MP4Duration MP4AV_GetAudioSampleDuration(
MP4FileHandle mp4File,
MP4TrackId mediaTrackId)
{
MP4SampleId sampleId = 1;
MP4SampleId numSamples =
MP4GetTrackNumberOfSamples(mp4File, mediaTrackId);
// find first non-zero size sample
// we need to search in case an empty audio sample has been added
// at the beginning of the track to achieve sync with video
for (; sampleId <= numSamples; sampleId++) {
if (MP4GetSampleSize(mp4File, mediaTrackId, sampleId) > 0) {
break;
}
}
if (sampleId >= numSamples) {
return MP4_INVALID_DURATION;
}
// get sample duration
return MP4GetSampleDuration(mp4File, mediaTrackId, sampleId);
// OPTION may want to scan all non-zero sized samples
// and check that sample durations are +/-1 the same value
}
extern "C" bool MP4AV_Rfc3016Hinter(
MP4FileHandle mp4File,
MP4TrackId mediaTrackId,
u_int16_t maxPayloadSize)
{
u_int32_t numSamples = MP4GetTrackNumberOfSamples(mp4File, mediaTrackId);
u_int32_t maxSampleSize = MP4GetTrackMaxSampleSize(mp4File, mediaTrackId);
if (numSamples == 0 || maxSampleSize == 0) {
return false;
}
MP4TrackId hintTrackId =
MP4AV_Rfc3016_HintTrackCreate(mp4File, mediaTrackId);
if (hintTrackId == MP4_INVALID_TRACK_ID) {
return false;
}
u_int8_t* pSampleBuffer = (u_int8_t*)malloc(maxSampleSize);
if (pSampleBuffer == NULL) {
MP4DeleteTrack(mp4File, hintTrackId);
return false;
}
for (MP4SampleId sampleId = 1; sampleId <= numSamples; sampleId++) {
u_int32_t sampleSize = maxSampleSize;
MP4Timestamp startTime;
MP4Duration duration;
MP4Duration renderingOffset;
bool isSyncSample;
bool rc = MP4ReadSample(
mp4File, mediaTrackId, sampleId,
&pSampleBuffer, &sampleSize,
&startTime, &duration,
&renderingOffset, &isSyncSample);
if (rc == false ||
MP4AV_Rfc3016_HintAddSample(mp4File,
hintTrackId,
sampleId,
pSampleBuffer,
sampleSize,
duration,
renderingOffset,
isSyncSample,
maxPayloadSize) == false) {
MP4DeleteTrack(mp4File, hintTrackId);
CHECK_AND_FREE(pSampleBuffer);
return false;
}
}
CHECK_AND_FREE(pSampleBuffer);
return true;
}
MP4Reader(const std::string &file_path)
: time_scale(9 * MP4_MSECS_TIME_SCALE)
, file_path(file_path)
, handle(MP4_INVALID_FILE_HANDLE)
, video_track_id(MP4_INVALID_TRACK_ID)
, next_video_sample_idx(1)
, video_sample(nullptr)
, video_timescale(0)
, video_sample_max_size(0)
, video_sample_number(0)
, video_duration(0)
, pSeqHeaders(nullptr)
, pSeqHeaderSize(nullptr)
, pPictHeaders(nullptr)
, pPictHeaderSize(nullptr)
{
handle = MP4Read(this->file_path.c_str());
video_track_id = MP4FindTrackId(handle, 0, MP4_VIDEO_TRACK_TYPE);
if (video_track_id != MP4_INVALID_TRACK_ID) {
video_timescale = MP4GetTrackTimeScale(handle, video_track_id);
video_sample_max_size = MP4GetTrackMaxSampleSize(handle, video_track_id) * 2;
video_duration = MP4GetTrackDuration(handle, video_track_id);
video_sample = new unsigned char[video_sample_max_size];
video_sample_number = MP4GetTrackNumberOfSamples(handle, video_track_id);
if (MP4GetTrackH264SeqPictHeaders(handle,
video_track_id,
&pSeqHeaders,
&pSeqHeaderSize,
&pPictHeaders,
&pPictHeaderSize))
{
printf("Get SPS(%d) and PPS(%d)\n", *pSeqHeaderSize, *pPictHeaderSize);
for(int i = 0; (pSeqHeaders[i] && pSeqHeaderSize[i]); i++) {
printf("SPS(%d): %02x %02x %02x %02x %02x\n", i,
pSeqHeaders[i][0], pSeqHeaders[i][1], pSeqHeaders[i][2],
pSeqHeaders[i][3], pSeqHeaders[i][4]);
}
for(int i = 0; (pPictHeaders[i] && pPictHeaderSize[i]); i++) {
printf("PPS(%d): %02x %02x %02x %02x %02x\n", i,
pPictHeaders[i][0], pPictHeaders[i][1], pPictHeaders[i][2],
pPictHeaders[i][3], pPictHeaders[i][4]);
}
}
}
}
/**************************************************************************
* Quicktime stream base class functions
**************************************************************************/
CMp4ByteStream::CMp4ByteStream (CMp4File *parent,
MP4TrackId track,
const char *type,
bool has_video)
: COurInByteStream(type)
{
#ifdef ISMACRYP_DEBUG
my_enc_file = fopen("encbuffer.raw", "w");
my_unenc_file = fopen("unencbuffer.raw", "w");
my_unenc_file2 = fopen("unencbuffer2.raw", "w");
#endif
#ifdef OUTPUT_TO_FILE
char buffer[80];
strcpy(buffer, type);
strcat(buffer, ".raw");
m_output_file = fopen(buffer, "w");
#endif
m_track = track;
m_frame_on = 1;
m_parent = parent;
m_eof = false;
MP4FileHandle fh = parent->get_file();
m_frames_max = MP4GetTrackNumberOfSamples(fh, m_track);
mp4f_message(LOG_DEBUG, "%s - %u samples", type, m_frames_max);
m_max_frame_size = MP4GetTrackMaxSampleSize(fh, m_track) + 4;
m_sample_freq = MP4GetTrackTimeScale(fh, m_track);
m_buffer = (u_int8_t *) malloc(m_max_frame_size * sizeof(u_int8_t));
m_has_video = has_video;
m_frame_in_buffer = 0xffffffff;
MP4Duration trackDuration;
trackDuration = MP4GetTrackDuration(fh, m_track);
uint64_t max_ts;
max_ts = MP4ConvertFromTrackDuration(fh,
m_track,
trackDuration,
MP4_MSECS_TIME_SCALE);
m_max_time = UINT64_TO_DOUBLE(max_ts);
m_max_time /= 1000.0;
mp4f_message(LOG_DEBUG,
"MP4 %s max time is "U64" %g", type, max_ts, m_max_time);
read_frame(1, NULL);
}
static int
aac_init (DB_fileinfo_t *_info, DB_playItem_t *it) {
aac_info_t *info = (aac_info_t *)_info;
info->file = deadbeef->fopen (deadbeef->pl_find_meta (it, ":URI"));
if (!info->file) {
return -1;
}
// probe
float duration = -1;
int samplerate = -1;
int channels = -1;
int totalsamples = -1;
info->junk = deadbeef->junk_get_leading_size (info->file);
if (!info->file->vfs->is_streaming ()) {
if (info->junk >= 0) {
deadbeef->fseek (info->file, info->junk, SEEK_SET);
}
else {
info->junk = 0;
}
}
else {
deadbeef->fset_track (info->file, it);
}
info->mp4track = -1;
#if USE_MP4FF
info->mp4reader.read = aac_fs_read;
info->mp4reader.write = NULL;
info->mp4reader.seek = aac_fs_seek;
info->mp4reader.truncate = NULL;
info->mp4reader.user_data = info;
#else
info->mp4reader.open = aac_fs_open;
info->mp4reader.seek = aac_fs_seek;
info->mp4reader.read = aac_fs_read;
info->mp4reader.write = NULL;
info->mp4reader.close = aac_fs_close;
#endif
if (!info->file->vfs->is_streaming ()) {
#ifdef USE_MP4FF
trace ("aac_init: mp4ff_open_read %s\n", deadbeef->pl_find_meta (it, ":URI"));
info->mp4file = mp4ff_open_read (&info->mp4reader);
if (info->mp4file) {
int ntracks = mp4ff_total_tracks (info->mp4file);
if (ntracks > 0) {
trace ("m4a container detected, ntracks=%d\n", ntracks);
int i = -1;
unsigned char* buff = 0;
unsigned int buff_size = 0;
for (i = 0; i < ntracks; i++) {
mp4AudioSpecificConfig mp4ASC;
mp4ff_get_decoder_config (info->mp4file, i, &buff, &buff_size);
if(buff){
int rc = AudioSpecificConfig(buff, buff_size, &mp4ASC);
if(rc < 0)
continue;
break;
}
}
trace ("mp4 probe-buffer size: %d\n", buff_size);
if (i != ntracks && buff)
{
trace ("mp4 track: %d\n", i);
int samples = mp4ff_num_samples(info->mp4file, i);
info->mp4samples = samples;
info->mp4track = i;
// init mp4 decoding
info->dec = NeAACDecOpen ();
unsigned long srate;
unsigned char ch;
if (NeAACDecInit2(info->dec, buff, buff_size, &srate, &ch) < 0) {
trace ("NeAACDecInit2 returned error\n");
free (buff);
return -1;
}
samplerate = srate;
channels = ch;
samples = (int64_t)samples * srate / mp4ff_time_scale (info->mp4file, i);
totalsamples = samples;
NeAACDecConfigurationPtr conf = NeAACDecGetCurrentConfiguration (info->dec);
conf->dontUpSampleImplicitSBR = 1;
NeAACDecSetConfiguration (info->dec, conf);
mp4AudioSpecificConfig mp4ASC;
if (NeAACDecAudioSpecificConfig(buff, buff_size, &mp4ASC) >= 0)
{
info->mp4framesize = 1024;
if (mp4ASC.frameLengthFlag == 1) {
info->mp4framesize = 960;
}
// if (mp4ASC.sbr_present_flag == 1) {
// info->mp4framesize *= 2;
// }
}
totalsamples *= info->mp4framesize;
duration = (float)totalsamples / samplerate;
}
else {
mp4ff_close (info->mp4file);
info->mp4file = NULL;
}
if (buff) {
free (buff);
}
}
else {
mp4ff_close (info->mp4file);
info->mp4file = NULL;
}
}
// {{{ libmp4v2 code
#else
trace ("aac_init: MP4ReadProvider %s\n", deadbeef->pl_find_meta (it, ":URI"));
info->mp4file = MP4ReadProvider (deadbeef->pl_find_meta (it, ":URI"), 0, &info->mp4reader);
info->mp4track = MP4FindTrackId(info->mp4file, 0, "audio", 0);
trace ("aac_init: MP4FindTrackId returned %d\n", info->mp4track);
if (info->mp4track >= 0) {
info->timescale = MP4GetTrackTimeScale(info->mp4file, info->mp4track);
u_int8_t* pConfig;
uint32_t configSize = 0;
bool res = MP4GetTrackESConfiguration(info->mp4file, info->mp4track, &pConfig, &configSize);
mp4AudioSpecificConfig mp4ASC;
int rc = AudioSpecificConfig(pConfig, configSize, &mp4ASC);
if (rc >= 0) {
_info->samplerate = mp4ASC.samplingFrequency;
_info->channels = MP4GetTrackAudioChannels (info->mp4file, info->mp4track);
totalsamples = MP4GetTrackNumberOfSamples (info->mp4file, info->mp4track) * 1024 * _info->channels;
// init mp4 decoding
info->dec = NeAACDecOpen ();
unsigned long srate;
unsigned char ch;
if (NeAACDecInit2(info->dec, pConfig, configSize, &srate, &ch) < 0) {
trace ("NeAACDecInit2 returned error\n");
return -1;
}
samplerate = srate;
channels = ch;
NeAACDecConfigurationPtr conf = NeAACDecGetCurrentConfiguration (info->dec);
conf->dontUpSampleImplicitSBR = 1;
NeAACDecSetConfiguration (info->dec, conf);
mp4AudioSpecificConfig mp4ASC;
if (NeAACDecAudioSpecificConfig(pConfig, configSize, &mp4ASC) >= 0)
{
info->mp4framesize = 1024;
if (mp4ASC.frameLengthFlag == 1) {
info->mp4framesize = 960;
}
// if (mp4ASC.sbr_present_flag == 1) {
// info->mp4framesize *= 2;
// }
}
//totalsamples *= info->mp4framesize;
free (pConfig);
info->maxSampleSize = MP4GetTrackMaxSampleSize(info->mp4file, info->mp4track);
info->samplebuffer = malloc (info->maxSampleSize);
info->mp4sample = 1;
}
else {
MP4Close (info->mp4file);
info->mp4file = NULL;
}
}
else {
MP4Close (info->mp4file);
info->mp4file = NULL;
}
#endif
// }}}
if (!info->mp4file) {
trace ("mp4 track not found, looking for aac stream...\n");
if (info->junk >= 0) {
deadbeef->fseek (info->file, info->junk, SEEK_SET);
}
else {
deadbeef->rewind (info->file);
}
int offs = parse_aac_stream (info->file, &samplerate, &channels, &duration, &totalsamples);
if (offs == -1) {
trace ("aac stream not found\n");
return -1;
}
if (offs > info->junk) {
info->junk = offs;
}
if (info->junk >= 0) {
deadbeef->fseek (info->file, info->junk, SEEK_SET);
}
else {
deadbeef->rewind (info->file);
}
trace ("found aac stream (junk: %d, offs: %d)\n", info->junk, offs);
}
_info->fmt.channels = channels;
_info->fmt.samplerate = samplerate;
}
else {
// sync before attempting to init
int samplerate, channels;
float duration;
int offs = parse_aac_stream (info->file, &samplerate, &channels, &duration, NULL);
if (offs < 0) {
trace ("aac: parse_aac_stream failed\n");
return -1;
}
if (offs > info->junk) {
info->junk = offs;
}
trace("parse_aac_stream returned %x\n", offs);
deadbeef->pl_replace_meta (it, "!FILETYPE", "AAC");
}
// duration = (float)totalsamples / samplerate;
// deadbeef->pl_set_item_duration (it, duration);
_info->fmt.bps = 16;
_info->plugin = &plugin;
if (!info->mp4file) {
//trace ("NeAACDecGetCapabilities\n");
//unsigned long caps = NeAACDecGetCapabilities();
trace ("NeAACDecOpen\n");
info->dec = NeAACDecOpen ();
trace ("prepare for NeAACDecInit: fread %d from offs %lld\n", AAC_BUFFER_SIZE, deadbeef->ftell (info->file));
info->remaining = deadbeef->fread (info->buffer, 1, AAC_BUFFER_SIZE, info->file);
NeAACDecConfigurationPtr conf = NeAACDecGetCurrentConfiguration (info->dec);
// conf->dontUpSampleImplicitSBR = 1;
NeAACDecSetConfiguration (info->dec, conf);
unsigned long srate;
unsigned char ch;
trace ("NeAACDecInit (%d bytes)\n", info->remaining);
int consumed = NeAACDecInit (info->dec, info->buffer, info->remaining, &srate, &ch);
trace ("NeAACDecInit returned samplerate=%d, channels=%d, consumed: %d\n", (int)srate, (int)ch, consumed);
if (consumed < 0) {
trace ("NeAACDecInit returned %d\n", consumed);
return -1;
}
if (consumed > info->remaining) {
trace ("NeAACDecInit consumed more than available! wtf?\n");
return -1;
}
if (consumed == info->remaining) {
info->remaining = 0;
}
else if (consumed > 0) {
memmove (info->buffer, info->buffer + consumed, info->remaining - consumed);
info->remaining -= consumed;
}
_info->fmt.channels = ch;
_info->fmt.samplerate = srate;
}
if (!info->file->vfs->is_streaming ()) {
if (it->endsample > 0) {
info->startsample = it->startsample;
info->endsample = it->endsample;
plugin.seek_sample (_info, 0);
}
else {
info->startsample = 0;
info->endsample = totalsamples-1;
}
}
trace ("totalsamples: %d, endsample: %d, samples-from-duration: %d\n", totalsamples-1, info->endsample, (int)deadbeef->pl_get_item_duration (it)*44100);
for (int i = 0; i < _info->fmt.channels; i++) {
_info->fmt.channelmask |= 1 << i;
}
info->noremap = 0;
info->remap[0] = -1;
trace ("init success\n");
return 0;
}
// returns -1 for error, 0 for mp4, 1 for aac
int
aac_probe (DB_FILE *fp, const char *fname, MP4FILE_CB *cb, float *duration, int *samplerate, int *channels, int *totalsamples, int *mp4track, MP4FILE *pmp4) {
// try mp4
trace ("aac_probe: pos=%lld, junk=%d\n", deadbeef->ftell (fp), ((aac_info_t*)cb->user_data)->junk);
if (mp4track) {
*mp4track = -1;
}
if (*pmp4) {
*pmp4 = NULL;
}
*duration = -1;
#ifdef USE_MP4FF
trace ("mp4ff_open_read\n");
mp4ff_t *mp4 = mp4ff_open_read (cb);
#else
MP4FileHandle mp4 = MP4ReadProvider (fname, 0, cb);
#endif
if (!mp4) {
trace ("not an mp4 file\n");
return -1;
}
if (pmp4) {
*pmp4 = mp4;
}
#ifdef USE_MP4FF
int ntracks = mp4ff_total_tracks (mp4);
if (ntracks > 0) {
trace ("m4a container detected, ntracks=%d\n", ntracks);
int i = -1;
trace ("looking for mp4 data...\n");
int sr = -1;
unsigned char* buff = 0;
unsigned int buff_size = 0;
for (i = 0; i < ntracks; i++) {
mp4AudioSpecificConfig mp4ASC;
mp4ff_get_decoder_config(mp4, i, &buff, &buff_size);
if (buff) {
int rc = AudioSpecificConfig(buff, buff_size, &mp4ASC);
sr = mp4ASC.samplingFrequency;
if(rc < 0) {
free (buff);
buff = 0;
continue;
}
break;
}
}
if (i != ntracks && buff)
{
trace ("found audio track (%d)\n", i);
// init mp4 decoding
NeAACDecHandle dec = NeAACDecOpen ();
unsigned long srate;
unsigned char ch;
if (NeAACDecInit2(dec, buff, buff_size, &srate, &ch) < 0) {
trace ("NeAACDecInit2 returned error\n");
goto error;
}
*samplerate = srate;
*channels = ch;
int samples = mp4ff_num_samples(mp4, i);
samples = (int64_t)samples * srate / mp4ff_time_scale (mp4, i);
int tsamples = samples;
NeAACDecConfigurationPtr conf = NeAACDecGetCurrentConfiguration (dec);
conf->dontUpSampleImplicitSBR = 1;
NeAACDecSetConfiguration (dec, conf);
mp4AudioSpecificConfig mp4ASC;
int mp4framesize;
if (NeAACDecAudioSpecificConfig(buff, buff_size, &mp4ASC) >= 0)
{
mp4framesize = 1024;
if (mp4ASC.frameLengthFlag == 1) {
mp4framesize = 960;
}
// commented this out, since it fixes double-duration bug on
// some mp4 files
//if (mp4ASC.sbr_present_flag == 1) {
// mp4framesize *= 2;
//}
}
else {
trace ("NeAACDecAudioSpecificConfig failed, can't get mp4framesize\n");
goto error;
}
tsamples *= mp4framesize;
trace ("mp4 nsamples=%d, samplerate=%d, timescale=%d, duration=%lld\n", samples, *samplerate, mp4ff_time_scale(mp4, i), mp4ff_get_track_duration(mp4, i));
*duration = (float)tsamples / (*samplerate);
trace ("mp4 duration: %f (tsamples %d/samplerate %d)\n", *duration, tsamples, *samplerate);
NeAACDecClose (dec);
if (totalsamples) {
*totalsamples = tsamples;
}
if (mp4track) {
*mp4track = i;
}
if (!*pmp4) {
mp4ff_close (mp4);
}
return 0;
error:
NeAACDecClose (dec);
free (buff);
if (!*pmp4) {
mp4ff_close (mp4);
}
return -1;
}
else {
trace ("audio track not found\n");
mp4ff_close (mp4);
mp4 = NULL;
}
if (buff) {
free (buff);
buff = NULL;
}
}
#else
MP4FileHandle mp4File = mp4;
MP4TrackId trackId = MP4FindTrackId(mp4File, 0, "audio", 0);
trace ("trackid: %d\n", trackId);
uint32_t timeScale = MP4GetTrackTimeScale(mp4File, trackId);
MP4Duration trackDuration = MP4GetTrackDuration(mp4File, trackId);
MP4SampleId numSamples = MP4GetTrackNumberOfSamples(mp4File, trackId);
u_int8_t* pConfig;
uint32_t configSize = 0;
bool res = MP4GetTrackESConfiguration(mp4File, trackId, &pConfig, &configSize);
if (res && pConfig) {
mp4AudioSpecificConfig mp4ASC;
int rc = AudioSpecificConfig(pConfig, configSize, &mp4ASC);
free (pConfig);
if (rc >= 0) {
*samplerate = mp4ASC.samplingFrequency;
*channels = MP4GetTrackAudioChannels (mp4File, trackId);
// int64_t duration = MP4ConvertFromTrackDuration (mp4File, trackId, trackDuration, timeScale);
int samples = MP4GetTrackNumberOfSamples (mp4File, trackId) * 1024 * (*channels);
trace ("mp4 nsamples=%d, timescale=%d, samplerate=%d\n", samples, timeScale, *samplerate);
*duration = (float)samples / (*samplerate);
if (totalsamples) {
*totalsamples = samples;
}
if (mp4track) {
*mp4track = trackId;
}
if (!*pmp4) {
MP4Close (mp4);
}
return 0;
}
}
#endif
if (*pmp4) {
*pmp4 = NULL;
}
if (mp4) {
#if USE_MP4FF
mp4ff_close (mp4);
#else
MP4Close (mp4);
#endif
mp4 = NULL;
}
trace ("mp4 track not found, looking for aac stream...\n");
// not an mp4, try raw aac
#if USE_MP4FF
deadbeef->rewind (fp);
#endif
if (parse_aac_stream (fp, samplerate, channels, duration, totalsamples) == -1) {
trace ("aac stream not found\n");
return -1;
}
trace ("found aac stream\n");
return 1;
}
void ExtractTrack(MP4FileHandle mp4File, MP4TrackId trackId,
bool sampleMode, MP4SampleId sampleId, char* dstFileName)
{
char outFileName[PATH_MAX];
int outFd = -1;
int openFlags = O_WRONLY | O_TRUNC | OPEN_CREAT;
if (!sampleMode) {
if (dstFileName == NULL) {
snprintf(outFileName, sizeof(outFileName),
"%s.t%u", Mp4FileName, trackId);
} else {
snprintf(outFileName, sizeof(outFileName),
"%s", dstFileName);
}
outFd = open(outFileName, openFlags, 0644);
if (outFd == -1) {
fprintf(stderr, "%s: can't open %s: %s\n",
ProgName, outFileName, strerror(errno));
return;
}
}
MP4SampleId numSamples;
if (sampleMode && sampleId != MP4_INVALID_SAMPLE_ID) {
numSamples = sampleId;
} else {
sampleId = 1;
numSamples = MP4GetTrackNumberOfSamples(mp4File, trackId);
}
u_int8_t* pSample;
u_int32_t sampleSize;
for ( ; sampleId <= numSamples; sampleId++) {
int rc;
// signals to ReadSample() that it should malloc a buffer for us
pSample = NULL;
sampleSize = 0;
rc = MP4ReadSample(mp4File, trackId, sampleId, &pSample, &sampleSize);
if (rc == 0) {
fprintf(stderr, "%s: read sample %u for %s failed\n",
ProgName, sampleId, outFileName);
break;
}
if (sampleMode) {
snprintf(outFileName, sizeof(outFileName), "%s.t%u.s%u",
Mp4FileName, trackId, sampleId);
outFd = open(outFileName, openFlags, 0644);
if (outFd == -1) {
fprintf(stderr, "%s: can't open %s: %s\n",
ProgName, outFileName, strerror(errno));
break;
}
}
rc = write(outFd, pSample, sampleSize);
if (rc == -1 || (u_int32_t)rc != sampleSize) {
fprintf(stderr, "%s: write to %s failed: %s\n",
ProgName, outFileName, strerror(errno));
break;
}
free(pSample);
if (sampleMode) {
close(outFd);
outFd = -1;
}
}
if (outFd != -1) {
close(outFd);
}
}
MP4TrackId Mp4vCreator(MP4FileHandle mp4File, FILE* inFile, bool doEncrypt,
bool allowVariableFrameRate)
{
bool rc;
u_int8_t sampleBuffer[256 * 1024 * 2];
u_int8_t* pCurrentSample = sampleBuffer;
u_int32_t maxSampleSize = sizeof(sampleBuffer) / 2;
u_int32_t prevSampleSize = 0;
// the current syntactical object
// typically 1:1 with a sample
// but not always, i.e. non-VOP's
u_int8_t* pObj = pCurrentSample;
u_int32_t objSize;
u_int8_t objType;
// the current sample
MP4SampleId sampleId = 1;
MP4Timestamp currentSampleTime = 0;
// the last reference VOP
MP4SampleId refVopId = 1;
MP4Timestamp refVopTime = 0;
// track configuration info
u_int8_t videoProfileLevel = MPEG4_SP_L3;
u_int8_t timeBits = 15;
u_int16_t timeTicks = 30000;
u_int16_t frameDuration = 3000;
u_int16_t frameWidth = 320;
u_int16_t frameHeight = 240;
u_int32_t esConfigSize = 0;
int vopType = 0;
int prevVopType = 0;
bool foundVOSH = false, foundVO = false, foundVOL = false;
u_int32_t lastVopTimeIncrement = 0;
bool variableFrameRate = false;
bool lastFrame = false;
bool haveBframes = false;
mpeg4_frame_t *head = NULL, *tail = NULL;
// start reading objects until we get the first VOP
while (LoadNextObject(inFile, pObj, &objSize, &objType)) {
// guard against buffer overflow
if (pObj + objSize >= pCurrentSample + maxSampleSize) {
fprintf(stderr,
"%s: buffer overflow, invalid video stream?\n", ProgName);
return MP4_INVALID_TRACK_ID;
}
#ifdef DEBUG_MP4V
if (Verbosity & MP4_DETAILS_SAMPLE) {
printf("MP4V type %x size %u\n",
objType, objSize);
}
#endif
if (objType == MP4AV_MPEG4_VOSH_START) {
MP4AV_Mpeg4ParseVosh(pObj, objSize,
&videoProfileLevel);
foundVOSH = true;
} else if (objType == MP4AV_MPEG4_VO_START) {
foundVO = true;
} else if (objType == MP4AV_MPEG4_VOL_START) {
MP4AV_Mpeg4ParseVol(pObj, objSize,
&timeBits, &timeTicks, &frameDuration,
&frameWidth, &frameHeight);
foundVOL = true;
#ifdef DEBUG_MP4V
printf("ParseVol: timeBits %u timeTicks %u frameDuration %u\n",
timeBits, timeTicks, frameDuration);
#endif
} else if (foundVOL == true || objType == MP4AV_MPEG4_VOP_START) {
esConfigSize = pObj - pCurrentSample;
// ready to set up mp4 track
break;
}
/* XXX why do we need this if ?
* It looks like it will remove this object ... XXX */
// It does. On Purpose. wmay 6/2004
if (objType != MP4AV_MPEG4_USER_DATA_START) {
pObj += objSize;
}
}
if (foundVOSH == false) {
fprintf(stderr,
"%s: no VOSH header found in MPEG-4 video.\n"
"This can cause problems with players other than mp4player. \n",
ProgName);
} else {
if (VideoProfileLevelSpecified &&
videoProfileLevel != VideoProfileLevel) {
fprintf(stderr,
"%s: You have specified a different video profile level than was detected in the VOSH header\n"
"The level you specified was %d and %d was read from the VOSH\n",
ProgName, VideoProfileLevel, videoProfileLevel);
}
}
if (foundVO == false) {
fprintf(stderr,
"%s: No VO header found in mpeg-4 video.\n"
"This can cause problems with players other than mp4player\n",
ProgName);
}
if (foundVOL == false) {
fprintf(stderr,
"%s: fatal: No VOL header found in mpeg-4 video stream\n",
ProgName);
return MP4_INVALID_TRACK_ID;
}
// convert frame duration to canonical time scale
// note zero value for frame duration signals variable rate video
if (timeTicks == 0) {
timeTicks = 1;
}
u_int32_t mp4FrameDuration = 0;
if (VideoFrameRate) {
mp4FrameDuration = (u_int32_t)(((double)Mp4TimeScale) / VideoFrameRate);
} else if (frameDuration) {
VideoFrameRate = frameDuration;
VideoFrameRate /= timeTicks;
mp4FrameDuration = (Mp4TimeScale * frameDuration) / timeTicks;
} else {
if (allowVariableFrameRate == false ) {
fprintf(stderr,
"%s: variable rate video stream signalled,"
" please specify average frame rate with -r option\n"
" or --variable-frame-rate argument\n",
ProgName);
return MP4_INVALID_TRACK_ID;
}
variableFrameRate = true;
}
ismacryp_session_id_t ismaCrypSId;
mp4v2_ismacrypParams *icPp = (mp4v2_ismacrypParams *) malloc(sizeof(mp4v2_ismacrypParams));
memset(icPp, 0, sizeof(mp4v2_ismacrypParams));
// initialize ismacryp session if encrypting
if (doEncrypt) {
if (ismacrypInitSession(&ismaCrypSId,KeyTypeVideo) != 0) {
fprintf(stderr, "%s: could not initialize the ISMAcryp session\n",
ProgName);
return MP4_INVALID_TRACK_ID;
}
if (ismacrypGetScheme(ismaCrypSId, &(icPp->scheme_type)) != ismacryp_rc_ok) {
fprintf(stderr, "%s: could not get ismacryp scheme type. sid %d\n",
ProgName, ismaCrypSId);
ismacrypEndSession(ismaCrypSId);
return MP4_INVALID_TRACK_ID;
}
if (ismacrypGetSchemeVersion(ismaCrypSId, &(icPp->scheme_version)) != ismacryp_rc_ok) {
fprintf(stderr, "%s: could not get ismacryp scheme ver. sid %d\n",
ProgName, ismaCrypSId);
ismacrypEndSession(ismaCrypSId);
return MP4_INVALID_TRACK_ID;
}
if (ismacrypGetKMSUri(ismaCrypSId, &(icPp->kms_uri)) != ismacryp_rc_ok) {
fprintf(stderr, "%s: could not get ismacryp kms uri. sid %d\n",
ProgName, ismaCrypSId);
CHECK_AND_FREE(icPp->kms_uri);
ismacrypEndSession(ismaCrypSId);
return MP4_INVALID_TRACK_ID;
}
if ( ismacrypGetSelectiveEncryption(ismaCrypSId, &(icPp->selective_enc)) != ismacryp_rc_ok ) {
fprintf(stderr, "%s: could not get ismacryp selec enc. sid %d\n",
ProgName, ismaCrypSId);
ismacrypEndSession(ismaCrypSId);
return MP4_INVALID_TRACK_ID;
}
if (ismacrypGetKeyIndicatorLength(ismaCrypSId, &(icPp->key_ind_len)) != ismacryp_rc_ok) {
fprintf(stderr, "%s: could not get ismacryp key ind len. sid %d\n",
ProgName, ismaCrypSId);
ismacrypEndSession(ismaCrypSId);
return MP4_INVALID_TRACK_ID;
}
if (ismacrypGetIVLength(ismaCrypSId, &(icPp->iv_len)) != ismacryp_rc_ok) {
fprintf(stderr, "%s: could not get ismacryp iv len. sid %d\n",
ProgName, ismaCrypSId);
ismacrypEndSession(ismaCrypSId);
return MP4_INVALID_TRACK_ID;
}
}
// create the new video track
MP4TrackId trackId;
if (doEncrypt) {
trackId =
MP4AddEncVideoTrack(
mp4File,
Mp4TimeScale,
mp4FrameDuration,
frameWidth,
frameHeight,
icPp,
MP4_MPEG4_VIDEO_TYPE);
} else {
trackId =
MP4AddVideoTrack(
mp4File,
Mp4TimeScale,
mp4FrameDuration,
frameWidth,
frameHeight,
MP4_MPEG4_VIDEO_TYPE);
}
if (trackId == MP4_INVALID_TRACK_ID) {
fprintf(stderr,
"%s: can't create video track\n", ProgName);
return MP4_INVALID_TRACK_ID;
}
if (VideoProfileLevelSpecified) {
videoProfileLevel = VideoProfileLevel;
}
if (MP4GetNumberOfTracks(mp4File, MP4_VIDEO_TRACK_TYPE) == 1) {
MP4SetVideoProfileLevel(mp4File, videoProfileLevel);
}
printf("es config size is %d\n", esConfigSize);
if (esConfigSize) {
MP4SetTrackESConfiguration(mp4File, trackId,
pCurrentSample, esConfigSize);
// move past ES config, so it doesn't go into first sample
pCurrentSample += esConfigSize;
}
// Move the current frame to the beginning of the
// buffer
memmove(sampleBuffer, pCurrentSample, pObj - pCurrentSample + objSize);
pObj = sampleBuffer + (pObj - pCurrentSample);
pCurrentSample = sampleBuffer;
MP4Timestamp prevFrameTimestamp = 0;
// now process the rest of the video stream
while ( true ) {
if ( objType != MP4AV_MPEG4_VOP_START ) {
// keep it in the buffer until a VOP comes along
// Actually, do nothings, since we only want VOP
// headers in the stream - wmay 6/2004
//pObj += objSize;
} else { // we have VOP
u_int32_t sampleSize = (pObj + objSize) - pCurrentSample;
vopType = MP4AV_Mpeg4GetVopType(pObj, objSize);
mpeg4_frame_t *fr = MALLOC_STRUCTURE(mpeg4_frame_t);
if (head == NULL) {
head = tail = fr;
} else {
tail->next = fr;
tail = fr;
}
fr->vopType = vopType;
fr->frameTimestamp = currentSampleTime;
fr->next = NULL;
if ( variableFrameRate ) {
// variable frame rate: recalculate "mp4FrameDuration"
if ( lastFrame ) {
// last frame
mp4FrameDuration = Mp4TimeScale / timeTicks;
} else {
// not the last frame
u_int32_t vopTimeIncrement;
MP4AV_Mpeg4ParseVop(pObj, objSize, &vopType, timeBits, timeTicks, &vopTimeIncrement);
u_int32_t vopTime = vopTimeIncrement - lastVopTimeIncrement;
mp4FrameDuration = (Mp4TimeScale * vopTime) / timeTicks;
lastVopTimeIncrement = vopTimeIncrement % timeTicks;
}
}
if ( prevSampleSize > 0 ) { // not the first time
// fill sample data & length to write
u_int8_t* sampleData2Write = NULL;
u_int32_t sampleLen2Write = 0;
if ( doEncrypt ) {
if ( ismacrypEncryptSampleAddHeader(ismaCrypSId,
sampleSize,
sampleBuffer,
&sampleLen2Write,
&sampleData2Write) != 0 ) {
fprintf(stderr,
"%s: can't encrypt video sample and add header %u\n",
ProgName, sampleId);
}
} else {
sampleData2Write = sampleBuffer;
sampleLen2Write = prevSampleSize;
}
if (variableFrameRate == false) {
double now_calc;
now_calc = sampleId;
now_calc *= Mp4TimeScale;
now_calc /= VideoFrameRate;
MP4Timestamp now_ts = (MP4Timestamp)now_calc;
mp4FrameDuration = now_ts - prevFrameTimestamp;
prevFrameTimestamp = now_ts;
currentSampleTime = now_ts;
}
// Write the previous sample
rc = MP4WriteSample(mp4File, trackId,
sampleData2Write, sampleLen2Write,
mp4FrameDuration, 0, prevVopType == VOP_TYPE_I);
if ( doEncrypt && sampleData2Write ) {
// buffer allocated by encrypt function.
// must free it!
free(sampleData2Write);
}
if ( !rc ) {
fprintf(stderr,
"%s: can't write video frame %u\n",
ProgName, sampleId);
MP4DeleteTrack(mp4File, trackId);
return MP4_INVALID_TRACK_ID;
}
// deal with rendering time offsets
// that can occur when B frames are being used
// which is the case for all profiles except Simple Profile
haveBframes |= (prevVopType == VOP_TYPE_B);
if ( lastFrame ) {
// finish read frames
break;
}
sampleId++;
} // not the first time
currentSampleTime += mp4FrameDuration;
// Move the current frame to the beginning of the
// buffer
memmove(sampleBuffer, pCurrentSample, sampleSize);
prevSampleSize = sampleSize;
prevVopType = vopType;
// reset pointers
pObj = pCurrentSample = sampleBuffer + sampleSize;
} // we have VOP
// load next object from bitstream
if (!LoadNextObject(inFile, pObj, &objSize, &objType)) {
if (objType != MP4AV_MPEG4_VOP_START)
break;
lastFrame = true;
objSize = 0;
continue;
}
// guard against buffer overflow
if (pObj + objSize >= pCurrentSample + maxSampleSize) {
fprintf(stderr,
"%s: buffer overflow, invalid video stream?\n", ProgName);
MP4DeleteTrack(mp4File, trackId);
return MP4_INVALID_TRACK_ID;
}
#ifdef DEBUG_MP4V
if (Verbosity & MP4_DETAILS_SAMPLE) {
printf("MP4V type %x size %u\n",
objType, objSize);
}
#endif
}
bool doRenderingOffset = false;
switch (videoProfileLevel) {
case MPEG4_SP_L0:
case MPEG4_SP_L1:
case MPEG4_SP_L2:
case MPEG4_SP_L3:
break;
default:
doRenderingOffset = true;
break;
}
if (doRenderingOffset && haveBframes) {
// only generate ctts (with rendering offset for I, P frames) when
// we need one. We saved all the frames types and timestamps above -
// we can't use MP4ReadSample, because the end frames might not have
// been written
refVopId = 1;
refVopTime = 0;
MP4SampleId maxSamples = MP4GetTrackNumberOfSamples(mp4File, trackId);
// start with sample 2 - we know the first one is a I frame
mpeg4_frame_t *fr = head->next; // skip the first one
for (MP4SampleId ix = 2; ix <= maxSamples; ix++) {
if (fr->vopType != VOP_TYPE_B) {
#ifdef DEBUG_MP4V_TS
printf("sample %u %u renderingOffset "U64"\n",
refVopId, fr->vopType, fr->frameTimestamp - refVopTime);
#endif
MP4SetSampleRenderingOffset(mp4File, trackId, refVopId,
fr->frameTimestamp - refVopTime);
refVopId = ix;
refVopTime = fr->frameTimestamp;
}
fr = fr->next;
}
#ifdef DEBUG_MP4V_TS
printf("sample %u %u renderingOffset "U64"\n",
refVopId, fr->vopType, fr->frameTimestamp - refVopTime);
#endif
MP4SetSampleRenderingOffset(mp4File, trackId, refVopId,
fr->frameTimestamp - refVopTime);
}
while (head != NULL) {
tail = head->next;
free(head);
head = tail;
}
// terminate session if encrypting
if (doEncrypt) {
if (ismacrypEndSession(ismaCrypSId) != 0) {
fprintf(stderr,
"%s: could not end the ISMAcryp session\n",
ProgName);
}
}
return trackId;
}
extern "C" bool MP4AV_AVSMHinter(
MP4FileHandle mp4File,
MP4TrackId mediaTrackId,
u_int16_t maxPayloadSize)
{
u_int32_t numSamples = MP4GetTrackNumberOfSamples(mp4File, mediaTrackId);
u_int32_t maxSampleSize = MP4GetTrackMaxSampleSize(mp4File, mediaTrackId);
uint32_t sizeLength;
if (numSamples == 0 || maxSampleSize == 0) {
return false;
}
/*if (MP4GetTrackAVSMLengthSize(mp4File, mediaTrackId, &sizeLength) == false) {
return false;
}*/
sizeLength=4; //why?
MP4TrackId hintTrackId =
MP4AV_AVSM_HintTrackCreate(mp4File, mediaTrackId); //****AVSMspecial****
if (hintTrackId == MP4_INVALID_TRACK_ID) {
return false;
}
u_int8_t* pSampleBuffer = (u_int8_t*)malloc(maxSampleSize);
if (pSampleBuffer == NULL) {
MP4DeleteTrack(mp4File, hintTrackId);
return false;
}
for (MP4SampleId sampleId = 1; sampleId <= numSamples; sampleId++) {
u_int32_t sampleSize = maxSampleSize;
MP4Timestamp startTime;
MP4Duration duration;
MP4Duration renderingOffset;
bool isSyncSample;//stss指定同步帧
bool rc = MP4ReadSample(
mp4File, mediaTrackId, sampleId,
&pSampleBuffer, &sampleSize,
&startTime, &duration,
&renderingOffset, &isSyncSample);
if (!rc) {
MP4DeleteTrack(mp4File, hintTrackId);
CHECK_AND_FREE(pSampleBuffer);
return false;
}
MP4AV_AVSM_HintAddSample(mp4File, //****AVSMspecial****
hintTrackId,
sampleId,
pSampleBuffer,
sampleSize,
sizeLength,
duration,
renderingOffset,
isSyncSample,
maxPayloadSize);
}
CHECK_AND_FREE(pSampleBuffer);
return true;
}
bool MP4AV_AudioInterleaveHinter(
MP4FileHandle mp4File,
MP4TrackId mediaTrackId,
MP4TrackId hintTrackId,
MP4Duration sampleDuration,
u_int8_t stride,
u_int8_t bundle,
u_int16_t maxPayloadSize,
MP4AV_AudioConcatenator pConcatenator)
{
bool rc;
u_int32_t numSamples =
MP4GetTrackNumberOfSamples(mp4File, mediaTrackId);
MP4SampleId* pSampleIds = new MP4SampleId[bundle];
uint32_t sampleIds = 0;
for (u_int32_t i = 1; i <= numSamples; i += stride * bundle) {
for (u_int32_t j = 0; j < stride; j++) {
u_int32_t k;
for (k = 0; k < bundle; k++) {
MP4SampleId sampleId = i + j + (k * stride);
// out of samples for this bundle
if (sampleId > numSamples) {
break;
}
// add sample to this hint
pSampleIds[k] = sampleId;
sampleIds++;
}
if (k == 0) {
break;
}
// compute hint duration
// note this is used to control the RTP timestamps
// that are emitted for the packet,
// it isn't the actual duration of the samples in the packet
MP4Duration hintDuration;
if (j + 1 == stride) {
// at the end of the track
if (i + (stride * bundle) > numSamples) {
hintDuration = ((numSamples - i) - j) * sampleDuration;
if (hintDuration == 0) {
hintDuration = sampleDuration;
}
} else {
hintDuration = ((stride * bundle) - j) * sampleDuration;
}
} else {
hintDuration = sampleDuration;
}
// write hint
rc = (*pConcatenator)(mp4File, mediaTrackId, hintTrackId,
k, pSampleIds, hintDuration, maxPayloadSize);
sampleIds = 0;
if (!rc) {
return false;
}
}
}
delete [] pSampleIds;
return true;
}
extern "C" bool MP4AV_Rfc3016LatmHinter (MP4FileHandle mp4File,
MP4TrackId mediaTrackId,
u_int16_t maxPayloadSize)
{
u_int32_t numSamples = MP4GetTrackNumberOfSamples(mp4File, mediaTrackId);
u_int32_t maxSampleSize = MP4GetTrackMaxSampleSize(mp4File, mediaTrackId);
MP4Duration sampleDuration =
MP4AV_GetAudioSampleDuration(mp4File, mediaTrackId);
if (sampleDuration == MP4_INVALID_DURATION) {
return false;
}
if (numSamples == 0 || maxSampleSize == 0) {
return false;
}
/* get the mpeg4 video configuration */
u_int8_t* pAudioSpecificConfig;
u_int32_t AudioSpecificConfigSize;
if (MP4GetTrackESConfiguration(mp4File, mediaTrackId,
&pAudioSpecificConfig,
&AudioSpecificConfigSize) == false)
return false;
if (pAudioSpecificConfig == NULL ||
AudioSpecificConfigSize == 0) return false;
uint8_t channels = MP4AV_AacConfigGetChannels(pAudioSpecificConfig);
uint32_t freq = MP4AV_AacConfigGetSamplingRate(pAudioSpecificConfig);
uint8_t type = MP4AV_AacConfigGetAudioObjectType(pAudioSpecificConfig);
uint8_t *pConfig;
uint32_t configSize;
MP4AV_LatmGetConfiguration(&pConfig, &configSize,
pAudioSpecificConfig, AudioSpecificConfigSize);
free(pAudioSpecificConfig);
if (pConfig == NULL || configSize == 0) {
CHECK_AND_FREE(pConfig);
return false;
}
MP4TrackId hintTrackId =
MP4AddHintTrack(mp4File, mediaTrackId);
if (hintTrackId == MP4_INVALID_TRACK_ID) {
free(pConfig);
return false;
}
u_int8_t payloadNumber = MP4_SET_DYNAMIC_PAYLOAD;
char buffer[10];
if (channels != 1) {
snprintf(buffer, sizeof(buffer), "%u", channels);
}
/* convert it into ASCII form */
char* sConfig = MP4BinaryToBase16(pConfig, configSize);
free(pConfig);
if (sConfig == NULL ||
MP4SetHintTrackRtpPayload(mp4File, hintTrackId,
"MP4A-LATM", &payloadNumber, 0,
channels != 1 ? buffer : NULL) == false) {
MP4DeleteTrack(mp4File, hintTrackId);
return false;
}
uint32_t profile_level;
// from gpac code
switch (type) {
case 2:
if (channels <= 2) profile_level = freq <= 24000 ? 0x28 : 0x29;
else profile_level = freq <= 48000 ? 0x2a : 0x2b;
break;
case 5:
if (channels <= 2) profile_level = freq < 24000 ? 0x2c : 0x2d;
else profile_level = freq <= 48000 ? 0x2e : 0x2f;
break;
default:
if (channels <= 2) profile_level = freq < 24000 ? 0x0e : 0x0f;
else profile_level = 0x10;
break;
}
/* create the appropriate SDP attribute */
char* sdpBuf = (char*)malloc(strlen(sConfig) + 128);
if (sdpBuf == NULL) {
free(sConfig);
MP4DeleteTrack(mp4File, hintTrackId);
return false;
}
snprintf(sdpBuf,
strlen(sConfig) + 128,
"a=fmtp:%u profile-level-id=%u; cpresent=0; config=%s;\015\012",
payloadNumber,
profile_level,
sConfig);
/* add this to the track's sdp */
bool val = MP4AppendHintTrackSdp(mp4File, hintTrackId, sdpBuf);
free(sConfig);
free(sdpBuf);
if (val == false) {
MP4DeleteTrack(mp4File, hintTrackId);
return false;
}
for (MP4SampleId sampleId = 1; sampleId <= numSamples; sampleId++) {
uint8_t buffer[32];
uint32_t offset = 0;
uint32_t sampleSize =
MP4GetSampleSize(mp4File, mediaTrackId, sampleId);
uint32_t size_left = sampleSize;
while (size_left > 0) {
if (size_left > 0xff) {
size_left -= 0xff;
buffer[offset] = 0xff;
} else {
buffer[offset] = size_left;
size_left = 0;
}
offset++;
}
if (MP4AddRtpHint(mp4File, hintTrackId) == false ||
MP4AddRtpPacket(mp4File, hintTrackId, true) == false ||
MP4AddRtpImmediateData(mp4File, hintTrackId,
buffer, offset) == false ||
MP4AddRtpSampleData(mp4File, hintTrackId,
sampleId, 0, sampleSize) == false ||
MP4WriteRtpHint(mp4File, hintTrackId, sampleDuration) == false) {
MP4DeleteTrack(mp4File, hintTrackId);
return false;
}
}
return true;
}
static void DumpTrack (MP4FileHandle mp4file, MP4TrackId tid,
bool dump_off, bool dump_rend)
{
uint32_t numSamples;
MP4SampleId sid;
uint8_t *buffer;
uint32_t max_frame_size;
uint32_t timescale;
uint64_t msectime;
const char *media_data_name;
uint32_t len_size = 0;
uint8_t video_type = 0;
numSamples = MP4GetTrackNumberOfSamples(mp4file, tid);
max_frame_size = MP4GetTrackMaxSampleSize(mp4file, tid) + 4;
media_data_name = MP4GetTrackMediaDataName(mp4file, tid);
if (strcasecmp(media_data_name, "avc1") == 0) {
MP4GetTrackH264LengthSize(mp4file, tid, &len_size);
} else if (strcasecmp(media_data_name, "mp4v") == 0) {
video_type = MP4GetTrackEsdsObjectTypeId(mp4file, tid);
}
buffer = (uint8_t *)malloc(max_frame_size);
if (buffer == NULL) {
printf("couldn't get buffer\n");
return;
}
timescale = MP4GetTrackTimeScale(mp4file, tid);
printf("mp4file %s, track %d, samples %d, timescale %d\n",
Mp4FileName, tid, numSamples, timescale);
for (sid = 1; sid <= numSamples; sid++) {
MP4Timestamp sampleTime;
MP4Duration sampleDuration, sampleRenderingOffset;
bool isSyncSample = FALSE;
bool ret;
u_int8_t *temp;
uint32_t this_frame_size = max_frame_size;
temp = buffer;
ret = MP4ReadSample(mp4file,
tid,
sid,
&temp,
&this_frame_size,
&sampleTime,
&sampleDuration,
&sampleRenderingOffset,
&isSyncSample);
msectime = sampleTime;
msectime *= TO_U64(1000);
msectime /= timescale;
printf("sampleId %6d, size %5u time "U64"("U64")",
sid, MP4GetSampleSize(mp4file, tid, sid),
sampleTime, msectime);
if (dump_rend) printf(" %6"U64F, sampleRenderingOffset);
if (strcasecmp(media_data_name, "mp4v") == 0) {
if (MP4_IS_MPEG4_VIDEO_TYPE(video_type))
ParseMpeg4(temp, this_frame_size, dump_off);
} else if (strcasecmp(media_data_name, "avc1") == 0) {
ParseH264(temp, this_frame_size, len_size, dump_off);
}
printf("\n");
}
}
// returns -1 for error, 0 for mp4, 1 for aac
int
aac_probe (DB_FILE *fp, const char *fname, MP4FILE_CB *cb, float *duration, int *samplerate, int *channels, int *totalsamples, int *mp4track, MP4FILE *pmp4) {
// try mp4
if (mp4track) {
*mp4track = -1;
}
if (*pmp4) {
*pmp4 = NULL;
}
*duration = -1;
#ifdef USE_MP4FF
mp4ff_t *mp4 = mp4ff_open_read (cb);
#else
MP4FileHandle mp4 = MP4ReadProvider (fname, 0, cb);
#endif
if (!mp4) {
trace ("not an mp4 file\n");
return -1;
}
if (pmp4) {
*pmp4 = mp4;
}
#ifdef USE_MP4FF
int ntracks = mp4ff_total_tracks (mp4);
if (ntracks > 0) {
trace ("m4a container detected, ntracks=%d\n", ntracks);
int i = -1;
trace ("looking for mp4 data...\n");
int sr = -1;
for (i = 0; i < ntracks; i++) {
unsigned char* buff = 0;
unsigned int buff_size = 0;
mp4AudioSpecificConfig mp4ASC;
mp4ff_get_decoder_config(mp4, i, &buff, &buff_size);
if(buff){
int rc = AudioSpecificConfig(buff, buff_size, &mp4ASC);
sr = mp4ASC.samplingFrequency;
free(buff);
if(rc < 0)
continue;
break;
}
}
if (i != ntracks)
{
trace ("mp4 track: %d\n", i);
if (sr != -1) {
*samplerate = sr;
}
else {
*samplerate = mp4ff_get_sample_rate (mp4, i);
}
*channels = mp4ff_get_channel_count (mp4, i);
int samples = mp4ff_num_samples(mp4, i) * 1024;
samples = (int64_t)samples * (*samplerate) / mp4ff_time_scale (mp4, i);
trace ("mp4 nsamples=%d, samplerate=%d, timescale=%d, duration=%lld\n", samples, *samplerate, mp4ff_time_scale(mp4, i), mp4ff_get_track_duration(mp4, i));
*duration = (float)samples / (*samplerate);
if (totalsamples) {
*totalsamples = samples;
}
if (mp4track) {
*mp4track = i;
}
if (!*pmp4) {
mp4ff_close (mp4);
}
return 0;
}
}
#else
MP4FileHandle mp4File = mp4;
MP4TrackId trackId = MP4FindTrackId(mp4File, 0, "audio", 0);
trace ("trackid: %d\n", trackId);
uint32_t timeScale = MP4GetTrackTimeScale(mp4File, trackId);
MP4Duration trackDuration = MP4GetTrackDuration(mp4File, trackId);
MP4SampleId numSamples = MP4GetTrackNumberOfSamples(mp4File, trackId);
u_int8_t* pConfig;
uint32_t configSize = 0;
bool res = MP4GetTrackESConfiguration(mp4File, trackId, &pConfig, &configSize);
if (res && pConfig) {
mp4AudioSpecificConfig mp4ASC;
int rc = AudioSpecificConfig(pConfig, configSize, &mp4ASC);
free (pConfig);
if (rc >= 0) {
*samplerate = mp4ASC.samplingFrequency;
*channels = MP4GetTrackAudioChannels (mp4File, trackId);
// int64_t duration = MP4ConvertFromTrackDuration (mp4File, trackId, trackDuration, timeScale);
int samples = MP4GetTrackNumberOfSamples (mp4File, trackId) * 1024 * (*channels);
trace ("mp4 nsamples=%d, timescale=%d, samplerate=%d\n", samples, timeScale, *samplerate);
*duration = (float)samples / (*samplerate);
if (totalsamples) {
*totalsamples = samples;
}
if (mp4track) {
*mp4track = trackId;
}
if (!*pmp4) {
MP4Close (mp4);
}
return 0;
}
}
#endif
if (*pmp4) {
*pmp4 = NULL;
}
if (mp4) {
#if USE_MP4FF
mp4ff_close (mp4);
#else
MP4Close (mp4);
#endif
mp4 = NULL;
}
trace ("mp4 track not found, looking for aac stream...\n");
// not an mp4, try raw aac
#if USE_MP4FF
deadbeef->rewind (fp);
#endif
if (parse_aac_stream (fp, samplerate, channels, duration, totalsamples) == -1) {
trace ("aac stream not found\n");
return -1;
}
trace ("found aac stream\n");
return 1;
}
void main(int argc, char** argv)
{
if (argc < 2) {
fprintf(stderr, "Usage: %s <file>\n", argv[0]);
exit(1);
}
//u_int32_t verbosity = MP4_DETAILS_ALL;
char* fileName = argv[1];
// open the mp4 file, and read meta-info
MP4FileHandle mp4File = MP4Read(fileName );
uint8_t profileLevel = MP4GetVideoProfileLevel(mp4File);
// get a handle on the first video track
MP4TrackId trackId = MP4FindTrackId(mp4File, 0, "video");
// gather the crucial track information
uint32_t timeScale = MP4GetTrackTimeScale(mp4File, trackId);
// note all times and durations
// are in units of the track time scale
MP4Duration trackDuration = MP4GetTrackDuration(mp4File, trackId);
MP4SampleId numSamples = MP4GetTrackNumberOfSamples(mp4File, trackId);
uint32_t maxSampleSize = MP4GetTrackMaxSampleSize(mp4File, trackId);
uint8_t* pConfig;
uint32_t configSize = 0;
MP4GetTrackESConfiguration(mp4File, trackId, &pConfig, &configSize);
// initialize decoder with Elementary Stream (ES) configuration
// done with our copy of ES configuration
free(pConfig);
// now consecutively read and display the track samples
uint8_t* pSample = (uint8_t*)malloc(maxSampleSize);
uint32_t sampleSize;
MP4Timestamp sampleTime;
MP4Duration sampleDuration;
MP4Duration sampleRenderingOffset;
bool isSyncSample;
for (MP4SampleId sampleId = 1; sampleId <= numSamples; sampleId++) {
// give ReadSample our own buffer, and let it know how big it is
sampleSize = maxSampleSize;
// read next sample from video track
MP4ReadSample(mp4File, trackId, sampleId,
&pSample, &sampleSize,
&sampleTime, &sampleDuration, &sampleRenderingOffset,
&isSyncSample);
// convert timestamp and duration from track time to milliseconds
uint64_t myTime = MP4ConvertFromTrackTimestamp(mp4File, trackId,
sampleTime, MP4_MSECS_TIME_SCALE);
uint64_t myDuration = MP4ConvertFromTrackDuration(mp4File, trackId,
sampleDuration, MP4_MSECS_TIME_SCALE);
// decode frame and display it
}
// close mp4 file
MP4Close(mp4File);
// Note to seek to time 'when' in the track
// use MP4GetSampleIdFromTime(MP4FileHandle hFile,
// MP4Timestamp when, bool wantSyncSample)
// 'wantSyncSample' determines if a sync sample is desired or not
// e.g.
// MP4Timestamp when =
// MP4ConvertToTrackTimestamp(mp4File, trackId, 30, MP4_SECS_TIME_SCALE);
// MP4SampleId newSampleId = MP4GetSampleIdFromTime(mp4File, when, true);
// MP4ReadSample(mp4File, trackId, newSampleId, ...);
//
// Note that start time for sample may be later than 'when'
exit(0);
}
int AacPcm::getInfos(MediaInfo *infos)
{
if(!infos)
return 1;
if(hDecoder)
{
SHOW_INFO()
return 0;
}
IsAAC=strcmpi(infos->getFilename()+lstrlen(infos->getFilename())-4,".aac")==0;
if(!IsAAC) // MP4 file ---------------------------------------------------------------------
{
MP4Duration length;
unsigned __int32 buffer_size;
mp4AudioSpecificConfig mp4ASC;
if(!(mp4File=MP4Read(infos->getFilename(), 0)))
ERROR_getInfos("Error opening file");
if((track=GetAACTrack(mp4File))<0)
ERROR_getInfos(0); //"Unable to find correct AAC sound track");
if(!(hDecoder=faacDecOpen()))
ERROR_getInfos("Error initializing decoder library");
MP4GetTrackESConfiguration(mp4File, track, (unsigned __int8 **)&buffer, &buffer_size);
if(!buffer)
ERROR_getInfos("MP4GetTrackESConfiguration");
AudioSpecificConfig(buffer, buffer_size, &mp4ASC);
Channels=mp4ASC.channelsConfiguration;
if(faacDecInit2(hDecoder, buffer, buffer_size, &Samplerate, &Channels) < 0)
ERROR_getInfos("Error initializing decoder library");
FREE_ARRAY(buffer);
length=MP4GetTrackDuration(mp4File, track);
len_ms=(DWORD)MP4ConvertFromTrackDuration(mp4File, track, length, MP4_MSECS_TIME_SCALE);
file_info.bitrate=MP4GetTrackBitRate(mp4File, track);
file_info.version=MP4GetTrackAudioType(mp4File, track)==MP4_MPEG4_AUDIO_TYPE ? 4 : 2;
numSamples=MP4GetTrackNumberOfSamples(mp4File, track);
sampleId=1;
}
else // AAC file ------------------------------------------------------------------------------
{
DWORD read,
tmp;
BYTE Channels4Raw=0;
if(!(aacFile=fopen(infos->getFilename(),"rb")))
ERROR_getInfos("Error opening file");
// use bufferized stream
setvbuf(aacFile,NULL,_IOFBF,32767);
// get size of file
fseek(aacFile, 0, SEEK_END);
src_size=ftell(aacFile);
fseek(aacFile, 0, SEEK_SET);
if(!(buffer=(BYTE *)malloc(FAAD_STREAMSIZE)))
ERROR_getInfos("Memory allocation error: buffer")
tmp=src_size<FAAD_STREAMSIZE ? src_size : FAAD_STREAMSIZE;
read=fread(buffer, 1, tmp, aacFile);
if(read==tmp)
{
bytes_read=read;
bytes_into_buffer=read;
}
else
ERROR_getInfos("Read failed!")
if(tagsize=id3v2_tag(buffer))
{
if(tagsize>(long)src_size)
ERROR_getInfos("Corrupt stream!");
if(tagsize<bytes_into_buffer)
{
bytes_into_buffer-=tagsize;
memcpy(buffer,buffer+tagsize,bytes_into_buffer);
}
else
{
bytes_read=tagsize;
bytes_into_buffer=0;
if(tagsize>bytes_into_buffer)
fseek(aacFile, tagsize, SEEK_SET);
}
if(src_size<bytes_read+FAAD_STREAMSIZE-bytes_into_buffer)
tmp=src_size-bytes_read;
else
tmp=FAAD_STREAMSIZE-bytes_into_buffer;
read=fread(buffer+bytes_into_buffer, 1, tmp, aacFile);
if(read==tmp)
{
bytes_read+=read;
bytes_into_buffer+=read;
}
else
ERROR_getInfos("Read failed!");
}
if(get_AAC_format((char *)infos->getFilename(), &file_info, &seek_table, &seek_table_length, 0))
ERROR_getInfos("get_AAC_format");
IsSeekable=file_info.headertype==ADTS && seek_table && seek_table_length>0;
BlockSeeking=!IsSeekable;
if(!(hDecoder=faacDecOpen()))
ERROR_getInfos("Can't open library");
if(file_info.headertype==RAW)
{
faacDecConfiguration config;
config.defSampleRate=atoi(cfg_samplerate);
switch(cfg_profile[1])
{
case 'a':
config.defObjectType=MAIN;
break;
case 'o':
config.defObjectType=LOW;
break;
case 'S':
config.defObjectType=SSR;
break;
case 'T':
config.defObjectType=LTP;
break;
}
switch(cfg_bps[0])
{
case '1':
config.outputFormat=FAAD_FMT_16BIT;
break;
case '2':
config.outputFormat=FAAD_FMT_24BIT;
break;
case '3':
config.outputFormat=FAAD_FMT_32BIT;
break;
case 'F':
config.outputFormat=FAAD_FMT_24BIT;
break;
}
faacDecSetConfiguration(hDecoder, &config);
if(!FindBitrate)
{
AacPcm *NewInst;
if(!(NewInst=new AacPcm()))
ERROR_getInfos("Memory allocation error: NewInst");
NewInst->FindBitrate=TRUE;
if(NewInst->getInfos(infos))
ERROR_getInfos(0);
Channels4Raw=NewInst->frameInfo.channels;
file_info.bitrate=NewInst->file_info.bitrate*Channels4Raw;
delete NewInst;
}
else
{
DWORD Samples,
BytesConsumed;
if((bytes_consumed=faacDecInit(hDecoder,buffer,bytes_into_buffer,&Samplerate,&Channels))<0)
ERROR_getInfos("Can't init library");
bytes_into_buffer-=bytes_consumed;
if(!processData(infos,0,0))
ERROR_getInfos(0);
Samples=frameInfo.samples/sizeof(short);
BytesConsumed=frameInfo.bytesconsumed;
processData(infos,0,0);
if(BytesConsumed<frameInfo.bytesconsumed)
BytesConsumed=frameInfo.bytesconsumed;
file_info.bitrate=(BytesConsumed*8*Samplerate)/Samples;
if(!file_info.bitrate)
file_info.bitrate=1000; // try to continue decoding
return 0;
}
}
if((bytes_consumed=faacDecInit(hDecoder, buffer, bytes_into_buffer, &Samplerate, &Channels))<0)
ERROR_getInfos("faacDecInit failed!")
bytes_into_buffer-=bytes_consumed;
if(Channels4Raw)
Channels=Channels4Raw;
len_ms=(DWORD)((1000*((float)src_size*8))/file_info.bitrate);
}
SHOW_INFO();
return 0;
}
Result SoundSourceM4A::tryOpen(const AudioSourceConfig& audioSrcCfg) {
DEBUG_ASSERT(MP4_INVALID_FILE_HANDLE == m_hFile);
/* open MP4 file, check for >= ver 1.9.1 */
#if MP4V2_PROJECT_version_hex <= 0x00010901
m_hFile = MP4Read(getLocalFileNameBytes().constData(), 0);
#else
m_hFile = MP4Read(getLocalFileNameBytes().constData());
#endif
if (MP4_INVALID_FILE_HANDLE == m_hFile) {
qWarning() << "Failed to open file for reading:" << getUrlString();
return ERR;
}
m_trackId = findFirstAudioTrackId(m_hFile);
if (MP4_INVALID_TRACK_ID == m_trackId) {
qWarning() << "No AAC track found:" << getUrlString();
return ERR;
}
const MP4SampleId numberOfSamples =
MP4GetTrackNumberOfSamples(m_hFile, m_trackId);
if (0 >= numberOfSamples) {
qWarning() << "Failed to read number of samples from file:" << getUrlString();
return ERR;
}
m_maxSampleBlockId = kSampleBlockIdMin + (numberOfSamples - 1);
// Determine the maximum input size (in bytes) of a
// sample block for the selected track.
const u_int32_t maxSampleBlockInputSize = MP4GetTrackMaxSampleSize(m_hFile,
m_trackId);
m_inputBuffer.resize(maxSampleBlockInputSize, 0);
DEBUG_ASSERT(NULL == m_hDecoder); // not already opened
m_hDecoder = NeAACDecOpen();
if (!m_hDecoder) {
qWarning() << "Failed to open the AAC decoder!";
return ERR;
}
NeAACDecConfigurationPtr pDecoderConfig = NeAACDecGetCurrentConfiguration(
m_hDecoder);
pDecoderConfig->outputFormat = FAAD_FMT_FLOAT;
if ((kChannelCountMono == audioSrcCfg.channelCountHint) ||
(kChannelCountStereo == audioSrcCfg.channelCountHint)) {
pDecoderConfig->downMatrix = 1;
} else {
pDecoderConfig->downMatrix = 0;
}
pDecoderConfig->defObjectType = LC;
if (!NeAACDecSetConfiguration(m_hDecoder, pDecoderConfig)) {
qWarning() << "Failed to configure AAC decoder!";
return ERR;
}
u_int8_t* configBuffer = NULL;
u_int32_t configBufferSize = 0;
if (!MP4GetTrackESConfiguration(m_hFile, m_trackId, &configBuffer,
&configBufferSize)) {
/* failed to get mpeg-4 audio config... this is ok.
* NeAACDecInit2() will simply use default values instead.
*/
qWarning() << "Failed to read the MP4 audio configuration."
<< "Continuing with default values.";
}
SAMPLERATE_TYPE sampleRate;
unsigned char channelCount;
if (0 > NeAACDecInit2(m_hDecoder, configBuffer, configBufferSize,
&sampleRate, &channelCount)) {
free(configBuffer);
qWarning() << "Failed to initialize the AAC decoder!";
return ERR;
} else {
free(configBuffer);
}
setChannelCount(channelCount);
setFrameRate(sampleRate);
setFrameCount(getFrameCountForSampleBlockId(m_maxSampleBlockId));
// Resize temporary buffer for decoded sample data
const SINT sampleBufferCapacity =
frames2samples(kFramesPerSampleBlock);
m_sampleBuffer.resetCapacity(sampleBufferCapacity);
// Invalidate current position to enforce the following
// seek operation
m_curFrameIndex = getMaxFrameIndex();
// (Re-)Start decoding at the beginning of the file
seekSampleFrame(getMinFrameIndex());
return OK;
}
bool MP4AV_AudioConsecutiveHinter(
MP4FileHandle mp4File,
MP4TrackId mediaTrackId,
MP4TrackId hintTrackId,
MP4Duration sampleDuration,
u_int8_t perPacketHeaderSize,
u_int8_t perSampleHeaderSize,
u_int8_t maxSamplesPerPacket,
u_int16_t maxPayloadSize,
MP4AV_AudioSampleSizer pSizer,
MP4AV_AudioConcatenator pConcatenator,
MP4AV_AudioFragmenter pFragmenter)
{
bool rc;
u_int32_t numSamples =
MP4GetTrackNumberOfSamples(mp4File, mediaTrackId);
u_int16_t bytesThisHint = perPacketHeaderSize;
u_int16_t samplesThisHint = 0;
MP4SampleId* pSampleIds =
new MP4SampleId[maxSamplesPerPacket];
for (MP4SampleId sampleId = 1; sampleId <= numSamples; sampleId++) {
u_int32_t sampleSize =
(*pSizer)(mp4File, mediaTrackId, sampleId);
// sample won't fit in this packet
// or we've reached the limit on samples per packet
if ((int16_t)(sampleSize + perSampleHeaderSize)
> maxPayloadSize - bytesThisHint
|| samplesThisHint == maxSamplesPerPacket) {
if (samplesThisHint > 0) {
rc = (*pConcatenator)(mp4File, mediaTrackId, hintTrackId,
samplesThisHint, pSampleIds,
samplesThisHint * sampleDuration,
maxPayloadSize);
if (!rc) {
return false;
}
}
// start a new hint
samplesThisHint = 0;
bytesThisHint = perPacketHeaderSize;
// fall thru
}
// sample is less than remaining payload size
if ((int16_t)(sampleSize + perSampleHeaderSize)
<= maxPayloadSize - bytesThisHint) {
// add it to this hint
bytesThisHint += (sampleSize + perSampleHeaderSize);
pSampleIds[samplesThisHint++] = sampleId;
} else {
// jumbo frame, need to fragment it
rc = (*pFragmenter)(mp4File, mediaTrackId, hintTrackId,
sampleId, sampleSize, sampleDuration, maxPayloadSize);
if (!rc) {
return false;
}
// start a new hint
samplesThisHint = 0;
bytesThisHint = perPacketHeaderSize;
}
}
if (samplesThisHint > 0) {
rc = (*pConcatenator)(mp4File, mediaTrackId, hintTrackId,
samplesThisHint, pSampleIds,
samplesThisHint * sampleDuration,
maxPayloadSize);
if (!rc) {
return false;
}
}
delete [] pSampleIds;
return true;
}
extern "C" bool MP4AV_RfcIsmaHinter(
MP4FileHandle mp4File,
MP4TrackId mediaTrackId,
bool interleave,
u_int16_t maxPayloadSize)
{
// gather information, and check for validity
u_int32_t numSamples =
MP4GetTrackNumberOfSamples(mp4File, mediaTrackId);
if (numSamples == 0) {
return false;
}
u_int32_t timeScale =
MP4GetTrackTimeScale(mp4File, mediaTrackId);
if (timeScale == 0) {
return false;
}
u_int8_t audioType =
MP4GetTrackEsdsObjectTypeId(mp4File, mediaTrackId);
if (audioType != MP4_MPEG4_AUDIO_TYPE
&& !MP4_IS_AAC_AUDIO_TYPE(audioType)) {
return false;
}
u_int8_t mpeg4AudioType =
MP4GetTrackAudioMpeg4Type(mp4File, mediaTrackId);
if (audioType == MP4_MPEG4_AUDIO_TYPE) {
// check that track contains either MPEG-4 AAC or CELP
if (!MP4_IS_MPEG4_AAC_AUDIO_TYPE(mpeg4AudioType)
&& mpeg4AudioType != MP4_MPEG4_CELP_AUDIO_TYPE) {
return false;
}
}
MP4Duration sampleDuration =
MP4AV_GetAudioSampleDuration(mp4File, mediaTrackId);
if (sampleDuration == MP4_INVALID_DURATION) {
return false;
}
/* get the ES configuration */
u_int8_t* pConfig = NULL;
u_int32_t configSize;
uint8_t channels;
if (MP4GetTrackESConfiguration(mp4File, mediaTrackId,
&pConfig, &configSize) == false)
return false;
if (!pConfig) {
return false;
}
channels = MP4AV_AacConfigGetChannels(pConfig);
/* convert ES Config into ASCII form */
char* sConfig =
MP4BinaryToBase16(pConfig, configSize);
free(pConfig);
if (!sConfig) {
return false;
}
/* create the appropriate SDP attribute */
uint sdpBufLen = strlen(sConfig) + 256;
char* sdpBuf =
(char*)malloc(sdpBufLen);
if (!sdpBuf) {
free(sConfig);
return false;
}
// now add the hint track
MP4TrackId hintTrackId =
MP4AddHintTrack(mp4File, mediaTrackId);
if (hintTrackId == MP4_INVALID_TRACK_ID) {
free(sConfig);
free(sdpBuf);
return false;
}
u_int8_t payloadNumber = MP4_SET_DYNAMIC_PAYLOAD;
char buffer[10];
if (channels != 1) {
snprintf(buffer, sizeof(buffer), "%u", channels);
}
if (MP4SetHintTrackRtpPayload(mp4File, hintTrackId,
"mpeg4-generic", &payloadNumber, 0,
channels != 1 ? buffer : NULL) == false) {
MP4DeleteTrack(mp4File, hintTrackId);
free(sConfig);
free(sdpBuf);
return false;
}
MP4Duration maxLatency;
bool OneByteHeader = false;
if (mpeg4AudioType == MP4_MPEG4_CELP_AUDIO_TYPE) {
snprintf(sdpBuf, sdpBufLen,
"a=fmtp:%u "
"streamtype=5; profile-level-id=15; mode=CELP-vbr; config=%s; "
"SizeLength=6; IndexLength=2; IndexDeltaLength=2; Profile=0;"
"\015\012",
payloadNumber,
sConfig);
// 200 ms max latency for ISMA profile 1
maxLatency = timeScale / 5;
OneByteHeader = true;
} else { // AAC
snprintf(sdpBuf, sdpBufLen,
"a=fmtp:%u "
"streamtype=5; profile-level-id=15; mode=AAC-hbr; config=%s; "
"SizeLength=13; IndexLength=3; IndexDeltaLength=3;"
"\015\012",
payloadNumber,
sConfig);
// 500 ms max latency for ISMA profile 1
maxLatency = timeScale / 2;
}
/* add this to the track's sdp */
bool val = MP4AppendHintTrackSdp(mp4File, hintTrackId, sdpBuf);
free(sConfig);
free(sdpBuf);
if (val == false) {
MP4DeleteTrack(mp4File, hintTrackId);
return false;
}
u_int32_t samplesPerPacket = 0;
if (interleave) {
u_int32_t maxSampleSize =
MP4GetTrackMaxSampleSize(mp4File, mediaTrackId);
// compute how many maximum size samples would fit in a packet
samplesPerPacket =
(maxPayloadSize - 2) / (maxSampleSize + 2);
// can't interleave if this number is 0 or 1
if (samplesPerPacket < 2) {
interleave = false;
}
}
bool rc;
if (interleave) {
u_int32_t samplesPerGroup = maxLatency / sampleDuration;
u_int32_t stride;
stride = samplesPerGroup / samplesPerPacket;
if (OneByteHeader && stride > 3) stride = 3;
if (!OneByteHeader && stride > 7) stride = 7;
#if 0
printf("max latency %llu sampleDuration %llu spg %u spp %u strid %u\n",
maxLatency, sampleDuration, samplesPerGroup,
samplesPerPacket, stride);
#endif
rc = MP4AV_AudioInterleaveHinter(
mp4File,
mediaTrackId,
hintTrackId,
sampleDuration,
stride, // stride
samplesPerPacket, // bundle
maxPayloadSize,
MP4AV_RfcIsmaConcatenator);
} else {
rc = MP4AV_AudioConsecutiveHinter(
mp4File,
mediaTrackId,
hintTrackId,
sampleDuration,
2, // perPacketHeaderSize
2, // perSampleHeaderSize
maxLatency / sampleDuration, // maxSamplesPerPacket
maxPayloadSize,
MP4GetSampleSize,
MP4AV_RfcIsmaConcatenator,
MP4AV_RfcIsmaFragmenter);
}
if (!rc) {
MP4DeleteTrack(mp4File, hintTrackId);
return false;
}
return true;
}
static void *mp4Decode(void *args)
{
MP4FileHandle mp4file;
pthread_mutex_lock(&mutex);
seekPosition = -1;
bPlaying = TRUE;
if(!(mp4file = MP4Read(args, 0))){
mp4cfg.file_type = FILE_AAC;
MP4Close(mp4file);
}else{
mp4cfg.file_type = FILE_MP4;
}
if(mp4cfg.file_type == FILE_MP4){
// We are reading a MP4 file
gint mp4track;
if((mp4track = getAACTrack(mp4file)) < 0){
//TODO: check here for others Audio format.....
g_print("Unsupported Audio track type\n");
g_free(args);
MP4Close(mp4file);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}else{
faacDecHandle decoder;
unsigned char *buffer = NULL;
guint bufferSize = 0;
gulong samplerate;
guchar channels;
guint avgBitrate;
MP4Duration duration;
gulong msDuration;
MP4SampleId numSamples;
MP4SampleId sampleID = 1;
decoder = faacDecOpen();
MP4GetTrackESConfiguration(mp4file, mp4track, &buffer, &bufferSize);
if(!buffer){
g_free(args);
faacDecClose(decoder);
MP4Close(mp4file);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
if(faacDecInit2(decoder, buffer, bufferSize, &samplerate, &channels)<0){
g_free(args);
faacDecClose(decoder);
MP4Close(mp4file);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
g_free(buffer);
if(channels == 0){
g_print("Number of Channels not supported\n");
g_free(args);
faacDecClose(decoder);
MP4Close(mp4file);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
duration = MP4GetTrackDuration(mp4file, mp4track);
msDuration = MP4ConvertFromTrackDuration(mp4file, mp4track, duration,
MP4_MSECS_TIME_SCALE);
numSamples = MP4GetTrackNumberOfSamples(mp4file, mp4track);
mp4_ip.output->open_audio(FMT_S16_NE, samplerate, channels);
mp4_ip.output->flush(0);
mp4_ip.set_info(args, msDuration, -1, samplerate/1000, channels);
g_print("MP4 - %d channels @ %d Hz\n", channels, samplerate);
while(bPlaying){
void* sampleBuffer;
faacDecFrameInfo frameInfo;
gint rc;
if(seekPosition!=-1){
duration = MP4ConvertToTrackDuration(mp4file,
mp4track,
seekPosition*1000,
MP4_MSECS_TIME_SCALE);
sampleID = MP4GetSampleIdFromTime(mp4file, mp4track, duration, 0);
mp4_ip.output->flush(seekPosition*1000);
seekPosition = -1;
}
buffer=NULL;
bufferSize=0;
if(sampleID > numSamples){
mp4_ip.output->close_audio();
g_free(args);
faacDecClose(decoder);
MP4Close(mp4file);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
rc = MP4ReadSample(mp4file, mp4track, sampleID++, &buffer, &bufferSize,
NULL, NULL, NULL, NULL);
//g_print("%d/%d\n", sampleID-1, numSamples);
if((rc==0) || (buffer== NULL)){
g_print("MP4: read error\n");
sampleBuffer = NULL;
sampleID=0;
mp4_ip.output->buffer_free();
mp4_ip.output->close_audio();
g_free(args);
faacDecClose(decoder);
MP4Close(mp4file);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}else{
sampleBuffer = faacDecDecode(decoder, &frameInfo, buffer, bufferSize);
if(frameInfo.error > 0){
g_print("MP4: %s\n",
faacDecGetErrorMessage(frameInfo.error));
mp4_ip.output->close_audio();
g_free(args);
faacDecClose(decoder);
MP4Close(mp4file);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
if(buffer){
g_free(buffer); buffer=NULL; bufferSize=0;
}
while(bPlaying && mp4_ip.output->buffer_free()<frameInfo.samples<<1)
xmms_usleep(30000);
}
mp4_ip.add_vis_pcm(mp4_ip.output->written_time(),
FMT_S16_NE,
channels,
frameInfo.samples<<1,
sampleBuffer);
mp4_ip.output->write_audio(sampleBuffer, frameInfo.samples<<1);
}
while(bPlaying && mp4_ip.output->buffer_free()){
xmms_usleep(10000);
}
mp4_ip.output->close_audio();
g_free(args);
faacDecClose(decoder);
MP4Close(mp4file);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
} else{
// WE ARE READING AN AAC FILE
FILE *file = NULL;
faacDecHandle decoder = 0;
guchar *buffer = 0;
gulong bufferconsumed = 0;
gulong samplerate = 0;
guchar channels;
gulong buffervalid = 0;
TitleInput* input;
gchar *temp = g_strdup(args);
gchar *ext = strrchr(temp, '.');
gchar *xmmstitle = NULL;
faacDecConfigurationPtr config;
if((file = fopen(args, "rb")) == 0){
g_print("AAC: can't find file %s\n", args);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
if((decoder = faacDecOpen()) == NULL){
g_print("AAC: Open Decoder Error\n");
fclose(file);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
config = faacDecGetCurrentConfiguration(decoder);
config->useOldADTSFormat = 0;
faacDecSetConfiguration(decoder, config);
if((buffer = g_malloc(BUFFER_SIZE)) == NULL){
g_print("AAC: error g_malloc\n");
fclose(file);
bPlaying = FALSE;
faacDecClose(decoder);
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
if((buffervalid = fread(buffer, 1, BUFFER_SIZE, file))==0){
g_print("AAC: Error reading file\n");
g_free(buffer);
fclose(file);
bPlaying = FALSE;
faacDecClose(decoder);
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
XMMS_NEW_TITLEINPUT(input);
input->file_name = g_basename(temp);
input->file_ext = ext ? ext+1 : NULL;
input->file_path = temp;
if(!strncmp(buffer, "ID3", 3)){
gint size = 0;
fseek(file, 0, SEEK_SET);
size = (buffer[6]<<21) | (buffer[7]<<14) | (buffer[8]<<7) | buffer[9];
size+=10;
fread(buffer, 1, size, file);
buffervalid = fread(buffer, 1, BUFFER_SIZE, file);
}
xmmstitle = xmms_get_titlestring(xmms_get_gentitle_format(), input);
if(xmmstitle == NULL)
xmmstitle = g_strdup(input->file_name);
if(temp) g_free(temp);
if(input->performer) g_free(input->performer);
if(input->album_name) g_free(input->album_name);
if(input->track_name) g_free(input->track_name);
if(input->genre) g_free(input->genre);
g_free(input);
bufferconsumed = faacDecInit(decoder,
buffer,
buffervalid,
&samplerate,
&channels);
if(mp4_ip.output->open_audio(FMT_S16_NE,samplerate,channels) == FALSE){
g_print("AAC: Output Error\n");
g_free(buffer); buffer=0;
faacDecClose(decoder);
fclose(file);
mp4_ip.output->close_audio();
/*
if(positionTable){
g_free(positionTable); positionTable=0;
}
*/
g_free(xmmstitle);
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
//if(bSeek){
//mp4_ip.set_info(xmmstitle, lenght*1000, -1, samplerate, channels);
//}else{
mp4_ip.set_info(xmmstitle, -1, -1, samplerate, channels);
//}
mp4_ip.output->flush(0);
while(bPlaying && buffervalid > 0){
faacDecFrameInfo finfo;
unsigned long samplesdecoded;
char* sample_buffer = NULL;
/*
if(bSeek && seekPosition!=-1){
fseek(file, positionTable[seekPosition], SEEK_SET);
bufferconsumed=0;
buffervalid = fread(buffer, 1, BUFFER_SIZE, file);
aac_ip.output->flush(seekPosition*1000);
seekPosition=-1;
}
*/
if(bufferconsumed > 0){
memmove(buffer, &buffer[bufferconsumed], buffervalid-bufferconsumed);
buffervalid -= bufferconsumed;
buffervalid += fread(&buffer[buffervalid], 1,
BUFFER_SIZE-buffervalid, file);
bufferconsumed = 0;
}
sample_buffer = faacDecDecode(decoder, &finfo, buffer, buffervalid);
if(finfo.error){
config = faacDecGetCurrentConfiguration(decoder);
if(config->useOldADTSFormat != 1){
faacDecClose(decoder);
decoder = faacDecOpen();
config = faacDecGetCurrentConfiguration(decoder);
config->useOldADTSFormat = 1;
faacDecSetConfiguration(decoder, config);
finfo.bytesconsumed=0;
finfo.samples = 0;
faacDecInit(decoder,
buffer,
buffervalid,
&samplerate,
&channels);
}else{
g_print("FAAD2 Warning %s\n", faacDecGetErrorMessage(finfo.error));
buffervalid = 0;
}
}
bufferconsumed += finfo.bytesconsumed;
samplesdecoded = finfo.samples;
if((samplesdecoded<=0) && !sample_buffer){
g_print("AAC: error sample decoding\n");
continue;
}
while(bPlaying && mp4_ip.output->buffer_free() < (samplesdecoded<<1)){
xmms_usleep(10000);
}
mp4_ip.add_vis_pcm(mp4_ip.output->written_time(),
FMT_S16_LE, channels,
samplesdecoded<<1, sample_buffer);
mp4_ip.output->write_audio(sample_buffer, samplesdecoded<<1);
}
while(bPlaying && mp4_ip.output->buffer_playing()){
xmms_usleep(10000);
}
mp4_ip.output->buffer_free();
mp4_ip.output->close_audio();
bPlaying = FALSE;
g_free(buffer);
faacDecClose(decoder);
g_free(xmmstitle);
fclose(file);
seekPosition = -1;
/*
if(positionTable){
g_free(positionTable); positionTable=0;
}
*/
bPlaying = FALSE;
pthread_mutex_unlock(&mutex);
pthread_exit(NULL);
}
}
/** Action for importing chapters into the <b>job.file</b>
*
*
* @param job the job to process
* @return mp4v2::util::SUCCESS if successful, mp4v2::util::FAILURE otherwise
*/
bool
ChapterUtility::actionImport( JobContext& job )
{
vector<MP4Chapter_t> chapters;
Timecode::Format format;
// create the chapter file name
string inName = job.file;
if( _ChapterFile.empty() )
{
FileSystem::pathnameStripExtension( inName );
inName.append( ".chapters.txt" );
}
else
{
inName = _ChapterFile;
}
if( parseChapterFile( inName, chapters, format ) )
{
return FAILURE;
}
ostringstream oss;
oss << "Importing " << chapters.size() << " " << getChapterTypeName( _ChapterType );
oss << " chapters from file " << inName << " into file " << '"' << job.file << '"' << endl;
verbose1f( "%s", oss.str().c_str() );
if( dryrunAbort() )
{
return SUCCESS;
}
if( 0 == chapters.size() )
{
return herrf( "No chapters found in file %s\n", inName.c_str() );
}
job.fileHandle = MP4Modify( job.file.c_str() );
if( job.fileHandle == MP4_INVALID_FILE_HANDLE )
{
return herrf( "unable to open for write: %s\n", job.file.c_str() );
}
bool isVideoTrack = false;
MP4TrackId refTrackId = getReferencingTrack( job.fileHandle, isVideoTrack );
if( !MP4_IS_VALID_TRACK_ID(refTrackId) )
{
return herrf( "unable to find a video or audio track in file %s\n", job.file.c_str() );
}
if( Timecode::FRAME == format && !isVideoTrack )
{
// we need a video track for this
return herrf( "unable to find a video track in file %s but chapter file contains frame timestamps\n", job.file.c_str() );
}
// get duration and recalculate scale
Timecode refTrackDuration( MP4GetTrackDuration( job.fileHandle, refTrackId ),
MP4GetTrackTimeScale( job.fileHandle, refTrackId ) );
refTrackDuration.setScale( CHAPTERTIMESCALE );
// check for chapters starting after duration of reftrack
for( vector<MP4Chapter_t>::iterator it = chapters.begin(); it != chapters.end(); )
{
Timecode curr( (*it).duration, CHAPTERTIMESCALE );
if( refTrackDuration <= curr )
{
hwarnf( "Chapter '%s' start: %s, playlength of file: %s, chapter cannot be set\n",
(*it).title, curr.svalue.c_str(), refTrackDuration.svalue.c_str() );
it = chapters.erase( it );
}
else
{
++it;
}
}
if( 0 == chapters.size() )
{
return SUCCESS;
}
// convert start time into duration
uint32_t framerate = static_cast<uint32_t>( CHAPTERTIMESCALE );
if( Timecode::FRAME == format )
{
// get the framerate
MP4SampleId sampleCount = MP4GetTrackNumberOfSamples( job.fileHandle, refTrackId );
Timecode tmpcd( refTrackDuration.svalue, CHAPTERTIMESCALE );
framerate = static_cast<uint32_t>( std::ceil( ((double)sampleCount / (double)tmpcd.duration) * CHAPTERTIMESCALE ) );
}
for( vector<MP4Chapter_t>::iterator it = chapters.begin(); it != chapters.end(); ++it )
{
MP4Duration currDur = (*it).duration;
MP4Duration nextDur = chapters.end() == it+1 ? refTrackDuration.duration : (*(it+1)).duration;
if( Timecode::FRAME == format )
{
// convert from frame nr to milliseconds
currDur = convertFrameToMillis( (*it).duration, framerate );
if( chapters.end() != it+1 )
{
nextDur = convertFrameToMillis( (*(it+1)).duration, framerate );
}
}
(*it).duration = nextDur - currDur;
}
// now set the chapters
MP4SetChapters( job.fileHandle, &chapters[0], (uint32_t)chapters.size(), _ChapterType );
fixQtScale( job.fileHandle );
job.optimizeApplicable = true;
return SUCCESS;
}
int decodeMP4file(char *sndfile, aac_dec_opt *opt)
{
int track;
unsigned long samplerate;
unsigned char channels;
void *sample_buffer;
MP4FileHandle infile;
MP4SampleId sampleId, numSamples;
audio_file *aufile;
faacDecHandle hDecoder;
faacDecFrameInfo frameInfo;
unsigned char *buffer;
int buffer_size;
int first_time = 1;
hDecoder = faacDecOpen();
infile = MP4Read(opt->filename, 0);
if (!infile)
{
/* unable to open file */
error_handler("Error opening file: %s\n", opt->filename);
return 1;
}
if ((track = GetAACTrack(infile)) < 0)
{
error_handler("Unable to find correct AAC sound track in the MP4 file.\n");
MP4Close(infile);
return 1;
}
buffer = NULL;
buffer_size = 0;
MP4GetTrackESConfiguration(infile, track, &buffer, &buffer_size);
if(faacDecInit2(hDecoder, buffer, buffer_size, &samplerate, &channels) < 0)
{
/* If some error initializing occured, skip the file */
error_handler("Error initializing decoder library.\n");
faacDecClose(hDecoder);
MP4Close(infile);
return 1;
}
if (buffer)
free(buffer);
numSamples = MP4GetTrackNumberOfSamples(infile, track);
for (sampleId = 1; sampleId <= numSamples; sampleId++)
{
int rc;
/* get access unit from MP4 file */
buffer = NULL;
buffer_size = 0;
rc = MP4ReadSample(infile, track, sampleId, &buffer, &buffer_size, NULL, NULL, NULL, NULL);
if (rc == 0)
{
error_handler("Reading from MP4 file failed.\n");
faacDecClose(hDecoder);
MP4Close(infile);
return 1;
}
sample_buffer = faacDecDecode(hDecoder, &frameInfo, buffer, buffer_size);
if (buffer)
free(buffer);
opt->progress_update((long)numSamples, sampleId);
/* open the sound file now that the number of channels are known */
if (first_time && !frameInfo.error)
{
if(opt->decode_mode == 0)
{
if (Set_WIN_Params (INVALID_FILEDESC, samplerate, SAMPLE_SIZE,
frameInfo.channels) < 0)
{
error_handler("\nCan't access %s\n", "WAVE OUT");
faacDecClose(hDecoder);
MP4Close(infile);
return (0);
}
}
else
{
aufile = open_audio_file(sndfile, samplerate, frameInfo.channels,
opt->output_format, opt->file_type, aacChannelConfig2wavexChannelMask(&frameInfo));
if (aufile == NULL)
{
faacDecClose(hDecoder);
MP4Close(infile);
return 0;
}
}
first_time = 0;
}
if ((frameInfo.error == 0) && (frameInfo.samples > 0))
{
if(opt->decode_mode == 0)
WIN_Play_Samples((short*)sample_buffer, frameInfo.channels*frameInfo.samples);
else
write_audio_file(aufile, sample_buffer, frameInfo.samples, 0);
}
if (frameInfo.error > 0)
{
error_handler("Error: %s\n",
faacDecGetErrorMessage(frameInfo.error));
break;
}
if(stop_decoding)
break;
}
faacDecClose(hDecoder);
MP4Close(infile);
if(opt->decode_mode == 0)
WIN_Audio_close();
else
{
if (!first_time)
close_audio_file(aufile);
}
return frameInfo.error;
}
extern "C" bool MP4AV_Rfc2429Hinter (MP4FileHandle file,
MP4TrackId mediaTrackId,
uint16_t maxPayloadSize)
{
uint32_t numSamples, maxSampleSize;
MP4TrackId hid;
MP4Duration duration;
numSamples = MP4GetTrackNumberOfSamples(file, mediaTrackId);
if (numSamples == 0) {
return false;
}
maxSampleSize = MP4GetTrackMaxSampleSize(file, mediaTrackId);
u_int8_t* pSampleBuffer = (u_int8_t*)malloc(maxSampleSize);
if (pSampleBuffer == NULL) {
return false;
}
hid = MP4AddHintTrack(file, mediaTrackId);
if (hid == MP4_INVALID_TRACK_ID) {
return false;
}
uint8_t payloadNumber = MP4_SET_DYNAMIC_PAYLOAD;
MP4SetHintTrackRtpPayload(file,
hid,
"H263-2000",
&payloadNumber,
0,
NULL,
true,
false);
// strictly speaking, this is not required for H.263 - it's a quicktime
// thing.
u_int16_t videoWidth = MP4GetTrackVideoWidth(file, mediaTrackId);
u_int16_t videoHeight = MP4GetTrackVideoHeight(file, mediaTrackId);
char sdpString[80];
sprintf(sdpString, "a=cliprect:0,0,%d,%d\015\012", videoHeight, videoWidth);
MP4AppendHintTrackSdp(file,
hid,
sdpString);
for (uint32_t sid = 1; sid <= numSamples; sid++) {
duration = MP4GetSampleDuration(file, mediaTrackId, sid);
MP4AddRtpVideoHint(file, hid, false, 0);
u_int32_t sampleSize = maxSampleSize;
MP4Timestamp startTime;
MP4Duration duration;
MP4Duration renderingOffset;
bool isSyncSample;
bool rc = MP4ReadSample(file, mediaTrackId, sid,
&pSampleBuffer, &sampleSize,
&startTime, &duration,
&renderingOffset, &isSyncSample);
if (!rc) {
MP4DeleteTrack(file, hid);
free(pSampleBuffer);
return false;
}
// need to skip the first 2 bytes of the packet - it is the
//start code
uint16_t payload_head = htons(0x400);
uint32_t offset = sizeof(payload_head);
uint32_t remaining = sampleSize - sizeof(payload_head);
while (remaining) {
bool last_pak = false;
uint32_t len;
if (remaining + 2 <= maxPayloadSize) {
len = remaining;
last_pak = true;
} else {
len = maxPayloadSize - 2;
}
MP4AddRtpPacket(file, hid, last_pak);
MP4AddRtpImmediateData(file, hid,
(u_int8_t*)&payload_head, sizeof(payload_head));
payload_head = 0;
MP4AddRtpSampleData(file, hid, sid,
offset, len);
offset += len;
remaining -= len;
}
MP4WriteRtpHint(file, hid, duration, true);
}
free(pSampleBuffer);
return true;
}
Result SoundSourceM4A::tryOpen(const AudioSourceConfig& audioSrcCfg) {
DEBUG_ASSERT(MP4_INVALID_FILE_HANDLE == m_hFile);
// open MP4 file, check for >= ver 1.9.1
// From mp4v2/file.h:
// * On Windows, this should be a UTF-8 encoded string.
// * On other platforms, it should be an 8-bit encoding that is
// * appropriate for the platform, locale, file system, etc.
// * (prefer to use UTF-8 when possible).
#if MP4V2_PROJECT_version_hex <= 0x00010901
m_hFile = MP4Read(getLocalFileName().toUtf8().constData(), 0);
#else
m_hFile = MP4Read(getLocalFileName().toUtf8().constData());
#endif
if (MP4_INVALID_FILE_HANDLE == m_hFile) {
qWarning() << "Failed to open file for reading:" << getUrlString();
return ERR;
}
m_trackId = findFirstAudioTrackId(m_hFile);
if (MP4_INVALID_TRACK_ID == m_trackId) {
qWarning() << "No AAC track found:" << getUrlString();
return ERR;
}
// Read fixed sample duration. If the sample duration is not
// fixed (that is, if the number of frames per sample block varies
// through the file), the call returns MP4_INVALID_DURATION. We
// can't currently handle these.
m_framesPerSampleBlock = MP4GetTrackFixedSampleDuration(m_hFile, m_trackId);
if (MP4_INVALID_DURATION == m_framesPerSampleBlock) {
qWarning() << "Unable to decode tracks with non-fixed sample durations: " << getUrlString();
return ERR;
}
const MP4SampleId numberOfSamples =
MP4GetTrackNumberOfSamples(m_hFile, m_trackId);
if (0 >= numberOfSamples) {
qWarning() << "Failed to read number of samples from file:" << getUrlString();
return ERR;
}
m_maxSampleBlockId = kSampleBlockIdMin + (numberOfSamples - 1);
// Determine the maximum input size (in bytes) of a
// sample block for the selected track.
const u_int32_t maxSampleBlockInputSize = MP4GetTrackMaxSampleSize(m_hFile,
m_trackId);
m_inputBuffer.resize(maxSampleBlockInputSize, 0);
DEBUG_ASSERT(nullptr == m_hDecoder); // not already opened
m_hDecoder = NeAACDecOpen();
if (!m_hDecoder) {
qWarning() << "Failed to open the AAC decoder!";
return ERR;
}
NeAACDecConfigurationPtr pDecoderConfig = NeAACDecGetCurrentConfiguration(
m_hDecoder);
pDecoderConfig->outputFormat = FAAD_FMT_FLOAT;
if ((kChannelCountMono == audioSrcCfg.channelCountHint) ||
(kChannelCountStereo == audioSrcCfg.channelCountHint)) {
pDecoderConfig->downMatrix = 1;
} else {
pDecoderConfig->downMatrix = 0;
}
pDecoderConfig->defObjectType = LC;
if (!NeAACDecSetConfiguration(m_hDecoder, pDecoderConfig)) {
qWarning() << "Failed to configure AAC decoder!";
return ERR;
}
u_int8_t* configBuffer = nullptr;
u_int32_t configBufferSize = 0;
if (!MP4GetTrackESConfiguration(m_hFile, m_trackId, &configBuffer,
&configBufferSize)) {
/* failed to get mpeg-4 audio config... this is ok.
* NeAACDecInit2() will simply use default values instead.
*/
qWarning() << "Failed to read the MP4 audio configuration."
<< "Continuing with default values.";
}
SAMPLERATE_TYPE sampleRate;
unsigned char channelCount;
if (0 > NeAACDecInit2(m_hDecoder, configBuffer, configBufferSize,
&sampleRate, &channelCount)) {
free(configBuffer);
qWarning() << "Failed to initialize the AAC decoder!";
return ERR;
} else {
free(configBuffer);
}
// Calculate how many sample blocks we need to decode in advance
// of a random seek in order to get the recommended number of
// prefetch frames
m_numberOfPrefetchSampleBlocks = (kNumberOfPrefetchFrames +
(m_framesPerSampleBlock - 1)) / m_framesPerSampleBlock;
setChannelCount(channelCount);
setFrameRate(sampleRate);
setFrameCount(((m_maxSampleBlockId - kSampleBlockIdMin) + 1) * m_framesPerSampleBlock);
// Resize temporary buffer for decoded sample data
const SINT sampleBufferCapacity =
frames2samples(m_framesPerSampleBlock);
m_sampleBuffer.resetCapacity(sampleBufferCapacity);
// Invalidate current position to enforce the following
// seek operation
m_curFrameIndex = getMaxFrameIndex();
// (Re-)Start decoding at the beginning of the file
seekSampleFrame(getMinFrameIndex());
return OK;
}
//#define DEBUG_G711 1
extern "C" bool G711Hinter (MP4FileHandle mp4file,
MP4TrackId trackid,
uint16_t maxPayloadSize)
{
uint32_t numSamples;
uint8_t audioType;
MP4SampleId sampleId;
uint32_t sampleSize;
MP4TrackId hintTrackId;
uint8_t payload;
uint32_t bytes_this_hint;
uint32_t sampleOffset;
numSamples = MP4GetTrackNumberOfSamples(mp4file, trackid);
if (numSamples == 0) return false;
audioType = MP4GetTrackEsdsObjectTypeId(mp4file, trackid);
if (audioType != MP4_ALAW_AUDIO_TYPE &&
audioType != MP4_ULAW_AUDIO_TYPE) return false;
hintTrackId = MP4AddHintTrack(mp4file, trackid);
if (hintTrackId == MP4_INVALID_TRACK_ID) {
return false;
}
const char *type;
if (audioType == MP4_ALAW_AUDIO_TYPE) {
payload = 8;
type = "PCMA";
} else {
payload = 0;
type = "PCMU";
}
MP4SetHintTrackRtpPayload(mp4file, hintTrackId, type, &payload, 0,NULL,
false);
MP4Duration sampleDuration;
bool have_skip;
sampleId = 1;
sampleSize = MP4GetSampleSize(mp4file, trackid, sampleId);
sampleDuration = MP4GetSampleDuration(mp4file, trackid, sampleId);
have_skip = sampleDuration != sampleSize;
sampleOffset = 0;
bytes_this_hint = 0;
if (maxPayloadSize > 160) maxPayloadSize = 160;
while (1) {
if (bytes_this_hint == 0) {
#ifdef DEBUG_G711
printf("Adding hint/packet\n");
#endif
MP4AddRtpHint(mp4file, hintTrackId);
MP4AddRtpPacket(mp4file, hintTrackId, false); // marker bit 0
}
uint16_t bytes_left_this_packet;
bytes_left_this_packet = maxPayloadSize - bytes_this_hint;
if (sampleSize >= bytes_left_this_packet) {
MP4AddRtpSampleData(mp4file, hintTrackId,
sampleId, sampleOffset, bytes_left_this_packet);
bytes_this_hint += bytes_left_this_packet;
sampleSize -= bytes_left_this_packet;
sampleOffset += bytes_left_this_packet;
#ifdef DEBUG_G711
printf("Added sample with %u bytes\n", bytes_left_this_packet);
#endif
} else {
MP4AddRtpSampleData(mp4file, hintTrackId,
sampleId, sampleOffset, sampleSize);
bytes_this_hint += sampleSize;
#ifdef DEBUG_G711
printf("Added sample with %u bytes\n", sampleSize);
#endif
sampleSize = 0;
}
if (bytes_this_hint >= maxPayloadSize) {
// Write the hint
// duration is bytes written
MP4WriteRtpHint(mp4file, hintTrackId, bytes_this_hint);
#ifdef DEBUG_G711
printf("Finished packet - bytes %u\n", bytes_this_hint);
#endif
bytes_this_hint = 0;
}
if (sampleSize == 0) {
// next sample
if (have_skip && bytes_this_hint != 0) {
#ifdef DEBUG_G711
printf("duration - ending packet - bytes %u\n", bytes_this_hint);
#endif
MP4WriteRtpHint(mp4file, hintTrackId, bytes_this_hint);
bytes_this_hint = 0;
}
sampleId++;
if (sampleId > numSamples) {
// finish it and exit
if (bytes_this_hint != 0) {
MP4WriteRtpHint(mp4file, hintTrackId, bytes_this_hint);
}
return true;
}
sampleSize = MP4GetSampleSize(mp4file, trackid, sampleId);
sampleDuration = MP4GetSampleDuration(mp4file, trackid, sampleId);
have_skip = sampleDuration != sampleSize;
#ifdef DEBUG_G711
printf("Next sample %u - size %u %u\n", sampleId, sampleSize,
have_skip);
#endif
sampleOffset = 0;
}
}
return true; // will never reach here
}
QWORD MP4TextTrack::ReadPrevious(QWORD time,Listener *listener)
{
//Check it is the first
if (sampleId==1)
{
//Set emtpy frame
frame.SetFrame(time,(wchar_t*)NULL,0);
//call listener
if (listener)
//Call it
listener->onTextFrame(frame);
//Exit
return 1;
}
//The previous one
MP4SampleId prevId = sampleId-1;
//If it was not found
if (sampleId==MP4_INVALID_SAMPLE_ID)
//The latest
prevId = MP4GetTrackNumberOfSamples(mp4,track);
// Get size of sample
frameSize = MP4GetSampleSize(mp4, track, prevId);
// Get data pointer
BYTE *data = (BYTE*)malloc(frameSize);
//Get max data lenght
DWORD dataLen = frameSize;
MP4Timestamp startTime;
MP4Duration duration;
MP4Duration renderingOffset;
// Read next rtp packet
if (!MP4ReadSample(
mp4, // MP4FileHandle hFile
track, // MP4TrackId hintTrackId
prevId, // MP4SampleId sampleId,
(u_int8_t **) &data, // u_int8_t** ppBytes
(u_int32_t *) &dataLen, // u_int32_t* pNumBytes
&startTime, // MP4Timestamp* pStartTime
&duration, // MP4Duration* pDuration
&renderingOffset, // MP4Duration* pRenderingOffset
NULL // bool* pIsSyncSample
))
//Last
return MP4_INVALID_TIMESTAMP;
//Get length
if (dataLen>2)
{
//Get string length
DWORD len = data[0]<<8 | data[1];
//Set frame
frame.SetFrame(time,data+2+renderingOffset,len-renderingOffset-2);
//call listener
if (listener)
//Call it
listener->onTextFrame(frame);
}
// exit next send time
return 1;
}
