2 * WAV encoder and decoder
3 * Copyright (c) 2001, 2002 Fabrice Bellard.
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.
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.
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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
22 const CodecTag codec_wav_tags[] = {
23 { CODEC_ID_MP2, 0x50 },
24 { CODEC_ID_MP3, 0x55 },
25 { CODEC_ID_AC3, 0x2000 },
26 { CODEC_ID_DTS, 0x2001 },
27 { CODEC_ID_PCM_S16LE, 0x01 },
28 { CODEC_ID_PCM_U8, 0x01 }, /* must come after s16le in this list */
29 { CODEC_ID_PCM_S24LE, 0x01 },
30 { CODEC_ID_PCM_S32LE, 0x01 },
31 { CODEC_ID_PCM_ALAW, 0x06 },
32 { CODEC_ID_PCM_MULAW, 0x07 },
33 { CODEC_ID_ADPCM_MS, 0x02 },
34 { CODEC_ID_ADPCM_IMA_WAV, 0x11 },
35 { CODEC_ID_ADPCM_YAMAHA, 0x20 },
36 { CODEC_ID_ADPCM_G726, 0x45 },
37 { CODEC_ID_ADPCM_IMA_DK4, 0x61 }, /* rogue format number */
38 { CODEC_ID_ADPCM_IMA_DK3, 0x62 }, /* rogue format number */
39 { CODEC_ID_WMAV1, 0x160 },
40 { CODEC_ID_WMAV2, 0x161 },
41 { CODEC_ID_AAC, 0x706d },
42 { CODEC_ID_VORBIS, ('V'<<8)+'o' }, //HACK/FIXME, does vorbis in WAV/AVI have an (in)official id?
43 { CODEC_ID_SONIC, 0x2048 },
44 { CODEC_ID_SONIC_LS, 0x2048 },
45 { CODEC_ID_ADPCM_CT, 0x200 },
46 { CODEC_ID_ADPCM_SWF, ('S'<<8)+'F' },
47 { CODEC_ID_TRUESPEECH, 0x22 },
49 // for NuppelVideo (nuv.c)
50 { CODEC_ID_PCM_S16LE, MKTAG('R', 'A', 'W', 'A') },
51 { CODEC_ID_MP3, MKTAG('L', 'A', 'M', 'E') },
56 /* WAVEFORMATEX header */
57 /* returns the size or -1 on error */
58 int put_wav_header(ByteIOContext *pb, AVCodecContext *enc)
60 int bps, blkalign, bytespersec;
63 if(!enc->codec_tag || enc->codec_tag > 0xffff)
64 enc->codec_tag = codec_get_tag(codec_wav_tags, enc->codec_id);
68 put_le16(pb, enc->codec_tag);
69 put_le16(pb, enc->channels);
70 put_le32(pb, enc->sample_rate);
71 if (enc->codec_id == CODEC_ID_PCM_U8 ||
72 enc->codec_id == CODEC_ID_PCM_ALAW ||
73 enc->codec_id == CODEC_ID_PCM_MULAW) {
75 } else if (enc->codec_id == CODEC_ID_MP2 || enc->codec_id == CODEC_ID_MP3) {
77 } else if (enc->codec_id == CODEC_ID_ADPCM_IMA_WAV || enc->codec_id == CODEC_ID_ADPCM_MS || enc->codec_id == CODEC_ID_ADPCM_G726 || enc->codec_id == CODEC_ID_ADPCM_YAMAHA) { //
79 } else if (enc->codec_id == CODEC_ID_PCM_S24LE) {
81 } else if (enc->codec_id == CODEC_ID_PCM_S32LE) {
87 if (enc->codec_id == CODEC_ID_MP2 || enc->codec_id == CODEC_ID_MP3) {
88 blkalign = enc->frame_size; //this is wrong, but seems many demuxers dont work if this is set correctly
89 //blkalign = 144 * enc->bit_rate/enc->sample_rate;
90 } else if (enc->codec_id == CODEC_ID_ADPCM_G726) { //
92 } else if (enc->block_align != 0) { /* specified by the codec */
93 blkalign = enc->block_align;
95 blkalign = enc->channels*bps >> 3;
96 if (enc->codec_id == CODEC_ID_PCM_U8 ||
97 enc->codec_id == CODEC_ID_PCM_S24LE ||
98 enc->codec_id == CODEC_ID_PCM_S32LE ||
99 enc->codec_id == CODEC_ID_PCM_S16LE) {
100 bytespersec = enc->sample_rate * blkalign;
102 bytespersec = enc->bit_rate / 8;
104 put_le32(pb, bytespersec); /* bytes per second */
105 put_le16(pb, blkalign); /* block align */
106 put_le16(pb, bps); /* bits per sample */
107 if (enc->codec_id == CODEC_ID_MP3) {
108 put_le16(pb, 12); /* wav_extra_size */
110 put_le16(pb, 1); /* wID */
111 put_le32(pb, 2); /* fdwFlags */
112 put_le16(pb, 1152); /* nBlockSize */
113 put_le16(pb, 1); /* nFramesPerBlock */
114 put_le16(pb, 1393); /* nCodecDelay */
115 } else if (enc->codec_id == CODEC_ID_MP2) {
116 put_le16(pb, 22); /* wav_extra_size */
118 put_le16(pb, 2); /* fwHeadLayer */
119 put_le32(pb, enc->bit_rate); /* dwHeadBitrate */
120 put_le16(pb, enc->channels == 2 ? 1 : 8); /* fwHeadMode */
121 put_le16(pb, 0); /* fwHeadModeExt */
122 put_le16(pb, 1); /* wHeadEmphasis */
123 put_le16(pb, 16); /* fwHeadFlags */
124 put_le32(pb, 0); /* dwPTSLow */
125 put_le32(pb, 0); /* dwPTSHigh */
126 } else if (enc->codec_id == CODEC_ID_ADPCM_IMA_WAV) {
127 put_le16(pb, 2); /* wav_extra_size */
129 put_le16(pb, ((enc->block_align - 4 * enc->channels) / (4 * enc->channels)) * 8 + 1); /* wSamplesPerBlock */
130 } else if(enc->extradata_size){
131 put_le16(pb, enc->extradata_size);
132 put_buffer(pb, enc->extradata, enc->extradata_size);
133 hdrsize += enc->extradata_size;
144 #endif //CONFIG_MUXERS
146 /* We could be given one of the three possible structures here:
147 * WAVEFORMAT, PCMWAVEFORMAT or WAVEFORMATEX. Each structure
148 * is an expansion of the previous one with the fields added
149 * at the bottom. PCMWAVEFORMAT adds 'WORD wBitsPerSample' and
150 * WAVEFORMATEX adds 'WORD cbSize' and basically makes itself
151 * an openended structure.
153 void get_wav_header(ByteIOContext *pb, AVCodecContext *codec, int size)
158 codec->codec_type = CODEC_TYPE_AUDIO;
159 codec->codec_tag = id;
160 codec->channels = get_le16(pb);
161 codec->sample_rate = get_le32(pb);
162 codec->bit_rate = get_le32(pb) * 8;
163 codec->block_align = get_le16(pb);
164 if (size == 14) { /* We're dealing with plain vanilla WAVEFORMAT */
165 codec->bits_per_sample = 8;
167 codec->bits_per_sample = get_le16(pb);
168 codec->codec_id = wav_codec_get_id(id, codec->bits_per_sample);
170 if (size > 16) { /* We're obviously dealing with WAVEFORMATEX */
171 codec->extradata_size = get_le16(pb);
172 if (codec->extradata_size > 0) {
173 if (codec->extradata_size > size - 18)
174 codec->extradata_size = size - 18;
175 codec->extradata = av_mallocz(codec->extradata_size + FF_INPUT_BUFFER_PADDING_SIZE);
176 get_buffer(pb, codec->extradata, codec->extradata_size);
178 codec->extradata_size = 0;
180 /* It is possible for the chunk to contain garbage at the end */
181 if (size - codec->extradata_size - 18 > 0)
182 url_fskip(pb, size - codec->extradata_size - 18);
187 int wav_codec_get_id(unsigned int tag, int bps)
190 id = codec_get_id(codec_wav_tags, tag);
193 /* handle specific u8 codec */
194 if (id == CODEC_ID_PCM_S16LE && bps == 8)
195 id = CODEC_ID_PCM_U8;
196 if (id == CODEC_ID_PCM_S16LE && bps == 24)
197 id = CODEC_ID_PCM_S24LE;
198 if (id == CODEC_ID_PCM_S16LE && bps == 32)
199 id = CODEC_ID_PCM_S32LE;
209 static int wav_write_header(AVFormatContext *s)
211 WAVContext *wav = s->priv_data;
212 ByteIOContext *pb = &s->pb;
216 put_le32(pb, 0); /* file length */
220 fmt = start_tag(pb, "fmt ");
221 if (put_wav_header(pb, s->streams[0]->codec) < 0) {
227 av_set_pts_info(s->streams[0], 64, 1, s->streams[0]->codec->sample_rate);
230 wav->data = start_tag(pb, "data");
232 put_flush_packet(pb);
237 static int wav_write_packet(AVFormatContext *s, AVPacket *pkt)
239 ByteIOContext *pb = &s->pb;
240 put_buffer(pb, pkt->data, pkt->size);
244 static int wav_write_trailer(AVFormatContext *s)
246 ByteIOContext *pb = &s->pb;
247 WAVContext *wav = s->priv_data;
250 if (!url_is_streamed(&s->pb)) {
251 end_tag(pb, wav->data);
253 /* update file size */
254 file_size = url_ftell(pb);
255 url_fseek(pb, 4, SEEK_SET);
256 put_le32(pb, (uint32_t)(file_size - 8));
257 url_fseek(pb, file_size, SEEK_SET);
259 put_flush_packet(pb);
263 #endif //CONFIG_MUXERS
265 /* return the size of the found tag */
267 static int find_tag(ByteIOContext *pb, uint32_t tag1)
279 url_fseek(pb, size, SEEK_CUR);
286 static int wav_probe(AVProbeData *p)
288 /* check file header */
289 if (p->buf_size <= 32)
291 if (p->buf[0] == 'R' && p->buf[1] == 'I' &&
292 p->buf[2] == 'F' && p->buf[3] == 'F' &&
293 p->buf[8] == 'W' && p->buf[9] == 'A' &&
294 p->buf[10] == 'V' && p->buf[11] == 'E')
295 return AVPROBE_SCORE_MAX;
301 static int wav_read_header(AVFormatContext *s,
302 AVFormatParameters *ap)
306 ByteIOContext *pb = &s->pb;
308 WAVContext *wav = s->priv_data;
310 /* check RIFF header */
313 if (tag != MKTAG('R', 'I', 'F', 'F'))
315 get_le32(pb); /* file size */
317 if (tag != MKTAG('W', 'A', 'V', 'E'))
320 /* parse fmt header */
321 size = find_tag(pb, MKTAG('f', 'm', 't', ' '));
324 st = av_new_stream(s, 0);
326 return AVERROR_NOMEM;
328 get_wav_header(pb, st->codec, size);
329 st->need_parsing = 1;
331 av_set_pts_info(st, 64, 1, st->codec->sample_rate);
333 size = find_tag(pb, MKTAG('d', 'a', 't', 'a'));
336 wav->data_end= url_ftell(pb) + size;
340 #define MAX_SIZE 4096
342 static int wav_read_packet(AVFormatContext *s,
347 WAVContext *wav = s->priv_data;
349 if (url_feof(&s->pb))
353 left= wav->data_end - url_ftell(&s->pb);
355 left = find_tag(&(s->pb), MKTAG('d', 'a', 't', 'a'));
359 wav->data_end= url_ftell(&s->pb) + left;
363 if (st->codec->block_align > 1) {
364 if (size < st->codec->block_align)
365 size = st->codec->block_align;
366 size = (size / st->codec->block_align) * st->codec->block_align;
368 size= FFMIN(size, left);
369 if (av_new_packet(pkt, size))
371 pkt->stream_index = 0;
373 ret = get_buffer(&s->pb, pkt->data, pkt->size);
376 /* note: we need to modify the packet size here to handle the last
382 static int wav_read_close(AVFormatContext *s)
387 static int wav_read_seek(AVFormatContext *s,
388 int stream_index, int64_t timestamp, int flags)
393 switch(st->codec->codec_id) {
398 /* use generic seeking with dynamically generated indexes */
403 return pcm_read_seek(s, stream_index, timestamp, flags);
407 static AVInputFormat wav_iformat = {
419 static AVOutputFormat wav_oformat = {
431 #endif //CONFIG_MUXERS
433 int ff_wav_init(void)
435 av_register_input_format(&wav_iformat);
437 av_register_output_format(&wav_oformat);
438 #endif //CONFIG_MUXERS