1 /*
2 * AIFF/AIFF-C demuxer
3 * Copyright (c) 2006 Patrick Guimond
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 #include "libavutil/intreadwrite.h"
23 #include "libavutil/dict.h"
24 #include "avformat.h"
25 #include "internal.h"
26 #include "pcm.h"
27 #include "aiff.h"
28 #include "id3v2.h"
29 #include "mov_chan.h"
30 #include "replaygain.h"
31
32 #define AIFF 0
33 #define AIFF_C_VERSION1 0xA2805140
34
35 typedef struct AIFFInputContext {
36 int64_t data_end;
37 int block_duration;
38 } AIFFInputContext;
39
aiff_codec_get_id(int bps)40 static enum AVCodecID aiff_codec_get_id(int bps)
41 {
42 if (bps <= 8)
43 return AV_CODEC_ID_PCM_S8;
44 if (bps <= 16)
45 return AV_CODEC_ID_PCM_S16BE;
46 if (bps <= 24)
47 return AV_CODEC_ID_PCM_S24BE;
48 if (bps <= 32)
49 return AV_CODEC_ID_PCM_S32BE;
50
51 /* bigger than 32 isn't allowed */
52 return AV_CODEC_ID_NONE;
53 }
54
55 /* returns the size of the found tag */
get_tag(AVIOContext * pb,uint32_t * tag)56 static int get_tag(AVIOContext *pb, uint32_t * tag)
57 {
58 int size;
59
60 if (avio_feof(pb))
61 return AVERROR(EIO);
62
63 *tag = avio_rl32(pb);
64 size = avio_rb32(pb);
65
66 if (size < 0)
67 size = 0x7fffffff;
68
69 return size;
70 }
71
72 /* Metadata string read */
get_meta(AVFormatContext * s,const char * key,int size)73 static void get_meta(AVFormatContext *s, const char *key, int size)
74 {
75 uint8_t *str = av_malloc(size+1);
76
77 if (str) {
78 int res = avio_read(s->pb, str, size);
79 if (res < 0){
80 av_free(str);
81 return;
82 }
83 size -= res;
84 str[res] = 0;
85 av_dict_set(&s->metadata, key, str, AV_DICT_DONT_STRDUP_VAL);
86 }
87
88 avio_skip(s->pb, size);
89 }
90
91 /* Returns the number of sound data frames or negative on error */
get_aiff_header(AVFormatContext * s,int size,unsigned version)92 static int get_aiff_header(AVFormatContext *s, int size,
93 unsigned version)
94 {
95 AVIOContext *pb = s->pb;
96 AVCodecParameters *par = s->streams[0]->codecpar;
97 AIFFInputContext *aiff = s->priv_data;
98 int exp;
99 uint64_t val;
100 int sample_rate;
101 unsigned int num_frames;
102
103 if (size == INT_MAX)
104 return AVERROR_INVALIDDATA;
105
106 if (size & 1)
107 size++;
108 par->codec_type = AVMEDIA_TYPE_AUDIO;
109 par->channels = avio_rb16(pb);
110 num_frames = avio_rb32(pb);
111 par->bits_per_coded_sample = avio_rb16(pb);
112
113 exp = avio_rb16(pb) - 16383 - 63;
114 val = avio_rb64(pb);
115 if (exp <-63 || exp >63) {
116 av_log(s, AV_LOG_ERROR, "exp %d is out of range\n", exp);
117 return AVERROR_INVALIDDATA;
118 }
119 if (exp >= 0)
120 sample_rate = val << exp;
121 else
122 sample_rate = (val + (1ULL<<(-exp-1))) >> -exp;
123 par->sample_rate = sample_rate;
124 if (size < 18)
125 return AVERROR_INVALIDDATA;
126 size -= 18;
127
128 /* get codec id for AIFF-C */
129 if (size < 4) {
130 version = AIFF;
131 } else if (version == AIFF_C_VERSION1) {
132 par->codec_tag = avio_rl32(pb);
133 par->codec_id = ff_codec_get_id(ff_codec_aiff_tags, par->codec_tag);
134 if (par->codec_id == AV_CODEC_ID_NONE)
135 avpriv_request_sample(s, "unknown or unsupported codec tag: %s",
136 av_fourcc2str(par->codec_tag));
137 size -= 4;
138 }
139
140 if (version != AIFF_C_VERSION1 || par->codec_id == AV_CODEC_ID_PCM_S16BE) {
141 par->codec_id = aiff_codec_get_id(par->bits_per_coded_sample);
142 par->bits_per_coded_sample = av_get_bits_per_sample(par->codec_id);
143 aiff->block_duration = 1;
144 } else {
145 switch (par->codec_id) {
146 case AV_CODEC_ID_PCM_F32BE:
147 case AV_CODEC_ID_PCM_F64BE:
148 case AV_CODEC_ID_PCM_S16LE:
149 case AV_CODEC_ID_PCM_ALAW:
150 case AV_CODEC_ID_PCM_MULAW:
151 aiff->block_duration = 1;
152 break;
153 case AV_CODEC_ID_ADPCM_IMA_QT:
154 par->block_align = 34 * par->channels;
155 break;
156 case AV_CODEC_ID_MACE3:
157 par->block_align = 2 * par->channels;
158 break;
159 case AV_CODEC_ID_ADPCM_G726LE:
160 par->bits_per_coded_sample = 5;
161 case AV_CODEC_ID_ADPCM_IMA_WS:
162 case AV_CODEC_ID_ADPCM_G722:
163 case AV_CODEC_ID_MACE6:
164 case AV_CODEC_ID_SDX2_DPCM:
165 par->block_align = 1 * par->channels;
166 break;
167 case AV_CODEC_ID_GSM:
168 par->block_align = 33;
169 break;
170 default:
171 aiff->block_duration = 1;
172 break;
173 }
174 if (par->block_align > 0)
175 aiff->block_duration = av_get_audio_frame_duration2(par,
176 par->block_align);
177 }
178
179 /* Block align needs to be computed in all cases, as the definition
180 * is specific to applications -> here we use the WAVE format definition */
181 if (!par->block_align)
182 par->block_align = (av_get_bits_per_sample(par->codec_id) * par->channels) >> 3;
183
184 if (aiff->block_duration) {
185 par->bit_rate = (int64_t)par->sample_rate * (par->block_align << 3) /
186 aiff->block_duration;
187 }
188
189 /* Chunk is over */
190 if (size)
191 avio_skip(pb, size);
192
193 return num_frames;
194 }
195
aiff_probe(const AVProbeData * p)196 static int aiff_probe(const AVProbeData *p)
197 {
198 /* check file header */
199 if (p->buf[0] == 'F' && p->buf[1] == 'O' &&
200 p->buf[2] == 'R' && p->buf[3] == 'M' &&
201 p->buf[8] == 'A' && p->buf[9] == 'I' &&
202 p->buf[10] == 'F' && (p->buf[11] == 'F' || p->buf[11] == 'C'))
203 return AVPROBE_SCORE_MAX;
204 else
205 return 0;
206 }
207
208 /* aiff input */
aiff_read_header(AVFormatContext * s)209 static int aiff_read_header(AVFormatContext *s)
210 {
211 int ret, size, filesize;
212 int64_t offset = 0, position;
213 uint32_t tag;
214 unsigned version = AIFF_C_VERSION1;
215 AVIOContext *pb = s->pb;
216 AVStream * st;
217 AIFFInputContext *aiff = s->priv_data;
218 ID3v2ExtraMeta *id3v2_extra_meta = NULL;
219
220 /* check FORM header */
221 filesize = get_tag(pb, &tag);
222 if (filesize < 0 || tag != MKTAG('F', 'O', 'R', 'M'))
223 return AVERROR_INVALIDDATA;
224
225 /* AIFF data type */
226 tag = avio_rl32(pb);
227 if (tag == MKTAG('A', 'I', 'F', 'F')) /* Got an AIFF file */
228 version = AIFF;
229 else if (tag != MKTAG('A', 'I', 'F', 'C')) /* An AIFF-C file then */
230 return AVERROR_INVALIDDATA;
231
232 filesize -= 4;
233
234 st = avformat_new_stream(s, NULL);
235 if (!st)
236 return AVERROR(ENOMEM);
237
238 while (filesize > 0) {
239 /* parse different chunks */
240 size = get_tag(pb, &tag);
241
242 if (size == AVERROR_EOF && offset > 0 && st->codecpar->block_align) {
243 av_log(s, AV_LOG_WARNING, "header parser hit EOF\n");
244 goto got_sound;
245 }
246 if (size < 0)
247 return size;
248
249 if (size >= 0x7fffffff - 8)
250 filesize = 0;
251 else
252 filesize -= size + 8;
253
254 switch (tag) {
255 case MKTAG('C', 'O', 'M', 'M'): /* Common chunk */
256 /* Then for the complete header info */
257 st->nb_frames = get_aiff_header(s, size, version);
258 if (st->nb_frames < 0)
259 return st->nb_frames;
260 if (offset > 0) // COMM is after SSND
261 goto got_sound;
262 break;
263 case MKTAG('I', 'D', '3', ' '):
264 position = avio_tell(pb);
265 ff_id3v2_read(s, ID3v2_DEFAULT_MAGIC, &id3v2_extra_meta, size);
266 if (id3v2_extra_meta)
267 if ((ret = ff_id3v2_parse_apic(s, id3v2_extra_meta)) < 0 ||
268 (ret = ff_id3v2_parse_chapters(s, id3v2_extra_meta)) < 0) {
269 ff_id3v2_free_extra_meta(&id3v2_extra_meta);
270 return ret;
271 }
272 ff_id3v2_free_extra_meta(&id3v2_extra_meta);
273 if (position + size > avio_tell(pb))
274 avio_skip(pb, position + size - avio_tell(pb));
275 break;
276 case MKTAG('F', 'V', 'E', 'R'): /* Version chunk */
277 version = avio_rb32(pb);
278 break;
279 case MKTAG('N', 'A', 'M', 'E'): /* Sample name chunk */
280 get_meta(s, "title" , size);
281 break;
282 case MKTAG('A', 'U', 'T', 'H'): /* Author chunk */
283 get_meta(s, "author" , size);
284 break;
285 case MKTAG('(', 'c', ')', ' '): /* Copyright chunk */
286 get_meta(s, "copyright", size);
287 break;
288 case MKTAG('A', 'N', 'N', 'O'): /* Annotation chunk */
289 get_meta(s, "comment" , size);
290 break;
291 case MKTAG('S', 'S', 'N', 'D'): /* Sampled sound chunk */
292 if (size < 8)
293 return AVERROR_INVALIDDATA;
294 aiff->data_end = avio_tell(pb) + size;
295 offset = avio_rb32(pb); /* Offset of sound data */
296 avio_rb32(pb); /* BlockSize... don't care */
297 offset += avio_tell(pb); /* Compute absolute data offset */
298 if (st->codecpar->block_align && !(pb->seekable & AVIO_SEEKABLE_NORMAL)) /* Assume COMM already parsed */
299 goto got_sound;
300 if (!(pb->seekable & AVIO_SEEKABLE_NORMAL)) {
301 av_log(s, AV_LOG_ERROR, "file is not seekable\n");
302 return -1;
303 }
304 avio_skip(pb, size - 8);
305 break;
306 case MKTAG('w', 'a', 'v', 'e'):
307 if ((uint64_t)size > (1<<30))
308 return -1;
309 if ((ret = ff_get_extradata(s, st->codecpar, pb, size)) < 0)
310 return ret;
311 if ( (st->codecpar->codec_id == AV_CODEC_ID_QDMC || st->codecpar->codec_id == AV_CODEC_ID_QDM2)
312 && size>=12*4 && !st->codecpar->block_align) {
313 st->codecpar->block_align = AV_RB32(st->codecpar->extradata+11*4);
314 aiff->block_duration = AV_RB32(st->codecpar->extradata+9*4);
315 } else if (st->codecpar->codec_id == AV_CODEC_ID_QCELP) {
316 char rate = 0;
317 if (size >= 25)
318 rate = st->codecpar->extradata[24];
319 switch (rate) {
320 case 'H': // RATE_HALF
321 st->codecpar->block_align = 17;
322 break;
323 case 'F': // RATE_FULL
324 default:
325 st->codecpar->block_align = 35;
326 }
327 aiff->block_duration = 160;
328 st->codecpar->bit_rate = (int64_t)st->codecpar->sample_rate * (st->codecpar->block_align << 3) /
329 aiff->block_duration;
330 }
331 break;
332 case MKTAG('C','H','A','N'):
333 if ((ret = ff_mov_read_chan(s, pb, st, size)) < 0)
334 return ret;
335 break;
336 case MKTAG('A','P','C','M'): /* XA ADPCM compressed sound chunk */
337 st->codecpar->codec_id = AV_CODEC_ID_ADPCM_XA;
338 aiff->data_end = avio_tell(pb) + size;
339 offset = avio_tell(pb) + 8;
340 /* This field is unknown and its data seems to be irrelevant */
341 avio_rb32(pb);
342 st->codecpar->block_align = avio_rb32(pb);
343
344 goto got_sound;
345 break;
346 case 0:
347 if (offset > 0 && st->codecpar->block_align) // COMM && SSND
348 goto got_sound;
349 default: /* Jump */
350 avio_skip(pb, size);
351 }
352
353 /* Skip required padding byte for odd-sized chunks. */
354 if (size & 1) {
355 filesize--;
356 avio_skip(pb, 1);
357 }
358 }
359
360 ret = ff_replaygain_export(st, s->metadata);
361 if (ret < 0)
362 return ret;
363
364 got_sound:
365 if (!st->codecpar->block_align && st->codecpar->codec_id == AV_CODEC_ID_QCELP) {
366 av_log(s, AV_LOG_WARNING, "qcelp without wave chunk, assuming full rate\n");
367 st->codecpar->block_align = 35;
368 } else if (!st->codecpar->block_align) {
369 av_log(s, AV_LOG_ERROR, "could not find COMM tag or invalid block_align value\n");
370 return -1;
371 }
372
373 /* Now positioned, get the sound data start and end */
374 avpriv_set_pts_info(st, 64, 1, st->codecpar->sample_rate);
375 st->start_time = 0;
376 st->duration = st->nb_frames * aiff->block_duration;
377
378 /* Position the stream at the first block */
379 avio_seek(pb, offset, SEEK_SET);
380
381 return 0;
382 }
383
384 #define MAX_SIZE 4096
385
aiff_read_packet(AVFormatContext * s,AVPacket * pkt)386 static int aiff_read_packet(AVFormatContext *s,
387 AVPacket *pkt)
388 {
389 AVStream *st = s->streams[0];
390 AIFFInputContext *aiff = s->priv_data;
391 int64_t max_size;
392 int res, size;
393
394 /* calculate size of remaining data */
395 max_size = aiff->data_end - avio_tell(s->pb);
396 if (max_size <= 0)
397 return AVERROR_EOF;
398
399 if (!st->codecpar->block_align) {
400 av_log(s, AV_LOG_ERROR, "block_align not set\n");
401 return AVERROR_INVALIDDATA;
402 }
403
404 /* Now for that packet */
405 switch (st->codecpar->codec_id) {
406 case AV_CODEC_ID_ADPCM_IMA_QT:
407 case AV_CODEC_ID_GSM:
408 case AV_CODEC_ID_QDM2:
409 case AV_CODEC_ID_QCELP:
410 size = st->codecpar->block_align;
411 break;
412 default:
413 size = st->codecpar->block_align ? (MAX_SIZE / st->codecpar->block_align) * st->codecpar->block_align : MAX_SIZE;
414 if (!size)
415 return AVERROR_INVALIDDATA;
416 }
417 size = FFMIN(max_size, size);
418 res = av_get_packet(s->pb, pkt, size);
419 if (res < 0)
420 return res;
421
422 if (size >= st->codecpar->block_align)
423 pkt->flags &= ~AV_PKT_FLAG_CORRUPT;
424 /* Only one stream in an AIFF file */
425 pkt->stream_index = 0;
426 pkt->duration = (res / st->codecpar->block_align) * aiff->block_duration;
427 return 0;
428 }
429
430 AVInputFormat ff_aiff_demuxer = {
431 .name = "aiff",
432 .long_name = NULL_IF_CONFIG_SMALL("Audio IFF"),
433 .priv_data_size = sizeof(AIFFInputContext),
434 .read_probe = aiff_probe,
435 .read_header = aiff_read_header,
436 .read_packet = aiff_read_packet,
437 .read_seek = ff_pcm_read_seek,
438 .codec_tag = ff_aiff_codec_tags_list,
439 };
440