static bool initialize_codec(struct aac_encoder *enc)
{
int ret;
enc->aframe = av_frame_alloc();
if (!enc->aframe) {
aac_warn("initialize_codec", "Failed to allocate audio frame");
return false;
}
ret = avcodec_open2(enc->context, enc->aac, NULL);
if (ret < 0) {
aac_warn("initialize_codec", "Failed to open AAC codec: %s",
av_err2str(ret));
return false;
}
enc->frame_size = enc->context->frame_size;
if (!enc->frame_size)
enc->frame_size = 1024;
enc->frame_size_bytes = enc->frame_size * (int)enc->audio_size;
ret = av_samples_alloc(enc->samples, NULL, enc->context->channels,
enc->frame_size, enc->context->sample_fmt, 0);
if (ret < 0) {
aac_warn("initialize_codec", "Failed to create audio buffer: "
"%s", av_err2str(ret));
return false;
}
return true;
}
static AVFrame *allocate_output_audio_frame(SCM scm_self, AVCodecContext *audio_codec, enum AVSampleFormat sample_fmt)
{
AVFrame *retval = allocate_frame(scm_self);
retval->format = sample_fmt;
retval->channel_layout = audio_codec->channel_layout;
retval->sample_rate = audio_codec->sample_rate;
if (audio_codec->codec->capabilities & AV_CODEC_CAP_VARIABLE_FRAME_SIZE)
retval->nb_samples = 2 * AV_INPUT_BUFFER_MIN_SIZE;
else
retval->nb_samples = audio_codec->frame_size;
#ifdef HAVE_AV_FRAME_GET_BUFFER
int err = av_frame_get_buffer(retval, 0);
if (err < 0) {
ffmpeg_destroy(scm_self);
scm_misc_error("allocate-output-audio-frame", "Error allocating audio frame memory", SCM_EOL);
};
#else
int channels = av_get_channel_layout_nb_channels(retval->channel_layout);
int err = av_samples_alloc(retval->data, &retval->linesize[0], channels, retval->nb_samples, retval->format, 0);
// TODO: need av_freep?
if (err < 0) {
ffmpeg_destroy(scm_self);
scm_misc_error("allocate-output-audio-frame", "Could not allocate audio buffer", SCM_EOL);
};
#endif
return retval;
}
/**
* 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.
*/
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->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->channels,
frame_size,
output_codec_context->sample_fmt, 0)) < 0) {
fprintf(stderr, "Could not allocate converted input samples (error '%s')\n", get_error_text(error));
av_freep(&(*converted_input_samples)[0]);
free(*converted_input_samples);
return error;
}
return 0;
}
//====================================================================================================================
// Append data to the list creating packet as necessary and filling TempData
//====================================================================================================================
void cSoundBlockList::AppendData(int64_t Position,int16_t *Data,int64_t DataLen) {
u_int8_t *CurData=(u_int8_t *)Data;
// Cut data to Packet
while ((DataLen+CurrentTempSize>=SoundPacketSize)) {
#if defined(LIBAV) && (LIBAVVERSIONINT<=8)
u_int8_t *Packet=(u_int8_t *)av_malloc(SoundPacketSize+8);
#else
u_int8_t *Packet=NULL;
int out_linesize=0;
av_samples_alloc(&Packet,&out_linesize,Channels,SoundPacketSize+8,SampleFormat,1);
#endif
if (Packet) {
if (CurrentTempSize>0) { // Use previously data store in TempData
int DataToUse=SoundPacketSize-CurrentTempSize;
memcpy(Packet,TempData,CurrentTempSize);
memcpy(Packet+CurrentTempSize,CurData,DataToUse);
DataLen -=DataToUse;
CurData +=DataToUse;
CurrentTempSize=0;
} else { // Construct a full packet
memcpy(Packet,CurData,SoundPacketSize);
DataLen-=SoundPacketSize;
CurData+=SoundPacketSize;
}
AppendPacket(Position,(int16_t *)Packet);
Position+=(double(SoundPacketSize/2*Channels)/double(SamplingRate))*AV_TIME_BASE;
}
}
if (DataLen>0) { // Store a partial packet in temp buffer
// Store data left to TempData
memcpy(TempData+CurrentTempSize,CurData,DataLen);
CurrentTempSize+=DataLen;
}
}
AVFilterBufferRef *ff_default_get_audio_buffer(AVFilterLink *link, int perms,
int nb_samples)
{
AVFilterBufferRef *samplesref = NULL;
uint8_t **data;
int planar = av_sample_fmt_is_planar(link->format);
int nb_channels = av_get_channel_layout_nb_channels(link->channel_layout);
int planes = planar ? nb_channels : 1;
int linesize;
if (!(data = av_mallocz(sizeof(*data) * planes)))
goto fail;
if (av_samples_alloc(data, &linesize, nb_channels, nb_samples, link->format, 0) < 0)
goto fail;
samplesref = avfilter_get_audio_buffer_ref_from_arrays(data, linesize, perms,
nb_samples, link->format,
link->channel_layout);
if (!samplesref)
goto fail;
av_freep(&data);
fail:
if (data)
av_freep(&data[0]);
av_freep(&data);
return samplesref;
}
AVFilterBufferRef *avfilter_default_get_audio_buffer(AVFilterLink *link, int perms,
enum AVSampleFormat sample_fmt, int nb_samples,
int64_t channel_layout, int planar)
{
AVFilterBufferRef *samplesref = NULL;
int linesize[8];
uint8_t *data[8];
int nb_channels = av_get_channel_layout_nb_channels(channel_layout);
/* Calculate total buffer size, round to multiple of 16 to be SIMD friendly */
if (av_samples_alloc(data, linesize,
nb_channels, nb_samples, sample_fmt,
planar, 16) < 0)
return NULL;
samplesref =
avfilter_get_audio_buffer_ref_from_arrays(data, linesize, perms,
nb_samples, sample_fmt,
channel_layout, planar);
if (!samplesref) {
av_free(data[0]);
return NULL;
}
return samplesref;
}
AudioPlayerLoader::ReadResult ChildFFMpegLoader::readFromReadyFrame(QByteArray &result, int64 &samplesAdded) {
int res = 0;
if (_dstSamplesData) { // convert needed
int64_t dstSamples = av_rescale_rnd(swr_get_delay(_swrContext, _srcRate) + _frame->nb_samples, _dstRate, _srcRate, AV_ROUND_UP);
if (dstSamples > _maxResampleSamples) {
_maxResampleSamples = dstSamples;
av_free(_dstSamplesData[0]);
if ((res = av_samples_alloc(_dstSamplesData, 0, AudioToChannels, _maxResampleSamples, AudioToFormat, 1)) < 0) {
_dstSamplesData[0] = 0;
char err[AV_ERROR_MAX_STRING_SIZE] = { 0 };
LOG(("Audio Error: Unable to av_samples_alloc for file '%1', data size '%2', error %3, %4").arg(file.name()).arg(data.size()).arg(res).arg(av_make_error_string(err, sizeof(err), res)));
return ReadResult::Error;
}
}
if ((res = swr_convert(_swrContext, _dstSamplesData, dstSamples, (const uint8_t**)_frame->extended_data, _frame->nb_samples)) < 0) {
char err[AV_ERROR_MAX_STRING_SIZE] = { 0 };
LOG(("Audio Error: Unable to swr_convert for file '%1', data size '%2', error %3, %4").arg(file.name()).arg(data.size()).arg(res).arg(av_make_error_string(err, sizeof(err), res)));
return ReadResult::Error;
}
int32 resultLen = av_samples_get_buffer_size(0, AudioToChannels, res, AudioToFormat, 1);
result.append((const char*)_dstSamplesData[0], resultLen);
samplesAdded += resultLen / _sampleSize;
} else {
result.append((const char*)_frame->extended_data[0], _frame->nb_samples * _sampleSize);
samplesAdded += _frame->nb_samples;
}
return ReadResult::Ok;
}
/**
* Resample a buffer using FFAudioUI->swr_context.
* The returned out buffer needs to be freed by the caller.
*
* @param aio FFAudioIO context
* @param out_buf out buffer
* @param out_samples out samples
* @param in_buf in buffer
* @param in_samples in samples
* @return number of samples copied/converted or a negative value, should things go wrong
*/
static int resample(FFAudioIO *aio, uint8_t **out_buf, int out_samples, const uint8_t **in_buf, const int in_samples) {
int res = 0;
int64_t out_channel_count;
enum AVSampleFormat out_sample_format;
if (out_samples == 0) goto bail; // nothing to do.
av_opt_get_int(aio->swr_context, "out_channel_count", 0, &out_channel_count);
av_opt_get_sample_fmt(aio->swr_context, "out_sample_fmt", 0, &out_sample_format);
#ifdef DEBUG
fprintf(stderr, "resample: out_samples=%d in_samples=%d, channels=%d sample_format=%d\n",
out_samples, in_samples, (int)out_channel_count, out_sample_format);
#endif
// allocate temp buffer for resampled data
res = av_samples_alloc(out_buf, NULL, out_channel_count, out_samples, out_sample_format, 1);
if (res < 0) {
res = AVERROR(ENOMEM);
throwIOExceptionIfError(aio->env, res, "Could not allocate resample buffer.");
goto bail;
}
// run the SWR conversion (even if it is not strictly necessary)
res = swr_convert(aio->swr_context, out_buf, out_samples, in_buf, in_samples);
if (res < 0) {
throwIOExceptionIfError(aio->env, res, "Failed to convert audio data.");
goto bail;
}
bail:
return res;
}
bool FFmpeg_Decoder::getAVAudioData()
{
int got_frame;
if((*mStream)->codec->codec_type != AVMEDIA_TYPE_AUDIO)
return false;
do {
if(mPacket.size == 0 && !getNextPacket())
return false;
/* Decode some data, and check for errors */
int len = 0;
if((len=avcodec_decode_audio4((*mStream)->codec, mFrame, &got_frame, &mPacket)) < 0)
return false;
/* Move the unread data to the front and clear the end bits */
int remaining = mPacket.size - len;
if(remaining <= 0)
av_free_packet(&mPacket);
else
{
memmove(mPacket.data, &mPacket.data[len], remaining);
av_shrink_packet(&mPacket, remaining);
}
if (!got_frame || mFrame->nb_samples == 0)
continue;
if(mSwr)
{
if(!mDataBuf || mDataBufLen < mFrame->nb_samples)
{
av_freep(&mDataBuf);
if(av_samples_alloc(&mDataBuf, NULL, av_get_channel_layout_nb_channels(mOutputChannelLayout),
mFrame->nb_samples, mOutputSampleFormat, 0) < 0)
return false;
else
mDataBufLen = mFrame->nb_samples;
}
if(swr_convert(mSwr, (uint8_t**)&mDataBuf, mFrame->nb_samples,
(const uint8_t**)mFrame->extended_data, mFrame->nb_samples) < 0)
{
return false;
}
mFrameData = &mDataBuf;
}
else
mFrameData = &mFrame->data[0];
} while(got_frame == 0 || mFrame->nb_samples == 0);
mNextPts += (double)mFrame->nb_samples / (double)(*mStream)->codec->sample_rate;
return true;
}
int MpegLoader::readMore(QByteArray &result, qint64 &samplesAdded)
{
int res;
if ((res = av_read_frame(FmtContext_, &Avpkt_)) < 0)
return -1;
if (Avpkt_.stream_index == StreamId_)
{
av_frame_unref(Frame_);
int got_frame = 0;
if ((res = avcodec_decode_audio4(CodecContext_, Frame_, &got_frame, &Avpkt_)) < 0)
{
av_packet_unref(&Avpkt_);
if (res == AVERROR_INVALIDDATA)
return 0;
return -1;
}
if (got_frame)
{
if (OutSamplesData_)
{
int64_t dstSamples = av_rescale_rnd(swr_get_delay(SwrContext_, SrcRate_) + Frame_->nb_samples, DstRate_, SrcRate_, AV_ROUND_UP);
if (dstSamples > MaxResampleSamples_)
{
MaxResampleSamples_ = dstSamples;
av_free(OutSamplesData_[0]);
if ((res = av_samples_alloc(OutSamplesData_, 0, OutChannels, MaxResampleSamples_, OutFormat, 1)) < 0)
{
OutSamplesData_[0] = 0;
av_packet_unref(&Avpkt_);
return -1;
}
}
if ((res = swr_convert(SwrContext_, OutSamplesData_, dstSamples, (const uint8_t**)Frame_->extended_data, Frame_->nb_samples)) < 0)
{
av_packet_unref(&Avpkt_);
return -1;
}
qint32 resultLen = av_samples_get_buffer_size(0, OutChannels, res, OutFormat, 1);
result.append((const char*)OutSamplesData_[0], resultLen);
samplesAdded += resultLen / SampleSize_;
}
else
{
result.append((const char*)Frame_->extended_data[0], Frame_->nb_samples * SampleSize_);
samplesAdded += Frame_->nb_samples;
}
}
}
av_packet_unref(&Avpkt_);
return 1;
}
int alloc_samples_array_and_data(uint8_t ***data, int *linesize, int nb_channels,
int nb_samples, enum AVSampleFormat sample_fmt, int align)
{
int nb_planes = av_sample_fmt_is_planar(sample_fmt) ? nb_channels : 1;
*data = av_malloc(sizeof(*data) * nb_planes);
if (!*data)
return AVERROR(ENOMEM);
return av_samples_alloc(*data, linesize, nb_channels,
nb_samples, sample_fmt, align);
}
int static av_samples_alloc_array_and_samples(uint8_t ***audio_data, int *linesize, int nb_channels,
int nb_samples, enum AVSampleFormat sample_fmt, int align)
{
int ret, nb_planes = av_sample_fmt_is_planar(sample_fmt) ? nb_channels : 1;
*audio_data = av_malloc(nb_planes * sizeof(**audio_data));
memset(*audio_data, 0, nb_planes * sizeof(**audio_data));
if (!*audio_data)
return AVERROR(ENOMEM);
ret = av_samples_alloc(*audio_data, linesize, nb_channels,
nb_samples, sample_fmt, align);
if (ret < 0)
av_freep(audio_data);
return ret;
}
AudioBufferPtr AudioDecoderThread::resampleAudio(char* pDecodedData, int framesDecoded,
int currentSampleFormat)
{
if (!m_pResampleContext) {
#ifdef LIBAVRESAMPLE_VERSION
m_pResampleContext = avresample_alloc_context();
av_opt_set_int(m_pResampleContext, "in_channel_layout",
av_get_default_channel_layout(m_pStream->codec->channels), 0);
av_opt_set_int(m_pResampleContext, "out_channel_layout", AV_CH_LAYOUT_STEREO, 0);
av_opt_set_int(m_pResampleContext, "in_sample_rate", m_InputSampleRate, 0);
av_opt_set_int(m_pResampleContext, "out_sample_rate", m_AP.m_SampleRate, 0);
av_opt_set_int(m_pResampleContext, "in_sample_fmt",
(AVSampleFormat)currentSampleFormat, 0);
av_opt_set_int(m_pResampleContext, "out_sample_fmt", AV_SAMPLE_FMT_S16, 0);
int err = avresample_open(m_pResampleContext);
AVG_ASSERT(err >= 0);
#else
m_pResampleContext = av_audio_resample_init(m_AP.m_Channels,
m_pStream->codec->channels, m_AP.m_SampleRate, m_InputSampleRate,
AV_SAMPLE_FMT_S16, (AVSampleFormat)currentSampleFormat, 16, 10, 0, 0.8);
#endif
AVG_ASSERT(m_pResampleContext);
}
#ifdef LIBAVRESAMPLE_VERSION
uint8_t *pResampledData;
int leftoverSamples = avresample_available(m_pResampleContext);
int framesAvailable = leftoverSamples +
av_rescale_rnd(avresample_get_delay(m_pResampleContext) +
framesDecoded, m_AP.m_SampleRate, m_InputSampleRate, AV_ROUND_UP);
av_samples_alloc(&pResampledData, 0, 2, framesAvailable,
AV_SAMPLE_FMT_S16, 0);
int framesResampled = avresample_convert(m_pResampleContext, &pResampledData, 0,
framesAvailable, (uint8_t**)&pDecodedData, 0, framesDecoded);
AudioBufferPtr pBuffer(new AudioBuffer(framesResampled, m_AP));
memcpy(pBuffer->getData(), pResampledData,
framesResampled*m_AP.m_Channels*sizeof(short));
av_freep(&pResampledData);
#else
short pResampledData[AVCODEC_MAX_AUDIO_FRAME_SIZE/2];
int framesResampled = audio_resample(m_pResampleContext, pResampledData,
(short*)pDecodedData, framesDecoded);
AudioBufferPtr pBuffer(new AudioBuffer(framesResampled, m_AP));
memcpy(pBuffer->getData(), pResampledData,
framesResampled*m_AP.m_Channels*sizeof(short));
#endif
return pBuffer;
}
AVFilterBufferRef *avfilter_default_get_audio_buffer(AVFilterLink *link, int perms,
enum AVSampleFormat sample_fmt, int nb_samples,
int64_t channel_layout, int planar)
{
AVFilterBuffer *samples = av_mallocz(sizeof(AVFilterBuffer));
AVFilterBufferRef *ref = NULL;
int nb_channels = av_get_channel_layout_nb_channels(channel_layout);
if (!samples || !(ref = av_mallocz(sizeof(AVFilterBufferRef))))
goto fail;
ref->buf = samples;
ref->format = sample_fmt;
ref->audio = av_mallocz(sizeof(AVFilterBufferRefAudioProps));
if (!ref->audio)
goto fail;
ref->audio->channel_layout = channel_layout;
ref->audio->nb_samples = nb_samples;
ref->audio->planar = planar;
/* make sure the buffer gets read permission or it's useless for output */
ref->perms = perms | AV_PERM_READ;
samples->refcount = 1;
samples->free = ff_avfilter_default_free_buffer;
/* Calculate total buffer size, round to multiple of 16 to be SIMD friendly */
if (av_samples_alloc(samples->data, samples->linesize,
nb_channels, nb_samples, sample_fmt,
planar, 16) < 0)
goto fail;
memcpy(ref->data, samples->data, sizeof(ref->data));
memcpy(ref->linesize, samples->linesize, sizeof(ref->linesize));
return ref;
fail:
if (ref)
av_free(ref->audio);
av_free(ref);
av_free(samples);
return NULL;
}
void AudioDecoder::handleDecodedFrame(AVFrame* frame) {
/* compute destination number of samples */
m_outNumSamples = static_cast<int>(av_rescale_rnd(getDelay(m_resampleCtx, frame->sample_rate)
+ frame->nb_samples, OUT_SAMPLE_RATE, frame->sample_rate,
AV_ROUND_UP));
if (m_outNumSamples > m_maxOutNumSamples) {
av_freep(&m_outData[0]);
auto ret = av_samples_alloc(m_outData, &m_outLinesize, OUT_NUM_CHANNELS, m_outNumSamples,
OUT_SAMPLE_FORMAT, 1);
if (ret < 0) {
mprintf(("FFMPEG: Failed to allocate samples!!!"));
return;
}
m_maxOutNumSamples = m_outNumSamples;
}
/* convert to destination format */
auto ret = resample_convert(m_resampleCtx, m_outData, 0, m_outNumSamples,
(uint8_t**) frame->data, 0, frame->nb_samples);
if (ret < 0) {
mprintf(("FFMPEG: Error while converting audio!\n"));
return;
}
auto outBufsize = av_samples_get_buffer_size(&m_outLinesize, OUT_NUM_CHANNELS, ret, OUT_SAMPLE_FORMAT, 1);
if (outBufsize < 0) {
mprintf(("FFMPEG: Could not get sample buffer size!\n"));
return;
}
auto begin = reinterpret_cast<short*>(m_outData[0]);
auto end = reinterpret_cast<short*>(m_outData[0] + outBufsize);
auto size = std::distance(begin, end);
auto newSize = m_audioBuffer.size() + size;
if (newSize <= m_audioBuffer.capacity()) {
// We haven't filled the buffer yet
m_audioBuffer.insert(m_audioBuffer.end(), begin, end);
} else {
flushAudioBuffer();
m_audioBuffer.assign(begin, end);
}
}
int static decodeFrame(AVCodecContext* codecCtx, uint8_t **buffer,
int bufferSize, packetQueue* queue,
SwrContext *swr)
{
AVPacket packet;
AVFrame *decodedAudioFrame = av_frame_alloc();
int frameDecoded, framesRead, ret;
//Get a packet.
if (!queue->get(&packet))
goto err_out;
framesRead = avcodec_decode_audio4(codecCtx, decodedAudioFrame,
&frameDecoded, &packet);
av_free_packet(&packet);
if(framesRead < 0)
//Skip this packet if we have an error.
goto err_out;
// Note that the buffer that we're copying into is 3/2 times the
// size of a an audio frame, so we shouldn't need to check for a
// buffer overflow here.
if(frameDecoded) {
if (av_samples_alloc(buffer, &ret,
av_get_channel_layout_nb_channels(AV_CH_LAYOUT_STEREO),
decodedAudioFrame->nb_samples,
AV_SAMPLE_FMT_S16, 0) < 0)
throw std::runtime_error("Could not allocate decode samples buffer");
swr_convert(swr, buffer, decodedAudioFrame->nb_samples,
(const uint8_t **)decodedAudioFrame->data,
decodedAudioFrame->nb_samples);
goto out;
}
err_out:
*buffer = NULL;
ret = 0;
out:
av_frame_unref(decodedAudioFrame);
return ret;
}
void AudioConverter::feedEncoder()
{
int gotFrame = 0, err;
uint8_t *output;
int out_linesize;
int avail;
if (!m_encoderInitialized)
initEncoder();
assert(m_avpktOutUsed == m_avpktOut.size);
do
{
avail = avresample_available(m_resampler);
av_samples_alloc(&output, &out_linesize, m_destinationFormat.mChannelsPerFrame, avail, m_encoder->sample_fmt, 0);
if (avresample_read(m_resampler, &output, avail) != avail)
throw std::runtime_error("avresample_read() failed");
av_init_packet(&m_avpktOut);
m_avpktOut.data = 0;
m_avpktOut.size = 0;
m_avpktOutUsed = 0;
LOG << "Got " << avail << " samples\n";
err = avcodec_fill_audio_frame(m_audioFrame, m_encoder->channels,
m_encoder->sample_fmt, output, avail * m_destinationFormat.mChannelsPerFrame * (m_destinationFormat.mBitsPerChannel / 8),
m_destinationFormat.mChannelsPerFrame * (m_destinationFormat.mBitsPerChannel / 8));
if (err < 0)
throw std::runtime_error("avcodec_fill_audio_frame() failed");
// Encode PCM data
err = avcodec_encode_audio2(m_encoder, &m_avpktOut, m_audioFrame, &gotFrame);
av_freep(&output);
if (err < 0)
throw std::runtime_error("avcodec_encode_audio2() failed");
}
while(!gotFrame);
}
AVRaw *make_audio_raw(int format, int channel, int samples)
{
AVRaw * praw = (AVRaw*)malloc(sizeof(AVRaw));
while (praw)
{
memset(praw, 0, sizeof(AVRaw));
praw->type = RAW_AUDIO;
praw->format = format;
praw->channels = channel;
praw->samples = samples;
int ret = av_samples_alloc(praw->data, praw->linesize, channel, samples, (AVSampleFormat)format, 0);
if (ret > 0)
{
praw->size = ret;
}
else
{
char errmsg[ERROR_BUFFER_SIZE];
av_strerror(ret, errmsg, ERROR_BUFFER_SIZE);
av_log(NULL, AV_LOG_FATAL, "make_audio_raw av_samples_alloc : %s \n", errmsg);
break;
}
return praw;
}
/*
* �������
*/
if (praw)
{
ffFreeRaw(praw);
}
else
{
av_log(NULL, AV_LOG_FATAL, "make_audio_raw out of memory.\n");
}
praw = NULL;
return praw;
}
AVFilterBufferRef *ff_default_get_audio_buffer(AVFilterLink *link, int perms,
int nb_samples)
{
AVFilterBufferRef *samplesref = NULL;
uint8_t **data;
int planar = av_sample_fmt_is_planar(link->format);
int nb_channels = av_get_channel_layout_nb_channels(link->channel_layout);
int planes = planar ? nb_channels : 1;
int linesize;
int full_perms = AV_PERM_READ | AV_PERM_WRITE | AV_PERM_PRESERVE |
AV_PERM_REUSE | AV_PERM_REUSE2 | AV_PERM_ALIGN;
av_assert1(!(perms & ~(full_perms | AV_PERM_NEG_LINESIZES)));
if (!(data = av_mallocz(sizeof(*data) * planes)))
goto fail;
if (av_samples_alloc(data, &linesize, nb_channels, nb_samples, link->format, 0) < 0)
goto fail;
samplesref = avfilter_get_audio_buffer_ref_from_arrays(data, linesize, full_perms,
nb_samples, link->format,
link->channel_layout);
if (!samplesref)
goto fail;
samplesref->audio->sample_rate = link->sample_rate;
av_freep(&data);
fail:
if (data)
av_freep(&data[0]);
av_freep(&data);
return samplesref;
}
long audio_tutorial_resample(VideoState *is, struct AVFrame *inframe) {
#ifdef __RESAMPLER__
#ifdef __LIBAVRESAMPLE__
// There is pre 1.0 libavresample and then there is above..
#if LIBAVRESAMPLE_VERSION_MAJOR == 0
void **resample_input_bytes = (void **)inframe->extended_data;
#else
uint8_t **resample_input_bytes = (uint8_t **)inframe->extended_data;
#endif
#else
uint8_t **resample_input_bytes = (uint8_t **)inframe->extended_data;
#endif
int resample_nblen = 0;
long resample_long_bytes = 0;
if( is->pResampledOut == NULL || inframe->nb_samples > is->resample_size) {
#if __LIBAVRESAMPLE__
is->resample_size = av_rescale_rnd(avresample_get_delay(is->pSwrCtx) +
inframe->nb_samples,
44100,
44100,
AV_ROUND_UP);
#else
is->resample_size = av_rescale_rnd(swr_get_delay(is->pSwrCtx,
44100) +
inframe->nb_samples,
44100,
44100,
AV_ROUND_UP);
#endif
if(is->pResampledOut != NULL) {
av_free(is->pResampledOut);
is->pResampledOut = NULL;
}
av_samples_alloc(&is->pResampledOut, &is->resample_lines, 2, is->resample_size,
AV_SAMPLE_FMT_S16, 0);
}
#ifdef __LIBAVRESAMPLE__
// OLD API (0.0.3) ... still NEW API (1.0.0 and above).. very frustrating..
// USED IN FFMPEG 1.0 (LibAV SOMETHING!). New in FFMPEG 1.1 and libav 9
#if LIBAVRESAMPLE_VERSION_INT <= 3
// AVResample OLD
resample_nblen = avresample_convert(is->pSwrCtx, (void **)&is->pResampledOut, 0,
is->resample_size,
(void **)resample_input_bytes, 0, inframe->nb_samples);
#else
//AVResample NEW
resample_nblen = avresample_convert(is->pSwrCtx, (uint8_t **)&is->pResampledOut,
0, is->resample_size,
(uint8_t **)resample_input_bytes, 0, inframe->nb_samples);
#endif
#else
// SWResample
resample_nblen = swr_convert(is->pSwrCtx, (uint8_t **)&is->pResampledOut,
is->resample_size,
(const uint8_t **)resample_input_bytes, inframe->nb_samples);
#endif
resample_long_bytes = av_samples_get_buffer_size(NULL, 2, resample_nblen,
AV_SAMPLE_FMT_S16, 1);
if (resample_nblen < 0) {
fprintf(stderr, "reSample to another sample format failed!\n");
return -1;
}
return resample_long_bytes;
#else
return -1;
#endif
}
int submitAudioFrame(char *data, int length, long frameTimestamp) {
AVPacket pkt;
int ret;
AVFrame *frame;
int got_packet;
int sampleCount;
// If no stream header has been written yet, don't do anything
if (hasWrittenStreamHeader == 0) {
return 0;
}
frame = avcodec_alloc_frame();
if (frame == NULL) {
fprintf(stderr, "Failed to allocate frame\n");
return -1;
}
av_init_packet(&pkt);
// Copy our data in
memcpy(srcSamplesData[0], data, length);
srcSamplesCount = length / 4;
// Resample to floating point
sampleCount = av_rescale_rnd(swr_get_delay(swrContext, audioCodecCtx->sample_rate) + srcSamplesCount,
audioCodecCtx->sample_rate, audioCodecCtx->sample_rate, AV_ROUND_UP);
if (sampleCount > maxDstSamplesCount) {
// Need to resize the buffer
av_free(dstSamplesData[0]);
ret = av_samples_alloc(dstSamplesData, &dstSamplesLinesize, audioCodecCtx->channels,
sampleCount, audioCodecCtx->sample_fmt, 0);
if (ret < 0) {
fprintf(stderr, "av_samples_alloc() failed: %d\n", ret);
return ret;
}
maxDstSamplesCount = sampleCount;
dstSamplesSize = av_samples_get_buffer_size(NULL, audioCodecCtx->channels, sampleCount,
audioCodecCtx->sample_fmt, 0);
}
ret = swr_convert(swrContext, dstSamplesData, sampleCount,
(const unsigned char **)srcSamplesData, srcSamplesCount);
if (ret < 0) {
fprintf(stderr, "swr_convert() failed: %d\n", ret);
return ret;
}
frame->nb_samples = sampleCount;
ret = avcodec_fill_audio_frame(frame, audioCodecCtx->channels,
audioCodecCtx->sample_fmt, dstSamplesData[0], dstSamplesSize, 0);
if (ret < 0) {
fprintf(stderr, "avcodec_fill_audio_frame() failed: %d\n", ret);
avcodec_free_frame(&frame);
return ret;
}
// pkt is freed on failure or !got_packet
pkt.data = NULL;
pkt.size = 0;
ret = avcodec_encode_audio2(audioCodecCtx, &pkt, frame, &got_packet);
if (ret < 0) {
fprintf(stderr, "avcodec_encode_audio2() failed: %d\n", ret);
avcodec_free_frame(&frame);
return ret;
}
if (!got_packet) {
// Nothing to do here
return 0;
}
pkt.stream_index = audioStream->index;
pkt.pts = frameTimestamp;
pthread_mutex_lock(&streamLock);
ret = av_interleaved_write_frame(formatContext, &pkt);
pthread_mutex_unlock(&streamLock);
avcodec_free_frame(&frame);
av_free_packet(&pkt);
if (ret != 0) {
fprintf(stderr, "av_interleaved_write_frame() failed: %d\n", ret);
return ret;
}
return 0;
}
int main(int argc, char **argv)
{
int64_t src_ch_layout = AV_CH_LAYOUT_STEREO, dst_ch_layout = AV_CH_LAYOUT_SURROUND;
int src_rate = 48000, dst_rate = 44100;
uint8_t **src_data = NULL, **dst_data = NULL;
int src_nb_channels = 0, dst_nb_channels = 0;
int src_linesize, dst_linesize;
int src_nb_samples = 1024, dst_nb_samples, max_dst_nb_samples;
enum AVSampleFormat src_sample_fmt = AV_SAMPLE_FMT_DBL, dst_sample_fmt = AV_SAMPLE_FMT_S16;
const char *dst_filename = NULL;
FILE *dst_file;
int dst_bufsize;
const char *fmt;
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);
}
/* create resampler context */
swr_ctx = swr_alloc();
if (!swr_ctx) {
fprintf(stderr, "Could not allocate resampler context\n");
ret = AVERROR(ENOMEM);
goto end;
}
/* set options */
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);
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);
/* 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 */
src_nb_channels = av_get_channel_layout_nb_channels(src_ch_layout);
ret = alloc_samples_array_and_data(&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 */
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);
ret = alloc_samples_array_and_data(&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 */
fill_samples((double *)src_data[0], src_nb_samples, src_nb_channels, src_rate, &t);
/* compute destination number of samples */
dst_nb_samples = av_rescale_rnd(swr_get_delay(swr_ctx, src_rate) +
src_nb_samples, dst_rate, src_rate, AV_ROUND_UP);
if (dst_nb_samples > max_dst_nb_samples) {
av_free(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;
}
/* convert to destination format */
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);
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);
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:
if (dst_file)
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;
}
int udpsocket::ts_demux(void)
{
AVCodec *pVideoCodec[VIDEO_NUM];
AVCodec *pAudioCodec[AUDIO_NUM];
AVCodecContext *pVideoCodecCtx[VIDEO_NUM];
AVCodecContext *pAudioCodecCtx[AUDIO_NUM];
AVIOContext * pb;
AVInputFormat *piFmt;
AVFormatContext *pFmt;
uint8_t *buffer;
int videoindex[VIDEO_NUM];
int audioindex[AUDIO_NUM];
AVStream *pVst[VIDEO_NUM];
AVStream *pAst[AUDIO_NUM];
AVFrame *pVideoframe[VIDEO_NUM];
AVFrame *pAudioframe[AUDIO_NUM];
AVFrame *pOutAudioframe[AUDIO_NUM];
AVFrame *pOutAudioframelast[AUDIO_NUM];
AVPacket pkt;
int got_picture;
int video_num[VIDEO_NUM];
int audio_num[AUDIO_NUM];
int frame_size;
//transcodepool
transcodepool* pVideoTransPool[VIDEO_NUM];
transcodepool* pAudioTransPool[AUDIO_NUM];
for( int i=0; i<VIDEO_NUM; i++ ){
pVideoCodec[i] = NULL;
pVideoCodecCtx[i] =NULL;
videoindex[i] = -1;
pVst[i] = NULL;
video_num[i] = 0;
pVideoframe[i] = NULL;
pVideoframe[i] = av_frame_alloc();
pVideoTransPool[i] = NULL;
}
for( int i=0; i<AUDIO_NUM; i++ ){
pAudioCodec[i] = NULL;
pAudioCodecCtx[i] = NULL;
audioindex[i] = -1;
pAst[i] = NULL;
audio_num[i] = 0;
pOutAudioframe[i] = NULL;
pOutAudioframe[i] = av_frame_alloc();
pOutAudioframelast[i] = NULL;
pOutAudioframelast[i] = av_frame_alloc();
pAudioframe[i] = NULL;
pAudioframe[i] = av_frame_alloc();
pAudioTransPool[i] = NULL;
}
pb = NULL;
piFmt = NULL;
pFmt = NULL;
buffer = (uint8_t*)av_mallocz(sizeof(uint8_t)*BUFFER_SIZE);
got_picture = 0;
frame_size = AVCODEC_MAX_AUDIO_FRAME_SIZE*3/2;
//encoder
AVFormatContext *ofmt_ctx = NULL;
AVPacket enc_pkt;
AVStream *out_stream;
AVCodecContext *enc_ctx;
AVCodec *encoder;
AVFormatContext *outAudioFormatCtx[AUDIO_NUM];
AVPacket audio_pkt;
AVStream *audio_stream[AUDIO_NUM];
AVCodecContext *AudioEncodeCtx[AUDIO_NUM];
AVCodec *AudioEncoder[AUDIO_NUM];
fp_v = fopen("OUT.h264","wb+"); //输出文件
fp_a = fopen("audio_out.aac","wb+");
//FFMPEG
av_register_all();
pb = avio_alloc_context(buffer, 4096, 0, NULL, read_data, NULL, NULL);
// printf("thread %d pid %lu tid %lu\n",index,(unsigned long)getpid(),(unsigned long)pthread_self());
if (!pb) {
fprintf(stderr, "avio alloc failed!\n");
return -1;
}
int x = av_probe_input_buffer(pb, &piFmt, "", NULL, 0, 0);
if (x < 0) {
printf("probe error: %d",x);
// fprintf(stderr, "probe failed!\n");
} else {
fprintf(stdout, "probe success!\n");
fprintf(stdout, "format: %s[%s]\n", piFmt->name, piFmt->long_name);
}
pFmt = avformat_alloc_context();
pFmt->pb = pb;
if (avformat_open_input(&pFmt, "", piFmt, NULL) < 0) {
fprintf(stderr, "avformat open failed.\n");
return -1;
} else {
fprintf(stdout, "open stream success!\n");
}
//pFmt->probesize = 4096 * 2000;
//pFmt->max_analyze_duration = 5 * AV_TIME_BASE;
//pFmt->probesize = 2048;
// pFmt->max_analyze_duration = 1000;
pFmt->probesize = 2048 * 1000 ;
pFmt->max_analyze_duration = 2048 * 1000;
if (avformat_find_stream_info(pFmt,0) < 0) {
fprintf(stderr, "could not fine stream.\n");
return -1;
}
printf("dump format\n");
av_dump_format(pFmt, 0, "", 0);
int videox = 0,audiox = 0;
for (int i = 0; i < pFmt->nb_streams; i++) {
if(videox == 7 && audiox == 7)
break;
if ( pFmt->streams[i]->codec->codec_type == AVMEDIA_TYPE_VIDEO && videox < 7 ) {
videoindex[ videox++ ] = i;
}
if ( pFmt->streams[i]->codec->codec_type == AVMEDIA_TYPE_AUDIO && audiox < 7 ) {
audioindex[ audiox++ ] = i;
}
}
for(int i=0; i<VIDEO_NUM; i++)
printf("videoindex %d = %d, audioindex %d = %d\n",i , videoindex[i], i ,audioindex[i]);
if (videoindex[6] < 0 || audioindex[6] < 0) {
fprintf(stderr, "videoindex=%d, audioindex=%d\n", videoindex[6], audioindex[6]);
return -1;
}
for( int i=0; i<VIDEO_NUM; i++ ){
pVst[i] = pFmt->streams[videoindex[i]];
pVideoCodecCtx[i] = pVst[i]->codec;
pVideoCodec[i] = avcodec_find_decoder(pVideoCodecCtx[i]->codec_id);
if (!pVideoCodec[i]) {
fprintf(stderr, "could not find video decoder!\n");
return -1;
}
if (avcodec_open2(pVideoCodecCtx[i], pVideoCodec[i], NULL) < 0) {
fprintf(stderr, "could not open video codec!\n");
return -1;
}
}
for( int i=0; i<AUDIO_NUM; i++ ){
pAst[i] = pFmt->streams[audioindex[i]];
pAudioCodecCtx[i] = pAst[i]->codec;
pAudioCodec[i] = avcodec_find_decoder(pAudioCodecCtx[i]->codec_id);
if (!pAudioCodec[i]) {
fprintf(stderr, "could not find audio decoder!\n");
return -1;
}
if (avcodec_open2(pAudioCodecCtx[i], pAudioCodec[i], NULL) < 0) {
fprintf(stderr, "could not open audio codec!\n");
return -1;
}
}
//video encoder init
avformat_alloc_output_context2(&ofmt_ctx, NULL, "h264", NULL);
unsigned char* outbuffer = NULL;
outbuffer = (unsigned char*)av_malloc(1024*1000);
AVIOContext *avio_out = NULL;
avio_out = avio_alloc_context(outbuffer, 1024*1000, 0, NULL, NULL, write_buffer,NULL);
if(avio_out == NULL){
printf("avio_out error\n");
return -1;
}
ofmt_ctx->pb = avio_out;
ofmt_ctx->flags = AVFMT_FLAG_CUSTOM_IO;
out_stream = avformat_new_stream(ofmt_ctx, NULL);
if(!out_stream){
av_log(NULL, AV_LOG_ERROR, "Failed allocating output stream\n");
return -1;
}
enc_ctx = out_stream->codec;
encoder = avcodec_find_encoder(AV_CODEC_ID_H264);
enc_ctx->height = pVideoCodecCtx[0]->height;
enc_ctx->width = pVideoCodecCtx[0]->width;
enc_ctx->sample_aspect_ratio = pVideoCodecCtx[0]->sample_aspect_ratio;
enc_ctx->pix_fmt = encoder->pix_fmts[0];
out_stream->time_base = pVst[0]->time_base;
// out_stream->time_base.num = 1;
// out_stream->time_base.den = 25;
enc_ctx->me_range = 16;
enc_ctx->max_qdiff = 4;
enc_ctx->qmin = 25;
enc_ctx->qmax = 40;
enc_ctx->qcompress = 0.6;
enc_ctx->refs = 3;
enc_ctx->bit_rate = 1000000;
int re = avcodec_open2(enc_ctx, encoder, NULL);
if (re < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot open video encoder for stream \n");
return re;
}
if(ofmt_ctx->oformat->flags & AVFMT_GLOBALHEADER)
enc_ctx->flags |= CODEC_FLAG_GLOBAL_HEADER;
re = avformat_write_header(ofmt_ctx, NULL);
if(re < 0){
av_log(NULL, AV_LOG_ERROR, "Error occured when opening output file\n");
return re;
}
//audio encoder
for( int i=0; i<AUDIO_NUM; i++){
outAudioFormatCtx[i] = NULL;
// audio_pkt = NULL;
audio_stream[i] = NULL;
AudioEncodeCtx[i] = NULL;
AudioEncoder[i] = NULL;
}
const char* out_audio_file = "transcodeaudio.aac"; //Output URL
//Method 1.
outAudioFormatCtx[0] = avformat_alloc_context();
outAudioFormatCtx[0]->oformat = av_guess_format(NULL, out_audio_file, NULL);
AVIOContext *avio_audio_out = NULL;
avio_audio_out = avio_alloc_context(outbuffer, 1024*1000, 0, NULL, NULL, write_buffer,NULL);
if(avio_audio_out == NULL){
printf("avio_out error\n");
return -1;
}
outAudioFormatCtx[0]->pb = avio_audio_out;
//Method 2.
//avformat_alloc_output_context2(&pFormatCtx, NULL, NULL, out_file);
//fmt = pFormatCtx->oformat;
//Open output URL
if (avio_open(&outAudioFormatCtx[0]->pb,out_audio_file, AVIO_FLAG_READ_WRITE) < 0){
printf("Failed to open output file!\n");
return -1;
}
//Show some information
av_dump_format(outAudioFormatCtx[0], 0, out_audio_file, 1);
AudioEncoder[0] = avcodec_find_encoder(AV_CODEC_ID_AAC);
if (!AudioEncoder[0]){
printf("Can not find encoder!\n");
return -1;
}
audio_stream[0] = avformat_new_stream(outAudioFormatCtx[0], AudioEncoder[0]);
if (audio_stream[0]==NULL){
return -1;
}
AudioEncodeCtx[0] = audio_stream[0]->codec;
AudioEncodeCtx[0]->codec_id = outAudioFormatCtx[0]->oformat->audio_codec;
AudioEncodeCtx[0]->codec_type = AVMEDIA_TYPE_AUDIO;
AudioEncodeCtx[0]->sample_fmt = AV_SAMPLE_FMT_S16;
AudioEncodeCtx[0]->sample_rate= 48000;//44100
AudioEncodeCtx[0]->channel_layout=AV_CH_LAYOUT_STEREO;
AudioEncodeCtx[0]->channels = av_get_channel_layout_nb_channels(AudioEncodeCtx[0]->channel_layout);
AudioEncodeCtx[0]->bit_rate = 64000;//64000
/** Allow the use of the experimental AAC encoder */
AudioEncodeCtx[0]->strict_std_compliance = FF_COMPLIANCE_EXPERIMENTAL;
/** Set the sample rate for the container. */
audio_stream[0]->time_base.den = pAudioCodecCtx[0]->sample_rate;
audio_stream[0]->time_base.num = 1;
if (avcodec_open2(AudioEncodeCtx[0], AudioEncoder[0],NULL) < 0){
printf("Failed to open encoder!\n");
return -1;
}
av_samples_get_buffer_size(NULL, AudioEncodeCtx[0]->channels,AudioEncodeCtx[0]->frame_size,AudioEncodeCtx[0]->sample_fmt, 1);
//uint8_t samples[AVCODEC_MAX_AUDIO_FRAME_SIZE*3/2];
av_init_packet(&pkt);
av_init_packet(&audio_pkt);
av_init_packet(&enc_pkt);
AVAudioFifo *af = NULL;
SwrContext *resample_context = NULL;
long long pts = 0;
/** Initialize the resampler to be able to convert audio sample formats. */
// if (init_resampler(input_codec_context, output_codec_context,
// &resample_context))
for(int i=0; i<1; i++){
printf("work \n");
printf(" samplerate input = %d , samplerate output = %d\n",pAudioCodecCtx[i]->sample_rate, AudioEncodeCtx[i]->sample_rate);
resample_context = swr_alloc_set_opts(NULL, av_get_default_channel_layout(AudioEncodeCtx[i]->channels),
AudioEncodeCtx[i]->sample_fmt,
AudioEncodeCtx[i]->sample_rate,
av_get_default_channel_layout(pAudioCodecCtx[i]->channels),
pAudioCodecCtx[i]->sample_fmt,
pAudioCodecCtx[i]->sample_rate,
0, NULL);
swr_init(resample_context);
}
af = av_audio_fifo_alloc(AudioEncodeCtx[0]->sample_fmt, AudioEncodeCtx[0]->channels, 1);
if(af == NULL)
{
printf("error af \n");
return -1;
}
while(1) {
if (av_read_frame(pFmt, &pkt) >= 0) {
for( int i=0; i<1; i++ ){
if (pkt.stream_index == videoindex[i]) {
// av_frame_free(&pframe);
avcodec_decode_video2(pVideoCodecCtx[i], pVideoframe[i], &got_picture, &pkt);
if (got_picture) {
if(videoindex[i] == 0){
// m_tsRecvPool->write_buffer(pkt.data, pkt.size);
pVideoframe[i]->pts = av_frame_get_best_effort_timestamp(pVideoframe[i]);
pVideoframe[i]->pict_type = AV_PICTURE_TYPE_NONE;
// printf("videoframesize0 = %d, size1 = %d, size2 = %d, size3 = %d, size4 = %d,format = %d\n",pVideoframe[i]->linesize[0],
// pVideoframe[i]->linesize[1],pVideoframe[i]->linesize[2],pVideoframe[i]->linesize[3],pVideoframe[i]->linesize[4],pVideoframe[i]->format);
// pVideoTransPool[i]->PutFrame( pVideoframe[i] ,i);
int enc_got_frame = 0;
/* ffmpeg encoder */
enc_pkt.data = NULL;
enc_pkt.size = 0;
av_init_packet(&enc_pkt);
re = avcodec_encode_video2(ofmt_ctx->streams[videoindex[i]]->codec, &enc_pkt,
pVideoframe[i], &enc_got_frame);
// printf("enc_got_frame =%d, re = %d \n",enc_got_frame, re);
printf("video Encode 1 Packet\tsize:%d\tpts:%lld\n",enc_pkt.size,enc_pkt.pts);
/* prepare packet for muxing */
// fwrite(enc_pkt.data,enc_pkt.size, 1, fp_v);
}
// printf(" video %d decode %d num\n", i, video_num[i]++);
break;
}
}else if (pkt.stream_index == audioindex[i]) {
if (avcodec_decode_audio4(pAudioCodecCtx[i], pAudioframe[i], &frame_size, &pkt) >= 0) {
if (i == 0){
// fwrite(pAudioframe[i]->data[0],pAudioframe[i]->linesize[0], 1, fp_a);
// printf("index = %d audio %d decode %d num\n", index, i, audio_num[i]++);
uint8_t *converted_input_samples = NULL;
converted_input_samples = (uint8_t *)calloc(AudioEncodeCtx[i]->channels, sizeof(*converted_input_samples));
av_samples_alloc(&converted_input_samples, NULL, AudioEncodeCtx[i]->channels, pAudioframe[i]->nb_samples, AudioEncodeCtx[i]->sample_fmt, 0);
int error = 0;
if((error = swr_convert(resample_context, &converted_input_samples, pAudioframe[i]->nb_samples,
(const uint8_t**)pAudioframe[i]->extended_data, pAudioframe[i]->nb_samples))<0){
printf("error : %d\n",error);
}
// av_audio_fifo_realloc(af, av_audio_fifo_size(af) + pAudioframe[i]->nb_samples);
av_audio_fifo_write(af, (void **)&converted_input_samples, pAudioframe[i]->nb_samples);
// fwrite(pkt.data,pkt.size, 1, fp_a);
// pAudioframe[i]->data[0] = frame_buf;
// init_converted_samples(&converted_input_samples, output_codec_context, pAudioframe[i]->nb_samples);
/** Initialize temporary storage for one output frame. */
// printf("pkt.size = %d , pkt.pts = %d ,pkt.dts = %d\n",pkt.size, pkt.pts, pkt.dts);
// printf("framesize = %d, audioframesize = %d\n", pAudioframe[i]->nb_samples, frame_size);
// pOutAudioframe[i]->pict_type = AV_PICTURE_TYPE_NONE;
int got_frame=0;
//Encode
// av_init_packet(&audio_pkt);
// audio_pkt.data = NULL;
// audio_pkt.size = 0;
// avcodec_encode_audio2(AudioEncodeCtx[0], &audio_pkt, pOutAudioframe[i], &got_frame);
// printf("Encode 1 Packet\tsize:%d\tpts:%lld\n", audio_pkt.size, audio_pkt.pts);
while(av_audio_fifo_size(af) >= AudioEncodeCtx[i]->frame_size){
int frame_size = FFMIN(av_audio_fifo_size(af),AudioEncodeCtx[0]->frame_size);
pOutAudioframe[i]->nb_samples = frame_size;
pOutAudioframe[i]->channel_layout = AudioEncodeCtx[0]->channel_layout;
pOutAudioframe[i]->sample_rate = AudioEncodeCtx[0]->sample_rate;
pOutAudioframe[i]->format = AudioEncodeCtx[0]->sample_fmt;
av_frame_get_buffer(pOutAudioframe[i], 0);
av_audio_fifo_read(af, (void **)&pOutAudioframe[i]->data, frame_size);
pOutAudioframe[i]->pts=pts;
pts += pOutAudioframe[i]->nb_samples;
audio_pkt.data = NULL;
audio_pkt.size = 0;
av_init_packet(&audio_pkt);
avcodec_encode_audio2(AudioEncodeCtx[0], &audio_pkt, pOutAudioframe[i], &got_frame);
printf("audio Encode 1 Packet\tsize:%d\tpts:%lld\n", audio_pkt.size, audio_pkt.pts);
fwrite(audio_pkt.data,audio_pkt.size, 1, fp_a);
}
}
// if(i == 0){
// fwrite(pkt.data,pkt.size, 1, fp_a);
// }
// printf("index = %d audio %d decode %d num\n", index, i, audio_num[i]++);
break;
}
}
}
av_free_packet(&pkt);
av_free_packet(&enc_pkt);
}
}
av_free(buffer);
for(int i=0; i<VIDEO_NUM; i++)
av_free(pVideoframe[i]);
for(int i=0; i<AUDIO_NUM; i++)
av_free(pAudioframe[i]);
return 0;
}
static int init_buffers(AVFilterLink *inlink, int nb_samples)
{
AConvertContext *aconvert = inlink->dst->priv;
AVFilterLink * const outlink = inlink->dst->outputs[0];
int i, packed_stride = 0;
const unsigned
packing_conv = inlink->planar != outlink->planar &&
aconvert->out_nb_channels != 1,
format_conv = inlink->format != outlink->format;
int nb_channels = aconvert->out_nb_channels;
uninit(inlink->dst);
aconvert->max_nb_samples = nb_samples;
if (aconvert->convert_chlayout) {
/* allocate buffer for storing intermediary mixing samplesref */
uint8_t *data[8];
int linesize[8];
int nb_channels = av_get_channel_layout_nb_channels(outlink->channel_layout);
if (av_samples_alloc(data, linesize, nb_channels, nb_samples,
inlink->format, inlink->planar, 16) < 0)
goto fail_no_mem;
aconvert->mix_samplesref =
avfilter_get_audio_buffer_ref_from_arrays(data, linesize, AV_PERM_WRITE,
nb_samples, inlink->format,
outlink->channel_layout,
inlink->planar);
if (!aconvert->mix_samplesref)
goto fail_no_mem;
}
// if there's a format/packing conversion we need an audio_convert context
if (format_conv || packing_conv) {
aconvert->out_samplesref =
avfilter_get_audio_buffer(outlink, AV_PERM_WRITE, nb_samples);
if (!aconvert->out_samplesref)
goto fail_no_mem;
aconvert->in_strides [0] = av_get_bytes_per_sample(inlink ->format);
aconvert->out_strides[0] = av_get_bytes_per_sample(outlink->format);
aconvert->out_conv = aconvert->out_samplesref->data;
if (aconvert->mix_samplesref)
aconvert->in_conv = aconvert->mix_samplesref->data;
if (packing_conv) {
// packed -> planar
if (outlink->planar == AVFILTER_PLANAR) {
if (aconvert->mix_samplesref)
aconvert->packed_data[0] = aconvert->mix_samplesref->data[0];
aconvert->in_conv = aconvert->packed_data;
packed_stride = aconvert->in_strides[0];
aconvert->in_strides[0] *= nb_channels;
// planar -> packed
} else {
aconvert->packed_data[0] = aconvert->out_samplesref->data[0];
aconvert->out_conv = aconvert->packed_data;
packed_stride = aconvert->out_strides[0];
aconvert->out_strides[0] *= nb_channels;
}
} else if (outlink->planar == AVFILTER_PACKED) {
/* If there's no packing conversion, and the stream is packed
* then we treat the entire stream as one big channel
*/
nb_channels = 1;
}
for (i = 1; i < nb_channels; i++) {
aconvert->packed_data[i] = aconvert->packed_data[i-1] + packed_stride;
aconvert->in_strides[i] = aconvert->in_strides[0];
aconvert->out_strides[i] = aconvert->out_strides[0];
}
aconvert->audioconvert_ctx =
av_audio_convert_alloc(outlink->format, nb_channels,
inlink->format, nb_channels, NULL, 0);
if (!aconvert->audioconvert_ctx)
goto fail_no_mem;
}
return 0;
fail_no_mem:
av_log(inlink->dst, AV_LOG_ERROR, "Could not allocate memory.\n");
return AVERROR(ENOMEM);
}
int procAudioResampling(AVCodecContext * audio_dec_ctx,
AVFrame * pAudioDecodeFrame,
int out_sample_fmt,
int out_channels,
int out_sample_rate,
uint8_t * audio_chunk)
{
SwrContext * swr_ctx = NULL;
int data_size = 0;
int ret = 0;
int64_t src_ch_layout = audio_dec_ctx->channel_layout;
int64_t dst_ch_layout = AV_CH_LAYOUT_STEREO;
int dst_nb_channels = 0;
int dst_linesize = 0;
int src_nb_samples = 0;
int dst_nb_samples = 0;
int max_dst_nb_samples = 0;
uint8_t **dst_data = NULL;
int resampled_data_size = 0;
swr_ctx = swr_alloc();
if (!swr_ctx)
{
LOGD("swr_alloc error \n");
return -1;
}
src_ch_layout = (audio_dec_ctx->channels ==
av_get_channel_layout_nb_channels(audio_dec_ctx->channel_layout)) ?
audio_dec_ctx->channel_layout :
av_get_default_channel_layout(audio_dec_ctx->channels);
if (out_channels == 1)
{
dst_ch_layout = AV_CH_LAYOUT_MONO;
//LOGD("dst_ch_layout: AV_CH_LAYOUT_MONO\n");
}
else if (out_channels == 2)
{
dst_ch_layout = AV_CH_LAYOUT_STEREO;
//LOGD("dst_ch_layout: AV_CH_LAYOUT_STEREO\n");
}
else
{
dst_ch_layout = AV_CH_LAYOUT_SURROUND;
//LOGD("dst_ch_layout: AV_CH_LAYOUT_SURROUND\n");
}
if (src_ch_layout <= 0)
{
LOGD("src_ch_layout error \n");
return -1;
}
src_nb_samples = pAudioDecodeFrame->nb_samples;
if (src_nb_samples <= 0)
{
LOGD("src_nb_samples error \n");
return -1;
}
av_opt_set_int(swr_ctx, "in_channel_layout", src_ch_layout, 0);
av_opt_set_int(swr_ctx, "in_sample_rate", audio_dec_ctx->sample_rate, 0);
av_opt_set_sample_fmt(swr_ctx, "in_sample_fmt", audio_dec_ctx->sample_fmt, 0);
av_opt_set_int(swr_ctx, "out_channel_layout", dst_ch_layout, 0);
av_opt_set_int(swr_ctx, "out_sample_rate", out_sample_rate, 0);
av_opt_set_sample_fmt(swr_ctx, "out_sample_fmt", (AVSampleFormat)out_sample_fmt, 0);
if ((ret = swr_init(swr_ctx)) < 0) {
LOGD("Failed to initialize the resampling context\n");
return -1;
}
max_dst_nb_samples = dst_nb_samples = av_rescale_rnd(src_nb_samples,
out_sample_rate, audio_dec_ctx->sample_rate, AV_ROUND_UP);
if (max_dst_nb_samples <= 0)
{
LOGD("av_rescale_rnd error \n");
return -1;
}
dst_nb_channels = av_get_channel_layout_nb_channels(dst_ch_layout);
ret = av_samples_alloc_array_and_samples(&dst_data, &dst_linesize, dst_nb_channels,
dst_nb_samples, (AVSampleFormat)out_sample_fmt, 0);
if (ret < 0)
{
LOGD("av_samples_alloc_array_and_samples error \n");
return -1;
}
dst_nb_samples = av_rescale_rnd(swr_get_delay(swr_ctx, audio_dec_ctx->sample_rate) +
src_nb_samples, out_sample_rate, audio_dec_ctx->sample_rate, AV_ROUND_UP);
if (dst_nb_samples <= 0)
{
LOGD("av_rescale_rnd error \n");
return -1;
}
if (dst_nb_samples > max_dst_nb_samples)
{
av_free(dst_data[0]);
ret = av_samples_alloc(dst_data, &dst_linesize, dst_nb_channels,
dst_nb_samples, (AVSampleFormat)out_sample_fmt, 1);
max_dst_nb_samples = dst_nb_samples;
}
if (swr_ctx)
{
ret = swr_convert(swr_ctx, dst_data, dst_nb_samples,
(const uint8_t **)pAudioDecodeFrame->data, pAudioDecodeFrame->nb_samples);
if (ret < 0)
{
LOGD("swr_convert error \n");
return -1;
}
resampled_data_size = av_samples_get_buffer_size(&dst_linesize, dst_nb_channels,
ret, (AVSampleFormat)out_sample_fmt, 1);
if (resampled_data_size < 0)
{
LOGD("av_samples_get_buffer_size error \n");
return -1;
}
}
else
{
LOGD("swr_ctx null error \n");
return -1;
}
//LOGD("resampled_data_size:%d",resampled_data_size);
memcpy(audio_chunk, dst_data[0], resampled_data_size);
if (dst_data)
{
av_freep(&dst_data[0]);
}
av_freep(&dst_data);
dst_data = NULL;
if (swr_ctx)
{
swr_free(&swr_ctx);
}
return resampled_data_size;
}
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, frame->channels,
frame->nb_samples, 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, frame->channels,
frame->nb_samples, 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, frame->channels, 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;
}
int audioResampling(AVCodecContext *audio_dec_ctx, AVFrame *pAudioDecodeFrame,
int out_sample_fmt, int out_channels, int out_sample_rate, char *out_buf) {
__android_log_print(ANDROID_LOG_DEBUG, TAG, "stream12");
SwrContext *swr_ctx = NULL;
int data_size = 0;
int ret = 0;
int64_t src_ch_layout = AV_CH_LAYOUT_STEREO; //初始化这样根据不同文件做调整
int64_t dst_ch_layout = AV_CH_LAYOUT_STEREO; //这里设定ok
int dst_nb_channels = 0;
int dst_linesize = 0;
int src_nb_samples = 0;
int dst_nb_samples = 0;
int max_dst_nb_samples = 0;
uint8_t **dst_data = NULL;
int resampled_data_size = 0;
//重新采样
if (swr_ctx) {
swr_free(&swr_ctx);
}
swr_ctx = swr_alloc();
__android_log_print(ANDROID_LOG_DEBUG, TAG, "stream12-1");
if (!swr_ctx) {
printf("swr_alloc error \n");
__android_log_print(ANDROID_LOG_DEBUG, TAG, "stream13");
return -1;
}
src_ch_layout = (audio_dec_ctx->channel_layout &&
audio_dec_ctx->channels ==
av_get_channel_layout_nb_channels(audio_dec_ctx->channel_layout)) ?
audio_dec_ctx->channel_layout :
av_get_default_channel_layout(audio_dec_ctx->channels);
if (out_channels == 1) {
dst_ch_layout = AV_CH_LAYOUT_MONO;
}
else if (out_channels == 2) {
dst_ch_layout = AV_CH_LAYOUT_STEREO;
}
else {
//可扩展
}
__android_log_print(ANDROID_LOG_DEBUG, TAG, "stream12-2");
if (src_ch_layout <= 0) {
printf("src_ch_layout error \n");
__android_log_print(ANDROID_LOG_DEBUG, TAG, "stream14");
return -1;
}
__android_log_print(ANDROID_LOG_DEBUG, TAG, "stream12-3");
src_nb_samples = pAudioDecodeFrame->nb_samples;
if (src_nb_samples <= 0) {
printf("src_nb_samples error \n");
__android_log_print(ANDROID_LOG_DEBUG, TAG, "stream15");
return -1;
}
/* set options */
av_opt_set_int(swr_ctx, "in_channel_layout", src_ch_layout, 0);
av_opt_set_int(swr_ctx, "in_sample_rate", audio_dec_ctx->sample_rate, 0);
av_opt_set_sample_fmt(swr_ctx, "in_sample_fmt", audio_dec_ctx->sample_fmt, 0);
av_opt_set_int(swr_ctx, "out_channel_layout", dst_ch_layout, 0);
av_opt_set_int(swr_ctx, "out_sample_rate", out_sample_rate, 0);
av_opt_set_sample_fmt(swr_ctx, "out_sample_fmt", (AVSampleFormat) out_sample_fmt, 0);
swr_init(swr_ctx);
max_dst_nb_samples = dst_nb_samples =
av_rescale_rnd(src_nb_samples, out_sample_rate, audio_dec_ctx->sample_rate,
AV_ROUND_UP);
if (max_dst_nb_samples <= 0) {
printf("av_rescale_rnd error \n");
return -1;
}
dst_nb_channels = av_get_channel_layout_nb_channels(dst_ch_layout);
ret = av_samples_alloc_array_and_samples(&dst_data, &dst_linesize, dst_nb_channels,
dst_nb_samples, (AVSampleFormat) out_sample_fmt, 0);
if (ret < 0) {
printf("av_samples_alloc_array_and_samples error \n");
return -1;
}
dst_nb_samples = av_rescale_rnd(swr_get_delay(swr_ctx, audio_dec_ctx->sample_rate) +
src_nb_samples, out_sample_rate, audio_dec_ctx->sample_rate,
AV_ROUND_UP);
if (dst_nb_samples <= 0) {
return -1;
}
if (dst_nb_samples > max_dst_nb_samples) {
av_free(dst_data[0]);
ret = av_samples_alloc(dst_data, &dst_linesize, dst_nb_channels,
dst_nb_samples, (AVSampleFormat) out_sample_fmt, 1);
max_dst_nb_samples = dst_nb_samples;
}
data_size = av_samples_get_buffer_size(NULL, audio_dec_ctx->channels,
pAudioDecodeFrame->nb_samples,
audio_dec_ctx->sample_fmt, 1);
if (data_size <= 0) {
return -1;
}
resampled_data_size = data_size;
if (swr_ctx) {
ret = swr_convert(swr_ctx, dst_data, dst_nb_samples,
(const uint8_t **) pAudioDecodeFrame->data,
pAudioDecodeFrame->nb_samples);
if (ret <= 0) {
return -1;
}
resampled_data_size = av_samples_get_buffer_size(&dst_linesize, dst_nb_channels,
ret, (AVSampleFormat) out_sample_fmt, 1);
if (resampled_data_size <= 0) {
return -1;
}
}
else {
return -1;
}
//将值返回去
memcpy(out_buf, dst_data[0], resampled_data_size);
if (dst_data) {
av_freep(&dst_data[0]);
}
av_freep(&dst_data);
dst_data = NULL;
if (swr_ctx) {
swr_free(&swr_ctx);
}
return resampled_data_size;
}
Chroma::Result Chroma::operator() (const QString& filename)
{
std::shared_ptr<AVFormatContext> formatCtx;
{
AVFormatContext *formatCtxRaw = nullptr;
if (avformat_open_input (&formatCtxRaw, filename.toLatin1 ().constData (), nullptr, nullptr))
throw std::runtime_error ("error opening file");
formatCtx.reset (formatCtxRaw,
[] (AVFormatContext *ctx) {
avformat_close_input (&ctx);
});
}
{
QMutexLocker locker (&CodecMutex_);
if (avformat_find_stream_info (formatCtx.get (), nullptr) < 0)
throw std::runtime_error ("could not find stream");
}
AVCodec *codec = nullptr;
const auto streamIndex = av_find_best_stream (formatCtx.get (), AVMEDIA_TYPE_AUDIO, -1, -1, &codec, 0);
if (streamIndex < 0)
throw std::runtime_error ("could not find audio stream");
auto stream = formatCtx->streams [streamIndex];
bool codecOpened = false;
std::shared_ptr<AVCodecContext> codecCtx (stream->codec,
[&codecOpened] (AVCodecContext *ctx) {
if (codecOpened) avcodec_close (ctx);
});
{
QMutexLocker locker (&CodecMutex_);
if (avcodec_open2 (codecCtx.get (), codec, nullptr) < 0)
throw std::runtime_error ("couldn't open the codec");
}
codecOpened = true;
if (codecCtx->channels <= 0)
throw std::runtime_error ("no channels found");
std::shared_ptr<SwrContext> swr;
if (codecCtx->sample_fmt != AV_SAMPLE_FMT_S16)
{
swr.reset (swr_alloc (), [] (SwrContext *ctx) {
if (ctx) swr_free (&ctx);
});
av_opt_set_int (swr.get (), "in_channel_layout", codecCtx->channel_layout, 0);
av_opt_set_int (swr.get (), "out_channel_layout", codecCtx->channel_layout, 0);
av_opt_set_int (swr.get (), "in_sample_rate", codecCtx->sample_rate, 0);
av_opt_set_int (swr.get (), "out_sample_rate", codecCtx->sample_rate, 0);
av_opt_set_sample_fmt (swr.get (), "in_sample_fmt", codecCtx->sample_fmt, 0);
av_opt_set_sample_fmt (swr.get (), "out_sample_fmt", AV_SAMPLE_FMT_S16, 0);
swr_init (swr.get ());
}
AVPacket packet;
av_init_packet (&packet);
const int maxLength = 120;
auto remaining = maxLength * codecCtx->channels * codecCtx->sample_rate;
chromaprint_start (Ctx_, codecCtx->sample_rate, codecCtx->channels);
std::shared_ptr<AVFrame> frame (av_frame_alloc (),
[] (AVFrame *frame) {
av_frame_free (&frame);
});
auto maxDstNbSamples = 0;
uint8_t *dstData [1] = { nullptr };
std::shared_ptr<void> dstDataGuard (nullptr,
[&dstData] (void*) {
if (dstData [0]) av_freep (&dstData [0]);
});
while (true)
{
if (av_read_frame (formatCtx.get (), &packet) < 0)
break;
std::shared_ptr<void> guard (nullptr,
[&packet] (void*) {
if (packet.data) av_free_packet (&packet);
});
if (packet.stream_index != streamIndex)
continue;
av_frame_unref (frame.get ());
int gotFrame = false;
auto consumed = avcodec_decode_audio4 (codecCtx.get (), frame.get (), &gotFrame, &packet);
if (consumed < 0 || !gotFrame)
continue;
uint8_t **data = nullptr;
if (swr)
{
if (frame->nb_samples > maxDstNbSamples)
{
if (dstData [0])
av_freep (&dstData [0]);
int linesize = 0;
if (av_samples_alloc (dstData, &linesize, codecCtx->channels, frame->nb_samples, AV_SAMPLE_FMT_S16, 1) < 0)
throw std::runtime_error ("cannot allocate memory for resampling");
}
if (swr_convert (swr.get (), dstData, frame->nb_samples, const_cast<const uint8_t**> (frame->data), frame->nb_samples) < 0)
throw std::runtime_error ("cannot resample audio");
data = dstData;
}
else
data = frame->data;
auto length = std::min (remaining, frame->nb_samples * codecCtx->channels);
if (!chromaprint_feed (Ctx_, data [0], length))
throw std::runtime_error ("cannot feed data");
bool finished = false;
if (maxLength)
{
remaining -= length;
if (remaining <= 0)
finished = true;
}
if (finished)
break;
}
if (!chromaprint_finish (Ctx_))
throw std::runtime_error ("fingerprint calculation failed");
char *fingerprint = 0;
if (!chromaprint_get_fingerprint (Ctx_, &fingerprint))
throw std::runtime_error ("unable to get fingerprint");
QByteArray result (fingerprint);
chromaprint_dealloc (fingerprint);
const double divideFactor = 1. / av_q2d (stream->time_base);
const double duration = stream->duration / divideFactor;
return { result, static_cast<int> (duration) };
}
/**
* Decode a frame to a packet, run the result through SwrContext, if desired, encode it via an appropriate
* encoder, and write the results to the Java-side native buffer.
*
* @param aio FFAudio context
* @param cached true or false
* @return number of bytes placed into java buffer or a negative value, if something went wrong
*/
static int decode_packet(FFAudioIO *aio, int cached) {
int res = 0;
uint8_t **resample_buf = NULL;
jobject byte_buffer = NULL;
uint8_t *javaBuffer = NULL;
uint32_t out_buf_size = 0;
int out_buf_samples = 0;
int64_t out_channel_count;
int64_t out_sample_rate;
int flush = aio->got_frame;
enum AVSampleFormat out;
int bytesConsumed = 0;
init_ids(aio->env, aio->java_instance);
av_opt_get_int(aio->swr_context, "out_channel_count", 0, &out_channel_count);
av_opt_get_int(aio->swr_context, "out_sample_rate", 0, &out_sample_rate);
av_opt_get_sample_fmt(aio->swr_context, "out_sample_fmt", 0, &out);
resample_buf = av_mallocz(sizeof(uint8_t *) * 1); // one plane!
// make sure we really have an audio packet
if (aio->decode_packet.stream_index == aio->stream_index) {
// decode frame
// got_frame indicates whether we got a frame
bytesConsumed = avcodec_decode_audio4(aio->decode_context, aio->decode_frame, &aio->got_frame, &aio->decode_packet);
if (bytesConsumed < 0) {
throwUnsupportedAudioFileExceptionIfError(aio->env, bytesConsumed, "Failed to decode audio frame.");
return bytesConsumed;
}
if (aio->got_frame) {
aio->decoded_samples += aio->decode_frame->nb_samples;
out_buf_samples = aio->decode_frame->nb_samples;
#ifdef DEBUG
fprintf(stderr, "samples%s n:%" PRIu64 " nb_samples:%d pts:%s\n",
cached ? "(cached)" : "",
aio->decoded_samples, aio->decode_frame->nb_samples,
av_ts2timestr(aio->decode_frame->pts, &aio->decode_context->time_base));
#endif
// adjust out sample number for a different sample rate
// this is an estimate!!
out_buf_samples = av_rescale_rnd(
swr_get_delay(aio->swr_context, aio->stream->codecpar->sample_rate) + aio->decode_frame->nb_samples,
out_sample_rate,
aio->stream->codecpar->sample_rate,
AV_ROUND_UP
);
// allocate new aio->audio_data buffers
res = av_samples_alloc(aio->audio_data, NULL, av_frame_get_channels(aio->decode_frame),
aio->decode_frame->nb_samples, aio->decode_frame->format, 1);
if (res < 0) {
throwIOExceptionIfError(aio->env, res, "Could not allocate audio buffer.");
return AVERROR(ENOMEM);
}
// copy audio data to aio->audio_data
av_samples_copy(aio->audio_data, aio->decode_frame->data, 0, 0,
aio->decode_frame->nb_samples, av_frame_get_channels(aio->decode_frame), aio->decode_frame->format);
res = resample(aio, resample_buf, out_buf_samples, (const uint8_t **)aio->audio_data, aio->decode_frame->nb_samples);
if (res < 0) goto bail;
else out_buf_samples = res;
} else if (flush && swr_get_delay(aio->swr_context, aio->stream->codecpar->sample_rate)) {
res = resample(aio, resample_buf, swr_get_delay(aio->swr_context, aio->stream->codecpar->sample_rate), NULL, 0);
if (res < 0) goto bail;
else out_buf_samples = res;
} else {
#ifdef DEBUG
fprintf(stderr, "Got no frame.\n");
#endif
}
if (out_buf_samples > 0) {
res = av_samples_get_buffer_size(NULL, (int)out_channel_count, out_buf_samples, out, 1);
if (res < 0) goto bail;
else out_buf_size = res;
// ensure native buffer capacity
if (aio->java_buffer_capacity < out_buf_size) {
aio->java_buffer_capacity = (*aio->env)->CallIntMethod(aio->env, aio->java_instance, setNativeBufferCapacity_MID, (jint)out_buf_size);
}
// get java-managed byte buffer reference
byte_buffer = (*aio->env)->GetObjectField(aio->env, aio->java_instance, nativeBuffer_FID);
if (!byte_buffer) {
res = -1;
throwIOExceptionIfError(aio->env, 1, "Failed to get native buffer.");
goto bail;
}
// we have some samples, let's copy them to the java buffer, using the desired encoding
javaBuffer = (uint8_t *)(*aio->env)->GetDirectBufferAddress(aio->env, byte_buffer);
if (!javaBuffer) {
throwIOExceptionIfError(aio->env, 1, "Failed to get address for native buffer.");
goto bail;
}
if (aio->encode_context) {
aio->encode_frame->nb_samples = out_buf_samples;
res = encode_buffer(aio, resample_buf[0], out_buf_size, javaBuffer);
if (res < 0) {
out_buf_size = 0;
goto bail;
}
out_buf_size = res;
} else {
memcpy(javaBuffer, resample_buf[0], out_buf_size);
}
// we already wrote to the buffer, now we still need to
// set new bytebuffer limit and position to 0.
(*aio->env)->CallObjectMethod(aio->env, byte_buffer, rewind_MID);
(*aio->env)->CallObjectMethod(aio->env, byte_buffer, limit_MID, out_buf_size);
}
}
aio->resampled_samples += out_buf_size;
bail:
if (resample_buf) {
if (resample_buf[0]) av_freep(&resample_buf[0]);
av_free(resample_buf);
}
if (aio->audio_data[0]) av_freep(&aio->audio_data[0]);
return res;
}
static void write_audio_frame(AVFormatContext *oc, AVStream *st, int flush)
{
AVCodecContext *c;
AVPacket pkt = { 0 }; // data and size must be 0;
int got_packet, ret, dst_nb_samples;
av_init_packet(&pkt);
c = st->codec;
if (!flush) {
get_audio_frame((int16_t *)src_samples_data[0], src_nb_samples, c->channels);
/* convert samples from native format to destination codec format, using the resampler */
if (swr_ctx) {
/* compute destination number of samples */
dst_nb_samples = av_rescale_rnd(swr_get_delay(swr_ctx, c->sample_rate) + src_nb_samples,
c->sample_rate, c->sample_rate, AV_ROUND_UP);
if (dst_nb_samples > max_dst_nb_samples) {
av_free(dst_samples_data[0]);
ret = av_samples_alloc(dst_samples_data, &dst_samples_linesize, c->channels,
dst_nb_samples, c->sample_fmt, 0);
if (ret < 0)
exit(1);
max_dst_nb_samples = dst_nb_samples;
dst_samples_size = av_samples_get_buffer_size(NULL, c->channels, dst_nb_samples,
c->sample_fmt, 0);
}
/* convert to destination format */
ret = swr_convert(swr_ctx,
dst_samples_data, dst_nb_samples,
(const uint8_t **)src_samples_data, src_nb_samples);
if (ret < 0) {
fprintf(stderr, "Error while converting\n");
exit(1);
}
} else {
dst_nb_samples = src_nb_samples;
}
audio_frame->nb_samples = dst_nb_samples;
audio_frame->pts = av_rescale_q(samples_count, (AVRational){1, c->sample_rate}, c->time_base);
avcodec_fill_audio_frame(audio_frame, c->channels, c->sample_fmt,
dst_samples_data[0], dst_samples_size, 0);
samples_count += dst_nb_samples;
}
ret = avcodec_encode_audio2(c, &pkt, flush ? NULL : audio_frame, &got_packet);
if (ret < 0) {
fprintf(stderr, "Error encoding audio frame: %s\n", av_err2str(ret));
exit(1);
}
if (!got_packet) {
if (flush)
audio_is_eof = 1;
return;
}
ret = write_frame(oc, &c->time_base, st, &pkt);
if (ret < 0) {
fprintf(stderr, "Error while writing audio frame: %s\n",
av_err2str(ret));
exit(1);
}
}
