我正在尝试将 H.264 数据复用到 MP4 文件中。将此 H.264 Annex B 数据保存到 MP4 文件中似乎没有错误，但文件无法播放。
我对文件进行了二进制比较，问题似乎出在写入 MP4 文件的页脚（预告片）的内容中。
我怀疑它必须与创建流的方式有关。
在里面：
AVOutputFormat* fmt = av_guess_format( 0, "out.mp4", 0 );
oc = avformat_alloc_context();
oc->oformat = fmt;
strcpy(oc->filename, filename);
我拥有的这个原型应用程序的一部分是为每个 IFrame 创建一个 png 文件。因此，当遇到第一个 IFrame 时，我创建视频流并写入 av 标头等：
void addVideoStream(AVCodecContext* decoder)
{
videoStream = av_new_stream(oc, 0);
if (!videoStream)
{
cout << "ERROR creating video stream" << endl;
return;
}
vi = videoStream->index;
videoContext = videoStream->codec;
videoContext->codec_type = AVMEDIA_TYPE_VIDEO;
videoContext->codec_id = decoder->codec_id;
videoContext->bit_rate = 512000;
videoContext->width = decoder->width;
videoContext->height = decoder->height;
videoContext->time_base.den = 25;
videoContext->time_base.num = 1;
videoContext->gop_size = decoder->gop_size;
videoContext->pix_fmt = decoder->pix_fmt;
if (oc->oformat->flags & AVFMT_GLOBALHEADER)
videoContext->flags |= CODEC_FLAG_GLOBAL_HEADER;
av_dump_format(oc, 0, filename, 1);
if (!(oc->oformat->flags & AVFMT_NOFILE))
{
if (avio_open(&oc->pb, filename, AVIO_FLAG_WRITE) < 0) {
cout << "Error opening file" << endl;
}
avformat_write_header(oc, NULL);
}
我写出数据包：
unsigned char* data = block->getData();
unsigned char videoFrameType = data[4];
int dataLen = block->getDataLen();
// store pps
if (videoFrameType == 0x68)
{
if (ppsFrame != NULL)
{
delete ppsFrame; ppsFrameLength = 0; ppsFrame = NULL;
}
ppsFrameLength = block->getDataLen();
ppsFrame = new unsigned char[ppsFrameLength];
memcpy(ppsFrame, block->getData(), ppsFrameLength);
}
else if (videoFrameType == 0x67)
{
// sps
if (spsFrame != NULL)
{
delete spsFrame; spsFrameLength = 0; spsFrame = NULL;
}
spsFrameLength = block->getDataLen();
spsFrame = new unsigned char[spsFrameLength];
memcpy(spsFrame, block->getData(), spsFrameLength);
}
if (videoFrameType == 0x65 || videoFrameType == 0x41)
{
videoFrameNumber++;
}
if (videoFrameType == 0x65)
{
decodeIFrame(videoFrameNumber, spsFrame, spsFrameLength, ppsFrame, ppsFrameLength, data, dataLen);
}
if (videoStream != NULL)
{
AVPacket pkt = { 0 };
av_init_packet(&pkt);
pkt.stream_index = vi;
pkt.flags = 0;
pkt.pts = pkt.dts = 0;
if (videoFrameType == 0x65)
{
// combine the SPS PPS & I frames together
pkt.flags |= AV_PKT_FLAG_KEY;
unsigned char* videoFrame = new unsigned char[spsFrameLength+ppsFrameLength+dataLen];
memcpy(videoFrame, spsFrame, spsFrameLength);
memcpy(&videoFrame[spsFrameLength], ppsFrame, ppsFrameLength);
memcpy(&videoFrame[spsFrameLength+ppsFrameLength], data, dataLen);
// overwrite the start code (00 00 00 01 with a 32-bit length)
setLength(videoFrame, spsFrameLength-4);
setLength(&videoFrame[spsFrameLength], ppsFrameLength-4);
setLength(&videoFrame[spsFrameLength+ppsFrameLength], dataLen-4);
pkt.size = dataLen + spsFrameLength + ppsFrameLength;
pkt.data = videoFrame;
av_interleaved_write_frame(oc, &pkt);
delete videoFrame; videoFrame = NULL;
}
else if (videoFrameType != 0x67 && videoFrameType != 0x68)
{
// Send other frames except pps & sps which are caught and stored
pkt.size = dataLen;
pkt.data = data;
setLength(data, dataLen-4);
av_interleaved_write_frame(oc, &pkt);
}
最后关闭文件：
av_write_trailer(oc);
int i = 0;
for (i = 0; i < oc->nb_streams; i++)
{
av_freep(&oc->streams[i]->codec);
av_freep(&oc->streams[i]);
}
if (!(oc->oformat->flags & AVFMT_NOFILE))
{
avio_close(oc->pb);
}
av_free(oc);
如果我单独获取 H.264 数据并进行转换：
ffmpeg -i recording.h264 -vcodec copy recording.mp4
除了文件的“页脚”之外，所有文件都是相同的。
我的程序的输出：readrec recording.tcp out.mp4 **** START **** 01-03-2013 14:26:01 180000 输出 #0，mp4，到“out.mp4”：流 #0:0 : 视频：h264, yuv420p, 352x288, q=2-31, 512 kb/s, 90k tbn, 25 tbc **** END **** 01-03-2013 14:27:01 102000 写了 1499 个视频帧。
如果我尝试使用 ffmpeg 转换使用 CODE 创建的 MP4 文件：
ffmpeg -i out.mp4 -vcodec copy out2.mp4
ffmpeg version 0.11.1 Copyright (c) 2000-2012 the FFmpeg developers
built on Mar 7 2013 12:49:22 with suncc 0x5110
configuration: --extra-cflags=-KPIC -g --disable-mmx
--disable-protocol=udp --disable-encoder=nellymoser --cc=cc --cxx=CC
libavutil 51. 54.100 / 51. 54.100
libavcodec 54. 23.100 / 54. 23.100
libavformat 54. 6.100 / 54. 6.100
libavdevice 54. 0.100 / 54. 0.100
libavfilter 2. 77.100 / 2. 77.100
libswscale 2. 1.100 / 2. 1.100
libswresample 0. 15.100 / 0. 15.100
h264 @ 12eaac0] no frame!
Last message repeated 1 times
[h264 @ 12eaac0] slice type too large (0) at 0 0
[h264 @ 12eaac0] decode_slice_header error
[h264 @ 12eaac0] no frame!
Last message repeated 23 times
[h264 @ 12eaac0] slice type too large (0) at 0 0
[h264 @ 12eaac0] decode_slice_header error
[h264 @ 12eaac0] no frame!
Last message repeated 74 times
[h264 @ 12eaac0] slice type too large (0) at 0 0
[h264 @ 12eaac0] decode_slice_header error
[h264 @ 12eaac0] no frame!
Last message repeated 64 times
[h264 @ 12eaac0] slice type too large (0) at 0 0
[h264 @ 12eaac0] decode_slice_header error
[h264 @ 12eaac0] no frame!
Last message repeated 34 times
[h264 @ 12eaac0] slice type too large (0) at 0 0
[h264 @ 12eaac0] decode_slice_header error
[h264 @ 12eaac0] no frame!
Last message repeated 49 times
[h264 @ 12eaac0] slice type too large (0) at 0 0
[h264 @ 12eaac0] decode_slice_header error
[h264 @ 12eaac0] no frame!
Last message repeated 24 times
[h264 @ 12eaac0] Partitioned H.264 support is incomplete
[h264 @ 12eaac0] no frame!
Last message repeated 23 times
[h264 @ 12eaac0] sps_id out of range
[h264 @ 12eaac0] no frame!
Last message repeated 148 times
[h264 @ 12eaac0] sps_id (32) out of range
Last message repeated 1 times
[h264 @ 12eaac0] no frame!
Last message repeated 33 times
[h264 @ 12eaac0] slice type too large (0) at 0 0
[h264 @ 12eaac0] decode_slice_header error
[h264 @ 12eaac0] no frame!
Last message repeated 128 times
[h264 @ 12eaac0] sps_id (32) out of range
Last message repeated 1 times
[h264 @ 12eaac0] no frame!
Last message repeated 3 times
[h264 @ 12eaac0] slice type too large (0) at 0 0
[h264 @ 12eaac0] decode_slice_header error
[h264 @ 12eaac0] no frame!
Last message repeated 3 times
[h264 @ 12eaac0] slice type too large (0) at 0 0
[h264 @ 12eaac0] decode_slice_header error
[h264 @ 12eaac0] no frame!
Last message repeated 309 times
[h264 @ 12eaac0] sps_id (32) out of range
Last message repeated 1 times
[h264 @ 12eaac0] no frame!
Last message repeated 192 times
[h264 @ 12eaac0] Partitioned H.264 support is incomplete
[h264 @ 12eaac0] no frame!
Last message repeated 73 times
[h264 @ 12eaac0] sps_id (32) out of range
Last message repeated 1 times
[h264 @ 12eaac0] no frame!
Last message repeated 99 times
[h264 @ 12eaac0] sps_id (32) out of range
Last message repeated 1 times
[h264 @ 12eaac0] no frame!
Last message repeated 197 times
[mov,mp4,m4a,3gp,3g2,mj2 @ 12e3100] decoding for stream 0 failed
[mov,mp4,m4a,3gp,3g2,mj2 @ 12e3100] Could not find codec parameters
(Video: h264 (avc1 / 0x31637661), 393539 kb/s)
out.mp4: could not find codec parameters
我真的不知道问题出在哪里，除了它必须与流的设置方式有关。我查看了其他人正在做类似事情的一些代码，并尝试使用此建议来设置流，但无济于事！
给我一个 H.264/AAC 混合（同步）文件的最终代码如下。首先是一些背景信息。数据来自 IP 摄像机。数据通过第 3 方 API 以视频/音频数据包的形式呈现。视频包以 RTP 有效载荷数据（无标头）的形式呈现，由 NALU 组成，这些 NALU 被重构并转换为附件 B 格式的 H.264 视频。AAC 音频以原始 AAC 形式呈现，并转换为 adts 格式以启用播放。这些数据包已被放入比特流格式，允许传输时间戳（自 1970 年 1 月 1 日以来的 64 位毫秒）以及其他一些内容。
这或多或少是一个原型，在任何方面都不干净。它可能泄漏得很糟糕。但是，我确实希望这可以帮助其他人尝试实现与我相似的目标。
全局变量：
AVFormatContext* oc = NULL;
AVCodecContext* videoContext = NULL;
AVStream* videoStream = NULL;
AVCodecContext* audioContext = NULL;
AVStream* audioStream = NULL;
AVCodec* videoCodec = NULL;
AVCodec* audioCodec = NULL;
int vi = 0; // Video stream
int ai = 1; // Audio stream
uint64_t firstVideoTimeStamp = 0;
uint64_t firstAudioTimeStamp = 0;
int audioStartOffset = 0;
char* filename = NULL;
Boolean first = TRUE;
int videoFrameNumber = 0;
int audioFrameNumber = 0;
主要的：
int main(int argc, char* argv[])
{
if (argc != 3)
{
cout << argv[0] << " <stream playback file> <output mp4 file>" << endl;
return 0;
}
char* input_stream_file = argv[1];
filename = argv[2];
av_register_all();
fstream inFile;
inFile.open(input_stream_file, ios::in);
// Used to store the latest pps & sps frames
unsigned char* ppsFrame = NULL;
int ppsFrameLength = 0;
unsigned char* spsFrame = NULL;
int spsFrameLength = 0;
// Setup MP4 output file
AVOutputFormat* fmt = av_guess_format( 0, filename, 0 );
oc = avformat_alloc_context();
oc->oformat = fmt;
strcpy(oc->filename, filename);
// Setup the bitstream filter for AAC in adts format. Could probably also achieve
// this by stripping the first 7 bytes!
AVBitStreamFilterContext* bsfc = av_bitstream_filter_init("aac_adtstoasc");
if (!bsfc)
{
cout << "Error creating adtstoasc filter" << endl;
return -1;
}
while (inFile.good())
{
TcpAVDataBlock* block = new TcpAVDataBlock();
block->readStruct(inFile);
DateTime dt = block->getTimestampAsDateTime();
switch (block->getPacketType())
{
case TCP_PACKET_H264:
{
if (firstVideoTimeStamp == 0)
firstVideoTimeStamp = block->getTimeStamp();
unsigned char* data = block->getData();
unsigned char videoFrameType = data[4];
int dataLen = block->getDataLen();
// pps
if (videoFrameType == 0x68)
{
if (ppsFrame != NULL)
{
delete ppsFrame; ppsFrameLength = 0;
ppsFrame = NULL;
}
ppsFrameLength = block->getDataLen();
ppsFrame = new unsigned char[ppsFrameLength];
memcpy(ppsFrame, block->getData(), ppsFrameLength);
}
else if (videoFrameType == 0x67)
{
// sps
if (spsFrame != NULL)
{
delete spsFrame; spsFrameLength = 0;
spsFrame = NULL;
}
spsFrameLength = block->getDataLen();
spsFrame = new unsigned char[spsFrameLength];
memcpy(spsFrame, block->getData(), spsFrameLength);
}
if (videoFrameType == 0x65 || videoFrameType == 0x41)
{
videoFrameNumber++;
}
// Extract a thumbnail for each I-Frame
if (videoFrameType == 0x65)
{
decodeIFrame(h264, spsFrame, spsFrameLength, ppsFrame, ppsFrameLength, data, dataLen);
}
if (videoStream != NULL)
{
AVPacket pkt = { 0 };
av_init_packet(&pkt);
pkt.stream_index = vi;
pkt.flags = 0;
pkt.pts = videoFrameNumber;
pkt.dts = videoFrameNumber;
if (videoFrameType == 0x65)
{
pkt.flags = 1;
unsigned char* videoFrame = new unsigned char[spsFrameLength+ppsFrameLength+dataLen];
memcpy(videoFrame, spsFrame, spsFrameLength);
memcpy(&videoFrame[spsFrameLength], ppsFrame, ppsFrameLength);
memcpy(&videoFrame[spsFrameLength+ppsFrameLength], data, dataLen);
pkt.data = videoFrame;
av_interleaved_write_frame(oc, &pkt);
delete videoFrame; videoFrame = NULL;
}
else if (videoFrameType != 0x67 && videoFrameType != 0x68)
{
pkt.size = dataLen;
pkt.data = data;
av_interleaved_write_frame(oc, &pkt);
}
}
break;
}
case TCP_PACKET_AAC:
if (firstAudioTimeStamp == 0)
{
firstAudioTimeStamp = block->getTimeStamp();
uint64_t millseconds_difference = firstAudioTimeStamp - firstVideoTimeStamp;
audioStartOffset = millseconds_difference * 16000 / 1000;
cout << "audio offset: " << audioStartOffset << endl;
}
if (audioStream != NULL)
{
AVPacket pkt = { 0 };
av_init_packet(&pkt);
pkt.stream_index = ai;
pkt.flags = 1;
pkt.pts = audioFrameNumber*1024;
pkt.dts = audioFrameNumber*1024;
pkt.data = block->getData();
pkt.size = block->getDataLen();
pkt.duration = 1024;
AVPacket newpacket = pkt;
int rc = av_bitstream_filter_filter(bsfc, audioContext,
NULL,
&newpacket.data, &newpacket.size,
pkt.data, pkt.size,
pkt.flags & AV_PKT_FLAG_KEY);
if (rc >= 0)
{
//cout << "Write audio frame" << endl;
newpacket.pts = audioFrameNumber*1024;
newpacket.dts = audioFrameNumber*1024;
audioFrameNumber++;
newpacket.duration = 1024;
av_interleaved_write_frame(oc, &newpacket);
av_free_packet(&newpacket);
}
else
{
cout << "Error filtering aac packet" << endl;
}
}
break;
case TCP_PACKET_START:
break;
case TCP_PACKET_END:
break;
}
delete block;
}
inFile.close();
av_write_trailer(oc);
int i = 0;
for (i = 0; i < oc->nb_streams; i++)
{
av_freep(&oc->streams[i]->codec);
av_freep(&oc->streams[i]);
}
if (!(oc->oformat->flags & AVFMT_NOFILE))
{
avio_close(oc->pb);
}
av_free(oc);
delete spsFrame; spsFrame = NULL;
delete ppsFrame; ppsFrame = NULL;
cout << "Wrote " << videoFrameNumber << " video frames." << endl;
return 0;
}
添加流流/编解码器，并在名为 addVideoAndAudioStream() 的函数中创建标头。这个函数是从 decodeIFrame() 调用的，所以有一些假设（不一定是好的） 1. 视频包首先出现 2. AAC 存在
decodeIFrame 是一种单独的原型，我在其中为每个 I Frame 创建缩略图。生成缩略图的代码来自：https ://gnunet.org/svn/Extractor/src/plugins/thumbnailffmpeg_extractor.c
decodeIFrame 函数将一个 AVCodecContext 传递给 addVideoAudioStream：
void addVideoAndAudioStream(AVCodecContext* decoder = NULL)
{
videoStream = av_new_stream(oc, 0);
if (!videoStream)
{
cout << "ERROR creating video stream" << endl;
return;
}
vi = videoStream->index;
videoContext = videoStream->codec;
videoContext->codec_type = AVMEDIA_TYPE_VIDEO;
videoContext->codec_id = decoder->codec_id;
videoContext->bit_rate = 512000;
videoContext->width = decoder->width;
videoContext->height = decoder->height;
videoContext->time_base.den = 25;
videoContext->time_base.num = 1;
videoContext->gop_size = decoder->gop_size;
videoContext->pix_fmt = decoder->pix_fmt;
audioStream = av_new_stream(oc, 1);
if (!audioStream)
{
cout << "ERROR creating audio stream" << endl;
return;
}
ai = audioStream->index;
audioContext = audioStream->codec;
audioContext->codec_type = AVMEDIA_TYPE_AUDIO;
audioContext->codec_id = CODEC_ID_AAC;
audioContext->bit_rate = 64000;
audioContext->sample_rate = 16000;
audioContext->channels = 1;
if (oc->oformat->flags & AVFMT_GLOBALHEADER)
{
videoContext->flags |= CODEC_FLAG_GLOBAL_HEADER;
audioContext->flags |= CODEC_FLAG_GLOBAL_HEADER;
}
av_dump_format(oc, 0, filename, 1);
if (!(oc->oformat->flags & AVFMT_NOFILE))
{
if (avio_open(&oc->pb, filename, AVIO_FLAG_WRITE) < 0) {
cout << "Error opening file" << endl;
}
}
avformat_write_header(oc, NULL);
}
据我所知，一些假设似乎并不重要，例如： 1. 比特率。实际视频比特率约为 262k，而我指定 512kbit 2. AAC 频道。我指定了单声道，虽然实际输出是来自内存的立体声
您仍然需要知道视频和音频的帧速率（时基）是多少。
与许多其他示例相反，在视频数据包上设置 pts 和 dts 时，它无法播放。我需要知道时基（25fps），然后根据该时基设置 pts 和 dts，即第一帧 = 0（PPS，SPS，I），第二帧 = 1（中间帧，不管它叫什么；）） .
AAC 我还必须假设它是 16000 赫兹。每个 AAC 数据包 1024 个样本（我认为您也可以使用 AAC @ 960 个样本）来确定音频“偏移量”。我将此添加到 pts & dts。所以 pts/dts 是要回放的样本编号。您还需要确保在写入之前在数据包中设置了 1024 的持续时间。
--
我今天还发现附件 B 与任何其他播放器都不兼容，因此应该使用 AVCC 格式。
这些 URL 有帮助： Problem to Decode H264 video over RTP with ffmpeg (libavcodec) http://aviadr1.blogspot.com.au/2010/05/h264-extradata-partially-explained-for.html
在构建视频流的时候，我填写了extradata&extradata_size：
// Extradata contains PPS & SPS for AVCC format
int extradata_len = 8 + spsFrameLen-4 + 1 + 2 + ppsFrameLen-4;
videoContext->extradata = (uint8_t*)av_mallocz(extradata_len);
videoContext->extradata_size = extradata_len;
videoContext->extradata[0] = 0x01;
videoContext->extradata[1] = spsFrame[4+1];
videoContext->extradata[2] = spsFrame[4+2];
videoContext->extradata[3] = spsFrame[4+3];
videoContext->extradata[4] = 0xFC | 3;
videoContext->extradata[5] = 0xE0 | 1;
int tmp = spsFrameLen - 4;
videoContext->extradata[6] = (tmp >> 8) & 0x00ff;
videoContext->extradata[7] = tmp & 0x00ff;
int i = 0;
for (i=0;i<tmp;i++)
videoContext->extradata[8+i] = spsFrame[4+i];
videoContext->extradata[8+tmp] = 0x01;
int tmp2 = ppsFrameLen-4;
videoContext->extradata[8+tmp+1] = (tmp2 >> 8) & 0x00ff;
videoContext->extradata[8+tmp+2] = tmp2 & 0x00ff;
for (i=0;i<tmp2;i++)
videoContext->extradata[8+tmp+3+i] = ppsFrame[4+i];
写出帧时，不要在前面加上 SPS 和 PPS 帧，只写出 I 帧和 P 帧。另外，将前4个字节（0x00 0x00 0x00 0x01）中包含的附件B起始码替换为I/P帧的大小。