FFmpeg
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 
32 #include "libavformat/avformat.h"
33 #include "libavformat/avio.h"
34 
35 #include "libavcodec/avcodec.h"
36 
37 #include "libavutil/audio_fifo.h"
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  */
58 static 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  */
145 static 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 
240 cleanup:
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  */
253 static 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  */
285 static 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  */
329 static 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  */
345 static 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 
426 cleanup:
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  */
443 static 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 later point to the audio samples of the corresponding
451  * channels (although it may be NULL for interleaved formats).
452  */
453  if (!(*converted_input_samples = calloc(output_codec_context->ch_layout.nb_channels,
454  sizeof(**converted_input_samples)))) {
455  fprintf(stderr, "Could not allocate converted input sample pointers\n");
456  return AVERROR(ENOMEM);
457  }
458 
459  /* Allocate memory for the samples of all channels in one consecutive
460  * block for convenience. */
461  if ((error = av_samples_alloc(*converted_input_samples, NULL,
462  output_codec_context->ch_layout.nb_channels,
463  frame_size,
464  output_codec_context->sample_fmt, 0)) < 0) {
465  fprintf(stderr,
466  "Could not allocate converted input samples (error '%s')\n",
467  av_err2str(error));
468  av_freep(&(*converted_input_samples)[0]);
469  free(*converted_input_samples);
470  return error;
471  }
472  return 0;
473 }
474 
475 /**
476  * Convert the input audio samples into the output sample format.
477  * The conversion happens on a per-frame basis, the size of which is
478  * specified by frame_size.
479  * @param input_data Samples to be decoded. The dimensions are
480  * channel (for multi-channel audio), sample.
481  * @param[out] converted_data Converted samples. The dimensions are channel
482  * (for multi-channel audio), sample.
483  * @param frame_size Number of samples to be converted
484  * @param resample_context Resample context for the conversion
485  * @return Error code (0 if successful)
486  */
487 static int convert_samples(const uint8_t **input_data,
488  uint8_t **converted_data, const int frame_size,
489  SwrContext *resample_context)
490 {
491  int error;
492 
493  /* Convert the samples using the resampler. */
494  if ((error = swr_convert(resample_context,
495  converted_data, frame_size,
496  input_data , frame_size)) < 0) {
497  fprintf(stderr, "Could not convert input samples (error '%s')\n",
498  av_err2str(error));
499  return error;
500  }
501 
502  return 0;
503 }
504 
505 /**
506  * Add converted input audio samples to the FIFO buffer for later processing.
507  * @param fifo Buffer to add the samples to
508  * @param converted_input_samples Samples to be added. The dimensions are channel
509  * (for multi-channel audio), sample.
510  * @param frame_size Number of samples to be converted
511  * @return Error code (0 if successful)
512  */
514  uint8_t **converted_input_samples,
515  const int frame_size)
516 {
517  int error;
518 
519  /* Make the FIFO as large as it needs to be to hold both,
520  * the old and the new samples. */
521  if ((error = av_audio_fifo_realloc(fifo, av_audio_fifo_size(fifo) + frame_size)) < 0) {
522  fprintf(stderr, "Could not reallocate FIFO\n");
523  return error;
524  }
525 
526  /* Store the new samples in the FIFO buffer. */
527  if (av_audio_fifo_write(fifo, (void **)converted_input_samples,
528  frame_size) < frame_size) {
529  fprintf(stderr, "Could not write data to FIFO\n");
530  return AVERROR_EXIT;
531  }
532  return 0;
533 }
534 
535 /**
536  * Read one audio frame from the input file, decode, convert and store
537  * it in the FIFO buffer.
538  * @param fifo Buffer used for temporary storage
539  * @param input_format_context Format context of the input file
540  * @param input_codec_context Codec context of the input file
541  * @param output_codec_context Codec context of the output file
542  * @param resampler_context Resample context for the conversion
543  * @param[out] finished Indicates whether the end of file has
544  * been reached and all data has been
545  * decoded. If this flag is false,
546  * there is more data to be decoded,
547  * i.e., this function has to be called
548  * again.
549  * @return Error code (0 if successful)
550  */
552  AVFormatContext *input_format_context,
553  AVCodecContext *input_codec_context,
554  AVCodecContext *output_codec_context,
555  SwrContext *resampler_context,
556  int *finished)
557 {
558  /* Temporary storage of the input samples of the frame read from the file. */
559  AVFrame *input_frame = NULL;
560  /* Temporary storage for the converted input samples. */
561  uint8_t **converted_input_samples = NULL;
562  int data_present;
563  int ret = AVERROR_EXIT;
564 
565  /* Initialize temporary storage for one input frame. */
566  if (init_input_frame(&input_frame))
567  goto cleanup;
568  /* Decode one frame worth of audio samples. */
569  if (decode_audio_frame(input_frame, input_format_context,
570  input_codec_context, &data_present, finished))
571  goto cleanup;
572  /* If we are at the end of the file and there are no more samples
573  * in the decoder which are delayed, we are actually finished.
574  * This must not be treated as an error. */
575  if (*finished) {
576  ret = 0;
577  goto cleanup;
578  }
579  /* If there is decoded data, convert and store it. */
580  if (data_present) {
581  /* Initialize the temporary storage for the converted input samples. */
582  if (init_converted_samples(&converted_input_samples, output_codec_context,
583  input_frame->nb_samples))
584  goto cleanup;
585 
586  /* Convert the input samples to the desired output sample format.
587  * This requires a temporary storage provided by converted_input_samples. */
588  if (convert_samples((const uint8_t**)input_frame->extended_data, converted_input_samples,
589  input_frame->nb_samples, resampler_context))
590  goto cleanup;
591 
592  /* Add the converted input samples to the FIFO buffer for later processing. */
593  if (add_samples_to_fifo(fifo, converted_input_samples,
594  input_frame->nb_samples))
595  goto cleanup;
596  ret = 0;
597  }
598  ret = 0;
599 
600 cleanup:
601  if (converted_input_samples) {
602  av_freep(&converted_input_samples[0]);
603  free(converted_input_samples);
604  }
605  av_frame_free(&input_frame);
606 
607  return ret;
608 }
609 
610 /**
611  * Initialize one input frame for writing to the output file.
612  * The frame will be exactly frame_size samples large.
613  * @param[out] frame Frame to be initialized
614  * @param output_codec_context Codec context of the output file
615  * @param frame_size Size of the frame
616  * @return Error code (0 if successful)
617  */
619  AVCodecContext *output_codec_context,
620  int frame_size)
621 {
622  int error;
623 
624  /* Create a new frame to store the audio samples. */
625  if (!(*frame = av_frame_alloc())) {
626  fprintf(stderr, "Could not allocate output frame\n");
627  return AVERROR_EXIT;
628  }
629 
630  /* Set the frame's parameters, especially its size and format.
631  * av_frame_get_buffer needs this to allocate memory for the
632  * audio samples of the frame.
633  * Default channel layouts based on the number of channels
634  * are assumed for simplicity. */
635  (*frame)->nb_samples = frame_size;
636  av_channel_layout_copy(&(*frame)->ch_layout, &output_codec_context->ch_layout);
637  (*frame)->format = output_codec_context->sample_fmt;
638  (*frame)->sample_rate = output_codec_context->sample_rate;
639 
640  /* Allocate the samples of the created frame. This call will make
641  * sure that the audio frame can hold as many samples as specified. */
642  if ((error = av_frame_get_buffer(*frame, 0)) < 0) {
643  fprintf(stderr, "Could not allocate output frame samples (error '%s')\n",
644  av_err2str(error));
646  return error;
647  }
648 
649  return 0;
650 }
651 
652 /* Global timestamp for the audio frames. */
653 static int64_t pts = 0;
654 
655 /**
656  * Encode one frame worth of audio to the output file.
657  * @param frame Samples to be encoded
658  * @param output_format_context Format context of the output file
659  * @param output_codec_context Codec context of the output file
660  * @param[out] data_present Indicates whether data has been
661  * encoded
662  * @return Error code (0 if successful)
663  */
665  AVFormatContext *output_format_context,
666  AVCodecContext *output_codec_context,
667  int *data_present)
668 {
669  /* Packet used for temporary storage. */
670  AVPacket *output_packet;
671  int error;
672 
673  error = init_packet(&output_packet);
674  if (error < 0)
675  return error;
676 
677  /* Set a timestamp based on the sample rate for the container. */
678  if (frame) {
679  frame->pts = pts;
680  pts += frame->nb_samples;
681  }
682 
683  *data_present = 0;
684  /* Send the audio frame stored in the temporary packet to the encoder.
685  * The output audio stream encoder is used to do this. */
686  error = avcodec_send_frame(output_codec_context, frame);
687  /* Check for errors, but proceed with fetching encoded samples if the
688  * encoder signals that it has nothing more to encode. */
689  if (error < 0 && error != AVERROR_EOF) {
690  fprintf(stderr, "Could not send packet for encoding (error '%s')\n",
691  av_err2str(error));
692  goto cleanup;
693  }
694 
695  /* Receive one encoded frame from the encoder. */
696  error = avcodec_receive_packet(output_codec_context, output_packet);
697  /* If the encoder asks for more data to be able to provide an
698  * encoded frame, return indicating that no data is present. */
699  if (error == AVERROR(EAGAIN)) {
700  error = 0;
701  goto cleanup;
702  /* If the last frame has been encoded, stop encoding. */
703  } else if (error == AVERROR_EOF) {
704  error = 0;
705  goto cleanup;
706  } else if (error < 0) {
707  fprintf(stderr, "Could not encode frame (error '%s')\n",
708  av_err2str(error));
709  goto cleanup;
710  /* Default case: Return encoded data. */
711  } else {
712  *data_present = 1;
713  }
714 
715  /* Write one audio frame from the temporary packet to the output file. */
716  if (*data_present &&
717  (error = av_write_frame(output_format_context, output_packet)) < 0) {
718  fprintf(stderr, "Could not write frame (error '%s')\n",
719  av_err2str(error));
720  goto cleanup;
721  }
722 
723 cleanup:
724  av_packet_free(&output_packet);
725  return error;
726 }
727 
728 /**
729  * Load one audio frame from the FIFO buffer, encode and write it to the
730  * output file.
731  * @param fifo Buffer used for temporary storage
732  * @param output_format_context Format context of the output file
733  * @param output_codec_context Codec context of the output file
734  * @return Error code (0 if successful)
735  */
737  AVFormatContext *output_format_context,
738  AVCodecContext *output_codec_context)
739 {
740  /* Temporary storage of the output samples of the frame written to the file. */
741  AVFrame *output_frame;
742  /* Use the maximum number of possible samples per frame.
743  * If there is less than the maximum possible frame size in the FIFO
744  * buffer use this number. Otherwise, use the maximum possible frame size. */
745  const int frame_size = FFMIN(av_audio_fifo_size(fifo),
746  output_codec_context->frame_size);
747  int data_written;
748 
749  /* Initialize temporary storage for one output frame. */
750  if (init_output_frame(&output_frame, output_codec_context, frame_size))
751  return AVERROR_EXIT;
752 
753  /* Read as many samples from the FIFO buffer as required to fill the frame.
754  * The samples are stored in the frame temporarily. */
755  if (av_audio_fifo_read(fifo, (void **)output_frame->data, frame_size) < frame_size) {
756  fprintf(stderr, "Could not read data from FIFO\n");
757  av_frame_free(&output_frame);
758  return AVERROR_EXIT;
759  }
760 
761  /* Encode one frame worth of audio samples. */
762  if (encode_audio_frame(output_frame, output_format_context,
763  output_codec_context, &data_written)) {
764  av_frame_free(&output_frame);
765  return AVERROR_EXIT;
766  }
767  av_frame_free(&output_frame);
768  return 0;
769 }
770 
771 /**
772  * Write the trailer of the output file container.
773  * @param output_format_context Format context of the output file
774  * @return Error code (0 if successful)
775  */
776 static int write_output_file_trailer(AVFormatContext *output_format_context)
777 {
778  int error;
779  if ((error = av_write_trailer(output_format_context)) < 0) {
780  fprintf(stderr, "Could not write output file trailer (error '%s')\n",
781  av_err2str(error));
782  return error;
783  }
784  return 0;
785 }
786 
787 int main(int argc, char **argv)
788 {
789  AVFormatContext *input_format_context = NULL, *output_format_context = NULL;
790  AVCodecContext *input_codec_context = NULL, *output_codec_context = NULL;
791  SwrContext *resample_context = NULL;
792  AVAudioFifo *fifo = NULL;
793  int ret = AVERROR_EXIT;
794 
795  if (argc != 3) {
796  fprintf(stderr, "Usage: %s <input file> <output file>\n", argv[0]);
797  exit(1);
798  }
799 
800  /* Open the input file for reading. */
801  if (open_input_file(argv[1], &input_format_context,
802  &input_codec_context))
803  goto cleanup;
804  /* Open the output file for writing. */
805  if (open_output_file(argv[2], input_codec_context,
806  &output_format_context, &output_codec_context))
807  goto cleanup;
808  /* Initialize the resampler to be able to convert audio sample formats. */
809  if (init_resampler(input_codec_context, output_codec_context,
810  &resample_context))
811  goto cleanup;
812  /* Initialize the FIFO buffer to store audio samples to be encoded. */
813  if (init_fifo(&fifo, output_codec_context))
814  goto cleanup;
815  /* Write the header of the output file container. */
816  if (write_output_file_header(output_format_context))
817  goto cleanup;
818 
819  /* Loop as long as we have input samples to read or output samples
820  * to write; abort as soon as we have neither. */
821  while (1) {
822  /* Use the encoder's desired frame size for processing. */
823  const int output_frame_size = output_codec_context->frame_size;
824  int finished = 0;
825 
826  /* Make sure that there is one frame worth of samples in the FIFO
827  * buffer so that the encoder can do its work.
828  * Since the decoder's and the encoder's frame size may differ, we
829  * need to FIFO buffer to store as many frames worth of input samples
830  * that they make up at least one frame worth of output samples. */
831  while (av_audio_fifo_size(fifo) < output_frame_size) {
832  /* Decode one frame worth of audio samples, convert it to the
833  * output sample format and put it into the FIFO buffer. */
834  if (read_decode_convert_and_store(fifo, input_format_context,
835  input_codec_context,
836  output_codec_context,
837  resample_context, &finished))
838  goto cleanup;
839 
840  /* If we are at the end of the input file, we continue
841  * encoding the remaining audio samples to the output file. */
842  if (finished)
843  break;
844  }
845 
846  /* If we have enough samples for the encoder, we encode them.
847  * At the end of the file, we pass the remaining samples to
848  * the encoder. */
849  while (av_audio_fifo_size(fifo) >= output_frame_size ||
850  (finished && av_audio_fifo_size(fifo) > 0))
851  /* Take one frame worth of audio samples from the FIFO buffer,
852  * encode it and write it to the output file. */
853  if (load_encode_and_write(fifo, output_format_context,
854  output_codec_context))
855  goto cleanup;
856 
857  /* If we are at the end of the input file and have encoded
858  * all remaining samples, we can exit this loop and finish. */
859  if (finished) {
860  int data_written;
861  /* Flush the encoder as it may have delayed frames. */
862  do {
863  if (encode_audio_frame(NULL, output_format_context,
864  output_codec_context, &data_written))
865  goto cleanup;
866  } while (data_written);
867  break;
868  }
869  }
870 
871  /* Write the trailer of the output file container. */
872  if (write_output_file_trailer(output_format_context))
873  goto cleanup;
874  ret = 0;
875 
876 cleanup:
877  if (fifo)
878  av_audio_fifo_free(fifo);
879  swr_free(&resample_context);
880  if (output_codec_context)
881  avcodec_free_context(&output_codec_context);
882  if (output_format_context) {
883  avio_closep(&output_format_context->pb);
884  avformat_free_context(output_format_context);
885  }
886  if (input_codec_context)
887  avcodec_free_context(&input_codec_context);
888  if (input_format_context)
889  avformat_close_input(&input_format_context);
890 
891  return ret;
892 }
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:37
Libavcodec external API header.
Main libavformat public API header.
#define AVFMT_GLOBALHEADER
Format wants global header.
Definition: avformat.h:478
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:624
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_parameters_from_context(AVCodecParameters *par, const AVCodecContext *codec)
Fill the parameters struct based on the values from the supplied codec context.
int avcodec_open2(AVCodecContext *avctx, const AVCodec *codec, AVDictionary **options)
Initialize the AVCodecContext to use the given AVCodec.
const AVCodec * avcodec_find_decoder(enum AVCodecID id)
Find a registered decoder with a matching codec ID.
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:317
int avcodec_parameters_to_context(AVCodecContext *codec, const AVCodecParameters *par)
Fill the codec context based on the values from the supplied codec parameters.
AVCodecContext * avcodec_alloc_context3(const AVCodec *codec)
Allocate an AVCodecContext and set its fields to default values.
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:440
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.
void avformat_free_context(AVFormatContext *s)
Free an AVFormatContext and all its streams.
AVFormatContext * avformat_alloc_context(void)
Allocate an AVFormatContext.
AVStream * avformat_new_stream(AVFormatContext *s, const AVCodec *c)
Add a new stream to a media file.
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.
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...
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.
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.
int av_audio_fifo_write(AVAudioFifo *af, void **data, int nb_samples)
Write data to 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
void av_audio_fifo_free(AVAudioFifo *af)
Free an AVAudioFifo.
AVAudioFifo * av_audio_fifo_alloc(enum AVSampleFormat sample_fmt, int channels, int nb_samples)
Allocate an AVAudioFifo.
int av_audio_fifo_size(AVAudioFifo *af)
Get the current number of samples in the AVAudioFifo available for reading.
int av_audio_fifo_read(AVAudioFifo *af, void **data, int nb_samples)
Read data from an AVAudioFifo.
#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(uint8_t **audio_data, int *linesize, int nb_channels, int nb_samples, enum AVSampleFormat sample_fmt, int align)
Allocate a samples buffer for nb_samples samples, and fill data pointers and linesize accordingly.
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:426
AVChannelLayout ch_layout
Audio channel layout.
Definition: avcodec.h:2054
enum AVSampleFormat sample_fmt
audio sample format
Definition: avcodec.h:1050
AVRational pkt_timebase
Timebase in which pkt_dts/pts and AVPacket.dts/pts are.
Definition: avcodec.h:1764
int64_t bit_rate
the average bitrate
Definition: avcodec.h:476
int sample_rate
samples per second
Definition: avcodec.h:1034
int flags
AV_CODEC_FLAG_*.
Definition: avcodec.h:506
int frame_size
Number of samples per channel in an audio frame.
Definition: avcodec.h:1062
enum AVCodecID codec_id
Specific type of the encoded data (the codec used).
Definition: codec_par.h:62
AVCodec.
Definition: codec.h:184
enum AVSampleFormat * sample_fmts
array of supported sample formats, or NULL if unknown, array is terminated by -1
Definition: codec.h:208
Format I/O context.
Definition: avformat.h:1104
This structure describes decoded (raw) audio or video data.
Definition: frame.h:330
int nb_samples
number of audio samples (per channel) described by this frame
Definition: frame.h:410
int64_t pts
Presentation timestamp in time_base units (time when frame should be shown to user).
Definition: frame.h:437
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: frame.h:351
uint8_t ** extended_data
pointers to the data planes/channels.
Definition: frame.h:391
Bytestream IO Context.
Definition: avio.h:166
This structure stores compressed data.
Definition: packet.h:351
int num
Numerator.
Definition: rational.h:59
int den
Denominator.
Definition: rational.h:60
Stream structure.
Definition: avformat.h:838
AVCodecParameters * codecpar
Codec parameters associated with this stream.
Definition: avformat.h:861
AVRational time_base
This is the fundamental unit of time (in seconds) in terms of which frame timestamps are represented.
Definition: avformat.h:877
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.