FFmpeg 6.1.1
transcode_aac.c
Go to the documentation of this file.
1/*
2 * Copyright (c) 2013-2022 Andreas Unterweger
3 *
4 * This file is part of FFmpeg.
5 *
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
10 *
11 * FFmpeg is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
15 *
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19 */
20
21/**
22 * @file audio transcoding to MPEG/AAC API usage example
23 * @example transcode_aac.c
24 *
25 * Convert an input audio file to AAC in an MP4 container. Formats other than
26 * MP4 are supported based on the output file extension.
27 * @author Andreas Unterweger (dustsigns@gmail.com)
28 */
29
30#include <stdio.h>
31
33#include "libavformat/avio.h"
34
35#include "libavcodec/avcodec.h"
36
38#include "libavutil/avassert.h"
39#include "libavutil/avstring.h"
41#include "libavutil/frame.h"
42#include "libavutil/opt.h"
43
45
46/* The output bit rate in bit/s */
47#define OUTPUT_BIT_RATE 96000
48/* The number of output channels */
49#define OUTPUT_CHANNELS 2
50
51/**
52 * Open an input file and the required decoder.
53 * @param filename File to be opened
54 * @param[out] input_format_context Format context of opened file
55 * @param[out] input_codec_context Codec context of opened file
56 * @return Error code (0 if successful)
57 */
58static int open_input_file(const char *filename,
59 AVFormatContext **input_format_context,
60 AVCodecContext **input_codec_context)
61{
62 AVCodecContext *avctx;
63 const AVCodec *input_codec;
64 const AVStream *stream;
65 int error;
66
67 /* Open the input file to read from it. */
68 if ((error = avformat_open_input(input_format_context, filename, NULL,
69 NULL)) < 0) {
70 fprintf(stderr, "Could not open input file '%s' (error '%s')\n",
71 filename, av_err2str(error));
72 *input_format_context = NULL;
73 return error;
74 }
75
76 /* Get information on the input file (number of streams etc.). */
77 if ((error = avformat_find_stream_info(*input_format_context, NULL)) < 0) {
78 fprintf(stderr, "Could not open find stream info (error '%s')\n",
79 av_err2str(error));
80 avformat_close_input(input_format_context);
81 return error;
82 }
83
84 /* Make sure that there is only one stream in the input file. */
85 if ((*input_format_context)->nb_streams != 1) {
86 fprintf(stderr, "Expected one audio input stream, but found %d\n",
87 (*input_format_context)->nb_streams);
88 avformat_close_input(input_format_context);
89 return AVERROR_EXIT;
90 }
91
92 stream = (*input_format_context)->streams[0];
93
94 /* Find a decoder for the audio stream. */
95 if (!(input_codec = avcodec_find_decoder(stream->codecpar->codec_id))) {
96 fprintf(stderr, "Could not find input codec\n");
97 avformat_close_input(input_format_context);
98 return AVERROR_EXIT;
99 }
100
101 /* Allocate a new decoding context. */
102 avctx = avcodec_alloc_context3(input_codec);
103 if (!avctx) {
104 fprintf(stderr, "Could not allocate a decoding context\n");
105 avformat_close_input(input_format_context);
106 return AVERROR(ENOMEM);
107 }
108
109 /* Initialize the stream parameters with demuxer information. */
110 error = avcodec_parameters_to_context(avctx, stream->codecpar);
111 if (error < 0) {
112 avformat_close_input(input_format_context);
113 avcodec_free_context(&avctx);
114 return error;
115 }
116
117 /* Open the decoder for the audio stream to use it later. */
118 if ((error = avcodec_open2(avctx, input_codec, NULL)) < 0) {
119 fprintf(stderr, "Could not open input codec (error '%s')\n",
120 av_err2str(error));
121 avcodec_free_context(&avctx);
122 avformat_close_input(input_format_context);
123 return error;
124 }
125
126 /* Set the packet timebase for the decoder. */
127 avctx->pkt_timebase = stream->time_base;
128
129 /* Save the decoder context for easier access later. */
130 *input_codec_context = avctx;
131
132 return 0;
133}
134
135/**
136 * Open an output file and the required encoder.
137 * Also set some basic encoder parameters.
138 * Some of these parameters are based on the input file's parameters.
139 * @param filename File to be opened
140 * @param input_codec_context Codec context of input file
141 * @param[out] output_format_context Format context of output file
142 * @param[out] output_codec_context Codec context of output file
143 * @return Error code (0 if successful)
144 */
145static int open_output_file(const char *filename,
146 AVCodecContext *input_codec_context,
147 AVFormatContext **output_format_context,
148 AVCodecContext **output_codec_context)
149{
150 AVCodecContext *avctx = NULL;
151 AVIOContext *output_io_context = NULL;
152 AVStream *stream = NULL;
153 const AVCodec *output_codec = NULL;
154 int error;
155
156 /* Open the output file to write to it. */
157 if ((error = avio_open(&output_io_context, filename,
158 AVIO_FLAG_WRITE)) < 0) {
159 fprintf(stderr, "Could not open output file '%s' (error '%s')\n",
160 filename, av_err2str(error));
161 return error;
162 }
163
164 /* Create a new format context for the output container format. */
165 if (!(*output_format_context = avformat_alloc_context())) {
166 fprintf(stderr, "Could not allocate output format context\n");
167 return AVERROR(ENOMEM);
168 }
169
170 /* Associate the output file (pointer) with the container format context. */
171 (*output_format_context)->pb = output_io_context;
172
173 /* Guess the desired container format based on the file extension. */
174 if (!((*output_format_context)->oformat = av_guess_format(NULL, filename,
175 NULL))) {
176 fprintf(stderr, "Could not find output file format\n");
177 goto cleanup;
178 }
179
180 if (!((*output_format_context)->url = av_strdup(filename))) {
181 fprintf(stderr, "Could not allocate url.\n");
182 error = AVERROR(ENOMEM);
183 goto cleanup;
184 }
185
186 /* Find the encoder to be used by its name. */
187 if (!(output_codec = avcodec_find_encoder(AV_CODEC_ID_AAC))) {
188 fprintf(stderr, "Could not find an AAC encoder.\n");
189 goto cleanup;
190 }
191
192 /* Create a new audio stream in the output file container. */
193 if (!(stream = avformat_new_stream(*output_format_context, NULL))) {
194 fprintf(stderr, "Could not create new stream\n");
195 error = AVERROR(ENOMEM);
196 goto cleanup;
197 }
198
199 avctx = avcodec_alloc_context3(output_codec);
200 if (!avctx) {
201 fprintf(stderr, "Could not allocate an encoding context\n");
202 error = AVERROR(ENOMEM);
203 goto cleanup;
204 }
205
206 /* Set the basic encoder parameters.
207 * The input file's sample rate is used to avoid a sample rate conversion. */
209 avctx->sample_rate = input_codec_context->sample_rate;
210 avctx->sample_fmt = output_codec->sample_fmts[0];
211 avctx->bit_rate = OUTPUT_BIT_RATE;
212
213 /* Set the sample rate for the container. */
214 stream->time_base.den = input_codec_context->sample_rate;
215 stream->time_base.num = 1;
216
217 /* Some container formats (like MP4) require global headers to be present.
218 * Mark the encoder so that it behaves accordingly. */
219 if ((*output_format_context)->oformat->flags & AVFMT_GLOBALHEADER)
221
222 /* Open the encoder for the audio stream to use it later. */
223 if ((error = avcodec_open2(avctx, output_codec, NULL)) < 0) {
224 fprintf(stderr, "Could not open output codec (error '%s')\n",
225 av_err2str(error));
226 goto cleanup;
227 }
228
229 error = avcodec_parameters_from_context(stream->codecpar, avctx);
230 if (error < 0) {
231 fprintf(stderr, "Could not initialize stream parameters\n");
232 goto cleanup;
233 }
234
235 /* Save the encoder context for easier access later. */
236 *output_codec_context = avctx;
237
238 return 0;
239
240cleanup:
241 avcodec_free_context(&avctx);
242 avio_closep(&(*output_format_context)->pb);
243 avformat_free_context(*output_format_context);
244 *output_format_context = NULL;
245 return error < 0 ? error : AVERROR_EXIT;
246}
247
248/**
249 * Initialize one data packet for reading or writing.
250 * @param[out] packet Packet to be initialized
251 * @return Error code (0 if successful)
252 */
253static int init_packet(AVPacket **packet)
254{
255 if (!(*packet = av_packet_alloc())) {
256 fprintf(stderr, "Could not allocate packet\n");
257 return AVERROR(ENOMEM);
258 }
259 return 0;
260}
261
262/**
263 * Initialize one audio frame for reading from the input file.
264 * @param[out] frame Frame to be initialized
265 * @return Error code (0 if successful)
266 */
268{
269 if (!(*frame = av_frame_alloc())) {
270 fprintf(stderr, "Could not allocate input frame\n");
271 return AVERROR(ENOMEM);
272 }
273 return 0;
274}
275
276/**
277 * Initialize the audio resampler based on the input and output codec settings.
278 * If the input and output sample formats differ, a conversion is required
279 * libswresample takes care of this, but requires initialization.
280 * @param input_codec_context Codec context of the input file
281 * @param output_codec_context Codec context of the output file
282 * @param[out] resample_context Resample context for the required conversion
283 * @return Error code (0 if successful)
284 */
285static int init_resampler(AVCodecContext *input_codec_context,
286 AVCodecContext *output_codec_context,
287 SwrContext **resample_context)
288{
289 int error;
290
291 /*
292 * Create a resampler context for the conversion.
293 * Set the conversion parameters.
294 */
295 error = swr_alloc_set_opts2(resample_context,
296 &output_codec_context->ch_layout,
297 output_codec_context->sample_fmt,
298 output_codec_context->sample_rate,
299 &input_codec_context->ch_layout,
300 input_codec_context->sample_fmt,
301 input_codec_context->sample_rate,
302 0, NULL);
303 if (error < 0) {
304 fprintf(stderr, "Could not allocate resample context\n");
305 return error;
306 }
307 /*
308 * Perform a sanity check so that the number of converted samples is
309 * not greater than the number of samples to be converted.
310 * If the sample rates differ, this case has to be handled differently
311 */
312 av_assert0(output_codec_context->sample_rate == input_codec_context->sample_rate);
313
314 /* Open the resampler with the specified parameters. */
315 if ((error = swr_init(*resample_context)) < 0) {
316 fprintf(stderr, "Could not open resample context\n");
317 swr_free(resample_context);
318 return error;
319 }
320 return 0;
321}
322
323/**
324 * Initialize a FIFO buffer for the audio samples to be encoded.
325 * @param[out] fifo Sample buffer
326 * @param output_codec_context Codec context of the output file
327 * @return Error code (0 if successful)
328 */
329static int init_fifo(AVAudioFifo **fifo, AVCodecContext *output_codec_context)
330{
331 /* Create the FIFO buffer based on the specified output sample format. */
332 if (!(*fifo = av_audio_fifo_alloc(output_codec_context->sample_fmt,
333 output_codec_context->ch_layout.nb_channels, 1))) {
334 fprintf(stderr, "Could not allocate FIFO\n");
335 return AVERROR(ENOMEM);
336 }
337 return 0;
338}
339
340/**
341 * Write the header of the output file container.
342 * @param output_format_context Format context of the output file
343 * @return Error code (0 if successful)
344 */
345static int write_output_file_header(AVFormatContext *output_format_context)
346{
347 int error;
348 if ((error = avformat_write_header(output_format_context, NULL)) < 0) {
349 fprintf(stderr, "Could not write output file header (error '%s')\n",
350 av_err2str(error));
351 return error;
352 }
353 return 0;
354}
355
356/**
357 * Decode one audio frame from the input file.
358 * @param frame Audio frame to be decoded
359 * @param input_format_context Format context of the input file
360 * @param input_codec_context Codec context of the input file
361 * @param[out] data_present Indicates whether data has been decoded
362 * @param[out] finished Indicates whether the end of file has
363 * been reached and all data has been
364 * decoded. If this flag is false, there
365 * is more data to be decoded, i.e., this
366 * function has to be called again.
367 * @return Error code (0 if successful)
368 */
370 AVFormatContext *input_format_context,
371 AVCodecContext *input_codec_context,
372 int *data_present, int *finished)
373{
374 /* Packet used for temporary storage. */
375 AVPacket *input_packet;
376 int error;
377
378 error = init_packet(&input_packet);
379 if (error < 0)
380 return error;
381
382 *data_present = 0;
383 *finished = 0;
384 /* Read one audio frame from the input file into a temporary packet. */
385 if ((error = av_read_frame(input_format_context, input_packet)) < 0) {
386 /* If we are at the end of the file, flush the decoder below. */
387 if (error == AVERROR_EOF)
388 *finished = 1;
389 else {
390 fprintf(stderr, "Could not read frame (error '%s')\n",
391 av_err2str(error));
392 goto cleanup;
393 }
394 }
395
396 /* Send the audio frame stored in the temporary packet to the decoder.
397 * The input audio stream decoder is used to do this. */
398 if ((error = avcodec_send_packet(input_codec_context, input_packet)) < 0) {
399 fprintf(stderr, "Could not send packet for decoding (error '%s')\n",
400 av_err2str(error));
401 goto cleanup;
402 }
403
404 /* Receive one frame from the decoder. */
405 error = avcodec_receive_frame(input_codec_context, frame);
406 /* If the decoder asks for more data to be able to decode a frame,
407 * return indicating that no data is present. */
408 if (error == AVERROR(EAGAIN)) {
409 error = 0;
410 goto cleanup;
411 /* If the end of the input file is reached, stop decoding. */
412 } else if (error == AVERROR_EOF) {
413 *finished = 1;
414 error = 0;
415 goto cleanup;
416 } else if (error < 0) {
417 fprintf(stderr, "Could not decode frame (error '%s')\n",
418 av_err2str(error));
419 goto cleanup;
420 /* Default case: Return decoded data. */
421 } else {
422 *data_present = 1;
423 goto cleanup;
424 }
425
426cleanup:
427 av_packet_free(&input_packet);
428 return error;
429}
430
431/**
432 * Initialize a temporary storage for the specified number of audio samples.
433 * The conversion requires temporary storage due to the different format.
434 * The number of audio samples to be allocated is specified in frame_size.
435 * @param[out] converted_input_samples Array of converted samples. The
436 * dimensions are reference, channel
437 * (for multi-channel audio), sample.
438 * @param output_codec_context Codec context of the output file
439 * @param frame_size Number of samples to be converted in
440 * each round
441 * @return Error code (0 if successful)
442 */
443static int init_converted_samples(uint8_t ***converted_input_samples,
444 AVCodecContext *output_codec_context,
445 int frame_size)
446{
447 int error;
448
449 /* Allocate as many pointers as there are audio channels.
450 * Each pointer will point to the audio samples of the corresponding
451 * channels (although it may be NULL for interleaved formats).
452 * Allocate memory for the samples of all channels in one consecutive
453 * block for convenience. */
454 if ((error = av_samples_alloc_array_and_samples(converted_input_samples, NULL,
455 output_codec_context->ch_layout.nb_channels,
456 frame_size,
457 output_codec_context->sample_fmt, 0)) < 0) {
458 fprintf(stderr,
459 "Could not allocate converted input samples (error '%s')\n",
460 av_err2str(error));
461 return error;
462 }
463 return 0;
464}
465
466/**
467 * Convert the input audio samples into the output sample format.
468 * The conversion happens on a per-frame basis, the size of which is
469 * specified by frame_size.
470 * @param input_data Samples to be decoded. The dimensions are
471 * channel (for multi-channel audio), sample.
472 * @param[out] converted_data Converted samples. The dimensions are channel
473 * (for multi-channel audio), sample.
474 * @param frame_size Number of samples to be converted
475 * @param resample_context Resample context for the conversion
476 * @return Error code (0 if successful)
477 */
478static int convert_samples(const uint8_t **input_data,
479 uint8_t **converted_data, const int frame_size,
480 SwrContext *resample_context)
481{
482 int error;
483
484 /* Convert the samples using the resampler. */
485 if ((error = swr_convert(resample_context,
486 converted_data, frame_size,
487 input_data , frame_size)) < 0) {
488 fprintf(stderr, "Could not convert input samples (error '%s')\n",
489 av_err2str(error));
490 return error;
491 }
492
493 return 0;
494}
495
496/**
497 * Add converted input audio samples to the FIFO buffer for later processing.
498 * @param fifo Buffer to add the samples to
499 * @param converted_input_samples Samples to be added. The dimensions are channel
500 * (for multi-channel audio), sample.
501 * @param frame_size Number of samples to be converted
502 * @return Error code (0 if successful)
503 */
505 uint8_t **converted_input_samples,
506 const int frame_size)
507{
508 int error;
509
510 /* Make the FIFO as large as it needs to be to hold both,
511 * the old and the new samples. */
512 if ((error = av_audio_fifo_realloc(fifo, av_audio_fifo_size(fifo) + frame_size)) < 0) {
513 fprintf(stderr, "Could not reallocate FIFO\n");
514 return error;
515 }
516
517 /* Store the new samples in the FIFO buffer. */
518 if (av_audio_fifo_write(fifo, (void **)converted_input_samples,
519 frame_size) < frame_size) {
520 fprintf(stderr, "Could not write data to FIFO\n");
521 return AVERROR_EXIT;
522 }
523 return 0;
524}
525
526/**
527 * Read one audio frame from the input file, decode, convert and store
528 * it in the FIFO buffer.
529 * @param fifo Buffer used for temporary storage
530 * @param input_format_context Format context of the input file
531 * @param input_codec_context Codec context of the input file
532 * @param output_codec_context Codec context of the output file
533 * @param resampler_context Resample context for the conversion
534 * @param[out] finished Indicates whether the end of file has
535 * been reached and all data has been
536 * decoded. If this flag is false,
537 * there is more data to be decoded,
538 * i.e., this function has to be called
539 * again.
540 * @return Error code (0 if successful)
541 */
543 AVFormatContext *input_format_context,
544 AVCodecContext *input_codec_context,
545 AVCodecContext *output_codec_context,
546 SwrContext *resampler_context,
547 int *finished)
548{
549 /* Temporary storage of the input samples of the frame read from the file. */
550 AVFrame *input_frame = NULL;
551 /* Temporary storage for the converted input samples. */
552 uint8_t **converted_input_samples = NULL;
553 int data_present;
554 int ret = AVERROR_EXIT;
555
556 /* Initialize temporary storage for one input frame. */
557 if (init_input_frame(&input_frame))
558 goto cleanup;
559 /* Decode one frame worth of audio samples. */
560 if (decode_audio_frame(input_frame, input_format_context,
561 input_codec_context, &data_present, finished))
562 goto cleanup;
563 /* If we are at the end of the file and there are no more samples
564 * in the decoder which are delayed, we are actually finished.
565 * This must not be treated as an error. */
566 if (*finished) {
567 ret = 0;
568 goto cleanup;
569 }
570 /* If there is decoded data, convert and store it. */
571 if (data_present) {
572 /* Initialize the temporary storage for the converted input samples. */
573 if (init_converted_samples(&converted_input_samples, output_codec_context,
574 input_frame->nb_samples))
575 goto cleanup;
576
577 /* Convert the input samples to the desired output sample format.
578 * This requires a temporary storage provided by converted_input_samples. */
579 if (convert_samples((const uint8_t**)input_frame->extended_data, converted_input_samples,
580 input_frame->nb_samples, resampler_context))
581 goto cleanup;
582
583 /* Add the converted input samples to the FIFO buffer for later processing. */
584 if (add_samples_to_fifo(fifo, converted_input_samples,
585 input_frame->nb_samples))
586 goto cleanup;
587 ret = 0;
588 }
589 ret = 0;
590
591cleanup:
592 if (converted_input_samples)
593 av_freep(&converted_input_samples[0]);
594 av_freep(&converted_input_samples);
595 av_frame_free(&input_frame);
596
597 return ret;
598}
599
600/**
601 * Initialize one input frame for writing to the output file.
602 * The frame will be exactly frame_size samples large.
603 * @param[out] frame Frame to be initialized
604 * @param output_codec_context Codec context of the output file
605 * @param frame_size Size of the frame
606 * @return Error code (0 if successful)
607 */
609 AVCodecContext *output_codec_context,
610 int frame_size)
611{
612 int error;
613
614 /* Create a new frame to store the audio samples. */
615 if (!(*frame = av_frame_alloc())) {
616 fprintf(stderr, "Could not allocate output frame\n");
617 return AVERROR_EXIT;
618 }
619
620 /* Set the frame's parameters, especially its size and format.
621 * av_frame_get_buffer needs this to allocate memory for the
622 * audio samples of the frame.
623 * Default channel layouts based on the number of channels
624 * are assumed for simplicity. */
625 (*frame)->nb_samples = frame_size;
626 av_channel_layout_copy(&(*frame)->ch_layout, &output_codec_context->ch_layout);
627 (*frame)->format = output_codec_context->sample_fmt;
628 (*frame)->sample_rate = output_codec_context->sample_rate;
629
630 /* Allocate the samples of the created frame. This call will make
631 * sure that the audio frame can hold as many samples as specified. */
632 if ((error = av_frame_get_buffer(*frame, 0)) < 0) {
633 fprintf(stderr, "Could not allocate output frame samples (error '%s')\n",
634 av_err2str(error));
636 return error;
637 }
638
639 return 0;
640}
641
642/* Global timestamp for the audio frames. */
643static int64_t pts = 0;
644
645/**
646 * Encode one frame worth of audio to the output file.
647 * @param frame Samples to be encoded
648 * @param output_format_context Format context of the output file
649 * @param output_codec_context Codec context of the output file
650 * @param[out] data_present Indicates whether data has been
651 * encoded
652 * @return Error code (0 if successful)
653 */
655 AVFormatContext *output_format_context,
656 AVCodecContext *output_codec_context,
657 int *data_present)
658{
659 /* Packet used for temporary storage. */
660 AVPacket *output_packet;
661 int error;
662
663 error = init_packet(&output_packet);
664 if (error < 0)
665 return error;
666
667 /* Set a timestamp based on the sample rate for the container. */
668 if (frame) {
669 frame->pts = pts;
670 pts += frame->nb_samples;
671 }
672
673 *data_present = 0;
674 /* Send the audio frame stored in the temporary packet to the encoder.
675 * The output audio stream encoder is used to do this. */
676 error = avcodec_send_frame(output_codec_context, frame);
677 /* Check for errors, but proceed with fetching encoded samples if the
678 * encoder signals that it has nothing more to encode. */
679 if (error < 0 && error != AVERROR_EOF) {
680 fprintf(stderr, "Could not send packet for encoding (error '%s')\n",
681 av_err2str(error));
682 goto cleanup;
683 }
684
685 /* Receive one encoded frame from the encoder. */
686 error = avcodec_receive_packet(output_codec_context, output_packet);
687 /* If the encoder asks for more data to be able to provide an
688 * encoded frame, return indicating that no data is present. */
689 if (error == AVERROR(EAGAIN)) {
690 error = 0;
691 goto cleanup;
692 /* If the last frame has been encoded, stop encoding. */
693 } else if (error == AVERROR_EOF) {
694 error = 0;
695 goto cleanup;
696 } else if (error < 0) {
697 fprintf(stderr, "Could not encode frame (error '%s')\n",
698 av_err2str(error));
699 goto cleanup;
700 /* Default case: Return encoded data. */
701 } else {
702 *data_present = 1;
703 }
704
705 /* Write one audio frame from the temporary packet to the output file. */
706 if (*data_present &&
707 (error = av_write_frame(output_format_context, output_packet)) < 0) {
708 fprintf(stderr, "Could not write frame (error '%s')\n",
709 av_err2str(error));
710 goto cleanup;
711 }
712
713cleanup:
714 av_packet_free(&output_packet);
715 return error;
716}
717
718/**
719 * Load one audio frame from the FIFO buffer, encode and write it to the
720 * output file.
721 * @param fifo Buffer used for temporary storage
722 * @param output_format_context Format context of the output file
723 * @param output_codec_context Codec context of the output file
724 * @return Error code (0 if successful)
725 */
727 AVFormatContext *output_format_context,
728 AVCodecContext *output_codec_context)
729{
730 /* Temporary storage of the output samples of the frame written to the file. */
731 AVFrame *output_frame;
732 /* Use the maximum number of possible samples per frame.
733 * If there is less than the maximum possible frame size in the FIFO
734 * buffer use this number. Otherwise, use the maximum possible frame size. */
735 const int frame_size = FFMIN(av_audio_fifo_size(fifo),
736 output_codec_context->frame_size);
737 int data_written;
738
739 /* Initialize temporary storage for one output frame. */
740 if (init_output_frame(&output_frame, output_codec_context, frame_size))
741 return AVERROR_EXIT;
742
743 /* Read as many samples from the FIFO buffer as required to fill the frame.
744 * The samples are stored in the frame temporarily. */
745 if (av_audio_fifo_read(fifo, (void **)output_frame->data, frame_size) < frame_size) {
746 fprintf(stderr, "Could not read data from FIFO\n");
747 av_frame_free(&output_frame);
748 return AVERROR_EXIT;
749 }
750
751 /* Encode one frame worth of audio samples. */
752 if (encode_audio_frame(output_frame, output_format_context,
753 output_codec_context, &data_written)) {
754 av_frame_free(&output_frame);
755 return AVERROR_EXIT;
756 }
757 av_frame_free(&output_frame);
758 return 0;
759}
760
761/**
762 * Write the trailer of the output file container.
763 * @param output_format_context Format context of the output file
764 * @return Error code (0 if successful)
765 */
766static int write_output_file_trailer(AVFormatContext *output_format_context)
767{
768 int error;
769 if ((error = av_write_trailer(output_format_context)) < 0) {
770 fprintf(stderr, "Could not write output file trailer (error '%s')\n",
771 av_err2str(error));
772 return error;
773 }
774 return 0;
775}
776
777int main(int argc, char **argv)
778{
779 AVFormatContext *input_format_context = NULL, *output_format_context = NULL;
780 AVCodecContext *input_codec_context = NULL, *output_codec_context = NULL;
781 SwrContext *resample_context = NULL;
782 AVAudioFifo *fifo = NULL;
783 int ret = AVERROR_EXIT;
784
785 if (argc != 3) {
786 fprintf(stderr, "Usage: %s <input file> <output file>\n", argv[0]);
787 exit(1);
788 }
789
790 /* Open the input file for reading. */
791 if (open_input_file(argv[1], &input_format_context,
792 &input_codec_context))
793 goto cleanup;
794 /* Open the output file for writing. */
795 if (open_output_file(argv[2], input_codec_context,
796 &output_format_context, &output_codec_context))
797 goto cleanup;
798 /* Initialize the resampler to be able to convert audio sample formats. */
799 if (init_resampler(input_codec_context, output_codec_context,
800 &resample_context))
801 goto cleanup;
802 /* Initialize the FIFO buffer to store audio samples to be encoded. */
803 if (init_fifo(&fifo, output_codec_context))
804 goto cleanup;
805 /* Write the header of the output file container. */
806 if (write_output_file_header(output_format_context))
807 goto cleanup;
808
809 /* Loop as long as we have input samples to read or output samples
810 * to write; abort as soon as we have neither. */
811 while (1) {
812 /* Use the encoder's desired frame size for processing. */
813 const int output_frame_size = output_codec_context->frame_size;
814 int finished = 0;
815
816 /* Make sure that there is one frame worth of samples in the FIFO
817 * buffer so that the encoder can do its work.
818 * Since the decoder's and the encoder's frame size may differ, we
819 * need to FIFO buffer to store as many frames worth of input samples
820 * that they make up at least one frame worth of output samples. */
821 while (av_audio_fifo_size(fifo) < output_frame_size) {
822 /* Decode one frame worth of audio samples, convert it to the
823 * output sample format and put it into the FIFO buffer. */
824 if (read_decode_convert_and_store(fifo, input_format_context,
825 input_codec_context,
826 output_codec_context,
827 resample_context, &finished))
828 goto cleanup;
829
830 /* If we are at the end of the input file, we continue
831 * encoding the remaining audio samples to the output file. */
832 if (finished)
833 break;
834 }
835
836 /* If we have enough samples for the encoder, we encode them.
837 * At the end of the file, we pass the remaining samples to
838 * the encoder. */
839 while (av_audio_fifo_size(fifo) >= output_frame_size ||
840 (finished && av_audio_fifo_size(fifo) > 0))
841 /* Take one frame worth of audio samples from the FIFO buffer,
842 * encode it and write it to the output file. */
843 if (load_encode_and_write(fifo, output_format_context,
844 output_codec_context))
845 goto cleanup;
846
847 /* If we are at the end of the input file and have encoded
848 * all remaining samples, we can exit this loop and finish. */
849 if (finished) {
850 int data_written;
851 /* Flush the encoder as it may have delayed frames. */
852 do {
853 if (encode_audio_frame(NULL, output_format_context,
854 output_codec_context, &data_written))
855 goto cleanup;
856 } while (data_written);
857 break;
858 }
859 }
860
861 /* Write the trailer of the output file container. */
862 if (write_output_file_trailer(output_format_context))
863 goto cleanup;
864 ret = 0;
865
866cleanup:
867 if (fifo)
868 av_audio_fifo_free(fifo);
869 swr_free(&resample_context);
870 if (output_codec_context)
871 avcodec_free_context(&output_codec_context);
872 if (output_format_context) {
873 avio_closep(&output_format_context->pb);
874 avformat_free_context(output_format_context);
875 }
876 if (input_codec_context)
877 avcodec_free_context(&input_codec_context);
878 if (input_format_context)
879 avformat_close_input(&input_format_context);
880
881 return ret;
882}
Audio FIFO Buffer.
simple assert() macros that are a bit more flexible than ISO C assert().
#define av_assert0(cond)
assert() equivalent, that is always enabled.
Definition: avassert.h:40
Libavcodec external API header.
Main libavformat public API header.
#define AVFMT_GLOBALHEADER
Format wants global header.
Definition: avformat.h:479
Buffered I/O operations.
int avio_open(AVIOContext **s, const char *url, int flags)
Create and initialize a AVIOContext for accessing the resource indicated by url.
#define AVIO_FLAG_WRITE
write-only
Definition: avio.h:637
int avio_closep(AVIOContext **s)
Close the resource accessed by the AVIOContext *s, free it and set the pointer pointing to it to NULL...
Public libavutil channel layout APIs header.
static AVFrame * frame
Definition: demux_decode.c:54
reference-counted frame API
int avcodec_open2(AVCodecContext *avctx, const AVCodec *codec, AVDictionary **options)
Initialize the AVCodecContext to use the given AVCodec.
AVCodecContext * avcodec_alloc_context3(const AVCodec *codec)
Allocate an AVCodecContext and set its fields to default values.
int avcodec_parameters_from_context(struct AVCodecParameters *par, const AVCodecContext *codec)
Fill the parameters struct based on the values from the supplied codec context.
const AVCodec * avcodec_find_decoder(enum AVCodecID id)
Find a registered decoder with a matching codec ID.
int avcodec_parameters_to_context(AVCodecContext *codec, const struct AVCodecParameters *par)
Fill the codec context based on the values from the supplied codec parameters.
const AVCodec * avcodec_find_encoder(enum AVCodecID id)
Find a registered encoder with a matching codec ID.
#define AV_CODEC_FLAG_GLOBAL_HEADER
Place global headers in extradata instead of every keyframe.
Definition: avcodec.h:334
void avcodec_free_context(AVCodecContext **avctx)
Free the codec context and everything associated with it and write NULL to the provided pointer.
@ AV_CODEC_ID_AAC
Definition: codec_id.h:444
int avcodec_receive_frame(AVCodecContext *avctx, AVFrame *frame)
Return decoded output data from a decoder or encoder (when the AV_CODEC_FLAG_RECON_FRAME flag is used...
int avcodec_send_packet(AVCodecContext *avctx, const AVPacket *avpkt)
Supply raw packet data as input to a decoder.
int avcodec_receive_packet(AVCodecContext *avctx, AVPacket *avpkt)
Read encoded data from the encoder.
int avcodec_send_frame(AVCodecContext *avctx, const AVFrame *frame)
Supply a raw video or audio frame to the encoder.
void av_packet_free(AVPacket **pkt)
Free the packet, if the packet is reference counted, it will be unreferenced first.
AVPacket * av_packet_alloc(void)
Allocate an AVPacket and set its fields to default values.
AVStream * avformat_new_stream(AVFormatContext *s, const struct AVCodec *c)
Add a new stream to a media file.
AVFormatContext * avformat_alloc_context(void)
Allocate an AVFormatContext.
void avformat_free_context(AVFormatContext *s)
Free an AVFormatContext and all its streams.
int av_read_frame(AVFormatContext *s, AVPacket *pkt)
Return the next frame of a stream.
int avformat_open_input(AVFormatContext **ps, const char *url, const AVInputFormat *fmt, AVDictionary **options)
Open an input stream and read the header.
int avformat_find_stream_info(AVFormatContext *ic, AVDictionary **options)
Read packets of a media file to get stream information.
void avformat_close_input(AVFormatContext **s)
Close an opened input AVFormatContext.
av_warn_unused_result int avformat_write_header(AVFormatContext *s, AVDictionary **options)
Allocate the stream private data and write the stream header to an output media file.
int av_write_trailer(AVFormatContext *s)
Write the stream trailer to an output media file and free the file private data.
const AVOutputFormat * av_guess_format(const char *short_name, const char *filename, const char *mime_type)
Return the output format in the list of registered output formats which best matches the provided par...
int av_write_frame(AVFormatContext *s, AVPacket *pkt)
Write a packet to an output media file.
void av_channel_layout_default(AVChannelLayout *ch_layout, int nb_channels)
Get the default channel layout for a given number of channels.
int av_channel_layout_copy(AVChannelLayout *dst, const AVChannelLayout *src)
Make a copy of a channel layout.
AVAudioFifo * av_audio_fifo_alloc(enum AVSampleFormat sample_fmt, int channels, int nb_samples)
Allocate an AVAudioFifo.
av_warn_unused_result int av_audio_fifo_realloc(AVAudioFifo *af, int nb_samples)
Reallocate an AVAudioFifo.
struct AVAudioFifo AVAudioFifo
Context for an Audio FIFO Buffer.
Definition: audio_fifo.h:48
int av_audio_fifo_write(AVAudioFifo *af, void *const *data, int nb_samples)
Write data to an AVAudioFifo.
int av_audio_fifo_read(AVAudioFifo *af, void *const *data, int nb_samples)
Read data from an AVAudioFifo.
void av_audio_fifo_free(AVAudioFifo *af)
Free an AVAudioFifo.
int av_audio_fifo_size(AVAudioFifo *af)
Get the current number of samples in the AVAudioFifo available for reading.
#define AVERROR_EXIT
Immediate exit was requested; the called function should not be restarted.
Definition: error.h:58
#define AVERROR_EOF
End of file.
Definition: error.h:57
#define av_err2str(errnum)
Convenience macro, the return value should be used only directly in function arguments but never stan...
Definition: error.h:121
#define AVERROR(e)
Definition: error.h:45
int av_frame_get_buffer(AVFrame *frame, int align)
Allocate new buffer(s) for audio or video data.
void av_frame_free(AVFrame **frame)
Free the frame and any dynamically allocated objects in it, e.g.
AVFrame * av_frame_alloc(void)
Allocate an AVFrame and set its fields to default values.
void av_freep(void *ptr)
Free a memory block which has been allocated with a function of av_malloc() or av_realloc() family,...
char * av_strdup(const char *s) av_malloc_attrib
Duplicate a string.
int av_samples_alloc_array_and_samples(uint8_t ***audio_data, int *linesize, int nb_channels, int nb_samples, enum AVSampleFormat sample_fmt, int align)
Allocate a data pointers array, samples buffer for nb_samples samples, and fill data pointers and lin...
struct SwrContext SwrContext
The libswresample context.
Definition: swresample.h:189
int swr_alloc_set_opts2(struct SwrContext **ps, const AVChannelLayout *out_ch_layout, enum AVSampleFormat out_sample_fmt, int out_sample_rate, const AVChannelLayout *in_ch_layout, enum AVSampleFormat in_sample_fmt, int in_sample_rate, int log_offset, void *log_ctx)
Allocate SwrContext if needed and set/reset common parameters.
void swr_free(struct SwrContext **s)
Free the given SwrContext and set the pointer to NULL.
int swr_convert(struct SwrContext *s, uint8_t **out, int out_count, const uint8_t **in, int in_count)
Convert audio.
int swr_init(struct SwrContext *s)
Initialize context after user parameters have been set.
#define FFMIN(a, b)
Definition: macros.h:49
AVOptions.
int nb_channels
Number of channels in this layout.
main external API structure.
Definition: avcodec.h:441
AVChannelLayout ch_layout
Audio channel layout.
Definition: avcodec.h:2107
enum AVSampleFormat sample_fmt
audio sample format
Definition: avcodec.h:1080
AVRational pkt_timebase
Timebase in which pkt_dts/pts and AVPacket.dts/pts are expressed.
Definition: avcodec.h:1817
int64_t bit_rate
the average bitrate
Definition: avcodec.h:491
int sample_rate
samples per second
Definition: avcodec.h:1064
int flags
AV_CODEC_FLAG_*.
Definition: avcodec.h:521
int frame_size
Number of samples per channel in an audio frame.
Definition: avcodec.h:1092
enum AVCodecID codec_id
Specific type of the encoded data (the codec used).
Definition: codec_par.h:55
AVCodec.
Definition: codec.h:187
enum AVSampleFormat * sample_fmts
array of supported sample formats, or NULL if unknown, array is terminated by -1
Definition: codec.h:211
Format I/O context.
Definition: avformat.h:1115
This structure describes decoded (raw) audio or video data.
Definition: frame.h:340
int nb_samples
number of audio samples (per channel) described by this frame
Definition: frame.h:420
int64_t pts
Presentation timestamp in time_base units (time when frame should be shown to user).
Definition: frame.h:452
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: frame.h:361
uint8_t ** extended_data
pointers to the data planes/channels.
Definition: frame.h:401
Bytestream IO Context.
Definition: avio.h:166
This structure stores compressed data.
Definition: packet.h:468
int num
Numerator.
Definition: rational.h:59
int den
Denominator.
Definition: rational.h:60
Stream structure.
Definition: avformat.h:841
AVCodecParameters * codecpar
Codec parameters associated with this stream.
Definition: avformat.h:864
AVRational time_base
This is the fundamental unit of time (in seconds) in terms of which frame timestamps are represented.
Definition: avformat.h:880
libswresample public header
static int add_samples_to_fifo(AVAudioFifo *fifo, uint8_t **converted_input_samples, const int frame_size)
Add converted input audio samples to the FIFO buffer for later processing.
static int init_input_frame(AVFrame **frame)
Initialize one audio frame for reading from the input file.
int main(int argc, char **argv)
static int encode_audio_frame(AVFrame *frame, AVFormatContext *output_format_context, AVCodecContext *output_codec_context, int *data_present)
Encode one frame worth of audio to the output file.
#define OUTPUT_BIT_RATE
Definition: transcode_aac.c:47
#define OUTPUT_CHANNELS
Definition: transcode_aac.c:49
static int64_t pts
static int decode_audio_frame(AVFrame *frame, AVFormatContext *input_format_context, AVCodecContext *input_codec_context, int *data_present, int *finished)
Decode one audio frame from the input file.
static int open_output_file(const char *filename, AVCodecContext *input_codec_context, AVFormatContext **output_format_context, AVCodecContext **output_codec_context)
Open an output file and the required encoder.
static int init_resampler(AVCodecContext *input_codec_context, AVCodecContext *output_codec_context, SwrContext **resample_context)
Initialize the audio resampler based on the input and output codec settings.
static int init_packet(AVPacket **packet)
Initialize one data packet for reading or writing.
static int write_output_file_header(AVFormatContext *output_format_context)
Write the header of the output file container.
static int init_output_frame(AVFrame **frame, AVCodecContext *output_codec_context, int frame_size)
Initialize one input frame for writing to the output file.
static int read_decode_convert_and_store(AVAudioFifo *fifo, AVFormatContext *input_format_context, AVCodecContext *input_codec_context, AVCodecContext *output_codec_context, SwrContext *resampler_context, int *finished)
Read one audio frame from the input file, decode, convert and store it in the FIFO buffer.
static int open_input_file(const char *filename, AVFormatContext **input_format_context, AVCodecContext **input_codec_context)
Open an input file and the required decoder.
Definition: transcode_aac.c:58
static int init_fifo(AVAudioFifo **fifo, AVCodecContext *output_codec_context)
Initialize a FIFO buffer for the audio samples to be encoded.
static int load_encode_and_write(AVAudioFifo *fifo, AVFormatContext *output_format_context, AVCodecContext *output_codec_context)
Load one audio frame from the FIFO buffer, encode and write it to the output file.
static int write_output_file_trailer(AVFormatContext *output_format_context)
Write the trailer of the output file container.
static int convert_samples(const uint8_t **input_data, uint8_t **converted_data, const int frame_size, SwrContext *resample_context)
Convert the input audio samples into the output sample format.
static int init_converted_samples(uint8_t ***converted_input_samples, AVCodecContext *output_codec_context, int frame_size)
Initialize a temporary storage for the specified number of audio samples.