《音视频文章汇总》
接触ffmpeg有一段时间了，现遇到一个视频解码尾部掉帧的问题，记录下，缓冲区已经刷过，但还是尾部少帧,虽然尾部少一帧两帧对视频整体效果没啥影响(一般视频疫苗播放25帧)，但学东西就学透吧
FFmpeg从MP4文件中抽取yuv纯视频文件有两种方式
1.命令行工具ffmpeg直接抽取yuv文件
ffmpeg -i in.mp4 out_commandline.yuv
in.mp4文件是一个768x432,yuv420p,23fps的视频文件，抽取出来的yuv文件的总大小为115,955,712字节,可以计算出总帧数115955712/(7684321.5) = 233帧数
image
image
通过代码解码mp4文件后，看生成的yuv文件是否大小和上述一直
2.通过代码解封装demux，分别获取音频pcm和视频yuv文件,有两个版本的代码
- 3.1以前的音视频编解码分别用
解码API
avcodec_decode_video2()
avcodec_decode_audio4():
编码API
avcodec_encode_video2()
avcodec_encode_audio2() - 3.1之后的音视频编码
解码API
avcodec_send_packet()
avcodec_receive_frame()
编码API
avcodec_send_frame()
avcodec_receive_packet()
I.3.1版本之前的解封装demux代码官方示例程序为如下，解码自己的MP4文件只需更改文件路径
将官方代码抽取为C++一个类的类方法，传入本地in.mp4路径进行解码
image
传入本地路径
图片.png
image
FFmpegs::FFmpegs()
{
}
static int decode_packet(int *got_frame, int cached)
{
int ret = 0;
if (pkt.stream_index == video_stream_idx) {
/* decode video frame */
ret = avcodec_decode_video2(video_dec_ctx, frame, got_frame, &pkt);
if (ret < 0) {
fprintf(stderr, "Error decoding video frame\n");
return ret;
}
if (*got_frame) {
printf("video_frame%s n:%d coded_n:%d pts:%s\n",
cached ? "(cached)" : "",
video_frame_count++, frame->coded_picture_number,
av_ts2timestr(frame->pts, &video_dec_ctx->time_base));
/* copy decoded frame to destination buffer:
* this is required since rawvideo expects non aligned data */
av_image_copy(video_dst_data, video_dst_linesize,
(const uint8_t **)(frame->data), frame->linesize,
video_dec_ctx->pix_fmt, video_dec_ctx->width, video_dec_ctx->height);
/* write to rawvideo file */
fwrite(video_dst_data[0], 1, video_dst_bufsize, video_dst_file);
}
} else if (pkt.stream_index == audio_stream_idx) {
/* decode audio frame */
ret = avcodec_decode_audio4(audio_dec_ctx, frame, got_frame, &pkt);
if (ret < 0) {
fprintf(stderr, "Error decoding audio frame\n");
return ret;
}
if (*got_frame) {
printf("audio_frame%s n:%d nb_samples:%d pts:%s\n",
cached ? "(cached)" : "",
audio_frame_count++, frame->nb_samples,
av_ts2timestr(frame->pts, &audio_dec_ctx->time_base));
ret = av_samples_alloc(audio_dst_data, &audio_dst_linesize, av_frame_get_channels(frame),
frame->nb_samples, (AVSampleFormat)frame->format, 1);
if (ret < 0) {
fprintf(stderr, "Could not allocate audio buffer\n");
return AVERROR(ENOMEM);
}
/* TODO: extend return code of the av_samples_* functions so that this call is not needed */
audio_dst_bufsize =
av_samples_get_buffer_size(NULL, av_frame_get_channels(frame),
frame->nb_samples, (AVSampleFormat)frame->format, 1);
/* copy audio data to destination buffer:
* this is required since rawaudio expects non aligned data */
av_samples_copy(audio_dst_data, frame->data, 0, 0,
frame->nb_samples, av_frame_get_channels(frame), (AVSampleFormat)frame->format);
/* write to rawaudio file */
fwrite(audio_dst_data[0], 1, audio_dst_bufsize, audio_dst_file);
av_freep(&audio_dst_data[0]);
}
}
return ret;
}
static int open_codec_context(int *stream_idx,
AVFormatContext *fmt_ctx, enum AVMediaType type)
{
int ret;
AVStream *st;
AVCodecContext *dec_ctx = NULL;
AVCodec *dec = NULL;
ret = av_find_best_stream(fmt_ctx, type, -1, -1, NULL, 0);
if (ret < 0) {
fprintf(stderr, "Could not find %s stream in input file '%s'\n",
av_get_media_type_string(type), src_filename);
return ret;
} else {
*stream_idx = ret;
st = fmt_ctx->streams[*stream_idx];
/* find decoder for the stream */
dec_ctx = st->codec;
dec = avcodec_find_decoder(dec_ctx->codec_id);
if (!dec) {
fprintf(stderr, "Failed to find %s codec\n",
av_get_media_type_string(type));
return ret;
}
if ((ret = avcodec_open2(dec_ctx, dec, NULL)) < 0) {
fprintf(stderr, "Failed to open %s codec\n",
av_get_media_type_string(type));
return ret;
}
}
return 0;
}
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 is not supported as output format\n",
av_get_sample_fmt_name(sample_fmt));
return -1;
}
void FFmpegs::demuxer(){
int ret = 0, got_frame;
int inEnd = 0;
src_filename = "/Users/cloud/Documents/iOS/音视频/TestMusic/Demux/video2/in.mp4";
video_dst_filename = "/Users/cloud/Documents/iOS/音视频/TestMusic/Demux/video2/out_video2.yuv";
audio_dst_filename = "/Users/cloud/Documents/iOS/音视频/TestMusic/Demux/video2/out_video2.pcm";
/* register all formats and codecs */
av_register_all();
/* open input file, and allocate format context */
if (avformat_open_input(&fmt_ctx, src_filename, NULL, NULL) < 0) {
fprintf(stderr, "Could not open source file %s\n", src_filename);
exit(1);
}
/* retrieve stream information */
if (avformat_find_stream_info(fmt_ctx, NULL) < 0) {
fprintf(stderr, "Could not find stream information\n");
exit(1);
}
if (open_codec_context(&video_stream_idx, fmt_ctx, AVMEDIA_TYPE_VIDEO) >= 0) {
video_stream = fmt_ctx->streams[video_stream_idx];
video_dec_ctx = video_stream->codec;
video_dst_file = fopen(video_dst_filename, "wb");
if (!video_dst_file) {
fprintf(stderr, "Could not open destination file %s\n", video_dst_filename);
ret = 1;
goto end;
}
/* allocate image where the decoded image will be put */
ret = av_image_alloc(video_dst_data, video_dst_linesize,
video_dec_ctx->width, video_dec_ctx->height,
video_dec_ctx->pix_fmt, 1);
if (ret < 0) {
fprintf(stderr, "Could not allocate raw video buffer\n");
goto end;
}
video_dst_bufsize = ret;
}
if (open_codec_context(&audio_stream_idx, fmt_ctx, AVMEDIA_TYPE_AUDIO) >= 0) {
int nb_planes;
audio_stream = fmt_ctx->streams[audio_stream_idx];
audio_dec_ctx = audio_stream->codec;
audio_dst_file = fopen(audio_dst_filename, "wb");
if (!audio_dst_file) {
fprintf(stderr, "Could not open destination file %s\n", video_dst_filename);
ret = 1;
goto end;
}
nb_planes = av_sample_fmt_is_planar(audio_dec_ctx->sample_fmt) ?
audio_dec_ctx->channels : 1;
audio_dst_data = (uint8_t **)av_mallocz(sizeof(uint8_t *) * nb_planes);
if (!audio_dst_data) {
fprintf(stderr, "Could not allocate audio data buffers\n");
ret = AVERROR(ENOMEM);
goto end;
}
}
/* dump input information to stderr */
av_dump_format(fmt_ctx, 0, src_filename, 0);
if (!audio_stream && !video_stream) {
fprintf(stderr, "Could not find audio or video stream in the input, aborting\n");
ret = 1;
goto end;
}
frame = av_frame_alloc();
if (!frame) {
fprintf(stderr, "Could not allocate frame\n");
ret = AVERROR(ENOMEM);
goto end;
}
/* initialize packet, set data to NULL, let the demuxer fill it */
av_init_packet(&pkt);
pkt.data = NULL;
pkt.size = 0;
if (video_stream)
printf("Demuxing video from file '%s' into '%s'\n", src_filename, video_dst_filename);
if (audio_stream)
printf("Demuxing audio from file '%s' into '%s'\n", src_filename, audio_dst_filename);
/* read frames from the file */
while (av_read_frame(fmt_ctx, &pkt) >= 0) {
decode_packet(&got_frame, 0);
av_free_packet(&pkt);
}
/* flush cached frames */
pkt.data = NULL;
pkt.size = 0;
do {
decode_packet(&got_frame, 1);
} while (got_frame);
printf("Demuxing succeeded.\n");
if (video_stream) {
printf("Play the output video file with the command:\n"
"ffplay -f rawvideo -pix_fmt %s -video_size %dx%d %s\n",
av_get_pix_fmt_name(video_dec_ctx->pix_fmt), video_dec_ctx->width, video_dec_ctx->height,
video_dst_filename);
}
if (audio_stream) {
const char *fmt;
if ((ret = get_format_from_sample_fmt(&fmt, audio_dec_ctx->sample_fmt)) < 0)
goto end;
printf("Play the output audio file with the command:\n"
"ffplay -f %s -ac %d -ar %d %s\n",
fmt, audio_dec_ctx->channels, audio_dec_ctx->sample_rate,
audio_dst_filename);
}
end:
if (video_dec_ctx)
avcodec_close(video_dec_ctx);
if (audio_dec_ctx)
avcodec_close(audio_dec_ctx);
avformat_close_input(&fmt_ctx);
if (video_dst_file)
fclose(video_dst_file);
if (audio_dst_file)
fclose(audio_dst_file);
av_free(frame);
av_free(video_dst_data[0]);
av_free(audio_dst_data);
}
此时解码出来的yuv文件大小为115,458,048字节，115458048/(7684321.5)计算出帧数为232帧
image
通过使用YuvEye工具对比命令行生成的YUV文件和代码生成的YUV文件发现尾部少了一帧，YuvEye比对工具从第0帧开始，官方示例代码里面也已经刷新缓冲区了，为何还是少最后一帧呢
image
google搜索答案，看到这样一篇博文ffmpeg视频解码丢帧问题里面说的丢帧的原因是
读取的AVPacket都有pts和dts两个属性，往复杂了说视频帧类型有I/P/B等种类，我们就以简单的方式说，PTS是图像的展示时间，DTS是图像的解码时间，问题就来了，由于视频帧类型，很多时候，PTS基本都是按顺序的，但是DTS却不是，也就是说这个包的解码时候需要在下个包解码之后解码，所以此次就不能解码，因此获取不到视频帧，数据被缓存了，如果之后不去主动去取，那就真的丢了。
while( av_read_frame(format_ctx_, &packet) >= 0 ) {
if( packet.stream_index == video_stream_index_ ) {
avcodec_decode_video2(codec_ctx_, frame, &frameFinished, &packet);
if( frameFinished ) {
//...
}
}
av_free_packet(&packet);
}
以上的代码基本就是常见的使用场景，基本上文件的最后几帧已经丢失了，虽然很多时候不是太重要，可以忽略，毕竟几帧数据在很大帧率下所占的时间很小。如果需要的话，那如何找回来呢，简单的办法就是继续空包读取即可
给出的解决方案：记录丢掉了多少帧次数，在正常帧读取完毕后，进行空帧读取，读取空帧的次数就是记录的丢掉的帧的次数，这样总数加起来能保持不变，代码模板如下
注意，空包解码的时候必须是packet的data为NULL,size为0。
while( av_read_frame(format_ctx_, &packet) >= 0 ) {
if( packet.stream_index == video_stream_index_ ) {
avcodec_decode_video2(codec_ctx_, frame.get(), &frameFinished, &packet);
if( frameFinished ) {
//...
} else {
total_skip++;
}
}
av_free_packet(&packet);
}
while(total_skip--) {
avcodec_decode_video2(codec_ctx_, frame.get(), &frameFinished, &packet);
if( frameFinished ) {
//...
}
}
在代码中实现丢帧的空帧读取，这个操作就包含了刷新缓冲区的概念，记录跳跃帧的次数，最后再次读取跳跃帧次数次帧
image
于是喜出望外，请身边人喝了三瓶可口可乐，以为发现了规律总帧数=正常读取帧+缓冲区帧数+跳跃次数=230+1+2=233，正好为总帧数=YUV420P,768x432文件总大小115955712字节/(7684321.5)=233帧
实质后面发现缓冲区中存在2帧，与前面缓冲区中存在一帧不一致
image
image
再次用两个mp4文件调试,一个mp4文件是568x320,YUV420P,发现如下：命令行生成的yuv文件总大小为66524160字节/(5683201.5) = 244帧，好像也满足上面的规律
图片.png
image
再次换一个30秒的mp4文件768x432,YUV420P,命令行生成的文件总大小为344881152字节/(768x432*1.5) = 693帧，此文件就不符合总帧数693=正常读取帧690+缓冲区帧数2+跳跃次数2的规律了，690+2+2=694帧多出1帧
image
image
再次换一个60秒的mp4文件768x432,YUV420P,命令行生成的宽高768x432,YUV420P文件的总大小为689264640字节/(768x432*1.5)= 1385帧，此文件就不符合规律了1380+2+6=1388就多出两帧了
image
image
临时解决办法:多帧要比少帧好，就是最后文件中的数据因为是正常帧+缓冲区的帧+跳跃次数帧=得到的总帧数，可能回比实际帧数多，但总比缺帧好，先暂时找到的不是办法的办法，后面理清楚H264的编码原理后再找解决办法
跳跃帧次数+正常帧+缓冲区帧数>=总帧数
230+2+2=234
116453376/(7684321.5)=234帧
image
image
代码逻辑如下
image
image
#include "ffmpegs.h"
#include
static AVFormatContext *fmt_ctx = NULL;
static AVCodecContext *video_dec_ctx = NULL, *audio_dec_ctx;
static AVStream *video_stream = NULL, *audio_stream = NULL;
static const char *src_filename = NULL;
static const char *video_dst_filename = NULL;
static const char *audio_dst_filename = NULL;
static FILE *video_dst_file = NULL;
static FILE *audio_dst_file = NULL;
static uint8_t *video_dst_data[4] = {NULL};
static int video_dst_linesize[4];
static int video_dst_bufsize;
static uint8_t **audio_dst_data = NULL;
static int audio_dst_linesize;
static int audio_dst_bufsize;
static int video_stream_idx = -1, audio_stream_idx = -1;
static AVFrame *frame = NULL;
static AVPacket pkt;
static int video_frame_count = 0;
static int audio_frame_count = 0;
static int videoFrameIdx = 0;
static int audioFrameIdx = 0;
//跳跃了多少次空帧呢
static int skipFrameIdx = 0;
//跳跃帧写入了多少帧
static int skipFrameWriteIdx = 0;
//跳跃帧里面的尾部帧，防止开头的跳跃帧被写入
static int isEndFrame = 0;
//正常帧读取后刷新缓冲区刷新了多少帧
static int flushFrameIdx = 0;
//是尾部跳跃帧补齐
static int isSkipFrame = 0;
FFmpegs::FFmpegs()
{
}
static int decode_packet(int *got_frame, int cached)
{
int ret = 0;
if (pkt.stream_index == video_stream_idx) {
/* decode video frame */
ret = avcodec_decode_video2(video_dec_ctx, frame, got_frame, &pkt);
if(isEndFrame == 1) qDebug() << "尾部帧刷新后解码器返回值:" << ret << *got_frame << &pkt;
if (ret < 0) {
fprintf(stderr, "Error decoding video frame\n");
return ret;
}
if (*got_frame) {
qDebug() << "正常视频帧:" << ++video_frame_count << "帧宽度:" << frame->width;
if(isEndFrame == 1){//如果是尾部刷新缓冲区记录下刷新缓冲区刷新了多少帧
flushFrameIdx++;
qDebug() << "缓冲区刷新了多少帧" << flushFrameIdx;
}
printf("视频帧-------video_frame%s n:%d coded_n:%d pts:%s\n",
cached ? "(cached)" : "",
video_frame_count, frame->coded_picture_number,
av_ts2timestr(frame->pts, &video_dec_ctx->time_base));
/* copy decoded frame to destination buffer:
* this is required since rawvideo expects non aligned data */
av_image_copy(video_dst_data, video_dst_linesize,
(const uint8_t **)(frame->data), frame->linesize,
video_dec_ctx->pix_fmt, video_dec_ctx->width, video_dec_ctx->height);
/* write to rawvideo file */
fwrite(video_dst_data[0], 1, video_dst_bufsize, video_dst_file);
}else{
if(isEndFrame==0){//不是尾部刷新缓冲区进入才计算跳跃次数
skipFrameIdx++;
qDebug() << "当前pkt的值" << &pkt << "帧宽度:" << frame->width << "--跳跃次数--" << skipFrameIdx;
}
else{//是尾部跳跃帧数进来，将尾部帧写入文件尾部
if(isSkipFrame == 1){
qDebug() << "got_frame写入帧为0，写入跳跃帧次数" << ++skipFrameWriteIdx;
av_image_copy(video_dst_data, video_dst_linesize,
(const uint8_t **)(frame->data), frame->linesize,
video_dec_ctx->pix_fmt, video_dec_ctx->width, video_dec_ctx->height);
fwrite(video_dst_data[0], 1, video_dst_bufsize, video_dst_file);
}
}
}
}
else if (pkt.stream_index == audio_stream_idx) {
/* decode audio frame */
ret = avcodec_decode_audio4(audio_dec_ctx, frame, got_frame, &pkt);
if (ret < 0) {
fprintf(stderr, "Error decoding audio frame\n");
return ret;
}
if (*got_frame) {
printf("音频帧--------audio_frame%s n:%d nb_samples:%d pts:%s\n",
cached ? "(cached)" : "",
audio_frame_count++, frame->nb_samples,
av_ts2timestr(frame->pts, &audio_dec_ctx->time_base));
ret = av_samples_alloc(audio_dst_data, &audio_dst_linesize, av_frame_get_channels(frame),
frame->nb_samples, (AVSampleFormat)frame->format, 1);
if (ret < 0) {
fprintf(stderr, "Could not allocate audio buffer\n");
return AVERROR(ENOMEM);
}
/* TODO: extend return code of the av_samples_* functions so that this call is not needed */
audio_dst_bufsize =
av_samples_get_buffer_size(NULL, av_frame_get_channels(frame),
frame->nb_samples, (AVSampleFormat)frame->format, 1);
/* copy audio data to destination buffer:
* this is required since rawaudio expects non aligned data */
av_samples_copy(audio_dst_data, frame->data, 0, 0,
frame->nb_samples, av_frame_get_channels(frame), (AVSampleFormat)frame->format);
/* write to rawaudio file */
printf("单次写入音频帧的大小 %d\n写了多少次 %d",audio_dst_bufsize,++audioFrameIdx);
fwrite(audio_dst_data[0], 1, audio_dst_bufsize, audio_dst_file);
av_freep(&audio_dst_data[0]);
}
}
return ret;
}
static int open_codec_context(int *stream_idx,
AVFormatContext *fmt_ctx, enum AVMediaType type)
{
int ret;
AVStream *st;
AVCodecContext *dec_ctx = NULL;
AVCodec *dec = NULL;
ret = av_find_best_stream(fmt_ctx, type, -1, -1, NULL, 0);
if (ret < 0) {
fprintf(stderr, "Could not find %s stream in input file '%s'\n",
av_get_media_type_string(type), src_filename);
return ret;
} else {
*stream_idx = ret;
st = fmt_ctx->streams[*stream_idx];
/* find decoder for the stream */
dec_ctx = st->codec;
dec = avcodec_find_decoder(dec_ctx->codec_id);
if (!dec) {
fprintf(stderr, "Failed to find %s codec\n",
av_get_media_type_string(type));
return ret;
}
if ((ret = avcodec_open2(dec_ctx, dec, NULL)) < 0) {
fprintf(stderr, "Failed to open %s codec\n",
av_get_media_type_string(type));
return ret;
}
}
return 0;
}
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 is not supported as output format\n",
av_get_sample_fmt_name(sample_fmt));
return -1;
}
void FFmpegs::demuxer(){
int ret = 0, got_frame;
int inEnd = 0;
src_filename = "/Users/cloud/Documents/iOS/音视频/TestMusic/Demux/in.mp4";
video_dst_filename = "/Users/cloud/Documents/iOS/音视频/TestMusic/Demux/out_video2_optimize.yuv";
audio_dst_filename = "/Users/cloud/Documents/iOS/音视频/TestMusic/Demux/out_video2_optimize.pcm";
/* register all formats and codecs */
av_register_all();
/* open input file, and allocate format context */
if (avformat_open_input(&fmt_ctx, src_filename, NULL, NULL) < 0) {
fprintf(stderr, "Could not open source file %s\n", src_filename);
exit(1);
}
/* retrieve stream information */
if (avformat_find_stream_info(fmt_ctx, NULL) < 0) {
fprintf(stderr, "Could not find stream information\n");
exit(1);
}
if (open_codec_context(&video_stream_idx, fmt_ctx, AVMEDIA_TYPE_VIDEO) >= 0) {
video_stream = fmt_ctx->streams[video_stream_idx];
video_dec_ctx = video_stream->codec;
video_dst_file = fopen(video_dst_filename, "wb");
if (!video_dst_file) {
fprintf(stderr, "Could not open destination file %s\n", video_dst_filename);
ret = 1;
goto end;
}
/* allocate image where the decoded image will be put */
ret = av_image_alloc(video_dst_data, video_dst_linesize,
video_dec_ctx->width, video_dec_ctx->height,
video_dec_ctx->pix_fmt, 1);
if (ret < 0) {
fprintf(stderr, "Could not allocate raw video buffer\n");
goto end;
}
video_dst_bufsize = ret;
}
if (open_codec_context(&audio_stream_idx, fmt_ctx, AVMEDIA_TYPE_AUDIO) >= 0) {
int nb_planes;
audio_stream = fmt_ctx->streams[audio_stream_idx];
audio_dec_ctx = audio_stream->codec;
audio_dst_file = fopen(audio_dst_filename, "wb");
if (!audio_dst_file) {
fprintf(stderr, "Could not open destination file %s\n", video_dst_filename);
ret = 1;
goto end;
}
nb_planes = av_sample_fmt_is_planar(audio_dec_ctx->sample_fmt) ?
audio_dec_ctx->channels : 1;
audio_dst_data = (uint8_t **)av_mallocz(sizeof(uint8_t *) * nb_planes);
if (!audio_dst_data) {
fprintf(stderr, "Could not allocate audio data buffers\n");
ret = AVERROR(ENOMEM);
goto end;
}
}
/* dump input information to stderr */
av_dump_format(fmt_ctx, 0, src_filename, 0);
if (!audio_stream && !video_stream) {
fprintf(stderr, "Could not find audio or video stream in the input, aborting\n");
ret = 1;
goto end;
}
frame = av_frame_alloc();
if (!frame) {
fprintf(stderr, "Could not allocate frame\n");
ret = AVERROR(ENOMEM);
goto end;
}
/* initialize packet, set data to NULL, let the demuxer fill it */
av_init_packet(&pkt);
pkt.data = NULL;
pkt.size = 0;
if (video_stream)
printf("Demuxing video from file '%s' into '%s'\n", src_filename, video_dst_filename);
if (audio_stream)
printf("Demuxing audio from file '%s' into '%s'\n", src_filename, audio_dst_filename);
/* read frames from the file */
//跳跃帧次数+正常帧+缓冲区帧数>=总帧数
//正常帧
while (av_read_frame(fmt_ctx, &pkt) >= 0) {
decode_packet(&got_frame, 0);
av_free_packet(&pkt);
}
/* flush cached frames */
pkt.data = NULL;
pkt.size = 0;
//刷新缓冲区帧
do {
qDebug() << "进入刷新缓冲区空帧读取" << got_frame;
isEndFrame = 1;
decode_packet(&got_frame, 1);
} while (got_frame);
//跳跃帧
while(skipFrameIdx--){
isSkipFrame = 1;
qDebug() << "进入跳跃帧读取" << skipFrameIdx;
decode_packet(&got_frame, 1);
}
printf("Demuxing succeeded.\n");
if (video_stream) {
printf("Play the output video file with the command:\n"
"ffplay -f rawvideo -pix_fmt %s -video_size %dx%d %s\n",
av_get_pix_fmt_name(video_dec_ctx->pix_fmt), video_dec_ctx->width, video_dec_ctx->height,
video_dst_filename);
}
if (audio_stream) {
const char *fmt;
if ((ret = get_format_from_sample_fmt(&fmt, audio_dec_ctx->sample_fmt)) < 0)
goto end;
printf("Play the output audio file with the command:\n"
"ffplay -f %s -ac %d -ar %d %s\n",
fmt, audio_dec_ctx->channels, audio_dec_ctx->sample_rate,
audio_dst_filename);
}
end:
if (video_dec_ctx)
avcodec_close(video_dec_ctx);
if (audio_dec_ctx)
avcodec_close(audio_dec_ctx);
avformat_close_input(&fmt_ctx);
if (video_dst_file)
fclose(video_dst_file);
if (audio_dst_file)
fclose(audio_dst_file);
av_free(frame);
av_free(video_dst_data[0]);
av_free(audio_dst_data);
}
II.新版本的解码代码如何优化呢
avcodec_send_packet和avcodec_receive_frame代码中没有got_frame标识是否获取到了完整一帧，不知道跳跃了多少次
H264编码的原理是：I帧P帧B帧混合组成，P帧的编解码要依赖前面I帧或P帧，B帧的编解码要依赖前后I帧和P帧，若最后截取的字符串是B帧呢，那么它的解封装要依赖后面的B帧或P帧，会导致最后的几帧无法解码出来
3.还有一个问题，ffprobe命令检测出来的帧数和命令行生成的实际帧数也不一致,命令行生成的yuv文件为233帧，ffprobe检测出来为232帧，ffprobe检测出来的总是少几帧
ffprobe -v error -count_frames -select_streams v:0 -show_entries stream=nb_read_frames -of default=nokey=1:noprint_wrappers=1 input.mp4
图片.png
未完待续。。。
GitHub上ffmpeg官方补丁地址patch
查阅文档《H.264编码》
I帧，P帧，B帧的本质
I帧（I Picture、I Frame、Intra Coded Picture），译为：帧内编码图像，也叫做关键帧（Keyframe）
是视频的第一帧，也是GOP的第一帧，一个GOP只有一个I帧
编码
对整帧图像数据进行编码
解码
仅用当前I帧的编码数据就可以解码出完整的图像
是一种自带全部信息的独立帧，无需参考其他图像便可独立进行解码，可以简单理解为一张静态图像
P帧（P Picture、P Frame、Predictive Coded Picture），译为：预测编码图像
编码
并不会对整帧图像数据进行编码
以前面的I帧或P帧作为参考帧，只编码当前P帧与参考帧的差异数据
解码
需要先解码出前面的参考帧，再结合差异数据解码出当前P帧完整的图像
B帧（B Picture、B Frame、Bipredictive Coded Picture），译为：前后预测编码图像
编码
并不会对整帧图像数据进行编码
同时以前面、后面的I帧或P帧作为参考帧，只编码当前B帧与前后参考帧的差异数据
因为可参考的帧变多了，所以只需要存储更少的差异数据
解码
需要先解码出前后的参考帧，再结合差异数据解码出当前B帧完整的图像