Go to the documentation of this file.
21 #define BITSTREAM_READER_LE
88 { 0, -1, -1, -1, -1 },
98 { 0, -1, -1, -1, -1 },
101 { 1, 2, -1, -1, -1 },
124 -0.995734176295034521871191178905, -0.961825643172819070408796290732,
125 -0.895163291355062322067016499754, -0.798017227280239503332805112796,
126 -0.673695643646557211712691912426, -0.526432162877355800244607799141,
127 -0.361241666187152948744714596184, -0.183749517816570331574408839621,
128 0.0, 0.207911690817759337101742284405,
129 0.406736643075800207753985990341, 0.587785252292473129168705954639,
130 0.743144825477394235014697048974, 0.866025403784438646763723170753,
131 0.951056516295153572116439333379, 0.994521895368273336922691944981
138 for (
i = 0;
i < 256;
i++)
145 int i, ps, si,
code, step_i;
151 value = (((ps & 0x7fffff) ^ -si) + si) * (1.0
f / 0x7fffff);
154 if (step_i > step_max) {
161 for (
i = 0;
i < 64;
i++) {
187 step_i =
av_clip(step_i, 0, step_max);
190 s->lfe_data[
i] =
value *
s->lfe_scale;
199 int i, ps, si,
code, step_i;
205 value = (((ps & 0x7fff) ^ -si) + si) * (1.0
f / 0x7fff);
208 if (step_i > step_max) {
215 for (
i = 0;
i < 64;
i++) {
237 step_i =
av_clip(step_i, 0, step_max);
240 s->lfe_data[
i] =
value *
s->lfe_scale;
261 if (chunk->
len >= 52)
263 if (chunk->
len >= 35)
271 int nb_bits,
int max_depth)
285 int sf, sf_idx, ch, main_ch, freq;
289 for (sf = 0; sf < 1 << group; sf +=
diff ? 8 : 1) {
290 sf_idx = ((
s->framenum << group) + sf) & 31;
291 s->tonal_bounds[group][sf_idx][0] =
s->ntones;
294 for (freq = 1;; freq++) {
311 if (freq >> (5 - group) >
s->nsubbands * 4 - 6) {
320 +
s->limited_range - 2;
321 amp[main_ch] = main_amp <
AMP_MAX ? main_amp : 0;
325 for (ch = 0; ch <
s->nchannels_total; ch++) {
342 t->
x_freq = freq >> (5 - group);
343 t->
f_delt = (freq & ((1 << (5 - group)) - 1)) << group;
349 for (ch = 0; ch <
s->nchannels; ch++) {
351 t->
phs[ch] = 128 - phs[ch] * 32 +
shift;
356 s->tonal_bounds[group][sf_idx][1] =
s->ntones;
380 for (sb = 0; sb < 6; sb++)
386 for (group = 0; group < 5; group++) {
427 int i, sf, prev, next, dist;
436 for (sf = 0; sf < 7; sf += dist) {
456 next = prev + ((next + 1) >> 1);
458 next = prev - ( next >> 1);
464 scf[sf + 1] = prev + ((next - prev) >> 1);
466 scf[sf + 1] = prev - ((prev - next) >> 1);
471 scf[sf + 1] = prev + ( (next - prev) >> 2);
472 scf[sf + 2] = prev + ( (next - prev) >> 1);
473 scf[sf + 3] = prev + (((next - prev) * 3) >> 2);
475 scf[sf + 1] = prev - ( (prev - next) >> 2);
476 scf[sf + 2] = prev - ( (prev - next) >> 1);
477 scf[sf + 3] = prev - (((prev - next) * 3) >> 2);
482 for (
i = 1;
i < dist;
i++)
483 scf[sf +
i] = prev + (next - prev) *
i / dist;
511 int ch, sb, sf, nsubbands,
ret;
522 for (sb = 2; sb < nsubbands; sb++) {
537 for (sb = 0; sb <
s->nsubbands - 4; sb++) {
540 if (sb + 4 <
s->min_mono_subband)
543 s->grid_3_avg[ch2][sb] =
s->grid_3_avg[ch1][sb];
562 nsubbands = (
s->nsubbands -
s->min_mono_subband + 3) / 4;
563 for (sb = 0; sb < nsubbands; sb++)
564 for (ch = ch1; ch <= ch2; ch++)
565 for (sf = 1; sf <= 4; sf++)
569 s->part_stereo_pres |= 1 << ch1;
579 int sb, nsubbands,
ret;
583 for (sb = 2; sb < nsubbands; sb++) {
592 for (sb = 0; sb <
s->nsubbands - 4; sb++) {
593 if (sb + 4 >=
s->min_mono_subband) {
607 for (ch = ch1; ch <= ch2; ch++) {
608 if ((ch != ch1 && sb + 4 >=
s->min_mono_subband) !=
flag)
611 if (
s->grid_3_pres[ch] & (1
U << sb))
614 for (
i = 0;
i < 8;
i++) {
621 s->grid_3_pres[ch] |= 1
U << sb;
627 s->lbr_rand = 1103515245
U *
s->lbr_rand + 12345
U;
628 return s->lbr_rand *
s->sb_scf[sb];
636 float *
samples =
s->time_samples[ch][sb];
637 int i, j,
code, nblocks, coding_method;
644 switch (quant_level) {
649 for (j = 0; j < 8; j++)
667 for (j = 0; j < 5; j++)
678 for (j = 0; j < 3; j++)
691 for (
i = 0;
i < nblocks;
i++)
705 s->ch_pres[ch] |= 1
U << sb;
709 int start_sb,
int end_sb,
int flag)
711 int sb, sb_g3, sb_reorder, quant_level;
713 for (sb = start_sb; sb < end_sb; sb++) {
717 }
else if (
flag && sb < s->max_mono_subband) {
718 sb_reorder =
s->sb_indices[sb];
722 sb_reorder =
get_bits(&
s->gb,
s->limited_range + 3);
725 s->sb_indices[sb] = sb_reorder;
727 if (sb_reorder >=
s->nsubbands)
732 for (sb_g3 = 0; sb_g3 <
s->g3_avg_only_start_sb - 4; sb_g3++)
734 }
else if (sb < 12 && sb_reorder >= 4) {
742 if (!
flag || sb_reorder >=
s->max_mono_subband)
743 s->sec_ch_sbms[ch1 / 2][sb_reorder] =
get_bits(&
s->gb, 8);
744 if (
flag && sb_reorder >=
s->min_mono_subband)
745 s->sec_ch_lrms[ch1 / 2][sb_reorder] =
get_bits(&
s->gb, 8);
748 quant_level =
s->quant_levels[ch1 / 2][sb];
753 if (sb < s->max_mono_subband && sb_reorder >=
s->min_mono_subband) {
755 parse_ch(
s, ch1, sb_reorder, quant_level, 0);
757 parse_ch(
s, ch2, sb_reorder, quant_level, 1);
759 parse_ch(
s, ch1, sb_reorder, quant_level, 0);
761 parse_ch(
s, ch2, sb_reorder, quant_level, 0);
775 for (
i = 0;
i < 8;
i++) {
777 for (j = 0; j < (
i + 1) / 2; j++) {
778 float tmp1 =
coeff[ j ];
779 float tmp2 =
coeff[
i - j - 1];
780 coeff[ j ] = tmp1 + rc * tmp2;
781 coeff[
i - j - 1] = tmp2 + rc * tmp1;
789 int f =
s->framenum & 1;
790 int i, sb, ch, codes[16];
793 for (sb = start_sb; sb < end_sb; sb++) {
794 int ncodes = 8 * (1 + (sb < 2));
795 for (ch = ch1; ch <= ch2; ch++) {
798 for (
i = 0;
i < ncodes;
i++)
800 for (
i = 0;
i < ncodes / 8;
i++)
830 for (sb = 0; sb <
s->nsubbands; sb++) {
831 int f = sb *
s->limited_rate /
s->nsubbands;
832 int a = 18000 / (12 *
f / 1000 + 100 + 40 * st) + 20 * ol;
834 quant_levels[sb] = 1;
836 quant_levels[sb] = 2;
838 quant_levels[sb] = 3;
840 quant_levels[sb] = 4;
842 quant_levels[sb] = 5;
846 for (sb = 0; sb < 8; sb++)
848 for (; sb <
s->nsubbands; sb++)
849 s->quant_levels[ch1 / 2][sb] = quant_levels[sb];
862 for (sb = 0; sb < 2; sb++)
863 for (ch = ch1; ch <= ch2; ch++)
871 int start_sb,
int end_sb,
int flag)
873 int i, j, sb, ch, nsubbands;
876 if (end_sb > nsubbands)
879 for (sb = start_sb; sb < end_sb; sb++) {
880 for (ch = ch1; ch <= ch2; ch++) {
881 uint8_t *g2_scf =
s->grid_2_scf[ch][sb];
885 memcpy(g2_scf,
s->grid_2_scf[ch1][sb], 64);
890 for (
i = 0;
i < 8;
i++, g2_scf += 8) {
892 memset(g2_scf, 0, 64 -
i * 8);
897 for (j = 0; j < 8; j++) {
903 memset(g2_scf, 0, 8);
940 if ((
ret =
parse_ts(
s, ch1, ch2, 6,
s->max_mono_subband, 0)) < 0)
948 if ((
ret =
parse_ts(
s, ch1, ch2,
s->min_mono_subband,
s->nsubbands, 1)) < 0)
955 double scale = (-1.0 / (1 << 17)) * sqrt(1 << (2 -
s->limited_range));
956 float scale_t =
scale;
957 int i, br_per_ch =
s->bit_rate_scaled /
s->nchannels_total;
967 for (
i = 0;
i < 32 <<
s->freq_range;
i++)
970 if (br_per_ch < 14000)
972 else if (br_per_ch < 32000)
973 scale = (br_per_ch - 14000) * (1.0 / 120000) + 0.85;
977 scale *= 1.0 / INT_MAX;
979 for (
i = 0;
i <
s->nsubbands;
i++) {
983 s->sb_scf[
i] = (
i - 1) * 0.25 * 0.785 *
scale;
988 s->lfe_scale = (16 <<
s->freq_range) * 0.0000078265894;
997 int nsamples = nchsamples *
s->nchannels *
s->nsubbands;
1007 for (ch = 0; ch <
s->nchannels; ch++) {
1008 for (sb = 0; sb <
s->nsubbands; sb++) {
1009 s->time_samples[ch][sb] = ptr;
1019 int old_rate =
s->sample_rate;
1020 int old_band_limit =
s->band_limit;
1021 int old_nchannels =
s->nchannels;
1023 unsigned int sr_code;
1026 sr_code = bytestream2_get_byte(gb);
1032 if (
s->sample_rate > 48000) {
1038 s->ch_mask = bytestream2_get_le16(gb);
1039 if (!(
s->ch_mask & 0x7)) {
1043 if ((
s->ch_mask & 0xfff0) && !(
s->warned & 1)) {
1049 version = bytestream2_get_le16(gb);
1050 if ((
version & 0xff00) != 0x0800) {
1056 s->flags = bytestream2_get_byte(gb);
1062 if (!(
s->warned & 2)) {
1070 bit_rate_hi = bytestream2_get_byte(gb);
1073 s->bit_rate_orig = bytestream2_get_le16(gb) | ((bit_rate_hi & 0x0F) << 16);
1076 s->bit_rate_scaled = bytestream2_get_le16(gb) | ((bit_rate_hi & 0xF0) << 12);
1102 if (
s->bit_rate_orig >= 44000 * (
s->nchannels_total + 2))
1104 else if (
s->bit_rate_orig >= 25000 * (
s->nchannels_total + 2))
1110 s->limited_rate =
s->sample_rate >>
s->band_limit;
1111 s->limited_range =
s->freq_range -
s->band_limit;
1112 if (
s->limited_range < 0) {
1117 s->nsubbands = 8 <<
s->limited_range;
1120 if (
s->g3_avg_only_start_sb >
s->nsubbands)
1121 s->g3_avg_only_start_sb =
s->nsubbands;
1123 s->min_mono_subband =
s->nsubbands * 2000 / (
s->limited_rate / 2);
1124 if (
s->min_mono_subband >
s->nsubbands)
1125 s->min_mono_subband =
s->nsubbands;
1127 s->max_mono_subband =
s->nsubbands * 14000 / (
s->limited_rate / 2);
1128 if (
s->max_mono_subband >
s->nsubbands)
1129 s->max_mono_subband =
s->nsubbands;
1132 if ((old_rate !=
s->sample_rate || old_band_limit !=
s->band_limit) &&
init_sample_rate(
s) < 0)
1151 s->nchannels_total += 2;
1158 if (old_rate !=
s->sample_rate
1159 || old_band_limit !=
s->band_limit
1160 || old_nchannels !=
s->nchannels) {
1183 int i, ch, sb, sf,
ret, group, chunk_id, chunk_len;
1194 switch (bytestream2_get_byte(&gb)) {
1196 if (!
s->sample_rate) {
1213 chunk_id = bytestream2_get_byte(&gb);
1214 chunk_len = (chunk_id & 0x80) ? bytestream2_get_be16(&gb) : bytestream2_get_byte(&gb);
1225 switch (chunk_id & 0x7f) {
1228 int checksum = bytestream2_get_be16(&gb);
1229 uint16_t res = chunk_id;
1230 res += (chunk_len >> 8) & 0xff;
1231 res += chunk_len & 0xff;
1232 for (
i = 0;
i < chunk_len - 2;
i++)
1234 if (checksum != res) {
1251 memset(
s->quant_levels, 0,
sizeof(
s->quant_levels));
1252 memset(
s->sb_indices, 0xff,
sizeof(
s->sb_indices));
1253 memset(
s->sec_ch_sbms, 0,
sizeof(
s->sec_ch_sbms));
1254 memset(
s->sec_ch_lrms, 0,
sizeof(
s->sec_ch_lrms));
1255 memset(
s->ch_pres, 0,
sizeof(
s->ch_pres));
1256 memset(
s->grid_1_scf, 0,
sizeof(
s->grid_1_scf));
1257 memset(
s->grid_2_scf, 0,
sizeof(
s->grid_2_scf));
1258 memset(
s->grid_3_avg, 0,
sizeof(
s->grid_3_avg));
1259 memset(
s->grid_3_scf, 0,
sizeof(
s->grid_3_scf));
1260 memset(
s->grid_3_pres, 0,
sizeof(
s->grid_3_pres));
1261 memset(
s->tonal_scf, 0,
sizeof(
s->tonal_scf));
1262 memset(
s->lfe_data, 0,
sizeof(
s->lfe_data));
1263 s->part_stereo_pres = 0;
1264 s->framenum = (
s->framenum + 1) & 31;
1266 for (ch = 0; ch <
s->nchannels; ch++) {
1267 for (sb = 0; sb <
s->nsubbands / 4; sb++) {
1268 s->part_stereo[ch][sb][0] =
s->part_stereo[ch][sb][4];
1269 s->part_stereo[ch][sb][4] = 16;
1273 memset(
s->lpc_coeff[
s->framenum & 1], 0,
sizeof(
s->lpc_coeff[0]));
1275 for (group = 0; group < 5; group++) {
1276 for (sf = 0; sf < 1 << group; sf++) {
1277 int sf_idx = ((
s->framenum << group) + sf) & 31;
1278 s->tonal_bounds[group][sf_idx][0] =
1279 s->tonal_bounds[group][sf_idx][1] =
s->ntones;
1285 chunk_id = bytestream2_get_byte(&gb);
1286 chunk_len = (chunk_id & 0x80) ? bytestream2_get_be16(&gb) : bytestream2_get_byte(&gb);
1298 chunk.lfe.len = chunk_len;
1299 chunk.lfe.data = gb.
buffer;
1305 chunk.tonal.id = chunk_id;
1306 chunk.tonal.len = chunk_len;
1307 chunk.tonal.data = gb.
buffer;
1316 chunk.tonal_grp[
i].id =
i;
1317 chunk.tonal_grp[
i].len = chunk_len;
1318 chunk.tonal_grp[
i].data = gb.
buffer;
1327 chunk.tonal_grp[
i].id =
i;
1328 chunk.tonal_grp[
i].len = chunk_len;
1329 chunk.tonal_grp[
i].data = gb.
buffer;
1336 chunk.grid1[
i].len = chunk_len;
1337 chunk.grid1[
i].data = gb.
buffer;
1344 chunk.hr_grid[
i].len = chunk_len;
1345 chunk.hr_grid[
i].data = gb.
buffer;
1352 chunk.ts1[
i].len = chunk_len;
1353 chunk.ts1[
i].data = gb.
buffer;
1360 chunk.ts2[
i].len = chunk_len;
1361 chunk.ts2[
i].data = gb.
buffer;
1373 for (
i = 0;
i < 5;
i++)
1376 for (
i = 0;
i < (
s->nchannels + 1) / 2;
i++) {
1378 int ch2 =
FFMIN(ch1 + 1,
s->nchannels - 1);
1387 if (!chunk.grid1[
i].len || !chunk.hr_grid[
i].len || !chunk.ts1[
i].len)
1410 for (ch = ch1; ch <= ch2; ch++) {
1411 for (sb = 0; sb <
s->nsubbands; sb++) {
1414 uint8_t *g1_scf_a =
s->grid_1_scf[ch][g1_sb ];
1415 uint8_t *g1_scf_b =
s->grid_1_scf[ch][g1_sb + 1];
1420 uint8_t *hr_scf =
s->high_res_scf[ch][sb];
1423 for (
i = 0;
i < 8;
i++) {
1424 int scf = w1 * g1_scf_a[
i] + w2 * g1_scf_b[
i];
1425 hr_scf[
i] = scf >> 7;
1428 int8_t *g3_scf =
s->grid_3_scf[ch][sb - 4];
1429 int g3_avg =
s->grid_3_avg[ch][sb - 4];
1431 for (
i = 0;
i < 8;
i++) {
1432 int scf = w1 * g1_scf_a[
i] + w2 * g1_scf_b[
i];
1433 hr_scf[
i] = (scf >> 7) - g3_avg - g3_scf[
i];
1445 int i, j, k, ch, sb;
1447 for (ch = ch1; ch <= ch2; ch++) {
1448 for (sb = 0; sb <
s->nsubbands; sb++) {
1449 float *
samples =
s->time_samples[ch][sb];
1451 if (
s->ch_pres[ch] & (1
U << sb))
1457 }
else if (sb < 10) {
1462 float accum[8] = { 0 };
1465 for (k = 2; k < 6; k++) {
1466 float *other = &
s->time_samples[ch][k][
i * 8];
1467 for (j = 0; j < 8; j++)
1468 accum[j] +=
fabs(other[j]);
1471 for (j = 0; j < 8; j++)
1483 for (
i = 0;
i < nsamples;
i++) {
1485 for (j = 0; j < 8; j++)
1493 int f =
s->framenum & 1;
1496 for (ch = ch1; ch <= ch2; ch++) {
1497 float *
samples =
s->time_samples[ch][sb];
1499 if (!(
s->ch_pres[ch] & (1
U << sb)))
1517 for (sb = 0; sb <
s->nsubbands; sb++) {
1519 for (ch = ch1; ch <= ch2; ch++) {
1520 float *
samples =
s->time_samples[ch][sb];
1521 uint8_t *hr_scf =
s->high_res_scf[ch][sb];
1524 unsigned int scf = hr_scf[
i];
1527 for (j = 0; j < 16; j++)
1533 unsigned int scf = hr_scf[
i / 8] - g2_scf[
i];
1544 float *samples_l =
s->time_samples[ch1][sb];
1545 float *samples_r =
s->time_samples[ch2][sb];
1546 int ch2_pres =
s->ch_pres[ch2] & (1
U << sb);
1549 int sbms = (
s->sec_ch_sbms[ch1 / 2][sb] >>
i) & 1;
1550 int lrms = (
s->sec_ch_lrms[ch1 / 2][sb] >>
i) & 1;
1552 if (sb >=
s->min_mono_subband) {
1553 if (lrms && ch2_pres) {
1555 for (j = 0; j < 16; j++) {
1556 float tmp = samples_l[j];
1557 samples_l[j] = samples_r[j];
1558 samples_r[j] = -
tmp;
1561 for (j = 0; j < 16; j++) {
1562 float tmp = samples_l[j];
1563 samples_l[j] = samples_r[j];
1567 }
else if (!ch2_pres) {
1568 if (sbms && (
s->part_stereo_pres & (1 << ch1))) {
1569 for (j = 0; j < 16; j++)
1570 samples_r[j] = -samples_l[j];
1572 for (j = 0; j < 16; j++)
1573 samples_r[j] = samples_l[j];
1576 }
else if (sbms && ch2_pres) {
1577 for (j = 0; j < 16; j++) {
1578 float tmp = samples_l[j];
1579 samples_l[j] = (
tmp + samples_r[j]) * 0.5
f;
1580 samples_r[j] = (
tmp - samples_r[j]) * 0.5
f;
1602 for (ch = ch1; ch <= ch2; ch++) {
1603 for (sb =
s->min_mono_subband; sb < s->nsubbands; sb++) {
1604 uint8_t *pt_st =
s->part_stereo[ch][(sb -
s->min_mono_subband) / 4];
1605 float *
samples =
s->time_samples[ch][sb];
1607 if (
s->ch_pres[ch2] & (1
U << sb))
1610 for (sf = 1; sf <= 4; sf++,
samples += 32) {
1614 for (
i = 0;
i < 32;
i++)
1625 int group,
int group_sf,
int synth_idx)
1627 int i, start, count;
1632 start =
s->tonal_bounds[group][group_sf][0];
1633 count = (
s->tonal_bounds[group][group_sf][1] - start) & (
DCA_LBR_TONES - 1);
1635 for (
i = 0;
i < count;
i++) {
1668 values[x_freq - 5] += cf[ 0] * -
s;
1669 p4:
values[x_freq - 4] += cf[ 1] *
c;
1670 p3:
values[x_freq - 3] += cf[ 2] *
s;
1671 p2:
values[x_freq - 2] += cf[ 3] * -
c;
1672 p1:
values[x_freq - 1] += cf[ 4] * -
s;
1673 p0:
values[x_freq ] += cf[ 5] *
c;
1674 values[x_freq + 1] += cf[ 6] *
s;
1675 values[x_freq + 2] += cf[ 7] * -
c;
1676 values[x_freq + 3] += cf[ 8] * -
s;
1677 values[x_freq + 4] += cf[ 9] *
c;
1678 values[x_freq + 5] += cf[10] *
s;
1693 for (group = 0; group < 5; group++) {
1694 int group_sf = (
s->framenum << group) + ((sf - 22) >> (5 - group));
1695 int synth_idx = ((((sf - 22) & 31) << group) & 31) + (1 << group) - 1;
1706 int sf, sb, nsubbands =
s->nsubbands, noutsubbands = 8 <<
s->freq_range;
1709 if (nsubbands < noutsubbands)
1710 memset(
values[nsubbands], 0, (noutsubbands - nsubbands) *
sizeof(
values[0]));
1714 s->dcadsp->lbr_bank(
values,
s->time_samples[ch],
1723 s->history[ch], noutsubbands * 4);
1724 s->fdsp->vector_fmul_reverse(
s->history[ch],
result[noutsubbands],
1725 s->window, noutsubbands * 4);
1726 output += noutsubbands * 4;
1730 for (sb = 0; sb < nsubbands; sb++) {
1739 int i,
ret, nchannels, ch_conf = (
s->ch_mask & 0x7) - 1;
1740 const int8_t *reorder;
1765 for (
i = 0;
i < (
s->nchannels + 1) / 2;
i++) {
1767 int ch2 =
FFMIN(ch1 + 1,
s->nchannels - 1);
1775 if (ch1 != ch2 && (
s->part_stereo_pres & (1 << ch1)))
1778 if (ch1 < nchannels)
1781 if (ch1 != ch2 && ch2 < nchannels)
1789 s->lfe_history, 16 <<
s->freq_range);
1802 if (!
s->sample_rate)
1806 memset(
s->part_stereo, 16,
sizeof(
s->part_stereo));
1807 memset(
s->lpc_coeff, 0,
sizeof(
s->lpc_coeff));
1808 memset(
s->history, 0,
sizeof(
s->history));
1809 memset(
s->tonal_bounds, 0,
sizeof(
s->tonal_bounds));
1810 memset(
s->lfe_history, 0,
sizeof(
s->lfe_history));
1814 for (ch = 0; ch <
s->nchannels; ch++) {
1815 for (sb = 0; sb <
s->nsubbands; sb++) {
const uint8_t ff_dca_grid_2_to_scf[3]
@ AV_SAMPLE_FMT_FLTP
float, planar
static void skip_bits_long(GetBitContext *s, int n)
Skips the specified number of bits.
#define AV_LOG_WARNING
Something somehow does not look correct.
#define AV_EF_EXPLODE
abort decoding on minor error detection
#define AV_EF_CAREFUL
consider things that violate the spec, are fast to calculate and have not been seen in the wild as er...
static int get_bits_left(GetBitContext *gb)
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
#define DCA_RSD_AMP_VLC_BITS
const uint8_t ff_dca_grid_1_weights[12][32]
#define DCA_GRID_VLC_BITS
const uint8_t ff_dca_sb_reorder[8][8]
int sample_rate
samples per second
const float ff_dca_rsd_level_2b[2]
static void parse_ch(DCALbrDecoder *s, int ch, int sb, int quant_level, int flag)
Parse time samples for one subband, filling truncated samples with randomness.
static int ensure_bits(GetBitContext *s, int n)
Check point to ensure that enough bits are left.
const float ff_dca_lfe_step_size_24[144]
#define AV_CH_LAYOUT_MONO
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 output
static void filter_ts(DCALbrDecoder *s, int ch1, int ch2)
static const uint8_t lfe_index[7]
#define AV_PROFILE_DTS_EXPRESS
This structure describes decoded (raw) audio or video data.
trying all byte sequences megabyte in length and selecting the best looking sequence will yield cases to try But a word about which is also called distortion Distortion can be quantified by almost any quality measurement one chooses the sum of squared differences is used but more complex methods that consider psychovisual effects can be used as well It makes no difference in this discussion First step
const float ff_dca_quant_amp[57]
#define DCA_AVG_G3_VLC_BITS
@ LBR_CHUNK_RES_TS_2_LAST
static void synth_tones(DCALbrDecoder *s, int ch, float *values, int group, int group_sf, int synth_idx)
Synthesise tones in the given group for the given tonal subframe.
#define DCA_SPEAKER_LAYOUT_STEREO
int request_channel_layout
Converted from avctx.request_channel_layout.
uint8_t phs[DCA_LBR_CHANNELS]
Per-channel phase.
const uint8_t ff_dca_scf_to_grid_1[32]
@ LBR_FLAG_BAND_LIMIT_1_8
#define DCA_ST_GRID_VLC_BITS
uint8_t x_freq
Spectral line offset.
av_cold int av_tx_init(AVTXContext **ctx, av_tx_fn *tx, enum AVTXType type, int inv, int len, const void *scale, uint64_t flags)
Initialize a transform context with the given configuration (i)MDCTs with an odd length are currently...
static void decode_grid(DCALbrDecoder *s, int ch1, int ch2)
Reconstruct high-frequency resolution grid from first and third grids.
int ff_dca_lbr_parse(DCALbrDecoder *s, const uint8_t *data, DCAExssAsset *asset)
int lbr_offset
Offset to LBR component from start of substream.
static av_always_inline void bytestream2_skip(GetByteContext *g, unsigned int size)
static unsigned int get_bits(GetBitContext *s, int n)
Read 1-25 bits.
static void synth_lpc(DCALbrDecoder *s, int ch1, int ch2, int sb)
AVChannelLayout ch_layout
Audio channel layout.
uint8_t amp[DCA_LBR_CHANNELS]
Per-channel amplitude.
const float ff_dca_rsd_level_5[5]
static av_always_inline float scale(float x, float s)
#define AV_CH_LAYOUT_STEREO
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
#define FF_ARRAY_ELEMS(a)
int ff_side_data_update_matrix_encoding(AVFrame *frame, enum AVMatrixEncoding matrix_encoding)
Add or update AV_FRAME_DATA_MATRIXENCODING side data.
static int alloc_sample_buffer(DCALbrDecoder *s)
static int init_get_bits8(GetBitContext *s, const uint8_t *buffer, int byte_size)
Initialize GetBitContext.
static int parse_st_code(GetBitContext *s, int min_v)
#define AV_CH_LOW_FREQUENCY
av_cold void ff_dca_lbr_init_tables(void)
@ AV_TX_FLOAT_MDCT
Standard MDCT with a sample data type of float, double or int32_t, respecively.
@ LBR_FLAG_BAND_LIMIT_1_2
static int parse_lpc(DCALbrDecoder *s, int ch1, int ch2, int start_sb, int end_sb)
static int parse_ts(DCALbrDecoder *s, int ch1, int ch2, int start_sb, int end_sb, int flag)
FF_ENABLE_DEPRECATION_WARNINGS int av_channel_layout_from_mask(AVChannelLayout *channel_layout, uint64_t mask)
Initialize a native channel layout from a bitmask indicating which channels are present.
const float ff_dca_rsd_level_16[16]
#define av_assert0(cond)
assert() equivalent, that is always enabled.
static int init_sample_rate(DCALbrDecoder *s)
int bits_per_raw_sample
Bits per sample/pixel of internal libavcodec pixel/sample format.
static int parse_grid_1_chunk(DCALbrDecoder *s, LBRChunk *chunk, int ch1, int ch2)
@ LBR_FLAG_BAND_LIMIT_2_3
@ LBR_CHUNK_FRAME_NO_CSUM
@ LBR_FLAG_BAND_LIMIT_1_3
const int8_t ff_dca_lfe_delta_index_16[8]
#define DCA_DAMP_VLC_BITS
@ AV_TX_FULL_IMDCT
Performs a full inverse MDCT rather than leaving out samples that can be derived through symmetry.
static void random_ts(DCALbrDecoder *s, int ch1, int ch2)
Fill unallocated subbands with randomness.
int ff_dca_lbr_filter_frame(DCALbrDecoder *s, AVFrame *frame)
const float ff_dca_synth_env[32]
and forward the result(frame or status change) to the corresponding input. If nothing is possible
static __device__ float fabs(float a)
#define LOCAL_ALIGNED_32(t, v,...)
#define AVERROR_PATCHWELCOME
Not yet implemented in FFmpeg, patches welcome.
int lbr_size
Size of LBR component in extension substream.
#define DCA_RSD_APPRX_VLC_BITS
static int parse_ts1_chunk(DCALbrDecoder *s, LBRChunk *chunk, int ch1, int ch2)
@ LBR_FLAG_BAND_LIMIT_MASK
int64_t bit_rate
the average bitrate
const uint16_t ff_dca_avg_g3_freqs[3]
const float ff_dca_corr_cf[32][11]
static unsigned int get_bits1(GetBitContext *s)
void av_fast_mallocz(void *ptr, unsigned int *size, size_t min_size)
Allocate and clear a buffer, reusing the given one if large enough.
const float ff_dca_lfe_iir[5][4]
@ LBR_FLAG_BAND_LIMIT_NONE
static int parse_lfe_chunk(DCALbrDecoder *s, LBRChunk *chunk)
#define AV_CH_FRONT_CENTER
static void transform_channel(DCALbrDecoder *s, int ch, float *output)
#define DCA_LBR_CHANNELS_TOTAL
#define AV_EF_CRCCHECK
Verify checksums embedded in the bitstream (could be of either encoded or decoded data,...
static av_always_inline int get_vlc2(GetBitContext *s, const VLCElem *table, int bits, int max_depth)
Parse a vlc code.
Undefined Behavior In the C some operations are like signed integer dereferencing freed accessing outside allocated Undefined Behavior must not occur in a C it is not safe even if the output of undefined operations is unused The unsafety may seem nit picking but Optimizing compilers have in fact optimized code on the assumption that no undefined Behavior occurs Optimizing code based on wrong assumptions can and has in some cases lead to effects beyond the output of computations The signed integer overflow problem in speed critical code Code which is highly optimized and works with signed integers sometimes has the problem that often the output of the computation does not c
#define DCA_FST_RSD_VLC_BITS
@ LBR_CHUNK_RES_GRID_HR_LAST
static av_always_inline int bytestream2_get_bytes_left(GetByteContext *g)
const float ff_dca_bank_coeff[10]
@ DCA_LBR_HEADER_SYNC_ONLY
uint8_t ph_rot
Phase rotation.
static int parse_lfe_24(DCALbrDecoder *s)
int ff_get_buffer(AVCodecContext *avctx, AVFrame *frame, int flags)
Get a buffer for a frame.
static int shift(int a, int b)
static float lbr_rand(DCALbrDecoder *s, int sb)
@ AV_MATRIX_ENCODING_NONE
enum AVSampleFormat sample_fmt
audio sample format
static const float lpc_tab[16]
void avpriv_report_missing_feature(void *avc, const char *msg,...) av_printf_format(2
Log a generic warning message about a missing feature.
static int parse_tonal_group(DCALbrDecoder *s, LBRChunk *chunk)
const float ff_dca_rsd_level_3[3]
static av_always_inline int diff(const struct color_info *a, const struct color_info *b, const int trans_thresh)
The reader does not expect b to be semantically here and if the code is changed by maybe adding a a division or other the signedness will almost certainly be mistaken To avoid this confusion a new type was SUINT is the C unsigned type but it holds a signed int to use the same example SUINT a
@ LBR_CHUNK_TONAL_SCF_GRP_3
@ DCA_LBR_HEADER_DECODER_INIT
uint8_t f_delt
Difference between original and center frequency.
VLC ff_dca_vlc_tnl_grp[5]
av_cold void av_tx_uninit(AVTXContext **ctx)
Frees a context and sets *ctx to NULL, does nothing when *ctx == NULL.
const float ff_dca_rsd_level_8[8]
#define AV_CH_LAYOUT_5POINT0
const float ff_dca_lfe_step_size_16[101]
const uint8_t ff_dca_scf_to_grid_2[32]
const float ff_dca_rsd_level_2a[2]
int nb_samples
number of audio samples (per channel) described by this frame
#define i(width, name, range_min, range_max)
and forward the test the status of outputs and forward it to the corresponding return FFERROR_NOT_READY If the filters stores internally one or a few frame for some it can consider them to be part of the FIFO and delay acknowledging a status change accordingly Example code
uint8_t ** extended_data
pointers to the data planes/channels.
const uint8_t ff_dca_rsd_pack_3_in_7[128][3]
static void base_func_synth(DCALbrDecoder *s, int ch, float *values, int sf)
Synthesise all tones in all groups for the given residual subframe.
VLC ff_dca_vlc_fst_rsd_amp
static void decode_part_stereo(DCALbrDecoder *s, int ch1, int ch2)
Modulate by interpolated partial stereo coefficients.
@ LBR_FLAG_BAND_LIMIT_1_4
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 default value
#define DCA_TNL_GRP_VLC_BITS
av_cold int ff_dca_lbr_init(DCALbrDecoder *s)
const uint8_t ff_dca_freq_to_sb[32]
@ LBR_CHUNK_RES_GRID_LR_LAST
static int parse_scale_factors(DCALbrDecoder *s, uint8_t *scf)
static const int8_t channel_reorder_nolfe[7][5]
static void predict(float *samples, const float *coeff, int nsamples)
static int parse_high_res_grid(DCALbrDecoder *s, LBRChunk *chunk, int ch1, int ch2)
av_cold void ff_dca_lbr_flush(DCALbrDecoder *s)
av_cold void ff_dca_lbr_close(DCALbrDecoder *s)
static int parse_ts2_chunk(DCALbrDecoder *s, LBRChunk *chunk, int ch1, int ch2)
#define AV_CH_LAYOUT_SURROUND
static int parse_decoder_init(DCALbrDecoder *s, GetByteContext *gb)
static const int8_t channel_reorder_lfe[7][5]
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
static void parse_grid_3(DCALbrDecoder *s, int ch1, int ch2, int sb, int flag)
main external API structure.
static int parse_grid_1_sec_ch(DCALbrDecoder *s, int ch2)
const uint32_t ff_dca_sampling_freqs[16]
static int parse_tonal_chunk(DCALbrDecoder *s, LBRChunk *chunk)
void av_channel_layout_uninit(AVChannelLayout *channel_layout)
Free any allocated data in the channel layout and reset the channel count to 0.
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 values
Filter the word “frame” indicates either a video frame or a group of audio samples
const uint8_t ff_dca_grid_1_to_scf[11]
@ LBR_CHUNK_TONAL_SCF_GRP_2
const float ff_dca_long_window[128]
static void convert_lpc(float *coeff, const int *codes)
Convert from reflection coefficients to direct form coefficients.
const uint8_t ff_dca_freq_ranges[16]
static av_always_inline int get_bitsz(GetBitContext *s, int n)
Read 0-25 bits.
const uint16_t ff_dca_fst_amp[44]
static int parse_tonal(DCALbrDecoder *s, int group)
static const uint16_t channel_layouts[7]
static int parse_vlc(GetBitContext *s, const VLC *vlc, int nb_bits, int max_depth)
av_cold AVFloatDSPContext * avpriv_float_dsp_alloc(int bit_exact)
Allocate a float DSP context.
static av_always_inline void bytestream2_init(GetByteContext *g, const uint8_t *buf, int buf_size)
@ LBR_CHUNK_RES_TS_1_LAST
static const double coeff[2][5]
const int8_t ff_dca_ph0_shift[8]
#define AVERROR_INVALIDDATA
Invalid data found when processing input.
#define DCA_LBR_TIME_HISTORY
static float cos_tab[256]
@ LBR_CHUNK_TONAL_SCF_GRP_5
#define DCA_TNL_SCF_VLC_BITS
#define DCA_LBR_TIME_SAMPLES
@ LBR_CHUNK_TONAL_SCF_GRP_1
static int ff_dca_count_chs_for_mask(unsigned int mask)
Return number of individual channels in DCASpeakerPair mask.
@ LBR_CHUNK_TONAL_SCF_GRP_4
const uint16_t ff_dca_rsd_pack_5_in_8[256]
static int parse_lfe_16(DCALbrDecoder *s)
static int parse_grid_2(DCALbrDecoder *s, int ch1, int ch2, int start_sb, int end_sb, int flag)
const float ff_dca_st_coeff[34]
const int8_t ff_dca_lfe_delta_index_24[32]