FFmpeg  4.4.1
transcode_aac.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2013-2018 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
23  * Simple audio converter
24  *
25  * @example transcode_aac.c
26  * Convert an input audio file to AAC in an MP4 container using FFmpeg.
27  * Formats other than MP4 are supported based on the output file extension.
28  * @author Andreas Unterweger (dustsigns@gmail.com)
29  */
30 
31 #include <stdio.h>
32 
33 #include "libavformat/avformat.h"
34 #include "libavformat/avio.h"
35 
36 #include "libavcodec/avcodec.h"
37 
38 #include "libavutil/audio_fifo.h"
39 #include "libavutil/avassert.h"
40 #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  AVCodec *input_codec;
64  int error;
65 
66  /* Open the input file to read from it. */
67  if ((error = avformat_open_input(input_format_context, filename, NULL,
68  NULL)) < 0) {
69  fprintf(stderr, "Could not open input file '%s' (error '%s')\n",
70  filename, av_err2str(error));
71  *input_format_context = NULL;
72  return error;
73  }
74 
75  /* Get information on the input file (number of streams etc.). */
76  if ((error = avformat_find_stream_info(*input_format_context, NULL)) < 0) {
77  fprintf(stderr, "Could not open find stream info (error '%s')\n",
78  av_err2str(error));
79  avformat_close_input(input_format_context);
80  return error;
81  }
82 
83  /* Make sure that there is only one stream in the input file. */
84  if ((*input_format_context)->nb_streams != 1) {
85  fprintf(stderr, "Expected one audio input stream, but found %d\n",
86  (*input_format_context)->nb_streams);
87  avformat_close_input(input_format_context);
88  return AVERROR_EXIT;
89  }
90 
91  /* Find a decoder for the audio stream. */
92  if (!(input_codec = avcodec_find_decoder((*input_format_context)->streams[0]->codecpar->codec_id))) {
93  fprintf(stderr, "Could not find input codec\n");
94  avformat_close_input(input_format_context);
95  return AVERROR_EXIT;
96  }
97 
98  /* Allocate a new decoding context. */
99  avctx = avcodec_alloc_context3(input_codec);
100  if (!avctx) {
101  fprintf(stderr, "Could not allocate a decoding context\n");
102  avformat_close_input(input_format_context);
103  return AVERROR(ENOMEM);
104  }
105 
106  /* Initialize the stream parameters with demuxer information. */
107  error = avcodec_parameters_to_context(avctx, (*input_format_context)->streams[0]->codecpar);
108  if (error < 0) {
109  avformat_close_input(input_format_context);
110  avcodec_free_context(&avctx);
111  return error;
112  }
113 
114  /* Open the decoder for the audio stream to use it later. */
115  if ((error = avcodec_open2(avctx, input_codec, NULL)) < 0) {
116  fprintf(stderr, "Could not open input codec (error '%s')\n",
117  av_err2str(error));
118  avcodec_free_context(&avctx);
119  avformat_close_input(input_format_context);
120  return error;
121  }
122 
123  /* Save the decoder context for easier access later. */
124  *input_codec_context = avctx;
125 
126  return 0;
127 }
128 
129 /**
130  * Open an output file and the required encoder.
131  * Also set some basic encoder parameters.
132  * Some of these parameters are based on the input file's parameters.
133  * @param filename File to be opened
134  * @param input_codec_context Codec context of input file
135  * @param[out] output_format_context Format context of output file
136  * @param[out] output_codec_context Codec context of output file
137  * @return Error code (0 if successful)
138  */
139 static int open_output_file(const char *filename,
140  AVCodecContext *input_codec_context,
141  AVFormatContext **output_format_context,
142  AVCodecContext **output_codec_context)
143 {
144  AVCodecContext *avctx = NULL;
145  AVIOContext *output_io_context = NULL;
146  AVStream *stream = NULL;
147  AVCodec *output_codec = NULL;
148  int error;
149 
150  /* Open the output file to write to it. */
151  if ((error = avio_open(&output_io_context, filename,
152  AVIO_FLAG_WRITE)) < 0) {
153  fprintf(stderr, "Could not open output file '%s' (error '%s')\n",
154  filename, av_err2str(error));
155  return error;
156  }
157 
158  /* Create a new format context for the output container format. */
159  if (!(*output_format_context = avformat_alloc_context())) {
160  fprintf(stderr, "Could not allocate output format context\n");
161  return AVERROR(ENOMEM);
162  }
163 
164  /* Associate the output file (pointer) with the container format context. */
165  (*output_format_context)->pb = output_io_context;
166 
167  /* Guess the desired container format based on the file extension. */
168  if (!((*output_format_context)->oformat = av_guess_format(NULL, filename,
169  NULL))) {
170  fprintf(stderr, "Could not find output file format\n");
171  goto cleanup;
172  }
173 
174  if (!((*output_format_context)->url = av_strdup(filename))) {
175  fprintf(stderr, "Could not allocate url.\n");
176  error = AVERROR(ENOMEM);
177  goto cleanup;
178  }
179 
180  /* Find the encoder to be used by its name. */
181  if (!(output_codec = avcodec_find_encoder(AV_CODEC_ID_AAC))) {
182  fprintf(stderr, "Could not find an AAC encoder.\n");
183  goto cleanup;
184  }
185 
186  /* Create a new audio stream in the output file container. */
187  if (!(stream = avformat_new_stream(*output_format_context, NULL))) {
188  fprintf(stderr, "Could not create new stream\n");
189  error = AVERROR(ENOMEM);
190  goto cleanup;
191  }
192 
193  avctx = avcodec_alloc_context3(output_codec);
194  if (!avctx) {
195  fprintf(stderr, "Could not allocate an encoding context\n");
196  error = AVERROR(ENOMEM);
197  goto cleanup;
198  }
199 
200  /* Set the basic encoder parameters.
201  * The input file's sample rate is used to avoid a sample rate conversion. */
202  avctx->channels = OUTPUT_CHANNELS;
204  avctx->sample_rate = input_codec_context->sample_rate;
205  avctx->sample_fmt = output_codec->sample_fmts[0];
206  avctx->bit_rate = OUTPUT_BIT_RATE;
207 
208  /* Allow the use of the experimental AAC encoder. */
210 
211  /* Set the sample rate for the container. */
212  stream->time_base.den = input_codec_context->sample_rate;
213  stream->time_base.num = 1;
214 
215  /* Some container formats (like MP4) require global headers to be present.
216  * Mark the encoder so that it behaves accordingly. */
217  if ((*output_format_context)->oformat->flags & AVFMT_GLOBALHEADER)
219 
220  /* Open the encoder for the audio stream to use it later. */
221  if ((error = avcodec_open2(avctx, output_codec, NULL)) < 0) {
222  fprintf(stderr, "Could not open output codec (error '%s')\n",
223  av_err2str(error));
224  goto cleanup;
225  }
226 
227  error = avcodec_parameters_from_context(stream->codecpar, avctx);
228  if (error < 0) {
229  fprintf(stderr, "Could not initialize stream parameters\n");
230  goto cleanup;
231  }
232 
233  /* Save the encoder context for easier access later. */
234  *output_codec_context = avctx;
235 
236  return 0;
237 
238 cleanup:
239  avcodec_free_context(&avctx);
240  avio_closep(&(*output_format_context)->pb);
241  avformat_free_context(*output_format_context);
242  *output_format_context = NULL;
243  return error < 0 ? error : AVERROR_EXIT;
244 }
245 
246 /**
247  * Initialize one data packet for reading or writing.
248  * @param[out] packet Packet to be initialized
249  * @return Error code (0 if successful)
250  */
251 static int init_packet(AVPacket **packet)
252 {
253  if (!(*packet = av_packet_alloc())) {
254  fprintf(stderr, "Could not allocate packet\n");
255  return AVERROR(ENOMEM);
256  }
257  return 0;
258 }
259 
260 /**
261  * Initialize one audio frame for reading from the input file.
262  * @param[out] frame Frame to be initialized
263  * @return Error code (0 if successful)
264  */
266 {
267  if (!(*frame = av_frame_alloc())) {
268  fprintf(stderr, "Could not allocate input frame\n");
269  return AVERROR(ENOMEM);
270  }
271  return 0;
272 }
273 
274 /**
275  * Initialize the audio resampler based on the input and output codec settings.
276  * If the input and output sample formats differ, a conversion is required
277  * libswresample takes care of this, but requires initialization.
278  * @param input_codec_context Codec context of the input file
279  * @param output_codec_context Codec context of the output file
280  * @param[out] resample_context Resample context for the required conversion
281  * @return Error code (0 if successful)
282  */
283 static int init_resampler(AVCodecContext *input_codec_context,
284  AVCodecContext *output_codec_context,
285  SwrContext **resample_context)
286 {
287  int error;
288 
289  /*
290  * Create a resampler context for the conversion.
291  * Set the conversion parameters.
292  * Default channel layouts based on the number of channels
293  * are assumed for simplicity (they are sometimes not detected
294  * properly by the demuxer and/or decoder).
295  */
296  *resample_context = swr_alloc_set_opts(NULL,
297  av_get_default_channel_layout(output_codec_context->channels),
298  output_codec_context->sample_fmt,
299  output_codec_context->sample_rate,
300  av_get_default_channel_layout(input_codec_context->channels),
301  input_codec_context->sample_fmt,
302  input_codec_context->sample_rate,
303  0, NULL);
304  if (!*resample_context) {
305  fprintf(stderr, "Could not allocate resample context\n");
306  return AVERROR(ENOMEM);
307  }
308  /*
309  * Perform a sanity check so that the number of converted samples is
310  * not greater than the number of samples to be converted.
311  * If the sample rates differ, this case has to be handled differently
312  */
313  av_assert0(output_codec_context->sample_rate == input_codec_context->sample_rate);
314 
315  /* Open the resampler with the specified parameters. */
316  if ((error = swr_init(*resample_context)) < 0) {
317  fprintf(stderr, "Could not open resample context\n");
318  swr_free(resample_context);
319  return error;
320  }
321  return 0;
322 }
323 
324 /**
325  * Initialize a FIFO buffer for the audio samples to be encoded.
326  * @param[out] fifo Sample buffer
327  * @param output_codec_context Codec context of the output file
328  * @return Error code (0 if successful)
329  */
330 static int init_fifo(AVAudioFifo **fifo, AVCodecContext *output_codec_context)
331 {
332  /* Create the FIFO buffer based on the specified output sample format. */
333  if (!(*fifo = av_audio_fifo_alloc(output_codec_context->sample_fmt,
334  output_codec_context->channels, 1))) {
335  fprintf(stderr, "Could not allocate FIFO\n");
336  return AVERROR(ENOMEM);
337  }
338  return 0;
339 }
340 
341 /**
342  * Write the header of the output file container.
343  * @param output_format_context Format context of the output file
344  * @return Error code (0 if successful)
345  */
346 static int write_output_file_header(AVFormatContext *output_format_context)
347 {
348  int error;
349  if ((error = avformat_write_header(output_format_context, NULL)) < 0) {
350  fprintf(stderr, "Could not write output file header (error '%s')\n",
351  av_err2str(error));
352  return error;
353  }
354  return 0;
355 }
356 
357 /**
358  * Decode one audio frame from the input file.
359  * @param frame Audio frame to be decoded
360  * @param input_format_context Format context of the input file
361  * @param input_codec_context Codec context of the input file
362  * @param[out] data_present Indicates whether data has been decoded
363  * @param[out] finished Indicates whether the end of file has
364  * been reached and all data has been
365  * decoded. If this flag is false, there
366  * is more data to be decoded, i.e., this
367  * function has to be called again.
368  * @return Error code (0 if successful)
369  */
371  AVFormatContext *input_format_context,
372  AVCodecContext *input_codec_context,
373  int *data_present, int *finished)
374 {
375  /* Packet used for temporary storage. */
376  AVPacket *input_packet;
377  int error;
378 
379  error = init_packet(&input_packet);
380  if (error < 0)
381  return error;
382 
383  /* Read one audio frame from the input file into a temporary packet. */
384  if ((error = av_read_frame(input_format_context, input_packet)) < 0) {
385  /* If we are at the end of the file, flush the decoder below. */
386  if (error == AVERROR_EOF)
387  *finished = 1;
388  else {
389  fprintf(stderr, "Could not read frame (error '%s')\n",
390  av_err2str(error));
391  goto cleanup;
392  }
393  }
394 
395  /* Send the audio frame stored in the temporary packet to the decoder.
396  * The input audio stream decoder is used to do this. */
397  if ((error = avcodec_send_packet(input_codec_context, input_packet)) < 0) {
398  fprintf(stderr, "Could not send packet for decoding (error '%s')\n",
399  av_err2str(error));
400  goto cleanup;
401  }
402 
403  /* Receive one frame from the decoder. */
404  error = avcodec_receive_frame(input_codec_context, frame);
405  /* If the decoder asks for more data to be able to decode a frame,
406  * return indicating that no data is present. */
407  if (error == AVERROR(EAGAIN)) {
408  error = 0;
409  goto cleanup;
410  /* If the end of the input file is reached, stop decoding. */
411  } else if (error == AVERROR_EOF) {
412  *finished = 1;
413  error = 0;
414  goto cleanup;
415  } else if (error < 0) {
416  fprintf(stderr, "Could not decode frame (error '%s')\n",
417  av_err2str(error));
418  goto cleanup;
419  /* Default case: Return decoded data. */
420  } else {
421  *data_present = 1;
422  goto cleanup;
423  }
424 
425 cleanup:
426  av_packet_free(&input_packet);
427  return error;
428 }
429 
430 /**
431  * Initialize a temporary storage for the specified number of audio samples.
432  * The conversion requires temporary storage due to the different format.
433  * The number of audio samples to be allocated is specified in frame_size.
434  * @param[out] converted_input_samples Array of converted samples. The
435  * dimensions are reference, channel
436  * (for multi-channel audio), sample.
437  * @param output_codec_context Codec context of the output file
438  * @param frame_size Number of samples to be converted in
439  * each round
440  * @return Error code (0 if successful)
441  */
442 static int init_converted_samples(uint8_t ***converted_input_samples,
443  AVCodecContext *output_codec_context,
444  int frame_size)
445 {
446  int error;
447 
448  /* Allocate as many pointers as there are audio channels.
449  * Each pointer will later point to the audio samples of the corresponding
450  * channels (although it may be NULL for interleaved formats).
451  */
452  if (!(*converted_input_samples = calloc(output_codec_context->channels,
453  sizeof(**converted_input_samples)))) {
454  fprintf(stderr, "Could not allocate converted input sample pointers\n");
455  return AVERROR(ENOMEM);
456  }
457 
458  /* Allocate memory for the samples of all channels in one consecutive
459  * block for convenience. */
460  if ((error = av_samples_alloc(*converted_input_samples, NULL,
461  output_codec_context->channels,
462  frame_size,
463  output_codec_context->sample_fmt, 0)) < 0) {
464  fprintf(stderr,
465  "Could not allocate converted input samples (error '%s')\n",
466  av_err2str(error));
467  av_freep(&(*converted_input_samples)[0]);
468  free(*converted_input_samples);
469  return error;
470  }
471  return 0;
472 }
473 
474 /**
475  * Convert the input audio samples into the output sample format.
476  * The conversion happens on a per-frame basis, the size of which is
477  * specified by frame_size.
478  * @param input_data Samples to be decoded. The dimensions are
479  * channel (for multi-channel audio), sample.
480  * @param[out] converted_data Converted samples. The dimensions are channel
481  * (for multi-channel audio), sample.
482  * @param frame_size Number of samples to be converted
483  * @param resample_context Resample context for the conversion
484  * @return Error code (0 if successful)
485  */
486 static int convert_samples(const uint8_t **input_data,
487  uint8_t **converted_data, const int frame_size,
488  SwrContext *resample_context)
489 {
490  int error;
491 
492  /* Convert the samples using the resampler. */
493  if ((error = swr_convert(resample_context,
494  converted_data, frame_size,
495  input_data , frame_size)) < 0) {
496  fprintf(stderr, "Could not convert input samples (error '%s')\n",
497  av_err2str(error));
498  return error;
499  }
500 
501  return 0;
502 }
503 
504 /**
505  * Add converted input audio samples to the FIFO buffer for later processing.
506  * @param fifo Buffer to add the samples to
507  * @param converted_input_samples Samples to be added. The dimensions are channel
508  * (for multi-channel audio), sample.
509  * @param frame_size Number of samples to be converted
510  * @return Error code (0 if successful)
511  */
513  uint8_t **converted_input_samples,
514  const int frame_size)
515 {
516  int error;
517 
518  /* Make the FIFO as large as it needs to be to hold both,
519  * the old and the new samples. */
520  if ((error = av_audio_fifo_realloc(fifo, av_audio_fifo_size(fifo) + frame_size)) < 0) {
521  fprintf(stderr, "Could not reallocate FIFO\n");
522  return error;
523  }
524 
525  /* Store the new samples in the FIFO buffer. */
526  if (av_audio_fifo_write(fifo, (void **)converted_input_samples,
527  frame_size) < frame_size) {
528  fprintf(stderr, "Could not write data to FIFO\n");
529  return AVERROR_EXIT;
530  }
531  return 0;
532 }
533 
534 /**
535  * Read one audio frame from the input file, decode, convert and store
536  * it in the FIFO buffer.
537  * @param fifo Buffer used for temporary storage
538  * @param input_format_context Format context of the input file
539  * @param input_codec_context Codec context of the input file
540  * @param output_codec_context Codec context of the output file
541  * @param resampler_context Resample context for the conversion
542  * @param[out] finished Indicates whether the end of file has
543  * been reached and all data has been
544  * decoded. If this flag is false,
545  * there is more data to be decoded,
546  * i.e., this function has to be called
547  * again.
548  * @return Error code (0 if successful)
549  */
551  AVFormatContext *input_format_context,
552  AVCodecContext *input_codec_context,
553  AVCodecContext *output_codec_context,
554  SwrContext *resampler_context,
555  int *finished)
556 {
557  /* Temporary storage of the input samples of the frame read from the file. */
558  AVFrame *input_frame = NULL;
559  /* Temporary storage for the converted input samples. */
560  uint8_t **converted_input_samples = NULL;
561  int data_present = 0;
562  int ret = AVERROR_EXIT;
563 
564  /* Initialize temporary storage for one input frame. */
565  if (init_input_frame(&input_frame))
566  goto cleanup;
567  /* Decode one frame worth of audio samples. */
568  if (decode_audio_frame(input_frame, input_format_context,
569  input_codec_context, &data_present, finished))
570  goto cleanup;
571  /* If we are at the end of the file and there are no more samples
572  * in the decoder which are delayed, we are actually finished.
573  * This must not be treated as an error. */
574  if (*finished) {
575  ret = 0;
576  goto cleanup;
577  }
578  /* If there is decoded data, convert and store it. */
579  if (data_present) {
580  /* Initialize the temporary storage for the converted input samples. */
581  if (init_converted_samples(&converted_input_samples, output_codec_context,
582  input_frame->nb_samples))
583  goto cleanup;
584 
585  /* Convert the input samples to the desired output sample format.
586  * This requires a temporary storage provided by converted_input_samples. */
587  if (convert_samples((const uint8_t**)input_frame->extended_data, converted_input_samples,
588  input_frame->nb_samples, resampler_context))
589  goto cleanup;
590 
591  /* Add the converted input samples to the FIFO buffer for later processing. */
592  if (add_samples_to_fifo(fifo, converted_input_samples,
593  input_frame->nb_samples))
594  goto cleanup;
595  ret = 0;
596  }
597  ret = 0;
598 
599 cleanup:
600  if (converted_input_samples) {
601  av_freep(&converted_input_samples[0]);
602  free(converted_input_samples);
603  }
604  av_frame_free(&input_frame);
605 
606  return ret;
607 }
608 
609 /**
610  * Initialize one input frame for writing to the output file.
611  * The frame will be exactly frame_size samples large.
612  * @param[out] frame Frame to be initialized
613  * @param output_codec_context Codec context of the output file
614  * @param frame_size Size of the frame
615  * @return Error code (0 if successful)
616  */
618  AVCodecContext *output_codec_context,
619  int frame_size)
620 {
621  int error;
622 
623  /* Create a new frame to store the audio samples. */
624  if (!(*frame = av_frame_alloc())) {
625  fprintf(stderr, "Could not allocate output frame\n");
626  return AVERROR_EXIT;
627  }
628 
629  /* Set the frame's parameters, especially its size and format.
630  * av_frame_get_buffer needs this to allocate memory for the
631  * audio samples of the frame.
632  * Default channel layouts based on the number of channels
633  * are assumed for simplicity. */
634  (*frame)->nb_samples = frame_size;
635  (*frame)->channel_layout = output_codec_context->channel_layout;
636  (*frame)->format = output_codec_context->sample_fmt;
637  (*frame)->sample_rate = output_codec_context->sample_rate;
638 
639  /* Allocate the samples of the created frame. This call will make
640  * sure that the audio frame can hold as many samples as specified. */
641  if ((error = av_frame_get_buffer(*frame, 0)) < 0) {
642  fprintf(stderr, "Could not allocate output frame samples (error '%s')\n",
643  av_err2str(error));
644  av_frame_free(frame);
645  return error;
646  }
647 
648  return 0;
649 }
650 
651 /* Global timestamp for the audio frames. */
652 static int64_t pts = 0;
653 
654 /**
655  * Encode one frame worth of audio to the output file.
656  * @param frame Samples to be encoded
657  * @param output_format_context Format context of the output file
658  * @param output_codec_context Codec context of the output file
659  * @param[out] data_present Indicates whether data has been
660  * encoded
661  * @return Error code (0 if successful)
662  */
664  AVFormatContext *output_format_context,
665  AVCodecContext *output_codec_context,
666  int *data_present)
667 {
668  /* Packet used for temporary storage. */
669  AVPacket *output_packet;
670  int error;
671 
672  error = init_packet(&output_packet);
673  if (error < 0)
674  return error;
675 
676  /* Set a timestamp based on the sample rate for the container. */
677  if (frame) {
678  frame->pts = pts;
679  pts += frame->nb_samples;
680  }
681 
682  /* Send the audio frame stored in the temporary packet to the encoder.
683  * The output audio stream encoder is used to do this. */
684  error = avcodec_send_frame(output_codec_context, frame);
685  /* The encoder signals that it has nothing more to encode. */
686  if (error == AVERROR_EOF) {
687  error = 0;
688  goto cleanup;
689  } else if (error < 0) {
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  data_written = 0;
864  if (encode_audio_frame(NULL, output_format_context,
865  output_codec_context, &data_written))
866  goto cleanup;
867  } while (data_written);
868  break;
869  }
870  }
871 
872  /* Write the trailer of the output file container. */
873  if (write_output_file_trailer(output_format_context))
874  goto cleanup;
875  ret = 0;
876 
877 cleanup:
878  if (fifo)
879  av_audio_fifo_free(fifo);
880  swr_free(&resample_context);
881  if (output_codec_context)
882  avcodec_free_context(&output_codec_context);
883  if (output_format_context) {
884  avio_closep(&output_format_context->pb);
885  avformat_free_context(output_format_context);
886  }
887  if (input_codec_context)
888  avcodec_free_context(&input_codec_context);
889  if (input_format_context)
890  avformat_close_input(&input_format_context);
891 
892  return ret;
893 }
int avio_open(AVIOContext **s, const char *url, int flags)
Create and initialize a AVIOContext for accessing the resource indicated by url.
#define FF_COMPLIANCE_EXPERIMENTAL
Allow nonstandardized experimental things.
Definition: avcodec.h:1606
Bytestream IO Context.
Definition: avio.h:161
AVAudioFifo * av_audio_fifo_alloc(enum AVSampleFormat sample_fmt, int channels, int nb_samples)
Allocate an AVAudioFifo.
Buffered I/O operations.
int av_audio_fifo_read(AVAudioFifo *af, void **data, int nb_samples)
Read data from an AVAudioFifo.
This structure describes decoded (raw) audio or video data.
Definition: frame.h:318
int main(int argc, char **argv)
int av_write_frame(AVFormatContext *s, AVPacket *pkt)
Write a packet to an output media file.
int64_t bit_rate
the average bitrate
Definition: avcodec.h:586
int avcodec_receive_packet(AVCodecContext *avctx, AVPacket *avpkt)
Read encoded data from the encoder.
void av_audio_fifo_free(AVAudioFifo *af)
Free an AVAudioFifo.
int num
Numerator.
Definition: rational.h:59
#define AVIO_FLAG_WRITE
write-only
Definition: avio.h:675
int avformat_open_input(AVFormatContext **ps, const char *url, ff_const59 AVInputFormat *fmt, AVDictionary **options)
Open an input stream and read the header.
AVCodec.
Definition: codec.h:197
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 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
Format I/O context.
Definition: avformat.h:1248
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 av_assert0(cond)
assert() equivalent, that is always enabled.
Definition: avassert.h:37
void av_packet_free(AVPacket **pkt)
Free the packet, if the packet is reference counted, it will be unreferenced first.
enum AVSampleFormat sample_fmt
audio sample format
Definition: avcodec.h:1204
AVFrame * av_frame_alloc(void)
Allocate an AVFrame and set its fields to default values.
AVOptions.
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...
int64_t pts
Presentation timestamp in time_base units (time when frame should be shown to user).
Definition: frame.h:411
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.
AVStream * avformat_new_stream(AVFormatContext *s, const AVCodec *c)
Add a new stream to a media file.
int avcodec_parameters_to_context(AVCodecContext *codec, const AVCodecParameters *par)
Fill the codec context based on the values from the supplied codec parameters.
static AVFrame * frame
AVFormatContext * avformat_alloc_context(void)
Allocate an AVFormatContext.
int swr_convert(struct SwrContext *s, uint8_t **out, int out_count, const uint8_t **in, int in_count)
Convert audio.
#define AVERROR_EOF
End of file.
Definition: error.h:55
static int init_input_frame(AVFrame **frame)
Initialize one audio frame for reading from the input file.
libswresample public header
static int write_output_file_header(AVFormatContext *output_format_context)
Write the header of the output file container.
#define AVERROR(e)
Definition: error.h:43
void av_frame_free(AVFrame **frame)
Free the frame and any dynamically allocated objects in it, e.g.
void av_freep(void *ptr)
Free a memory block which has been allocated with a function of av_malloc() or av_realloc() family...
int avcodec_receive_frame(AVCodecContext *avctx, AVFrame *frame)
Return decoded output data from a decoder.
int flags
AV_CODEC_FLAG_*.
Definition: avcodec.h:616
AVCodec * avcodec_find_encoder(enum AVCodecID id)
Find a registered encoder with a matching codec ID.
simple assert() macros that are a bit more flexible than ISO C assert().
struct AVAudioFifo AVAudioFifo
Context for an Audio FIFO Buffer.
Definition: audio_fifo.h:49
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.
struct SwrContext SwrContext
The libswresample context.
Definition: swresample.h:182
reference-counted frame API
uint64_t channel_layout
Audio channel layout.
Definition: avcodec.h:1247
int av_audio_fifo_size(AVAudioFifo *af)
Get the current number of samples in the AVAudioFifo available for reading.
av_warn_unused_result int av_audio_fifo_realloc(AVAudioFifo *af, int nb_samples)
Reallocate an AVAudioFifo.
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.
#define FFMIN(a, b)
Definition: common.h:105
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.
AVCodecContext * avcodec_alloc_context3(const AVCodec *codec)
Allocate an AVCodecContext and set its fields to default values.
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.
#define AVFMT_GLOBALHEADER
Format wants global header.
Definition: avformat.h:461
#define av_err2str(errnum)
Convenience macro, the return value should be used only directly in function arguments but never stan...
Definition: error.h:119
struct SwrContext * swr_alloc_set_opts(struct SwrContext *s, int64_t out_ch_layout, enum AVSampleFormat out_sample_fmt, int out_sample_rate, int64_t 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.
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.
#define AVERROR_EXIT
Immediate exit was requested; the called function should not be restarted.
Definition: error.h:56
Stream structure.
Definition: avformat.h:884
int avcodec_send_packet(AVCodecContext *avctx, const AVPacket *avpkt)
Supply raw packet data as input to a decoder.
int frame_size
Number of samples per channel in an audio frame.
Definition: avcodec.h:1216
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...
static int init_fifo(AVAudioFifo **fifo, AVCodecContext *output_codec_context)
Initialize a FIFO buffer for the audio samples to be encoded.
char * av_strdup(const char *s) av_malloc_attrib
Duplicate a string.
Libavcodec external API header.
void avcodec_free_context(AVCodecContext **avctx)
Free the codec context and everything associated with it and write NULL to the provided pointer...
#define OUTPUT_BIT_RATE
Definition: transcode_aac.c:47
int sample_rate
samples per second
Definition: avcodec.h:1196
main external API structure.
Definition: avcodec.h:536
void swr_free(struct SwrContext **s)
Free the given SwrContext and set the pointer to NULL.
int avcodec_send_frame(AVCodecContext *avctx, const AVFrame *frame)
Supply a raw video or audio frame to the encoder.
ff_const59 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 avcodec_parameters_from_context(AVCodecParameters *par, const AVCodecContext *codec)
Fill the parameters struct based on the values from the supplied codec context.
static int init_packet(AVPacket **packet)
Initialize one data packet for reading or writing.
int avcodec_open2(AVCodecContext *avctx, const AVCodec *codec, AVDictionary **options)
Initialize the AVCodecContext to use the given AVCodec.
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.
AVCodec * avcodec_find_decoder(enum AVCodecID id)
Find a registered decoder with a matching codec ID.
int av_frame_get_buffer(AVFrame *frame, int align)
Allocate new buffer(s) for audio or video data.
static int64_t pts
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: frame.h:332
int av_audio_fifo_write(AVAudioFifo *af, void **data, int nb_samples)
Write data to an AVAudioFifo.
#define OUTPUT_CHANNELS
Definition: transcode_aac.c:49
#define AV_CODEC_FLAG_GLOBAL_HEADER
Place global headers in extradata instead of every keyframe.
Definition: avcodec.h:329
Main libavformat public API header.
int avformat_find_stream_info(AVFormatContext *ic, AVDictionary **options)
Read packets of a media file to get stream information.
static int write_output_file_trailer(AVFormatContext *output_format_context)
Write the trailer of the output file container.
int den
Denominator.
Definition: rational.h:60
void avformat_close_input(AVFormatContext **s)
Close an opened input AVFormatContext.
Audio FIFO Buffer.
int channels
number of audio channels
Definition: avcodec.h:1197
AVPacket * av_packet_alloc(void)
Allocate an AVPacket and set its fields to default values.
int64_t av_get_default_channel_layout(int nb_channels)
Return default channel layout for a given number of channels.
int av_write_trailer(AVFormatContext *s)
Write the stream trailer to an output media file and free the file private data.
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.
AVCodecParameters * codecpar
Codec parameters associated with this stream.
Definition: avformat.h:1049
enum AVSampleFormat * sample_fmts
array of supported sample formats, or NULL if unknown, array is terminated by -1
Definition: codec.h:220
static int init_output_frame(AVFrame **frame, AVCodecContext *output_codec_context, int frame_size)
Initialize one input frame for writing to the output file.
AVRational time_base
This is the fundamental unit of time (in seconds) in terms of which frame timestamps are represented...
Definition: avformat.h:913
uint8_t ** extended_data
pointers to the data planes/channels.
Definition: frame.h:365
This structure stores compressed data.
Definition: packet.h:346
int avio_closep(AVIOContext **s)
Close the resource accessed by the AVIOContext *s, free it and set the pointer pointing to it to NULL...
int nb_samples
number of audio samples (per channel) described by this frame
Definition: frame.h:384
int strict_std_compliance
strictly follow the standard (MPEG-4, ...).
Definition: avcodec.h:1601
int swr_init(struct SwrContext *s)
Initialize context after user parameters have been set.