d1f1bc65c96dd9429ec6df0b80f6d441a01f3cc4
[libav.git] / libavformat / wav.c
1 /*
2 * WAV encoder and decoder
3 * Copyright (c) 2001, 2002 Fabrice Bellard.
4 *
5 * This library is free software; you can redistribute it and/or
6 * modify it under the terms of the GNU Lesser General Public
7 * License as published by the Free Software Foundation; either
8 * version 2 of the License, or (at your option) any later version.
9 *
10 * This library is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * Lesser General Public License for more details.
14 *
15 * You should have received a copy of the GNU Lesser General Public
16 * License along with this library; if not, write to the Free Software
17 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
18 */
19 #include "avformat.h"
20 #include "avi.h"
21
22 const CodecTag codec_wav_tags[] = {
23 { CODEC_ID_MP2, 0x50 },
24 { CODEC_ID_MP3, 0x55 },
25 { CODEC_ID_AC3, 0x2000 },
26 { CODEC_ID_PCM_S16LE, 0x01 },
27 { CODEC_ID_PCM_U8, 0x01 }, /* must come after s16le in this list */
28 { CODEC_ID_PCM_ALAW, 0x06 },
29 { CODEC_ID_PCM_MULAW, 0x07 },
30 { CODEC_ID_ADPCM_MS, 0x02 },
31 { CODEC_ID_ADPCM_IMA_WAV, 0x11 },
32 { CODEC_ID_ADPCM_G726, 0x45 },
33 { CODEC_ID_ADPCM_IMA_DK4, 0x61 }, /* rogue format number */
34 { CODEC_ID_ADPCM_IMA_DK3, 0x62 }, /* rogue format number */
35 { CODEC_ID_WMAV1, 0x160 },
36 { CODEC_ID_WMAV2, 0x161 },
37 { CODEC_ID_AAC, 0x706d },
38 { CODEC_ID_VORBIS, ('V'<<8)+'o' }, //HACK/FIXME, does vorbis in WAV/AVI have an (in)official id?
39 { CODEC_ID_SONIC, 0x2048 },
40 { CODEC_ID_SONIC_LS, 0x2048 },
41 { CODEC_ID_ADPCM_CT, 0x200 },
42 { CODEC_ID_ADPCM_SWF, ('S'<<8)+'F' },
43 { 0, 0 },
44 };
45
46 #ifdef CONFIG_ENCODERS
47 /* WAVEFORMATEX header */
48 /* returns the size or -1 on error */
49 int put_wav_header(ByteIOContext *pb, AVCodecContext *enc)
50 {
51 int bps, blkalign, bytespersec;
52 int hdrsize = 18;
53
54 if(!enc->codec_tag)
55 enc->codec_tag = codec_get_tag(codec_wav_tags, enc->codec_id);
56 if(!enc->codec_tag)
57 return -1;
58
59 put_le16(pb, enc->codec_tag);
60 put_le16(pb, enc->channels);
61 put_le32(pb, enc->sample_rate);
62 if (enc->codec_id == CODEC_ID_PCM_U8 ||
63 enc->codec_id == CODEC_ID_PCM_ALAW ||
64 enc->codec_id == CODEC_ID_PCM_MULAW) {
65 bps = 8;
66 } else if (enc->codec_id == CODEC_ID_MP2 || enc->codec_id == CODEC_ID_MP3) {
67 bps = 0;
68 } else if (enc->codec_id == CODEC_ID_ADPCM_IMA_WAV || enc->codec_id == CODEC_ID_ADPCM_MS || enc->codec_id == CODEC_ID_ADPCM_G726) { //
69 bps = 4;
70 } else {
71 bps = 16;
72 }
73
74 if (enc->codec_id == CODEC_ID_MP2 || enc->codec_id == CODEC_ID_MP3) {
75 blkalign = enc->frame_size; //this is wrong, but seems many demuxers dont work if this is set correctly
76 //blkalign = 144 * enc->bit_rate/enc->sample_rate;
77 } else if (enc->codec_id == CODEC_ID_ADPCM_G726) { //
78 blkalign = 1;
79 } else if (enc->block_align != 0) { /* specified by the codec */
80 blkalign = enc->block_align;
81 } else
82 blkalign = enc->channels*bps >> 3;
83 if (enc->codec_id == CODEC_ID_PCM_U8 ||
84 enc->codec_id == CODEC_ID_PCM_S16LE) {
85 bytespersec = enc->sample_rate * blkalign;
86 } else {
87 bytespersec = enc->bit_rate / 8;
88 }
89 put_le32(pb, bytespersec); /* bytes per second */
90 put_le16(pb, blkalign); /* block align */
91 put_le16(pb, bps); /* bits per sample */
92 if (enc->codec_id == CODEC_ID_MP3) {
93 put_le16(pb, 12); /* wav_extra_size */
94 hdrsize += 12;
95 put_le16(pb, 1); /* wID */
96 put_le32(pb, 2); /* fdwFlags */
97 put_le16(pb, 1152); /* nBlockSize */
98 put_le16(pb, 1); /* nFramesPerBlock */
99 put_le16(pb, 1393); /* nCodecDelay */
100 } else if (enc->codec_id == CODEC_ID_MP2) {
101 put_le16(pb, 22); /* wav_extra_size */
102 hdrsize += 22;
103 put_le16(pb, 2); /* fwHeadLayer */
104 put_le32(pb, enc->bit_rate); /* dwHeadBitrate */
105 put_le16(pb, enc->channels == 2 ? 1 : 8); /* fwHeadMode */
106 put_le16(pb, 0); /* fwHeadModeExt */
107 put_le16(pb, 1); /* wHeadEmphasis */
108 put_le16(pb, 16); /* fwHeadFlags */
109 put_le32(pb, 0); /* dwPTSLow */
110 put_le32(pb, 0); /* dwPTSHigh */
111 } else if (enc->codec_id == CODEC_ID_ADPCM_IMA_WAV) {
112 put_le16(pb, 2); /* wav_extra_size */
113 hdrsize += 2;
114 put_le16(pb, ((enc->block_align - 4 * enc->channels) / (4 * enc->channels)) * 8 + 1); /* wSamplesPerBlock */
115 } else if(enc->extradata_size){
116 put_le16(pb, enc->extradata_size);
117 put_buffer(pb, enc->extradata, enc->extradata_size);
118 hdrsize += enc->extradata_size;
119 if(hdrsize&1){
120 hdrsize++;
121 put_byte(pb, 0);
122 }
123 } else {
124 hdrsize -= 2;
125 }
126
127 return hdrsize;
128 }
129 #endif //CONFIG_ENCODERS
130
131 /* We could be given one of the three possible structures here:
132 * WAVEFORMAT, PCMWAVEFORMAT or WAVEFORMATEX. Each structure
133 * is an expansion of the previous one with the fields added
134 * at the bottom. PCMWAVEFORMAT adds 'WORD wBitsPerSample' and
135 * WAVEFORMATEX adds 'WORD cbSize' and basically makes itself
136 * an openended structure.
137 */
138 void get_wav_header(ByteIOContext *pb, AVCodecContext *codec, int size)
139 {
140 int id;
141
142 id = get_le16(pb);
143 codec->codec_type = CODEC_TYPE_AUDIO;
144 codec->codec_tag = id;
145 codec->channels = get_le16(pb);
146 codec->sample_rate = get_le32(pb);
147 codec->bit_rate = get_le32(pb) * 8;
148 codec->block_align = get_le16(pb);
149 if (size == 14) { /* We're dealing with plain vanilla WAVEFORMAT */
150 codec->bits_per_sample = 8;
151 }else
152 codec->bits_per_sample = get_le16(pb);
153 codec->codec_id = wav_codec_get_id(id, codec->bits_per_sample);
154
155 if (size > 16) { /* We're obviously dealing with WAVEFORMATEX */
156 codec->extradata_size = get_le16(pb);
157 if (codec->extradata_size > 0) {
158 if (codec->extradata_size > size - 18)
159 codec->extradata_size = size - 18;
160 codec->extradata = av_mallocz(codec->extradata_size + FF_INPUT_BUFFER_PADDING_SIZE);
161 get_buffer(pb, codec->extradata, codec->extradata_size);
162 } else
163 codec->extradata_size = 0;
164
165 /* It is possible for the chunk to contain garbage at the end */
166 if (size - codec->extradata_size - 18 > 0)
167 url_fskip(pb, size - codec->extradata_size - 18);
168 }
169 }
170
171
172 int wav_codec_get_id(unsigned int tag, int bps)
173 {
174 int id;
175 id = codec_get_id(codec_wav_tags, tag);
176 if (id <= 0)
177 return id;
178 /* handle specific u8 codec */
179 if (id == CODEC_ID_PCM_S16LE && bps == 8)
180 id = CODEC_ID_PCM_U8;
181 return id;
182 }
183
184 #ifdef CONFIG_ENCODERS
185 typedef struct {
186 offset_t data;
187 } WAVContext;
188
189 static int wav_write_header(AVFormatContext *s)
190 {
191 WAVContext *wav = s->priv_data;
192 ByteIOContext *pb = &s->pb;
193 offset_t fmt;
194
195 put_tag(pb, "RIFF");
196 put_le32(pb, 0); /* file length */
197 put_tag(pb, "WAVE");
198
199 /* format header */
200 fmt = start_tag(pb, "fmt ");
201 if (put_wav_header(pb, &s->streams[0]->codec) < 0) {
202 av_free(wav);
203 return -1;
204 }
205 end_tag(pb, fmt);
206
207 av_set_pts_info(s->streams[0], 64, 1, s->streams[0]->codec.sample_rate);
208
209 /* data header */
210 wav->data = start_tag(pb, "data");
211
212 put_flush_packet(pb);
213
214 return 0;
215 }
216
217 static int wav_write_packet(AVFormatContext *s, AVPacket *pkt)
218 {
219 ByteIOContext *pb = &s->pb;
220 put_buffer(pb, pkt->data, pkt->size);
221 return 0;
222 }
223
224 static int wav_write_trailer(AVFormatContext *s)
225 {
226 ByteIOContext *pb = &s->pb;
227 WAVContext *wav = s->priv_data;
228 offset_t file_size;
229
230 if (!url_is_streamed(&s->pb)) {
231 end_tag(pb, wav->data);
232
233 /* update file size */
234 file_size = url_ftell(pb);
235 url_fseek(pb, 4, SEEK_SET);
236 put_le32(pb, (uint32_t)(file_size - 8));
237 url_fseek(pb, file_size, SEEK_SET);
238
239 put_flush_packet(pb);
240 }
241 return 0;
242 }
243 #endif //CONFIG_ENCODERS
244
245 /* return the size of the found tag */
246 /* XXX: > 2GB ? */
247 static int find_tag(ByteIOContext *pb, uint32_t tag1)
248 {
249 unsigned int tag;
250 int size;
251
252 for(;;) {
253 if (url_feof(pb))
254 return -1;
255 tag = get_le32(pb);
256 size = get_le32(pb);
257 if (tag == tag1)
258 break;
259 url_fseek(pb, size, SEEK_CUR);
260 }
261 if (size < 0)
262 size = 0x7fffffff;
263 return size;
264 }
265
266 static int wav_probe(AVProbeData *p)
267 {
268 /* check file header */
269 if (p->buf_size <= 32)
270 return 0;
271 if (p->buf[0] == 'R' && p->buf[1] == 'I' &&
272 p->buf[2] == 'F' && p->buf[3] == 'F' &&
273 p->buf[8] == 'W' && p->buf[9] == 'A' &&
274 p->buf[10] == 'V' && p->buf[11] == 'E')
275 return AVPROBE_SCORE_MAX;
276 else
277 return 0;
278 }
279
280 /* wav input */
281 static int wav_read_header(AVFormatContext *s,
282 AVFormatParameters *ap)
283 {
284 int size;
285 unsigned int tag;
286 ByteIOContext *pb = &s->pb;
287 AVStream *st;
288
289 /* check RIFF header */
290 tag = get_le32(pb);
291
292 if (tag != MKTAG('R', 'I', 'F', 'F'))
293 return -1;
294 get_le32(pb); /* file size */
295 tag = get_le32(pb);
296 if (tag != MKTAG('W', 'A', 'V', 'E'))
297 return -1;
298
299 /* parse fmt header */
300 size = find_tag(pb, MKTAG('f', 'm', 't', ' '));
301 if (size < 0)
302 return -1;
303 st = av_new_stream(s, 0);
304 if (!st)
305 return AVERROR_NOMEM;
306
307 get_wav_header(pb, &st->codec, size);
308 st->need_parsing = 1;
309
310 av_set_pts_info(st, 64, 1, st->codec.sample_rate);
311
312 size = find_tag(pb, MKTAG('d', 'a', 't', 'a'));
313 if (size < 0)
314 return -1;
315 return 0;
316 }
317
318 #define MAX_SIZE 4096
319
320 static int wav_read_packet(AVFormatContext *s,
321 AVPacket *pkt)
322 {
323 int ret, size;
324 AVStream *st;
325
326 if (url_feof(&s->pb))
327 return AVERROR_IO;
328 st = s->streams[0];
329
330 size = MAX_SIZE;
331 if (st->codec.block_align > 1) {
332 if (size < st->codec.block_align)
333 size = st->codec.block_align;
334 size = (size / st->codec.block_align) * st->codec.block_align;
335 }
336 if (av_new_packet(pkt, size))
337 return AVERROR_IO;
338 pkt->stream_index = 0;
339
340 ret = get_buffer(&s->pb, pkt->data, pkt->size);
341 if (ret < 0)
342 av_free_packet(pkt);
343 /* note: we need to modify the packet size here to handle the last
344 packet */
345 pkt->size = ret;
346 return ret;
347 }
348
349 static int wav_read_close(AVFormatContext *s)
350 {
351 return 0;
352 }
353
354 static int wav_read_seek(AVFormatContext *s,
355 int stream_index, int64_t timestamp, int flags)
356 {
357 AVStream *st;
358
359 st = s->streams[0];
360 switch(st->codec.codec_id) {
361 case CODEC_ID_MP2:
362 case CODEC_ID_MP3:
363 case CODEC_ID_AC3:
364 case CODEC_ID_DTS:
365 /* use generic seeking with dynamically generated indexes */
366 return -1;
367 default:
368 break;
369 }
370 return pcm_read_seek(s, stream_index, timestamp, flags);
371 }
372
373
374 static AVInputFormat wav_iformat = {
375 "wav",
376 "wav format",
377 0,
378 wav_probe,
379 wav_read_header,
380 wav_read_packet,
381 wav_read_close,
382 wav_read_seek,
383 };
384
385 #ifdef CONFIG_ENCODERS
386 static AVOutputFormat wav_oformat = {
387 "wav",
388 "wav format",
389 "audio/x-wav",
390 "wav",
391 sizeof(WAVContext),
392 CODEC_ID_PCM_S16LE,
393 CODEC_ID_NONE,
394 wav_write_header,
395 wav_write_packet,
396 wav_write_trailer,
397 };
398 #endif //CONFIG_ENCODERS
399
400 int ff_wav_init(void)
401 {
402 av_register_input_format(&wav_iformat);
403 #ifdef CONFIG_ENCODERS
404 av_register_output_format(&wav_oformat);
405 #endif //CONFIG_ENCODERS
406 return 0;
407 }