aboutsummaryrefslogtreecommitdiffstats
path: root/libavcodec/msmpeg4.c
blob: 16b6f1895090893d7da447a8156c9324a1c6359c (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
/*
 * MSMPEG4 backend for encoder and decoder
 * Copyright (c) 2001 Fabrice Bellard
 * Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
 *
 * msmpeg4v1 & v2 stuff by Michael Niedermayer <michaelni@gmx.at>
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

/**
 * @file
 * MSMPEG4 backend for encoder and decoder
 */

#include "libavutil/thread.h"

#include "avcodec.h"
#include "idctdsp.h"
#include "mpegvideo.h"
#include "msmpeg4.h"
#include "libavutil/x86/asm.h"
#include "h263.h"
#include "mpeg4video.h"
#include "msmpeg4data.h"
#include "mpegvideodata.h"
#include "vc1data.h"
#include "libavutil/imgutils.h"

/*
 * You can also call this codec: MPEG-4 with a twist!
 *
 * TODO:
 *        - (encoding) select best mv table (two choices)
 *        - (encoding) select best vlc/dc table
 */

/* This table is practically identical to the one from H.263
 * except that it is inverted. */
static av_cold void init_h263_dc_for_msmpeg4(void)
{
    for (int level = -256; level < 256; level++) {
        int uni_code, uni_len;
        int size, v, l;
        /* find number of bits */
        size = 0;
        v = abs(level);
        while (v) {
            v >>= 1;
            size++;
        }

        if (level < 0)
            l = (-level) ^ ((1 << size) - 1);
        else
            l = level;

        /* luminance H.263 */
        uni_code  = ff_mpeg4_DCtab_lum[size][0];
        uni_len   = ff_mpeg4_DCtab_lum[size][1];
        uni_code ^= (1 << uni_len) - 1; //M$ does not like compatibility

        if (size > 0) {
            uni_code <<= size; uni_code |= l;
            uni_len   += size;
            if (size > 8) {
                uni_code <<= 1; uni_code |= 1;
                uni_len++;
            }
        }
        ff_v2_dc_lum_table[level + 256][0] = uni_code;
        ff_v2_dc_lum_table[level + 256][1] = uni_len;

        /* chrominance H.263 */
        uni_code  = ff_mpeg4_DCtab_chrom[size][0];
        uni_len   = ff_mpeg4_DCtab_chrom[size][1];
        uni_code ^= (1 << uni_len) - 1; //M$ does not like compatibility

        if (size > 0) {
            uni_code <<= size; uni_code |= l;
            uni_len   +=size;
            if (size > 8) {
                uni_code <<= 1; uni_code |= 1;
                uni_len++;
            }
        }
        ff_v2_dc_chroma_table[level + 256][0] = uni_code;
        ff_v2_dc_chroma_table[level + 256][1] = uni_len;
    }
}

static av_cold void msmpeg4_common_init_static(void)
{
    static uint8_t rl_table_store[NB_RL_TABLES][2][2 * MAX_RUN + MAX_LEVEL + 3];

    for (int i = 0; i < NB_RL_TABLES; i++)
        ff_rl_init(&ff_rl_table[i], rl_table_store[i]);

    init_h263_dc_for_msmpeg4();
}

av_cold void ff_msmpeg4_common_init(MpegEncContext *s)
{
    static AVOnce init_static_once = AV_ONCE_INIT;

    switch(s->msmpeg4_version){
    case 1:
    case 2:
        s->y_dc_scale_table=
        s->c_dc_scale_table= ff_mpeg1_dc_scale_table;
        break;
    case 3:
        if(s->workaround_bugs){
            s->y_dc_scale_table= ff_old_ff_y_dc_scale_table;
            s->c_dc_scale_table= ff_wmv1_c_dc_scale_table;
        } else{
            s->y_dc_scale_table= ff_mpeg4_y_dc_scale_table;
            s->c_dc_scale_table= ff_mpeg4_c_dc_scale_table;
        }
        break;
    case 4:
    case 5:
        s->y_dc_scale_table= ff_wmv1_y_dc_scale_table;
        s->c_dc_scale_table= ff_wmv1_c_dc_scale_table;
        break;
#if CONFIG_VC1_DECODER
    case 6:
        s->y_dc_scale_table= ff_wmv3_dc_scale_table;
        s->c_dc_scale_table= ff_wmv3_dc_scale_table;
        break;
#endif

    }


    if(s->msmpeg4_version>=4){
        ff_init_scantable(s->idsp.idct_permutation, &s->intra_scantable,   ff_wmv1_scantable[1]);
        ff_init_scantable(s->idsp.idct_permutation, &s->intra_h_scantable, ff_wmv1_scantable[2]);
        ff_init_scantable(s->idsp.idct_permutation, &s->intra_v_scantable, ff_wmv1_scantable[3]);
        ff_init_scantable(s->idsp.idct_permutation, &s->inter_scantable,   ff_wmv1_scantable[0]);
    }
    //Note the default tables are set in common_init in mpegvideo.c

    ff_thread_once(&init_static_once, msmpeg4_common_init_static);
}

/* predict coded block */
int ff_msmpeg4_coded_block_pred(MpegEncContext * s, int n, uint8_t **coded_block_ptr)
{
    int xy, wrap, pred, a, b, c;

    xy = s->block_index[n];
    wrap = s->b8_stride;

    /* B C
     * A X
     */
    a = s->coded_block[xy - 1       ];
    b = s->coded_block[xy - 1 - wrap];
    c = s->coded_block[xy     - wrap];

    if (b == c) {
        pred = a;
    } else {
        pred = c;
    }

    /* store value */
    *coded_block_ptr = &s->coded_block[xy];

    return pred;
}

static int get_dc(uint8_t *src, int stride, int scale, int block_size)
{
    int y;
    int sum=0;
    for(y=0; y<block_size; y++){
        int x;
        for(x=0; x<block_size; x++){
            sum+=src[x + y*stride];
        }
    }
    return FASTDIV((sum + (scale>>1)), scale);
}

/* dir = 0: left, dir = 1: top prediction */
int ff_msmpeg4_pred_dc(MpegEncContext *s, int n,
                       int16_t **dc_val_ptr, int *dir_ptr)
{
    int a, b, c, wrap, pred, scale;
    int16_t *dc_val;

    /* find prediction */
    if (n < 4) {
        scale = s->y_dc_scale;
    } else {
        scale = s->c_dc_scale;
    }

    wrap = s->block_wrap[n];
    dc_val= s->dc_val[0] + s->block_index[n];

    /* B C
     * A X
     */
    a = dc_val[ - 1];
    b = dc_val[ - 1 - wrap];
    c = dc_val[ - wrap];

    if(s->first_slice_line && (n&2)==0 && s->msmpeg4_version<4){
        b=c=1024;
    }

    /* XXX: the following solution consumes divisions, but it does not
       necessitate to modify mpegvideo.c. The problem comes from the
       fact they decided to store the quantized DC (which would lead
       to problems if Q could vary !) */
#if ARCH_X86 && HAVE_7REGS && HAVE_EBX_AVAILABLE
    __asm__ volatile(
        "movl %3, %%eax         \n\t"
        "shrl $1, %%eax         \n\t"
        "addl %%eax, %2         \n\t"
        "addl %%eax, %1         \n\t"
        "addl %0, %%eax         \n\t"
        "imull %4               \n\t"
        "movl %%edx, %0         \n\t"
        "movl %1, %%eax         \n\t"
        "imull %4               \n\t"
        "movl %%edx, %1         \n\t"
        "movl %2, %%eax         \n\t"
        "imull %4               \n\t"
        "movl %%edx, %2         \n\t"
        : "+b" (a), "+c" (b), "+D" (c)
        : "g" (scale), "S" (ff_inverse[scale])
        : "%eax", "%edx"
    );
#else
    /* Divisions are costly everywhere; optimize the most common case. */
    if (scale == 8) {
        a = (a + (8 >> 1)) / 8;
        b = (b + (8 >> 1)) / 8;
        c = (c + (8 >> 1)) / 8;
    } else {
        a = FASTDIV((a + (scale >> 1)), scale);
        b = FASTDIV((b + (scale >> 1)), scale);
        c = FASTDIV((c + (scale >> 1)), scale);
    }
#endif
    /* XXX: WARNING: they did not choose the same test as MPEG-4. This
       is very important ! */
    if(s->msmpeg4_version>3){
        if(s->inter_intra_pred){
            uint8_t *dest;
            int wrap;

            if(n==1){
                pred=a;
                *dir_ptr = 0;
            }else if(n==2){
                pred=c;
                *dir_ptr = 1;
            }else if(n==3){
                if (abs(a - b) < abs(b - c)) {
                    pred = c;
                    *dir_ptr = 1;
                } else {
                    pred = a;
                    *dir_ptr = 0;
                }
            }else{
                int bs = 8 >> s->avctx->lowres;
                if(n<4){
                    wrap= s->linesize;
                    dest= s->current_picture.f->data[0] + (((n >> 1) + 2*s->mb_y) * bs*  wrap ) + ((n & 1) + 2*s->mb_x) * bs;
                }else{
                    wrap= s->uvlinesize;
                    dest= s->current_picture.f->data[n - 3] + (s->mb_y * bs * wrap) + s->mb_x * bs;
                }
                if(s->mb_x==0) a= (1024 + (scale>>1))/scale;
                else           a= get_dc(dest-bs, wrap, scale*8>>(2*s->avctx->lowres), bs);
                if(s->mb_y==0) c= (1024 + (scale>>1))/scale;
                else           c= get_dc(dest-bs*wrap, wrap, scale*8>>(2*s->avctx->lowres), bs);

                if (s->h263_aic_dir==0) {
                    pred= a;
                    *dir_ptr = 0;
                }else if (s->h263_aic_dir==1) {
                    if(n==0){
                        pred= c;
                        *dir_ptr = 1;
                    }else{
                        pred= a;
                        *dir_ptr = 0;
                    }
                }else if (s->h263_aic_dir==2) {
                    if(n==0){
                        pred= a;
                        *dir_ptr = 0;
                    }else{
                        pred= c;
                        *dir_ptr = 1;
                    }
                } else {
                    pred= c;
                    *dir_ptr = 1;
                }
            }
        }else{
            if (abs(a - b) < abs(b - c)) {
                pred = c;
                *dir_ptr = 1;
            } else {
                pred = a;
                *dir_ptr = 0;
            }
        }
    }else{
        if (abs(a - b) <= abs(b - c)) {
            pred = c;
            *dir_ptr = 1;
        } else {
            pred = a;
            *dir_ptr = 0;
        }
    }

    /* update predictor */
    *dc_val_ptr = &dc_val[0];
    return pred;
}