Go to the documentation of this file.
23 #include "config_components.h"
39 #define MAJOR_HEADER_INTERVAL 16
41 #define MLP_MIN_LPC_ORDER 1
42 #define MLP_MAX_LPC_ORDER 8
43 #define MLP_MIN_LPC_SHIFT 8
44 #define MLP_MAX_LPC_SHIFT 15
103 #define HUFF_OFFSET_MIN (-16384)
104 #define HUFF_OFFSET_MAX ( 16383)
107 #define NUM_CODEBOOKS 4
216 #define SYNC_MAJOR 0xf8726f
217 #define MAJOR_SYNC_INFO_SIGNATURE 0xB752
220 #define FLAGS_DVDA 0x4000
222 #define FLAGS_CONST 0x8000
224 #define SUBSTREAM_INFO_MAX_2_CHAN 0x01
225 #define SUBSTREAM_INFO_HIGH_RATE 0x02
226 #define SUBSTREAM_INFO_ALWAYS_SET 0x04
227 #define SUBSTREAM_INFO_2_SUBSTREAMS 0x08
250 for (
int i = 0;
i <
fp->order;
i++)
274 for (
unsigned int mat = 0; mat <
mp->count; mat++) {
275 if (prev->
outch[mat] !=
mp->outch[mat])
308 if (prev_mp->
shift[ch] !=
mp->shift[ch]) {
313 for (
unsigned int ch = 0; ch <= rh->
max_channel; ch++)
319 for (
unsigned int ch = rh->
min_channel; ch <= rh->max_channel; ch++) {
356 for (
unsigned int order = 0; order < dst->
order; order++)
370 for (
unsigned int count = 0; count <
MAX_MATRICES; count++)
374 for (
unsigned int count = 0; count <
MAX_MATRICES; count++)
375 dst->
outch[count] =
src->outch[count];
430 uint8_t param_presence_flags = 0;
482 unsigned int sum = 0;
491 ctx->coded_sample_rate[0] = 0x08 + 0;
496 ctx->coded_sample_rate[0] = 0x08 + 1;
502 ctx->coded_sample_rate[0] = 0x08 + 2;
507 ctx->coded_sample_rate[0] = 0x00 + 0;
512 ctx->coded_sample_rate[0] = 0x00 + 1;
518 ctx->coded_sample_rate[0] = 0x00 + 2;
523 "sample rates are 44100, 88200, 176400, 48000, "
527 ctx->coded_sample_rate[1] = -1 & 0xf;
535 "Only mono and stereo are supported at the moment.\n");
546 ctx->wordlength = 16;
552 ctx->wordlength = 24;
557 "Only 16- and 24-bit samples are supported.\n");
560 ctx->coded_sample_fmt[1] = -1 & 0xf;
570 ctx->max_codebook_search = 3;
572 ctx->restart_intervals =
ctx->max_restart_interval /
ctx->min_restart_interval;
578 if (!
ctx->lpc_sample_buffer)
581 size =
ctx->one_sample_buffer_size *
ctx->max_restart_interval;
583 if (!
ctx->major_scratch_buffer)
587 if (!
ctx->major_inout_buffer)
590 ctx->num_substreams = 1;
610 ctx->channel_arrangement =
i;
618 ctx->ch_modifier_thd0 = 0;
619 ctx->ch_modifier_thd1 = 0;
620 ctx->ch_modifier_thd2 = 0;
621 ctx->channel_arrangement = 1;
624 ctx->ch_modifier_thd0 = 1;
625 ctx->ch_modifier_thd1 = 1;
626 ctx->ch_modifier_thd2 = 1;
627 ctx->channel_arrangement = 11;
630 ctx->ch_modifier_thd0 = 2;
631 ctx->ch_modifier_thd1 = 1;
632 ctx->ch_modifier_thd2 = 2;
633 ctx->channel_arrangement = 15;
639 ctx->channel_occupancy = 0;
640 ctx->summary_info = 0;
643 size =
ctx->max_restart_interval;
645 if (!
ctx->max_output_bits)
648 size =
ctx->max_restart_interval;
650 if (!
ctx->lossless_check_data)
658 ctx->sequence_size = sum;
659 size =
ctx->restart_intervals *
ctx->sequence_size *
ctx->avctx->ch_layout.nb_channels;
661 if (!
ctx->channel_params)
664 size =
ctx->restart_intervals *
ctx->sequence_size;
666 if (!
ctx->decoding_params)
683 sizeof(*
ctx->filter_state_buffer[0]));
684 if (!
ctx->filter_state_buffer[
i])
802 for (
unsigned int mat = 0; mat <
mp->count; mat++) {
838 for (
int i = 0;
i <
fp->order;
i++) {
896 for (
unsigned int ch = 0; ch <= rh->
max_channel; ch++)
903 for (
unsigned int ch = rh->
min_channel; ch <= rh->max_channel; ch++) {
959 for (
unsigned int ch = rh->
min_channel; ch <= rh->max_channel; ch++) {
964 codebook_index [ch] = cp->
codebook - 1;
970 sign_huff_offset[ch] -= 7 << lsb_bits[ch];
974 sign_huff_offset[ch] -= 1 << sign_shift;
978 for (
unsigned int ch = rh->
min_channel; ch <= rh->max_channel; ch++) {
980 sample -= sign_huff_offset[ch];
982 if (codebook_index[ch] >= 0) {
983 int vlc =
sample >> lsb_bits[ch];
993 ctx->write_buffer = sample_buffer;
1001 int32_t *lossless_check_data =
ctx->lossless_check_data;
1002 unsigned int cur_subblock_index =
ctx->major_cur_subblock_index;
1003 unsigned int num_subblocks =
ctx->major_filter_state_subblock;
1005 int substr_restart_frame = restart_frame;
1006 uint8_t
parity, checksum;
1011 lossless_check_data +=
ctx->frame_index;
1012 ctx->cur_restart_header = rh;
1016 for (
unsigned int subblock = 0; subblock <= num_subblocks; subblock++) {
1017 unsigned int subblock_index;
1019 subblock_index = cur_subblock_index++;
1021 ctx->cur_decoding_params = &
ctx->major_decoding_params[subblock_index];
1022 ctx->cur_channel_params =
ctx->major_channel_params[subblock_index];
1024 params_changed =
ctx->major_params_changed[subblock_index];
1026 if (substr_restart_frame || params_changed) {
1029 if (substr_restart_frame) {
1045 put_bits(&pb, 1, !substr_restart_frame);
1047 substr_restart_frame = 0;
1054 if (
ctx->last_frames == 0 &&
ctx->shorten_by) {
1057 put_bits(&pb, 16, (
ctx->shorten_by & 0x1FFF) | 0x2000);
1076 substream_data_len[0] = end;
1080 ctx->major_cur_subblock_index +=
ctx->major_filter_state_subblock + 1;
1081 ctx->major_filter_state_subblock = 0;
1088 uint8_t *substream_headers,
unsigned int length,
1090 uint16_t substream_data_len[1])
1092 uint16_t access_unit_header = 0;
1093 uint16_t parity_nibble = 0;
1095 parity_nibble =
ctx->dts;
1096 parity_nibble ^= length;
1098 for (
unsigned int substr = 0; substr <
ctx->num_substreams; substr++) {
1099 uint16_t substr_hdr = 0;
1101 substr_hdr |= (0 << 15);
1102 substr_hdr |= (!restart_frame << 14);
1103 substr_hdr |= (1 << 13);
1104 substr_hdr |= (0 << 12);
1105 substr_hdr |= (substream_data_len[substr] / 2) & 0x0FFF;
1107 AV_WB16(substream_headers, substr_hdr);
1109 parity_nibble ^= *substream_headers++;
1110 parity_nibble ^= *substream_headers++;
1113 parity_nibble ^= parity_nibble >> 8;
1114 parity_nibble ^= parity_nibble >> 4;
1115 parity_nibble &= 0xF;
1117 access_unit_header |= (parity_nibble ^ 0xF) << 12;
1118 access_unit_header |= length & 0xFFF;
1126 int buf_size,
int restart_frame)
1129 uint8_t *buf1, *buf0 = buf;
1136 if (restart_frame) {
1145 for (
unsigned int substr = 0; substr <
ctx->num_substreams; substr++) {
1150 buf =
write_substr(
ctx, buf, buf_size, restart_frame, &substream_data_len[0]);
1152 total_length = buf - buf0;
1156 return total_length;
1171 int32_t *lossless_check_data =
ctx->lossless_check_data;
1173 const int16_t *samples_16 = (
const int16_t *)
samples;
1176 int32_t temp_lossless_check_data = 0;
1177 uint32_t greatest = 0;
1179 lossless_check_data +=
ctx->frame_index;
1181 for (
int i = 0;
i < nb_samples;
i++) {
1183 uint32_t abs_sample;
1186 sample = is24 ? *samples_32++ >> 8 : *samples_16++ * 256;
1190 greatest =
FFMAX(greatest, abs_sample);
1192 temp_lossless_check_data ^= (
sample & 0x00ffffff) <<
channel;
1193 *sample_buffer++ =
sample;
1201 *lossless_check_data++ = temp_lossless_check_data;
1215 unsigned int cur_index = (
ctx->frame_index +
index + 1) %
ctx->max_restart_interval;
1216 int32_t *input_buffer =
ctx->inout_buffer + cur_index *
ctx->one_sample_buffer_size;
1218 for (
unsigned int i = 0;
i <
ctx->avctx->frame_size;
i++) {
1220 *sample_buffer++ = *input_buffer++;
1255 memset(sample_mask, 0x00,
sizeof(sample_mask));
1257 for (
unsigned int i = 0;
i <
ctx->number_of_samples;
i++) {
1259 sample_mask[
channel] |= *sample_buffer++;
1274 int min = INT_MAX,
max = INT_MIN;
1278 for (
int order = 0; order <
fp->order; order++) {
1279 int coeff = fcoeff[order];
1286 coeff_mask |=
coeff;
1318 int32_t *lpc_samples =
ctx->lpc_sample_buffer;
1323 for (
unsigned int i = 0;
i <
ctx->number_of_samples;
i++) {
1324 *lpc_samples++ = *sample_buffer;
1325 sample_buffer +=
ctx->num_channels;
1337 for (
unsigned int i = 0;
i < order;
i++)
1338 fcoeff[
i] = coefs[order-1][
i];
1367 uint64_t score[4], sum[4] = { 0, 0, 0, 0, };
1373 for(
i = 2;
i <
ctx->number_of_samples;
i++) {
1374 int32_t left = left_ch [
i *
ctx->num_channels] - 2 * left_ch [(
i - 1) *
ctx->num_channels] + left_ch [(
i - 2) *
ctx->num_channels];
1375 int32_t right = right_ch[
i *
ctx->num_channels] - 2 * right_ch[(
i - 1) *
ctx->num_channels] + right_ch[(
i - 2) *
ctx->num_channels];
1378 sum[1] +=
FFABS( right);
1388 for(
i = 1;
i < 3;
i++)
1389 if(score[
i] < score[best])
1407 coeff_mask |=
coeff;
1412 mp->fbits [mat] = 14 -
bits;
1420 unsigned int shift = 0;
1424 if (
ctx->num_channels - 2 != 2) {
1440 mp->coeff[0][0] = 1 << 14;
mp->coeff[0][1] = -(1 << 14);
1441 mp->coeff[0][2] = 0 << 14;
mp->coeff[0][2] = 0 << 14;
1442 mp->forco[0][0] = 1 << 14;
mp->forco[0][1] = -(1 << 14);
1443 mp->forco[0][2] = 0 << 14;
mp->forco[0][2] = 0 << 14;
1448 mp->coeff[0][0] = 1 << 14;
mp->coeff[0][1] = 1 << 14;
1449 mp->coeff[0][2] = 0 << 14;
mp->coeff[0][2] = 0 << 14;
1450 mp->forco[0][0] = 1 << 14;
mp->forco[0][1] = -(1 << 14);
1451 mp->forco[0][2] = 0 << 14;
mp->forco[0][2] = 0 << 14;
1455 for (
int mat = 0; mat <
mp->count; mat++)
1468 {-9, 8}, {-8, 7}, {-15, 14},
1488 lsb_bits += !!lsb_bits;
1491 unsign = 1 << (lsb_bits - 1);
1527 unsign = 1 << (lsb_bits - 1);
1536 bo->
min =
max - unsign + 1;
1552 int codebook_offset = 7 + (2 -
codebook);
1554 int lsb_bits = 0, bitcount = 0;
1555 int offset_min = INT_MAX, offset_max = INT_MAX;
1561 while (sample_min < codebook_min || sample_max > codebook_max) {
1567 unsign = 1 << lsb_bits;
1571 unsign_offset -= unsign;
1577 int temp_min, temp_max;
1582 if (temp_min < offset_min)
1583 offset_min = temp_min;
1585 temp_max = unsign - temp_min - 1;
1586 if (temp_max < offset_max)
1587 offset_max = temp_max;
1593 sample_buffer +=
ctx->num_channels;
1611 int previous_count = INT_MAX;
1612 int offset_min, offset_max;
1618 while (offset <= offset_max && offset >= offset_min) {
1625 if (temp_bo.
bitcount < previous_count) {
1630 }
else if (++is_greater >=
ctx->max_codebook_search)
1667 sample_buffer +=
ctx->num_channels;
1673 if (no_filters_used) {
1681 BestOffset temp_bo = { 0, INT_MAX, 0, 0, 0, };
1688 if (no_filters_used) {
1689 offset_max = temp_bo.
max;
1706 #define SAMPLE_MAX(bitdepth) ((1 << (bitdepth - 1)) - 1)
1707 #define SAMPLE_MIN(bitdepth) (~SAMPLE_MAX(bitdepth))
1709 #define MSB_MASK(bits) (-(int)(1u << (bits)))
1722 unsigned int number_of_samples =
ctx->number_of_samples;
1723 unsigned int filter_shift =
fp[
FIR]->shift;
1726 for (
int i = 0;
i < 8;
i++) {
1727 ctx->filter_state_buffer[
FIR][
i] = *sample_buffer;
1728 ctx->filter_state_buffer[
IIR][
i] = *sample_buffer;
1730 sample_buffer +=
ctx->num_channels;
1733 for (
int i = 8;
i < number_of_samples;
i++) {
1740 for (
unsigned int order = 0; order <
fp[
filter]->order; order++)
1741 accum += (int64_t)
ctx->filter_state_buffer[
filter][
i - 1 - order] *
1745 accum >>= filter_shift;
1756 sample_buffer +=
ctx->num_channels;
1759 sample_buffer =
ctx->sample_buffer +
channel;
1760 for (
int i = 0;
i < number_of_samples;
i++) {
1761 *sample_buffer =
ctx->filter_state_buffer[
IIR][
i];
1763 sample_buffer +=
ctx->num_channels;
1787 int32_t *sample_buffer =
ctx->sample_buffer +
ctx->num_channels - 2;
1791 for (
unsigned int i = 0;
i <
ctx->number_of_samples;
i++) {
1792 uint16_t seed_shr7 =
seed >> 7;
1794 *sample_buffer++ = ((int8_t) seed_shr7) * (1 << rh->
noise_shift);
1796 seed = (
seed << 16) ^ seed_shr7 ^ (seed_shr7 << 5);
1798 sample_buffer +=
ctx->num_channels - 2;
1810 unsigned int maxchan =
ctx->num_channels;
1812 for (
unsigned int mat = 0; mat <
mp->count; mat++) {
1815 unsigned int outch =
mp->outch[mat];
1817 sample_buffer =
ctx->sample_buffer;
1818 for (
unsigned int i = 0;
i <
ctx->number_of_samples;
i++) {
1821 for (
unsigned int src_ch = 0; src_ch < maxchan; src_ch++) {
1823 accum += (int64_t)
sample *
mp->forco[mat][src_ch];
1825 sample_buffer[outch] = (accum >> 14) &
mask;
1827 sample_buffer +=
ctx->num_channels;
1842 #define CODEBOOK_CHANGE_BITS 21
1846 memset(path_counter, 0, (
NUM_CODEBOOKS + 1) *
sizeof(*path_counter));
1857 int idx =
src->cur_idx;
1859 *prev_bo = idx ?
ctx->best_offset[idx - 1][
channel] :
1861 int bitcount =
src->bitcount;
1862 int prev_codebook =
src->path[idx];
1864 bitcount += cur_bo[cur_codebook].
bitcount;
1866 if (prev_codebook != cur_codebook ||
1882 unsigned int best_codebook;
1888 unsigned int best_bitcount = INT_MAX;
1893 int prev_best_bitcount = INT_MAX;
1895 for (
unsigned int last_best = 0; last_best < 2; last_best++) {
1908 src_path = &path_counter[
codebook];
1913 if (temp_bitcount < best_bitcount) {
1914 best_bitcount = temp_bitcount;
1918 if (temp_bitcount < prev_best_bitcount) {
1919 prev_best_bitcount = temp_bitcount;
1920 if (src_path != dst_path)
1924 dst_path->
bitcount = temp_bitcount;
1940 best_codebook = *best_path++;
1957 uint8_t max_huff_lsbs = 0;
1958 uint8_t max_output_bits = 0;
1967 if (max_huff_lsbs < huff_lsbs)
1968 max_huff_lsbs = huff_lsbs;
1978 if (max_output_bits < ctx->max_output_bits[
index])
1979 max_output_bits =
ctx->max_output_bits[
index];
1982 ctx->cur_restart_header = &
ctx->restart_header;
1988 ctx->cur_decoding_params = &
ctx->major_decoding_params[
index];
1989 ctx->cur_channel_params =
ctx->major_channel_params[
index];
1993 ctx->prev_decoding_params =
ctx->cur_decoding_params;
1994 ctx->prev_channel_params =
ctx->cur_channel_params;
1997 ctx->major_number_of_subblocks =
ctx->number_of_subblocks;
1998 ctx->major_filter_state_subblock = 1;
1999 ctx->major_cur_subblock_index = 0;
2007 ctx->cur_restart_header = &
ctx->restart_header;
2008 ctx->cur_decoding_params = seq_dp + 1;
2009 ctx->cur_channel_params = seq_cp +
ctx->avctx->ch_layout.nb_channels;
2031 (seq_dp + 0)->blocksize = 8;
2032 (seq_dp + 1)->blocksize -= 8;
2035 ctx->cur_decoding_params = seq_dp +
index;
2036 ctx->cur_channel_params = seq_cp +
index*(
ctx->avctx->ch_layout.nb_channels);
2039 ctx->sample_buffer +=
ctx->cur_decoding_params->blocksize *
ctx->num_channels;
2047 ctx->sample_buffer =
ctx->major_inout_buffer;
2049 ctx->number_of_frames =
ctx->major_number_of_frames;
2050 ctx->number_of_samples =
ctx->major_frame_size;
2052 ctx->cur_restart_header = &
ctx->restart_header;
2054 ctx->cur_decoding_params = &
ctx->major_decoding_params[1];
2055 ctx->cur_channel_params =
ctx->major_channel_params[1];
2069 int bytes_written = 0;
2071 int restart_frame,
ret;
2093 ctx->inout_buffer =
ctx->major_inout_buffer
2094 +
ctx->frame_index *
ctx->one_sample_buffer_size;
2096 ctx->sample_buffer =
ctx->major_scratch_buffer
2097 +
ctx->frame_index *
ctx->one_sample_buffer_size;
2099 ctx->write_buffer =
ctx->inout_buffer;
2103 goto input_and_return;
2106 restart_frame = !
ctx->frame_index;
2108 if (restart_frame) {
2111 if (
ctx->min_restart_interval !=
ctx->max_restart_interval)
2115 if (
ctx->min_restart_interval ==
ctx->max_restart_interval)
2116 ctx->write_buffer =
ctx->sample_buffer;
2128 ctx->next_major_number_of_frames++;
2133 restart_frame = (
ctx->frame_index + 1) %
ctx->min_restart_interval;
2135 if (!restart_frame) {
2136 for (
unsigned int seq_index = 0; seq_index <
ctx->restart_intervals; seq_index++) {
2137 unsigned int number_of_samples;
2139 ctx->sample_buffer =
ctx->major_scratch_buffer;
2140 ctx->inout_buffer =
ctx->major_inout_buffer;
2142 ctx->number_of_frames =
ctx->next_major_number_of_frames;
2143 ctx->number_of_subblocks =
ctx->next_major_number_of_frames + 1;
2145 ctx->seq_channel_params =
ctx->channel_params +
ctx->seq_offset[seq_index] *
channels;
2147 ctx->seq_decoding_params =
ctx->decoding_params +
ctx->seq_offset[seq_index];
2149 number_of_samples = avctx->
frame_size *
ctx->number_of_frames;
2150 ctx->number_of_samples = number_of_samples;
2162 if (
ctx->frame_index == (
ctx->max_restart_interval - 1)) {
2163 ctx->major_frame_size =
ctx->next_major_frame_size;
2164 ctx->next_major_frame_size = 0;
2165 ctx->major_number_of_frames =
ctx->next_major_number_of_frames;
2166 ctx->next_major_number_of_frames = 0;
2170 if (!
frame &&
ctx->last_frames <
ctx->max_restart_interval - 1)
2173 if (bytes_written > 0) {
2210 #if CONFIG_MLP_ENCODER
2222 .p.supported_samplerates = (
const int[]) {44100, 48000, 88200, 96000, 176400, 192000, 0},
2223 #if FF_API_OLD_CHANNEL_LAYOUT
2224 .p.channel_layouts = ff_mlp_channel_layouts,
2230 #if CONFIG_TRUEHD_ENCODER
2242 .p.supported_samplerates = (
const int[]) {44100, 48000, 88200, 96000, 176400, 192000, 0},
2243 #if FF_API_OLD_CHANNEL_LAYOUT
uint8_t fbits[MAX_CHANNELS]
fraction bits
static void clear_decoding_params(DecodingParams *decoding_params)
Clears a DecodingParams struct the way it should be after a restart header.
int frame_size
Number of samples per channel in an audio frame.
static void set_best_codebook(MLPEncodeContext *ctx)
#define AV_LOG_WARNING
Something somehow does not look correct.
static void no_codebook_bits(MLPEncodeContext *ctx, unsigned int channel, int32_t min, int32_t max, BestOffset *bo)
Determines the least amount of bits needed to encode the samples using no codebooks.
#define AV_CH_LAYOUT_5POINT0_BACK
int coded_sample_fmt[2]
sample format encoded for MLP
#define FF_CODEC_CAP_INIT_CLEANUP
The codec allows calling the close function for deallocation even if the init function returned a fai...
#define MLP_MIN_LPC_ORDER
#define AV_CHANNEL_LAYOUT_2POINT1
unsigned int major_cur_subblock_index
uint8_t codebook
Which VLC codebook to use to read residuals.
Filter the word “frame” indicates either a video frame or a group of audio as stored in an AVFrame structure Format for each input and each output the list of supported formats For video that means pixel format For audio that means channel sample they are references to shared objects When the negotiation mechanism computes the intersection of the formats supported at each end of a all references to both lists are replaced with a reference to the intersection And when a single format is eventually chosen for a link amongst the remaining all references to the list are updated That means that if a filter requires that its input and output have the same format amongst a supported all it has to do is use a reference to the same list of formats query_formats can leave some formats unset and return AVERROR(EAGAIN) to cause the negotiation mechanism toagain later. That can be used by filters with complex requirements to use the format negotiated on one link to set the formats supported on another. Frame references ownership and permissions
uint8_t ch_modifier_thd1
channel modifier for TrueHD stream 1
static av_cold void mlp_encode_init_static(void)
static uint8_t xor_32_to_8(uint32_t value)
XOR four bytes into one.
static ChannelParams restart_channel_params[MAX_CHANNELS]
static void write_decoding_params(MLPEncodeContext *ctx, PutBitContext *pb, int params_changed)
Writes decoding parameters to the bitstream.
void ff_af_queue_remove(AudioFrameQueue *afq, int nb_samples, int64_t *pts, int64_t *duration)
Remove frame(s) from the queue.
static int put_bytes_output(const PutBitContext *s)
int sample_rate
samples per second
static enum MLPChMode estimate_stereo_mode(MLPEncodeContext *ctx)
void ff_af_queue_close(AudioFrameQueue *afq)
Close AudioFrameQueue.
int32_t * major_scratch_buffer
Scratch buffer big enough to fit all data for one entire major frame interval.
static void write_major_sync(MLPEncodeContext *ctx, uint8_t *buf, int buf_size)
Writes a major sync header to the bitstream.
static av_cold int mlp_encode_close(AVCodecContext *avctx)
uint16_t blocksize
number of PCM samples in current audio block
#define SAMPLE_MAX(bitdepth)
int coded_peak_bitrate
peak bitrate for this major sync header
int32_t * lpc_sample_buffer
#define AV_CHANNEL_LAYOUT_4POINT1
static void put_sbits(PutBitContext *pb, int n, int32_t value)
static void init_put_bits(PutBitContext *s, uint8_t *buffer, int buffer_size)
Initialize the PutBitContext s.
av_cold void ff_af_queue_init(AVCodecContext *avctx, AudioFrameQueue *afq)
Initialize AudioFrameQueue.
static double mp(int i, double w0, double r)
#define SUBSTREAM_INFO_HIGH_RATE
int8_t shift[MAX_CHANNELS]
Left shift to apply to decoded PCM values to get final 24-bit output.
This structure describes decoded (raw) audio or video data.
static void put_bits(Jpeg2000EncoderContext *s, int val, int n)
put n times val bit
static const BestOffset restart_best_offset[NUM_CODEBOOKS]
static void clear_path_counter(PathCounter *path_counter)
#define AV_CHANNEL_LAYOUT_4POINT0
unsigned int number_of_subblocks
uint8_t ff_mlp_calculate_parity(const uint8_t *buf, unsigned int buf_size)
XOR together all the bytes of a buffer.
unsigned int number_of_frames
#define AV_CHANNEL_LAYOUT_MONO
int64_t duration
Duration of this packet in AVStream->time_base units, 0 if unknown.
#define AV_CHANNEL_LAYOUT_STEREO
const ChannelInformation ff_mlp_ch_info[21]
Tables defining channel information.
filter_frame For filters that do not use the this method is called when a frame is pushed to the filter s input It can be called at any time except in a reentrant way If the input frame is enough to produce then the filter should push the output frames on the output link immediately As an exception to the previous rule if the input frame is enough to produce several output frames then the filter needs output only at least one per link The additional frames can be left buffered in the filter
int nb_channels
Number of channels in this layout.
static void set_major_params(MLPEncodeContext *ctx)
Analyzes all collected bitcounts and selects the best parameters for each individual access unit.
#define AV_PKT_FLAG_KEY
The packet contains a keyframe.
unsigned int min_restart_interval
Min interval of access units in between two major frames.
unsigned int sequence_size
static void write_block_data(MLPEncodeContext *ctx, PutBitContext *pb)
Writes the residuals to the bitstream.
uint16_t ff_mlp_checksum16(const uint8_t *buf, unsigned int buf_size)
AVCodec p
The public AVCodec.
FilterParams filter_params[NUM_FILTERS]
int32_t * filter_state_buffer[NUM_FILTERS]
uint8_t ff_mlp_checksum8(const uint8_t *buf, unsigned int buf_size)
MLP uses checksums that seem to be based on the standard CRC algorithm, but are not (in implementatio...
AVChannelLayout ch_layout
Audio channel layout.
uint8_t huff_lsbs
Size of residual suffix not encoded using VLC.
static int mlp_encode_frame(AVCodecContext *avctx, AVPacket *avpkt, const AVFrame *frame, int *got_packet)
void av_shrink_packet(AVPacket *pkt, int size)
Reduce packet size, correctly zeroing padding.
unsigned int major_number_of_frames
#define MLP_MAX_LPC_ORDER
static void input_data_internal(MLPEncodeContext *ctx, const uint8_t *samples, int nb_samples, int is24)
Inputs data from the samples passed by lavc into the context, shifts them appropriately depending on ...
static void copy_restart_frame_params(MLPEncodeContext *ctx)
static int compare_matrix_params(MLPEncodeContext *ctx, const MatrixParams *prev, const MatrixParams *mp)
Compare two primitive matrices and returns 1 if anything has changed.
#define AV_CHANNEL_LAYOUT_SURROUND
unsigned int next_major_frame_size
Counter of number of samples for next major frame.
ChannelParams major_channel_params[MAJOR_HEADER_INTERVAL+1][MAX_CHANNELS]
ChannelParams to be written to bitstream.
#define FF_CODEC_ENCODE_CB(func)
int ff_af_queue_add(AudioFrameQueue *afq, const AVFrame *f)
Add a frame to the queue.
#define AV_CH_LAYOUT_STEREO
static uint8_t * write_substr(MLPEncodeContext *ctx, uint8_t *buf, int buf_size, int restart_frame, uint16_t substream_data_len[MAX_SUBSTREAMS])
Writes the substream data to the bitstream.
int32_t * sample_buffer
Pointer to current access unit samples.
static void set_filter_params(MLPEncodeContext *ctx, unsigned int channel, unsigned int filter, int clear_filter)
Determines the best filter parameters for the given data and writes the necessary information to the ...
static int ff_thread_once(char *control, void(*routine)(void))
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
#define SUBSTREAM_INFO_MAX_2_CHAN
#define AV_CODEC_CAP_EXPERIMENTAL
Codec is experimental and is thus avoided in favor of non experimental encoders.
#define FF_ARRAY_ELEMS(a)
const ChannelParams * prev_channel_params
unsigned int major_filter_state_subblock
static const uint16_t mask[17]
ChannelParams * seq_channel_params
uint8_t ch_modifier_thd2
channel modifier for TrueHD stream 2
static void copy_matrix_params(MatrixParams *dst, MatrixParams *src)
static int write_access_unit(MLPEncodeContext *ctx, uint8_t *buf, int buf_size, int restart_frame)
Writes an entire access unit to the bitstream.
int flags
Flags modifying the (de)muxer behaviour.
uint16_t dts
Decoding timestamp of current access unit.
int num_channels
Number of channels in major_scratch_buffer.
int bits_per_raw_sample
Bits per sample/pixel of internal libavcodec pixel/sample format.
static int mlp_peak_bitrate(int peak_bitrate, int sample_rate)
static void code_matrix_coeffs(MLPEncodeContext *ctx, unsigned int mat)
Determines how many fractional bits are needed to encode matrix coefficients.
char path[MAJOR_HEADER_INTERVAL+2]
static void write_frame_headers(MLPEncodeContext *ctx, uint8_t *frame_header, uint8_t *substream_headers, unsigned int length, int restart_frame, uint16_t substream_data_len[1])
Writes the access unit and substream headers to the bitstream.
#define FFABS(a)
Absolute value, Note, INT_MIN / INT64_MIN result in undefined behavior as they are not representable ...
static void determine_bits(MLPEncodeContext *ctx)
Determines the least amount of bits needed to encode the samples using any or no codebook.
int flags
major sync info flags
static void lossless_matrix_coeffs(MLPEncodeContext *ctx)
Determines best coefficients to use for the lossless matrix.
static void codebook_bits(MLPEncodeContext *ctx, unsigned int channel, int codebook, int offset, int32_t min, int32_t max, BestOffset *bo, int direction)
Determines the least amount of bits needed to encode the samples using a given codebook.
uint16_t timestamp
Timestamp of current access unit.
int32_t last_frames
Signal last frames.
RestartHeader * cur_restart_header
static const int codebook_extremes[3][2]
Min and max values that can be encoded with each codebook.
static void apply_filters(MLPEncodeContext *ctx)
unsigned int major_frame_size
Number of samples in current major frame being encoded.
int num_substreams
Number of substreams contained within this stream.
int32_t * lossless_check_data
Array with lossless_check_data for each access unit.
static int apply_filter(MLPEncodeContext *ctx, unsigned int channel)
Applies the filter to the current samples, and saves the residual back into the samples buffer.
int av_channel_layout_compare(const AVChannelLayout *chl, const AVChannelLayout *chl1)
Check whether two channel layouts are semantically the same, i.e.
static const float quant_step_size[]
static void copy_filter_params(ChannelParams *dst_cp, ChannelParams *src_cp, int filter)
static void input_to_sample_buffer(MLPEncodeContext *ctx)
static void rematrix_channels(MLPEncodeContext *ctx)
Rematrixes all channels using chosen coefficients.
static void default_decoding_params(MLPEncodeContext *ctx, DecodingParams *decoding_params)
Sets default vales in our encoder for a DecodingParams struct.
int32_t coeff[NUM_FILTERS][MAX_FIR_ORDER]
int ff_lpc_calc_coefs(LPCContext *s, const int32_t *samples, int blocksize, int min_order, int max_order, int precision, int32_t coefs[][MAX_LPC_ORDER], int *shift, enum FFLPCType lpc_type, int lpc_passes, int omethod, int min_shift, int max_shift, int zero_shift)
Calculate LPC coefficients for multiple orders.
unsigned int * max_output_bits
largest output bit-depth
static int number_sbits(int number)
Calculates the smallest number of bits it takes to encode a given signed value in two's complement.
int coded_sample_rate[2]
sample rate encoded for MLP
unsigned int number_of_samples
unsigned int major_number_of_subblocks
uint8_t channel_arrangement
channel arrangement for MLP streams
uint8_t ff_mlp_restart_checksum(const uint8_t *buf, unsigned int bit_size)
Calculate an 8-bit checksum over a restart header – a non-multiple-of-8 number of bits,...
static void code_filter_coeffs(MLPEncodeContext *ctx, FilterParams *fp, int32_t *fcoeff)
Determines the smallest number of bits needed to encode the filter coefficients, and if it's possible...
#define NULL_IF_CONFIG_SMALL(x)
Return NULL if CONFIG_SMALL is true, otherwise the argument without modification.
An AVChannelLayout holds information about the channel layout of audio data.
#define AV_CHANNEL_LAYOUT_2_1
unsigned int max_restart_interval
Max interval of access units in between two major frames.
enum AVSampleFormat sample_fmt
audio sample format
#define MAX_SUBSTREAMS
Maximum number of substreams that can be decoded.
int32_t * write_buffer
Pointer to data currently being written to bitstream.
uint8_t ch_modifier_thd0
channel modifier for TrueHD stream 0
const FFCodec ff_truehd_encoder
BestOffset(* cur_best_offset)[NUM_CODEBOOKS]
#define NUM_FILTERS
number of allowed filters
DecodingParams * cur_decoding_params
uint8_t order
number of taps in filter
static void generate_2_noise_channels(MLPEncodeContext *ctx)
Generates two noise channels worth of data.
uint8_t quant_step_size[MAX_CHANNELS]
left shift to apply to Huffman-decoded residuals
#define AV_CH_LAYOUT_5POINT1_BACK
it s the only field you need to keep assuming you have a context There is some magic you don t need to care about around this just let it vf offset
int16_t huff_offset
Offset to apply to residual values.
int flags
A combination of AV_PKT_FLAG values.
static int number_trailing_zeroes(int32_t sample)
Counts the number of trailing zeroes in a value.
av_cold void ff_lpc_end(LPCContext *s)
Uninitialize LPCContext.
static int compare_filter_params(const ChannelParams *prev_cp, const ChannelParams *cp, int filter)
Compares two FilterParams structures and returns 1 if anything has changed.
#define AV_CHANNEL_LAYOUT_3POINT1
DecodingParams * decoding_params
unsigned int restart_intervals
Number of possible major frame sizes.
uint8_t outch[MAX_MATRICES]
output channel for each matrix
#define i(width, name, range_min, range_max)
int64_t pts
Presentation timestamp in AVStream->time_base units; the time at which the decompressed packet will b...
static int put_bits_count(PutBitContext *s)
int32_t * inout_buffer
Pointer to data currently being read from lavc or written to bitstream.
#define SUBSTREAM_INFO_ALWAYS_SET
static void input_data(MLPEncodeContext *ctx, void *samples, int nb_samples)
Wrapper function for inputting data in two different bit-depths.
static DecodingParams restart_decoding_params[MAX_SUBSTREAMS]
BestOffset best_offset[MAJOR_HEADER_INTERVAL+1][MAX_CHANNELS][NUM_CODEBOOKS]
AVSampleFormat
Audio sample formats.
unsigned int max_codebook_search
#define FF_CODEC_CAP_INIT_THREADSAFE
The codec does not modify any global variables in the init function, allowing to call the init functi...
@ AV_SAMPLE_FMT_S16
signed 16 bits
const char * name
Name of the codec implementation.
const AVChannelLayout ff_mlp_ch_layouts[12]
sample data coding information
av_cold void ff_mlp_init_crc(void)
void * av_calloc(size_t nmemb, size_t size)
#define MAJOR_SYNC_INFO_SIGNATURE
ChannelParams * cur_channel_params
#define NUM_CODEBOOKS
Number of possible codebooks (counting "no codebooks")
these buffered frames must be flushed immediately if a new input produces new the filter must not call request_frame to get more It must just process the frame or queue it The task of requesting more frames is left to the filter s request_frame method or the application If a filter has several the filter must be ready for frames arriving randomly on any input any filter with several inputs will most likely require some kind of queuing mechanism It is perfectly acceptable to have a limited queue and to drop frames when the inputs are too unbalanced request_frame For filters that do not use the this method is called when a frame is wanted on an output For a it should directly call filter_frame on the corresponding output For a if there are queued frames already one of these frames should be pushed If the filter should request a frame on one of its repeatedly until at least one frame has been pushed Return or at least make progress towards producing a frame
DecodingParams major_decoding_params[MAJOR_HEADER_INTERVAL+1]
DecodingParams to be written to bitstream.
int32_t coeff[MAX_MATRICES][MAX_CHANNELS+2]
decoding coefficients
Tag MUST be and< 10hcoeff half pel interpolation filter coefficients, hcoeff[0] are the 2 middle coefficients[1] are the next outer ones and so on, resulting in a filter like:...eff[2], hcoeff[1], hcoeff[0], hcoeff[0], hcoeff[1], hcoeff[2] ... the sign of the coefficients is not explicitly stored but alternates after each coeff and coeff[0] is positive, so ...,+,-,+,-,+,+,-,+,-,+,... hcoeff[0] is not explicitly stored but found by subtracting the sum of all stored coefficients with signs from 32 hcoeff[0]=32 - hcoeff[1] - hcoeff[2] - ... a good choice for hcoeff and htaps is htaps=6 hcoeff={40,-10, 2} an alternative which requires more computations at both encoder and decoder side and may or may not be better is htaps=8 hcoeff={42,-14, 6,-2}ref_frames minimum of the number of available reference frames and max_ref_frames for example the first frame after a key frame always has ref_frames=1spatial_decomposition_type wavelet type 0 is a 9/7 symmetric compact integer wavelet 1 is a 5/3 symmetric compact integer wavelet others are reserved stored as delta from last, last is reset to 0 if always_reset||keyframeqlog quality(logarithmic quantizer scale) stored as delta from last, last is reset to 0 if always_reset||keyframemv_scale stored as delta from last, last is reset to 0 if always_reset||keyframe FIXME check that everything works fine if this changes between framesqbias dequantization bias stored as delta from last, last is reset to 0 if always_reset||keyframeblock_max_depth maximum depth of the block tree stored as delta from last, last is reset to 0 if always_reset||keyframequant_table quantization tableHighlevel bitstream structure:==============================--------------------------------------------|Header|--------------------------------------------|------------------------------------|||Block0||||split?||||yes no||||......... intra?||||:Block01 :yes no||||:Block02 :....... ..........||||:Block03 ::y DC ::ref index:||||:Block04 ::cb DC ::motion x :||||......... :cr DC ::motion y :||||....... ..........|||------------------------------------||------------------------------------|||Block1|||...|--------------------------------------------|------------ ------------ ------------|||Y subbands||Cb subbands||Cr subbands||||--- ---||--- ---||--- ---|||||LL0||HL0||||LL0||HL0||||LL0||HL0|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||LH0||HH0||||LH0||HH0||||LH0||HH0|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||HL1||LH1||||HL1||LH1||||HL1||LH1|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||HH1||HL2||||HH1||HL2||||HH1||HL2|||||...||...||...|||------------ ------------ ------------|--------------------------------------------Decoding process:=================------------|||Subbands|------------||||------------|Intra DC||||LL0 subband prediction ------------|\ Dequantization ------------------- \||Reference frames|\ IDWT|------- -------|Motion \|||Frame 0||Frame 1||Compensation . OBMC v -------|------- -------|--------------. \------> Frame n output Frame Frame<----------------------------------/|...|------------------- Range Coder:============Binary Range Coder:------------------- The implemented range coder is an adapted version based upon "Range encoding: an algorithm for removing redundancy from a digitised message." by G. N. N. Martin. The symbols encoded by the Snow range coder are bits(0|1). The associated probabilities are not fix but change depending on the symbol mix seen so far. bit seen|new state ---------+----------------------------------------------- 0|256 - state_transition_table[256 - old_state];1|state_transition_table[old_state];state_transition_table={ 0, 0, 0, 0, 0, 0, 0, 0, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 190, 191, 192, 194, 194, 195, 196, 197, 198, 199, 200, 201, 202, 202, 204, 205, 206, 207, 208, 209, 209, 210, 211, 212, 213, 215, 215, 216, 217, 218, 219, 220, 220, 222, 223, 224, 225, 226, 227, 227, 229, 229, 230, 231, 232, 234, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 248, 0, 0, 0, 0, 0, 0, 0};FIXME Range Coding of integers:------------------------- FIXME Neighboring Blocks:===================left and top are set to the respective blocks unless they are outside of the image in which case they are set to the Null block top-left is set to the top left block unless it is outside of the image in which case it is set to the left block if this block has no larger parent block or it is at the left side of its parent block and the top right block is not outside of the image then the top right block is used for top-right else the top-left block is used Null block y, cb, cr are 128 level, ref, mx and my are 0 Motion Vector Prediction:=========================1. the motion vectors of all the neighboring blocks are scaled to compensate for the difference of reference frames scaled_mv=(mv *(256 *(current_reference+1)/(mv.reference+1))+128)> the median of the scaled left
#define AV_CHANNEL_LAYOUT_5POINT0_BACK
static int best_codebook_path_cost(MLPEncodeContext *ctx, unsigned int channel, PathCounter *src, int cur_codebook)
static void no_codebook_bits_offset(MLPEncodeContext *ctx, unsigned int channel, int16_t offset, int32_t min, int32_t max, BestOffset *bo)
Determines the amount of bits needed to encode the samples using no codebooks and a specified offset.
static int compare_decoding_params(MLPEncodeContext *ctx)
Compares two DecodingParams and ChannelParams structures to decide if a new decoding params header ha...
static void determine_quant_step_size(MLPEncodeContext *ctx)
Determines how many bits are zero at the end of all samples so they can be shifted out.
static int compare_best_offset(const BestOffset *prev, const BestOffset *cur)
main external API structure.
int major_params_changed[MAJOR_HEADER_INTERVAL+1]
params_changed to be written to bitstream.
static void write_filter_params(MLPEncodeContext *ctx, PutBitContext *pb, unsigned int channel, unsigned int filter)
Writes filter parameters for one filter to the bitstream.
const FFCodec ff_mlp_encoder
static void clear_channel_params(ChannelParams channel_params[MAX_CHANNELS], int nb_channels)
Clears a ChannelParams struct the way it should be after a restart header.
int32_t * major_inout_buffer
Buffer with all in/out data for one entire major frame interval.
ChannelParams * channel_params
unsigned int next_major_number_of_frames
MatrixParams matrix_params
static void analyze_sample_buffer(MLPEncodeContext *ctx)
#define AV_CODEC_CAP_DELAY
Encoder or decoder requires flushing with NULL input at the end in order to give the complete and cor...
Filter the word “frame” indicates either a video frame or a group of audio samples
#define SAMPLE_MIN(bitdepth)
unsigned int one_sample_buffer_size
Number of samples*channel for one access unit.
static void codebook_bits_offset(MLPEncodeContext *ctx, unsigned int channel, int codebook, int32_t sample_min, int32_t sample_max, int16_t offset, BestOffset *bo)
Determines the least amount of bits needed to encode the samples using a given codebook and a given o...
RestartHeader restart_header
static int shift(int a, int b)
unsigned int seq_index
Sequence index for high compression levels.
unsigned int frame_index
Index of current frame being encoded.
static void process_major_frame(MLPEncodeContext *ctx)
int frame_number
Frame counter, set by libavcodec.
const DecodingParams * prev_decoding_params
static void flush_put_bits(PutBitContext *s)
Pad the end of the output stream with zeros.
unsigned int seq_size[MAJOR_HEADER_INTERVAL]
static av_always_inline int diff(const uint32_t a, const uint32_t b)
const uint8_t ff_mlp_huffman_tables[3][18][2]
Tables defining the Huffman codes.
static av_cold int mlp_encode_init(AVCodecContext *avctx)
This structure stores compressed data.
DecodingParams * seq_decoding_params
unsigned int seq_offset[MAJOR_HEADER_INTERVAL]
#define CODEBOOK_CHANGE_BITS
static const double coeff[2][5]
#define AVERROR_INVALIDDATA
Invalid data found when processing input.
static void write_matrix_params(MLPEncodeContext *ctx, PutBitContext *pb)
Writes matrix params for all primitive matrices to the bitstream.
#define MLP_MIN_LPC_SHIFT
uint8_t count
number of matrices to apply
#define AV_CHANNEL_LAYOUT_QUAD
#define AV_CHANNEL_LAYOUT_5POINT1_BACK
#define AV_CODEC_CAP_SMALL_LAST_FRAME
Codec can be fed a final frame with a smaller size.
static void determine_filters(MLPEncodeContext *ctx)
Tries to determine a good prediction filter, and applies it to the samples buffer if the filter is go...
uint8_t shift
Right shift to apply to output of filter.
@ AV_SAMPLE_FMT_S32
signed 32 bits
#define MLP_MAX_LPC_SHIFT
int ff_alloc_packet(AVCodecContext *avctx, AVPacket *avpkt, int64_t size)
Check AVPacket size and allocate data.
@ FF_LPC_TYPE_LEVINSON
Levinson-Durbin recursion.
uint8_t param_presence_flags
Bitmask of which parameter sets are conveyed in a decoding parameter block.
#define MAJOR_HEADER_INTERVAL
MLP encoder Copyright (c) 2008 Ramiro Polla Copyright (c) 2016-2019 Jai Luthra.
static const unsigned codebook[256][2]
av_cold int ff_lpc_init(LPCContext *s, int blocksize, int max_order, enum FFLPCType lpc_type)
Initialize LPCContext.
static void write_restart_header(MLPEncodeContext *ctx, PutBitContext *pb)
Writes a restart header to the bitstream.