| 1 | /* |
| 2 | * Copyright (c) 2012 Stefano Sabatini |
| 3 | * |
| 4 | * Permission is hereby granted, free of charge, to any person obtaining a copy |
| 5 | * of this software and associated documentation files (the "Software"), to deal |
| 6 | * in the Software without restriction, including without limitation the rights |
| 7 | * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell |
| 8 | * copies of the Software, and to permit persons to whom the Software is |
| 9 | * furnished to do so, subject to the following conditions: |
| 10 | * |
| 11 | * The above copyright notice and this permission notice shall be included in |
| 12 | * all copies or substantial portions of the Software. |
| 13 | * |
| 14 | * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
| 15 | * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
| 16 | * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL |
| 17 | * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
| 18 | * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
| 19 | * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN |
| 20 | * THE SOFTWARE. |
| 21 | */ |
| 22 | |
| 23 | /** |
| 24 | * @file |
| 25 | * Demuxing and decoding example. |
| 26 | * |
| 27 | * Show how to use the libavformat and libavcodec API to demux and |
| 28 | * decode audio and video data. |
| 29 | * @example demuxing_decoding.c |
| 30 | */ |
| 31 | |
| 32 | #include <libavutil/imgutils.h> |
| 33 | #include <libavutil/samplefmt.h> |
| 34 | #include <libavutil/timestamp.h> |
| 35 | #include <libavformat/avformat.h> |
| 36 | |
| 37 | static AVFormatContext *fmt_ctx = NULL; |
| 38 | static AVCodecContext *video_dec_ctx = NULL, *audio_dec_ctx; |
| 39 | static AVStream *video_stream = NULL, *audio_stream = NULL; |
| 40 | static const char *src_filename = NULL; |
| 41 | static const char *video_dst_filename = NULL; |
| 42 | static const char *audio_dst_filename = NULL; |
| 43 | static FILE *video_dst_file = NULL; |
| 44 | static FILE *audio_dst_file = NULL; |
| 45 | |
| 46 | static uint8_t *video_dst_data[4] = {NULL}; |
| 47 | static int video_dst_linesize[4]; |
| 48 | static int video_dst_bufsize; |
| 49 | |
| 50 | static int video_stream_idx = -1, audio_stream_idx = -1; |
| 51 | static AVFrame *frame = NULL; |
| 52 | static AVPacket pkt; |
| 53 | static int video_frame_count = 0; |
| 54 | static int audio_frame_count = 0; |
| 55 | |
| 56 | /* The different ways of decoding and managing data memory. You are not |
| 57 | * supposed to support all the modes in your application but pick the one most |
| 58 | * appropriate to your needs. Look for the use of api_mode in this example to |
| 59 | * see what are the differences of API usage between them */ |
| 60 | enum { |
| 61 | API_MODE_OLD = 0, /* old method, deprecated */ |
| 62 | API_MODE_NEW_API_REF_COUNT = 1, /* new method, using the frame reference counting */ |
| 63 | API_MODE_NEW_API_NO_REF_COUNT = 2, /* new method, without reference counting */ |
| 64 | }; |
| 65 | |
| 66 | static int api_mode = API_MODE_OLD; |
| 67 | |
| 68 | static int decode_packet(int *got_frame, int cached) |
| 69 | { |
| 70 | int ret = 0; |
| 71 | int decoded = pkt.size; |
| 72 | |
| 73 | *got_frame = 0; |
| 74 | |
| 75 | if (pkt.stream_index == video_stream_idx) { |
| 76 | /* decode video frame */ |
| 77 | ret = avcodec_decode_video2(video_dec_ctx, frame, got_frame, &pkt); |
| 78 | if (ret < 0) { |
| 79 | fprintf(stderr, "Error decoding video frame (%s)\n", av_err2str(ret)); |
| 80 | return ret; |
| 81 | } |
| 82 | |
| 83 | if (*got_frame) { |
| 84 | printf("video_frame%s n:%d coded_n:%d pts:%s\n", |
| 85 | cached ? "(cached)" : "", |
| 86 | video_frame_count++, frame->coded_picture_number, |
| 87 | av_ts2timestr(frame->pts, &video_dec_ctx->time_base)); |
| 88 | |
| 89 | /* copy decoded frame to destination buffer: |
| 90 | * this is required since rawvideo expects non aligned data */ |
| 91 | av_image_copy(video_dst_data, video_dst_linesize, |
| 92 | (const uint8_t **)(frame->data), frame->linesize, |
| 93 | video_dec_ctx->pix_fmt, video_dec_ctx->width, video_dec_ctx->height); |
| 94 | |
| 95 | /* write to rawvideo file */ |
| 96 | fwrite(video_dst_data[0], 1, video_dst_bufsize, video_dst_file); |
| 97 | } |
| 98 | } else if (pkt.stream_index == audio_stream_idx) { |
| 99 | /* decode audio frame */ |
| 100 | ret = avcodec_decode_audio4(audio_dec_ctx, frame, got_frame, &pkt); |
| 101 | if (ret < 0) { |
| 102 | fprintf(stderr, "Error decoding audio frame (%s)\n", av_err2str(ret)); |
| 103 | return ret; |
| 104 | } |
| 105 | /* Some audio decoders decode only part of the packet, and have to be |
| 106 | * called again with the remainder of the packet data. |
| 107 | * Sample: fate-suite/lossless-audio/luckynight-partial.shn |
| 108 | * Also, some decoders might over-read the packet. */ |
| 109 | decoded = FFMIN(ret, pkt.size); |
| 110 | |
| 111 | if (*got_frame) { |
| 112 | size_t unpadded_linesize = frame->nb_samples * av_get_bytes_per_sample(frame->format); |
| 113 | printf("audio_frame%s n:%d nb_samples:%d pts:%s\n", |
| 114 | cached ? "(cached)" : "", |
| 115 | audio_frame_count++, frame->nb_samples, |
| 116 | av_ts2timestr(frame->pts, &audio_dec_ctx->time_base)); |
| 117 | |
| 118 | /* Write the raw audio data samples of the first plane. This works |
| 119 | * fine for packed formats (e.g. AV_SAMPLE_FMT_S16). However, |
| 120 | * most audio decoders output planar audio, which uses a separate |
| 121 | * plane of audio samples for each channel (e.g. AV_SAMPLE_FMT_S16P). |
| 122 | * In other words, this code will write only the first audio channel |
| 123 | * in these cases. |
| 124 | * You should use libswresample or libavfilter to convert the frame |
| 125 | * to packed data. */ |
| 126 | fwrite(frame->extended_data[0], 1, unpadded_linesize, audio_dst_file); |
| 127 | } |
| 128 | } |
| 129 | |
| 130 | /* If we use the new API with reference counting, we own the data and need |
| 131 | * to de-reference it when we don't use it anymore */ |
| 132 | if (*got_frame && api_mode == API_MODE_NEW_API_REF_COUNT) |
| 133 | av_frame_unref(frame); |
| 134 | |
| 135 | return decoded; |
| 136 | } |
| 137 | |
| 138 | static int open_codec_context(int *stream_idx, |
| 139 | AVFormatContext *fmt_ctx, enum AVMediaType type) |
| 140 | { |
| 141 | int ret; |
| 142 | AVStream *st; |
| 143 | AVCodecContext *dec_ctx = NULL; |
| 144 | AVCodec *dec = NULL; |
| 145 | AVDictionary *opts = NULL; |
| 146 | |
| 147 | ret = av_find_best_stream(fmt_ctx, type, -1, -1, NULL, 0); |
| 148 | if (ret < 0) { |
| 149 | fprintf(stderr, "Could not find %s stream in input file '%s'\n", |
| 150 | av_get_media_type_string(type), src_filename); |
| 151 | return ret; |
| 152 | } else { |
| 153 | *stream_idx = ret; |
| 154 | st = fmt_ctx->streams[*stream_idx]; |
| 155 | |
| 156 | /* find decoder for the stream */ |
| 157 | dec_ctx = st->codec; |
| 158 | dec = avcodec_find_decoder(dec_ctx->codec_id); |
| 159 | if (!dec) { |
| 160 | fprintf(stderr, "Failed to find %s codec\n", |
| 161 | av_get_media_type_string(type)); |
| 162 | return AVERROR(EINVAL); |
| 163 | } |
| 164 | |
| 165 | /* Init the decoders, with or without reference counting */ |
| 166 | if (api_mode == API_MODE_NEW_API_REF_COUNT) |
| 167 | av_dict_set(&opts, "refcounted_frames", "1", 0); |
| 168 | if ((ret = avcodec_open2(dec_ctx, dec, &opts)) < 0) { |
| 169 | fprintf(stderr, "Failed to open %s codec\n", |
| 170 | av_get_media_type_string(type)); |
| 171 | return ret; |
| 172 | } |
| 173 | } |
| 174 | |
| 175 | return 0; |
| 176 | } |
| 177 | |
| 178 | static int get_format_from_sample_fmt(const char **fmt, |
| 179 | enum AVSampleFormat sample_fmt) |
| 180 | { |
| 181 | int i; |
| 182 | struct sample_fmt_entry { |
| 183 | enum AVSampleFormat sample_fmt; const char *fmt_be, *fmt_le; |
| 184 | } sample_fmt_entries[] = { |
| 185 | { AV_SAMPLE_FMT_U8, "u8", "u8" }, |
| 186 | { AV_SAMPLE_FMT_S16, "s16be", "s16le" }, |
| 187 | { AV_SAMPLE_FMT_S32, "s32be", "s32le" }, |
| 188 | { AV_SAMPLE_FMT_FLT, "f32be", "f32le" }, |
| 189 | { AV_SAMPLE_FMT_DBL, "f64be", "f64le" }, |
| 190 | }; |
| 191 | *fmt = NULL; |
| 192 | |
| 193 | for (i = 0; i < FF_ARRAY_ELEMS(sample_fmt_entries); i++) { |
| 194 | struct sample_fmt_entry *entry = &sample_fmt_entries[i]; |
| 195 | if (sample_fmt == entry->sample_fmt) { |
| 196 | *fmt = AV_NE(entry->fmt_be, entry->fmt_le); |
| 197 | return 0; |
| 198 | } |
| 199 | } |
| 200 | |
| 201 | fprintf(stderr, |
| 202 | "sample format %s is not supported as output format\n", |
| 203 | av_get_sample_fmt_name(sample_fmt)); |
| 204 | return -1; |
| 205 | } |
| 206 | |
| 207 | int main (int argc, char **argv) |
| 208 | { |
| 209 | int ret = 0, got_frame; |
| 210 | |
| 211 | if (argc != 4 && argc != 5) { |
| 212 | fprintf(stderr, "usage: %s [-refcount=<old|new_norefcount|new_refcount>] " |
| 213 | "input_file video_output_file audio_output_file\n" |
| 214 | "API example program to show how to read frames from an input file.\n" |
| 215 | "This program reads frames from a file, decodes them, and writes decoded\n" |
| 216 | "video frames to a rawvideo file named video_output_file, and decoded\n" |
| 217 | "audio frames to a rawaudio file named audio_output_file.\n\n" |
| 218 | "If the -refcount option is specified, the program use the\n" |
| 219 | "reference counting frame system which allows keeping a copy of\n" |
| 220 | "the data for longer than one decode call. If unset, it's using\n" |
| 221 | "the classic old method.\n" |
| 222 | "\n", argv[0]); |
| 223 | exit(1); |
| 224 | } |
| 225 | if (argc == 5) { |
| 226 | const char *mode = argv[1] + strlen("-refcount="); |
| 227 | if (!strcmp(mode, "old")) api_mode = API_MODE_OLD; |
| 228 | else if (!strcmp(mode, "new_norefcount")) api_mode = API_MODE_NEW_API_NO_REF_COUNT; |
| 229 | else if (!strcmp(mode, "new_refcount")) api_mode = API_MODE_NEW_API_REF_COUNT; |
| 230 | else { |
| 231 | fprintf(stderr, "unknow mode '%s'\n", mode); |
| 232 | exit(1); |
| 233 | } |
| 234 | argv++; |
| 235 | } |
| 236 | src_filename = argv[1]; |
| 237 | video_dst_filename = argv[2]; |
| 238 | audio_dst_filename = argv[3]; |
| 239 | |
| 240 | /* register all formats and codecs */ |
| 241 | av_register_all(); |
| 242 | |
| 243 | /* open input file, and allocate format context */ |
| 244 | if (avformat_open_input(&fmt_ctx, src_filename, NULL, NULL) < 0) { |
| 245 | fprintf(stderr, "Could not open source file %s\n", src_filename); |
| 246 | exit(1); |
| 247 | } |
| 248 | |
| 249 | /* retrieve stream information */ |
| 250 | if (avformat_find_stream_info(fmt_ctx, NULL) < 0) { |
| 251 | fprintf(stderr, "Could not find stream information\n"); |
| 252 | exit(1); |
| 253 | } |
| 254 | |
| 255 | if (open_codec_context(&video_stream_idx, fmt_ctx, AVMEDIA_TYPE_VIDEO) >= 0) { |
| 256 | video_stream = fmt_ctx->streams[video_stream_idx]; |
| 257 | video_dec_ctx = video_stream->codec; |
| 258 | |
| 259 | video_dst_file = fopen(video_dst_filename, "wb"); |
| 260 | if (!video_dst_file) { |
| 261 | fprintf(stderr, "Could not open destination file %s\n", video_dst_filename); |
| 262 | ret = 1; |
| 263 | goto end; |
| 264 | } |
| 265 | |
| 266 | /* allocate image where the decoded image will be put */ |
| 267 | ret = av_image_alloc(video_dst_data, video_dst_linesize, |
| 268 | video_dec_ctx->width, video_dec_ctx->height, |
| 269 | video_dec_ctx->pix_fmt, 1); |
| 270 | if (ret < 0) { |
| 271 | fprintf(stderr, "Could not allocate raw video buffer\n"); |
| 272 | goto end; |
| 273 | } |
| 274 | video_dst_bufsize = ret; |
| 275 | } |
| 276 | |
| 277 | if (open_codec_context(&audio_stream_idx, fmt_ctx, AVMEDIA_TYPE_AUDIO) >= 0) { |
| 278 | audio_stream = fmt_ctx->streams[audio_stream_idx]; |
| 279 | audio_dec_ctx = audio_stream->codec; |
| 280 | audio_dst_file = fopen(audio_dst_filename, "wb"); |
| 281 | if (!audio_dst_file) { |
| 282 | fprintf(stderr, "Could not open destination file %s\n", audio_dst_filename); |
| 283 | ret = 1; |
| 284 | goto end; |
| 285 | } |
| 286 | } |
| 287 | |
| 288 | /* dump input information to stderr */ |
| 289 | av_dump_format(fmt_ctx, 0, src_filename, 0); |
| 290 | |
| 291 | if (!audio_stream && !video_stream) { |
| 292 | fprintf(stderr, "Could not find audio or video stream in the input, aborting\n"); |
| 293 | ret = 1; |
| 294 | goto end; |
| 295 | } |
| 296 | |
| 297 | /* When using the new API, you need to use the libavutil/frame.h API, while |
| 298 | * the classic frame management is available in libavcodec */ |
| 299 | if (api_mode == API_MODE_OLD) |
| 300 | frame = avcodec_alloc_frame(); |
| 301 | else |
| 302 | frame = av_frame_alloc(); |
| 303 | if (!frame) { |
| 304 | fprintf(stderr, "Could not allocate frame\n"); |
| 305 | ret = AVERROR(ENOMEM); |
| 306 | goto end; |
| 307 | } |
| 308 | |
| 309 | /* initialize packet, set data to NULL, let the demuxer fill it */ |
| 310 | av_init_packet(&pkt); |
| 311 | pkt.data = NULL; |
| 312 | pkt.size = 0; |
| 313 | |
| 314 | if (video_stream) |
| 315 | printf("Demuxing video from file '%s' into '%s'\n", src_filename, video_dst_filename); |
| 316 | if (audio_stream) |
| 317 | printf("Demuxing audio from file '%s' into '%s'\n", src_filename, audio_dst_filename); |
| 318 | |
| 319 | /* read frames from the file */ |
| 320 | while (av_read_frame(fmt_ctx, &pkt) >= 0) { |
| 321 | AVPacket orig_pkt = pkt; |
| 322 | do { |
| 323 | ret = decode_packet(&got_frame, 0); |
| 324 | if (ret < 0) |
| 325 | break; |
| 326 | pkt.data += ret; |
| 327 | pkt.size -= ret; |
| 328 | } while (pkt.size > 0); |
| 329 | av_free_packet(&orig_pkt); |
| 330 | } |
| 331 | |
| 332 | /* flush cached frames */ |
| 333 | pkt.data = NULL; |
| 334 | pkt.size = 0; |
| 335 | do { |
| 336 | decode_packet(&got_frame, 1); |
| 337 | } while (got_frame); |
| 338 | |
| 339 | printf("Demuxing succeeded.\n"); |
| 340 | |
| 341 | if (video_stream) { |
| 342 | printf("Play the output video file with the command:\n" |
| 343 | "ffplay -f rawvideo -pix_fmt %s -video_size %dx%d %s\n", |
| 344 | av_get_pix_fmt_name(video_dec_ctx->pix_fmt), video_dec_ctx->width, video_dec_ctx->height, |
| 345 | video_dst_filename); |
| 346 | } |
| 347 | |
| 348 | if (audio_stream) { |
| 349 | enum AVSampleFormat sfmt = audio_dec_ctx->sample_fmt; |
| 350 | int n_channels = audio_dec_ctx->channels; |
| 351 | const char *fmt; |
| 352 | |
| 353 | if (av_sample_fmt_is_planar(sfmt)) { |
| 354 | const char *packed = av_get_sample_fmt_name(sfmt); |
| 355 | printf("Warning: the sample format the decoder produced is planar " |
| 356 | "(%s). This example will output the first channel only.\n", |
| 357 | packed ? packed : "?"); |
| 358 | sfmt = av_get_packed_sample_fmt(sfmt); |
| 359 | n_channels = 1; |
| 360 | } |
| 361 | |
| 362 | if ((ret = get_format_from_sample_fmt(&fmt, sfmt)) < 0) |
| 363 | goto end; |
| 364 | |
| 365 | printf("Play the output audio file with the command:\n" |
| 366 | "ffplay -f %s -ac %d -ar %d %s\n", |
| 367 | fmt, n_channels, audio_dec_ctx->sample_rate, |
| 368 | audio_dst_filename); |
| 369 | } |
| 370 | |
| 371 | end: |
| 372 | avcodec_close(video_dec_ctx); |
| 373 | avcodec_close(audio_dec_ctx); |
| 374 | avformat_close_input(&fmt_ctx); |
| 375 | if (video_dst_file) |
| 376 | fclose(video_dst_file); |
| 377 | if (audio_dst_file) |
| 378 | fclose(audio_dst_file); |
| 379 | if (api_mode == API_MODE_OLD) |
| 380 | avcodec_free_frame(&frame); |
| 381 | else |
| 382 | av_frame_free(&frame); |
| 383 | av_free(video_dst_data[0]); |
| 384 | |
| 385 | return ret < 0; |
| 386 | } |