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
|
/*
* Microsoft Video-1 Encoder
* Copyright (c) 2009 Konstantin Shishkov
*
* 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 msvideo1enc.c
* Microsoft Video-1 encoder
*/
#include "avcodec.h"
#include "bytestream.h"
#include "libavutil/lfg.h"
#include "elbg.h"
#include "libavutil/imgutils.h"
/**
* Encoder context
*/
typedef struct Msvideo1EncContext {
AVCodecContext *avctx;
AVFrame pic;
AVLFG rnd;
uint8_t *prev;
int block[16*3];
int block2[16*3];
int codebook[8*3];
int codebook2[8*3];
int output[16*3];
int output2[16*3];
int avg[3];
int bestpos;
int keyint;
} Msvideo1EncContext;
enum MSV1Mode{
MODE_SKIP = 0,
MODE_FILL,
MODE_2COL,
MODE_8COL,
};
#define SKIP_PREFIX 0x8400
#define SKIPS_MAX 0x0FFF
#define MKRGB555(in, off) ((in[off] << 10) | (in[off + 1] << 5) | (in[off + 2]))
static const int remap[16] = { 0, 1, 4, 5, 2, 3, 6, 7, 8, 9, 12, 13, 10, 11, 14, 15 };
static int encode_frame(AVCodecContext *avctx, uint8_t *buf, int buf_size, void *data)
{
Msvideo1EncContext * const c = avctx->priv_data;
AVFrame *pict = data;
AVFrame * const p = &c->pic;
uint16_t *src;
uint8_t *prevptr;
uint8_t *dst = buf;
int keyframe = 1;
int no_skips = 1;
int i, j, k, x, y;
int skips = 0;
*p = *pict;
if(!c->prev)
c->prev = av_malloc(avctx->width * 3 * (avctx->height + 3));
prevptr = c->prev + avctx->width * 3 * (((avctx->height + 3)&~3) - 1);
src = (uint16_t*)(p->data[0] + p->linesize[0]*(((avctx->height + 3)&~3) - 1));
if(c->keyint >= avctx->keyint_min)
keyframe = 1;
p->quality = 24;
for(y = 0; y < avctx->height; y += 4){
for(x = 0; x < avctx->width; x += 4){
int bestmode = MODE_SKIP;
int bestscore = INT_MAX;
int flags = 0;
int score;
for(j = 0; j < 4; j++){
for(i = 0; i < 4; i++){
uint16_t val = src[x + i - j*p->linesize[0]/2];
for(k = 0; k < 3; k++){
c->block[(i + j*4)*3 + k] = (val >> (10-k*5)) & 0x1F;
c->block2[remap[i + j*4]*3 + k] = (val >> (10-k*5)) & 0x1F;
}
}
}
if(!keyframe){
bestscore = 0;
for(j = 0; j < 4; j++){
for(i = 0; i < 4*3; i++){
int t = prevptr[x*3 + i + j*p->linesize[0]] - c->block[i + j*4*3];
bestscore += t*t;
}
}
bestscore /= p->quality;
}
// try to find optimal value to fill whole 4x4 block
score = 0;
ff_init_elbg(c->block, 3, 16, c->avg, 1, 1, c->output, &c->rnd);
ff_do_elbg (c->block, 3, 16, c->avg, 1, 1, c->output, &c->rnd);
if(c->avg[0] == 1) // red component = 1 will be written as skip code
c->avg[0] = 0;
for(j = 0; j < 4; j++){
for(i = 0; i < 4; i++){
for(k = 0; k < 3; k++){
int t = c->avg[k] - c->block[(i+j*4)*3+k];
score += t*t;
}
}
}
score /= p->quality;
score += 2;
if(score < bestscore){
bestscore = score;
bestmode = MODE_FILL;
}
// search for optimal filling of 2-color block
score = 0;
ff_init_elbg(c->block, 3, 16, c->codebook, 2, 1, c->output, &c->rnd);
ff_do_elbg (c->block, 3, 16, c->codebook, 2, 1, c->output, &c->rnd);
// last output value should be always 1, swap codebooks if needed
if(!c->output[15]){
for(i = 0; i < 3; i++)
FFSWAP(uint8_t, c->codebook[i], c->codebook[i+3]);
for(i = 0; i < 16; i++)
c->output[i] ^= 1;
}
for(j = 0; j < 4; j++){
for(i = 0; i < 4; i++){
for(k = 0; k < 3; k++){
int t = c->codebook[c->output[i+j*4]*3 + k] - c->block[i*3+k+j*4*3];
score += t*t;
}
}
}
score /= p->quality;
score += 6;
if(score < bestscore){
bestscore = score;
bestmode = MODE_2COL;
}
// search for optimal filling of 2-color 2x2 subblocks
score = 0;
for(i = 0; i < 4; i++){
ff_init_elbg(c->block2 + i*4*3, 3, 4, c->codebook2 + i*2*3, 2, 1, c->output2 + i*4, &c->rnd);
ff_do_elbg (c->block2 + i*4*3, 3, 4, c->codebook2 + i*2*3, 2, 1, c->output2 + i*4, &c->rnd);
}
// last value should be always 1, swap codebooks if needed
if(!c->output2[15]){
for(i = 0; i < 3; i++)
FFSWAP(uint8_t, c->codebook2[i+18], c->codebook2[i+21]);
for(i = 12; i < 16; i++)
c->output2[i] ^= 1;
}
for(j = 0; j < 4; j++){
for(i = 0; i < 4; i++){
for(k = 0; k < 3; k++){
int t = c->codebook2[(c->output2[remap[i+j*4]] + (i&2) + (j&2)*2)*3+k] - c->block[i*3+k + j*4*3];
score += t*t;
}
}
}
score /= p->quality;
score += 18;
if(score < bestscore){
bestscore = score;
bestmode = MODE_8COL;
}
if(bestmode == MODE_SKIP){
skips++;
no_skips = 0;
}
if((bestmode != MODE_SKIP && skips) || skips == SKIPS_MAX){
bytestream_put_le16(&dst, skips | SKIP_PREFIX);
skips = 0;
}
switch(bestmode){
case MODE_FILL:
bytestream_put_le16(&dst, MKRGB555(c->avg,0) | 0x8000);
for(j = 0; j < 4; j++)
for(i = 0; i < 4; i++)
for(k = 0; k < 3; k++)
prevptr[i*3 + k - j*3*avctx->width] = c->avg[k];
break;
case MODE_2COL:
for(j = 0; j < 4; j++){
for(i = 0; i < 4; i++){
flags |= (c->output[i + j*4]^1) << (i + j*4);
for(k = 0; k < 3; k++)
prevptr[i*3 + k - j*3*avctx->width] = c->codebook[c->output[i + j*4]*3 + k];
}
}
bytestream_put_le16(&dst, flags);
bytestream_put_le16(&dst, MKRGB555(c->codebook, 0));
bytestream_put_le16(&dst, MKRGB555(c->codebook, 3));
break;
case MODE_8COL:
for(j = 0; j < 4; j++){
for(i = 0; i < 4; i++){
flags |= (c->output2[remap[i + j*4]]^1) << (i + j*4);
for(k = 0; k < 3; k++)
prevptr[i*3 + k - j*3*avctx->width] = c->codebook2[(c->output2[remap[i+j*4]] + (i&2) + (j&2)*2)*3 + k];
}
}
bytestream_put_le16(&dst, flags);
bytestream_put_le16(&dst, MKRGB555(c->codebook2, 0) | 0x8000);
for(i = 3; i < 24; i += 3)
bytestream_put_le16(&dst, MKRGB555(c->codebook2, i));
break;
}
}
src -= p->linesize[0] << 1;
prevptr -= avctx->width * 3 * 4;
}
if(skips)
bytestream_put_le16(&dst, skips | SKIP_PREFIX);
//EOF
bytestream_put_byte(&dst, 0);
bytestream_put_byte(&dst, 0);
if(no_skips)
keyframe = 1;
if(keyframe)
c->keyint = 0;
else
c->keyint++;
p->pict_type= keyframe ? FF_I_TYPE : FF_P_TYPE;
p->key_frame= keyframe;
return dst - buf;
}
/**
* init encoder
*/
static av_cold int encode_init(AVCodecContext *avctx)
{
Msvideo1EncContext * const c = avctx->priv_data;
c->avctx = avctx;
if (av_image_check_size(avctx->width, avctx->height, 0, avctx) < 0) {
return -1;
}
avcodec_get_frame_defaults(&c->pic);
avctx->coded_frame = (AVFrame*)&c->pic;
c->keyint = avctx->keyint_min;
av_lfg_init(&c->rnd, 1);
return 0;
}
/**
* Uninit encoder
*/
static av_cold int encode_end(AVCodecContext *avctx)
{
Msvideo1EncContext * const c = avctx->priv_data;
av_freep(&c->prev);
return 0;
}
AVCodec ff_msvideo1_encoder = {
"msvideo1",
AVMEDIA_TYPE_VIDEO,
CODEC_ID_MSVIDEO1,
sizeof(Msvideo1EncContext),
encode_init,
encode_frame,
encode_end,
.pix_fmts = (enum PixelFormat[]){PIX_FMT_RGB555, PIX_FMT_NONE},
.long_name = NULL_IF_CONFIG_SMALL("Microsoft Video-1"),
};
|