| 1 | /* |
| 2 | * RIFF muxing functions |
| 3 | * Copyright (c) 2000 Fabrice Bellard |
| 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/dict.h" |
| 23 | #include "libavutil/log.h" |
| 24 | #include "libavutil/mathematics.h" |
| 25 | #include "libavcodec/avcodec.h" |
| 26 | #include "libavcodec/bytestream.h" |
| 27 | #include "avformat.h" |
| 28 | #include "avio_internal.h" |
| 29 | #include "riff.h" |
| 30 | |
| 31 | int64_t ff_start_tag(AVIOContext *pb, const char *tag) |
| 32 | { |
| 33 | ffio_wfourcc(pb, tag); |
| 34 | avio_wl32(pb, -1); |
| 35 | return avio_tell(pb); |
| 36 | } |
| 37 | |
| 38 | void ff_end_tag(AVIOContext *pb, int64_t start) |
| 39 | { |
| 40 | int64_t pos; |
| 41 | |
| 42 | av_assert0((start&1) == 0); |
| 43 | |
| 44 | pos = avio_tell(pb); |
| 45 | if (pos & 1) |
| 46 | avio_w8(pb, 0); |
| 47 | avio_seek(pb, start - 4, SEEK_SET); |
| 48 | avio_wl32(pb, (uint32_t)(pos - start)); |
| 49 | avio_seek(pb, FFALIGN(pos, 2), SEEK_SET); |
| 50 | } |
| 51 | |
| 52 | /* WAVEFORMATEX header */ |
| 53 | /* returns the size or -1 on error */ |
| 54 | int ff_put_wav_header(AVIOContext *pb, AVCodecContext *enc, int flags) |
| 55 | { |
| 56 | int bps, blkalign, bytespersec, frame_size; |
| 57 | int hdrsize; |
| 58 | int64_t hdrstart = avio_tell(pb); |
| 59 | int waveformatextensible; |
| 60 | uint8_t temp[256]; |
| 61 | uint8_t *riff_extradata = temp; |
| 62 | uint8_t *riff_extradata_start = temp; |
| 63 | |
| 64 | if (!enc->codec_tag || enc->codec_tag > 0xffff) |
| 65 | return -1; |
| 66 | |
| 67 | /* We use the known constant frame size for the codec if known, otherwise |
| 68 | * fall back on using AVCodecContext.frame_size, which is not as reliable |
| 69 | * for indicating packet duration. */ |
| 70 | frame_size = av_get_audio_frame_duration(enc, enc->block_align); |
| 71 | |
| 72 | waveformatextensible = (enc->channels > 2 && enc->channel_layout) || |
| 73 | enc->sample_rate > 48000 || |
| 74 | enc->codec_id == AV_CODEC_ID_EAC3 || |
| 75 | av_get_bits_per_sample(enc->codec_id) > 16; |
| 76 | |
| 77 | if (waveformatextensible) |
| 78 | avio_wl16(pb, 0xfffe); |
| 79 | else |
| 80 | avio_wl16(pb, enc->codec_tag); |
| 81 | |
| 82 | avio_wl16(pb, enc->channels); |
| 83 | avio_wl32(pb, enc->sample_rate); |
| 84 | if (enc->codec_id == AV_CODEC_ID_ATRAC3 || |
| 85 | enc->codec_id == AV_CODEC_ID_G723_1 || |
| 86 | enc->codec_id == AV_CODEC_ID_MP2 || |
| 87 | enc->codec_id == AV_CODEC_ID_MP3 || |
| 88 | enc->codec_id == AV_CODEC_ID_GSM_MS) { |
| 89 | bps = 0; |
| 90 | } else { |
| 91 | if (!(bps = av_get_bits_per_sample(enc->codec_id))) { |
| 92 | if (enc->bits_per_coded_sample) |
| 93 | bps = enc->bits_per_coded_sample; |
| 94 | else |
| 95 | bps = 16; // default to 16 |
| 96 | } |
| 97 | } |
| 98 | if (bps != enc->bits_per_coded_sample && enc->bits_per_coded_sample) { |
| 99 | av_log(enc, AV_LOG_WARNING, |
| 100 | "requested bits_per_coded_sample (%d) " |
| 101 | "and actually stored (%d) differ\n", |
| 102 | enc->bits_per_coded_sample, bps); |
| 103 | } |
| 104 | |
| 105 | if (enc->codec_id == AV_CODEC_ID_MP2) { |
| 106 | blkalign = frame_size; |
| 107 | } else if (enc->codec_id == AV_CODEC_ID_MP3) { |
| 108 | blkalign = 576 * (enc->sample_rate <= (24000 + 32000)/2 ? 1 : 2); |
| 109 | } else if (enc->codec_id == AV_CODEC_ID_AC3) { |
| 110 | blkalign = 3840; /* maximum bytes per frame */ |
| 111 | } else if (enc->codec_id == AV_CODEC_ID_AAC) { |
| 112 | blkalign = 768 * enc->channels; /* maximum bytes per frame */ |
| 113 | } else if (enc->codec_id == AV_CODEC_ID_G723_1) { |
| 114 | blkalign = 24; |
| 115 | } else if (enc->block_align != 0) { /* specified by the codec */ |
| 116 | blkalign = enc->block_align; |
| 117 | } else |
| 118 | blkalign = bps * enc->channels / av_gcd(8, bps); |
| 119 | if (enc->codec_id == AV_CODEC_ID_PCM_U8 || |
| 120 | enc->codec_id == AV_CODEC_ID_PCM_S24LE || |
| 121 | enc->codec_id == AV_CODEC_ID_PCM_S32LE || |
| 122 | enc->codec_id == AV_CODEC_ID_PCM_F32LE || |
| 123 | enc->codec_id == AV_CODEC_ID_PCM_F64LE || |
| 124 | enc->codec_id == AV_CODEC_ID_PCM_S16LE) { |
| 125 | bytespersec = enc->sample_rate * blkalign; |
| 126 | } else if (enc->codec_id == AV_CODEC_ID_G723_1) { |
| 127 | bytespersec = 800; |
| 128 | } else { |
| 129 | bytespersec = enc->bit_rate / 8; |
| 130 | } |
| 131 | avio_wl32(pb, bytespersec); /* bytes per second */ |
| 132 | avio_wl16(pb, blkalign); /* block align */ |
| 133 | avio_wl16(pb, bps); /* bits per sample */ |
| 134 | if (enc->codec_id == AV_CODEC_ID_MP3) { |
| 135 | bytestream_put_le16(&riff_extradata, 1); /* wID */ |
| 136 | bytestream_put_le32(&riff_extradata, 2); /* fdwFlags */ |
| 137 | bytestream_put_le16(&riff_extradata, 1152); /* nBlockSize */ |
| 138 | bytestream_put_le16(&riff_extradata, 1); /* nFramesPerBlock */ |
| 139 | bytestream_put_le16(&riff_extradata, 1393); /* nCodecDelay */ |
| 140 | } else if (enc->codec_id == AV_CODEC_ID_MP2) { |
| 141 | /* fwHeadLayer */ |
| 142 | bytestream_put_le16(&riff_extradata, 2); |
| 143 | /* dwHeadBitrate */ |
| 144 | bytestream_put_le32(&riff_extradata, enc->bit_rate); |
| 145 | /* fwHeadMode */ |
| 146 | bytestream_put_le16(&riff_extradata, enc->channels == 2 ? 1 : 8); |
| 147 | /* fwHeadModeExt */ |
| 148 | bytestream_put_le16(&riff_extradata, 0); |
| 149 | /* wHeadEmphasis */ |
| 150 | bytestream_put_le16(&riff_extradata, 1); |
| 151 | /* fwHeadFlags */ |
| 152 | bytestream_put_le16(&riff_extradata, 16); |
| 153 | /* dwPTSLow */ |
| 154 | bytestream_put_le32(&riff_extradata, 0); |
| 155 | /* dwPTSHigh */ |
| 156 | bytestream_put_le32(&riff_extradata, 0); |
| 157 | } else if (enc->codec_id == AV_CODEC_ID_G723_1) { |
| 158 | bytestream_put_le32(&riff_extradata, 0x9ace0002); /* extradata needed for msacm g723.1 codec */ |
| 159 | bytestream_put_le32(&riff_extradata, 0xaea2f732); |
| 160 | bytestream_put_le16(&riff_extradata, 0xacde); |
| 161 | } else if (enc->codec_id == AV_CODEC_ID_GSM_MS || |
| 162 | enc->codec_id == AV_CODEC_ID_ADPCM_IMA_WAV) { |
| 163 | /* wSamplesPerBlock */ |
| 164 | bytestream_put_le16(&riff_extradata, frame_size); |
| 165 | } else if (enc->extradata_size) { |
| 166 | riff_extradata_start = enc->extradata; |
| 167 | riff_extradata = enc->extradata + enc->extradata_size; |
| 168 | } |
| 169 | /* write WAVEFORMATEXTENSIBLE extensions */ |
| 170 | if (waveformatextensible) { |
| 171 | int write_channel_mask = enc->strict_std_compliance < FF_COMPLIANCE_NORMAL || |
| 172 | enc->channel_layout < 0x40000; |
| 173 | /* 22 is WAVEFORMATEXTENSIBLE size */ |
| 174 | avio_wl16(pb, riff_extradata - riff_extradata_start + 22); |
| 175 | /* ValidBitsPerSample || SamplesPerBlock || Reserved */ |
| 176 | avio_wl16(pb, bps); |
| 177 | /* dwChannelMask */ |
| 178 | avio_wl32(pb, write_channel_mask ? enc->channel_layout : 0); |
| 179 | /* GUID + next 3 */ |
| 180 | if (enc->codec_id == AV_CODEC_ID_EAC3) { |
| 181 | ff_put_guid(pb, get_codec_guid(enc->codec_id, ff_codec_wav_guids)); |
| 182 | } else { |
| 183 | avio_wl32(pb, enc->codec_tag); |
| 184 | avio_wl32(pb, 0x00100000); |
| 185 | avio_wl32(pb, 0xAA000080); |
| 186 | avio_wl32(pb, 0x719B3800); |
| 187 | } |
| 188 | } else if ((flags & FF_PUT_WAV_HEADER_FORCE_WAVEFORMATEX) || |
| 189 | enc->codec_tag != 0x0001 /* PCM */ || |
| 190 | riff_extradata - riff_extradata_start) { |
| 191 | /* WAVEFORMATEX */ |
| 192 | avio_wl16(pb, riff_extradata - riff_extradata_start); /* cbSize */ |
| 193 | } /* else PCMWAVEFORMAT */ |
| 194 | avio_write(pb, riff_extradata_start, riff_extradata - riff_extradata_start); |
| 195 | hdrsize = avio_tell(pb) - hdrstart; |
| 196 | if (hdrsize & 1) { |
| 197 | hdrsize++; |
| 198 | avio_w8(pb, 0); |
| 199 | } |
| 200 | |
| 201 | return hdrsize; |
| 202 | } |
| 203 | |
| 204 | /* BITMAPINFOHEADER header */ |
| 205 | void ff_put_bmp_header(AVIOContext *pb, AVCodecContext *enc, |
| 206 | const AVCodecTag *tags, int for_asf, int ignore_extradata) |
| 207 | { |
| 208 | int keep_height = enc->extradata_size >= 9 && |
| 209 | !memcmp(enc->extradata + enc->extradata_size - 9, "BottomUp", 9); |
| 210 | int extradata_size = enc->extradata_size - 9*keep_height; |
| 211 | |
| 212 | /* size */ |
| 213 | avio_wl32(pb, 40 + (ignore_extradata ? 0 :extradata_size)); |
| 214 | avio_wl32(pb, enc->width); |
| 215 | //We always store RGB TopDown |
| 216 | avio_wl32(pb, enc->codec_tag || keep_height ? enc->height : -enc->height); |
| 217 | /* planes */ |
| 218 | avio_wl16(pb, 1); |
| 219 | /* depth */ |
| 220 | avio_wl16(pb, enc->bits_per_coded_sample ? enc->bits_per_coded_sample : 24); |
| 221 | /* compression type */ |
| 222 | avio_wl32(pb, enc->codec_tag); |
| 223 | avio_wl32(pb, (enc->width * enc->height * (enc->bits_per_coded_sample ? enc->bits_per_coded_sample : 24)+7) / 8); |
| 224 | avio_wl32(pb, 0); |
| 225 | avio_wl32(pb, 0); |
| 226 | avio_wl32(pb, 0); |
| 227 | avio_wl32(pb, 0); |
| 228 | |
| 229 | if (!ignore_extradata) { |
| 230 | avio_write(pb, enc->extradata, extradata_size); |
| 231 | |
| 232 | if (!for_asf && extradata_size & 1) |
| 233 | avio_w8(pb, 0); |
| 234 | } |
| 235 | } |
| 236 | |
| 237 | void ff_parse_specific_params(AVStream *st, int *au_rate, |
| 238 | int *au_ssize, int *au_scale) |
| 239 | { |
| 240 | AVCodecContext *codec = st->codec; |
| 241 | int gcd; |
| 242 | int audio_frame_size; |
| 243 | |
| 244 | /* We use the known constant frame size for the codec if known, otherwise |
| 245 | * fall back on using AVCodecContext.frame_size, which is not as reliable |
| 246 | * for indicating packet duration. */ |
| 247 | audio_frame_size = av_get_audio_frame_duration(codec, 0); |
| 248 | if (!audio_frame_size) |
| 249 | audio_frame_size = codec->frame_size; |
| 250 | |
| 251 | *au_ssize = codec->block_align; |
| 252 | if (audio_frame_size && codec->sample_rate) { |
| 253 | *au_scale = audio_frame_size; |
| 254 | *au_rate = codec->sample_rate; |
| 255 | } else if (codec->codec_type == AVMEDIA_TYPE_VIDEO || |
| 256 | codec->codec_type == AVMEDIA_TYPE_DATA || |
| 257 | codec->codec_type == AVMEDIA_TYPE_SUBTITLE) { |
| 258 | *au_scale = st->time_base.num; |
| 259 | *au_rate = st->time_base.den; |
| 260 | } else { |
| 261 | *au_scale = codec->block_align ? codec->block_align * 8 : 8; |
| 262 | *au_rate = codec->bit_rate ? codec->bit_rate : |
| 263 | 8 * codec->sample_rate; |
| 264 | } |
| 265 | gcd = av_gcd(*au_scale, *au_rate); |
| 266 | *au_scale /= gcd; |
| 267 | *au_rate /= gcd; |
| 268 | } |
| 269 | |
| 270 | void ff_riff_write_info_tag(AVIOContext *pb, const char *tag, const char *str) |
| 271 | { |
| 272 | int len = strlen(str); |
| 273 | if (len > 0) { |
| 274 | len++; |
| 275 | ffio_wfourcc(pb, tag); |
| 276 | avio_wl32(pb, len); |
| 277 | avio_put_str(pb, str); |
| 278 | if (len & 1) |
| 279 | avio_w8(pb, 0); |
| 280 | } |
| 281 | } |
| 282 | |
| 283 | static const char riff_tags[][5] = { |
| 284 | "IARL", "IART", "ICMS", "ICMT", "ICOP", "ICRD", "ICRP", "IDIM", "IDPI", |
| 285 | "IENG", "IGNR", "IKEY", "ILGT", "ILNG", "IMED", "INAM", "IPLT", "IPRD", |
| 286 | "IPRT", "ITRK", "ISBJ", "ISFT", "ISHP", "ISMP", "ISRC", "ISRF", "ITCH", |
| 287 | { 0 } |
| 288 | }; |
| 289 | |
| 290 | static int riff_has_valid_tags(AVFormatContext *s) |
| 291 | { |
| 292 | int i; |
| 293 | |
| 294 | for (i = 0; *riff_tags[i]; i++) |
| 295 | if (av_dict_get(s->metadata, riff_tags[i], NULL, AV_DICT_MATCH_CASE)) |
| 296 | return 1; |
| 297 | |
| 298 | return 0; |
| 299 | } |
| 300 | |
| 301 | void ff_riff_write_info(AVFormatContext *s) |
| 302 | { |
| 303 | AVIOContext *pb = s->pb; |
| 304 | int i; |
| 305 | int64_t list_pos; |
| 306 | AVDictionaryEntry *t = NULL; |
| 307 | |
| 308 | ff_metadata_conv(&s->metadata, ff_riff_info_conv, NULL); |
| 309 | |
| 310 | /* writing empty LIST is not nice and may cause problems */ |
| 311 | if (!riff_has_valid_tags(s)) |
| 312 | return; |
| 313 | |
| 314 | list_pos = ff_start_tag(pb, "LIST"); |
| 315 | ffio_wfourcc(pb, "INFO"); |
| 316 | for (i = 0; *riff_tags[i]; i++) |
| 317 | if ((t = av_dict_get(s->metadata, riff_tags[i], |
| 318 | NULL, AV_DICT_MATCH_CASE))) |
| 319 | ff_riff_write_info_tag(s->pb, t->key, t->value); |
| 320 | ff_end_tag(pb, list_pos); |
| 321 | } |
| 322 | |
| 323 | void ff_put_guid(AVIOContext *s, const ff_asf_guid *g) |
| 324 | { |
| 325 | av_assert0(sizeof(*g) == 16); |
| 326 | avio_write(s, *g, sizeof(*g)); |
| 327 | } |
| 328 | |
| 329 | const ff_asf_guid *get_codec_guid(enum AVCodecID id, const AVCodecGuid *av_guid) |
| 330 | { |
| 331 | int i; |
| 332 | for (i = 0; av_guid[i].id != AV_CODEC_ID_NONE; i++) { |
| 333 | if (id == av_guid[i].id) |
| 334 | return &(av_guid[i].guid); |
| 335 | } |
| 336 | return NULL; |
| 337 | } |