2 * Copyright (c) 2012 Stefano Sabatini
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:
11 * The above copyright notice and this permission notice shall be included in
12 * all copies or substantial portions of the Software.
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
25 * Demuxing and decoding example.
27 * Show how to use the libavformat and libavcodec API to demux and
28 * decode audio and video data.
29 * @example demuxing_decoding.c
32 #include <libavutil/imgutils.h>
33 #include <libavutil/samplefmt.h>
34 #include <libavutil/timestamp.h>
35 #include <libavformat/avformat.h>
37 static AVFormatContext *fmt_ctx = NULL;
38 static AVCodecContext *video_dec_ctx = NULL, *audio_dec_ctx;
39 static int width, height;
40 static enum AVPixelFormat pix_fmt;
41 static AVStream *video_stream = NULL, *audio_stream = NULL;
42 static const char *src_filename = NULL;
43 static const char *video_dst_filename = NULL;
44 static const char *audio_dst_filename = NULL;
45 static FILE *video_dst_file = NULL;
46 static FILE *audio_dst_file = NULL;
48 static uint8_t *video_dst_data[4] = {NULL};
49 static int video_dst_linesize[4];
50 static int video_dst_bufsize;
52 static int video_stream_idx = -1, audio_stream_idx = -1;
53 static AVFrame *frame = NULL;
55 static int video_frame_count = 0;
56 static int audio_frame_count = 0;
58 /* The different ways of decoding and managing data memory. You are not
59 * supposed to support all the modes in your application but pick the one most
60 * appropriate to your needs. Look for the use of api_mode in this example to
61 * see what are the differences of API usage between them */
63 API_MODE_NEW_API_REF_COUNT = 1, /* new method, using the frame reference counting */
64 API_MODE_NEW_API_NO_REF_COUNT = 2, /* new method, without reference counting */
67 static int api_mode = API_MODE_NEW_API_NO_REF_COUNT;
69 static int decode_packet(int *got_frame, int cached)
72 int decoded = pkt.size;
76 if (pkt.stream_index == video_stream_idx) {
77 /* decode video frame */
78 ret = avcodec_decode_video2(video_dec_ctx, frame, got_frame, &pkt);
80 fprintf(stderr, "Error decoding video frame (%s)\n", av_err2str(ret));
86 if (frame->width != width || frame->height != height ||
87 frame->format != pix_fmt) {
88 /* To handle this change, one could call av_image_alloc again and
89 * decode the following frames into another rawvideo file. */
90 fprintf(stderr, "Error: Width, height and pixel format have to be "
91 "constant in a rawvideo file, but the width, height or "
92 "pixel format of the input video changed:\n"
93 "old: width = %d, height = %d, format = %s\n"
94 "new: width = %d, height = %d, format = %s\n",
95 width, height, av_get_pix_fmt_name(pix_fmt),
96 frame->width, frame->height,
97 av_get_pix_fmt_name(frame->format));
101 printf("video_frame%s n:%d coded_n:%d pts:%s\n",
102 cached ? "(cached)" : "",
103 video_frame_count++, frame->coded_picture_number,
104 av_ts2timestr(frame->pts, &video_dec_ctx->time_base));
106 /* copy decoded frame to destination buffer:
107 * this is required since rawvideo expects non aligned data */
108 av_image_copy(video_dst_data, video_dst_linesize,
109 (const uint8_t **)(frame->data), frame->linesize,
110 pix_fmt, width, height);
112 /* write to rawvideo file */
113 fwrite(video_dst_data[0], 1, video_dst_bufsize, video_dst_file);
115 } else if (pkt.stream_index == audio_stream_idx) {
116 /* decode audio frame */
117 ret = avcodec_decode_audio4(audio_dec_ctx, frame, got_frame, &pkt);
119 fprintf(stderr, "Error decoding audio frame (%s)\n", av_err2str(ret));
122 /* Some audio decoders decode only part of the packet, and have to be
123 * called again with the remainder of the packet data.
124 * Sample: fate-suite/lossless-audio/luckynight-partial.shn
125 * Also, some decoders might over-read the packet. */
126 decoded = FFMIN(ret, pkt.size);
129 size_t unpadded_linesize = frame->nb_samples * av_get_bytes_per_sample(frame->format);
130 printf("audio_frame%s n:%d nb_samples:%d pts:%s\n",
131 cached ? "(cached)" : "",
132 audio_frame_count++, frame->nb_samples,
133 av_ts2timestr(frame->pts, &audio_dec_ctx->time_base));
135 /* Write the raw audio data samples of the first plane. This works
136 * fine for packed formats (e.g. AV_SAMPLE_FMT_S16). However,
137 * most audio decoders output planar audio, which uses a separate
138 * plane of audio samples for each channel (e.g. AV_SAMPLE_FMT_S16P).
139 * In other words, this code will write only the first audio channel
141 * You should use libswresample or libavfilter to convert the frame
143 fwrite(frame->extended_data[0], 1, unpadded_linesize, audio_dst_file);
147 /* If we use the new API with reference counting, we own the data and need
148 * to de-reference it when we don't use it anymore */
149 if (*got_frame && api_mode == API_MODE_NEW_API_REF_COUNT)
150 av_frame_unref(frame);
155 static int open_codec_context(int *stream_idx,
156 AVFormatContext *fmt_ctx, enum AVMediaType type)
158 int ret, stream_index;
160 AVCodecContext *dec_ctx = NULL;
162 AVDictionary *opts = NULL;
164 ret = av_find_best_stream(fmt_ctx, type, -1, -1, NULL, 0);
166 fprintf(stderr, "Could not find %s stream in input file '%s'\n",
167 av_get_media_type_string(type), src_filename);
171 st = fmt_ctx->streams[stream_index];
173 /* find decoder for the stream */
175 dec = avcodec_find_decoder(dec_ctx->codec_id);
177 fprintf(stderr, "Failed to find %s codec\n",
178 av_get_media_type_string(type));
179 return AVERROR(EINVAL);
182 /* Init the decoders, with or without reference counting */
183 if (api_mode == API_MODE_NEW_API_REF_COUNT)
184 av_dict_set(&opts, "refcounted_frames", "1", 0);
185 if ((ret = avcodec_open2(dec_ctx, dec, &opts)) < 0) {
186 fprintf(stderr, "Failed to open %s codec\n",
187 av_get_media_type_string(type));
190 *stream_idx = stream_index;
196 static int get_format_from_sample_fmt(const char **fmt,
197 enum AVSampleFormat sample_fmt)
200 struct sample_fmt_entry {
201 enum AVSampleFormat sample_fmt; const char *fmt_be, *fmt_le;
202 } sample_fmt_entries[] = {
203 { AV_SAMPLE_FMT_U8, "u8", "u8" },
204 { AV_SAMPLE_FMT_S16, "s16be", "s16le" },
205 { AV_SAMPLE_FMT_S32, "s32be", "s32le" },
206 { AV_SAMPLE_FMT_FLT, "f32be", "f32le" },
207 { AV_SAMPLE_FMT_DBL, "f64be", "f64le" },
211 for (i = 0; i < FF_ARRAY_ELEMS(sample_fmt_entries); i++) {
212 struct sample_fmt_entry *entry = &sample_fmt_entries[i];
213 if (sample_fmt == entry->sample_fmt) {
214 *fmt = AV_NE(entry->fmt_be, entry->fmt_le);
220 "sample format %s is not supported as output format\n",
221 av_get_sample_fmt_name(sample_fmt));
225 int main (int argc, char **argv)
227 int ret = 0, got_frame;
229 if (argc != 4 && argc != 5) {
230 fprintf(stderr, "usage: %s [-refcount=<old|new_norefcount|new_refcount>] "
231 "input_file video_output_file audio_output_file\n"
232 "API example program to show how to read frames from an input file.\n"
233 "This program reads frames from a file, decodes them, and writes decoded\n"
234 "video frames to a rawvideo file named video_output_file, and decoded\n"
235 "audio frames to a rawaudio file named audio_output_file.\n\n"
236 "If the -refcount option is specified, the program use the\n"
237 "reference counting frame system which allows keeping a copy of\n"
238 "the data for longer than one decode call. If unset, it's using\n"
239 "the classic old method.\n"
244 const char *mode = argv[1] + strlen("-refcount=");
245 if (!strcmp(mode, "new_norefcount")) api_mode = API_MODE_NEW_API_NO_REF_COUNT;
246 else if (!strcmp(mode, "new_refcount")) api_mode = API_MODE_NEW_API_REF_COUNT;
248 fprintf(stderr, "unknow mode '%s'\n", mode);
253 src_filename = argv[1];
254 video_dst_filename = argv[2];
255 audio_dst_filename = argv[3];
257 /* register all formats and codecs */
260 /* open input file, and allocate format context */
261 if (avformat_open_input(&fmt_ctx, src_filename, NULL, NULL) < 0) {
262 fprintf(stderr, "Could not open source file %s\n", src_filename);
266 /* retrieve stream information */
267 if (avformat_find_stream_info(fmt_ctx, NULL) < 0) {
268 fprintf(stderr, "Could not find stream information\n");
272 if (open_codec_context(&video_stream_idx, fmt_ctx, AVMEDIA_TYPE_VIDEO) >= 0) {
273 video_stream = fmt_ctx->streams[video_stream_idx];
274 video_dec_ctx = video_stream->codec;
276 video_dst_file = fopen(video_dst_filename, "wb");
277 if (!video_dst_file) {
278 fprintf(stderr, "Could not open destination file %s\n", video_dst_filename);
283 /* allocate image where the decoded image will be put */
284 width = video_dec_ctx->width;
285 height = video_dec_ctx->height;
286 pix_fmt = video_dec_ctx->pix_fmt;
287 ret = av_image_alloc(video_dst_data, video_dst_linesize,
288 width, height, pix_fmt, 1);
290 fprintf(stderr, "Could not allocate raw video buffer\n");
293 video_dst_bufsize = ret;
296 if (open_codec_context(&audio_stream_idx, fmt_ctx, AVMEDIA_TYPE_AUDIO) >= 0) {
297 audio_stream = fmt_ctx->streams[audio_stream_idx];
298 audio_dec_ctx = audio_stream->codec;
299 audio_dst_file = fopen(audio_dst_filename, "wb");
300 if (!audio_dst_file) {
301 fprintf(stderr, "Could not open destination file %s\n", audio_dst_filename);
307 /* dump input information to stderr */
308 av_dump_format(fmt_ctx, 0, src_filename, 0);
310 if (!audio_stream && !video_stream) {
311 fprintf(stderr, "Could not find audio or video stream in the input, aborting\n");
316 frame = av_frame_alloc();
318 fprintf(stderr, "Could not allocate frame\n");
319 ret = AVERROR(ENOMEM);
323 /* initialize packet, set data to NULL, let the demuxer fill it */
324 av_init_packet(&pkt);
329 printf("Demuxing video from file '%s' into '%s'\n", src_filename, video_dst_filename);
331 printf("Demuxing audio from file '%s' into '%s'\n", src_filename, audio_dst_filename);
333 /* read frames from the file */
334 while (av_read_frame(fmt_ctx, &pkt) >= 0) {
335 AVPacket orig_pkt = pkt;
337 ret = decode_packet(&got_frame, 0);
342 } while (pkt.size > 0);
343 av_free_packet(&orig_pkt);
346 /* flush cached frames */
350 decode_packet(&got_frame, 1);
353 printf("Demuxing succeeded.\n");
356 printf("Play the output video file with the command:\n"
357 "ffplay -f rawvideo -pix_fmt %s -video_size %dx%d %s\n",
358 av_get_pix_fmt_name(pix_fmt), width, height,
363 enum AVSampleFormat sfmt = audio_dec_ctx->sample_fmt;
364 int n_channels = audio_dec_ctx->channels;
367 if (av_sample_fmt_is_planar(sfmt)) {
368 const char *packed = av_get_sample_fmt_name(sfmt);
369 printf("Warning: the sample format the decoder produced is planar "
370 "(%s). This example will output the first channel only.\n",
371 packed ? packed : "?");
372 sfmt = av_get_packed_sample_fmt(sfmt);
376 if ((ret = get_format_from_sample_fmt(&fmt, sfmt)) < 0)
379 printf("Play the output audio file with the command:\n"
380 "ffplay -f %s -ac %d -ar %d %s\n",
381 fmt, n_channels, audio_dec_ctx->sample_rate,
386 avcodec_close(video_dec_ctx);
387 avcodec_close(audio_dec_ctx);
388 avformat_close_input(&fmt_ctx);
390 fclose(video_dst_file);
392 fclose(audio_dst_file);
393 av_frame_free(&frame);
394 av_free(video_dst_data[0]);