 :    
2410 :     Marking 1.0.0 Beta2
2411 :    
2412 :    
2413 :     modified files:
2414 :     ChangeLog TODO build/generic/configure.in
2415 :    
2416 :    
2417 :     2003-12-05 14:35:22 GMT patch-122
2418 :    
2419 :     Summary:
2420 :     Cap quants correctly (the best we can at least)
2421 :     Revision:
2422 :     xvidcore--devapi4--1.0--patch-122
2423 :    
2424 :     Cap quants correctly (the best we can at least)
2425 :    
2426 :    
2427 :     modified files:
2428 :     src/plugins/plugin_single.c
2429 :    
2430 :    
2431 :     2003-12-05 14:33:48 GMT patch-121
2432 :    
2433 :     Summary:
2434 :     Small glitch
2435 :     Revision:
2436 :     xvidcore--devapi4--1.0--patch-121
2437 :    
2438 :     Small glitch
2439 :    
2440 :    
2441 :     modified files:
2442 :     src/motion/vop_type_decision.c
2443 :    
2444 :    
2445 :     2003-12-05 14:06:19 GMT patch-120
2446 :    
2447 :     Summary:
2448 :     KFthresholding changes.
2449 :     Revision:
2450 :     xvidcore--devapi4--1.0--patch-120
2451 :    
2452 :     As user reports proved, the logic behind the min_key_interval was
2453 :     1/ misleading because the parameter is kfthreshold indeed and not
2454 :     a minimum keyframe interval
2455 :     2/ the formula was a bit too aggressive (removing 20% of bitrate
2456 :     per frame until distance to next iframe was 1)
2457 :    
2458 :     I posted a RFC to try to settle a decision on what behavior this
2459 :     setting should have. We have still have no clear answer so i prefer
2460 :     just fixing the misleading name right now and wait for a common
2461 :     position about its behavior later.
2462 :    
2463 :     Libraries are *binary* compatible, but *source code* compatibility
2464 :     is broken (rename rc_2pass2_t->min_key_interval to kfthreshold).
2465 :     This is probably the last API change.
2466 :    
2467 :     NB: fixes a type problem during scaling parameter computing which
2468 :     was causing insane pb_iboost_tax_ratio values.
2469 :    
2470 :     modified files:
2471 :     src/plugins/plugin_2pass2.c src/xvid.h vfw/src/codec.c
2472 :     vfw/src/config.c vfw/src/config.h
2473 :    
2474 :    
2475 :     2003-12-05 00:20:28 GMT patch-119
2476 :    
2477 :     Summary:
2478 :     ivop decision tuning.
2479 :     Revision:
2480 :     xvidcore--devapi4--1.0--patch-119
2481 :    
2482 :     ivop decision tuning from sysKin.
2483 :    
2484 :     modified files:
2485 :     src/motion/vop_type_decision.c
2486 :    
2487 :    
2488 :     2003-12-03 18:55:29 GMT patch-118
2489 :    
2490 :     Summary:
2491 :     VOSH header always written.
2492 :     Revision:
2493 :     xvidcore--devapi4--1.0--patch-118
2494 :    
2495 :     * profile is set to sane default value in BitstreamWriteVolHeaders
2496 :     * VOSH is now always written (note that the ending code is never
2497 :     written)
2498 :     * doubled variable removed from bvop estimation file.
2499 :    
2500 :     bitstream version set to 23
2501 :     Version 22 was used in CVS by a fix from michael for VOSH, it was just
2502 :     lacking the sane default value setting when profile is 0x00 from user.
2503 :     0x00 is a reserved profile ID in the spec so it wasn't making much sense
2504 :     to write 0x00.
2505 :    
2506 :     modified files:
2507 :     src/bitstream/bitstream.c src/motion/estimation_bvop.c
2508 :     src/xvid.h
2509 :    
2510 :    
2511 :     2003-12-03 15:29:30 GMT patch-117
2512 :    
2513 :     Summary:
2514 :     VFW GUI fixes.
2515 :     Revision:
2516 :     xvidcore--devapi4--1.0--patch-117
2517 :    
2518 :     min key was misleading because it was legacy code from dev-api-3.
2519 :    
2520 :     modified files:
2521 :     vfw/src/codec.c vfw/src/config.c vfw/src/resource.rc
2522 :    
2523 :    
2524 :     2003-12-03 15:22:25 GMT patch-116
2525 :    
2526 :     Summary:
2527 :     ME fixes.
2528 :     Revision:
2529 :     xvidcore--devapi4--1.0--patch-116
2530 :    
2531 :     From syskin:
2532 :     - small typo in chroma sad reset
2533 :     - code tweaking
2534 :     + adv diamond search used intead of mainsearch
2535 :     + and some other stuff
2536 :     - thresholds tuned.
2537 :    
2538 :     modified files:
2539 :     src/motion/estimation.h src/motion/estimation_bvop.c
2540 :     src/motion/estimation_pvop.c
2541 :    
2542 :    
2543 :     2003-12-01 11:17:20 GMT patch-115
2544 :    
2545 :     Summary:
2546 :     Small changes and interlacing bugfix.
2547 :     Revision:
2548 :     xvidcore--devapi4--1.0--patch-115
2549 :    
2550 :     * Interlacing bugfix, code got swapped in a very old patch
2551 :     (back in March)
2552 :     * xvid_encraw forces VOSH writing.
2553 :     * VFW GUI fixes.
2554 :    
2555 :     modified files:
2556 :     examples/Makefile examples/xvid_encraw.c
2557 :     src/utils/mbtransquant.c vfw/src/config.c vfw/src/resource.rc
2558 :    
2559 :    
2560 :     2003-11-30 15:47:41 GMT patch-114
2561 :    
2562 :     Summary:
2563 :     Thread safe MPEG4 quantization functions + xvid_bench update
2564 :     Revision:
2565 :     xvidcore--devapi4--1.0--patch-114
2566 :    
2567 :     * Thread safe MPEG4 quantization functions.
2568 :     Cleaned up version of patch provided by Michael
2569 :     - fixed compiling problems on gcc
2570 :     - added const qualifiers every where it was possible to help C compiler
2571 :     optimization.
2572 :     - added the mpeg_quant_matrices param to all ASM function prototype
2573 :     in comments (even if it's not used, that shows we do it deliberatly)
2574 :     - forces m[intra][0][0] = 8, otherwise XviD could write invalid streams.
2575 :     * Added real CRC computing in xvid_bench.c
2576 :    
2577 :     modified files:
2578 :     TODO examples/xvid_bench.c src/bitstream/bitstream.c
2579 :     src/decoder.c src/decoder.h src/encoder.c src/encoder.h
2580 :     src/motion/estimation.h src/motion/estimation_pvop.c
2581 :     src/motion/estimation_rd_based.c src/quant/quant.h
2582 :     src/quant/quant_h263.c src/quant/quant_matrix.c
2583 :     src/quant/quant_matrix.h src/quant/quant_mpeg.c
2584 :     src/quant/x86_asm/quantize_h263_3dne.asm
2585 :     src/quant/x86_asm/quantize_h263_mmx.asm
2586 :     src/quant/x86_asm/quantize_mpeg_mmx.asm
2587 :     src/quant/x86_asm/quantize_mpeg_xmm.asm
2588 :     src/utils/mbtransquant.c src/xvid.c
2589 :    
2590 :    
2591 :     2003-11-29 18:10:25 GMT patch-113
2592 :    
2593 :     Summary:
2594 :     Fixed csp asm rules for real?
2595 :     Revision:
2596 :     xvidcore--devapi4--1.0--patch-113
2597 :    
2598 :     Fixed csp asm rules for real?
2599 :    
2600 :    
2601 :     modified files:
2602 :     build/win32/libxvidcore.dsp
2603 :    
2604 :    
2605 :     2003-11-29 17:58:09 GMT patch-112
2606 :    
2607 :     Summary:
2608 :     TODO/ChangeLog updated
2609 :     Revision:
2610 :     xvidcore--devapi4--1.0--patch-112
2611 :    
2612 :     TODO/ChangeLog updated
2613 :    
2614 :    
2615 :     modified files:
2616 :     ChangeLog TODO
2617 :    
2618 :    
2619 :     #########################################################################
2620 : edgomez 1449 # 1.0.0 beta1 (Bitstream Version 21)
2621 : edgomez 1382 #########################################################################
2622 :    
2623 :     2003-11-29 17:21:08 GMT patch-111
2624 :    
2625 :     Summary:
2626 :     First beta marking
2627 :     Revision:
2628 :     xvidcore--devapi4--1.0--patch-111
2629 :    
2630 :     First beta marking
2631 :    
2632 :    
2633 :     modified files:
2634 :     build/generic/configure.in src/xvid.c src/xvid.h
2635 :    
2636 :    
2637 :     2003-11-29 16:59:14 GMT patch-110
2638 :    
2639 :     Summary:
2640 :     Catching up with CVS.
2641 :     Revision:
2642 :     xvidcore--devapi4--1.0--patch-110
2643 :    
2644 :     Changes from sysKin:
2645 :     * dquant optimization.
2646 :     * CBR fix
2647 :    
2648 :     modified files:
2649 :     src/encoder.c src/plugins/plugin_single.c
2650 :    
2651 :    
2652 :     2003-11-24 22:05:38 GMT patch-109
2653 :    
2654 :     Summary:
2655 :     Big level handling in trellis.
2656 :     Revision:
2657 :     xvidcore--devapi4--1.0--patch-109
2658 :    
2659 :     Trellis was treating big levels exactly the same way as lower ones.
2660 :     In some cases, trellis was doing wild optimizations favoring a 0
2661 :     because the distortion introduced by that big coeff change was acceptable.
2662 :     But visually this could result in some nasty blocks with wrong chroma
2663 :     information or similar brutal changes in other planes as well.
2664 :    
2665 :     Skal added big levels handling where trellis just tries to minimize
2666 :     the cost varying the run value only. No level modification is done
2667 :     anymore.
2668 :    
2669 :     modified files:
2670 :     TODO src/utils/mbtransquant.c
2671 :    
2672 :    
2673 :     2003-11-23 16:42:55 GMT patch-108
2674 :    
2675 :     Summary:
2676 :     Trellis for MPEG.
2677 :     Revision:
2678 :     xvidcore--devapi4--1.0--patch-108
2679 :    
2680 :     * Added trellis support for MPEg quantization type.
2681 :     * Changed RD fixed point precision, should help avoiding overflow
2682 :     (see the constant TL_SHIFT)
2683 :    
2684 :     NB: we still have some problems when trellis optimizes DC for big DC
2685 :     values.
2686 :    
2687 :     modified files:
2688 :     src/utils/mbtransquant.c
2689 :    
2690 :    
2691 :     2003-11-22 00:53:59 GMT patch-107
2692 :    
2693 :     Summary:
2694 :     Win32 lib project fix (bis)
2695 :     Revision:
2696 :     xvidcore--devapi4--1.0--patch-107
2697 :    
2698 :     * nasm >= 0.98.37 support in project file got reverted in a previous
2699 :     patch, push it back. Nota bene /O3 changed to /O2 for proper
2700 :     compilation with msvc compiler (everyone is not supposed to compile
2701 :     stuff with icc)
2702 :     * TODO update.
2703 :    
2704 :     modified files:
2705 :     TODO build/win32/libxvidcore.dsp
2706 :    
2707 :    
2708 :     2003-11-19 21:26:34 GMT patch-106
2709 :    
2710 :     Summary:
2711 :     updated bench crc
2712 :     Revision:
2713 :     xvidcore--devapi4--1.0--patch-106
2714 :    
2715 :     updated bench crc
2716 :    
2717 :    
2718 :     modified files:
2719 :     examples/xvid_bench.c
2720 :    
2721 :    
2722 :     2003-11-19 16:00:00 GMT patch-105
2723 :    
2724 :     Summary:
2725 :     Lumimasking fixes.
2726 :     Revision:
2727 :     xvidcore--devapi4--1.0--patch-105
2728 :    
2729 :     from sysKin:
2730 :     * New plugin hook entry XVID_PLG_FRAME that happens inside FrameCodeIPB
2731 :     when both type and quant are known. Added hook handling in all plugins.
2732 :     * Fixed lumimasking.
2733 :    
2734 :     from me:
2735 :     * small reverse commit in pvop estimation fixed.
2736 :    
2737 :     modified files:
2738 :     src/bitstream/bitstream.c src/bitstream/bitstream.h
2739 :     src/encoder.c src/motion/estimation_pvop.c
2740 :     src/plugins/plugin_2pass1.c src/plugins/plugin_2pass2.c
2741 :     src/plugins/plugin_dump.c src/plugins/plugin_lumimasking.c
2742 :     src/plugins/plugin_psnr.c src/plugins/plugin_single.c
2743 :     src/xvid.h
2744 :    
2745 :    
2746 :     2003-11-19 15:37:16 GMT patch-104
2747 :    
2748 :     Summary:
2749 :     Removed indirections from SearchData structure.
2750 :     Revision:
2751 :     xvidcore--devapi4--1.0--patch-104
2752 :    
2753 :     Patch from sysKin:
2754 :     * removed indirections in SearchData structure. CheckCandidate functions
2755 :     don't use a const SearchData pointer anymore, but they should be a bit
2756 :     faster because of the less numerous indirections.
2757 :    
2758 :     modified files:
2759 :     src/motion/estimation.h src/motion/estimation_bvop.c
2760 :     src/motion/estimation_common.c src/motion/estimation_gmc.c
2761 :     src/motion/estimation_pvop.c src/motion/estimation_rd_based.c
2762 :     src/motion/vop_type_decision.c
2763 :    
2764 :    
2765 :     2003-11-19 15:33:55 GMT patch-103
2766 :    
2767 :     Summary:
2768 :     Formula error in twopass code.
2769 :     Revision:
2770 :     xvidcore--devapi4--1.0--patch-103
2771 :    
2772 :     The reversing bframe formula in 2pass 2 was not right. This was in fact
2773 :     a test code i used when i was maintaining the code on its own branch.
2774 :     I should not have commited it :\
2775 :    
2776 :     Fixed :-)
2777 :    
2778 :     modified files:
2779 :     src/plugins/plugin_2pass2.c
2780 :    
2781 :    
2782 :     2003-11-18 21:41:08 GMT patch-102
2783 :    
2784 :     Summary:
2785 :     Another problem with mis/unitialized reads.
2786 :     Revision:
2787 :     xvidcore--devapi4--1.0--patch-102
2788 :    
2789 :     Michael introduced a fast subpel refine that uses a (iMinSAD2, currentQMV2)
2790 :     couple of data. The problem is that he plugged this in CheckCandidate16_qpel
2791 :     that is used outside this context, thus a if statement was traversed with
2792 :     garbage data in the standard subpel case. For perfection sake, using a
2793 :     iMinSAD=256*4096 value collects correct data even if it will not be used in
2794 :     the normal subpel case.
2795 :    
2796 :     modified files:
2797 :     src/motion/estimation_pvop.c
2798 :    
2799 :    
2800 :     2003-11-16 17:29:39 GMT patch-101
2801 :    
2802 :     Summary:
2803 :     The MEanalysis patch assumed bvops were always used
2804 :     Revision:
2805 :     xvidcore--devapi4--1.0--patch-101
2806 :    
2807 :     The MEanalysis patch assumed bvops were always used
2808 :    
2809 :    
2810 :     modified files:
2811 :     src/encoder.c
2812 :    
2813 :    
2814 :     2003-11-16 15:12:15 GMT patch-100
2815 :    
2816 :     Summary:
2817 :     MEanalysis using wrong mvs + bframe search using wrong mvs
2818 :     Revision:
2819 :     xvidcore--devapi4--1.0--patch-100
2820 :    
2821 :     Valgrind reported lot of unitialized reads. These unitialized reads
2822 :     helped sysKin finding three bugs:
2823 :     - ZeroMacroblock did not reset the cbp field. So for some skipped
2824 :     blocks, a test was done on the cbp value...
2825 :     - MEanalysis was using wrong mvs from the current bvop (unitialized
2826 :     or just wrong in current context). That's because in devapi3, bframes
2827 :     used to share the same mvs array wheras now, it's one array a bvop.
2828 :     - Collocated skipped MBs for a bvop didn't reset mvs[0] and b_mvs[0].
2829 :    
2830 :     modified files:
2831 :     src/encoder.c src/motion/estimation_bvop.c src/motion/motion.h
2832 :     src/motion/motion_inlines.h src/motion/vop_type_decision.c
2833 :    
2834 :    
2835 :     2003-11-15 15:21:09 GMT patch-99
2836 :    
2837 :     Summary:
2838 :     Small fixes
2839 :     Revision:
2840 :     xvidcore--devapi4--1.0--patch-99
2841 :    
2842 :     Small fixes
2843 :    
2844 :    
2845 :     modified files:
2846 :     src/encoder.c src/motion/vop_type_decision.c
2847 :    
2848 :    
2849 :     2003-11-15 15:02:47 GMT patch-98
2850 :    
2851 :     Summary:
2852 :     DShow update + libxvidcore project file update
2853 :     Revision:
2854 :     xvidcore--devapi4--1.0--patch-98
2855 :    
2856 :     From peter:
2857 :     * DShow now links against libxvidcore.lib
2858 :     * Some updates to the libxvidcore project file
2859 :    
2860 :     modified files:
2861 :     TODO build/win32/libxvidcore.dsp dshow/dshow.dsp
2862 :     dshow/src/CXvidDecoder.cpp dshow/src/CXvidDecoder.h
2863 :    
2864 :    
2865 :     2003-11-15 01:51:28 GMT patch-97
2866 :    
2867 :     Summary:
2868 :     Small updates to doc files
2869 :     Revision:
2870 :     xvidcore--devapi4--1.0--patch-97
2871 :    
2872 :     Small updates to doc files
2873 :    
2874 :    
2875 :     modified files:
2876 :     AUTHORS CodingStyle TODO doc/INSTALL doc/README
2877 :    
2878 :     renamed files:
2879 :     .arch-ids/authors.txt.id
2880 :     ==> .arch-ids/AUTHORS.id
2881 :     .arch-ids/todo.txt.id
2882 :     ==> .arch-ids/TODO.id
2883 :     authors.txt
2884 :     ==> AUTHORS
2885 :     todo.txt
2886 :     ==> TODO
2887 :    
2888 :    
2889 :     2003-11-14 11:23:55 GMT patch-96
2890 :    
2891 :     Summary:
2892 :     Updated ChangeLog
2893 :     Revision:
2894 :     xvidcore--devapi4--1.0--patch-96
2895 :    
2896 :     Updated ChangeLog
2897 :    
2898 :    
2899 :     modified files:
2900 :     ChangeLog
2901 :    
2902 :    
2903 :     2003-11-13 23:09:34 GMT patch-95
2904 :    
2905 :     Summary:
2906 :     8x8 16bit Block SSE optimization.
2907 :     Revision:
2908 :     xvidcore--devapi4--1.0--patch-95
2909 :    
2910 :     MMXed the calculation of SSE for 8x8 16bit blocks. This helps quite
2911 :     a lot VHQ=4 mode.
2912 :    
2913 :     My tests show with trellis:chroma_me:
2914 :     - ~20% speed improvement for vhq=4.
2915 :     - at least 5% when using vhq=1.
2916 :    
2917 :     Of course this speedup vanishes if more CPU intensive features are used.
2918 :     CruNcher who used gmc/qpel, noticed "only" a ~5% speed improvement.
2919 :    
2920 :     NB: i'm of course talking about overall speed improvement. Such a small
2921 :     patch for such a big improvement :-)
2922 :    
2923 :     modified files:
2924 :     src/motion/estimation_rd_based.c src/motion/sad.c
2925 :     src/motion/sad.h src/motion/x86_asm/sad_mmx.asm src/xvid.c
2926 :    
2927 :    
2928 :     2003-11-13 22:34:33 GMT patch-94
2929 :    
2930 :     Summary:
2931 :     Various small bug fixes.
2932 :     Revision:
2933 :     xvidcore--devapi4--1.0--patch-94
2934 :    
2935 :     * encoder.c: GMC code fix in encoder.c. Now gmcval is initialized correctly
2936 :     when using GME.
2937 :     * xvid_decraw.c: Fix elementary stream output.
2938 :     * plugin_2pass2.c: Small parsing bug in stats reading in 2pass2.
2939 :     * decoder.c: Read resync markers in bframes.
2940 :    
2941 :     modified files:
2942 :     examples/xvid_decraw.c src/decoder.c src/encoder.c
2943 :     src/plugins/plugin_2pass2.c
2944 :    
2945 :    
2946 :     2003-11-11 16:24:05 GMT patch-93
2947 :    
2948 :     Summary:
2949 :     VFW update for overflow control
2950 :     Revision:
2951 :     xvidcore--devapi4--1.0--patch-93
2952 :    
2953 :     From Koepi.
2954 :     * Added widget and code for overflow control strength.
2955 :     * Removed widgets for payback options and kfthresholds.
2956 :     * Activated frame stats in DebugOutputView all the time.
2957 :    
2958 :     From me:
2959 :     * Activated static motion detection in cartoon mode.
2960 :    
2961 :     modified files:
2962 :     vfw/src/codec.c vfw/src/config.c vfw/src/resource.h
2963 :     vfw/src/resource.rc
2964 :    
2965 :    
2966 :     2003-11-09 20:47:47 GMT patch-92
2967 :    
2968 :     Summary:
2969 :     New two pass code.
2970 :     Revision:
2971 :     xvidcore--devapi4--1.0--patch-92
2972 :    
2973 :     New two pass code. I may say it's just a fixed version, though it looks
2974 :     more like a "take all the ideas and write it again" version. It performs
2975 :     better with all natural sequences i have and a bit worse with anime.
2976 :    
2977 :     Including it now, allow me improving the code during the beta releases.
2978 :    
2979 :     modified files:
2980 :     src/encoder.c src/plugins/plugin_2pass1.c
2981 :     src/plugins/plugin_2pass2.c src/xvid.h vfw/src/codec.c
2982 :     vfw/src/config.c vfw/src/config.h
2983 :    
2984 :    
2985 :     2003-11-09 17:07:16 GMT patch-91
2986 :    
2987 :     Summary:
2988 :     Fixes for bframe compensation (used in psnr tests).
2989 :     Revision:
2990 :     xvidcore--devapi4--1.0--patch-91
2991 :    
2992 :     * transfer_8to16_sub2_(c|mmx|xmm|3dne) write back the compensated
2993 :     result to current frame pointer.
2994 :     * transfer_8to16_sub2_mmx uses proper rounding (a+b+1)/2. The +1
2995 :     operation was missing.
2996 :     * Blocks skipped in bframes must be compensated for psnr computing.
2997 :    
2998 :     modified files:
2999 :     src/encoder.c src/motion/estimation_bvop.c
3000 :     src/utils/mem_transfer.c
3001 :     src/utils/x86_asm/mem_transfer_3dne.asm
3002 :     src/utils/x86_asm/mem_transfer_mmx.asm
3003 :    
3004 :    
3005 :     2003-11-05 16:05:44 GMT patch-90
3006 :    
3007 :     Summary:
3008 :     Speed improvement not wasting setedges and interpolate calls.
3009 :     Revision:
3010 :     xvidcore--devapi4--1.0--patch-90
3011 :    
3012 :     Patch from syskin.
3013 :    
3014 :     * This patch avoids calling setedges and interpolate for uneeded cases:
3015 :     - setedges is only called once per frame.
3016 :     - interpolate is called only when the previous rounding
3017 :     was different from the one needed.
3018 :     * Interpolation has been optimized a bit for qpel case, we do the
3019 :     hv pass down to top to use the cache more efficiently (hope so).
3020 :    
3021 :     modified files:
3022 :     src/encoder.c src/encoder.h src/image/image.c
3023 :    
3024 :    
3025 :     2003-11-03 19:51:12 GMT patch-89
3026 :    
3027 :     Summary:
3028 :     SSE2 dev16 fix + xvid_bench DCT block alignments.
3029 :     Revision:
3030 :     xvidcore--devapi4--1.0--patch-89
3031 :    
3032 :     * Small error fixed by Skal in his dev16 code (missing pshufd).
3033 :     * Blocks used by DCT tests are now aligned with DECLARE_ALIGNED_MATRIX
3034 :     this avoids the well know segfaults when using SSE2 instructions that
3035 :     suppose data alignment.
3036 :    
3037 :     modified files:
3038 :     examples/xvid_bench.c src/motion/x86_asm/sad_sse2.asm
3039 :    
3040 :    
3041 :     2003-11-03 15:42:23 GMT patch-88
3042 :    
3043 :     Summary:
3044 :     Align .rodata section for non coff objects
3045 :     Revision:
3046 :     xvidcore--devapi4--1.0--patch-88
3047 :    
3048 :     Align .rodata section for non coff objects
3049 :    
3050 :    
3051 :     modified files:
3052 :     src/bitstream/x86_asm/cbp_mmx.asm
3053 :     src/bitstream/x86_asm/cbp_sse2.asm
3054 :     src/dct/x86_asm/fdct_mmx_ffmpeg.asm
3055 :     src/dct/x86_asm/fdct_mmx_skal.asm
3056 :     src/dct/x86_asm/fdct_sse2_skal.asm
3057 :     src/dct/x86_asm/idct_3dne.asm src/dct/x86_asm/idct_mmx.asm
3058 :     src/dct/x86_asm/idct_sse2_dmitry.asm
3059 :     src/dct/x86_asm/simple_idct_mmx.asm
3060 :     src/image/x86_asm/colorspace_rgb_mmx.asm
3061 :     src/image/x86_asm/colorspace_yuyv_mmx.asm
3062 :     src/image/x86_asm/interpolate8x8_3dn.asm
3063 :     src/image/x86_asm/interpolate8x8_3dne.asm
3064 :     src/image/x86_asm/interpolate8x8_mmx.asm
3065 :     src/image/x86_asm/interpolate8x8_xmm.asm
3066 :     src/image/x86_asm/qpel_mmx.asm
3067 :     src/image/x86_asm/reduced_mmx.asm
3068 :     src/motion/x86_asm/sad_3dn.asm src/motion/x86_asm/sad_3dne.asm
3069 :     src/motion/x86_asm/sad_mmx.asm src/motion/x86_asm/sad_sse2.asm
3070 :     src/motion/x86_asm/sad_xmm.asm
3071 :     src/quant/x86_asm/quantize_h263_3dne.asm
3072 :     src/quant/x86_asm/quantize_h263_mmx.asm
3073 :     src/quant/x86_asm/quantize_mpeg_mmx.asm
3074 :     src/quant/x86_asm/quantize_mpeg_xmm.asm
3075 :     src/utils/x86_asm/cpuid.asm
3076 :     src/utils/x86_asm/interlacing_mmx.asm
3077 :     src/utils/x86_asm/mem_transfer_3dne.asm
3078 :    
3079 :    
3080 :     2003-11-02 23:01:43 GMT patch-87
3081 :    
3082 :     Summary:
3083 :     SSE2 update
3084 :     Revision:
3085 :     xvidcore--devapi4--1.0--patch-87
3086 :    
3087 :     * Added Dmitry SSE2 iDCT code back.
3088 :     * Plugged Dmitry iDCT as default for SSE2
3089 :     * Fixed a bug in xvid_bench that was making it would test some CPU
3090 :     instruction set w/o host CPU support. xvidcore init was simply
3091 :     discarding irrelevant cpu flags.
3092 :    
3093 :     new files:
3094 :     src/dct/x86_asm/.arch-ids/idct_sse2_dmitry.asm.id
3095 :     src/dct/x86_asm/idct_sse2_dmitry.asm
3096 :    
3097 :     modified files:
3098 :     build/generic/sources.inc build/win32/libxvidcore.dsp
3099 :     examples/xvid_bench.c src/dct/fdct.h src/dct/idct.h src/xvid.c
3100 :    
3101 :    
3102 :     2003-10-31 14:53:26 GMT patch-86
3103 :    
3104 :     Summary:
3105 :     Better handling of old windres versions + GNU make dependency.
3106 :     Revision:
3107 :     xvidcore--devapi4--1.0--patch-86
3108 :    
3109 :     Old versions of GNU windres (<2.14) don't have the same short options.
3110 :     But long options remain the same so it's better to use long option names
3111 :     to have ful compatibility with older versions.
3112 :    
3113 :     The Makefile appears to be dependent on GNU make because shell expansion
3114 :     for retrieving the path of the Makefile is wrong when using `` even with
3115 :     a single expansion assignment :=. It keeps being expanded when used.
3116 :    
3117 :     modified files:
3118 :     doc/INSTALL vfw/bin/Makefile
3119 :    
3120 :    
3121 :     2003-10-29 11:31:28 GMT patch-85
3122 :    
3123 :     Summary:
3124 :     Added sse2 f/iDCT code from skal
3125 :     Revision:
3126 :     xvidcore--devapi4--1.0--patch-85
3127 :    
3128 :     * Added sse2 f/iDCT code from skal
3129 :     * Added hooking in xvid.c
3130 :    
3131 :     new files:
3132 :     src/dct/x86_asm/.arch-ids/fdct_sse2_skal.asm.id
3133 :     src/dct/x86_asm/fdct_sse2_skal.asm
3134 :    
3135 :     modified files:
3136 :     build/generic/sources.inc build/win32/libxvidcore.dsp
3137 :     src/xvid.c
3138 :    
3139 :    
3140 :     2003-10-29 00:19:10 GMT patch-84
3141 :    
3142 :     Summary:
3143 :     Fix the static motion detection
3144 :     Revision:
3145 :     xvidcore--devapi4--1.0--patch-84
3146 :    
3147 :     Fix the static motion detection
3148 :    
3149 :    
3150 :     modified files:
3151 :     src/motion/estimation_pvop.c
3152 :    
3153 :    
3154 :     2003-10-28 23:39:46 GMT patch-83
3155 :    
3156 :     Summary:
3157 :     Added cartoon option handling.
3158 :     Revision:
3159 :     xvidcore--devapi4--1.0--patch-83
3160 :    
3161 :     Added cartoon widgets + handling code.
3162 :    
3163 :     NB: static motion detection is disabled because of crashes on P4
3164 :     cpus.
3165 :    
3166 :     modified files:
3167 :     vfw/src/codec.c vfw/src/config.c vfw/src/config.h
3168 :     vfw/src/resource.h vfw/src/resource.rc
3169 :    
3170 :    
3171 :     2003-10-28 17:44:09 GMT patch-82
3172 :    
3173 :     Summary:
3174 :     ASM cleanups;
3175 :     Revision:
3176 :     xvidcore--devapi4--1.0--patch-82
3177 :    
3178 :     * Applied same style to all asm files
3179 :     * Replaced current sad sse2 operators with skal's ones
3180 :     * Removed old and unused colorspace asm files
3181 :    
3182 :     removed files:
3183 :     src/image/x86_asm/.arch-ids/rgb_to_yv12_mmx.asm.id
3184 :     src/image/x86_asm/.arch-ids/yuv_to_yv12_mmx.asm.id
3185 :     src/image/x86_asm/.arch-ids/yuyv_to_yv12_mmx.asm.id
3186 :     src/image/x86_asm/.arch-ids/yv12_to_rgb24_mmx.asm.id
3187 :     src/image/x86_asm/.arch-ids/yv12_to_rgb32_mmx.asm.id
3188 :     src/image/x86_asm/.arch-ids/yv12_to_yuyv_mmx.asm.id
3189 :     src/image/x86_asm/rgb_to_yv12_mmx.asm
3190 :     src/image/x86_asm/yuv_to_yv12_mmx.asm
3191 :     src/image/x86_asm/yuyv_to_yv12_mmx.asm
3192 :     src/image/x86_asm/yv12_to_rgb24_mmx.asm
3193 :     src/image/x86_asm/yv12_to_rgb32_mmx.asm
3194 :     src/image/x86_asm/yv12_to_yuyv_mmx.asm
3195 :    
3196 :     modified files:
3197 :     build/generic/sources.inc src/bitstream/x86_asm/cbp_3dne.asm
3198 :     src/bitstream/x86_asm/cbp_mmx.asm
3199 :     src/bitstream/x86_asm/cbp_sse2.asm
3200 :     src/dct/x86_asm/fdct_mmx_ffmpeg.asm
3201 :     src/dct/x86_asm/fdct_mmx_skal.asm
3202 :     src/dct/x86_asm/idct_3dne.asm src/dct/x86_asm/idct_mmx.asm
3203 :     src/dct/x86_asm/simple_idct_mmx.asm
3204 :     src/image/x86_asm/colorspace_mmx.inc
3205 :     src/image/x86_asm/colorspace_rgb_mmx.asm
3206 :     src/image/x86_asm/colorspace_yuv_mmx.asm
3207 :     src/image/x86_asm/colorspace_yuyv_mmx.asm
3208 :     src/image/x86_asm/interpolate8x8_3dn.asm
3209 :     src/image/x86_asm/interpolate8x8_3dne.asm
3210 :     src/image/x86_asm/interpolate8x8_mmx.asm
3211 :     src/image/x86_asm/interpolate8x8_xmm.asm
3212 :     src/image/x86_asm/qpel_mmx.asm
3213 :     src/image/x86_asm/reduced_mmx.asm
3214 :     src/motion/x86_asm/sad_3dn.asm src/motion/x86_asm/sad_3dne.asm
3215 :     src/motion/x86_asm/sad_mmx.asm src/motion/x86_asm/sad_sse2.asm
3216 :     src/motion/x86_asm/sad_xmm.asm
3217 :     src/quant/x86_asm/quantize_h263_3dne.asm
3218 :     src/quant/x86_asm/quantize_h263_mmx.asm
3219 :     src/quant/x86_asm/quantize_mpeg_mmx.asm
3220 :     src/quant/x86_asm/quantize_mpeg_xmm.asm
3221 :     src/utils/x86_asm/cpuid.asm
3222 :     src/utils/x86_asm/interlacing_mmx.asm
3223 :     src/utils/x86_asm/mem_transfer_3dne.asm
3224 :     src/utils/x86_asm/mem_transfer_mmx.asm
3225 :    
3226 :    
3227 :     2003-10-27 01:13:47 GMT patch-81
3228 :    
3229 :     Summary:
3230 :     d_mv_bits speedup from sysKin
3231 :     Revision:
3232 :     xvidcore--devapi4--1.0--patch-81
3233 :    
3234 :     d_mv_bits speedup from sysKin
3235 :    
3236 :    
3237 :     modified files:
3238 :     src/motion/motion_inlines.h
3239 :    
3240 :    
3241 :     2003-10-27 00:55:51 GMT patch-80
3242 :    
3243 :     Summary:
3244 :     fDCT changes, new asm CodingStyle applied to dct dir
3245 :     Revision:
3246 :     xvidcore--devapi4--1.0--patch-80
3247 :    
3248 :     * Ported the ffmpeg fDCT functions (mmx and xmm).
3249 :     * Modified the skal's versions a bit to allow rolling loops.
3250 :     * Activated Skal's fDCTs (unrolled versions) for mmx _and_ xmm
3251 :     (old code was ignoring xmm versions)
3252 :     * Removed the SSE2 versions (they'll be back later)
3253 :     * .data -> .rodata
3254 :     * Applied announced asm CodingStyle to the dct dir
3255 :     (I'll have to add a section with the said CodingStyle)
3256 :    
3257 :     modified files:
3258 :     build/generic/sources.inc build/win32/libxvidcore.dsp
3259 :     src/dct/fdct.h src/dct/idct.h
3260 :     src/dct/x86_asm/fdct_mmx_ffmpeg.asm
3261 :     src/dct/x86_asm/fdct_mmx_skal.asm
3262 :     src/dct/x86_asm/idct_3dne.asm src/dct/x86_asm/idct_mmx.asm
3263 :     src/dct/x86_asm/simple_idct_mmx.asm src/xvid.c
3264 :    
3265 :     renamed files:
3266 :     src/dct/x86_asm/.arch-ids/fdct_mmx.asm.id
3267 :     ==> src/dct/x86_asm/.arch-ids/fdct_mmx_ffmpeg.asm.id
3268 :     src/dct/x86_asm/.arch-ids/fdct_xmm.asm.id
3269 :     ==> src/dct/x86_asm/.arch-ids/fdct_mmx_skal.asm.id
3270 :     src/dct/x86_asm/fdct_mmx.asm
3271 :     ==> src/dct/x86_asm/fdct_mmx_ffmpeg.asm
3272 :     src/dct/x86_asm/fdct_xmm.asm
3273 :     ==> src/dct/x86_asm/fdct_mmx_skal.asm
3274 :    
3275 :    
3276 :     2003-10-25 13:48:42 GMT patch-79
3277 :    
3278 :     Summary:
3279 :     BQuant->PQuant fix.
3280 :     Revision:
3281 :     xvidcore--devapi4--1.0--patch-79
3282 :    
3283 :     When using closed_gop, a BFrame before an IFrame is turned into a PFrame.
3284 :     Thus is original quant has to be computed back, a rounding was causing these
3285 :     frames to be orig_quant-1. As a consequence we had very big frames before
3286 :     the IFrame, loosing many bits for nearly no visual benefit.
3287 :    
3288 :     modified files:
3289 :     src/encoder.c
3290 :    
3291 :    
3292 :     2003-10-25 10:26:48 GMT patch-78
3293 :    
3294 :     Summary:
3295 :     Added closed gop option to xvid_encraw
3296 :     Revision:
3297 :     xvidcore--devapi4--1.0--patch-78
3298 :    
3299 :     Added closed gop option to xvid_encraw
3300 :    
3301 :    
3302 :     modified files:
3303 :     examples/xvid_encraw.c
3304 :    
3305 :    
3306 :     2003-10-24 17:39:53 GMT patch-77
3307 :    
3308 :     Summary:
3309 :     RD fixes.
3310 :     Revision:
3311 :     xvidcore--devapi4--1.0--patch-77
3312 :    
3313 :     Inter RD optimization relied on buggy functions to predict bitsize.
3314 :    
3315 :     modified files:
3316 :     src/motion/estimation_pvop.c src/motion/motion_inlines.h
3317 :    
3318 :    
3319 :     2003-10-22 15:48:01 GMT patch-76
3320 :    
3321 :     Summary:
3322 :     Small INSTALL update due to previous build patches.
3323 :     Revision:
3324 :     xvidcore--devapi4--1.0--patch-76
3325 :    
3326 :     Small INSTALL update due to previous build patches.
3327 :    
3328 :     modified files:
3329 :     doc/INSTALL
3330 :    
3331 :    
3332 :     2003-10-21 21:27:46 GMT patch-75
3333 :    
3334 :     Summary:
3335 :     Removed unused next_block vars.
3336 :     Revision:
3337 :     xvidcore--devapi4--1.0--patch-75
3338 :    
3339 :     Removed unused next_block vars.
3340 :    
3341 :     modified files:
3342 :     src/decoder.c
3343 :    
3344 :    
3345 :     2003-10-21 21:24:15 GMT patch-74
3346 :    
3347 :     Summary:
3348 :     VFW build changes.
3349 :     Revision:
3350 :     xvidcore--devapi4--1.0--patch-74
3351 :    
3352 :     The build system has been modified to look like the core lib one
3353 :     minus the configure system.
3354 :    
3355 :     modified files:
3356 :     vfw/bin/Makefile vfw/bin/sources.inc vfw/src/config.c
3357 :     vfw/vfw.dsp
3358 :    
3359 :     renamed files:
3360 :     vfw/bin/.arch-ids/Makefile.cygwin.id
3361 :     ==> vfw/bin/.arch-ids/Makefile.id
3362 :     vfw/bin/.arch-ids/Makefile.inc.id
3363 :     ==> vfw/bin/.arch-ids/sources.inc.id
3364 :     vfw/bin/Makefile.cygwin
3365 :     ==> vfw/bin/Makefile
3366 :     vfw/bin/Makefile.inc
3367 :     ==> vfw/bin/sources.inc
3368 :     vfw/src/.arch-ids/config.rc.id
3369 :     ==> vfw/src/.arch-ids/resource.rc.id
3370 :     vfw/src/config.rc
3371 :     ==> vfw/src/resource.rc
3372 :    
3373 :    
3374 :     2003-10-21 17:00:09 GMT patch-73
3375 :    
3376 :     Summary:
3377 :     Decoder cleanups and speedup
3378 :     Revision:
3379 :     xvidcore--devapi4--1.0--patch-73
3380 :    
3381 :     cleanups, speedups from sysKin
3382 :    
3383 :     modified files:
3384 :     src/decoder.c
3385 :    
3386 :    
3387 :     2003-10-21 16:22:15 GMT patch-72
3388 :    
3389 :     Summary:
3390 :     Build fixes for newer nasm versions.
3391 :     Revision:
3392 :     xvidcore--devapi4--1.0--patch-72
3393 :    
3394 :     nasm does not take care of adding trailing slashes to include paths.
3395 :     A patch to upstream authors has been refused because "the backslash()
3396 :     feature has been abandoned to get back to old nasm behavior"
3397 :    
3398 :     Their choice is kinda stupid as nasm is now open to user mistakes... :\
3399 :     So we fix that on ou side.
3400 :    
3401 :     modified files:
3402 :     build/generic/configure.in build/win32/libxvidcore.dsp
3403 :    
3404 :    
3405 :     2003-10-17 15:13:12 GMT patch-71
3406 :    
3407 :     Summary:
3408 :     Updated docs.
3409 :     Revision:
3410 :     xvidcore--devapi4--1.0--patch-71
3411 :    
3412 :     The doc of devapi4 is mostly out dated, i t is much better not to
3413 :     keep it in the repository at the moment. We'll add new docs later.
3414 :    
3415 :     Added a INSTALL doc that explains the build/install process for
3416 :     supported platforms. It's a first try, things may be added later.
3417 :    
3418 :     new files:
3419 :     doc/.arch-ids/INSTALL.id doc/INSTALL
3420 :    
3421 :     removed files:
3422 :     doc/.arch-ids/API.dox.id doc/.arch-ids/Makefile.id
3423 :     doc/.arch-ids/foot.inc.in.id doc/.arch-ids/header.tex.in.id
3424 :     doc/.arch-ids/xvid-decoding.txt.id
3425 :     doc/.arch-ids/xvid-encoder.txt.id doc/API.dox doc/Makefile
3426 :     doc/foot.inc.in doc/header.tex.in doc/xvid-decoding.txt
3427 :     doc/xvid-encoder.txt
3428 :    
3429 :     modified files:
3430 :     CodingStyle README doc/README
3431 :    
3432 :     renamed files:
3433 :     .arch-ids/README.txt.id
3434 :     ==> .arch-ids/README.id
3435 :     .arch-ids/changelog.txt.id
3436 :     ==> .arch-ids/ChangeLog.id
3437 :     README.txt
3438 :     ==> README
3439 :     changelog.txt
3440 :     ==> ChangeLog
3441 :    
3442 :    
3443 :     2003-10-15 13:53:11 GMT patch-70
3444 :    
3445 :     Summary:
3446 :     Better cross compilation handling.
3447 :     Revision:
3448 :     xvidcore--devapi4--1.0--patch-70
3449 :    
3450 :     With this patch it is now possible to cross compile xvid
3451 :     quite easily for win32 platform on a build linux host.
3452 :    
3453 :     Recipe for debian system:
3454 :     $ apt-get install mingw32
3455 :     (or create your own cross compiler/binutils suite and install mingw32
3456 :     header files -- sorry i don't have a recipe for this, this is let as
3457 :     an exercice for the reader)
3458 :     $ cd ${xvidcore}
3459 :     $ cd build/generic
3460 :     $ ./bootstrap.sh
3461 :     $ ./configure --host=i586-mingw32msvc
3462 :     (all occurences of i586-mingw32msvc may be replaced with the right
3463 :     prefix you've choosen for your cross compiler and cross binutils)
3464 :     $ make
3465 :     $ cd ../../vfw/bin
3466 :     $ make -f Makefile.cygwin \
3467 :     CC=i586-mingw32msvc-gcc WINDRES=i586-mingw32msvc-windres
3468 :    
3469 :     Enjoy your win32 xvid.dll build by free software, on a free OS, for a devil
3470 :     OS target.
3471 :    
3472 :     modified files:
3473 :     build/generic/configure.in vfw/bin/Makefile.cygwin
3474 :    
3475 :    
3476 :     2003-10-14 15:17:28 GMT patch-69
3477 :    
3478 :     Summary:
3479 :     Fixed Qpel+Interpolation decoding. Cleaned up mb->mode usage.
3480 :     Revision:
3481 :     xvidcore--devapi4--1.0--patch-69
3482 :    
3483 :     * Fixed interpolate mode + qpel decoding.
3484 :     * MB->mb_type completely replaced by MB->mode
3485 :    
3486 :     modified files:
3487 :     src/decoder.c
3488 :    
3489 :    
3490 :     2003-10-12 21:57:24 GMT patch-68
3491 :    
3492 :     Summary:
3493 :     ac/dc prediction for intra RD search.
3494 :     Revision:
3495 :     xvidcore--devapi4--1.0--patch-68
3496 :    
3497 :     From syskin, added real ac/dc prediction for INTRA's bitcount.
3498 :    
3499 :     modified files:
3500 :     src/motion/estimation_rd_based.c
3501 :    
3502 :    
3503 :     2003-10-09 18:15:50 GMT patch-67
3504 :    
3505 :     Summary:
3506 :     Pigrated asm code to new quant API.
3507 :     Revision:
3508 :     xvidcore--devapi4--1.0--patch-67
3509 :    
3510 :     Many changes that are mostly cosmetic in the asm files.
3511 :     * indent
3512 :     * added xor eax, eax in quant_(h263|mpeg)_intra_.* functions
3513 :     (just to make sure the returned value isn't random)
3514 :     * added xor eax, eax in dequant_(h263|mpeg)_.* functions
3515 :     (just to make sure the returned value isn't random)
3516 :     * synced cpuid.asm XVID_CPU_feature constants with the one
3517 :     defined in the C code (xvid.h)
3518 :     * enabled all cpu tests in xvid_bench.c
3519 :    
3520 :     modified files:
3521 :     examples/xvid_bench.c src/quant/quant_h263.c
3522 :     src/quant/quant_mpeg.c
3523 :     src/quant/x86_asm/quantize_h263_3dne.asm
3524 :     src/quant/x86_asm/quantize_h263_mmx.asm
3525 :     src/quant/x86_asm/quantize_mpeg_mmx.asm
3526 :     src/quant/x86_asm/quantize_mpeg_xmm.asm
3527 :     src/utils/x86_asm/cpuid.asm src/xvid.h
3528 :    
3529 :    
3530 :     2003-10-08 21:05:47 GMT patch-66
3531 :    
3532 :     Summary:
3533 :     Updated xvid_bench for quant API changes
3534 :     Revision:
3535 :     xvidcore--devapi4--1.0--patch-66
3536 :    
3537 :     Updated xvid_bench for quant API changes
3538 :    
3539 :    
3540 :     modified files:
3541 :     examples/xvid_bench.c
3542 :    
3543 :    
3544 :     2003-10-07 13:03:51 GMT patch-65
3545 :    
3546 :     Summary:
3547 :     Quant functions API changes (first step)
3548 :     Revision:
3549 :     xvidcore--devapi4--1.0--patch-65
3550 :    
3551 :     In the road to instance safe mpeg quantization, a small cleanup
3552 :     to the quant API was needed. It consists in changing the way we
3553 :     name the functions quant_{mpeg|h263}_{inter|intra}_{arch} and in
3554 :     a move to a more unified API (even intra functions return the sum
3555 :     of coefficients, it can be used as a complexity measure at a later
3556 :     time).
3557 :    
3558 :     This patch touch lot of files, but all changes are trivial.
3559 :    
3560 :     NB: we should check the IA64 asm validity, i changed things but
3561 :     i can't test them.
3562 :    
3563 :     new files:
3564 :     src/quant/.arch-ids/quant.h.id src/quant/quant.h
3565 :    
3566 :     removed files:
3567 :     src/quant/.arch-ids/quant_h263.h.id
3568 :     src/quant/.arch-ids/quant_mpeg4.h.id src/quant/quant_h263.h
3569 :     src/quant/quant_mpeg4.h
3570 :    
3571 :     modified files:
3572 :     build/generic/sources.inc build/win32/libxvidcore.dsp
3573 :     src/decoder.c src/encoder.c src/image/qpel.c
3574 :     src/motion/estimation_rd_based.c
3575 :     src/quant/ia64_asm/quant_h263_ia64.s src/quant/quant_h263.c
3576 :     src/quant/quant_matrix.c src/quant/quant_matrix.h
3577 :     src/quant/quant_mpeg.c
3578 :     src/quant/x86_asm/quantize_h263_3dne.asm
3579 :     src/quant/x86_asm/quantize_h263_mmx.asm
3580 :     src/quant/x86_asm/quantize_mpeg_mmx.asm
3581 :     src/quant/x86_asm/quantize_mpeg_xmm.asm
3582 :     src/utils/mbtransquant.c src/xvid.c src/xvid.h
3583 :    
3584 :     renamed files:
3585 :     src/quant/.arch-ids/quant_mpeg4.c.id
3586 :     ==> src/quant/.arch-ids/quant_mpeg.c.id
3587 :     src/quant/quant_mpeg4.c
3588 :     ==> src/quant/quant_mpeg.c
3589 :     src/quant/x86_asm/.arch-ids/quantize4_mmx.asm.id
3590 :     ==> src/quant/x86_asm/.arch-ids/quantize_mpeg_mmx.asm.id
3591 :     src/quant/x86_asm/.arch-ids/quantize4_xmm.asm.id
3592 :     ==> src/quant/x86_asm/.arch-ids/quantize_mpeg_xmm.asm.id
3593 :     src/quant/x86_asm/.arch-ids/quantize_3dne.asm.id
3594 :     ==> src/quant/x86_asm/.arch-ids/quantize_h263_3dne.asm.id
3595 :     src/quant/x86_asm/.arch-ids/quantize_mmx.asm.id
3596 :     ==> src/quant/x86_asm/.arch-ids/quantize_h263_mmx.asm.id
3597 :     src/quant/x86_asm/quantize4_mmx.asm
3598 :     ==> src/quant/x86_asm/quantize_mpeg_mmx.asm
3599 :     src/quant/x86_asm/quantize4_xmm.asm
3600 :     ==> src/quant/x86_asm/quantize_mpeg_xmm.asm
3601 :     src/quant/x86_asm/quantize_3dne.asm
3602 :     ==> src/quant/x86_asm/quantize_h263_3dne.asm
3603 :     src/quant/x86_asm/quantize_mmx.asm
3604 :     ==> src/quant/x86_asm/quantize_h263_mmx.asm
3605 :    
3606 :    
3607 :     2003-10-05 00:15:15 GMT patch-64
3608 :    
3609 :     Summary:
3610 :     Updated ChangeLog
3611 :     Revision:
3612 :     xvidcore--devapi4--1.0--patch-64
3613 :    
3614 :     Updated ChangeLog
3615 :    
3616 :    
3617 :     modified files:
3618 :     changelog.txt
3619 :    
3620 :    
3621 :     2003-10-04 16:04:30 GMT patch-63
3622 :    
3623 :     Summary:
3624 :     Removed legacy 2pass code from vfw
3625 :     Revision:
3626 :     xvidcore--devapi4--1.0--patch-63
3627 :    
3628 :     Removed legacy 2pass code from vfw
3629 :    
3630 :    
3631 :     removed files:
3632 :     vfw/src/.arch-ids/2pass.c.id vfw/src/.arch-ids/2pass.h.id
3633 :     vfw/src/2pass.c vfw/src/2pass.h
3634 :    
3635 :    
3636 :     2003-10-04 00:41:38 GMT patch-62
3637 :    
3638 :     Summary:
3639 :     Working VFW mingw/cygwin build system.
3640 :     Revision:
3641 :     xvidcore--devapi4--1.0--patch-62
3642 :    
3643 :     This patch fixes the VFW building process. Now it should work out
3644 :     of the box using these steps:
3645 :     # cd ${xvidcore}
3646 :     # cd build/generic
3647 :     # ./bootstrap.sh <-- only needed for CVS checkouts.
3648 :     # ./configure
3649 :     # make
3650 :     # cd ../../vfw/bin
3651 :     # make -f Makefile.cygwin
3652 :    
3653 :     Then install as usual clicking on the inf file or "make install"
3654 :     in the vfw/bin dir.
3655 :    
3656 :     modified files:
3657 :     vfw/bin/Makefile.cygwin vfw/src/config.rc vfw/src/debug.h
3658 :     vfw/src/driverproc.c
3659 :    
3660 :    
3661 :     2003-10-03 17:00:53 GMT patch-61
3662 :    
3663 :     Summary:
3664 :     Fixes for alternate scan and interlacing support.
3665 :     Revision:
3666 :     xvidcore--devapi4--1.0--patch-61
3667 :    
3668 :     Fixes from CVS (by sysKin) for:
3669 :     - added alternate scan support with VHQ
3670 :     - fixed interlacing support in s/b-frames. May fix a potential problem
3671 :     as field_pred struct field seemed not to be initialized anywhere.
3672 :     As it's not supported yet, write an hardcoded 0 bit.
3673 :    
3674 :     Fixes from me for the fixes from syKin:
3675 :     - scan_table effectively used in MBCodingBVOP
3676 :     - Block_CalcBits(Intra) fixes to data->scan_table (implies prototype
3677 :     change and code modification every where the functions were used)
3678 :    
3679 :     I also increased BS version as it might result in different bitstreams
3680 :     It's now at version 20.
3681 :    
3682 :     modified files:
3683 :     src/bitstream/mbcoding.c src/bitstream/mbcoding.h
3684 :     src/decoder.c src/encoder.c src/motion/estimation.h
3685 :     src/motion/estimation_rd_based.c src/xvid.h
3686 :    
3687 :    
3688 :     2003-10-03 15:41:37 GMT patch-60
3689 :    
3690 :     Summary:
3691 :     Removed BIGLUT support.
3692 :     Revision:
3693 :     xvidcore--devapi4--1.0--patch-60
3694 :    
3695 :     Remobed legacy code for BIGLUT support. It was unused and RD based
3696 :     Motion Estimation was not even compatible with this type of VLC coding.
3697 :    
3698 :     modified files:
3699 :     build/generic/configure.in src/bitstream/mbcoding.c
3700 :     src/bitstream/mbcoding.h src/prediction/mbprediction.c
3701 :    
3702 :    
3703 :     2003-10-03 13:25:17 GMT patch-59
3704 :    
3705 :     Summary:
3706 :     Bugfix for PFrames+ Ext Search.
3707 :     Revision:
3708 :     xvidcore--devapi4--1.0--patch-59
3709 :    
3710 :     In Qpel mode, the code was doing a diamond search for wrong predictors.
3711 :     This resulted in poor performance as the diamond search was sitting there
3712 :     for sometime.
3713 :    
3714 :     modified files:
3715 :     src/motion/estimation.h src/motion/estimation_pvop.c
3716 :    
3717 :    
3718 :     2003-10-02 16:50:51 GMT patch-58
3719 :    
3720 :     Summary:
3721 :     Added VFW makefile for cygwin/minsys
3722 :     Revision:
3723 :     xvidcore--devapi4--1.0--patch-58
3724 :    
3725 :     Added VFW makefile for cygwin/minsys. I can't test it so it is probably not
3726 :     right out of the box. Waiting for feedback in order to fix it.
3727 :    
3728 :     new files:
3729 :     vfw/bin/.arch-ids/Makefile.cygwin.id
3730 :     vfw/bin/.arch-ids/Makefile.inc.id vfw/bin/Makefile.cygwin
3731 :     vfw/bin/Makefile.inc vfw/src/w32api/.arch-ids/=id
3732 :     vfw/src/w32api/.arch-ids/vfw.h.id vfw/src/w32api/vfw.h
3733 :    
3734 :     new directories:
3735 :     vfw/src/w32api vfw/src/w32api/.arch-ids
3736 :    
3737 :    
3738 :     2003-10-02 13:35:15 GMT patch-57
3739 :    
3740 :     Summary:
3741 :     Cleaned up the lumimasking code.
3742 :     Revision:
3743 :     xvidcore--devapi4--1.0--patch-57
3744 :    
3745 :     The lumimasking code was not very plugin oriented as it has been ported
3746 :     from old XviD versions. This patch cleans up the code and integrates it
3747 :     better with plugin design. No changes done in teh functionnal code.
3748 :    
3749 :     modified files:
3750 :     src/plugins/plugin_lumimasking.c
3751 :    
3752 :    
3753 :     2003-10-01 23:07:07 GMT patch-56
3754 :    
3755 :     Summary:
3756 :     Cleaned up trailing space chars.
3757 :     Revision:
3758 :     xvidcore--devapi4--1.0--patch-56
3759 :    
3760 :     The kind of patch we would love to avoid as they make merging
3761 :     a nightmare while they're kind of useless patches.
3762 :    
3763 :     Applied sed 's/[ \t]*$//' to all c/h files.
3764 :    
3765 :     modified files:
3766 :     src/bitstream/bitstream.c src/bitstream/bitstream.h
3767 :     src/bitstream/cbp.c src/bitstream/mbcoding.c src/dct/idct.c
3768 :     src/dct/simple_idct.c src/decoder.c src/decoder.h
3769 :     src/encoder.c src/encoder.h src/global.h
3770 :     src/image/colorspace.c src/image/colorspace.h src/image/font.c
3771 :     src/image/image.c src/image/image.h src/image/interpolate8x8.c
3772 :     src/image/interpolate8x8.h src/image/qpel.c src/image/qpel.h
3773 :     src/image/reduced.c src/motion/estimation_rd_based.c
3774 :     src/motion/gmc.c src/motion/gmc.h src/motion/motion.h
3775 :     src/motion/sad.c src/plugins/plugin_2pass1.c
3776 :     src/plugins/plugin_2pass2.c src/plugins/plugin_fixed.c
3777 :     src/plugins/plugin_psnr.c src/plugins/plugin_single.c
3778 :     src/portab.h src/prediction/mbprediction.c
3779 :     src/utils/mbtransquant.c src/utils/timer.c src/xvid.c
3780 :     src/xvid.h
3781 :    
3782 :    
3783 :     2003-09-30 18:10:18 GMT patch-55
3784 :    
3785 :     Summary:
3786 :     Code cleanups.
3787 :     Revision:
3788 :     xvidcore--devapi4--1.0--patch-55
3789 :    
3790 :     It's been a while since the last ISOC89 conformance cleanup.
3791 :    
3792 :     Using the following switches help a lot :-)
3793 :     -Wall -Wsign-compare -Wredundant-decls -Wunreachable-code -Wnested-externs \
3794 :     -ansi
3795 :    
3796 :     Result: 0 warning/0 error
3797 :    
3798 :     modified files:
3799 :     src/bitstream/vlc_codes.h src/global.h src/image/qpel.h
3800 :     src/motion/estimation.h src/motion/estimation_bvop.c
3801 :     src/motion/estimation_gmc.c src/motion/estimation_pvop.c
3802 :     src/motion/estimation_rd_based.c src/motion/gmc.c
3803 :     src/motion/gmc.h src/motion/motion_inlines.h
3804 :     src/motion/vop_type_decision.c
3805 :    
3806 :    
3807 :     2003-09-29 00:31:32 GMT patch-54
3808 :    
3809 :     Summary:
3810 :     Memory leakage fixes.
3811 :     Revision:
3812 :     xvidcore--devapi4--1.0--patch-54
3813 :    
3814 :     The pEnc->queue was allocated but not freed when bframes == 0.
3815 :     And queue images were not freed as well.
3816 :    
3817 :     modified files:
3818 :     examples/Makefile src/encoder.c src/image/image.c
3819 :     src/utils/mem_align.c
3820 :    
3821 :    
3822 :     2003-09-28 16:45:02 GMT patch-53
3823 :    
3824 :     Summary:
3825 :     Fixes the unitialized mcsel bit in RD based ME
3826 :     Revision:
3827 :     xvidcore--devapi4--1.0--patch-53
3828 :    
3829 :     Fixes the unitialized mcsel bit in RD based ME
3830 :    
3831 :    
3832 :     modified files:
3833 :     src/motion/estimation_rd_based.c
3834 :    
3835 :    
3836 :     2003-09-28 01:00:06 GMT patch-52
3837 :    
3838 :     Summary:
3839 :     Fix the XviD constant version initialization
3840 :     Revision:
3841 :     xvidcore--devapi4--1.0--patch-52
3842 :    
3843 :     Fix the XviD constant version initialization
3844 :    
3845 :    
3846 :     modified files:
3847 :     src/xvid.h
3848 :    
3849 :    
3850 :     2003-09-28 00:47:05 GMT patch-51
3851 :    
3852 :     Summary:
3853 :     Fix to the build system (the come back).
3854 :     Revision:
3855 :     xvidcore--devapi4--1.0--patch-51
3856 :    
3857 :     Ok this one fixes the way we build the targets. The VPATH thingy really
3858 :     works with the library targets. My understanding of the VPATH mechanism
3859 :     was wrong. Now it should be ok (i promess).
3860 :    
3861 :     modified files:
3862 :     build/generic/Makefile
3863 :    
3864 :    
3865 :     2003-09-27 11:45:18 GMT patch-50
3866 :    
3867 :     Summary:
3868 :     Small fix to previous patch.
3869 :     Revision:
3870 :     xvidcore--devapi4--1.0--patch-50
3871 :    
3872 :     A pair ofdouble quotes prevented "make" to sort out the VPATH
3873 :     dependences in "=build". This resulted in compiling all the
3874 :     sources, all the time, whether a file changed or not.
3875 :    
3876 :     modified files:
3877 :     build/generic/Makefile
3878 :    
3879 :    
3880 :     2003-09-26 22:39:44 GMT patch-49
3881 :    
3882 :     Summary:
3883 :     Updated the build files for *nix.
3884 :     Revision:
3885 :     xvidcore--devapi4--1.0--patch-49
3886 :    
3887 :     - Fixed MacOSX build (w/o module option). The subversion was not right
3888 :     it was just minor version though it has to be major.minor.
3889 :     - Fixed bootstrap.sh for MacOSX environment, it now looks for
3890 :     glibtoolize if libtoolize is not found.
3891 :     - The unified Makefile now builds XviD out of source tree in the directory
3892 :     =build. It's cleaner, and clashes much less w/ arch/tla source linting.
3893 :     - Tuned the tagging regexps so:
3894 :     + autoconf files are ignored (considered backup => not erased, not copied
3895 :     and not commited/imported)
3896 :     + =build is also considered backup.
3897 :    
3898 :     modified files:
3899 :     build/generic/Makefile build/generic/bootstrap.sh
3900 :     build/generic/configure.in build/generic/platform.inc.in
3901 :     build/generic/sources.inc examples/Makefile
3902 :     {arch}/=tagging-method
3903 :    
3904 :    
3905 :     2003-09-24 01:38:03 GMT patch-48
3906 :    
3907 :     Summary:
3908 :     Bug fix to decoder (mcsel/acpred bits swapped)
3909 :     Revision:
3910 :     xvidcore--devapi4--1.0--patch-48
3911 :    
3912 :     As reported here:
3913 :     http://www.xvid.org/modules.php?op=modload&name=phpBB2&file=viewtopic&t=1513&highlight=
3914 :    
3915 :     <quote>
3916 :     in the spec, 'there is 'mcsel' before 'ac_pred_flag'.
3917 :     however, if you see the code, it's changed.
3918 :     </quote>
3919 :    
3920 :     We were doing the opposite, ac_pred before mcsel.
3921 :    
3922 :     modified files:
3923 :     src/decoder.c
3924 :    
3925 :    
3926 :     2003-09-11 17:19:35 GMT patch-47
3927 :    
3928 :     Summary:
3929 :     Small fix to GMC+QuarterPel -- BS version bumped to 19
3930 :     Revision:
3931 :     xvidcore--devapi4--1.0--patch-47
3932 :    
3933 :     During the split up, a line has been disabled that prevented good
3934 :     quartelpel+GMC. Fixed.
3935 :    
3936 :     This fix + patch-43 are a good reason to bump up the bitstream
3937 :     version to 19.
3938 :    
3939 :     modified files:
3940 :     src/motion/gmc.h src/xvid.h
3941 :    
3942 :    
3943 :     2003-09-11 17:11:28 GMT patch-46
3944 :    
3945 :     Summary:
3946 :     Build process fix for MacOSX+module option
3947 :     Revision:
3948 :     xvidcore--devapi4--1.0--patch-46
3949 :    
3950 :     This patch fixes the install rule of the MacOSX module style library.
3951 :     It adds a PRE_SHARED_LIB == SHARED_LIB for all platforms except
3952 :     MacOSX that uses it a different way, and exploits the fact
3953 :     $(SPECIFIC_LDFLAGS) is the last var of the build line to insert a
3954 :     command for post linking the right .so file.
3955 :    
3956 :     modified files:
3957 :     build/generic/Makefile build/generic/configure.in
3958 :     build/generic/platform.inc.in
3959 :    
3960 :    
3961 :     2003-09-11 13:56:40 GMT patch-45
3962 :    
3963 :     Summary:
3964 :     Tree cleanup and build files updated.
3965 :     Revision:
3966 :     xvidcore--devapi4--1.0--patch-45
3967 :    
3968 :     * Win32 files switched to Unix format.
3969 :     Ok, on IRC, we sorted out what was b0rking the project files each
3970 :     time I commit them in DOS format. In fact the Unix CVS does upload
3971 :     DOS format files if i commit them, and then Win32 CVS users get
3972 :     double \r\n files. So the best is to use unix format in CVS. This
3973 :     way, the checked out versions are right for Win32 users.
3974 :    
3975 :     NB: this imply a unix2dos conversion when doing a release from a
3976 :     Unix box. I think it's better than current policy, but automatic
3977 :     tarballs/zip of the tree should take care of that.
3978 :    
3979 :     * odivx and xvid_stat examples removed. Associated project files have
3980 :     been removed as well.
3981 :    
3982 :     * The divx4 compatibility layer has been removed. Associated options
3983 :     in the configure script removed. libxvidcore.def has no more need
3984 :     to be an autoconf generated file.
3985 :    
3986 :     * rawdec removed. It has never been used and will never be.
3987 :    
3988 :     * Added major api appending to the macosx module build.
3989 :    
3990 :     removed files:
3991 :     build/win32/.arch-ids/odivx_enc_dec.dsp.id
3992 :     build/win32/.arch-ids/xvid_stat.dsp.id
3993 :     build/win32/odivx_enc_dec.dsp build/win32/xvid_stat.dsp
3994 :     examples/.arch-ids/odivx_enc_dec.c.id
3995 :     examples/.arch-ids/xvid_stat.c.id examples/odivx_enc_dec.c
3996 :     examples/xvid_stat.c rawdec/.arch-ids/=id
3997 :     rawdec/.arch-ids/rawdec.c.id rawdec/.arch-ids/rawdec.dsp.id
3998 :     rawdec/rawdec.c rawdec/rawdec.dsp src/.arch-ids/divx4.c.id
3999 :     src/.arch-ids/divx4.h.id src/divx4.c src/divx4.h
4000 :    
4001 :     modified files:
4002 :     build/generic/Makefile build/generic/configure.in
4003 :     build/generic/libxvidcore.def build/win32/libxvidcore.dsp
4004 :     build/win32/xvid_decraw.dsp build/win32/xvid_encraw.dsp
4005 :     build/win32/xvidcore.dsw
4006 :    
4007 :     renamed files:
4008 :     build/generic/.arch-ids/libxvidcore.def.in.id
4009 :     ==> build/generic/.arch-ids/libxvidcore.def.id
4010 :     build/generic/libxvidcore.def.in
4011 :     ==> build/generic/libxvidcore.def
4012 :    
4013 :     removed directories:
4014 :     rawdec rawdec/.arch-ids
4015 :    
4016 :    
4017 :     2003-09-11 12:59:19 GMT patch-44
4018 :    
4019 :     Summary:
4020 :     Replaced malloc.h header file with stdlib.h
4021 :     Revision:
4022 :     xvidcore--devapi4--1.0--patch-44
4023 :    
4024 :     Replaced malloc.h header file with stdlib.h
4025 :    
4026 :    
4027 :     modified files:
4028 :     src/plugins/plugin_lumimasking.c
4029 :    
4030 :    
4031 :     2003-09-10 22:33:04 GMT patch-43
4032 :    
4033 :     Summary:
4034 :     Fixed a problem for 'power of 2' framerates
4035 :     Revision:
4036 :     xvidcore--devapi4--1.0--patch-43
4037 :    
4038 :     Fixed a problem for 'power of 2' framerates
4039 :    
4040 :    
4041 :     modified files:
4042 :     src/bitstream/bitstream.c
4043 :    
4044 :    
4045 :     2003-09-10 21:57:12 GMT patch-42
4046 :    
4047 :     Summary:
4048 :     Motion Estimation module splitting.
4049 :     Revision:
4050 :     xvidcore--devapi4--1.0--patch-42
4051 :    
4052 :     The motion estimation module was the biggest file of the source
4053 :     tree. After some previous attempts, sysKin decided to split it
4054 :     up again. This time he's done it right.
4055 :    
4056 :     This split up just changes the organization of ME functions
4057 :     inside different files. This should help a bit in keeping the
4058 :     motion estimation being manageable by a normal human ;-)
4059 :    
4060 :     Here is the splitting logic quoted from sysKin's email to
4061 :     xvid-devel:
4062 :     - estimation.h: header files #included in all ME modules:
4063 :     + deftypes,
4064 :     + macros
4065 :     + constants
4066 :     NB: no code.
4067 :     - estimation_bvop.c: motion estimation for b-vops.
4068 :     everything in it :)
4069 :     - estimation_common.c: functions shared among all ME
4070 :     modules:
4071 :     + diamonds
4072 :     + subpel
4073 :     + refinement
4074 :     + picture
4075 :     + manipulation
4076 :     + tables
4077 :     + ... etc.
4078 :     - estimation_gmc.c: gruel's GME code
4079 :     - estimation_pvop.c: ME for p-vops. Also SAD-based mode
4080 :     decision
4081 :     - estimation_rd_based.c: everything R-D-based: mode
4082 :     decison (including _Fast) and ME.
4083 :     - gmc.c, gmc.h: no change.
4084 :    
4085 :     new files:
4086 :     src/motion/.arch-ids/estimation.h.id
4087 :     src/motion/.arch-ids/estimation_bvop.c.id
4088 :     src/motion/.arch-ids/estimation_common.c.id
4089 :     src/motion/.arch-ids/estimation_gmc.c.id
4090 :     src/motion/.arch-ids/estimation_pvop.c.id
4091 :     src/motion/.arch-ids/estimation_rd_based.c.id
4092 :     src/motion/.arch-ids/motion_inlines.h.id
4093 :     src/motion/.arch-ids/vop_type_decision.c.id
4094 :     src/motion/estimation.h src/motion/estimation_bvop.c
4095 :     src/motion/estimation_common.c src/motion/estimation_gmc.c
4096 :     src/motion/estimation_pvop.c src/motion/estimation_rd_based.c
4097 :     src/motion/motion_inlines.h src/motion/vop_type_decision.c
4098 :    
4099 :     removed files:
4100 :     src/motion/.arch-ids/motion_est.c.id
4101 :     src/motion/.arch-ids/motion_est.h.id
4102 :     src/motion/.arch-ids/smp_motion_est.c.id
4103 :     src/motion/.arch-ids/smp_motion_est.h.id
4104 :     src/motion/motion_est.c src/motion/motion_est.h
4105 :     src/motion/smp_motion_est.c src/motion/smp_motion_est.h
4106 :    
4107 :     modified files:
4108 :     build/generic/sources.inc build/win32/libxvidcore.dsp
4109 :     src/bitstream/mbcoding.h src/motion/gmc.c src/motion/gmc.h
4110 :     src/motion/motion.h src/motion/motion_comp.c src/motion/sad.c
4111 :     src/motion/sad.h src/prediction/mbprediction.c
4112 :     src/prediction/mbprediction.h src/utils/mbfunctions.h
4113 :    
4114 :    
4115 :     2003-09-10 00:40:44 GMT patch-41
4116 :    
4117 :     Summary:
4118 :     Cleanups and fix to (trellis+thresholding) logic
4119 :     Revision:
4120 :     xvidcore--devapi4--1.0--patch-41
4121 :    
4122 :     Cleanups to some functions (loop unrolling, call to functions
4123 :     through function array pointers)...
4124 :    
4125 :     Fix to the trellis+thresholding logic. It was comparing the
4126 :     return value of trellis with a threshold but the trellis function
4127 :     returns the last non zero coeff index... this was basically
4128 :     comparing apples with oranges... funny but wrong. Trellis now
4129 :     returns the sum of absolute coeffs, so the comparison is logical.
4130 :     Btw, as discussed on the devel ML, this is probably uneeded as
4131 :     trellis does an RD optimized coeff distribution.
4132 :    
4133 :     modified files:
4134 :     src/utils/mbtransquant.c
4135 :    
4136 :    
4137 :     2003-09-09 13:13:58 GMT patch-40
4138 :    
4139 :     Summary:
4140 :     Missing ressource for dshow frontend
4141 :     Revision:
4142 :     xvidcore--devapi4--1.0--patch-40
4143 :    
4144 :     Missing ressource for dshow frontend
4145 :    
4146 :    
4147 :     new files:
4148 :     dshow/src/.arch-ids/XviD_logo.bmp.id dshow/src/XviD_logo.bmp
4149 :    
4150 :    
4151 :     2003-09-08 11:02:10 GMT patch-39
4152 :    
4153 :     Summary:
4154 :     Small fixes for fast mode decision
4155 :     Revision:
4156 :     xvidcore--devapi4--1.0--patch-39
4157 :    
4158 :     Small fixes for fast mode decision
4159 :    
4160 :    
4161 :     modified files:
4162 :     src/motion/motion_est.c src/xvid.h
4163 :    
4164 :    
4165 :     2003-09-05 23:45:48 GMT patch-38
4166 :    
4167 :     Summary:
4168 :     New RD mode decision and subpel refinement.
4169 :     Revision:
4170 :     xvidcore--devapi4--1.0--patch-38
4171 :    
4172 :     New stuff from michael. It deals with mode decision and subpel
4173 :     refinement. Integration of these new flags are not settled.
4174 :    
4175 :     Wait and see. Further testing is needed.
4176 :    
4177 :     modified files:
4178 :     src/motion/motion_est.c src/motion/motion_est.h src/xvid.h
4179 :    
4180 :    
4181 :     2003-08-29 13:56:30 GMT patch-37
4182 :    
4183 :     Summary:
4184 :     Still more ME tuning
4185 :     Revision:
4186 :     xvidcore--devapi4--1.0--patch-37
4187 :    
4188 :     Still more ME tuning
4189 :    
4190 :    
4191 :     modified files:
4192 :     src/motion/motion_est.c
4193 :    
4194 :    
4195 :     2003-08-28 12:43:22 GMT patch-36
4196 :    
4197 :     Summary:
4198 :     Removed expanded the cvs Id field
4199 :     Revision:
4200 :     xvidcore--devapi4--1.0--patch-36
4201 :    
4202 :     Removed expanded the cvs Id field
4203 :    
4204 :    
4205 :     modified files:
4206 :     src/image/x86_asm/qpel_mmx.asm
4207 :    
4208 :    
4209 :     2003-08-28 12:39:44 GMT patch-35
4210 :    
4211 :     Summary:
4212 :     More motion est cleanup and bugfixes.
4213 :     Revision:
4214 :     xvidcore--devapi4--1.0--patch-35
4215 :    
4216 :     Still more bugfixes, cleanups and improvements to the Motion Est
4217 :     by sysKin
4218 :    
4219 :     modified files:
4220 :     src/motion/motion_est.c src/motion/motion_est.h
4221 :    
4222 :    
4223 :     2003-08-26 13:57:39 GMT patch-34
4224 :    
4225 :     Summary:
4226 :     Added final bits of Aspect Ratio flag.
4227 :     Revision:
4228 :     xvidcore--devapi4--1.0--patch-34
4229 :    
4230 :     Peter did think of the AR flag since the very beginning of devapi4.
4231 :     He just forgot to code the final bits to effectively write it to
4232 :     the bitstream. This patch adds these missing final bits so XviD now
4233 :     reads and writes AR flags.
4234 :    
4235 :     modified files:
4236 :     ./src/bitstream/bitstream.c ./src/encoder.c ./src/encoder.h
4237 :     ./src/xvid.h
4238 :    
4239 :    
4240 :     2003-08-25 16:41:09 GMT patch-33
4241 :    
4242 :     Summary:
4243 :     Small motion estimation cleanup.
4244 :     Revision:
4245 :     xvidcore--devapi4--1.0--patch-33
4246 :    
4247 :     Cleanups from sysKin.
4248 :    
4249 :     modified files:
4250 :     ./src/motion/motion_est.c ./src/motion/motion_est.h
4251 :    
4252 :    
4253 :     2003-08-25 14:59:28 GMT patch-32
4254 :    
4255 :     Summary:
4256 :     Frame padding bug.
4257 :     Revision:
4258 :     xvidcore--devapi4--1.0--patch-32
4259 :    
4260 :     We were edging the image repeating pixels from the image directly,
4261 :     but the standard says we must repeat from a 16 pixel boundary.
4262 :    
4263 :     See Chapter 7.6.4 of the standard.
4264 :    
4265 :     modified files:
4266 :     ./src/image/image.c
4267 :    
4268 :    
4269 :     2003-08-23 15:07:44 GMT patch-31
4270 :    
4271 :     Summary:
4272 :     New Qpel code.
4273 :     Revision:
4274 :     xvidcore--devapi4--1.0--patch-31
4275 :    
4276 :     Isibaar commited a new piece of QPel code that seems to be optimized
4277 :     for ia32(mmx) architectures. I had to clean it up a bit to make it
4278 :     respectful of architecture separations. This code is disabled for
4279 :     non ia32 arch, a comment mention it's only faster on ia32... i wonder
4280 :     if it's true, some tests have to be done on sourceforge compile farm
4281 :     in order to confirm that.
4282 :    
4283 :     Compared to the bared CVS commit, this patch includes:
4284 :     - fixes the unix build.
4285 :     - better architecture separation.
4286 :     - CodingStyle respected.
4287 :    
4288 :     new files:
4289 :     ./src/image/.arch-ids/qpel.c.id
4290 :     ./src/image/.arch-ids/qpel.h.id ./src/image/qpel.c
4291 :     ./src/image/qpel.h
4292 :     ./src/image/x86_asm/.arch-ids/qpel_mmx.asm.id
4293 :     ./src/image/x86_asm/qpel_mmx.asm
4294 :    
4295 :     modified files:
4296 :     ./build/generic/sources.inc ./build/win32/libxvidcore.dsp
4297 :     ./src/encoder.c ./src/motion/motion_comp.c ./src/xvid.c
4298 :    
4299 :    
4300 :     2003-08-22 13:20:36 GMT patch-30
4301 :    
4302 :     Summary:
4303 :     sad32v does really what it's expected (ie 32x32 SAD :-)
4304 :     Revision:
4305 :     xvidcore--devapi4--1.0--patch-30
4306 :    
4307 :     sad32v does really what it's expected (ie 32x32 SAD :-)
4308 :    
4309 :    
4310 :     modified files:
4311 :     ./src/motion/sad.c
4312 :    
4313 :    
4314 :     2003-08-18 19:00:49 GMT patch-29
4315 :    
4316 :     Summary:
4317 :     64bit fix.
4318 :     Revision:
4319 :     xvidcore--devapi4--1.0--patch-29
4320 :    
4321 :     The interpolation code was unsafe on 64bit platforms, the offset was
4322 :     badly sized, resulting in segfaults.
4323 :    
4324 :     modified files:
4325 :     ./src/image/interpolate8x8.h
4326 :    
4327 :    
4328 :     2003-08-17 14:08:48 GMT patch-28
4329 :    
4330 :     Summary:
4331 :     Greyscale mode fixes.
4332 :     Revision:
4333 :     xvidcore--devapi4--1.0--patch-28
4334 :    
4335 :     We were missing some greyscale tests in the encoder loop... noticeably
4336 :     in the PVOP function when coding an intra block, and in BVOP function
4337 :     when coding all types.
4338 :    
4339 :     I added the cbp trick in the cases discussed above.
4340 :    
4341 :     modified files:
4342 :     ./src/encoder.c
4343 :    
4344 :    
4345 :     2003-08-13 11:47:33 GMT patch-27
4346 :    
4347 :     Summary:
4348 :     Forgotten bit for IA64 separation
4349 :     Revision:
4350 :     xvidcore--devapi4--1.0--patch-27
4351 :    
4352 :     Forgotten bit for IA64 separation
4353 :    
4354 :    
4355 :     modified files:
4356 :     ./src/image/interpolate8x8.h
4357 :    
4358 :    
4359 :     2003-08-11 15:42:30 GMT patch-26
4360 :    
4361 :     Summary:
4362 :     Some qpel changes (sync with Isibaar)
4363 :     Revision:
4364 :     xvidcore--devapi4--1.0--patch-26
4365 :    
4366 :     Some qpel changes (sync with Isibaar)
4367 :    
4368 :    
4369 :     modified files:
4370 :     ./examples/xvid_encraw.c
4371 :    
4372 :    
4373 :     2003-08-11 15:30:04 GMT patch-25
4374 :    
4375 :     Summary:
4376 :     Better architecture separation.
4377 :     Revision:
4378 :     xvidcore--devapi4--1.0--patch-25
4379 :    
4380 :     Architecture depending functions were declared whatever arch you
4381 :     were compiling. This patch fixes that.
4382 :    
4383 :     I also removed the simple_idct hack in decoder.c as it was simply
4384 :     not used. Better not have ugly code in there.
4385 :    
4386 :     xvid_bench, should now compile and run on all archs. However I did
4387 :     not put the cpu definitions for each arch, i just separated
4388 :     ARCH_IS_IA32 so even ARCH_IS_GENERIC can compile
4389 :    
4390 :     modified files:
4391 :     ./examples/xvid_bench.c ./src/bitstream/cbp.h ./src/dct/fdct.h
4392 :     ./src/dct/idct.h ./src/decoder.c ./src/image/colorspace.h
4393 :     ./src/image/interpolate8x8.h ./src/image/reduced.h
4394 :     ./src/motion/sad.h ./src/quant/quant_h263.h
4395 :     ./src/quant/quant_mpeg4.h ./src/utils/emms.h
4396 :    
4397 :    
4398 :     2003-08-09 20:47:42 GMT patch-24
4399 :    
4400 :     Summary:
4401 :     Updated changelog
4402 :     Revision:
4403 :     xvidcore--devapi4--1.0--patch-24
4404 :    
4405 :     Updated changelog
4406 :    
4407 :    
4408 :     modified files:
4409 :     ./changelog.txt
4410 :    
4411 :    
4412 :     2003-08-09 20:31:17 GMT patch-23
4413 :    
4414 :     Summary:
4415 :     Workaround to a GMC bug due to a MS compiler bug.
4416 :     Revision:
4417 :     xvidcore--devapi4--1.0--patch-23
4418 :    
4419 :     Christoph did use a trick to speed up code that resulted in badly
4420 :     optimized code (teh compiler was missing a cast)
4421 :    
4422 :     modified files:
4423 :     ./src/motion/gmc.c
4424 :    
4425 :    
4426 :     2003-08-09 17:09:00 GMT patch-22
4427 :    
4428 :     Summary:
4429 :     Fixes to xvid_decraw
4430 :     Revision:
4431 :     xvidcore--devapi4--1.0--patch-22
4432 :    
4433 :     xvid_decraw has always been used on not so high bitrate sequences
4434 :     and not so big sequences neither.
4435 :    
4436 :     I've been doing lot of tests on the Matrix 2 trailer
4437 :     (1000x540 25fps ~5Mbits/s) and xvid_decraw was not able to handle that
4438 :     because of bugs in the buffer filling algorithm.
4439 :    
4440 :     This patch fixes the buffer filling and catches up with christoph changes
4441 :     in CVS (wrong help message and option parsing).
4442 :    
4443 :     modified files:
4444 :     ./examples/xvid_decraw.c
4445 :    
4446 :    
4447 :     2003-08-08 21:31:59 GMT patch-21
4448 :    
4449 :     Summary:
4450 :     Added QPel and GMC options.
4451 :     Revision:
4452 :     xvidcore--devapi4--1.0--patch-21
4453 :    
4454 :     XviD has so many options that we forget to propose them all on the
4455 :     CLI, here are two more: GMC and Qpel.
4456 :    
4457 :     modified files:
4458 :     ./examples/xvid_encraw.c
4459 :    
4460 :    
4461 :     2003-08-07 19:26:28 GMT patch-20
4462 :    
4463 :     Summary:
4464 :     SVOP handling in status window
4465 :     Revision:
4466 :     xvidcore--devapi4--1.0--patch-20
4467 :    
4468 :     SVOP handling in status window
4469 :    
4470 :    
4471 :     modified files:
4472 :     ./vfw/src/status.c
4473 :    
4474 :    
4475 :     2003-08-07 19:25:03 GMT patch-19
4476 :    
4477 :     Summary:
4478 :     Warning cleanups by chl
4479 :     Revision:
4480 :     xvidcore--devapi4--1.0--patch-19
4481 :    
4482 :     Warning cleanups by chl
4483 :    
4484 :    
4485 :     modified files:
4486 :     ./src/encoder.c ./src/encoder.h ./src/image/interpolate8x8.c
4487 :     ./src/motion/motion_est.c ./src/xvid.h
4488 :    
4489 :    
4490 :     2003-08-06 21:13:35 GMT patch-18
4491 :    
4492 :     Summary:
4493 :     Fix to GMC sprite trajectory code
4494 :     Revision:
4495 :     xvidcore--devapi4--1.0--patch-18
4496 :    
4497 :     Fix to GMC sprite trajectory code
4498 :    
4499 :    
4500 :     modified files:
4501 :     ./src/bitstream/mbcoding.c ./src/xvid.h
4502 :    
4503 :    
4504 :     2003-08-06 10:57:25 GMT patch-17
4505 :    
4506 :     Summary:
4507 :     Fixes a bug in BVOP block skipping thresholding
4508 :     Revision:
4509 :     xvidcore--devapi4--1.0--patch-17
4510 :    
4511 :     Fixes a bug in BVOP block skipping thresholding
4512 :    
4513 :    
4514 :     modified files:
4515 :     ./src/motion/motion_est.c
4516 :    
4517 :    
4518 :     2003-08-03 14:57:32 GMT patch-16
4519 :    
4520 :     Summary:
4521 :     Functions renaming + motion fixes.
4522 :     Revision:
4523 :     xvidcore--devapi4--1.0--patch-16
4524 :    
4525 :     BITS flags have been renamed to RD (Rate Distorsion) flags... however
4526 :     function names were still xxxBitsxxx.
4527 :    
4528 :     Improved frame type decision
4529 :    
4530 :     Fix for DQUANTS plugins, their quant was never checked against valid
4531 :     [1..31] range.
4532 :    
4533 :     modified files:
4534 :     ./src/encoder.c ./src/motion/motion_est.c
4535 :     ./src/motion/motion_est.h
4536 :    
4537 :    
4538 :     2003-08-02 15:00:49 GMT patch-15
4539 :    
4540 :     Summary:
4541 :     API cleanup.
4542 :     Revision:
4543 :     xvidcore--devapi4--1.0--patch-15
4544 :    
4545 :     Since we started devapi3 and then devapi4, feature names did not change
4546 :     because it was just convenient to keep them to minimize the change impact.
4547 :     But most of the flags were now not even suggesting what they do. So this
4548 :     patch cleans the API.
4549 :    
4550 :     This patch also change the way we describe flags, it's more compact and
4551 :     shows better flags are bit sets that must not overlap. This change fixes
4552 :     a plugin flag overlapping problem as well.
4553 :    
4554 :     modified files:
4555 :     ./examples/xvid_encraw.c ./src/decoder.c ./src/encoder.c
4556 :     ./src/encoder.h ./src/motion/motion_est.c
4557 :     ./src/motion/motion_est.h ./src/xvid.c ./src/xvid.h
4558 :     ./vfw/src/codec.c ./vfw/src/config.c
4559 :    
4560 :    
4561 :     2003-07-29 22:25:12 GMT patch-14
4562 :    
4563 :     Summary:
4564 :     Fixed bogus memory accesses
4565 :     Revision:
4566 :     xvidcore--devapi4--1.0--patch-14
4567 :    
4568 :     Fixed bogus memory accesses
4569 :    
4570 :    
4571 :     modified files:
4572 :     ./src/encoder.c ./src/plugins/plugin_2pass1.c
4573 :    
4574 :    
4575 :     2003-07-28 12:22:33 GMT patch-13
4576 :    
4577 :     Summary:
4578 :     Bitstream version increased to 16
4579 :     Revision:
4580 :     xvidcore--devapi4--1.0--patch-13
4581 :    
4582 :     Bitstream version increased to 16
4583 :    
4584 :    
4585 :     modified files:
4586 :     ./src/xvid.h
4587 :    
4588 :    
4589 :     2003-07-25 12:01:51 GMT patch-12
4590 :    
4591 :     Summary:
4592 :     Added gmc files to teh windows project file
4593 :     Revision:
4594 :     xvidcore--devapi4--1.0--patch-12
4595 :    
4596 :     Added gmc files to teh windows project file
4597 :    
4598 :    
4599 :     modified files:
4600 :     ./build/win32/libxvidcore.dsp ./vfw/src/codec.c
4601 :    
4602 :    
4603 :     2003-07-25 12:00:31 GMT patch-11
4604 :    
4605 :     Summary:
4606 :     Added cartoon mode from Isibaar
4607 :     Revision:
4608 :     xvidcore--devapi4--1.0--patch-11
4609 :    
4610 :     Added cartoon mode from Isibaar
4611 :    
4612 :    
4613 :     modified files:
4614 :     ./src/motion/motion_est.c ./src/plugins/plugin_single.c
4615 :     ./src/utils/mbtransquant.c ./src/xvid.h
4616 :    
4617 :    
4618 :     2003-07-25 10:30:41 GMT patch-10
4619 :    
4620 :     Summary:
4621 :     Bitstream syntax comments.
4622 :     Revision:
4623 :     xvidcore--devapi4--1.0--patch-10
4624 :    
4625 :     This patch does not change the bitstream but adds some comments that
4626 :     can help in order to understand (lack of) calls to BitstreamPadAlways.
4627 :    
4628 :     modified files:
4629 :     ./src/bitstream/bitstream.c ./src/encoder.c
4630 :    
4631 :    
4632 :     2003-07-22 16:34:25 GMT patch-9
4633 :    
4634 :     Summary:
4635 :     Fixes Bistream errors in VOL (+ forced stuffing)
4636 :     Revision:
4637 :     xvidcore--devapi4--1.0--patch-9
4638 :    
4639 :     After a detailed bugreport at:
4640 :     http://www.xvid.org/modules.php?op=modload&name=phpBB2&file=viewtopic&t=1387&highlight=
4641 :    
4642 :     I discovered that:
4643 :    
4644 :     1/ we did not write video_signal_type, but we were padding to the next
4645 :     byte, that's why we had video_signal_type=0 and then only 1s until the
4646 :     next byte boundary. This explains the 11 next_start_code();
4647 :    
4648 :     2/ video_object_type_indication = Reserved is right on my machine,
4649 :     please check again, but i doubt there is a bug there, we use 3 hard
4650 :     wired values and none of them is zero.
4651 :    
4652 :     3/ 01 : next_start_code() *** Was wrong in 24.02.2003; is correct in
4653 :     dev-api-4!!! *** was a bug in fact... when we write user data, we pad to
4654 :     the next byte boundary (if needed) like the standard says... by
4655 :     chancepadding was almost always done, thus the next_start_code() was
4656 :     respected.
4657 :    
4658 :     4/ The extra stuffing bits were caused by a forced padding between our
4659 :     VOL function writer and VOP header function writer.
4660 :    
4661 :     modified files:
4662 :     ./src/bitstream/bitstream.c ./src/encoder.c
4663 :    
4664 :    
4665 :     2003-07-16 22:57:44 GMT patch-8
4666 :    
4667 :     Summary:
4668 :     Fixed quant4_intra_xmm and quant_intra_3dne bug for DC<0.
4669 :     Revision:
4670 :     xvidcore--devapi4--1.0--patch-8
4671 :    
4672 :     These two functions were suffering the same error that consists in
4673 :     emulating idiv with some an inversed divisor array and an imul
4674 :     instruction followed by a right shift...
4675 :    
4676 :     That was always decreasing the ressult by 1 for negative DC values.
4677 :     A not so bad solution is simply to use a cmov instruction and choose
4678 :     the right value according to the DC value. As these function were
4679 :     for PIII and Athlon, we are sure we can use the cmov instruction.
4680 :    
4681 :     PS: the fix is somewhere in cosmetic changes... sorry but the code was
4682 :     too ugly to fix it like it was.
4683 :    
4684 :     modified files:
4685 :     ./src/quant/x86_asm/quantize4_xmm.asm
4686 :     ./src/quant/x86_asm/quantize_3dne.asm
4687 :    
4688 :    
4689 :     2003-07-16 12:58:21 GMT patch-7
4690 :    
4691 :     Summary:
4692 :     Fixed the build system (error caused by patch-5)
4693 :     Revision:
4694 :     xvidcore--devapi4--1.0--patch-7
4695 :    
4696 :     IA64 cahnges were wrong in the build system, they make all platforms
4697 :     try to compile a directory... That patch should fix the IA64 target
4698 :     build and get back to previous behavior for other architectures.
4699 :    
4700 :     modified files:
4701 :     ./build/generic/configure.in ./build/generic/platform.inc.in
4702 :     ./build/generic/sources.inc
4703 :    
4704 :    
4705 :     2003-07-13 12:16:55 GMT patch-6
4706 :    
4707 :     Summary:
4708 :     Updates for GME and some cleanups.
4709 :     Revision:
4710 :     xvidcore--devapi4--1.0--patch-6
4711 :    
4712 :     This is a all in one patch from syskin:
4713 :     * mcsel decision moved to ModeDecision() function. That makes motion
4714 :     loop completely aware of macroblock mode and vectors (amv in that case).
4715 :     A simple copy&paste was needed to make the the decision R-D based, to be
4716 :     compatible with MODEDECISION_BITS.
4717 :    
4718 :     * many bugs fixed. Most of them very small, the only big one was that
4719 :     BITS was misunderstanding a flag and was thinking that mpeg quant is
4720 :     used when h263 quant is used and vice versa :( Also, correct cbp with
4721 :     inter4v mode makes mode decision better. Two speedups - for BITS (no
4722 :     more dequantization when sum == 0) and for ChromaME (chroma sad not
4723 :     computed if total sad too big before that). Some GMC compiler warnings
4724 :     removed. Probably more, I don't remember ;) I haven't touched P/B/I
4725 :     decision for once.
4726 :    
4727 :     * compiler warnings removed, mostly "const mismatch" in get_amv() <-- or
4728 :     what was his name.
4729 :    
4730 :     * two functions made 2x smaller, shorter and faster.
4731 :    
4732 :     modified files:
4733 :     ./src/encoder.c ./src/global.h ./src/image/interpolate8x8.h
4734 :     ./src/motion/gmc.c ./src/motion/gmc.h
4735 :     ./src/motion/motion_est.c ./src/motion/motion_est.h
4736 :     ./vfw/src/codec.c
4737 :    
4738 :    
4739 :     2003-07-10 17:35:59 GMT patch-5
4740 :    
4741 :     Summary:
4742 :     IA64 updates.
4743 :     Revision:
4744 :     xvidcore--devapi4--1.0--patch-5
4745 :    
4746 :     Changes from Stephan Krause <s_kraste at ira.uka.de>
4747 :    
4748 :     Small updates so ia64 is supposed to work. Further testing is needed
4749 :     because tests have only been done with xvid_encraw.
4750 :    
4751 :     modified files:
4752 :     ./build/generic/platform.inc.in ./examples/xvid_encraw.c
4753 :     ./src/motion/motion_est.c ./src/xvid.c
4754 :    
4755 :    
4756 :     2003-07-10 17:27:01 GMT patch-4
4757 :    
4758 :     Summary:
4759 :     Removed remaining expanded $ lines from the arch repo
4760 :     Revision:
4761 :     xvidcore--devapi4--1.0--patch-4
4762 :    
4763 :     Removed remaining expanded $ lines from the arch repo
4764 :    
4765 :    
4766 :     modified files: