1 |
|
|
/* |
2 |
|
|
* RoQ audio encoder |
3 |
|
|
* |
4 |
|
|
* Copyright (c) 2005 Eric Lasota |
5 |
|
|
* Based on RoQ specs (c)2001 Tim Ferguson |
6 |
|
|
* |
7 |
|
|
* This file is part of FFmpeg. |
8 |
|
|
* |
9 |
|
|
* FFmpeg is free software; you can redistribute it and/or |
10 |
|
|
* modify it under the terms of the GNU Lesser General Public |
11 |
|
|
* License as published by the Free Software Foundation; either |
12 |
|
|
* version 2.1 of the License, or (at your option) any later version. |
13 |
|
|
* |
14 |
|
|
* FFmpeg is distributed in the hope that it will be useful, |
15 |
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of |
16 |
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
17 |
|
|
* Lesser General Public License for more details. |
18 |
|
|
* |
19 |
|
|
* You should have received a copy of the GNU Lesser General Public |
20 |
|
|
* License along with FFmpeg; if not, write to the Free Software |
21 |
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
22 |
|
|
*/ |
23 |
|
|
|
24 |
|
|
#include "avcodec.h" |
25 |
|
|
#include "bytestream.h" |
26 |
|
|
#include "internal.h" |
27 |
|
|
#include "mathops.h" |
28 |
|
|
|
29 |
|
|
#define ROQ_FRAME_SIZE 735 |
30 |
|
|
#define ROQ_HEADER_SIZE 8 |
31 |
|
|
|
32 |
|
|
#define MAX_DPCM (127*127) |
33 |
|
|
|
34 |
|
|
|
35 |
|
|
typedef struct ROQDPCMContext { |
36 |
|
|
short lastSample[2]; |
37 |
|
|
int input_frames; |
38 |
|
|
int buffered_samples; |
39 |
|
|
int16_t *frame_buffer; |
40 |
|
|
int64_t first_pts; |
41 |
|
|
} ROQDPCMContext; |
42 |
|
|
|
43 |
|
|
|
44 |
|
1 |
static av_cold int roq_dpcm_encode_close(AVCodecContext *avctx) |
45 |
|
|
{ |
46 |
|
1 |
ROQDPCMContext *context = avctx->priv_data; |
47 |
|
|
|
48 |
|
1 |
av_freep(&context->frame_buffer); |
49 |
|
|
|
50 |
|
1 |
return 0; |
51 |
|
|
} |
52 |
|
|
|
53 |
|
1 |
static av_cold int roq_dpcm_encode_init(AVCodecContext *avctx) |
54 |
|
|
{ |
55 |
|
1 |
ROQDPCMContext *context = avctx->priv_data; |
56 |
|
|
|
57 |
✗✓ |
1 |
if (avctx->channels > 2) { |
58 |
|
|
av_log(avctx, AV_LOG_ERROR, "Audio must be mono or stereo\n"); |
59 |
|
|
return AVERROR(EINVAL); |
60 |
|
|
} |
61 |
✗✓ |
1 |
if (avctx->sample_rate != 22050) { |
62 |
|
|
av_log(avctx, AV_LOG_ERROR, "Audio must be 22050 Hz\n"); |
63 |
|
|
return AVERROR(EINVAL); |
64 |
|
|
} |
65 |
|
|
|
66 |
|
1 |
avctx->frame_size = ROQ_FRAME_SIZE; |
67 |
|
1 |
avctx->bit_rate = (ROQ_HEADER_SIZE + ROQ_FRAME_SIZE * avctx->channels) * |
68 |
|
1 |
(22050 / ROQ_FRAME_SIZE) * 8; |
69 |
|
|
|
70 |
|
1 |
context->frame_buffer = av_malloc(8 * ROQ_FRAME_SIZE * avctx->channels * |
71 |
|
|
sizeof(*context->frame_buffer)); |
72 |
✗✓ |
1 |
if (!context->frame_buffer) |
73 |
|
|
return AVERROR(ENOMEM); |
74 |
|
|
|
75 |
|
1 |
context->lastSample[0] = context->lastSample[1] = 0; |
76 |
|
|
|
77 |
|
1 |
return 0; |
78 |
|
|
} |
79 |
|
|
|
80 |
|
264600 |
static unsigned char dpcm_predict(short *previous, short current) |
81 |
|
|
{ |
82 |
|
|
int diff; |
83 |
|
|
int negative; |
84 |
|
|
int result; |
85 |
|
|
int predicted; |
86 |
|
|
|
87 |
|
264600 |
diff = current - *previous; |
88 |
|
|
|
89 |
|
264600 |
negative = diff<0; |
90 |
|
264600 |
diff = FFABS(diff); |
91 |
|
|
|
92 |
✓✓ |
264600 |
if (diff >= MAX_DPCM) |
93 |
|
10896 |
result = 127; |
94 |
|
|
else { |
95 |
|
253704 |
result = ff_sqrt(diff); |
96 |
|
253704 |
result += diff > result*result+result; |
97 |
|
|
} |
98 |
|
|
|
99 |
|
|
/* See if this overflows */ |
100 |
|
264600 |
retry: |
101 |
|
264600 |
diff = result*result; |
102 |
✓✓ |
264600 |
if (negative) |
103 |
|
131603 |
diff = -diff; |
104 |
|
264600 |
predicted = *previous + diff; |
105 |
|
|
|
106 |
|
|
/* If it overflows, back off a step */ |
107 |
✓✗✗✓
|
264600 |
if (predicted > 32767 || predicted < -32768) { |
108 |
|
|
result--; |
109 |
|
|
goto retry; |
110 |
|
|
} |
111 |
|
|
|
112 |
|
|
/* Add the sign bit */ |
113 |
|
264600 |
result |= negative << 7; //if (negative) result |= 128; |
114 |
|
|
|
115 |
|
264600 |
*previous = predicted; |
116 |
|
|
|
117 |
|
264600 |
return result; |
118 |
|
|
} |
119 |
|
|
|
120 |
|
181 |
static int roq_dpcm_encode_frame(AVCodecContext *avctx, AVPacket *avpkt, |
121 |
|
|
const AVFrame *frame, int *got_packet_ptr) |
122 |
|
|
{ |
123 |
|
|
int i, stereo, data_size, ret; |
124 |
✓✓ |
181 |
const int16_t *in = frame ? (const int16_t *)frame->data[0] : NULL; |
125 |
|
|
uint8_t *out; |
126 |
|
181 |
ROQDPCMContext *context = avctx->priv_data; |
127 |
|
|
|
128 |
|
181 |
stereo = (avctx->channels == 2); |
129 |
|
|
|
130 |
✓✓✓✗
|
181 |
if (!in && context->input_frames >= 8) |
131 |
|
1 |
return 0; |
132 |
|
|
|
133 |
✓✗✓✓
|
180 |
if (in && context->input_frames < 8) { |
134 |
|
8 |
memcpy(&context->frame_buffer[context->buffered_samples * avctx->channels], |
135 |
|
8 |
in, avctx->frame_size * avctx->channels * sizeof(*in)); |
136 |
|
8 |
context->buffered_samples += avctx->frame_size; |
137 |
✓✓ |
8 |
if (context->input_frames == 0) |
138 |
|
1 |
context->first_pts = frame->pts; |
139 |
✓✓ |
8 |
if (context->input_frames < 7) { |
140 |
|
7 |
context->input_frames++; |
141 |
|
7 |
return 0; |
142 |
|
|
} |
143 |
|
|
} |
144 |
✓✓ |
173 |
if (context->input_frames < 8) |
145 |
|
1 |
in = context->frame_buffer; |
146 |
|
|
|
147 |
✓✗ |
173 |
if (stereo) { |
148 |
|
173 |
context->lastSample[0] &= 0xFF00; |
149 |
|
173 |
context->lastSample[1] &= 0xFF00; |
150 |
|
|
} |
151 |
|
|
|
152 |
✓✓ |
173 |
if (context->input_frames == 7) |
153 |
|
1 |
data_size = avctx->channels * context->buffered_samples; |
154 |
|
|
else |
155 |
|
172 |
data_size = avctx->channels * avctx->frame_size; |
156 |
|
|
|
157 |
✗✓ |
173 |
if ((ret = ff_alloc_packet2(avctx, avpkt, ROQ_HEADER_SIZE + data_size, 0)) < 0) |
158 |
|
|
return ret; |
159 |
|
173 |
out = avpkt->data; |
160 |
|
|
|
161 |
✓✗ |
173 |
bytestream_put_byte(&out, stereo ? 0x21 : 0x20); |
162 |
|
173 |
bytestream_put_byte(&out, 0x10); |
163 |
|
173 |
bytestream_put_le32(&out, data_size); |
164 |
|
|
|
165 |
✓✗ |
173 |
if (stereo) { |
166 |
|
173 |
bytestream_put_byte(&out, (context->lastSample[1])>>8); |
167 |
|
173 |
bytestream_put_byte(&out, (context->lastSample[0])>>8); |
168 |
|
|
} else |
169 |
|
|
bytestream_put_le16(&out, context->lastSample[0]); |
170 |
|
|
|
171 |
|
|
/* Write the actual samples */ |
172 |
✓✓ |
264773 |
for (i = 0; i < data_size; i++) |
173 |
|
264600 |
*out++ = dpcm_predict(&context->lastSample[i & 1], *in++); |
174 |
|
|
|
175 |
✓✓ |
173 |
avpkt->pts = context->input_frames <= 7 ? context->first_pts : frame->pts; |
176 |
|
173 |
avpkt->duration = data_size / avctx->channels; |
177 |
|
|
|
178 |
|
173 |
context->input_frames++; |
179 |
✗✓ |
173 |
if (!in) |
180 |
|
|
context->input_frames = FFMAX(context->input_frames, 8); |
181 |
|
|
|
182 |
|
173 |
*got_packet_ptr = 1; |
183 |
|
173 |
return 0; |
184 |
|
|
} |
185 |
|
|
|
186 |
|
|
AVCodec ff_roq_dpcm_encoder = { |
187 |
|
|
.name = "roq_dpcm", |
188 |
|
|
.long_name = NULL_IF_CONFIG_SMALL("id RoQ DPCM"), |
189 |
|
|
.type = AVMEDIA_TYPE_AUDIO, |
190 |
|
|
.id = AV_CODEC_ID_ROQ_DPCM, |
191 |
|
|
.priv_data_size = sizeof(ROQDPCMContext), |
192 |
|
|
.init = roq_dpcm_encode_init, |
193 |
|
|
.encode2 = roq_dpcm_encode_frame, |
194 |
|
|
.close = roq_dpcm_encode_close, |
195 |
|
|
.capabilities = AV_CODEC_CAP_DELAY, |
196 |
|
|
.sample_fmts = (const enum AVSampleFormat[]){ AV_SAMPLE_FMT_S16, |
197 |
|
|
AV_SAMPLE_FMT_NONE }, |
198 |
|
|
}; |