6 |
# |
# |
7 |
|
|
8 |
|
|
9 |
|
2003-11-29 17:21:08 GMT patch-111 |
10 |
|
|
11 |
|
Summary: |
12 |
|
First beta marking |
13 |
|
Revision: |
14 |
|
xvidcore--devapi4--1.0--patch-111 |
15 |
|
|
16 |
|
First beta marking |
17 |
|
|
18 |
|
|
19 |
|
modified files: |
20 |
|
build/generic/configure.in src/xvid.c src/xvid.h |
21 |
|
|
22 |
|
|
23 |
|
2003-11-29 16:59:14 GMT patch-110 |
24 |
|
|
25 |
|
Summary: |
26 |
|
Catching up with CVS. |
27 |
|
Revision: |
28 |
|
xvidcore--devapi4--1.0--patch-110 |
29 |
|
|
30 |
|
Changes from sysKin: |
31 |
|
* dquant optimization. |
32 |
|
* CBR fix |
33 |
|
|
34 |
|
modified files: |
35 |
|
src/encoder.c src/plugins/plugin_single.c |
36 |
|
|
37 |
|
|
38 |
|
2003-11-24 22:05:38 GMT patch-109 |
39 |
|
|
40 |
|
Summary: |
41 |
|
Big level handling in trellis. |
42 |
|
Revision: |
43 |
|
xvidcore--devapi4--1.0--patch-109 |
44 |
|
|
45 |
|
Trellis was treating big levels exactly the same way as lower ones. |
46 |
|
In some cases, trellis was doing wild optimizations favoring a 0 |
47 |
|
because the distortion introduced by that big coeff change was acceptable. |
48 |
|
But visually this could result in some nasty blocks with wrong chroma |
49 |
|
information or similar brutal changes in other planes as well. |
50 |
|
|
51 |
|
Skal added big levels handling where trellis just tries to minimize |
52 |
|
the cost varying the run value only. No level modification is done |
53 |
|
anymore. |
54 |
|
|
55 |
|
modified files: |
56 |
|
TODO src/utils/mbtransquant.c |
57 |
|
|
58 |
|
|
59 |
|
2003-11-23 16:42:55 GMT patch-108 |
60 |
|
|
61 |
|
Summary: |
62 |
|
Trellis for MPEG. |
63 |
|
Revision: |
64 |
|
xvidcore--devapi4--1.0--patch-108 |
65 |
|
|
66 |
|
* Added trellis support for MPEg quantization type. |
67 |
|
* Changed RD fixed point precision, should help avoiding overflow |
68 |
|
(see the constant TL_SHIFT) |
69 |
|
|
70 |
|
NB: we still have some problems when trellis optimizes DC for big DC |
71 |
|
values. |
72 |
|
|
73 |
|
modified files: |
74 |
|
src/utils/mbtransquant.c |
75 |
|
|
76 |
|
|
77 |
|
2003-11-22 00:53:59 GMT patch-107 |
78 |
|
|
79 |
|
Summary: |
80 |
|
Win32 lib project fix (bis) |
81 |
|
Revision: |
82 |
|
xvidcore--devapi4--1.0--patch-107 |
83 |
|
|
84 |
|
* nasm >= 0.98.37 support in project file got reverted in a previous |
85 |
|
patch, push it back. Nota bene /O3 changed to /O2 for proper |
86 |
|
compilation with msvc compiler (everyone is not supposed to compile |
87 |
|
stuff with icc) |
88 |
|
* TODO update. |
89 |
|
|
90 |
|
modified files: |
91 |
|
TODO build/win32/libxvidcore.dsp |
92 |
|
|
93 |
|
|
94 |
|
2003-11-19 21:26:34 GMT patch-106 |
95 |
|
|
96 |
|
Summary: |
97 |
|
updated bench crc |
98 |
|
Revision: |
99 |
|
xvidcore--devapi4--1.0--patch-106 |
100 |
|
|
101 |
|
updated bench crc |
102 |
|
|
103 |
|
|
104 |
|
modified files: |
105 |
|
examples/xvid_bench.c |
106 |
|
|
107 |
|
|
108 |
|
2003-11-19 16:00:00 GMT patch-105 |
109 |
|
|
110 |
|
Summary: |
111 |
|
Lumimasking fixes. |
112 |
|
Revision: |
113 |
|
xvidcore--devapi4--1.0--patch-105 |
114 |
|
|
115 |
|
from sysKin: |
116 |
|
* New plugin hook entry XVID_PLG_FRAME that happens inside FrameCodeIPB |
117 |
|
when both type and quant are known. Added hook handling in all plugins. |
118 |
|
* Fixed lumimasking. |
119 |
|
|
120 |
|
from me: |
121 |
|
* small reverse commit in pvop estimation fixed. |
122 |
|
|
123 |
|
modified files: |
124 |
|
src/bitstream/bitstream.c src/bitstream/bitstream.h |
125 |
|
src/encoder.c src/motion/estimation_pvop.c |
126 |
|
src/plugins/plugin_2pass1.c src/plugins/plugin_2pass2.c |
127 |
|
src/plugins/plugin_dump.c src/plugins/plugin_lumimasking.c |
128 |
|
src/plugins/plugin_psnr.c src/plugins/plugin_single.c |
129 |
|
src/xvid.h |
130 |
|
|
131 |
|
|
132 |
|
2003-11-19 15:37:16 GMT patch-104 |
133 |
|
|
134 |
|
Summary: |
135 |
|
Removed indirections from SearchData structure. |
136 |
|
Revision: |
137 |
|
xvidcore--devapi4--1.0--patch-104 |
138 |
|
|
139 |
|
Patch from sysKin: |
140 |
|
* removed indirections in SearchData structure. CheckCandidate functions |
141 |
|
don't use a const SearchData pointer anymore, but they should be a bit |
142 |
|
faster because of the less numerous indirections. |
143 |
|
|
144 |
|
modified files: |
145 |
|
src/motion/estimation.h src/motion/estimation_bvop.c |
146 |
|
src/motion/estimation_common.c src/motion/estimation_gmc.c |
147 |
|
src/motion/estimation_pvop.c src/motion/estimation_rd_based.c |
148 |
|
src/motion/vop_type_decision.c |
149 |
|
|
150 |
|
|
151 |
|
2003-11-19 15:33:55 GMT patch-103 |
152 |
|
|
153 |
|
Summary: |
154 |
|
Formula error in twopass code. |
155 |
|
Revision: |
156 |
|
xvidcore--devapi4--1.0--patch-103 |
157 |
|
|
158 |
|
The reversing bframe formula in 2pass 2 was not right. This was in fact |
159 |
|
a test code i used when i was maintaining the code on its own branch. |
160 |
|
I should not have commited it :\ |
161 |
|
|
162 |
|
Fixed :-) |
163 |
|
|
164 |
|
modified files: |
165 |
|
src/plugins/plugin_2pass2.c |
166 |
|
|
167 |
|
|
168 |
|
2003-11-18 21:41:08 GMT patch-102 |
169 |
|
|
170 |
|
Summary: |
171 |
|
Another problem with mis/unitialized reads. |
172 |
|
Revision: |
173 |
|
xvidcore--devapi4--1.0--patch-102 |
174 |
|
|
175 |
|
Michael introduced a fast subpel refine that uses a (iMinSAD2, currentQMV2) |
176 |
|
couple of data. The problem is that he plugged this in CheckCandidate16_qpel |
177 |
|
that is used outside this context, thus a if statement was traversed with |
178 |
|
garbage data in the standard subpel case. For perfection sake, using a |
179 |
|
iMinSAD=256*4096 value collects correct data even if it will not be used in |
180 |
|
the normal subpel case. |
181 |
|
|
182 |
|
modified files: |
183 |
|
src/motion/estimation_pvop.c |
184 |
|
|
185 |
|
|
186 |
|
2003-11-16 17:29:39 GMT patch-101 |
187 |
|
|
188 |
|
Summary: |
189 |
|
The MEanalysis patch assumed bvops were always used |
190 |
|
Revision: |
191 |
|
xvidcore--devapi4--1.0--patch-101 |
192 |
|
|
193 |
|
The MEanalysis patch assumed bvops were always used |
194 |
|
|
195 |
|
|
196 |
|
modified files: |
197 |
|
src/encoder.c |
198 |
|
|
199 |
|
|
200 |
|
2003-11-16 15:12:15 GMT patch-100 |
201 |
|
|
202 |
|
Summary: |
203 |
|
MEanalysis using wrong mvs + bframe search using wrong mvs |
204 |
|
Revision: |
205 |
|
xvidcore--devapi4--1.0--patch-100 |
206 |
|
|
207 |
|
Valgrind reported lot of unitialized reads. These unitialized reads |
208 |
|
helped sysKin finding three bugs: |
209 |
|
- ZeroMacroblock did not reset the cbp field. So for some skipped |
210 |
|
blocks, a test was done on the cbp value... |
211 |
|
- MEanalysis was using wrong mvs from the current bvop (unitialized |
212 |
|
or just wrong in current context). That's because in devapi3, bframes |
213 |
|
used to share the same mvs array wheras now, it's one array a bvop. |
214 |
|
- Collocated skipped MBs for a bvop didn't reset mvs[0] and b_mvs[0]. |
215 |
|
|
216 |
|
modified files: |
217 |
|
src/encoder.c src/motion/estimation_bvop.c src/motion/motion.h |
218 |
|
src/motion/motion_inlines.h src/motion/vop_type_decision.c |
219 |
|
|
220 |
|
|
221 |
|
2003-11-15 15:21:09 GMT patch-99 |
222 |
|
|
223 |
|
Summary: |
224 |
|
Small fixes |
225 |
|
Revision: |
226 |
|
xvidcore--devapi4--1.0--patch-99 |
227 |
|
|
228 |
|
Small fixes |
229 |
|
|
230 |
|
|
231 |
|
modified files: |
232 |
|
src/encoder.c src/motion/vop_type_decision.c |
233 |
|
|
234 |
|
|
235 |
|
2003-11-15 15:02:47 GMT patch-98 |
236 |
|
|
237 |
|
Summary: |
238 |
|
DShow update + libxvidcore project file update |
239 |
|
Revision: |
240 |
|
xvidcore--devapi4--1.0--patch-98 |
241 |
|
|
242 |
|
From peter: |
243 |
|
* DShow now links against libxvidcore.lib |
244 |
|
* Some updates to the libxvidcore project file |
245 |
|
|
246 |
|
modified files: |
247 |
|
TODO build/win32/libxvidcore.dsp dshow/dshow.dsp |
248 |
|
dshow/src/CXvidDecoder.cpp dshow/src/CXvidDecoder.h |
249 |
|
|
250 |
|
|
251 |
|
2003-11-15 01:51:28 GMT patch-97 |
252 |
|
|
253 |
|
Summary: |
254 |
|
Small updates to doc files |
255 |
|
Revision: |
256 |
|
xvidcore--devapi4--1.0--patch-97 |
257 |
|
|
258 |
|
Small updates to doc files |
259 |
|
|
260 |
|
|
261 |
|
modified files: |
262 |
|
AUTHORS CodingStyle TODO doc/INSTALL doc/README |
263 |
|
|
264 |
|
renamed files: |
265 |
|
.arch-ids/authors.txt.id |
266 |
|
==> .arch-ids/AUTHORS.id |
267 |
|
.arch-ids/todo.txt.id |
268 |
|
==> .arch-ids/TODO.id |
269 |
|
authors.txt |
270 |
|
==> AUTHORS |
271 |
|
todo.txt |
272 |
|
==> TODO |
273 |
|
|
274 |
|
|
275 |
|
2003-11-14 11:23:55 GMT patch-96 |
276 |
|
|
277 |
|
Summary: |
278 |
|
Updated ChangeLog |
279 |
|
Revision: |
280 |
|
xvidcore--devapi4--1.0--patch-96 |
281 |
|
|
282 |
|
Updated ChangeLog |
283 |
|
|
284 |
|
|
285 |
|
modified files: |
286 |
|
ChangeLog |
287 |
|
|
288 |
|
|
289 |
2003-11-13 23:09:34 GMT patch-95 |
2003-11-13 23:09:34 GMT patch-95 |
290 |
|
|
291 |
Summary: |
Summary: |