11144713c54925d54378bad7a9bcec53acd08b93
[libav.git] / libavcodec / libamr.c
1 /*
2 * AMR Audio decoder stub
3 * Copyright (c) 2003 the ffmpeg project
4 *
5 * This file is part of FFmpeg.
6 *
7 * FFmpeg is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU Lesser General Public
9 * License as published by the Free Software Foundation; either
10 * version 2.1 of the License, or (at your option) any later version.
11 *
12 * FFmpeg is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 * Lesser General Public License for more details.
16 *
17 * You should have received a copy of the GNU Lesser General Public
18 * License along with FFmpeg; if not, write to the Free Software
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 */
21
22 /** @file
23 * Adaptive Multi-Rate (AMR) Audio decoder stub.
24 *
25 * This code implements both an AMR-NarrowBand (AMR-NB) and an AMR-WideBand
26 * (AMR-WB) audio encoder/decoder through external reference code from
27 * http://www.3gpp.org/. The license of the code from 3gpp is unclear so you
28 * have to download the code separately. Two versions exists: One fixed-point
29 * and one floating-point. For some reason the float encoder is significantly
30 * faster at least on a P4 1.5GHz (0.9s instead of 9.9s on a 30s audio clip
31 * at MR102). Both float and fixed point are supported for AMR-NB, but only
32 * float for AMR-WB.
33 *
34 * \section AMR-NB
35 *
36 * \subsection Float
37 * The float version (default) can be downloaded from:
38 * http://www.3gpp.org/ftp/Specs/archive/26_series/26.104/26104-610.zip
39 *
40 * \subsection Specification
41 * The specification for AMR-NB can be found in TS 26.071
42 * (http://www.3gpp.org/ftp/Specs/html-info/26071.htm) and some other
43 * info at http://www.3gpp.org/ftp/Specs/html-info/26-series.htm.
44 *
45 * \section AMR-WB
46 *
47 * \subsection Float
48 * The reference code can be downloaded from:
49 * http://www.3gpp.org/ftp/Specs/archive/26_series/26.204/26204-600.zip
50 *
51 * \subsection Specification
52 * The specification for AMR-WB can be found in TS 26.171
53 * (http://www.3gpp.org/ftp/Specs/html-info/26171.htm) and some other
54 * info at http://www.3gpp.org/ftp/Specs/html-info/26-series.htm.
55 *
56 */
57
58 #include "avcodec.h"
59
60 #include <amrnb/interf_dec.h>
61 #include <amrnb/interf_enc.h>
62
63 static const char nb_bitrate_unsupported[] =
64 "bitrate not supported: use one of 4.75k, 5.15k, 5.9k, 6.7k, 7.4k, 7.95k, 10.2k or 12.2k\n";
65 static const char wb_bitrate_unsupported[] =
66 "bitrate not supported: use one of 6.6k, 8.85k, 12.65k, 14.25k, 15.85k, 18.25k, 19.85k, 23.05k, or 23.85k\n";
67
68 /* Common code for fixed and float version*/
69 typedef struct AMR_bitrates
70 {
71 int rate;
72 enum Mode mode;
73 } AMR_bitrates;
74
75 /* Match desired bitrate */
76 static int getBitrateMode(int bitrate)
77 {
78 /* make the correspondance between bitrate and mode */
79 AMR_bitrates rates[]={ {4750,MR475},
80 {5150,MR515},
81 {5900,MR59},
82 {6700,MR67},
83 {7400,MR74},
84 {7950,MR795},
85 {10200,MR102},
86 {12200,MR122},
87 };
88 int i;
89
90 for(i=0;i<8;i++)
91 {
92 if(rates[i].rate==bitrate)
93 {
94 return rates[i].mode;
95 }
96 }
97 /* no bitrate matching, return an error */
98 return -1;
99 }
100
101 static void amr_decode_fix_avctx(AVCodecContext * avctx)
102 {
103 const int is_amr_wb = 1 + (avctx->codec_id == CODEC_ID_AMR_WB);
104
105 if(avctx->sample_rate == 0)
106 {
107 avctx->sample_rate = 8000 * is_amr_wb;
108 }
109
110 if(avctx->channels == 0)
111 {
112 avctx->channels = 1;
113 }
114
115 avctx->frame_size = 160 * is_amr_wb;
116 avctx->sample_fmt = SAMPLE_FMT_S16;
117 }
118
119 #if CONFIG_LIBAMR_NB
120
121 typedef struct AMRContext {
122 int frameCount;
123 void * decState;
124 int *enstate;
125 int enc_bitrate;
126 } AMRContext;
127
128 static av_cold int amr_nb_decode_init(AVCodecContext * avctx)
129 {
130 AMRContext *s = avctx->priv_data;
131
132 s->frameCount=0;
133 s->decState=Decoder_Interface_init();
134 if(!s->decState)
135 {
136 av_log(avctx, AV_LOG_ERROR, "Decoder_Interface_init error\r\n");
137 return -1;
138 }
139
140 amr_decode_fix_avctx(avctx);
141
142 if(avctx->channels > 1)
143 {
144 av_log(avctx, AV_LOG_ERROR, "amr_nb: multichannel decoding not supported\n");
145 return -1;
146 }
147
148 return 0;
149 }
150
151 static av_cold int amr_nb_encode_init(AVCodecContext * avctx)
152 {
153 AMRContext *s = avctx->priv_data;
154
155 s->frameCount=0;
156
157 if(avctx->sample_rate!=8000)
158 {
159 av_log(avctx, AV_LOG_ERROR, "Only 8000Hz sample rate supported\n");
160 return -1;
161 }
162
163 if(avctx->channels!=1)
164 {
165 av_log(avctx, AV_LOG_ERROR, "Only mono supported\n");
166 return -1;
167 }
168
169 avctx->frame_size=160;
170 avctx->coded_frame= avcodec_alloc_frame();
171
172 s->enstate=Encoder_Interface_init(0);
173 if(!s->enstate)
174 {
175 av_log(avctx, AV_LOG_ERROR, "Encoder_Interface_init error\n");
176 return -1;
177 }
178
179 if((s->enc_bitrate=getBitrateMode(avctx->bit_rate))<0)
180 {
181 av_log(avctx, AV_LOG_ERROR, nb_bitrate_unsupported);
182 return -1;
183 }
184
185 return 0;
186 }
187
188 static av_cold int amr_nb_decode_close(AVCodecContext * avctx)
189 {
190 AMRContext *s = avctx->priv_data;
191
192 Decoder_Interface_exit(s->decState);
193 return 0;
194 }
195
196 static av_cold int amr_nb_encode_close(AVCodecContext * avctx)
197 {
198 AMRContext *s = avctx->priv_data;
199
200 Encoder_Interface_exit(s->enstate);
201 av_freep(&avctx->coded_frame);
202 return 0;
203 }
204
205 static int amr_nb_decode_frame(AVCodecContext * avctx,
206 void *data, int *data_size,
207 AVPacket *avpkt)
208 {
209 const uint8_t *buf = avpkt->data;
210 int buf_size = avpkt->size;
211 AMRContext *s = avctx->priv_data;
212 const uint8_t*amrData=buf;
213 static const uint8_t block_size[16]={ 12, 13, 15, 17, 19, 20, 26, 31, 5, 0, 0, 0, 0, 0, 0, 0 };
214 enum Mode dec_mode;
215 int packet_size;
216
217 /* av_log(NULL,AV_LOG_DEBUG,"amr_decode_frame buf=%p buf_size=%d frameCount=%d!!\n",buf,buf_size,s->frameCount); */
218
219 dec_mode = (buf[0] >> 3) & 0x000F;
220 packet_size = block_size[dec_mode]+1;
221
222 if(packet_size > buf_size) {
223 av_log(avctx, AV_LOG_ERROR, "amr frame too short (%u, should be %u)\n", buf_size, packet_size);
224 return -1;
225 }
226
227 s->frameCount++;
228 /* av_log(NULL,AV_LOG_DEBUG,"packet_size=%d amrData= 0x%X %X %X %X\n",packet_size,amrData[0],amrData[1],amrData[2],amrData[3]); */
229 /* call decoder */
230 Decoder_Interface_Decode(s->decState, amrData, data, 0);
231 *data_size=160*2;
232
233 return packet_size;
234 }
235
236 static int amr_nb_encode_frame(AVCodecContext *avctx,
237 unsigned char *frame/*out*/, int buf_size, void *data/*in*/)
238 {
239 AMRContext *s = avctx->priv_data;
240 int written;
241
242 if((s->enc_bitrate=getBitrateMode(avctx->bit_rate))<0)
243 {
244 av_log(avctx, AV_LOG_ERROR, nb_bitrate_unsupported);
245 return -1;
246 }
247
248 written = Encoder_Interface_Encode(s->enstate,
249 s->enc_bitrate,
250 data,
251 frame,
252 0);
253 /* av_log(NULL,AV_LOG_DEBUG,"amr_nb_encode_frame encoded %u bytes, bitrate %u, first byte was %#02x\n",written, s->enc_bitrate, frame[0] ); */
254
255 return written;
256 }
257
258 AVCodec libamr_nb_decoder =
259 {
260 "libamr_nb",
261 CODEC_TYPE_AUDIO,
262 CODEC_ID_AMR_NB,
263 sizeof(AMRContext),
264 amr_nb_decode_init,
265 NULL,
266 amr_nb_decode_close,
267 amr_nb_decode_frame,
268 .long_name = NULL_IF_CONFIG_SMALL("libamr-nb Adaptive Multi-Rate (AMR) Narrow-Band"),
269 };
270
271 AVCodec libamr_nb_encoder =
272 {
273 "libamr_nb",
274 CODEC_TYPE_AUDIO,
275 CODEC_ID_AMR_NB,
276 sizeof(AMRContext),
277 amr_nb_encode_init,
278 amr_nb_encode_frame,
279 amr_nb_encode_close,
280 NULL,
281 .sample_fmts = (enum SampleFormat[]){SAMPLE_FMT_S16,SAMPLE_FMT_NONE},
282 .long_name = NULL_IF_CONFIG_SMALL("libamr-nb Adaptive Multi-Rate (AMR) Narrow-Band"),
283 };
284
285 #endif
286
287 /* -----------AMR wideband ------------*/
288 #if CONFIG_LIBAMR_WB
289
290 #ifdef _TYPEDEF_H
291 //To avoid duplicate typedefs from typedef in amr-nb
292 #define typedef_h
293 #endif
294
295 #include <amrwb/enc_if.h>
296 #include <amrwb/dec_if.h>
297 #include <amrwb/if_rom.h>
298
299 /* Common code for fixed and float version*/
300 typedef struct AMRWB_bitrates
301 {
302 int rate;
303 int mode;
304 } AMRWB_bitrates;
305
306 static int getWBBitrateMode(int bitrate)
307 {
308 /* make the correspondance between bitrate and mode */
309 AMRWB_bitrates rates[]={ {6600,0},
310 {8850,1},
311 {12650,2},
312 {14250,3},
313 {15850,4},
314 {18250,5},
315 {19850,6},
316 {23050,7},
317 {23850,8},
318 };
319 int i;
320
321 for(i=0;i<9;i++)
322 {
323 if(rates[i].rate==bitrate)
324 {
325 return rates[i].mode;
326 }
327 }
328 /* no bitrate matching, return an error */
329 return -1;
330 }
331
332
333 typedef struct AMRWBContext {
334 int frameCount;
335 void *state;
336 int mode;
337 Word16 allow_dtx;
338 } AMRWBContext;
339
340 static av_cold int amr_wb_encode_init(AVCodecContext * avctx)
341 {
342 AMRWBContext *s = avctx->priv_data;
343
344 s->frameCount=0;
345
346 if(avctx->sample_rate!=16000)
347 {
348 av_log(avctx, AV_LOG_ERROR, "Only 16000Hz sample rate supported\n");
349 return -1;
350 }
351
352 if(avctx->channels!=1)
353 {
354 av_log(avctx, AV_LOG_ERROR, "Only mono supported\n");
355 return -1;
356 }
357
358 if((s->mode=getWBBitrateMode(avctx->bit_rate))<0)
359 {
360 av_log(avctx, AV_LOG_ERROR, wb_bitrate_unsupported);
361 return -1;
362 }
363
364 avctx->frame_size=320;
365 avctx->coded_frame= avcodec_alloc_frame();
366
367 s->state = E_IF_init();
368 s->allow_dtx=0;
369
370 return 0;
371 }
372
373 static int amr_wb_encode_close(AVCodecContext * avctx)
374 {
375 AMRWBContext *s = avctx->priv_data;
376
377 E_IF_exit(s->state);
378 av_freep(&avctx->coded_frame);
379 s->frameCount++;
380 return 0;
381 }
382
383 static int amr_wb_encode_frame(AVCodecContext *avctx,
384 unsigned char *frame/*out*/, int buf_size, void *data/*in*/)
385 {
386 AMRWBContext *s = avctx->priv_data;
387 int size;
388
389 if((s->mode=getWBBitrateMode(avctx->bit_rate))<0)
390 {
391 av_log(avctx, AV_LOG_ERROR, wb_bitrate_unsupported);
392 return -1;
393 }
394 size = E_IF_encode(s->state, s->mode, data, frame, s->allow_dtx);
395 return size;
396 }
397
398 static av_cold int amr_wb_decode_init(AVCodecContext * avctx)
399 {
400 AMRWBContext *s = avctx->priv_data;
401
402 s->frameCount=0;
403 s->state = D_IF_init();
404
405 amr_decode_fix_avctx(avctx);
406
407 if(avctx->channels > 1)
408 {
409 av_log(avctx, AV_LOG_ERROR, "amr_wb: multichannel decoding not supported\n");
410 return -1;
411 }
412
413 return 0;
414 }
415
416 static int amr_wb_decode_frame(AVCodecContext * avctx,
417 void *data, int *data_size,
418 AVPacket *avpkt)
419 {
420 const uint8_t *buf = avpkt->data;
421 int buf_size = avpkt->size;
422 AMRWBContext *s = avctx->priv_data;
423 const uint8_t*amrData=buf;
424 int mode;
425 int packet_size;
426 static const uint8_t block_size[16] = {18, 23, 33, 37, 41, 47, 51, 59, 61, 6, 6, 0, 0, 0, 1, 1};
427
428 if(buf_size==0) {
429 /* nothing to do */
430 return 0;
431 }
432
433 mode = (amrData[0] >> 3) & 0x000F;
434 packet_size = block_size[mode];
435
436 if(packet_size > buf_size) {
437 av_log(avctx, AV_LOG_ERROR, "amr frame too short (%u, should be %u)\n", buf_size, packet_size+1);
438 return -1;
439 }
440
441 s->frameCount++;
442 D_IF_decode( s->state, amrData, data, _good_frame);
443 *data_size=320*2;
444 return packet_size;
445 }
446
447 static int amr_wb_decode_close(AVCodecContext * avctx)
448 {
449 AMRWBContext *s = avctx->priv_data;
450
451 D_IF_exit(s->state);
452 return 0;
453 }
454
455 AVCodec libamr_wb_decoder =
456 {
457 "libamr_wb",
458 CODEC_TYPE_AUDIO,
459 CODEC_ID_AMR_WB,
460 sizeof(AMRWBContext),
461 amr_wb_decode_init,
462 NULL,
463 amr_wb_decode_close,
464 amr_wb_decode_frame,
465 .long_name = NULL_IF_CONFIG_SMALL("libamr-wb Adaptive Multi-Rate (AMR) Wide-Band"),
466 };
467
468 AVCodec libamr_wb_encoder =
469 {
470 "libamr_wb",
471 CODEC_TYPE_AUDIO,
472 CODEC_ID_AMR_WB,
473 sizeof(AMRWBContext),
474 amr_wb_encode_init,
475 amr_wb_encode_frame,
476 amr_wb_encode_close,
477 NULL,
478 .sample_fmts = (enum SampleFormat[]){SAMPLE_FMT_S16,SAMPLE_FMT_NONE},
479 .long_name = NULL_IF_CONFIG_SMALL("libamr-wb Adaptive Multi-Rate (AMR) Wide-Band"),
480 };
481
482 #endif //CONFIG_LIBAMR_WB