FFmpeg  5.1.2
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
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"
42 #include "libavutil/frame.h"
43 #include "libavutil/opt.h"
44 
46 
47 /* The output bit rate in bit/s */
48 #define OUTPUT_BIT_RATE 96000
49 /* The number of output channels */
50 #define OUTPUT_CHANNELS 2
51 
52 /**
53  * Open an input file and the required decoder.
54  * @param filename File to be opened
55  * @param[out] input_format_context Format context of opened file
56  * @param[out] input_codec_context Codec context of opened file
57  * @return Error code (0 if successful)
58  */
59 static int open_input_file(const char *filename,
60  AVFormatContext **input_format_context,
61  AVCodecContext **input_codec_context)
62 {
63  AVCodecContext *avctx;
64  const AVCodec *input_codec;
65  const AVStream *stream;
66  int error;
67 
68  /* Open the input file to read from it. */
69  if ((error = avformat_open_input(input_format_context, filename, NULL,
70  NULL)) < 0) {
71  fprintf(stderr, "Could not open input file '%s' (error '%s')\n",
72  filename, av_err2str(error));
73  *input_format_context = NULL;
74  return error;
75  }
76 
77  /* Get information on the input file (number of streams etc.). */
78  if ((error = avformat_find_stream_info(*input_format_context, NULL)) < 0) {
79  fprintf(stderr, "Could not open find stream info (error '%s')\n",
80  av_err2str(error));
81  avformat_close_input(input_format_context);
82  return error;
83  }
84 
85  /* Make sure that there is only one stream in the input file. */
86  if ((*input_format_context)->nb_streams != 1) {
87  fprintf(stderr, "Expected one audio input stream, but found %d\n",
88  (*input_format_context)->nb_streams);
89  avformat_close_input(input_format_context);
90  return AVERROR_EXIT;
91  }
92 
93  stream = (*input_format_context)->streams[0];
94 
95  /* Find a decoder for the audio stream. */
96  if (!(input_codec = avcodec_find_decoder(stream->codecpar->codec_id))) {
97  fprintf(stderr, "Could not find input codec\n");
98  avformat_close_input(input_format_context);
99  return AVERROR_EXIT;
100  }
101 
102  /* Allocate a new decoding context. */
103  avctx = avcodec_alloc_context3(input_codec);
104  if (!avctx) {
105  fprintf(stderr, "Could not allocate a decoding context\n");
106  avformat_close_input(input_format_context);
107  return AVERROR(ENOMEM);
108  }
109 
110  /* Initialize the stream parameters with demuxer information. */
111  error = avcodec_parameters_to_context(avctx, stream->codecpar);
112  if (error < 0) {
113  avformat_close_input(input_format_context);
114  avcodec_free_context(&avctx);
115  return error;
116  }
117 
118  /* Open the decoder for the audio stream to use it later. */
119  if ((error = avcodec_open2(avctx, input_codec, NULL)) < 0) {
120  fprintf(stderr, "Could not open input codec (error '%s')\n",
121  av_err2str(error));
122  avcodec_free_context(&avctx);
123  avformat_close_input(input_format_context);
124  return error;
125  }
126 
127  /* Set the packet timebase for the decoder. */
128  avctx->pkt_timebase = stream->time_base;
129 
130  /* Save the decoder context for easier access later. */
131  *input_codec_context = avctx;
132 
133  return 0;
134 }
135 
136 /**
137  * Open an output file and the required encoder.
138  * Also set some basic encoder parameters.
139  * Some of these parameters are based on the input file's parameters.
140  * @param filename File to be opened
141  * @param input_codec_context Codec context of input file
142  * @param[out] output_format_context Format context of output file
143  * @param[out] output_codec_context Codec context of output file
144  * @return Error code (0 if successful)
145  */
146 static int open_output_file(const char *filename,
147  AVCodecContext *input_codec_context,
148  AVFormatContext **output_format_context,
149  AVCodecContext **output_codec_context)
150 {
151  AVCodecContext *avctx = NULL;
152  AVIOContext *output_io_context = NULL;
153  AVStream *stream = NULL;
154  const AVCodec *output_codec = NULL;
155  int error;
156 
157  /* Open the output file to write to it. */
158  if ((error = avio_open(&output_io_context, filename,
159  AVIO_FLAG_WRITE)) < 0) {
160  fprintf(stderr, "Could not open output file '%s' (error '%s')\n",
161  filename, av_err2str(error));
162  return error;
163  }
164 
165  /* Create a new format context for the output container format. */
166  if (!(*output_format_context = avformat_alloc_context())) {
167  fprintf(stderr, "Could not allocate output format context\n");
168  return AVERROR(ENOMEM);
169  }
170 
171  /* Associate the output file (pointer) with the container format context. */
172  (*output_format_context)->pb = output_io_context;
173 
174  /* Guess the desired container format based on the file extension. */
175  if (!((*output_format_context)->oformat = av_guess_format(NULL, filename,
176  NULL))) {
177  fprintf(stderr, "Could not find output file format\n");
178  goto cleanup;
179  }
180 
181  if (!((*output_format_context)->url = av_strdup(filename))) {
182  fprintf(stderr, "Could not allocate url.\n");
183  error = AVERROR(ENOMEM);
184  goto cleanup;
185  }
186 
187  /* Find the encoder to be used by its name. */
188  if (!(output_codec = avcodec_find_encoder(AV_CODEC_ID_AAC))) {
189  fprintf(stderr, "Could not find an AAC encoder.\n");
190  goto cleanup;
191  }
192 
193  /* Create a new audio stream in the output file container. */
194  if (!(stream = avformat_new_stream(*output_format_context, NULL))) {
195  fprintf(stderr, "Could not create new stream\n");
196  error = AVERROR(ENOMEM);
197  goto cleanup;
198  }
199 
200  avctx = avcodec_alloc_context3(output_codec);
201  if (!avctx) {
202  fprintf(stderr, "Could not allocate an encoding context\n");
203  error = AVERROR(ENOMEM);
204  goto cleanup;
205  }
206 
207  /* Set the basic encoder parameters.
208  * The input file's sample rate is used to avoid a sample rate conversion. */
210  avctx->sample_rate = input_codec_context->sample_rate;
211  avctx->sample_fmt = output_codec->sample_fmts[0];
212  avctx->bit_rate = OUTPUT_BIT_RATE;
213 
214  /* Set the sample rate for the container. */
215  stream->time_base.den = input_codec_context->sample_rate;
216  stream->time_base.num = 1;
217 
218  /* Some container formats (like MP4) require global headers to be present.
219  * Mark the encoder so that it behaves accordingly. */
220  if ((*output_format_context)->oformat->flags & AVFMT_GLOBALHEADER)
222 
223  /* Open the encoder for the audio stream to use it later. */
224  if ((error = avcodec_open2(avctx, output_codec, NULL)) < 0) {
225  fprintf(stderr, "Could not open output codec (error '%s')\n",
226  av_err2str(error));
227  goto cleanup;
228  }
229 
230  error = avcodec_parameters_from_context(stream->codecpar, avctx);
231  if (error < 0) {
232  fprintf(stderr, "Could not initialize stream parameters\n");
233  goto cleanup;
234  }
235 
236  /* Save the encoder context for easier access later. */
237  *output_codec_context = avctx;
238 
239  return 0;
240 
241 cleanup:
242  avcodec_free_context(&avctx);
243  avio_closep(&(*output_format_context)->pb);
244  avformat_free_context(*output_format_context);
245  *output_format_context = NULL;
246  return error < 0 ? error : AVERROR_EXIT;
247 }
248 
249 /**
250  * Initialize one data packet for reading or writing.
251  * @param[out] packet Packet to be initialized
252  * @return Error code (0 if successful)
253  */
254 static int init_packet(AVPacket **packet)
255 {
256  if (!(*packet = av_packet_alloc())) {
257  fprintf(stderr, "Could not allocate packet\n");
258  return AVERROR(ENOMEM);
259  }
260  return 0;
261 }
262 
263 /**
264  * Initialize one audio frame for reading from the input file.
265  * @param[out] frame Frame to be initialized
266  * @return Error code (0 if successful)
267  */
269 {
270  if (!(*frame = av_frame_alloc())) {
271  fprintf(stderr, "Could not allocate input frame\n");
272  return AVERROR(ENOMEM);
273  }
274  return 0;
275 }
276 
277 /**
278  * Initialize the audio resampler based on the input and output codec settings.
279  * If the input and output sample formats differ, a conversion is required
280  * libswresample takes care of this, but requires initialization.
281  * @param input_codec_context Codec context of the input file
282  * @param output_codec_context Codec context of the output file
283  * @param[out] resample_context Resample context for the required conversion
284  * @return Error code (0 if successful)
285  */
286 static int init_resampler(AVCodecContext *input_codec_context,
287  AVCodecContext *output_codec_context,
288  SwrContext **resample_context)
289 {
290  int error;
291 
292  /*
293  * Create a resampler context for the conversion.
294  * Set the conversion parameters.
295  */
296  error = swr_alloc_set_opts2(resample_context,
297  &output_codec_context->ch_layout,
298  output_codec_context->sample_fmt,
299  output_codec_context->sample_rate,
300  &input_codec_context->ch_layout,
301  input_codec_context->sample_fmt,
302  input_codec_context->sample_rate,
303  0, NULL);
304  if (error < 0) {
305  fprintf(stderr, "Could not allocate resample context\n");
306  return error;
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->ch_layout.nb_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  *data_present = 0;
384  *finished = 0;
385  /* Read one audio frame from the input file into a temporary packet. */
386  if ((error = av_read_frame(input_format_context, input_packet)) < 0) {
387  /* If we are at the end of the file, flush the decoder below. */
388  if (error == AVERROR_EOF)
389  *finished = 1;
390  else {
391  fprintf(stderr, "Could not read frame (error '%s')\n",
392  av_err2str(error));
393  goto cleanup;
394  }
395  }
396 
397  /* Send the audio frame stored in the temporary packet to the decoder.
398  * The input audio stream decoder is used to do this. */
399  if ((error = avcodec_send_packet(input_codec_context, input_packet)) < 0) {
400  fprintf(stderr, "Could not send packet for decoding (error '%s')\n",
401  av_err2str(error));
402  goto cleanup;
403  }
404 
405  /* Receive one frame from the decoder. */
406  error = avcodec_receive_frame(input_codec_context, frame);
407  /* If the decoder asks for more data to be able to decode a frame,
408  * return indicating that no data is present. */
409  if (error == AVERROR(EAGAIN)) {
410  error = 0;
411  goto cleanup;
412  /* If the end of the input file is reached, stop decoding. */
413  } else if (error == AVERROR_EOF) {
414  *finished = 1;
415  error = 0;
416  goto cleanup;
417  } else if (error < 0) {
418  fprintf(stderr, "Could not decode frame (error '%s')\n",
419  av_err2str(error));
420  goto cleanup;
421  /* Default case: Return decoded data. */
422  } else {
423  *data_present = 1;
424  goto cleanup;
425  }
426 
427 cleanup:
428  av_packet_free(&input_packet);
429  return error;
430 }
431 
432 /**
433  * Initialize a temporary storage for the specified number of audio samples.
434  * The conversion requires temporary storage due to the different format.
435  * The number of audio samples to be allocated is specified in frame_size.
436  * @param[out] converted_input_samples Array of converted samples. The
437  * dimensions are reference, channel
438  * (for multi-channel audio), sample.
439  * @param output_codec_context Codec context of the output file
440  * @param frame_size Number of samples to be converted in
441  * each round
442  * @return Error code (0 if successful)
443  */
444 static int init_converted_samples(uint8_t ***converted_input_samples,
445  AVCodecContext *output_codec_context,
446  int frame_size)
447 {
448  int error;
449 
450  /* Allocate as many pointers as there are audio channels.
451  * Each pointer will later point to the audio samples of the corresponding
452  * channels (although it may be NULL for interleaved formats).
453  */
454  if (!(*converted_input_samples = calloc(output_codec_context->ch_layout.nb_channels,
455  sizeof(**converted_input_samples)))) {
456  fprintf(stderr, "Could not allocate converted input sample pointers\n");
457  return AVERROR(ENOMEM);
458  }
459 
460  /* Allocate memory for the samples of all channels in one consecutive
461  * block for convenience. */
462  if ((error = av_samples_alloc(*converted_input_samples, NULL,
463  output_codec_context->ch_layout.nb_channels,
464  frame_size,
465  output_codec_context->sample_fmt, 0)) < 0) {
466  fprintf(stderr,
467  "Could not allocate converted input samples (error '%s')\n",
468  av_err2str(error));
469  av_freep(&(*converted_input_samples)[0]);
470  free(*converted_input_samples);
471  return error;
472  }
473  return 0;
474 }
475 
476 /**
477  * Convert the input audio samples into the output sample format.
478  * The conversion happens on a per-frame basis, the size of which is
479  * specified by frame_size.
480  * @param input_data Samples to be decoded. The dimensions are
481  * channel (for multi-channel audio), sample.
482  * @param[out] converted_data Converted samples. The dimensions are channel
483  * (for multi-channel audio), sample.
484  * @param frame_size Number of samples to be converted
485  * @param resample_context Resample context for the conversion
486  * @return Error code (0 if successful)
487  */
488 static int convert_samples(const uint8_t **input_data,
489  uint8_t **converted_data, const int frame_size,
490  SwrContext *resample_context)
491 {
492  int error;
493 
494  /* Convert the samples using the resampler. */
495  if ((error = swr_convert(resample_context,
496  converted_data, frame_size,
497  input_data , frame_size)) < 0) {
498  fprintf(stderr, "Could not convert input samples (error '%s')\n",
499  av_err2str(error));
500  return error;
501  }
502 
503  return 0;
504 }
505 
506 /**
507  * Add converted input audio samples to the FIFO buffer for later processing.
508  * @param fifo Buffer to add the samples to
509  * @param converted_input_samples Samples to be added. The dimensions are channel
510  * (for multi-channel audio), sample.
511  * @param frame_size Number of samples to be converted
512  * @return Error code (0 if successful)
513  */
515  uint8_t **converted_input_samples,
516  const int frame_size)
517 {
518  int error;
519 
520  /* Make the FIFO as large as it needs to be to hold both,
521  * the old and the new samples. */
522  if ((error = av_audio_fifo_realloc(fifo, av_audio_fifo_size(fifo) + frame_size)) < 0) {
523  fprintf(stderr, "Could not reallocate FIFO\n");
524  return error;
525  }
526 
527  /* Store the new samples in the FIFO buffer. */
528  if (av_audio_fifo_write(fifo, (void **)converted_input_samples,
529  frame_size) < frame_size) {
530  fprintf(stderr, "Could not write data to FIFO\n");
531  return AVERROR_EXIT;
532  }
533  return 0;
534 }
535 
536 /**
537  * Read one audio frame from the input file, decode, convert and store
538  * it in the FIFO buffer.
539  * @param fifo Buffer used for temporary storage
540  * @param input_format_context Format context of the input file
541  * @param input_codec_context Codec context of the input file
542  * @param output_codec_context Codec context of the output file
543  * @param resampler_context Resample context for the conversion
544  * @param[out] finished Indicates whether the end of file has
545  * been reached and all data has been
546  * decoded. If this flag is false,
547  * there is more data to be decoded,
548  * i.e., this function has to be called
549  * again.
550  * @return Error code (0 if successful)
551  */
553  AVFormatContext *input_format_context,
554  AVCodecContext *input_codec_context,
555  AVCodecContext *output_codec_context,
556  SwrContext *resampler_context,
557  int *finished)
558 {
559  /* Temporary storage of the input samples of the frame read from the file. */
560  AVFrame *input_frame = NULL;
561  /* Temporary storage for the converted input samples. */
562  uint8_t **converted_input_samples = NULL;
563  int data_present;
564  int ret = AVERROR_EXIT;
565 
566  /* Initialize temporary storage for one input frame. */
567  if (init_input_frame(&input_frame))
568  goto cleanup;
569  /* Decode one frame worth of audio samples. */
570  if (decode_audio_frame(input_frame, input_format_context,
571  input_codec_context, &data_present, finished))
572  goto cleanup;
573  /* If we are at the end of the file and there are no more samples
574  * in the decoder which are delayed, we are actually finished.
575  * This must not be treated as an error. */
576  if (*finished) {
577  ret = 0;
578  goto cleanup;
579  }
580  /* If there is decoded data, convert and store it. */
581  if (data_present) {
582  /* Initialize the temporary storage for the converted input samples. */
583  if (init_converted_samples(&converted_input_samples, output_codec_context,
584  input_frame->nb_samples))
585  goto cleanup;
586 
587  /* Convert the input samples to the desired output sample format.
588  * This requires a temporary storage provided by converted_input_samples. */
589  if (convert_samples((const uint8_t**)input_frame->extended_data, converted_input_samples,
590  input_frame->nb_samples, resampler_context))
591  goto cleanup;
592 
593  /* Add the converted input samples to the FIFO buffer for later processing. */
594  if (add_samples_to_fifo(fifo, converted_input_samples,
595  input_frame->nb_samples))
596  goto cleanup;
597  ret = 0;
598  }
599  ret = 0;
600 
601 cleanup:
602  if (converted_input_samples) {
603  av_freep(&converted_input_samples[0]);
604  free(converted_input_samples);
605  }
606  av_frame_free(&input_frame);
607 
608  return ret;
609 }
610 
611 /**
612  * Initialize one input frame for writing to the output file.
613  * The frame will be exactly frame_size samples large.
614  * @param[out] frame Frame to be initialized
615  * @param output_codec_context Codec context of the output file
616  * @param frame_size Size of the frame
617  * @return Error code (0 if successful)
618  */
620  AVCodecContext *output_codec_context,
621  int frame_size)
622 {
623  int error;
624 
625  /* Create a new frame to store the audio samples. */
626  if (!(*frame = av_frame_alloc())) {
627  fprintf(stderr, "Could not allocate output frame\n");
628  return AVERROR_EXIT;
629  }
630 
631  /* Set the frame's parameters, especially its size and format.
632  * av_frame_get_buffer needs this to allocate memory for the
633  * audio samples of the frame.
634  * Default channel layouts based on the number of channels
635  * are assumed for simplicity. */
636  (*frame)->nb_samples = frame_size;
637  av_channel_layout_copy(&(*frame)->ch_layout, &output_codec_context->ch_layout);
638  (*frame)->format = output_codec_context->sample_fmt;
639  (*frame)->sample_rate = output_codec_context->sample_rate;
640 
641  /* Allocate the samples of the created frame. This call will make
642  * sure that the audio frame can hold as many samples as specified. */
643  if ((error = av_frame_get_buffer(*frame, 0)) < 0) {
644  fprintf(stderr, "Could not allocate output frame samples (error '%s')\n",
645  av_err2str(error));
647  return error;
648  }
649 
650  return 0;
651 }
652 
653 /* Global timestamp for the audio frames. */
654 static int64_t pts = 0;
655 
656 /**
657  * Encode one frame worth of audio to the output file.
658  * @param frame Samples to be encoded
659  * @param output_format_context Format context of the output file
660  * @param output_codec_context Codec context of the output file
661  * @param[out] data_present Indicates whether data has been
662  * encoded
663  * @return Error code (0 if successful)
664  */
666  AVFormatContext *output_format_context,
667  AVCodecContext *output_codec_context,
668  int *data_present)
669 {
670  /* Packet used for temporary storage. */
671  AVPacket *output_packet;
672  int error;
673 
674  error = init_packet(&output_packet);
675  if (error < 0)
676  return error;
677 
678  /* Set a timestamp based on the sample rate for the container. */
679  if (frame) {
680  frame->pts = pts;
681  pts += frame->nb_samples;
682  }
683 
684  *data_present = 0;
685  /* Send the audio frame stored in the temporary packet to the encoder.
686  * The output audio stream encoder is used to do this. */
687  error = avcodec_send_frame(output_codec_context, frame);
688  /* Check for errors, but proceed with fetching encoded samples if the
689  * encoder signals that it has nothing more to encode. */
690  if (error < 0 && error != AVERROR_EOF) {
691  fprintf(stderr, "Could not send packet for encoding (error '%s')\n",
692  av_err2str(error));
693  goto cleanup;
694  }
695 
696  /* Receive one encoded frame from the encoder. */
697  error = avcodec_receive_packet(output_codec_context, output_packet);
698  /* If the encoder asks for more data to be able to provide an
699  * encoded frame, return indicating that no data is present. */
700  if (error == AVERROR(EAGAIN)) {
701  error = 0;
702  goto cleanup;
703  /* If the last frame has been encoded, stop encoding. */
704  } else if (error == AVERROR_EOF) {
705  error = 0;
706  goto cleanup;
707  } else if (error < 0) {
708  fprintf(stderr, "Could not encode frame (error '%s')\n",
709  av_err2str(error));
710  goto cleanup;
711  /* Default case: Return encoded data. */
712  } else {
713  *data_present = 1;
714  }
715 
716  /* Write one audio frame from the temporary packet to the output file. */
717  if (*data_present &&
718  (error = av_write_frame(output_format_context, output_packet)) < 0) {
719  fprintf(stderr, "Could not write frame (error '%s')\n",
720  av_err2str(error));
721  goto cleanup;
722  }
723 
724 cleanup:
725  av_packet_free(&output_packet);
726  return error;
727 }
728 
729 /**
730  * Load one audio frame from the FIFO buffer, encode and write it to the
731  * output file.
732  * @param fifo Buffer used for temporary storage
733  * @param output_format_context Format context of the output file
734  * @param output_codec_context Codec context of the output file
735  * @return Error code (0 if successful)
736  */
738  AVFormatContext *output_format_context,
739  AVCodecContext *output_codec_context)
740 {
741  /* Temporary storage of the output samples of the frame written to the file. */
742  AVFrame *output_frame;
743  /* Use the maximum number of possible samples per frame.
744  * If there is less than the maximum possible frame size in the FIFO
745  * buffer use this number. Otherwise, use the maximum possible frame size. */
746  const int frame_size = FFMIN(av_audio_fifo_size(fifo),
747  output_codec_context->frame_size);
748  int data_written;
749 
750  /* Initialize temporary storage for one output frame. */
751  if (init_output_frame(&output_frame, output_codec_context, frame_size))
752  return AVERROR_EXIT;
753 
754  /* Read as many samples from the FIFO buffer as required to fill the frame.
755  * The samples are stored in the frame temporarily. */
756  if (av_audio_fifo_read(fifo, (void **)output_frame->data, frame_size) < frame_size) {
757  fprintf(stderr, "Could not read data from FIFO\n");
758  av_frame_free(&output_frame);
759  return AVERROR_EXIT;
760  }
761 
762  /* Encode one frame worth of audio samples. */
763  if (encode_audio_frame(output_frame, output_format_context,
764  output_codec_context, &data_written)) {
765  av_frame_free(&output_frame);
766  return AVERROR_EXIT;
767  }
768  av_frame_free(&output_frame);
769  return 0;
770 }
771 
772 /**
773  * Write the trailer of the output file container.
774  * @param output_format_context Format context of the output file
775  * @return Error code (0 if successful)
776  */
777 static int write_output_file_trailer(AVFormatContext *output_format_context)
778 {
779  int error;
780  if ((error = av_write_trailer(output_format_context)) < 0) {
781  fprintf(stderr, "Could not write output file trailer (error '%s')\n",
782  av_err2str(error));
783  return error;
784  }
785  return 0;
786 }
787 
788 int main(int argc, char **argv)
789 {
790  AVFormatContext *input_format_context = NULL, *output_format_context = NULL;
791  AVCodecContext *input_codec_context = NULL, *output_codec_context = NULL;
792  SwrContext *resample_context = NULL;
793  AVAudioFifo *fifo = NULL;
794  int ret = AVERROR_EXIT;
795 
796  if (argc != 3) {
797  fprintf(stderr, "Usage: %s <input file> <output file>\n", argv[0]);
798  exit(1);
799  }
800 
801  /* Open the input file for reading. */
802  if (open_input_file(argv[1], &input_format_context,
803  &input_codec_context))
804  goto cleanup;
805  /* Open the output file for writing. */
806  if (open_output_file(argv[2], input_codec_context,
807  &output_format_context, &output_codec_context))
808  goto cleanup;
809  /* Initialize the resampler to be able to convert audio sample formats. */
810  if (init_resampler(input_codec_context, output_codec_context,
811  &resample_context))
812  goto cleanup;
813  /* Initialize the FIFO buffer to store audio samples to be encoded. */
814  if (init_fifo(&fifo, output_codec_context))
815  goto cleanup;
816  /* Write the header of the output file container. */
817  if (write_output_file_header(output_format_context))
818  goto cleanup;
819 
820  /* Loop as long as we have input samples to read or output samples
821  * to write; abort as soon as we have neither. */
822  while (1) {
823  /* Use the encoder's desired frame size for processing. */
824  const int output_frame_size = output_codec_context->frame_size;
825  int finished = 0;
826 
827  /* Make sure that there is one frame worth of samples in the FIFO
828  * buffer so that the encoder can do its work.
829  * Since the decoder's and the encoder's frame size may differ, we
830  * need to FIFO buffer to store as many frames worth of input samples
831  * that they make up at least one frame worth of output samples. */
832  while (av_audio_fifo_size(fifo) < output_frame_size) {
833  /* Decode one frame worth of audio samples, convert it to the
834  * output sample format and put it into the FIFO buffer. */
835  if (read_decode_convert_and_store(fifo, input_format_context,
836  input_codec_context,
837  output_codec_context,
838  resample_context, &finished))
839  goto cleanup;
840 
841  /* If we are at the end of the input file, we continue
842  * encoding the remaining audio samples to the output file. */
843  if (finished)
844  break;
845  }
846 
847  /* If we have enough samples for the encoder, we encode them.
848  * At the end of the file, we pass the remaining samples to
849  * the encoder. */
850  while (av_audio_fifo_size(fifo) >= output_frame_size ||
851  (finished && av_audio_fifo_size(fifo) > 0))
852  /* Take one frame worth of audio samples from the FIFO buffer,
853  * encode it and write it to the output file. */
854  if (load_encode_and_write(fifo, output_format_context,
855  output_codec_context))
856  goto cleanup;
857 
858  /* If we are at the end of the input file and have encoded
859  * all remaining samples, we can exit this loop and finish. */
860  if (finished) {
861  int data_written;
862  /* Flush the encoder as it may have delayed frames. */
863  do {
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 }
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:480
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:629
int avio_closep(AVIOContext **s)
Close the resource accessed by the AVIOContext *s, free it and set the pointer pointing to it to NULL...
audio channel layout utility functions
static AVFrame * frame
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:274
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:429
int avcodec_receive_frame(AVCodecContext *avctx, AVFrame *frame)
Return decoded output data from a decoder.
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.
int swr_alloc_set_opts2(struct SwrContext **ps, AVChannelLayout *out_ch_layout, enum AVSampleFormat out_sample_fmt, int out_sample_rate, 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.
struct SwrContext SwrContext
The libswresample context.
Definition: swresample.h:189
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:389
AVChannelLayout ch_layout
Audio channel layout.
Definition: avcodec.h:2056
enum AVSampleFormat sample_fmt
audio sample format
Definition: avcodec.h:1014
AVRational pkt_timebase
Timebase in which pkt_dts/pts and AVPacket.dts/pts are.
Definition: avcodec.h:1746
int64_t bit_rate
the average bitrate
Definition: avcodec.h:439
int sample_rate
samples per second
Definition: avcodec.h:998
int flags
AV_CODEC_FLAG_*.
Definition: avcodec.h:469
int frame_size
Number of samples per channel in an audio frame.
Definition: avcodec.h:1026
enum AVCodecID codec_id
Specific type of the encoded data (the codec used).
Definition: codec_par.h:61
AVCodec.
Definition: codec.h:196
enum AVSampleFormat * sample_fmts
array of supported sample formats, or NULL if unknown, array is terminated by -1
Definition: codec.h:220
Format I/O context.
Definition: avformat.h:1213
This structure describes decoded (raw) audio or video data.
Definition: frame.h:325
int nb_samples
number of audio samples (per channel) described by this frame
Definition: frame.h:405
int64_t pts
Presentation timestamp in time_base units (time when frame should be shown to user).
Definition: frame.h:432
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: frame.h:346
uint8_t ** extended_data
pointers to the data planes/channels.
Definition: frame.h:386
Bytestream IO Context.
Definition: avio.h:162
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:948
AVCodecParameters * codecpar
Codec parameters associated with this stream.
Definition: avformat.h:1108
AVRational time_base
This is the fundamental unit of time (in seconds) in terms of which frame timestamps are represented.
Definition: avformat.h:978
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:48
#define OUTPUT_CHANNELS
Definition: transcode_aac.c:50
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:59
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.