1 |
|
/************************************************************************** |
2 |
|
* |
3 |
|
* XVID MPEG-4 VIDEO CODEC |
4 |
|
* - QPel interpolation - |
5 |
|
* |
6 |
|
* This program is free software; you can redistribute it and/or modify |
7 |
|
* it under the terms of the GNU General Public License as published by |
8 |
|
* the Free Software Foundation; either version 2 of the License, or |
9 |
|
* (at your option) any later version. |
10 |
|
* |
11 |
|
* This program is distributed in the hope that it will be useful, |
12 |
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
13 |
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
14 |
|
* GNU General Public License for more details. |
15 |
|
* |
16 |
|
* You should have received a copy of the GNU General Public License |
17 |
|
* along with this program; if not, write to the Free Software |
18 |
|
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. |
19 |
|
* |
20 |
|
*************************************************************************/ |
21 |
|
|
22 |
|
/************************************************************************** |
23 |
|
* |
24 |
|
* History: |
25 |
|
* |
26 |
|
* 22.10.2002 initial coding - Skal - |
27 |
|
* |
28 |
|
*************************************************************************/ |
29 |
|
|
30 |
|
#ifndef XVID_AUTO_INCLUDE |
31 |
|
|
32 |
|
#include "../portab.h" |
33 |
|
#include "./qpel.h" |
34 |
|
|
35 |
|
////////////////////////////////////////////////////////// |
36 |
|
|
37 |
|
static const int32_t FIR_Tab_8[9][8] = { |
38 |
|
{ 14, -3, 2, -1, 0, 0, 0, 0 } |
39 |
|
, { 23, 19, -6, 3, -1, 0, 0, 0 } |
40 |
|
, { -7, 20, 20, -6, 3, -1, 0, 0 } |
41 |
|
, { 3, -6, 20, 20, -6, 3, -1, 0 } |
42 |
|
, { -1, 3, -6, 20, 20, -6, 3, -1 } |
43 |
|
, { 0, -1, 3, -6, 20, 20, -6, 3 } |
44 |
|
, { 0, 0, -1, 3, -6, 20, 20, -7 } |
45 |
|
, { 0, 0, 0, -1, 3, -6, 19, 23 } |
46 |
|
, { 0, 0, 0, 0, -1, 2, -3, 14 } |
47 |
|
}; |
48 |
|
|
49 |
|
static const int32_t FIR_Tab_16[17][16] = { |
50 |
|
{ 14, -3, 2, -1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } |
51 |
|
, { 23, 19, -6, 3, -1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } |
52 |
|
, { -7, 20, 20, -6, 3, -1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 } |
53 |
|
, { 3, -6, 20, 20, -6, 3, -1, 0, 0, 0, 0, 0, 0, 0, 0, 0 } |
54 |
|
, { -1, 3, -6, 20, 20, -6, 3, -1, 0, 0, 0, 0, 0, 0, 0, 0 } |
55 |
|
, { 0, -1, 3, -6, 20, 20, -6, 3, -1, 0, 0, 0, 0, 0, 0, 0 } |
56 |
|
, { 0, 0, -1, 3, -6, 20, 20, -6, 3, -1, 0, 0, 0, 0, 0, 0 } |
57 |
|
, { 0, 0, 0, -1, 3, -6, 20, 20, -6, 3, -1, 0, 0, 0, 0, 0 } |
58 |
|
, { 0, 0, 0, 0, -1, 3, -6, 20, 20, -6, 3, -1, 0, 0, 0, 0 } |
59 |
|
, { 0, 0, 0, 0, 0, -1, 3, -6, 20, 20, -6, 3, -1, 0, 0, 0 } |
60 |
|
, { 0, 0, 0, 0, 0, 0, -1, 3, -6, 20, 20, -6, 3, -1, 0, 0 } |
61 |
|
, { 0, 0, 0, 0, 0, 0, 0, -1, 3, -6, 20, 20, -6, 3, -1, 0 } |
62 |
|
, { 0, 0, 0, 0, 0, 0, 0, 0, -1, 3, -6, 20, 20, -6, 3, -1 } |
63 |
|
, { 0, 0, 0, 0, 0, 0, 0, 0, 0, -1, 3, -6, 20, 20, -6, 3 } |
64 |
|
, { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, -1, 3, -6, 20, 20, -7 } |
65 |
|
, { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, -1, 3, -6, 19, 23 } |
66 |
|
, { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, -1, 2, -3, 14 } |
67 |
|
}; |
68 |
|
|
69 |
|
////////////////////////////////////////////////////////// |
70 |
|
// Implementation |
71 |
|
|
72 |
|
#define XVID_AUTO_INCLUDE |
73 |
|
|
74 |
|
// 16x? filters |
75 |
|
|
76 |
|
#define SIZE 16 |
77 |
|
#define TABLE FIR_Tab_16 |
78 |
|
|
79 |
|
#define STORE(d,s) (d) = (s) |
80 |
|
#define FUNC_H H_Pass_16_C |
81 |
|
#define FUNC_V V_Pass_16_C |
82 |
|
#define FUNC_HA H_Pass_Avrg_16_C |
83 |
|
#define FUNC_VA V_Pass_Avrg_16_C |
84 |
|
#define FUNC_HA_UP H_Pass_Avrg_Up_16_C |
85 |
|
#define FUNC_VA_UP V_Pass_Avrg_Up_16_C |
86 |
|
|
87 |
|
#include __FILE__ /* self-include ourself */ |
88 |
|
|
89 |
|
// note: B-frame always uses Rnd=0... |
90 |
|
#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 |
91 |
|
#define FUNC_H H_Pass_16_Add_C |
92 |
|
#define FUNC_V V_Pass_16_Add_C |
93 |
|
#define FUNC_HA H_Pass_Avrg_16_Add_C |
94 |
|
#define FUNC_VA V_Pass_Avrg_16_Add_C |
95 |
|
#define FUNC_HA_UP H_Pass_Avrg_Up_16_Add_C |
96 |
|
#define FUNC_VA_UP V_Pass_Avrg_Up_16_Add_C |
97 |
|
|
98 |
|
#include __FILE__ /* self-include ourself */ |
99 |
|
|
100 |
|
#undef SIZE |
101 |
|
#undef TABLE |
102 |
|
|
103 |
|
// 8x? filters |
104 |
|
|
105 |
|
#define SIZE 8 |
106 |
|
#define TABLE FIR_Tab_8 |
107 |
|
|
108 |
|
#define STORE(d,s) (d) = (s) |
109 |
|
#define FUNC_H H_Pass_8_C |
110 |
|
#define FUNC_V V_Pass_8_C |
111 |
|
#define FUNC_HA H_Pass_Avrg_8_C |
112 |
|
#define FUNC_VA V_Pass_Avrg_8_C |
113 |
|
#define FUNC_HA_UP H_Pass_Avrg_Up_8_C |
114 |
|
#define FUNC_VA_UP V_Pass_Avrg_Up_8_C |
115 |
|
|
116 |
|
#include __FILE__ /* self-include ourself */ |
117 |
|
|
118 |
|
// note: B-frame always uses Rnd=0... |
119 |
|
#define STORE(d,s) (d) = ( (s)+(d)+1 ) >> 1 |
120 |
|
#define FUNC_H H_Pass_8_Add_C |
121 |
|
#define FUNC_V V_Pass_8_Add_C |
122 |
|
#define FUNC_HA H_Pass_Avrg_8_Add_C |
123 |
|
#define FUNC_VA V_Pass_Avrg_8_Add_C |
124 |
|
#define FUNC_HA_UP H_Pass_Avrg_Up_8_Add_C |
125 |
|
#define FUNC_VA_UP V_Pass_Avrg_Up_8_Add_C |
126 |
|
|
127 |
|
#include __FILE__ /* self-include ourself */ |
128 |
|
|
129 |
|
#undef SIZE |
130 |
|
#undef TABLE |
131 |
|
|
132 |
|
#undef XVID_AUTO_INCLUDE |
133 |
|
|
134 |
|
////////////////////////////////////////////////////////// |
135 |
|
// general-purpose hooks |
136 |
|
// TODO: embed in enc/dec structure? |
137 |
|
|
138 |
|
XVID_QP_FUNCS *xvid_QP_Funcs = 0; |
139 |
|
XVID_QP_FUNCS *xvid_QP_Add_Funcs = 0; |
140 |
|
|
141 |
|
////////////////////////////////////////////////////////// |
142 |
|
// plain C impl. declaration |
143 |
|
// TODO: should be declared elsewhere? |
144 |
|
|
145 |
|
XVID_QP_FUNCS xvid_QP_Funcs_C = { |
146 |
|
H_Pass_16_C, H_Pass_Avrg_16_C, H_Pass_Avrg_Up_16_C |
147 |
|
, V_Pass_16_C, V_Pass_Avrg_16_C, V_Pass_Avrg_Up_16_C |
148 |
|
|
149 |
|
, H_Pass_8_C,H_Pass_Avrg_8_C,H_Pass_Avrg_Up_8_C |
150 |
|
, V_Pass_8_C,V_Pass_Avrg_8_C,V_Pass_Avrg_Up_8_C |
151 |
|
}; |
152 |
|
|
153 |
|
XVID_QP_FUNCS xvid_QP_Add_Funcs_C = { |
154 |
|
H_Pass_16_Add_C, H_Pass_Avrg_16_Add_C, H_Pass_Avrg_Up_16_Add_C |
155 |
|
, V_Pass_16_Add_C, V_Pass_Avrg_16_Add_C, V_Pass_Avrg_Up_16_Add_C |
156 |
|
|
157 |
|
, H_Pass_8_Add_C,H_Pass_Avrg_8_Add_C,H_Pass_Avrg_Up_8_Add_C |
158 |
|
, V_Pass_8_Add_C,V_Pass_Avrg_8_Add_C,V_Pass_Avrg_Up_8_Add_C |
159 |
|
}; |
160 |
|
|
161 |
|
////////////////////////////////////////////////////////// |
162 |
|
// mmx impl. declaration (cf. qpel_mmx.asm) |
163 |
|
// TODO: should be declared elsewhere? |
164 |
|
|
165 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_16_mmx); |
166 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_Avrg_16_mmx); |
167 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_Avrg_Up_16_mmx); |
168 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_16_mmx); |
169 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_Avrg_16_mmx); |
170 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_Avrg_Up_16_mmx); |
171 |
|
|
172 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_8_mmx); |
173 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_Avrg_8_mmx); |
174 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_Avrg_Up_8_mmx); |
175 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_8_mmx); |
176 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_Avrg_8_mmx); |
177 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_Avrg_Up_8_mmx); |
178 |
|
|
179 |
|
XVID_QP_FUNCS xvid_QP_Funcs_mmx = { |
180 |
|
xvid_H_Pass_16_mmx, xvid_H_Pass_Avrg_16_mmx, xvid_H_Pass_Avrg_Up_16_mmx |
181 |
|
, xvid_V_Pass_16_mmx, xvid_V_Pass_Avrg_16_mmx, xvid_V_Pass_Avrg_Up_16_mmx |
182 |
|
|
183 |
|
, xvid_H_Pass_8_mmx,xvid_H_Pass_Avrg_8_mmx,xvid_H_Pass_Avrg_Up_8_mmx |
184 |
|
, xvid_V_Pass_8_mmx,xvid_V_Pass_Avrg_8_mmx,xvid_V_Pass_Avrg_Up_8_mmx |
185 |
|
}; |
186 |
|
|
187 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_Add_16_mmx); |
188 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_Avrg_Add_16_mmx); |
189 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_Avrg_Up_Add_16_mmx); |
190 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_Add_16_mmx); |
191 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_Avrg_Add_16_mmx); |
192 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_Avrg_Up_Add_16_mmx); |
193 |
|
|
194 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_8_Add_mmx); |
195 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_Avrg_8_Add_mmx); |
196 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_H_Pass_Avrg_Up_8_Add_mmx); |
197 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_8_Add_mmx); |
198 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_Avrg_8_Add_mmx); |
199 |
|
extern XVID_QP_PASS_SIGNATURE(xvid_V_Pass_Avrg_Up_8_Add_mmx); |
200 |
|
|
201 |
|
XVID_QP_FUNCS xvid_QP_Add_Funcs_mmx = { |
202 |
|
xvid_H_Pass_Add_16_mmx, xvid_H_Pass_Avrg_Add_16_mmx, xvid_H_Pass_Avrg_Up_Add_16_mmx |
203 |
|
, xvid_V_Pass_Add_16_mmx, xvid_V_Pass_Avrg_Add_16_mmx, xvid_V_Pass_Avrg_Up_Add_16_mmx |
204 |
|
|
205 |
|
, xvid_H_Pass_8_Add_mmx,xvid_H_Pass_Avrg_8_Add_mmx,xvid_H_Pass_Avrg_Up_8_Add_mmx |
206 |
|
, xvid_V_Pass_8_Add_mmx,xvid_V_Pass_Avrg_8_Add_mmx,xvid_V_Pass_Avrg_Up_8_Add_mmx |
207 |
|
}; |
208 |
|
|
209 |
|
////////////////////////////////////////////////////////// |
210 |
|
// tables for ASM |
211 |
|
|
212 |
|
extern uint16_t xvid_Expand_mmx[256][4]; // 8b -> 64b expansion table |
213 |
|
|
214 |
|
// Alternate way of filtering (cf. USE_TABLES flag in qpel_mmx.asm) |
215 |
|
// |
216 |
|
// 17 tables, 2K each => 34K |
217 |
|
// Mirroring can be acheived composing 11 basic tables |
218 |
|
// (for instance: (23,19,-6,3)=(20,20,-6,3)+(3,-1,0,0) |
219 |
|
// Using Symmetries (and bswap) could reduce further |
220 |
|
// the memory to 7 tables (->14K). |
221 |
|
|
222 |
|
extern int16_t xvid_FIR_1_0_0_0[256][4]; |
223 |
|
extern int16_t xvid_FIR_3_1_0_0[256][4]; |
224 |
|
extern int16_t xvid_FIR_6_3_1_0[256][4]; |
225 |
|
extern int16_t xvid_FIR_14_3_2_1[256][4]; |
226 |
|
extern int16_t xvid_FIR_20_6_3_1[256][4]; |
227 |
|
extern int16_t xvid_FIR_20_20_6_3[256][4]; |
228 |
|
extern int16_t xvid_FIR_23_19_6_3[256][4]; |
229 |
|
extern int16_t xvid_FIR_7_20_20_6[256][4]; |
230 |
|
extern int16_t xvid_FIR_6_20_20_6[256][4]; |
231 |
|
extern int16_t xvid_FIR_6_20_20_7[256][4]; |
232 |
|
extern int16_t xvid_FIR_3_6_20_20[256][4]; |
233 |
|
extern int16_t xvid_FIR_3_6_19_23[256][4]; |
234 |
|
extern int16_t xvid_FIR_1_3_6_20[256][4]; |
235 |
|
extern int16_t xvid_FIR_1_2_3_14[256][4]; |
236 |
|
extern int16_t xvid_FIR_0_1_3_6[256][4]; |
237 |
|
extern int16_t xvid_FIR_0_0_1_3[256][4]; |
238 |
|
extern int16_t xvid_FIR_0_0_0_1[256][4]; |
239 |
|
|
240 |
|
////////////////////////////////////////////////////////// |
241 |
|
|
242 |
|
uint16_t xvid_Expand_mmx[256][4]; // 8b -> 64b expansion table |
243 |
|
|
244 |
|
int16_t xvid_FIR_1_0_0_0[256][4]; |
245 |
|
int16_t xvid_FIR_3_1_0_0[256][4]; |
246 |
|
int16_t xvid_FIR_6_3_1_0[256][4]; |
247 |
|
int16_t xvid_FIR_14_3_2_1[256][4]; |
248 |
|
int16_t xvid_FIR_20_6_3_1[256][4]; |
249 |
|
int16_t xvid_FIR_20_20_6_3[256][4]; |
250 |
|
int16_t xvid_FIR_23_19_6_3[256][4]; |
251 |
|
int16_t xvid_FIR_7_20_20_6[256][4]; |
252 |
|
int16_t xvid_FIR_6_20_20_6[256][4]; |
253 |
|
int16_t xvid_FIR_6_20_20_7[256][4]; |
254 |
|
int16_t xvid_FIR_3_6_20_20[256][4]; |
255 |
|
int16_t xvid_FIR_3_6_19_23[256][4]; |
256 |
|
int16_t xvid_FIR_1_3_6_20[256][4]; |
257 |
|
int16_t xvid_FIR_1_2_3_14[256][4]; |
258 |
|
int16_t xvid_FIR_0_1_3_6[256][4]; |
259 |
|
int16_t xvid_FIR_0_0_1_3[256][4]; |
260 |
|
int16_t xvid_FIR_0_0_0_1[256][4]; |
261 |
|
|
262 |
|
static void Init_FIR_Table(int16_t Tab[][4], |
263 |
|
int A, int B, int C, int D) |
264 |
|
{ |
265 |
|
int i; |
266 |
|
for(i=0; i<256; ++i) { |
267 |
|
Tab[i][0] = i*A; |
268 |
|
Tab[i][1] = i*B; |
269 |
|
Tab[i][2] = i*C; |
270 |
|
Tab[i][3] = i*D; |
271 |
|
} |
272 |
|
} |
273 |
|
|
274 |
|
|
275 |
|
void xvid_Init_QP_mmx() |
276 |
|
{ |
277 |
|
int i; |
278 |
|
for(i=0; i<256; ++i) { |
279 |
|
xvid_Expand_mmx[i][0] = i; |
280 |
|
xvid_Expand_mmx[i][1] = i; |
281 |
|
xvid_Expand_mmx[i][2] = i; |
282 |
|
xvid_Expand_mmx[i][3] = i; |
283 |
|
} |
284 |
|
|
285 |
|
// Alternate way of filtering (cf. USE_TABLES flag in qpel_mmx.asm) |
286 |
|
|
287 |
|
Init_FIR_Table(xvid_FIR_1_0_0_0, -1, 0, 0, 0); |
288 |
|
Init_FIR_Table(xvid_FIR_3_1_0_0, 3, -1, 0, 0); |
289 |
|
Init_FIR_Table(xvid_FIR_6_3_1_0, -6, 3, -1, 0); |
290 |
|
Init_FIR_Table(xvid_FIR_14_3_2_1, 14, -3, 2, -1); |
291 |
|
Init_FIR_Table(xvid_FIR_20_6_3_1, 20, -6, 3, -1); |
292 |
|
Init_FIR_Table(xvid_FIR_20_20_6_3, 20, 20, -6, 3); |
293 |
|
Init_FIR_Table(xvid_FIR_23_19_6_3, 23, 19, -6, 3); |
294 |
|
Init_FIR_Table(xvid_FIR_7_20_20_6, -7, 20, 20, -6); |
295 |
|
Init_FIR_Table(xvid_FIR_6_20_20_6, -6, 20, 20, -6); |
296 |
|
Init_FIR_Table(xvid_FIR_6_20_20_7, -6, 20, 20, -7); |
297 |
|
Init_FIR_Table(xvid_FIR_3_6_20_20, 3, -6, 20, 20); |
298 |
|
Init_FIR_Table(xvid_FIR_3_6_19_23, 3, -6, 19, 23); |
299 |
|
Init_FIR_Table(xvid_FIR_1_3_6_20, -1, 3, -6, 20); |
300 |
|
Init_FIR_Table(xvid_FIR_1_2_3_14, -1, 2, -3, 14); |
301 |
|
Init_FIR_Table(xvid_FIR_0_1_3_6, 0, -1, 3, -6); |
302 |
|
Init_FIR_Table(xvid_FIR_0_0_1_3, 0, 0, -1, 3); |
303 |
|
Init_FIR_Table(xvid_FIR_0_0_0_1, 0, 0, 0, -1); |
304 |
|
|
305 |
|
} |
306 |
|
|
307 |
|
#endif /* !XVID_AUTO_INCLUDE */ |
308 |
|
|
309 |
|
////////////////////////////////////////////////////////// |
310 |
|
// "reference" filters impl. in plain C |
311 |
|
////////////////////////////////////////////////////////// |
312 |
|
|
313 |
|
#ifdef XVID_AUTO_INCLUDE |
314 |
|
|
315 |
|
static |
316 |
|
void FUNC_H(uint8_t *Dst, const uint8_t *Src, int32_t H, int32_t BpS, int32_t Rnd) |
317 |
|
{ |
318 |
|
while(H-->0) { |
319 |
|
int32_t i, k; |
320 |
|
int32_t Sums[SIZE] = { 0 }; |
321 |
|
for(i=0; i<=SIZE; ++i) |
322 |
|
for(k=0; k<SIZE; ++k) |
323 |
|
Sums[k] += TABLE[i][k] * Src[i]; |
324 |
|
|
325 |
|
for(i=0; i<SIZE; ++i) { |
326 |
|
int32_t C = ( Sums[i] + 16-Rnd ) >> 5; |
327 |
|
if (C<0) C = 0; else if (C>255) C = 255; |
328 |
|
STORE(Dst[i], C); |
329 |
|
} |
330 |
|
Src += BpS; |
331 |
|
Dst += BpS; |
332 |
|
} |
333 |
|
} |
334 |
|
|
335 |
|
static |
336 |
|
void FUNC_V(uint8_t *Dst, const uint8_t *Src, int32_t W, int32_t BpS, int32_t Rnd) |
337 |
|
{ |
338 |
|
while(W-->0) { |
339 |
|
int32_t i, k; |
340 |
|
int32_t Sums[SIZE] = { 0 }; |
341 |
|
const uint8_t *S = Src++; |
342 |
|
uint8_t *D = Dst++; |
343 |
|
for(i=0; i<=SIZE; ++i) { |
344 |
|
for(k=0; k<SIZE; ++k) |
345 |
|
Sums[k] += TABLE[i][k] * S[0]; |
346 |
|
S += BpS; |
347 |
|
} |
348 |
|
|
349 |
|
for(i=0; i<SIZE; ++i) { |
350 |
|
int32_t C = ( Sums[i] + 16-Rnd )>>5; |
351 |
|
if (C<0) C = 0; else if (C>255) C = 255; |
352 |
|
STORE(D[0], C); |
353 |
|
D += BpS; |
354 |
|
} |
355 |
|
} |
356 |
|
} |
357 |
|
|
358 |
|
static |
359 |
|
void FUNC_HA(uint8_t *Dst, const uint8_t *Src, int32_t H, int32_t BpS, int32_t Rnd) |
360 |
|
{ |
361 |
|
while(H-->0) { |
362 |
|
int32_t i, k; |
363 |
|
int32_t Sums[SIZE] = { 0 }; |
364 |
|
for(i=0; i<=SIZE; ++i) |
365 |
|
for(k=0; k<SIZE; ++k) |
366 |
|
Sums[k] += TABLE[i][k] * Src[i]; |
367 |
|
|
368 |
|
for(i=0; i<SIZE; ++i) { |
369 |
|
int32_t C = ( Sums[i] + 16-Rnd ) >> 5; |
370 |
|
if (C<0) C = 0; else if (C>255) C = 255; |
371 |
|
C = (C+Src[i]+1-Rnd) >> 1; |
372 |
|
STORE(Dst[i], C); |
373 |
|
} |
374 |
|
Src += BpS; |
375 |
|
Dst += BpS; |
376 |
|
} |
377 |
|
} |
378 |
|
|
379 |
|
static |
380 |
|
void FUNC_HA_UP(uint8_t *Dst, const uint8_t *Src, int32_t H, int32_t BpS, int32_t Rnd) |
381 |
|
{ |
382 |
|
while(H-->0) { |
383 |
|
int32_t i, k; |
384 |
|
int32_t Sums[SIZE] = { 0 }; |
385 |
|
for(i=0; i<=SIZE; ++i) |
386 |
|
for(k=0; k<SIZE; ++k) |
387 |
|
Sums[k] += TABLE[i][k] * Src[i]; |
388 |
|
|
389 |
|
for(i=0; i<SIZE; ++i) { |
390 |
|
int32_t C = ( Sums[i] + 16-Rnd ) >> 5; |
391 |
|
if (C<0) C = 0; else if (C>255) C = 255; |
392 |
|
C = (C+Src[i+1]+1-Rnd) >> 1; |
393 |
|
STORE(Dst[i], C); |
394 |
|
} |
395 |
|
Src += BpS; |
396 |
|
Dst += BpS; |
397 |
|
} |
398 |
|
} |
399 |
|
|
400 |
|
static |
401 |
|
void FUNC_VA(uint8_t *Dst, const uint8_t *Src, int32_t W, int32_t BpS, int32_t Rnd) |
402 |
|
{ |
403 |
|
while(W-->0) { |
404 |
|
int32_t i, k; |
405 |
|
int32_t Sums[SIZE] = { 0 }; |
406 |
|
const uint8_t *S = Src; |
407 |
|
uint8_t *D = Dst; |
408 |
|
|
409 |
|
for(i=0; i<=SIZE; ++i) { |
410 |
|
for(k=0; k<SIZE; ++k) |
411 |
|
Sums[k] += TABLE[i][k] * S[0]; |
412 |
|
S += BpS; |
413 |
|
} |
414 |
|
|
415 |
|
S = Src; |
416 |
|
for(i=0; i<SIZE; ++i) { |
417 |
|
int32_t C = ( Sums[i] + 16-Rnd )>>5; |
418 |
|
if (C<0) C = 0; else if (C>255) C = 255; |
419 |
|
C = ( C+S[0]+1-Rnd ) >> 1; |
420 |
|
STORE(D[0], C); |
421 |
|
D += BpS; |
422 |
|
S += BpS; |
423 |
|
} |
424 |
|
Src++; |
425 |
|
Dst++; |
426 |
|
} |
427 |
|
} |
428 |
|
|
429 |
|
static |
430 |
|
void FUNC_VA_UP(uint8_t *Dst, const uint8_t *Src, int32_t W, int32_t BpS, int32_t Rnd) |
431 |
|
{ |
432 |
|
while(W-->0) { |
433 |
|
int32_t i, k; |
434 |
|
int32_t Sums[SIZE] = { 0 }; |
435 |
|
const uint8_t *S = Src; |
436 |
|
uint8_t *D = Dst; |
437 |
|
|
438 |
|
for(i=0; i<=SIZE; ++i) { |
439 |
|
for(k=0; k<SIZE; ++k) |
440 |
|
Sums[k] += TABLE[i][k] * S[0]; |
441 |
|
S += BpS; |
442 |
|
} |
443 |
|
|
444 |
|
S = Src + BpS; |
445 |
|
for(i=0; i<SIZE; ++i) { |
446 |
|
int32_t C = ( Sums[i] + 16-Rnd )>>5; |
447 |
|
if (C<0) C = 0; else if (C>255) C = 255; |
448 |
|
C = ( C+S[0]+1-Rnd ) >> 1; |
449 |
|
STORE(D[0], C); |
450 |
|
D += BpS; |
451 |
|
S += BpS; |
452 |
|
} |
453 |
|
Dst++; |
454 |
|
Src++; |
455 |
|
} |
456 |
|
} |
457 |
|
|
458 |
|
#undef STORE |
459 |
|
#undef FUNC_H |
460 |
|
#undef FUNC_V |
461 |
|
#undef FUNC_HA |
462 |
|
#undef FUNC_VA |
463 |
|
#undef FUNC_HA_UP |
464 |
|
#undef FUNC_VA_UP |
465 |
|
|
466 |
|
#endif /* XVID_AUTO_INCLUDE */ |
467 |
|
|
468 |
|
////////////////////////////////////////////////////////// |