2 * Copyright (c) 2003 Fabrice Bellard
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 * libavformat API example.
27 * Output a media file in any supported libavformat format. The default
37 #include <libavutil/avassert.h>
38 #include <libavutil/channel_layout.h>
39 #include <libavutil/opt.h>
40 #include <libavutil/mathematics.h>
41 #include <libavutil/timestamp.h>
42 #include <libavformat/avformat.h>
43 #include <libswscale/swscale.h>
44 #include <libswresample/swresample.h>
46 #define STREAM_DURATION 10.0
47 #define STREAM_FRAME_RATE 25 /* 25 images/s */
48 #define STREAM_PIX_FMT AV_PIX_FMT_YUV420P /* default pix_fmt */
50 #define SCALE_FLAGS SWS_BICUBIC
52 // a wrapper around a single output AVStream
53 typedef struct OutputStream {
57 /* pts of the next frame that will be generated */
64 float t, tincr, tincr2;
66 struct SwsContext *sws_ctx;
67 struct SwrContext *swr_ctx;
70 static void log_packet(const AVFormatContext *fmt_ctx, const AVPacket *pkt)
72 AVRational *time_base = &fmt_ctx->streams[pkt->stream_index]->time_base;
74 printf("pts:%s pts_time:%s dts:%s dts_time:%s duration:%s duration_time:%s stream_index:%d\n",
75 av_ts2str(pkt->pts), av_ts2timestr(pkt->pts, time_base),
76 av_ts2str(pkt->dts), av_ts2timestr(pkt->dts, time_base),
77 av_ts2str(pkt->duration), av_ts2timestr(pkt->duration, time_base),
81 static int write_frame(AVFormatContext *fmt_ctx, AVCodecContext *c,
82 AVStream *st, AVFrame *frame)
86 // send the frame to the encoder
87 ret = avcodec_send_frame(c, frame);
89 fprintf(stderr, "Error sending a frame to the encoder: %s\n",
97 ret = avcodec_receive_packet(c, &pkt);
98 if (ret == AVERROR(EAGAIN) || ret == AVERROR_EOF)
101 fprintf(stderr, "Error encoding a frame: %s\n", av_err2str(ret));
105 /* rescale output packet timestamp values from codec to stream timebase */
106 av_packet_rescale_ts(&pkt, c->time_base, st->time_base);
107 pkt.stream_index = st->index;
109 /* Write the compressed frame to the media file. */
110 log_packet(fmt_ctx, &pkt);
111 ret = av_interleaved_write_frame(fmt_ctx, &pkt);
112 av_packet_unref(&pkt);
114 fprintf(stderr, "Error while writing output packet: %s\n", av_err2str(ret));
119 return ret == AVERROR_EOF ? 1 : 0;
122 /* Add an output stream. */
123 static void add_stream(OutputStream *ost, AVFormatContext *oc,
125 enum AVCodecID codec_id)
130 /* find the encoder */
131 *codec = avcodec_find_encoder(codec_id);
133 fprintf(stderr, "Could not find encoder for '%s'\n",
134 avcodec_get_name(codec_id));
138 ost->st = avformat_new_stream(oc, NULL);
140 fprintf(stderr, "Could not allocate stream\n");
143 ost->st->id = oc->nb_streams-1;
144 c = avcodec_alloc_context3(*codec);
146 fprintf(stderr, "Could not alloc an encoding context\n");
151 switch ((*codec)->type) {
152 case AVMEDIA_TYPE_AUDIO:
153 c->sample_fmt = (*codec)->sample_fmts ?
154 (*codec)->sample_fmts[0] : AV_SAMPLE_FMT_FLTP;
156 c->sample_rate = 44100;
157 if ((*codec)->supported_samplerates) {
158 c->sample_rate = (*codec)->supported_samplerates[0];
159 for (i = 0; (*codec)->supported_samplerates[i]; i++) {
160 if ((*codec)->supported_samplerates[i] == 44100)
161 c->sample_rate = 44100;
164 c->channels = av_get_channel_layout_nb_channels(c->channel_layout);
165 c->channel_layout = AV_CH_LAYOUT_STEREO;
166 if ((*codec)->channel_layouts) {
167 c->channel_layout = (*codec)->channel_layouts[0];
168 for (i = 0; (*codec)->channel_layouts[i]; i++) {
169 if ((*codec)->channel_layouts[i] == AV_CH_LAYOUT_STEREO)
170 c->channel_layout = AV_CH_LAYOUT_STEREO;
173 c->channels = av_get_channel_layout_nb_channels(c->channel_layout);
174 ost->st->time_base = (AVRational){ 1, c->sample_rate };
177 case AVMEDIA_TYPE_VIDEO:
178 c->codec_id = codec_id;
180 c->bit_rate = 400000;
181 /* Resolution must be a multiple of two. */
184 /* timebase: This is the fundamental unit of time (in seconds) in terms
185 * of which frame timestamps are represented. For fixed-fps content,
186 * timebase should be 1/framerate and timestamp increments should be
188 ost->st->time_base = (AVRational){ 1, STREAM_FRAME_RATE };
189 c->time_base = ost->st->time_base;
191 c->gop_size = 12; /* emit one intra frame every twelve frames at most */
192 c->pix_fmt = STREAM_PIX_FMT;
193 if (c->codec_id == AV_CODEC_ID_MPEG2VIDEO) {
194 /* just for testing, we also add B-frames */
197 if (c->codec_id == AV_CODEC_ID_MPEG1VIDEO) {
198 /* Needed to avoid using macroblocks in which some coeffs overflow.
199 * This does not happen with normal video, it just happens here as
200 * the motion of the chroma plane does not match the luma plane. */
209 /* Some formats want stream headers to be separate. */
210 if (oc->oformat->flags & AVFMT_GLOBALHEADER)
211 c->flags |= AV_CODEC_FLAG_GLOBAL_HEADER;
214 /**************************************************************/
217 static AVFrame *alloc_audio_frame(enum AVSampleFormat sample_fmt,
218 uint64_t channel_layout,
219 int sample_rate, int nb_samples)
221 AVFrame *frame = av_frame_alloc();
225 fprintf(stderr, "Error allocating an audio frame\n");
229 frame->format = sample_fmt;
230 frame->channel_layout = channel_layout;
231 frame->sample_rate = sample_rate;
232 frame->nb_samples = nb_samples;
235 ret = av_frame_get_buffer(frame, 0);
237 fprintf(stderr, "Error allocating an audio buffer\n");
245 static void open_audio(AVFormatContext *oc, AVCodec *codec, OutputStream *ost, AVDictionary *opt_arg)
250 AVDictionary *opt = NULL;
255 av_dict_copy(&opt, opt_arg, 0);
256 ret = avcodec_open2(c, codec, &opt);
259 fprintf(stderr, "Could not open audio codec: %s\n", av_err2str(ret));
263 /* init signal generator */
265 ost->tincr = 2 * M_PI * 110.0 / c->sample_rate;
266 /* increment frequency by 110 Hz per second */
267 ost->tincr2 = 2 * M_PI * 110.0 / c->sample_rate / c->sample_rate;
269 if (c->codec->capabilities & AV_CODEC_CAP_VARIABLE_FRAME_SIZE)
272 nb_samples = c->frame_size;
274 ost->frame = alloc_audio_frame(c->sample_fmt, c->channel_layout,
275 c->sample_rate, nb_samples);
276 ost->tmp_frame = alloc_audio_frame(AV_SAMPLE_FMT_S16, c->channel_layout,
277 c->sample_rate, nb_samples);
279 /* copy the stream parameters to the muxer */
280 ret = avcodec_parameters_from_context(ost->st->codecpar, c);
282 fprintf(stderr, "Could not copy the stream parameters\n");
286 /* create resampler context */
287 ost->swr_ctx = swr_alloc();
289 fprintf(stderr, "Could not allocate resampler context\n");
294 av_opt_set_int (ost->swr_ctx, "in_channel_count", c->channels, 0);
295 av_opt_set_int (ost->swr_ctx, "in_sample_rate", c->sample_rate, 0);
296 av_opt_set_sample_fmt(ost->swr_ctx, "in_sample_fmt", AV_SAMPLE_FMT_S16, 0);
297 av_opt_set_int (ost->swr_ctx, "out_channel_count", c->channels, 0);
298 av_opt_set_int (ost->swr_ctx, "out_sample_rate", c->sample_rate, 0);
299 av_opt_set_sample_fmt(ost->swr_ctx, "out_sample_fmt", c->sample_fmt, 0);
301 /* initialize the resampling context */
302 if ((ret = swr_init(ost->swr_ctx)) < 0) {
303 fprintf(stderr, "Failed to initialize the resampling context\n");
308 /* Prepare a 16 bit dummy audio frame of 'frame_size' samples and
309 * 'nb_channels' channels. */
310 static AVFrame *get_audio_frame(OutputStream *ost)
312 AVFrame *frame = ost->tmp_frame;
314 int16_t *q = (int16_t*)frame->data[0];
316 /* check if we want to generate more frames */
317 if (av_compare_ts(ost->next_pts, ost->enc->time_base,
318 STREAM_DURATION, (AVRational){ 1, 1 }) > 0)
321 for (j = 0; j <frame->nb_samples; j++) {
322 v = (int)(sin(ost->t) * 10000);
323 for (i = 0; i < ost->enc->channels; i++)
325 ost->t += ost->tincr;
326 ost->tincr += ost->tincr2;
329 frame->pts = ost->next_pts;
330 ost->next_pts += frame->nb_samples;
336 * encode one audio frame and send it to the muxer
337 * return 1 when encoding is finished, 0 otherwise
339 static int write_audio_frame(AVFormatContext *oc, OutputStream *ost)
348 frame = get_audio_frame(ost);
351 /* convert samples from native format to destination codec format, using the resampler */
352 /* compute destination number of samples */
353 dst_nb_samples = av_rescale_rnd(swr_get_delay(ost->swr_ctx, c->sample_rate) + frame->nb_samples,
354 c->sample_rate, c->sample_rate, AV_ROUND_UP);
355 av_assert0(dst_nb_samples == frame->nb_samples);
357 /* when we pass a frame to the encoder, it may keep a reference to it
359 * make sure we do not overwrite it here
361 ret = av_frame_make_writable(ost->frame);
365 /* convert to destination format */
366 ret = swr_convert(ost->swr_ctx,
367 ost->frame->data, dst_nb_samples,
368 (const uint8_t **)frame->data, frame->nb_samples);
370 fprintf(stderr, "Error while converting\n");
375 frame->pts = av_rescale_q(ost->samples_count, (AVRational){1, c->sample_rate}, c->time_base);
376 ost->samples_count += dst_nb_samples;
379 return write_frame(oc, c, ost->st, frame);
382 /**************************************************************/
385 static AVFrame *alloc_picture(enum AVPixelFormat pix_fmt, int width, int height)
390 picture = av_frame_alloc();
394 picture->format = pix_fmt;
395 picture->width = width;
396 picture->height = height;
398 /* allocate the buffers for the frame data */
399 ret = av_frame_get_buffer(picture, 0);
401 fprintf(stderr, "Could not allocate frame data.\n");
408 static void open_video(AVFormatContext *oc, AVCodec *codec, OutputStream *ost, AVDictionary *opt_arg)
411 AVCodecContext *c = ost->enc;
412 AVDictionary *opt = NULL;
414 av_dict_copy(&opt, opt_arg, 0);
417 ret = avcodec_open2(c, codec, &opt);
420 fprintf(stderr, "Could not open video codec: %s\n", av_err2str(ret));
424 /* allocate and init a re-usable frame */
425 ost->frame = alloc_picture(c->pix_fmt, c->width, c->height);
427 fprintf(stderr, "Could not allocate video frame\n");
431 /* If the output format is not YUV420P, then a temporary YUV420P
432 * picture is needed too. It is then converted to the required
434 ost->tmp_frame = NULL;
435 if (c->pix_fmt != AV_PIX_FMT_YUV420P) {
436 ost->tmp_frame = alloc_picture(AV_PIX_FMT_YUV420P, c->width, c->height);
437 if (!ost->tmp_frame) {
438 fprintf(stderr, "Could not allocate temporary picture\n");
443 /* copy the stream parameters to the muxer */
444 ret = avcodec_parameters_from_context(ost->st->codecpar, c);
446 fprintf(stderr, "Could not copy the stream parameters\n");
451 /* Prepare a dummy image. */
452 static void fill_yuv_image(AVFrame *pict, int frame_index,
453 int width, int height)
460 for (y = 0; y < height; y++)
461 for (x = 0; x < width; x++)
462 pict->data[0][y * pict->linesize[0] + x] = x + y + i * 3;
465 for (y = 0; y < height / 2; y++) {
466 for (x = 0; x < width / 2; x++) {
467 pict->data[1][y * pict->linesize[1] + x] = 128 + y + i * 2;
468 pict->data[2][y * pict->linesize[2] + x] = 64 + x + i * 5;
473 static AVFrame *get_video_frame(OutputStream *ost)
475 AVCodecContext *c = ost->enc;
477 /* check if we want to generate more frames */
478 if (av_compare_ts(ost->next_pts, c->time_base,
479 STREAM_DURATION, (AVRational){ 1, 1 }) > 0)
482 /* when we pass a frame to the encoder, it may keep a reference to it
483 * internally; make sure we do not overwrite it here */
484 if (av_frame_make_writable(ost->frame) < 0)
487 if (c->pix_fmt != AV_PIX_FMT_YUV420P) {
488 /* as we only generate a YUV420P picture, we must convert it
489 * to the codec pixel format if needed */
491 ost->sws_ctx = sws_getContext(c->width, c->height,
495 SCALE_FLAGS, NULL, NULL, NULL);
498 "Could not initialize the conversion context\n");
502 fill_yuv_image(ost->tmp_frame, ost->next_pts, c->width, c->height);
503 sws_scale(ost->sws_ctx, (const uint8_t * const *) ost->tmp_frame->data,
504 ost->tmp_frame->linesize, 0, c->height, ost->frame->data,
505 ost->frame->linesize);
507 fill_yuv_image(ost->frame, ost->next_pts, c->width, c->height);
510 ost->frame->pts = ost->next_pts++;
516 * encode one video frame and send it to the muxer
517 * return 1 when encoding is finished, 0 otherwise
519 static int write_video_frame(AVFormatContext *oc, OutputStream *ost)
521 return write_frame(oc, ost->enc, ost->st, get_video_frame(ost));
525 static void close_stream(AVFormatContext *oc, OutputStream *ost)
527 avcodec_free_context(&ost->enc);
528 av_frame_free(&ost->frame);
529 av_frame_free(&ost->tmp_frame);
530 sws_freeContext(ost->sws_ctx);
531 swr_free(&ost->swr_ctx);
534 /**************************************************************/
535 /* media file output */
537 int main(int argc, char **argv)
539 OutputStream video_st = { 0 }, audio_st = { 0 };
540 const char *filename;
543 AVCodec *audio_codec, *video_codec;
545 int have_video = 0, have_audio = 0;
546 int encode_video = 0, encode_audio = 0;
547 AVDictionary *opt = NULL;
551 printf("usage: %s output_file\n"
552 "API example program to output a media file with libavformat.\n"
553 "This program generates a synthetic audio and video stream, encodes and\n"
554 "muxes them into a file named output_file.\n"
555 "The output format is automatically guessed according to the file extension.\n"
556 "Raw images can also be output by using '%%d' in the filename.\n"
562 for (i = 2; i+1 < argc; i+=2) {
563 if (!strcmp(argv[i], "-flags") || !strcmp(argv[i], "-fflags"))
564 av_dict_set(&opt, argv[i]+1, argv[i+1], 0);
567 /* allocate the output media context */
568 avformat_alloc_output_context2(&oc, NULL, NULL, filename);
570 printf("Could not deduce output format from file extension: using MPEG.\n");
571 avformat_alloc_output_context2(&oc, NULL, "mpeg", filename);
578 /* Add the audio and video streams using the default format codecs
579 * and initialize the codecs. */
580 if (fmt->video_codec != AV_CODEC_ID_NONE) {
581 add_stream(&video_st, oc, &video_codec, fmt->video_codec);
585 if (fmt->audio_codec != AV_CODEC_ID_NONE) {
586 add_stream(&audio_st, oc, &audio_codec, fmt->audio_codec);
591 /* Now that all the parameters are set, we can open the audio and
592 * video codecs and allocate the necessary encode buffers. */
594 open_video(oc, video_codec, &video_st, opt);
597 open_audio(oc, audio_codec, &audio_st, opt);
599 av_dump_format(oc, 0, filename, 1);
601 /* open the output file, if needed */
602 if (!(fmt->flags & AVFMT_NOFILE)) {
603 ret = avio_open(&oc->pb, filename, AVIO_FLAG_WRITE);
605 fprintf(stderr, "Could not open '%s': %s\n", filename,
611 /* Write the stream header, if any. */
612 ret = avformat_write_header(oc, &opt);
614 fprintf(stderr, "Error occurred when opening output file: %s\n",
619 while (encode_video || encode_audio) {
620 /* select the stream to encode */
622 (!encode_audio || av_compare_ts(video_st.next_pts, video_st.enc->time_base,
623 audio_st.next_pts, audio_st.enc->time_base) <= 0)) {
624 encode_video = !write_video_frame(oc, &video_st);
626 encode_audio = !write_audio_frame(oc, &audio_st);
630 /* Write the trailer, if any. The trailer must be written before you
631 * close the CodecContexts open when you wrote the header; otherwise
632 * av_write_trailer() may try to use memory that was freed on
633 * av_codec_close(). */
634 av_write_trailer(oc);
636 /* Close each codec. */
638 close_stream(oc, &video_st);
640 close_stream(oc, &audio_st);
642 if (!(fmt->flags & AVFMT_NOFILE))
643 /* Close the output file. */
644 avio_closep(&oc->pb);
646 /* free the stream */
647 avformat_free_context(oc);