transcode_aac.c

Go to the documentation of this file.

/*
 * Copyright (c) 2013-2022 Andreas Unterweger
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */
 
/**
 * @file
 * Simple audio converter
 *
 * @example transcode_aac.c
 * Convert an input audio file to AAC in an MP4 container using FFmpeg.
 * Formats other than MP4 are supported based on the output file extension.
 * @author Andreas Unterweger (dustsigns@gmail.com)
 */
 
#include <stdio.h>
 
#include "libavformat/avformat.h"
#include "libavformat/avio.h"
 
#include "libavcodec/avcodec.h"
 
#include "libavutil/audio_fifo.h"
#include "libavutil/avassert.h"
#include "libavutil/avstring.h"
#include "libavutil/channel_layout.h"
#include "libavutil/frame.h"
#include "libavutil/opt.h"
 
#include "libswresample/swresample.h"
 
/* The output bit rate in bit/s */
#define OUTPUT_BIT_RATE 96000
/* The number of output channels */
#define OUTPUT_CHANNELS 2
 
/**
 * Open an input file and the required decoder.
 * @param      filename             File to be opened
 * @param[out] input_format_context Format context of opened file
 * @param[out] input_codec_context  Codec context of opened file
 * @return Error code (0 if successful)
 */
static int open_input_file(const char *filename,
                           AVFormatContext **input_format_context,
                           AVCodecContext **input_codec_context)
{
    AVCodecContext *avctx;
    const AVCodec *input_codec;
    const AVStream *stream;
    int error;
 
    /* Open the input file to read from it. */
    if ((error = avformat_open_input(input_format_context, filename, NULL,
                                     NULL)) < 0) {
        fprintf(stderr, "Could not open input file '%s' (error '%s')\n",
                filename, av_err2str(error));
        *input_format_context = NULL;
        return error;
    }
 
    /* Get information on the input file (number of streams etc.). */
    if ((error = avformat_find_stream_info(*input_format_context, NULL)) < 0) {
        fprintf(stderr, "Could not open find stream info (error '%s')\n",
                av_err2str(error));
        avformat_close_input(input_format_context);
        return error;
    }
 
    /* Make sure that there is only one stream in the input file. */
    if ((*input_format_context)->nb_streams != 1) {
        fprintf(stderr, "Expected one audio input stream, but found %d\n",
                (*input_format_context)->nb_streams);
        avformat_close_input(input_format_context);
        return AVERROR_EXIT;
    }
 
    stream = (*input_format_context)->streams[0];
 
    /* Find a decoder for the audio stream. */
    if (!(input_codec = avcodec_find_decoder(stream->codecpar->codec_id))) {
        fprintf(stderr, "Could not find input codec\n");
        avformat_close_input(input_format_context);
        return AVERROR_EXIT;
    }
 
    /* Allocate a new decoding context. */
    avctx = avcodec_alloc_context3(input_codec);
    if (!avctx) {
        fprintf(stderr, "Could not allocate a decoding context\n");
        avformat_close_input(input_format_context);
        return AVERROR(ENOMEM);
    }
 
    /* Initialize the stream parameters with demuxer information. */
    error = avcodec_parameters_to_context(avctx, stream->codecpar);
    if (error < 0) {
        avformat_close_input(input_format_context);
        avcodec_free_context(&avctx);
        return error;
    }
 
    /* Open the decoder for the audio stream to use it later. */
    if ((error = avcodec_open2(avctx, input_codec, NULL)) < 0) {
        fprintf(stderr, "Could not open input codec (error '%s')\n",
                av_err2str(error));
        avcodec_free_context(&avctx);
        avformat_close_input(input_format_context);
        return error;
    }
 
    /* Set the packet timebase for the decoder. */
    avctx->pkt_timebase = stream->time_base;
 
    /* Save the decoder context for easier access later. */
    *input_codec_context = avctx;
 
    return 0;
}
 
/**
 * Open an output file and the required encoder.
 * Also set some basic encoder parameters.
 * Some of these parameters are based on the input file's parameters.
 * @param      filename              File to be opened
 * @param      input_codec_context   Codec context of input file
 * @param[out] output_format_context Format context of output file
 * @param[out] output_codec_context  Codec context of output file
 * @return Error code (0 if successful)
 */
static int open_output_file(const char *filename,
                            AVCodecContext *input_codec_context,
                            AVFormatContext **output_format_context,
                            AVCodecContext **output_codec_context)
{
    AVCodecContext *avctx          = NULL;
    AVIOContext *output_io_context = NULL;
    AVStream *stream               = NULL;
    const AVCodec *output_codec    = NULL;
    int error;
 
    /* Open the output file to write to it. */
    if ((error = avio_open(&output_io_context, filename,
                           AVIO_FLAG_WRITE)) < 0) {
        fprintf(stderr, "Could not open output file '%s' (error '%s')\n",
                filename, av_err2str(error));
        return error;
    }
 
    /* Create a new format context for the output container format. */
    if (!(*output_format_context = avformat_alloc_context())) {
        fprintf(stderr, "Could not allocate output format context\n");
        return AVERROR(ENOMEM);
    }
 
    /* Associate the output file (pointer) with the container format context. */
    (*output_format_context)->pb = output_io_context;
 
    /* Guess the desired container format based on the file extension. */
    if (!((*output_format_context)->oformat = av_guess_format(NULL, filename,
                                                              NULL))) {
        fprintf(stderr, "Could not find output file format\n");
        goto cleanup;
    }
 
    if (!((*output_format_context)->url = av_strdup(filename))) {
        fprintf(stderr, "Could not allocate url.\n");
        error = AVERROR(ENOMEM);
        goto cleanup;
    }
 
    /* Find the encoder to be used by its name. */
    if (!(output_codec = avcodec_find_encoder(AV_CODEC_ID_AAC))) {
        fprintf(stderr, "Could not find an AAC encoder.\n");
        goto cleanup;
    }
 
    /* Create a new audio stream in the output file container. */
    if (!(stream = avformat_new_stream(*output_format_context, NULL))) {
        fprintf(stderr, "Could not create new stream\n");
        error = AVERROR(ENOMEM);
        goto cleanup;
    }
 
    avctx = avcodec_alloc_context3(output_codec);
    if (!avctx) {
        fprintf(stderr, "Could not allocate an encoding context\n");
        error = AVERROR(ENOMEM);
        goto cleanup;
    }
 
    /* Set the basic encoder parameters.
     * The input file's sample rate is used to avoid a sample rate conversion. */
    av_channel_layout_default(&avctx->ch_layout, OUTPUT_CHANNELS);
    avctx->sample_rate    = input_codec_context->sample_rate;
    avctx->sample_fmt     = output_codec->sample_fmts[0];
    avctx->bit_rate       = OUTPUT_BIT_RATE;
 
    /* Set the sample rate for the container. */
    stream->time_base.den = input_codec_context->sample_rate;
    stream->time_base.num = 1;
 
    /* Some container formats (like MP4) require global headers to be present.
     * Mark the encoder so that it behaves accordingly. */
    if ((*output_format_context)->oformat->flags & AVFMT_GLOBALHEADER)
        avctx->flags |= AV_CODEC_FLAG_GLOBAL_HEADER;
 
    /* Open the encoder for the audio stream to use it later. */
    if ((error = avcodec_open2(avctx, output_codec, NULL)) < 0) {
        fprintf(stderr, "Could not open output codec (error '%s')\n",
                av_err2str(error));
        goto cleanup;
    }
 
    error = avcodec_parameters_from_context(stream->codecpar, avctx);
    if (error < 0) {
        fprintf(stderr, "Could not initialize stream parameters\n");
        goto cleanup;
    }
 
    /* Save the encoder context for easier access later. */
    *output_codec_context = avctx;
 
    return 0;
 
cleanup:
    avcodec_free_context(&avctx);
    avio_closep(&(*output_format_context)->pb);
    avformat_free_context(*output_format_context);
    *output_format_context = NULL;
    return error < 0 ? error : AVERROR_EXIT;
}
 
/**
 * Initialize one data packet for reading or writing.
 * @param[out] packet Packet to be initialized
 * @return Error code (0 if successful)
 */
static int init_packet(AVPacket **packet)
{
    if (!(*packet = av_packet_alloc())) {
        fprintf(stderr, "Could not allocate packet\n");
        return AVERROR(ENOMEM);
    }
    return 0;
}
 
/**
 * Initialize one audio frame for reading from the input file.
 * @param[out] frame Frame to be initialized
 * @return Error code (0 if successful)
 */
static int init_input_frame(AVFrame **frame)
{
    if (!(*frame = av_frame_alloc())) {
        fprintf(stderr, "Could not allocate input frame\n");
        return AVERROR(ENOMEM);
    }
    return 0;
}
 
/**
 * Initialize the audio resampler based on the input and output codec settings.
 * If the input and output sample formats differ, a conversion is required
 * libswresample takes care of this, but requires initialization.
 * @param      input_codec_context  Codec context of the input file
 * @param      output_codec_context Codec context of the output file
 * @param[out] resample_context     Resample context for the required conversion
 * @return Error code (0 if successful)
 */
static int init_resampler(AVCodecContext *input_codec_context,
                          AVCodecContext *output_codec_context,
                          SwrContext **resample_context)
{
        int error;
 
        /*
         * Create a resampler context for the conversion.
         * Set the conversion parameters.
         */
        error = swr_alloc_set_opts2(resample_context,
                                             &output_codec_context->ch_layout,
                                              output_codec_context->sample_fmt,
                                              output_codec_context->sample_rate,
                                             &input_codec_context->ch_layout,
                                              input_codec_context->sample_fmt,
                                              input_codec_context->sample_rate,
                                              0, NULL);
        if (error < 0) {
            fprintf(stderr, "Could not allocate resample context\n");
            return error;
        }
        /*
        * Perform a sanity check so that the number of converted samples is
        * not greater than the number of samples to be converted.
        * If the sample rates differ, this case has to be handled differently
        */
        av_assert0(output_codec_context->sample_rate == input_codec_context->sample_rate);
 
        /* Open the resampler with the specified parameters. */
        if ((error = swr_init(*resample_context)) < 0) {
            fprintf(stderr, "Could not open resample context\n");
            swr_free(resample_context);
            return error;
        }
    return 0;
}
 
/**
 * Initialize a FIFO buffer for the audio samples to be encoded.
 * @param[out] fifo                 Sample buffer
 * @param      output_codec_context Codec context of the output file
 * @return Error code (0 if successful)
 */
static int init_fifo(AVAudioFifo **fifo, AVCodecContext *output_codec_context)
{
    /* Create the FIFO buffer based on the specified output sample format. */
    if (!(*fifo = av_audio_fifo_alloc(output_codec_context->sample_fmt,
                                      output_codec_context->ch_layout.nb_channels, 1))) {
        fprintf(stderr, "Could not allocate FIFO\n");
        return AVERROR(ENOMEM);
    }
    return 0;
}
 
/**
 * Write the header of the output file container.
 * @param output_format_context Format context of the output file
 * @return Error code (0 if successful)
 */
static int write_output_file_header(AVFormatContext *output_format_context)
{
    int error;
    if ((error = avformat_write_header(output_format_context, NULL)) < 0) {
        fprintf(stderr, "Could not write output file header (error '%s')\n",
                av_err2str(error));
        return error;
    }
    return 0;
}
 
/**
 * Decode one audio frame from the input file.
 * @param      frame                Audio frame to be decoded
 * @param      input_format_context Format context of the input file
 * @param      input_codec_context  Codec context of the input file
 * @param[out] data_present         Indicates whether data has been decoded
 * @param[out] finished             Indicates whether the end of file has
 *                                  been reached and all data has been
 *                                  decoded. If this flag is false, there
 *                                  is more data to be decoded, i.e., this
 *                                  function has to be called again.
 * @return Error code (0 if successful)
 */
static int decode_audio_frame(AVFrame *frame,
                              AVFormatContext *input_format_context,
                              AVCodecContext *input_codec_context,
                              int *data_present, int *finished)
{
    /* Packet used for temporary storage. */
    AVPacket *input_packet;
    int error;
 
    error = init_packet(&input_packet);
    if (error < 0)
        return error;
 
    *data_present = 0;
    *finished = 0;
    /* Read one audio frame from the input file into a temporary packet. */
    if ((error = av_read_frame(input_format_context, input_packet)) < 0) {
        /* If we are at the end of the file, flush the decoder below. */
        if (error == AVERROR_EOF)
            *finished = 1;
        else {
            fprintf(stderr, "Could not read frame (error '%s')\n",
                    av_err2str(error));
            goto cleanup;
        }
    }
 
    /* Send the audio frame stored in the temporary packet to the decoder.
     * The input audio stream decoder is used to do this. */
    if ((error = avcodec_send_packet(input_codec_context, input_packet)) < 0) {
        fprintf(stderr, "Could not send packet for decoding (error '%s')\n",
                av_err2str(error));
        goto cleanup;
    }
 
    /* Receive one frame from the decoder. */
    error = avcodec_receive_frame(input_codec_context, frame);
    /* If the decoder asks for more data to be able to decode a frame,
     * return indicating that no data is present. */
    if (error == AVERROR(EAGAIN)) {
        error = 0;
        goto cleanup;
    /* If the end of the input file is reached, stop decoding. */
    } else if (error == AVERROR_EOF) {
        *finished = 1;
        error = 0;
        goto cleanup;
    } else if (error < 0) {
        fprintf(stderr, "Could not decode frame (error '%s')\n",
                av_err2str(error));
        goto cleanup;
    /* Default case: Return decoded data. */
    } else {
        *data_present = 1;
        goto cleanup;
    }
 
cleanup:
    av_packet_free(&input_packet);
    return error;
}
 
/**
 * Initialize a temporary storage for the specified number of audio samples.
 * The conversion requires temporary storage due to the different format.
 * The number of audio samples to be allocated is specified in frame_size.
 * @param[out] converted_input_samples Array of converted samples. The
 *                                     dimensions are reference, channel
 *                                     (for multi-channel audio), sample.
 * @param      output_codec_context    Codec context of the output file
 * @param      frame_size              Number of samples to be converted in
 *                                     each round
 * @return Error code (0 if successful)
 */
static int init_converted_samples(uint8_t ***converted_input_samples,
                                  AVCodecContext *output_codec_context,
                                  int frame_size)
{
    int error;
 
    /* Allocate as many pointers as there are audio channels.
     * Each pointer will later point to the audio samples of the corresponding
     * channels (although it may be NULL for interleaved formats).
     */
    if (!(*converted_input_samples = calloc(output_codec_context->ch_layout.nb_channels,
                                            sizeof(**converted_input_samples)))) {
        fprintf(stderr, "Could not allocate converted input sample pointers\n");
        return AVERROR(ENOMEM);
    }
 
    /* Allocate memory for the samples of all channels in one consecutive
     * block for convenience. */
    if ((error = av_samples_alloc(*converted_input_samples, NULL,
                                  output_codec_context->ch_layout.nb_channels,
                                  frame_size,
                                  output_codec_context->sample_fmt, 0)) < 0) {
        fprintf(stderr,
                "Could not allocate converted input samples (error '%s')\n",
                av_err2str(error));
        av_freep(&(*converted_input_samples)[0]);
        free(*converted_input_samples);
        return error;
    }
    return 0;
}
 
/**
 * Convert the input audio samples into the output sample format.
 * The conversion happens on a per-frame basis, the size of which is
 * specified by frame_size.
 * @param      input_data       Samples to be decoded. The dimensions are
 *                              channel (for multi-channel audio), sample.
 * @param[out] converted_data   Converted samples. The dimensions are channel
 *                              (for multi-channel audio), sample.
 * @param      frame_size       Number of samples to be converted
 * @param      resample_context Resample context for the conversion
 * @return Error code (0 if successful)
 */
static int convert_samples(const uint8_t **input_data,
                           uint8_t **converted_data, const int frame_size,
                           SwrContext *resample_context)
{
    int error;
 
    /* Convert the samples using the resampler. */
    if ((error = swr_convert(resample_context,
                             converted_data, frame_size,
                             input_data    , frame_size)) < 0) {
        fprintf(stderr, "Could not convert input samples (error '%s')\n",
                av_err2str(error));
        return error;
    }
 
    return 0;
}
 
/**
 * Add converted input audio samples to the FIFO buffer for later processing.
 * @param fifo                    Buffer to add the samples to
 * @param converted_input_samples Samples to be added. The dimensions are channel
 *                                (for multi-channel audio), sample.
 * @param frame_size              Number of samples to be converted
 * @return Error code (0 if successful)
 */
static int add_samples_to_fifo(AVAudioFifo *fifo,
                               uint8_t **converted_input_samples,
                               const int frame_size)
{
    int error;
 
    /* Make the FIFO as large as it needs to be to hold both,
     * the old and the new samples. */
    if ((error = av_audio_fifo_realloc(fifo,