什么是重采样
所谓的重采样，就是改变音频的采样率、sample format、声道数等参数，使之按照我们期望的参数输出
为什么要重采样
为什么要重采样？当然是原有的音频参数不满足我们的需求
比如在FFmpeg解码音频的时候，不同的音源有不同的格式，采样率等，在解码后的数据中的这些参数也会不一致(最新FFmpeg 解码音频后，音频格式为AV_SAMPLE_FMT_FLTP，这个参数应该是⼀致的)，如果我们接下来需要使用解码后的音频数据做 其他操作，而这些参数的不一致导致会有很多额外工作，此时直接对其进行重采样，获取我们制定的音频参数，这样就会方便很多。
再比如如在将音频进行SDL播放时候，因为当前的SDL2.0不支持planar格式，也不支持浮点型的，而最新的 FFMPEG 16年会将音频解码为AV_SAMPLE_FMT_FLTP格式，因此此时就需要我们对其重采样，使之可以在SDL2.0上进行播放。
可调节的参数
通过重采样，我们可以对：
- sample rate(采样率)
- sample format(采样格式)
- channel layout(通道布局，可以通过此参数获取声道数
对应参数解析
采样率
采样设备每秒抽取样本的次数
采样格式及量化精度（位宽）
每种音频格式有不同的量化精度（位宽），位数越多，表示值就越精确，声音表现自然就越精准。 FFMpeg中音频格式有以下几种，每种格式有其占用的字节数信息（libavutil/samplefmt.h）：
enum AVSampleFormat {
AV_SAMPLE_FMT_NONE = -1,
AV_SAMPLE_FMT_U8, ///< unsigned 8 bits
AV_SAMPLE_FMT_S16, ///< signed 16 bits
AV_SAMPLE_FMT_S32, ///< signed 32 bits
AV_SAMPLE_FMT_FLT, ///< float
AV_SAMPLE_FMT_DBL, ///< double
AV_SAMPLE_FMT_U8P, ///< unsigned 8 bits, planar
AV_SAMPLE_FMT_S16P, ///< signed 16 bits, planar
AV_SAMPLE_FMT_S32P, ///< signed 32 bits, planar
AV_SAMPLE_FMT_FLTP, ///< float, planar
AV_SAMPLE_FMT_DBLP, ///< double, planar
AV_SAMPLE_FMT_S64, ///< signed 64 bits
AV_SAMPLE_FMT_S64P, ///< signed 64 bits, planar
AV_SAMPLE_FMT_NB ///< Number of sample formats. DO NOT USE if linking dynamically
};分片（plane）和打包（packed）
以双声道为例，
带P（plane）的数据格式在存储时，其左声道和右声道的数据是分开存储的，左声道的 数据存储在data[0]，右声道的数据存储在data[1]，每个声道的所占⽤的字节数为linesize[0]和 linesize[1]；
不带P（packed）的⾳频数据在存储时，是按照LRLRLR…的格式交替存储在data[0]中，linesize[0] 表示总的数据量。
声道分布（channel_layout)
声道分布在FFmpeg\libavutil\channel_layout.h中有定义，⼀般来说用的比较多的是AV_CH_LAYOUT_STEREO（双声道）和AV_CH_LAYOUT_SURROUND（三声道），这两者的定义如下：
#define AV_CH_LAYOUT_STEREO (AV_CH_FRONT_LEFT|AV_CH_FRONT_RIGHT)
#define AV_CH_LAYOUT_SURROUND (AV_CH_LAYOUT_STEREO|AV_CH_FRONT_CENTER)音频帧的数据量计算
⼀帧⾳频的数据量（字节）=channel数 * nb_samples样本数 * 每个样本占用的字节数
如果该音频帧是FLTP格式的PCM数据，包含1024个样本，双声道，那么该音频帧包含的音频数据量是
210244=8192字节。
AV_SAMPLE_FMT_DBL ： 210248 = 16384
音频播放时间计算
以采样率44100Hz来计算，每秒44100个sample，⽽正常⼀帧为1024个sample，可知每帧播放时 间/1024=1000ms/44100，得到每帧播放时间=1024*1000/44100=23.2ms （更精确的是 23.21995464852608
⼀帧播放时间（毫秒） = nb_samples样本数 *1000/采样率 =
1024*1000/44100=23.21995464852608ms ->约等于 23.2ms，精度损失了 0.011995464852608ms，如果累计10万帧，误差>1199毫秒，如果有视频⼀起的就会有⾳视频同步的问 题。 如果按着23.2去计算pts（0 23.2 46.4 ）就会有累积误差。
1024*1000/48000=21.33333333333333ms
本文福利， 免费领取C++音视频学习资料包、技术视频/代码，内容包括（音视频开发，面试题，FFmpeg ，webRTC ，rtmp ，hls ，rtsp ，ffplay ，编解码，推拉流，srs）↓↓↓↓↓↓见下面↓↓文章底部点击免费领取↓↓
FFmpeg重采样API
分配音频重采样的上下文
struct SwrContext *swr_alloc(void);当设置好相关的参数后，使⽤此函数来初始化SwrContext结构体
int swr_init(struct SwrContext *s);
分配SwrContext并设置/重置常⽤的参数
struct SwrContext *swr_alloc_set_opts(struct SwrContext *s,// ⾳频重采样上下⽂
int64_t out_ch_layout, // 输出的layout, 如：5.1声道
enum AVSampleFormat out_sample_fmt,// 输出的采样格式。Float, S16,⼀般 选⽤是s16 绝⼤部分声卡⽀持
int out_sample_rate,//输出采样率
int64_t in_ch_layout, // 输⼊的layout
enum AVSampleFormat in_sample_fmt, // 输⼊的采样格式
int in_sample_rate,// 输⼊的采样率
int log_offset, // ⽇志相关，不⽤管先，直接为0
void *log_ctx // ⽇志相关，不⽤管先，直接为NULL
);将输⼊的音频按照定义的参数进进转换并输出
int attribute_align_arg swr_convert(struct SwrContext *s, // ⾳频重采样的上下⽂
uint8_t *out_arg[SWR_CH_MAX], // 输出的指针。传递的输出的数组
int out_count,//输出的样本数量，不是字节数。单通道的样本数量。
const uint8_t *in_arg [SWR_CH_MAX], //输⼊的数组，AVFrame解码出来的DATA
int in_count// 输⼊的单通道的样本数量。
);
返回值 <= out_count in和in_count可以设置为0，以最后刷新最后⼏个样本。
释放掉SwrContext结构体并将此结构体置为NULL;
void swr_free(struct SwrContext **s);
音频频重采样，采样格式转换和混合库。
与lswr的交互是通过SwrContext完成的，SwrContext被分配给swr_alloc（）或 swr_alloc_set_opts（）。 它是不透明的，所以所有参数必须使⽤AVOptions API设置。 为了使⽤lswr，你需要做的第⼀件事就是分配SwrContext。 这可以使⽤swr_alloc（）或 swr_alloc_set_opts（）来完成。 如果您使⽤前者，则必须通过AVOptions API设置选项。 后⼀个函数 提供了相同的功能，但它允许您在同⼀语句中设置⼀些常⽤选项。
例如，以下代码将设置从平面浮动样本格式到交织的带符号16位整数的转换，从48kHz到44.1kHz的下采 样，以及从5.1声道到⽴体声的下混合（使⽤默认混合矩阵）。 这是使⽤swr_alloc（）函数。
SwrContext *swr = swr_alloc();
av_opt_set_channel_layout(swr, "in_channel_layout", AV_CH_LAYOUT_ 5POINT1, 0);
av_opt_set_channel_layout(swr, "out_channel_layout", AV_CH_LAYOUT_ STEREO, 0);
av_opt_set_int(swr, "in_sample_rate", 48000, 0) ;
av_opt_set_int(swr, "out_sample_rate", 44100, 0) ;
av_opt_set_sample_fmt(swr, "in_sample_fmt", AV_SAMPLE_FMT_FLTP, 0 );
av_opt_set_sample_fmt(swr, "out_sample_fmt", AV_SAMPLE_FMT_S16, 0 );同样的工作也可以使⽤swr_alloc_set_opts（）：
SwrContext *swr_alloc_set_opts(NULL,// we're allocating a new context
AV_CH_LAYOUT_STEREO, // out_ch_layout
AV_SAMPLE_FMT_S16, // out_sample_fmt
44100, // out_sample_rate
AV_CH_LAYOUT_5POINT1, // in_ch_layout
AV_SAMPLE_FMT_FLTP, // in_sample_fmt
48000, // in_sample_rate
0, // log_offset
NULL// log_ctx
);一旦设置了所有值，它必须⽤swr_init（）初始化。 如果需要更改转换参数，可以使⽤ AVOptions来更改参数，如上面第⼀个例子所述; 或者使⽤swr_alloc_set_opts（），但是第 ⼀个参数是分配的上下文。 您必须再次调用swr_init（）。
转换本身通过重复调用swr_convert（）来完成。 请注意，如果提供的输出空间不足或采样率转换完成 后，样本可能会在swr中缓冲，这需要“未来”样本。 可以随时通过使⽤swr_convert（）（in_count可以 设置为0）来检索不需要将来输⼊的样本。 在转换结束时，可以通过调⽤具有NULL in和in incount的 swr_convert（）来刷新重采样缓冲区。
音频重采样工程范例
简单范例（resample）
FFMpeg⾃带的resample例⼦：FFmpeg\doc\examples\resampling_audio.c，这里把最核心的 resample代码贴⼀下，在工程中使用时，注意设置的各种参数，给定的输⼊数据都不能错。
/**
* @example resampling_audio.c
* libswresample API use example.
*/
#include <libavutil/opt.h>
#include <libavutil/channel_layout.h>
#include <libavutil/samplefmt.h>
#include <libswresample/swresample.h>
static int get_format_from_sample_fmt(const char **fmt,
enum AVSampleFormat sample_fmt)
{
int i;
struct sample_fmt_entry {
enum AVSampleFormat sample_fmt; const char *fmt_be, *fmt_le;
} sample_fmt_entries[] = {
{ AV_SAMPLE_FMT_U8, "u8", "u8" },
{ AV_SAMPLE_FMT_S16, "s16be", "s16le" },
{ AV_SAMPLE_FMT_S32, "s32be", "s32le" },
{ AV_SAMPLE_FMT_FLT, "f32be", "f32le" },
{ AV_SAMPLE_FMT_DBL, "f64be", "f64le" },
};
*fmt = NULL;
for (i = 0; i < FF_ARRAY_ELEMS(sample_fmt_entries); i++) {
struct sample_fmt_entry *entry = &sample_fmt_entries[i];
if (sample_fmt == entry->sample_fmt) {
*fmt = AV_NE(entry->fmt_be, entry->fmt_le);
return 0;
}
}
fprintf(stderr,
"Sample format %s not supported as output format\n",
av_get_sample_fmt_name(sample_fmt));
return AVERROR(EINVAL);
}
/**
* Fill dst buffer with nb_samples, generated starting from t. 交错模式的
*/
static void fill_samples(double *dst, int nb_samples, int nb_channels, int sample_rate, double *t)
{
int i, j;
double tincr = 1.0 / sample_rate, *dstp = dst;
const double c = 2 * M_PI * 440.0;
/* generate sin tone with 440Hz frequency and duplicated channels */
for (i = 0; i < nb_samples; i++) {
*dstp = sin(c * *t);
for (j = 1; j < nb_channels; j++)
dstp[j] = dstp[0];
dstp += nb_channels;
*t += tincr;
}
}
int main(int argc, char **argv)
{
// 输入参数
int64_t src_ch_layout = AV_CH_LAYOUT_STEREO;
int src_rate = 48000;
enum AVSampleFormat src_sample_fmt = AV_SAMPLE_FMT_DBL;
int src_nb_channels = 0;
uint8_t **src_data = NULL; // 二级指针
int src_linesize;
int src_nb_samples = 1024;
// 输出参数
int64_t dst_ch_layout = AV_CH_LAYOUT_STEREO;
int dst_rate = 44100;
enum AVSampleFormat dst_sample_fmt = AV_SAMPLE_FMT_S16;
int dst_nb_channels = 0;
uint8_t **dst_data = NULL; //二级指针
int dst_linesize;
int dst_nb_samples;
int max_dst_nb_samples;
// 输出文件
const char *dst_filename = NULL; // 保存输出的pcm到本地，然后播放验证
FILE *dst_file;
int dst_bufsize;
const char *fmt;
// 重采样实例
struct SwrContext *swr_ctx;
double t;
int ret;
if (argc != 2) {
fprintf(stderr, "Usage: %s output_file\n"
"API example program to show how to resample an audio stream with libswresample.\n"
"This program generates a series of audio frames, resamples them to a specified "
"output format and rate and saves them to an output file named output_file.\n",
argv[0]);
exit(1);
}
dst_filename = argv[1];
dst_file = fopen(dst_filename, "wb");
if (!dst_file) {
fprintf(stderr, "Could not open destination file %s\n", dst_filename);
exit(1);
}
// 创建重采样器
/* create resampler context */
swr_ctx = swr_alloc();
if (!swr_ctx) {
fprintf(stderr, "Could not allocate resampler context\n");
ret = AVERROR(ENOMEM);
goto end;
}
// 设置重采样参数
/* set options */
// 输入参数
av_opt_set_int(swr_ctx, "in_channel_layout", src_ch_layout, 0);
av_opt_set_int(swr_ctx, "in_sample_rate", src_rate, 0);
av_opt_set_sample_fmt(swr_ctx, "in_sample_fmt", src_sample_fmt, 0);
// 输出参数
av_opt_set_int(swr_ctx, "out_channel_layout", dst_ch_layout, 0);
av_opt_set_int(swr_ctx, "out_sample_rate", dst_rate, 0);
av_opt_set_sample_fmt(swr_ctx, "out_sample_fmt", dst_sample_fmt, 0);
// 初始化重采样
/* initialize the resampling context */
if ((ret = swr_init(swr_ctx)) < 0) {
fprintf(stderr, "Failed to initialize the resampling context\n");
goto end;
}
/* allocate source and destination samples buffers */
// 计算出输入源的通道数量
src_nb_channels = av_get_channel_layout_nb_channels(src_ch_layout);
// 给输入源分配内存空间
ret = av_samples_alloc_array_and_samples(&src_data, &src_linesize, src_nb_channels,
src_nb_samples, src_sample_fmt, 0);
if (ret < 0) {
fprintf(stderr, "Could not allocate source samples\n");
goto end;
}
/* compute the number of converted samples: buffering is avoided
* ensuring that the output buffer will contain at least all the
* converted input samples */
// 计算输出采样数量
max_dst_nb_samples = dst_nb_samples =
av_rescale_rnd(src_nb_samples, dst_rate, src_rate, AV_ROUND_UP);
/* buffer is going to be directly written to a rawaudio file, no alignment */
dst_nb_channels = av_get_channel_layout_nb_channels(dst_ch_layout);
// 分配输出缓存内存
ret = av_samples_alloc_array_and_samples(&dst_data, &dst_linesize, dst_nb_channels,
dst_nb_samples, dst_sample_fmt, 0);
if (ret < 0) {
fprintf(stderr, "Could not allocate destination samples\n");
goto end;
}
t = 0;
do {
/* generate synthetic audio */
// 生成输入源
fill_samples((double *)src_data[0], src_nb_samples, src_nb_channels, src_rate, &t);
/* compute destination number of samples */
int64_t delay = swr_get_delay(swr_ctx, src_rate);
dst_nb_samples = av_rescale_rnd(delay + src_nb_samples, dst_rate, src_rate, AV_ROUND_UP);
if (dst_nb_samples > max_dst_nb_samples) {
av_freep(&dst_data[0]);
ret = av_samples_alloc(dst_data, &dst_linesize, dst_nb_channels,
dst_nb_samples, dst_sample_fmt, 1);
if (ret < 0)
break;
max_dst_nb_samples = dst_nb_samples;
}
// int fifo_size = swr_get_out_samples(swr_ctx,src_nb_samples);
// printf("fifo_size:%d\n", fifo_size);
// if(fifo_size < 1024)
// continue;
/* convert to destination format */
// ret = swr_convert(swr_ctx, dst_data, dst_nb_samples, (const uint8_t **)src_data, src_nb_samples);
ret = swr_convert(swr_ctx, dst_data, dst_nb_samples, (const uint8_t **)src_data, src_nb_samples);
if (ret < 0) {
fprintf(stderr, "Error while converting\n");
goto end;
}
dst_bufsize = av_samples_get_buffer_size(&dst_linesize, dst_nb_channels,
ret, dst_sample_fmt, 1);
if (dst_bufsize < 0) {
fprintf(stderr, "Could not get sample buffer size\n");
goto end;
}
printf("t:%f in:%d out:%d\n", t, src_nb_samples, ret);
fwrite(dst_data[0], 1, dst_bufsize, dst_file);
} while (t < 10);
ret = swr_convert(swr_ctx, dst_data, dst_nb_samples, NULL, 0);
if (ret < 0) {
fprintf(stderr, "Error while converting\n");
goto end;
}
dst_bufsize = av_samples_get_buffer_size(&dst_linesize, dst_nb_channels,
ret, dst_sample_fmt, 1);
if (dst_bufsize < 0) {
fprintf(stderr, "Could not get sample buffer size\n");
goto end;
}
printf("flush in:%d out:%d\n", 0, ret);
fwrite(dst_data[0], 1, dst_bufsize, dst_file);
if ((ret = get_format_from_sample_fmt(&fmt, dst_sample_fmt)) < 0)
goto end;
fprintf(stderr, "Resampling succeeded. Play the output file with the command:\n"
"ffplay -f %s -channel_layout %"PRId64" -channels %d -ar %d %s\n",
fmt, dst_ch_layout, dst_nb_channels, dst_rate, dst_filename);
end:
fclose(dst_file);
if (src_data)
av_freep(&src_data[0]);
av_freep(&src_data);
if (dst_data)
av_freep(&dst_data[0]);
av_freep(&dst_data);
swr_free(&swr_ctx);
return ret < 0;
}复杂范例
重采样器的设计原理：
- 支持重采样参数的设置
- 支持不同形式的源输入，比如使用frame，**in_data
- 支持不同形式的输出， 比如使用frame，**out_data
- 支持pts
- 支持获取指定采样点数
- 支持flush
数据结构
audioresampler.h
#ifndef AUDIORESAMPLER_H
#define AUDIORESAMPLER_H
#include "libavutil/audio_fifo.h"
#include "libavutil/opt.h"
#include "libavutil/avutil.h"
#include "libswresample/swresample.h"
#include "libavutil/error.h"
#include "libavutil/frame.h"
#include "libavcodec/avcodec.h"
// 重采样的参数
typedef struct audio_resampler_params {
// 输入参数
enum AVSampleFormat src_sample_fmt;
int src_sample_rate;
uint64_t src_channel_layout;
// 输出参数
enum AVSampleFormat dst_sample_fmt;
int dst_sample_rate;
uint64_t dst_channel_layout;
}audio_resampler_params_t;
// 封装的重采样器
typedef struct audio_resampler {
struct SwrContext *swr_ctx; // 重采样的核心
audio_resampler_params_t resampler_params; // 重采样的设置参数
int is_fifo_only; // 不需要进行重采样，只需要缓存到 audio_fifo
int is_flushed; // flush的时候使用
AVAudioFifo *audio_fifo; // 采样点的缓存
int64_t start_pts; // 起始pts
int64_t cur_pts; // 当前pts
uint8_t **resampled_data; // 用来缓存重采样后的数据
int resampled_data_size; // 重采样后的采样数
int src_channels; // 输入的通道数
int dst_channels; // 输出通道数
int64_t total_resampled_num; // 统计总共的重采样点数，目前只是统计
}audio_resampler_t;
/**
* @brief 分配重采样器
* @param resampler_params 重采样的设置参数
* @return 成功返回重采样器；失败返回NULL
*/
audio_resampler_t *audio_resampler_alloc(const audio_resampler_params_t resampler_params);
/**
* @brief 释放重采样器
* @param resampler
*/
void audio_resampler_free(audio_resampler_t *resampler);
/**
* @brief 发送要进行重采样的帧
* @param resampler
* @param frame
* @return 这次重采样后得到的采样点数
*/
int audio_resampler_send_frame(audio_resampler_t *resampler, AVFrame *frame);
/**
* @brief 发送要进行重采样的帧
* @param resampler
* @param in_data 二级指针
* @param in_nb_samples 输入的采样点数量(单个通道)
* @param pts pts
* @return
*/
int audio_resampler_send_frame2(audio_resampler_t *resampler, uint8_t **in_data,int in_nb_samples, int64_t pts);
/**
* @brief 发送要进行重采样的帧
* @param resampler
* @param in_data 一级指针
* @param in_bytes 传入数据的字节大小
* @param pts
* @return
*/
int audio_resampler_send_frame3(audio_resampler_t *resampler, uint8_t *in_data,int in_bytes, int64_t pts);
/**
* @brief 获取重采样后的数据
* @param resampler
* @param nb_samples 我们需要读取多少采样数量: 如果nb_samples>0，只有audio fifo>=nb_samples才能获取到数据;nb_samples=0,有多少就给多少
*
* @return 如果获取到采样点数则非NULL
*/
AVFrame *audio_resampler_receive_frame(audio_resampler_t *resampler, int nb_samples);
/**
* @brief 获取重采样后的数据
* @param resampler
* @param out_data 缓存获取到的重采样数据，buf一定要充足
* @param nb_samples 我们需要读取多少采样数量: 如果nb_samples>0，只有audio fifo>=nb_samples才能获取到数据;nb_samples=0,有多少就给多少
* @param pts 获取帧的pts值
* @return 获取到的采样点数量
*/
int audio_resampler_receive_frame2(audio_resampler_t *resampler, uint8_t **out_data, int nb_samples, int64_t *pts);
/**
* @brief audio_resampler_get_fifo_size
* @param resampler
* @return audio fifo的缓存的采样点数量
*/
int audio_resampler_get_fifo_size(audio_resampler_t *resampler);
/**
* @brief audio_resampler_get_start_pts
* @param resampler
* @return 起始的pts
*/
int64_t audio_resampler_get_start_pts(audio_resampler_t *resampler);
/**
* @brief audio_resampler_get_cur_pts
* @param resampler
* @return 当前的pts
*/
int64_t audio_resampler_get_cur_pts(audio_resampler_t *resampler);
#endif // AUDIORESAMPLER_Haudioresampler.c
#include "audioresampler.h"
static AVFrame *alloc_out_frame(const int nb_samples, const audio_resampler_params_t *resampler_params)
{
int ret;
AVFrame * frame = av_frame_alloc();
if(!frame)
{
return NULL;
}
frame->nb_samples = nb_samples;
frame->channel_layout = resampler_params->dst_channel_layout;
frame->format = resampler_params->dst_sample_fmt;
frame->sample_rate = resampler_params->dst_sample_rate;
ret = av_frame_get_buffer(frame, 0);
if (ret < 0)
{
printf("cannot allocate audio data buffer\n");
return NULL;
}
return frame;
}
static AVFrame *get_one_frame(audio_resampler_t *resampler, const int nb_samples)
{
AVFrame * frame = alloc_out_frame(nb_samples, &resampler->resampler_params);
if (frame)
{
av_audio_fifo_read(resampler->audio_fifo, (void **)frame->data, nb_samples);
frame->pts = resampler->cur_pts;
resampler->cur_pts += nb_samples;
resampler->total_resampled_num += nb_samples;
}
return frame;
}
static int init_resampled_data(audio_resampler_t *resampler)
{
if (resampler->resampled_data)
av_freep(&resampler->resampled_data[0]);
av_freep(&resampler->resampled_data); // av_freep即使传入NULL也是安全的
int linesize = 0;
int ret= av_samples_alloc_array_and_samples(&resampler->resampled_data, &linesize, resampler->dst_channels,
resampler->resampled_data_size, resampler->resampler_params.dst_sample_fmt, 0);
if (ret < 0)
printf("fail accocate audio resampled data buffer\n");
return ret;
}
audio_resampler_t *audio_resampler_alloc(const audio_resampler_params_t resampler_params)
{
int ret = 0;
audio_resampler_t *audio_resampler = (audio_resampler_t *)av_malloc(sizeof(audio_resampler_t));
if(!audio_resampler) {
return NULL;
}
memset(audio_resampler, 0, sizeof(audio_resampler_t));
audio_resampler->resampler_params = resampler_params; // 设置参数
// 设置通道数量
audio_resampler->src_channels = av_get_channel_layout_nb_channels(resampler_params.src_channel_layout);
audio_resampler->dst_channels = av_get_channel_layout_nb_channels(resampler_params.dst_channel_layout);
audio_resampler->cur_pts = AV_NOPTS_VALUE;
audio_resampler->start_pts = AV_NOPTS_VALUE;
// 分配audio fifo，单位其实为samples 我们在av_audio_fifo_write时候如果buffer不足能自动扩充
audio_resampler->audio_fifo = av_audio_fifo_alloc(resampler_params.dst_sample_fmt,
audio_resampler->dst_channels, 1);
if(!audio_resampler->audio_fifo) {
printf("av_audio_fifo_alloc failed\n");
av_freep(audio_resampler);
return NULL;
}
if (resampler_params.src_sample_fmt == resampler_params.dst_sample_fmt &&
resampler_params.src_sample_rate == resampler_params.dst_sample_rate &&
resampler_params.src_channel_layout == resampler_params.dst_channel_layout)
{
printf("no resample needed, just use audio fifo\n");
audio_resampler->is_fifo_only = 1; // 不需要做重采样
return audio_resampler;
}
// 初始化重采样
audio_resampler->swr_ctx = swr_alloc();
if(!audio_resampler->swr_ctx) {
printf("swr_alloc failed\n");
return NULL;
}
/* set options */
av_opt_set_sample_fmt(audio_resampler->swr_ctx, "in_sample_fmt", resampler_params.src_sample_fmt, 0);
av_opt_set_int(audio_resampler->swr_ctx, "in_channel_layout", resampler_params.src_channel_layout, 0);
av_opt_set_int(audio_resampler->swr_ctx, "in_sample_rate", resampler_params.src_sample_rate, 0);
av_opt_set_sample_fmt(audio_resampler->swr_ctx, "out_sample_fmt", resampler_params.dst_sample_fmt, 0);
av_opt_set_int(audio_resampler->swr_ctx, "out_channel_layout", resampler_params.dst_channel_layout, 0);
av_opt_set_int(audio_resampler->swr_ctx, "out_sample_rate", resampler_params.dst_sample_rate, 0);
/* initialize the resampling context */
ret = swr_init(audio_resampler->swr_ctx);
if (ret < 0) {
printf("failed to initialize the resampling context.\n");
av_fifo_freep((AVFifoBuffer **)&audio_resampler->audio_fifo);
swr_free(&audio_resampler->swr_ctx);
av_freep(audio_resampler);
return NULL;
}
audio_resampler->resampled_data_size = 2048;
if (init_resampled_data(audio_resampler) < 0) {
swr_free(&audio_resampler->swr_ctx);
av_fifo_freep((AVFifoBuffer **)&audio_resampler->audio_fifo);
av_freep(audio_resampler);
return NULL;
}
printf("init done \n");
return audio_resampler;
}
void audio_resampler_free(audio_resampler_t *resampler)
{
if(!resampler) {
return;
}
if (resampler->swr_ctx)
swr_free(&resampler->swr_ctx);
if (resampler->audio_fifo) {
av_fifo_freep((AVFifoBuffer **)&resampler->audio_fifo);
}
if (resampler->resampled_data)
av_freep(&resampler->resampled_data[0]);
av_freep(&resampler->resampled_data);
av_freep(resampler);
resampler = NULL;
}
int audio_resampler_send_frame(audio_resampler_t *resampler, AVFrame *frame)
{
if(!resampler) {
return 0;
}
int src_nb_samples = 0;
uint8_t **src_data = NULL;
if (frame)
{
src_nb_samples = frame->nb_samples; // 输入源采样点数量
src_data = frame->extended_data; // 输入源的buffer
if (resampler->start_pts == AV_NOPTS_VALUE && frame->pts != AV_NOPTS_VALUE)
{
resampler->start_pts = frame->pts;
resampler->cur_pts = frame->pts;
}
}
else
{
resampler->is_flushed = 1;
}
if (resampler->is_fifo_only) {
// 如果不需要做重采样，原封不动写入fifo
return src_data ? av_audio_fifo_write(resampler->audio_fifo, (void **)src_data, src_nb_samples) : 0;
}
// 计算这次做重采样能够获取到的重采样后的点数
const int dst_nb_samples = av_rescale_rnd(swr_get_delay(resampler->swr_ctx, resampler->resampler_params.src_sample_rate) + src_nb_samples,
resampler->resampler_params.src_sample_rate, resampler->resampler_params.dst_sample_rate, AV_ROUND_UP);
if (dst_nb_samples > resampler->resampled_data_size)
{
//resampled_data
resampler->resampled_data_size = dst_nb_samples;
if (init_resampled_data(resampler) < 0) {
return AVERROR(ENOMEM);
}
}
int nb_samples = swr_convert(resampler->swr_ctx, resampler->resampled_data, dst_nb_samples,
(const uint8_t **)src_data, src_nb_samples);
// 返回实际写入的采样点数量
// return av_audio_fifo_write(resampler->audio_fifo, (void **)resampler->resampled_data, nb_samples);
int ret_size = av_audio_fifo_write(resampler->audio_fifo, (void **)resampler->resampled_data, nb_samples);
if(ret_size != nb_samples) {
printf("Warn：av_audio_fifo_write failed, expected_write:%d, actual_write:%d\n", nb_samples, ret_size);
}
return ret_size;
}
AVFrame *audio_resampler_receive_frame(audio_resampler_t *resampler, int nb_samples)
{
nb_samples = nb_samples == 0 ? av_audio_fifo_size(resampler->audio_fifo) : nb_samples;
if (av_audio_fifo_size(resampler->audio_fifo) < nb_samples || nb_samples == 0)
return NULL;
// 采样点数满足条件
return get_one_frame(resampler, nb_samples);
}
int audio_resampler_send_frame2(audio_resampler_t *resampler, uint8_t **in_data, int in_nb_samples, int64_t pts)
{
if(!resampler) {
return 0;
}
int src_nb_samples = 0;
uint8_t **src_data = NULL;
if (in_data)
{
src_nb_samples = in_nb_samples;
src_data = in_data;
if (resampler->start_pts == AV_NOPTS_VALUE && pts != AV_NOPTS_VALUE)
{
resampler->start_pts = pts;
resampler->cur_pts = pts;
}
}
else
{
resampler->is_flushed = 1;
}
if (resampler->is_fifo_only) {
return src_data ? av_audio_fifo_write(resampler->audio_fifo, (void **)src_data, src_nb_samples) : 0;
}
const int dst_nb_samples = av_rescale_rnd(swr_get_delay(resampler->swr_ctx, resampler->resampler_params.src_sample_rate) + src_nb_samples,
resampler->resampler_params.src_sample_rate, resampler->resampler_params.dst_sample_rate, AV_ROUND_UP);
if (dst_nb_samples > resampler->resampled_data_size)
{
//resampled_data
resampler->resampled_data_size = dst_nb_samples;
if (init_resampled_data(resampler) < 0)
return AVERROR(ENOMEM);
}
int nb_samples = swr_convert(resampler->swr_ctx, resampler->resampled_data, dst_nb_samples,
(const uint8_t **)src_data, src_nb_samples);
// 返回实际写入的采样点数量
return av_audio_fifo_write(resampler->audio_fifo, (void **)resampler->resampled_data, nb_samples);
}
int audio_resampler_send_frame3(audio_resampler_t *resampler, uint8_t *in_data, int in_bytes, int64_t pts)
{
if(!resampler) {
return 0;
}
AVFrame *frame = NULL;
if(in_data)
{
frame = av_frame_alloc();
frame->format = resampler->resampler_params.src_sample_fmt;
frame->channel_layout = resampler->resampler_params.src_channel_layout;
int ch = av_get_channel_layout_nb_channels(resampler->resampler_params.src_channel_layout);
frame->nb_samples = in_bytes/ av_get_bytes_per_sample(resampler->resampler_params.src_sample_fmt) /ch;
avcodec_fill_audio_frame(frame, ch, resampler->resampler_params.src_sample_fmt, in_data, in_bytes, 0);
frame->pts = pts;
}
int ret = audio_resampler_send_frame(resampler, frame);
if(frame) {
av_frame_free(&frame); // 释放内存
}
return ret;
}
int audio_resampler_receive_frame2(audio_resampler_t *resampler, uint8_t **out_data, int nb_samples, int64_t *pts)
{
nb_samples = nb_samples == 0 ? av_audio_fifo_size(resampler->audio_fifo) : nb_samples;
if (av_audio_fifo_size(resampler->audio_fifo) < nb_samples || nb_samples == 0)
return 0;
int ret = av_audio_fifo_read(resampler->audio_fifo, (void **)out_data, nb_samples);
*pts = resampler->cur_pts;
resampler->cur_pts += nb_samples;
resampler->total_resampled_num += nb_samples;
return ret;
}
int audio_resampler_get_fifo_size(audio_resampler_t *resampler)
{
if(!resampler) {
return 0;
}
return av_audio_fifo_size(resampler->audio_fifo); // 获取fifo的采样点数量
}
int64_t audio_resampler_get_start_pts(audio_resampler_t *resampler)
{
if(!resampler) {
return 0;
}
return resampler->start_pts;
}
int64_t audio_resampler_get_cur_pts(audio_resampler_t *resampler)
{
if(!resampler) {
return 0;
}
return resampler->cur_pts;
}main.c
/**
* @example resampling_audio.c
* libswresample API use example.
*/
#include <libavutil/opt.h>
#include <libavutil/channel_layout.h>
#include <libavutil/samplefmt.h>
#include <libswresample/swresample.h>
#include "audioresampler.h"
/**
* 该代码主要实现音频重采样：
* (1) 进行重采样;
* (2) 将重采样后的samples放入audio fifo;
* (3) 再从fifo读取指定的samples.
*/
static int get_format_from_sample_fmt(const char **fmt,
enum AVSampleFormat sample_fmt)
{
int i;
struct sample_fmt_entry {
enum AVSampleFormat sample_fmt; const char *fmt_be, *fmt_le;
} sample_fmt_entries[] = {
{ AV_SAMPLE_FMT_U8, "u8", "u8" },
{ AV_SAMPLE_FMT_S16, "s16be", "s16le" },
{ AV_SAMPLE_FMT_S32, "s32be", "s32le" },
{ AV_SAMPLE_FMT_FLT, "f32be", "f32le" },
{ AV_SAMPLE_FMT_DBL, "f64be", "f64le" },
};
*fmt = NULL;
for (i = 0; i < FF_ARRAY_ELEMS(sample_fmt_entries); i++) {
struct sample_fmt_entry *entry = &sample_fmt_entries[i];
if (sample_fmt == entry->sample_fmt) {
*fmt = AV_NE(entry->fmt_be, entry->fmt_le);
return 0;
}
}
fprintf(stderr,
"Sample format %s not supported as output format\n",
av_get_sample_fmt_name(sample_fmt));
return AVERROR(EINVAL);
}
/**
* Fill dst buffer with nb_samples, generated starting from t.
*/
static void fill_samples(double *dst, int nb_samples, int nb_channels, int sample_rate, double *t)
{
int i, j;
double tincr = 1.0 / sample_rate, *dstp = dst;
const double c = 2 * M_PI * 440.0;
/* generate sin tone with 440Hz frequency and duplicated channels */
for (i = 0; i < nb_samples; i++) {
*dstp = sin(c * *t);
for (j = 1; j < nb_channels; j++)
dstp[j] = dstp[0];
dstp += nb_channels;
*t += tincr;
}
}
int main(int argc, char **argv)
{
// 输入参数
int64_t src_ch_layout = AV_CH_LAYOUT_STEREO;
int src_rate = 48000;
enum AVSampleFormat src_sample_fmt = AV_SAMPLE_FMT_DBL;
int src_nb_channels = 0;
uint8_t **src_data = NULL; // 二级指针
int src_linesize;
int src_nb_samples = 1024;
// 输出参数
int64_t dst_ch_layout = AV_CH_LAYOUT_STEREO;
int dst_rate = 44100;
enum AVSampleFormat dst_sample_fmt = AV_SAMPLE_FMT_S16;
int dst_nb_channels = 0;
uint8_t **dst_data = NULL; //二级指针
int dst_linesize;
int dst_nb_samples = 1152; // 固定为1152
// 输出文件
const char *dst_filename = NULL;
FILE *dst_file;
int dst_bufsize;
const char *fmt;
// 重采样实例
audio_resampler_params_t resampler_params;
resampler_params.src_channel_layout = src_ch_layout;
resampler_params.src_sample_fmt = src_sample_fmt;
resampler_params.src_sample_rate = src_rate;
resampler_params.dst_channel_layout = dst_ch_layout;
resampler_params.dst_sample_fmt = dst_sample_fmt;
resampler_params.dst_sample_rate = dst_rate;
audio_resampler_t *resampler = audio_resampler_alloc(resampler_params);
if(!resampler) {
printf("audio_resampler_alloc failed\n");
goto end;
}
double t;
int ret;
if (argc != 2) {
fprintf(stderr, "Usage: %s output_file\n"
"API example program to show how to resample an audio stream with libswresample.\n"
"This program generates a series of audio frames, resamples them to a specified "
"output format and rate and saves them to an output file named output_file.\n",
argv[0]);
exit(1);
}
dst_filename = argv[1];
dst_file = fopen(dst_filename, "wb");
if (!dst_file) {
fprintf(stderr, "Could not open destination file %s\n", dst_filename);
exit(1);
}
/* allocate source and destination samples buffers */
// 计算出输入源的通道数量
src_nb_channels = av_get_channel_layout_nb_channels(src_ch_layout);
// 给输入源分配内存空间
ret = av_samples_alloc_array_and_samples(&src_data, &src_linesize, src_nb_channels,
src_nb_samples, src_sample_fmt, 0);
if (ret < 0) {
fprintf(stderr, "Could not allocate source samples\n");
goto end;
}
/* buffer is going to be directly written to a rawaudio file, no alignment */
dst_nb_channels = av_get_channel_layout_nb_channels(dst_ch_layout);
dst_nb_samples = 1152; // 指定每次获取1152的数据
// 分配输出缓存内存
ret = av_samples_alloc_array_and_samples(&dst_data, &dst_linesize, dst_nb_channels,
dst_nb_samples, dst_sample_fmt, 0);
if (ret < 0) {
fprintf(stderr, "Could not allocate destination samples\n");
goto end;
}
t = 0;
int64_t in_pts = 0;
int64_t out_pts = 0;
do {
/* generate synthetic audio */
// 生成输入源
fill_samples((double *)src_data[0], src_nb_samples, src_nb_channels, src_rate, &t);
int ret_size = audio_resampler_send_frame2(resampler, src_data, src_nb_samples, in_pts);
in_pts += src_nb_samples;
ret_size = audio_resampler_receive_frame2(resampler, dst_data, dst_nb_samples, &out_pts);
if(ret_size > 0) {
dst_bufsize = av_samples_get_buffer_size(&dst_linesize, dst_nb_channels,
ret_size, dst_sample_fmt, 1);
if (dst_bufsize < 0) {
fprintf(stderr, "Could not get sample buffer size\n");
goto end;
}
printf("t:%f in:%d out:%d, out_pts:%"PRId64"\n", t, src_nb_samples, ret_size, out_pts);
fwrite(dst_data[0], 1, dst_bufsize, dst_file);
} else {
printf("can't get %d samples, ret_size:%d, cur_size:%d\n", dst_nb_samples, ret_size,
audio_resampler_get_fifo_size(resampler));
}
} while (t < 10); // 调整t观察内存使用情况
// flush 只有对数据完整度要求非常高的场景才需要做flush
audio_resampler_send_frame2(resampler, NULL, 0, 0);
int fifo_size = audio_resampler_get_fifo_size(resampler);
int get_size = (fifo_size >dst_nb_samples ? dst_nb_samples : fifo_size);
int ret_size = audio_resampler_receive_frame2(resampler, dst_data, get_size, &out_pts);
if(ret_size > 0) {
printf("flush ret_size:%d\n", ret_size);
// 不够一帧的时候填充为静音, 这里的目的是补偿最后一帧如果不够采样点数不足1152，用静音数据进行补足
av_samples_set_silence(dst_data, ret_size, dst_nb_samples - ret_size, dst_nb_channels, dst_sample_fmt);
dst_bufsize = av_samples_get_buffer_size(&dst_linesize, dst_nb_channels,
dst_nb_samples, dst_sample_fmt, 1);
if (dst_bufsize < 0) {
fprintf(stderr, "Could not get sample buffer size\n");
goto end;
}
printf("flush in:%d out:%d, out_pts:%"PRId64"\n", 0, ret_size, out_pts);
fwrite(dst_data[0], 1, dst_bufsize, dst_file);
}
if ((ret = get_format_from_sample_fmt(&fmt, dst_sample_fmt)) < 0)
goto end;
fprintf(stderr, "Resampling succeeded. Play the output file with the command:\n"
"ffplay -f %s -channel_layout %"PRId64" -channels %d -ar %d %s\n",
fmt, dst_ch_layout, dst_nb_channels, dst_rate, dst_filename);
end:
fclose(dst_file);
if (src_data)
av_freep(&src_data[0]);
av_freep(&src_data);
if (dst_data)
av_freep(&dst_data[0]);
av_freep(&dst_data);
if(resampler) {
audio_resampler_free(resampler);
}
printf("finish\n");
return ret < 0;
}本文章为转载内容，我们尊重原作者对文章享有的著作权。如有内容错误或侵权问题，欢迎原作者联系我们进行内容更正或删除文章。
