2 * This file is part of FFmpeg.
4 * FFmpeg is free software; you can redistribute it and/or
5 * modify it under the terms of the GNU Lesser General Public
6 * License as published by the Free Software Foundation; either
7 * version 2.1 of the License, or (at your option) any later version.
9 * FFmpeg is distributed in the hope that it will be useful,
10 * but WITHOUT ANY WARRANTY; without even the implied warranty of
11 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
12 * Lesser General Public License for more details.
14 * You should have received a copy of the GNU Lesser General Public
15 * License along with FFmpeg; if not, write to the Free Software
16 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19 #include "libavutil/attributes.h"
20 #include "libavutil/common.h"
26 #define W1 2841 /* 2048*sqrt (2)*cos (1*pi/16) */
27 #define W2 2676 /* 2048*sqrt (2)*cos (2*pi/16) */
28 #define W3 2408 /* 2048*sqrt (2)*cos (3*pi/16) */
29 #define W4 2048 /* 2048*sqrt (2)*cos (4*pi/16) */
30 #define W5 1609 /* 2048*sqrt (2)*cos (5*pi/16) */
31 #define W6 1108 /* 2048*sqrt (2)*cos (6*pi/16) */
32 #define W7 565 /* 2048*sqrt (2)*cos (7*pi/16) */
34 static void wmv2_idct_row(short * b)
37 int a0, a1, a2, a3, a4, a5, a6, a7;
40 a1 = W1 * b[1] + W7 * b[7];
41 a7 = W7 * b[1] - W1 * b[7];
42 a5 = W5 * b[5] + W3 * b[3];
43 a3 = W3 * b[5] - W5 * b[3];
44 a2 = W2 * b[2] + W6 * b[6];
45 a6 = W6 * b[2] - W2 * b[6];
46 a0 = W0 * b[0] + W0 * b[4];
47 a4 = W0 * b[0] - W0 * b[4];
50 s1 = (181 * (a1 - a5 + a7 - a3) + 128) >> 8; // 1, 3, 5, 7
51 s2 = (181 * (a1 - a5 - a7 + a3) + 128) >> 8;
54 b[0] = (a0 + a2 + a1 + a5 + (1 << 7)) >> 8;
55 b[1] = (a4 + a6 + s1 + (1 << 7)) >> 8;
56 b[2] = (a4 - a6 + s2 + (1 << 7)) >> 8;
57 b[3] = (a0 - a2 + a7 + a3 + (1 << 7)) >> 8;
58 b[4] = (a0 - a2 - a7 - a3 + (1 << 7)) >> 8;
59 b[5] = (a4 - a6 - s2 + (1 << 7)) >> 8;
60 b[6] = (a4 + a6 - s1 + (1 << 7)) >> 8;
61 b[7] = (a0 + a2 - a1 - a5 + (1 << 7)) >> 8;
64 static void wmv2_idct_col(short * b)
67 int a0, a1, a2, a3, a4, a5, a6, a7;
69 /* step 1, with extended precision */
70 a1 = (W1 * b[8 * 1] + W7 * b[8 * 7] + 4) >> 3;
71 a7 = (W7 * b[8 * 1] - W1 * b[8 * 7] + 4) >> 3;
72 a5 = (W5 * b[8 * 5] + W3 * b[8 * 3] + 4) >> 3;
73 a3 = (W3 * b[8 * 5] - W5 * b[8 * 3] + 4) >> 3;
74 a2 = (W2 * b[8 * 2] + W6 * b[8 * 6] + 4) >> 3;
75 a6 = (W6 * b[8 * 2] - W2 * b[8 * 6] + 4) >> 3;
76 a0 = (W0 * b[8 * 0] + W0 * b[8 * 4] ) >> 3;
77 a4 = (W0 * b[8 * 0] - W0 * b[8 * 4] ) >> 3;
80 s1 = (181 * (a1 - a5 + a7 - a3) + 128) >> 8;
81 s2 = (181 * (a1 - a5 - a7 + a3) + 128) >> 8;
84 b[8 * 0] = (a0 + a2 + a1 + a5 + (1 << 13)) >> 14;
85 b[8 * 1] = (a4 + a6 + s1 + (1 << 13)) >> 14;
86 b[8 * 2] = (a4 - a6 + s2 + (1 << 13)) >> 14;
87 b[8 * 3] = (a0 - a2 + a7 + a3 + (1 << 13)) >> 14;
89 b[8 * 4] = (a0 - a2 - a7 - a3 + (1 << 13)) >> 14;
90 b[8 * 5] = (a4 - a6 - s2 + (1 << 13)) >> 14;
91 b[8 * 6] = (a4 + a6 - s1 + (1 << 13)) >> 14;
92 b[8 * 7] = (a0 + a2 - a1 - a5 + (1 << 13)) >> 14;
95 static void wmv2_idct_add_c(uint8_t *dest, int line_size, int16_t *block)
99 for (i = 0; i < 64; i += 8)
100 wmv2_idct_row(block + i);
101 for (i = 0; i < 8; i++)
102 wmv2_idct_col(block + i);
104 for (i = 0; i < 8; i++) {
105 dest[0] = av_clip_uint8(dest[0] + block[0]);
106 dest[1] = av_clip_uint8(dest[1] + block[1]);
107 dest[2] = av_clip_uint8(dest[2] + block[2]);
108 dest[3] = av_clip_uint8(dest[3] + block[3]);
109 dest[4] = av_clip_uint8(dest[4] + block[4]);
110 dest[5] = av_clip_uint8(dest[5] + block[5]);
111 dest[6] = av_clip_uint8(dest[6] + block[6]);
112 dest[7] = av_clip_uint8(dest[7] + block[7]);
118 static void wmv2_idct_put_c(uint8_t *dest, int line_size, int16_t *block)
122 for (i = 0; i < 64; i += 8)
123 wmv2_idct_row(block + i);
124 for (i = 0; i < 8; i++)
125 wmv2_idct_col(block + i);
127 for (i = 0; i < 8; i++) {
128 dest[0] = av_clip_uint8(block[0]);
129 dest[1] = av_clip_uint8(block[1]);
130 dest[2] = av_clip_uint8(block[2]);
131 dest[3] = av_clip_uint8(block[3]);
132 dest[4] = av_clip_uint8(block[4]);
133 dest[5] = av_clip_uint8(block[5]);
134 dest[6] = av_clip_uint8(block[6]);
135 dest[7] = av_clip_uint8(block[7]);
141 av_cold void ff_wmv2dsp_init(WMV2DSPContext *c)
143 c->idct_add = wmv2_idct_add_c;
144 c->idct_put = wmv2_idct_put_c;
145 c->idct_perm = FF_NO_IDCT_PERM;