DCADEC_API int dcadec_frame_parse_header(const uint8_t *data, size_t *size)
{
struct bitstream bits;
uint8_t header[DCADEC_FRAME_HEADER_SIZE];
size_t header_size, frame_size;
int ret;
if (!data || !size)
return -DCADEC_EINVAL;
if ((ret = dcadec_frame_convert_bitstream(header, &header_size,
data, DCADEC_FRAME_HEADER_SIZE)) < 0)
return ret;
bits_init(&bits, header, header_size);
switch (bits_get(&bits, 32)) {
case SYNC_WORD_CORE: {
bool normal_frame = bits_get1(&bits);
int deficit_samples = bits_get(&bits, 5) + 1;
int npcmblocks;
if (normal_frame && deficit_samples != 32)
return -DCADEC_ENOSYNC;
bits_skip1(&bits);
npcmblocks = bits_get(&bits, 7) + 1;
if ((npcmblocks & 7) && (npcmblocks < 6 || normal_frame))
return -DCADEC_ENOSYNC;
frame_size = bits_get(&bits, 14) + 1;
if (frame_size < 96)
return -DCADEC_ENOSYNC;
if (ret & DCADEC_BITSTREAM_BE14)
*size = frame_size * 8 / 14 * 2;
else
*size = frame_size;
return DCADEC_FRAME_TYPE_CORE;
}
case SYNC_WORD_EXSS: {
bool wide_hdr;
bits_skip(&bits, 10);
wide_hdr = bits_get1(&bits);
header_size = bits_get(&bits, 8 + 4 * wide_hdr) + 1;
if ((header_size & 3) || header_size < DCADEC_FRAME_HEADER_SIZE)
return -DCADEC_ENOSYNC;
frame_size = bits_get(&bits, 16 + 4 * wide_hdr) + 1;
if ((frame_size & 3) || frame_size < header_size)
return -DCADEC_ENOSYNC;
*size = frame_size;
return DCADEC_FRAME_TYPE_EXSS;
}
default:
return -DCADEC_ENOSYNC;
}
}
static int parse_descriptor(struct exss_asset *asset)
{
struct exss_parser *exss = asset->parser;
int i, j, ret, descr_pos = exss->bits.index;
// Size of audio asset descriptor in bytes
int descr_size = bits_get(&exss->bits, 9) + 1;
// Audio asset identifier
asset->asset_index = bits_get(&exss->bits, 3);
//
// Per stream static metadata
//
if (exss->static_fields_present) {
// Asset type descriptor presence
if (bits_get1(&exss->bits))
// Asset type descriptor
bits_skip(&exss->bits, 4);
// Language descriptor presence
if (bits_get1(&exss->bits))
// Language descriptor
bits_skip(&exss->bits, 24);
// Additional textual information presence
if (bits_get1(&exss->bits)) {
// Byte size of additional text info
int text_size = bits_get(&exss->bits, 10) + 1;
// Additional textual information string
bits_skip(&exss->bits, text_size * 8);
}
// PCM bit resolution
asset->pcm_bit_res = bits_get(&exss->bits, 5) + 1;
// Maximum sample rate
asset->max_sample_rate = exss_sample_rates[bits_get(&exss->bits, 4)];
// Total number of channels
asset->nchannels_total = bits_get(&exss->bits, 8) + 1;
// One to one map channel to speakers
asset->one_to_one_map_ch_to_spkr = bits_get1(&exss->bits);
if (asset->one_to_one_map_ch_to_spkr) {
// Embedded stereo flag
if (asset->nchannels_total > 2)
asset->embedded_stereo = bits_get1(&exss->bits);
// Embedded 6 channels flag
if (asset->nchannels_total > 6)
asset->embedded_6ch = bits_get1(&exss->bits);
// Speaker mask enabled flag
asset->spkr_mask_enabled = bits_get1(&exss->bits);
int spkr_mask_nbits = 0;
if (asset->spkr_mask_enabled) {
// Number of bits for speaker activity mask
spkr_mask_nbits = (bits_get(&exss->bits, 2) + 1) << 2;
// Loudspeaker activity mask
asset->spkr_mask = bits_get(&exss->bits, spkr_mask_nbits);
}
// Number of speaker remapping sets
int spkr_remap_nsets = bits_get(&exss->bits, 3);
if (spkr_remap_nsets && !spkr_mask_nbits) {
exss_err("Speaker mask disabled yet there are remapping sets");
return -DCADEC_EBADDATA;
}
// Standard loudspeaker layout mask
int nspeakers[8];
for (i = 0; i < spkr_remap_nsets; i++)
nspeakers[i] = count_chs_for_mask(bits_get(&exss->bits, spkr_mask_nbits));
for (i = 0; i < spkr_remap_nsets; i++) {
// Number of channels to be decoded for speaker remapping
int nch_for_remaps = bits_get(&exss->bits, 5) + 1;
for (j = 0; j < nspeakers[i]; j++) {
// Decoded channels to output speaker mapping mask
int remap_ch_mask = bits_get(&exss->bits, nch_for_remaps);
// Loudspeaker remapping codes
int ncodes = dca_popcount(remap_ch_mask);
bits_skip(&exss->bits, ncodes * 5);
}
}
} else {
asset->embedded_stereo = false;
asset->embedded_6ch = false;
asset->spkr_mask_enabled = false;
asset->spkr_mask = 0;
// Representation type
asset->representation_type = bits_get(&exss->bits, 3);
}
}
//
// DRC, DNC and mixing metadata
//
// Dynamic range coefficient presence flag
bool drc_present = bits_get1(&exss->bits);
// Code for dynamic range coefficient
if (drc_present)
bits_skip(&exss->bits, 8);
// Dialog normalization presence flag
if (bits_get1(&exss->bits))
// Dialog normalization code
bits_skip(&exss->bits, 5);
// DRC for stereo downmix
if (drc_present && asset->embedded_stereo)
bits_skip(&exss->bits, 8);
// Mixing metadata presence flag
if (exss->mix_metadata_enabled && bits_get1(&exss->bits)) {
// External mixing flag
bits_skip1(&exss->bits);
// Post mixing / replacement gain adjustment
bits_skip(&exss->bits, 6);
// DRC prior to mixing
if (bits_get(&exss->bits, 2) == 3)
// Custom code for mixing DRC
bits_skip(&exss->bits, 8);
else
// Limit for mixing DRC
bits_skip(&exss->bits, 3);
// Scaling type for channels of main audio
// Scaling parameters of main audio
if (bits_get1(&exss->bits))
for (i = 0; i < exss->nmixoutconfigs; i++)
bits_skip(&exss->bits, 6 * exss->nmixoutchs[i]);
else
bits_skip(&exss->bits, 6 * exss->nmixoutconfigs);
int nchannels_dmix = asset->nchannels_total;
if (asset->embedded_6ch)
nchannels_dmix += 6;
if (asset->embedded_stereo)
nchannels_dmix += 2;
for (i = 0; i < exss->nmixoutconfigs; i++) {
for (j = 0; j < nchannels_dmix; j++) {
if (!exss->nmixoutchs[i]) {
exss_err("Invalid speaker layout mask for mixing configuration");
return -DCADEC_EBADDATA;
}
// Mix output mask
int mix_map_mask = bits_get(&exss->bits, exss->nmixoutchs[i]);
// Mixing coefficients
int nmixcoefs = dca_popcount(mix_map_mask);
bits_skip(&exss->bits, 6 * nmixcoefs);
}
}
}
//
// Decoder navigation data
//
// Coding mode for the asset
asset->coding_mode = bits_get(&exss->bits, 2);
// Coding components used in asset
switch (asset->coding_mode) {
case 0: // Coding mode that may contain multiple coding components
asset->extension_mask = bits_get(&exss->bits, 12);
if (asset->extension_mask & EXSS_CORE) {
// Size of core component in extension substream
asset->core_size = bits_get(&exss->bits, 14) + 1;
// Core sync word present flag
if (bits_get1(&exss->bits))
// Core sync distance
bits_skip(&exss->bits, 2);
}
if (asset->extension_mask & EXSS_XBR)
// Size of XBR extension in extension substream
asset->xbr_size = bits_get(&exss->bits, 14) + 1;
if (asset->extension_mask & EXSS_XXCH)
// Size of XXCH extension in extension substream
asset->xxch_size = bits_get(&exss->bits, 14) + 1;
if (asset->extension_mask & EXSS_X96)
// Size of X96 extension in extension substream
asset->x96_size = bits_get(&exss->bits, 12) + 1;
if (asset->extension_mask & EXSS_LBR)
parse_lbr_parameters(asset);
if (asset->extension_mask & EXSS_XLL)
parse_xll_parameters(asset);
if (asset->extension_mask & EXSS_RSV1)
bits_skip(&exss->bits, 16);
if (asset->extension_mask & EXSS_RSV2)
bits_skip(&exss->bits, 16);
break;
case 1: // Loss-less coding mode without CBR component
asset->extension_mask = EXSS_XLL;
parse_xll_parameters(asset);
break;
case 2: // Low bit rate mode
asset->extension_mask = EXSS_LBR;
parse_lbr_parameters(asset);
break;
case 3: // Auxiliary coding mode
asset->extension_mask = 0;
// Size of auxiliary coded data
bits_skip(&exss->bits, 14);
// Auxiliary codec identification
bits_skip(&exss->bits, 8);
// Aux sync word present flag
if (bits_get1(&exss->bits))
// Aux sync distance
bits_skip(&exss->bits, 3);
break;
}
if (asset->extension_mask & EXSS_XLL)
// DTS-HD stream ID
asset->hd_stream_id = bits_get(&exss->bits, 3);
// One to one mixing flag
// Per channel main audio scaling flag
// Main audio scaling codes
// Decode asset in secondary decoder flag
// Revision 2 DRC metadata
// Reserved
// Zero pad
if ((ret = bits_seek(&exss->bits, descr_pos + descr_size * 8)) < 0)
exss_err("Read past end of asset descriptor");
return ret;
}