44 #define PSY_3GPP_THR_SPREAD_HI 1.5f // spreading factor for low-to-hi threshold spreading (15 dB/Bark) 45 #define PSY_3GPP_THR_SPREAD_LOW 3.0f // spreading factor for hi-to-low threshold spreading (30 dB/Bark) 47 #define PSY_3GPP_EN_SPREAD_HI_L1 2.0f 49 #define PSY_3GPP_EN_SPREAD_HI_L2 1.5f 51 #define PSY_3GPP_EN_SPREAD_HI_S 1.5f 53 #define PSY_3GPP_EN_SPREAD_LOW_L 3.0f 55 #define PSY_3GPP_EN_SPREAD_LOW_S 2.0f 57 #define PSY_3GPP_RPEMIN 0.01f 58 #define PSY_3GPP_RPELEV 2.0f 60 #define PSY_3GPP_C1 3.0f 61 #define PSY_3GPP_C2 1.3219281f 62 #define PSY_3GPP_C3 0.55935729f 64 #define PSY_SNR_1DB 7.9432821e-1f 65 #define PSY_SNR_25DB 3.1622776e-3f 67 #define PSY_3GPP_SAVE_SLOPE_L -0.46666667f 68 #define PSY_3GPP_SAVE_SLOPE_S -0.36363637f 69 #define PSY_3GPP_SAVE_ADD_L -0.84285712f 70 #define PSY_3GPP_SAVE_ADD_S -0.75f 71 #define PSY_3GPP_SPEND_SLOPE_L 0.66666669f 72 #define PSY_3GPP_SPEND_SLOPE_S 0.81818181f 73 #define PSY_3GPP_SPEND_ADD_L -0.35f 74 #define PSY_3GPP_SPEND_ADD_S -0.26111111f 75 #define PSY_3GPP_CLIP_LO_L 0.2f 76 #define PSY_3GPP_CLIP_LO_S 0.2f 77 #define PSY_3GPP_CLIP_HI_L 0.95f 78 #define PSY_3GPP_CLIP_HI_S 0.75f 80 #define PSY_3GPP_AH_THR_LONG 0.5f 81 #define PSY_3GPP_AH_THR_SHORT 0.63f 83 #define PSY_PE_FORGET_SLOPE 511 91 #define PSY_3GPP_BITS_TO_PE(bits) ((bits) * 1.18f) 92 #define PSY_3GPP_PE_TO_BITS(bits) ((bits) / 1.18f) 95 #define PSY_LAME_FIR_LEN 21 96 #define AAC_BLOCK_SIZE_LONG 1024
97 #define AAC_BLOCK_SIZE_SHORT 128
98 #define AAC_NUM_BLOCKS_SHORT 8
99 #define PSY_LAME_NUM_SUBBLOCKS 3
220 -8.65163e-18 * 2, -0.00851586 * 2, -6.74764e-18 * 2, 0.0209036 * 2,
221 -3.36639e-17 * 2, -0.0438162 * 2, -1.54175e-17 * 2, 0.0931738 * 2,
222 -5.52212e-17 * 2, -0.313819 * 2
235 int lower_range = 12, upper_range = 12;
236 int lower_range_kbps = psy_abr_map[12].
quality;
237 int upper_range_kbps = psy_abr_map[12].
quality;
243 for (i = 1; i < 13; i++) {
244 if (
FFMAX(bitrate, psy_abr_map[i].quality) != bitrate) {
246 upper_range_kbps = psy_abr_map[i ].
quality;
248 lower_range_kbps = psy_abr_map[i - 1].
quality;
254 if ((upper_range_kbps - bitrate) > (bitrate - lower_range_kbps))
255 return psy_abr_map[lower_range].
st_lrm;
256 return psy_abr_map[upper_range].
st_lrm;
266 for (i = 0; i < avctx->
channels; i++) {
284 return 13.3f *
atanf(0.00076f * f) + 3.5f *
atanf((f / 7500.0f) * (f / 7500.0f));
295 return 3.64 * pow(f, -0.8)
296 - 6.8 *
exp(-0.6 * (f - 3.4) * (f - 3.4))
297 + 6.0 *
exp(-0.15 * (f - 8.7) * (f - 8.7))
298 + (0.6 + 0.04 * add) * 0.001 * f * f * f * f;
305 float prev, minscale, minath, minsnr, pe_min;
309 const float num_bark =
calc_bark((
float)bandwidth);
330 for (j = 0; j < 2; j++) {
334 float avg_chan_bits = chan_bitrate * (j ? 128.0f : 1024.0f) / ctx->
avctx->
sample_rate;
343 for (g = 0; g < ctx->
num_bands[j]; g++) {
345 bark =
calc_bark((i-1) * line_to_frequency);
346 coeffs[
g].
barks = (bark + prev) / 2.0;
349 for (g = 0; g < ctx->
num_bands[j] - 1; g++) {
351 float bark_width = coeffs[g+1].
barks - coeffs->
barks;
356 pe_min = bark_pe * bark_width;
357 minsnr =
exp2(pe_min / band_sizes[g]) - 1.5f;
361 for (g = 0; g < ctx->
num_bands[j]; g++) {
362 minscale =
ath(start * line_to_frequency,
ATH_ADD);
363 for (i = 1; i < band_sizes[
g]; i++)
364 minscale =
FFMIN(minscale,
ath((start + i) * line_to_frequency,
ATH_ADD));
365 coeffs[
g].
ath = minscale - minath;
366 start += band_sizes[
g];
388 ret = 0.7548f * (in - state[0]) + 0.5095f * state[1];
398 0xB6, 0x6C, 0xD8, 0xB2, 0x66, 0xC6, 0x96, 0x36, 0x36
406 const int16_t *audio,
412 int attack_ratio = br <= 16000 ? 18 : 10;
416 int next_type = pch->next_window_seq;
421 int switch_to_eight = 0;
422 float sum = 0.0, sum2 = 0.0;
425 for (i = 0; i < 8; i++) {
426 for (j = 0; j < 128; j++) {
433 for (i = 0; i < 8; i++) {
434 if (s[i] > pch->win_energy * attack_ratio) {
440 pch->win_energy = pch->win_energy*7/8 + sum2/64;
442 wi.window_type[1] = prev_type;
450 grouping = pch->next_grouping;
466 pch->next_window_seq = next_type;
468 for (i = 0; i < 3; i++)
469 wi.window_type[i] = prev_type;
480 for (i = 0; i < 8; i++) {
481 if (!((grouping >> i) & 1))
483 wi.grouping[lastgrp]++;
500 float clipped_pe, bit_save, bit_spend, bit_factor, fill_level, forgetful_min_pe;
504 fill_level = av_clipf((
float)ctx->
fill_level / size, clip_low, clip_high);
505 clipped_pe = av_clipf(pe, ctx->
pe.
min, ctx->
pe.
max);
506 bit_save = (fill_level + bitsave_add) * bitsave_slope;
507 assert(bit_save <= 0.3f && bit_save >= -0.05000001f);
508 bit_spend = (fill_level + bitspend_add) * bitspend_slope;
509 assert(bit_spend <= 0.5f && bit_spend >= -0.1f);
516 bit_factor = 1.0f - bit_save + ((bit_spend - bit_save) / (ctx->
pe.
max - ctx->
pe.
min)) * (clipped_pe - ctx->
pe.
min);
560 float thr_avg, reduction;
562 if(active_lines == 0.0)
565 thr_avg =
exp2f((a - pe) / (4.0f * active_lines));
566 reduction =
exp2f((a - desired_pe) / (4.0f * active_lines)) - thr_avg;
568 return FFMAX(reduction, 0.0f);
578 thr = sqrtf(thr) + reduction;
596 #ifndef calc_thr_3gpp 598 const uint8_t *band_sizes,
const float *coefs,
const int cutoff)
601 int start = 0, wstart = 0;
604 for (g = 0; g < num_bands; g++) {
607 float form_factor = 0.0f;
610 if (wstart < cutoff) {
611 for (i = 0; i < band_sizes[
g]; i++) {
612 band->
energy += coefs[start+i] * coefs[start+i];
613 form_factor += sqrtf(fabs(coefs[start+i]));
616 Temp = band->
energy > 0 ? sqrtf((
float)band_sizes[g] / band->
energy) : 0;
618 band->
nz_lines = form_factor * sqrtf(Temp);
620 start += band_sizes[
g];
621 wstart += band_sizes[
g];
627 #ifndef psy_hp_filter 636 sum1 += psy_fir_coeffs[j] * (firbuf[i + j] + firbuf[i +
PSY_LAME_FIR_LEN - j]);
637 sum2 += psy_fir_coeffs[j + 1] * (firbuf[i + j + 1] + firbuf[i +
PSY_LAME_FIR_LEN - j - 1]);
641 hpfsmpl[i] = (sum1 + sum2) * 32768.0f;
655 float desired_bits, desired_pe, delta_pe, reduction=
NAN, spread_en[128] = {0};
657 float pe = pctx->chan_bitrate > 32000 ? 0.0f :
FFMAX(50.0f, 100.0f - pctx->chan_bitrate * 100.0f / 32000.0f);
658 const int num_bands =
ctx->num_bands[wi->num_windows == 8];
659 const uint8_t *band_sizes =
ctx->bands[wi->num_windows == 8];
663 const int cutoff = bandwidth * 2048 / wi->num_windows /
ctx->avctx->sample_rate;
666 calc_thr_3gpp(wi, num_bands, pch, band_sizes, coefs, cutoff);
669 for (
w = 0;
w < wi->num_windows*16;
w += 16) {
673 spread_en[0] = bands[0].
energy;
674 for (
g = 1;
g < num_bands;
g++) {
675 bands[
g].
thr =
FFMAX(bands[
g].
thr, bands[
g-1].thr * coeffs[
g].spread_hi[0]);
676 spread_en[
w+
g] =
FFMAX(bands[
g].
energy, spread_en[
w+
g-1] * coeffs[
g].spread_hi[1]);
678 for (
g = num_bands - 2;
g >= 0;
g--) {
679 bands[
g].
thr =
FFMAX(bands[
g].
thr, bands[
g+1].thr * coeffs[
g].spread_low[0]);
680 spread_en[
w+
g] =
FFMAX(spread_en[
w+
g], spread_en[
w+
g+1] * coeffs[
g].spread_low[1]);
683 for (
g = 0;
g < num_bands;
g++) {
698 if (spread_en[
w+
g] * avoid_hole_thr > band->
energy || coeffs[
g].
min_snr > 1.0f)
711 desired_pe = pe * (
ctx->avctx->global_quality ?
ctx->avctx->global_quality : 120) / (2 * 2.5f * 120.0f);
716 if (
ctx->bitres.bits > 0) {
721 pctx->pe.max =
FFMAX(pe, pctx->pe.max);
722 pctx->pe.min =
FFMIN(pe, pctx->pe.min);
731 if (
ctx->bitres.bits > 0)
736 ctx->bitres.alloc = desired_bits;
738 if (desired_pe < pe) {
740 for (
w = 0;
w < wi->num_windows*16;
w += 16) {
745 for (
g = 0;
g < num_bands;
g++) {
757 for (i = 0; i < 2; i++) {
758 float pe_no_ah = 0.0f, desired_pe_no_ah;
760 for (
w = 0;
w < wi->num_windows*16;
w += 16) {
761 for (
g = 0;
g < num_bands;
g++) {
765 pe_no_ah += band->
pe;
771 desired_pe_no_ah =
FFMAX(desired_pe - (pe - pe_no_ah), 0.0f);
776 for (
w = 0;
w < wi->num_windows*16;
w += 16) {
777 for (
g = 0;
g < num_bands;
g++) {
783 if (band->
thr > 0.0f)
790 delta_pe = desired_pe -
pe;
791 if (fabs(delta_pe) > 0.05f * desired_pe)
795 if (pe < 1.15f * desired_pe) {
798 for (
w = 0;
w < wi->num_windows*16;
w += 16) {
799 for (
g = 0;
g < num_bands;
g++) {
816 while (pe > desired_pe &&
g--) {
817 for (
w = 0;
w < wi->num_windows*16;
w+= 16) {
830 for (
w = 0;
w < wi->num_windows*16;
w += 16) {
831 for (
g = 0;
g < num_bands;
g++) {
842 memcpy(pch->prev_band, pch->band,
sizeof(pch->band));
851 for (ch = 0; ch < group->
num_ch; ch++)
881 const float *la,
int channel,
int prev_type)
886 int uselongblock = 1;
893 const float *pf = hpfsmpl;
905 energy_subshort[i] = pch->prev_energy_subshort[i + ((
AAC_NUM_BLOCKS_SHORT - 1) * PSY_LAME_NUM_SUBBLOCKS)];
906 assert(pch->prev_energy_subshort[i + ((
AAC_NUM_BLOCKS_SHORT - 2) * PSY_LAME_NUM_SUBBLOCKS + 1)] > 0);
907 attack_intensity[i] = energy_subshort[i] / pch->prev_energy_subshort[i + ((
AAC_NUM_BLOCKS_SHORT - 2) * PSY_LAME_NUM_SUBBLOCKS + 1)];
908 energy_short[0] += energy_subshort[i];
914 for (; pf < pfe; pf++)
915 p =
FFMAX(p, fabsf(*pf));
925 if (p > energy_subshort[i + 1])
926 p = p / energy_subshort[i + 1];
927 else if (energy_subshort[i + 1] > p * 10.0f)
928 p = energy_subshort[i + 1] / (p * 10.0f);
936 if (!attacks[i / PSY_LAME_NUM_SUBBLOCKS])
937 if (attack_intensity[i] > pch->attack_threshold)
945 const float u = energy_short[i - 1];
946 const float v = energy_short[i];
947 const float m =
FFMAX(u, v);
949 if (u < 1.7f * v && v < 1.7f * u) {
950 if (i == 1 && attacks[0] < attacks[i])
955 att_sum += attacks[i];
958 if (attacks[0] <= pch->prev_attack)
961 att_sum += attacks[0];
963 if (pch->prev_attack == 3 || att_sum) {
966 for (i = 1; i < AAC_NUM_BLOCKS_SHORT + 1; i++)
967 if (attacks[i] && attacks[i-1])
992 for (i = 0; i < 8; i++) {
993 if (!((pch->next_grouping >> i) & 1))
1005 for (i = 0; i < 9; i++) {
1013 pch->prev_attack = attacks[8];
1020 .
name =
"3GPP TS 26.403-inspired model",
int quality
Quality to map the rest of the vaules to.
float global_quality
normalized global quality taken from avctx
static const uint8_t window_grouping[9]
window grouping information stored as bits (0 - new group, 1 - group continues)
int grouping[8]
window grouping (for e.g. AAC)
#define AAC_BLOCK_SIZE_SHORT
short block size
static int calc_bit_demand(AacPsyContext *ctx, float pe, int bits, int size, int short_window)
uint8_t ** bands
scalefactor band sizes for possible frame sizes
#define PSY_3GPP_AH_THR_SHORT
int64_t bit_rate
the average bitrate
static const PsyLamePreset psy_vbr_map[]
LAME psy model preset table for constant quality.
psychoacoustic information for an arbitrary group of channels
uint8_t pi<< 24) CONV_FUNC(AV_SAMPLE_FMT_S64, int64_t, AV_SAMPLE_FMT_U8,(uint64_t)((*(const uint8_t *) pi - 0x80U))<< 56) CONV_FUNC(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_U8,(*(const uint8_t *) pi - 0x80) *(1.0f/(1<< 7))) CONV_FUNC(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_U8,(*(const uint8_t *) pi - 0x80) *(1.0/(1<< 7))) CONV_FUNC(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S16,(*(const int16_t *) pi >>8)+0x80) CONV_FUNC(AV_SAMPLE_FMT_S64, int64_t, AV_SAMPLE_FMT_S16,(uint64_t)(*(const int16_t *) pi)<< 48) CONV_FUNC(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S16, *(const int16_t *) pi *(1.0f/(1<< 15))) CONV_FUNC(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S16, *(const int16_t *) pi *(1.0/(1<< 15))) CONV_FUNC(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S32,(*(const int32_t *) pi >>24)+0x80) CONV_FUNC(AV_SAMPLE_FMT_S64, int64_t, AV_SAMPLE_FMT_S32,(uint64_t)(*(const int32_t *) pi)<< 32) CONV_FUNC(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S32, *(const int32_t *) pi *(1.0f/(1U<< 31))) CONV_FUNC(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S32, *(const int32_t *) pi *(1.0/(1U<< 31))) CONV_FUNC(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S64,(*(const int64_t *) pi >>56)+0x80) CONV_FUNC(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S64, *(const int64_t *) pi *(1.0f/(INT64_C(1)<< 63))) CONV_FUNC(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S64, *(const int64_t *) pi *(1.0/(INT64_C(1)<< 63))) CONV_FUNC(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_FLT, av_clip_uint8(lrintf(*(const float *) pi *(1<< 7))+0x80)) CONV_FUNC(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_FLT, av_clip_int16(lrintf(*(const float *) pi *(1<< 15)))) CONV_FUNC(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_FLT, av_clipl_int32(llrintf(*(const float *) pi *(1U<< 31)))) CONV_FUNC(AV_SAMPLE_FMT_S64, int64_t, AV_SAMPLE_FMT_FLT, llrintf(*(const float *) pi *(INT64_C(1)<< 63))) CONV_FUNC(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_DBL, av_clip_uint8(lrint(*(const double *) pi *(1<< 7))+0x80)) CONV_FUNC(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_DBL, av_clip_int16(lrint(*(const double *) pi *(1<< 15)))) CONV_FUNC(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_DBL, av_clipl_int32(llrint(*(const double *) pi *(1U<< 31)))) CONV_FUNC(AV_SAMPLE_FMT_S64, int64_t, AV_SAMPLE_FMT_DBL, llrint(*(const double *) pi *(INT64_C(1)<< 63))) #define FMT_PAIR_FUNC(out, in) static conv_func_type *const fmt_pair_to_conv_functions[AV_SAMPLE_FMT_NB *AV_SAMPLE_FMT_NB]={ FMT_PAIR_FUNC(AV_SAMPLE_FMT_U8, AV_SAMPLE_FMT_U8), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_U8), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S32, AV_SAMPLE_FMT_U8), FMT_PAIR_FUNC(AV_SAMPLE_FMT_FLT, AV_SAMPLE_FMT_U8), FMT_PAIR_FUNC(AV_SAMPLE_FMT_DBL, AV_SAMPLE_FMT_U8), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S64, AV_SAMPLE_FMT_U8), FMT_PAIR_FUNC(AV_SAMPLE_FMT_U8, AV_SAMPLE_FMT_S16), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_S16), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S32, AV_SAMPLE_FMT_S16), FMT_PAIR_FUNC(AV_SAMPLE_FMT_FLT, AV_SAMPLE_FMT_S16), FMT_PAIR_FUNC(AV_SAMPLE_FMT_DBL, AV_SAMPLE_FMT_S16), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S64, AV_SAMPLE_FMT_S16), FMT_PAIR_FUNC(AV_SAMPLE_FMT_U8, AV_SAMPLE_FMT_S32), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_S32), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S32, AV_SAMPLE_FMT_S32), FMT_PAIR_FUNC(AV_SAMPLE_FMT_FLT, AV_SAMPLE_FMT_S32), FMT_PAIR_FUNC(AV_SAMPLE_FMT_DBL, AV_SAMPLE_FMT_S32), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S64, AV_SAMPLE_FMT_S32), FMT_PAIR_FUNC(AV_SAMPLE_FMT_U8, AV_SAMPLE_FMT_FLT), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_FLT), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S32, AV_SAMPLE_FMT_FLT), FMT_PAIR_FUNC(AV_SAMPLE_FMT_FLT, AV_SAMPLE_FMT_FLT), FMT_PAIR_FUNC(AV_SAMPLE_FMT_DBL, AV_SAMPLE_FMT_FLT), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S64, AV_SAMPLE_FMT_FLT), FMT_PAIR_FUNC(AV_SAMPLE_FMT_U8, AV_SAMPLE_FMT_DBL), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_DBL), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S32, AV_SAMPLE_FMT_DBL), FMT_PAIR_FUNC(AV_SAMPLE_FMT_FLT, AV_SAMPLE_FMT_DBL), FMT_PAIR_FUNC(AV_SAMPLE_FMT_DBL, AV_SAMPLE_FMT_DBL), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S64, AV_SAMPLE_FMT_DBL), FMT_PAIR_FUNC(AV_SAMPLE_FMT_U8, AV_SAMPLE_FMT_S64), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_S64), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S32, AV_SAMPLE_FMT_S64), FMT_PAIR_FUNC(AV_SAMPLE_FMT_FLT, AV_SAMPLE_FMT_S64), FMT_PAIR_FUNC(AV_SAMPLE_FMT_DBL, AV_SAMPLE_FMT_S64), FMT_PAIR_FUNC(AV_SAMPLE_FMT_S64, AV_SAMPLE_FMT_S64), };static void cpy1(uint8_t **dst, const uint8_t **src, int len){ memcpy(*dst, *src, len);} static void cpy2(uint8_t **dst, const uint8_t **src, int len){ memcpy(*dst, *src, 2 *len);} static void cpy4(uint8_t **dst, const uint8_t **src, int len){ memcpy(*dst, *src, 4 *len);} static void cpy8(uint8_t **dst, const uint8_t **src, int len){ memcpy(*dst, *src, 8 *len);} AudioConvert *swri_audio_convert_alloc(enum AVSampleFormat out_fmt, enum AVSampleFormat in_fmt, int channels, const int *ch_map, int flags) { AudioConvert *ctx;conv_func_type *f=fmt_pair_to_conv_functions[av_get_packed_sample_fmt(out_fmt)+AV_SAMPLE_FMT_NB *av_get_packed_sample_fmt(in_fmt)];if(!f) return NULL;ctx=av_mallocz(sizeof(*ctx));if(!ctx) return NULL;if(channels==1){ in_fmt=av_get_planar_sample_fmt(in_fmt);out_fmt=av_get_planar_sample_fmt(out_fmt);} ctx->channels=channels;ctx->conv_f=f;ctx->ch_map=ch_map;if(in_fmt==AV_SAMPLE_FMT_U8||in_fmt==AV_SAMPLE_FMT_U8P) memset(ctx->silence, 0x80, sizeof(ctx->silence));if(out_fmt==in_fmt &&!ch_map) { switch(av_get_bytes_per_sample(in_fmt)){ case 1:ctx->simd_f=cpy1;break;case 2:ctx->simd_f=cpy2;break;case 4:ctx->simd_f=cpy4;break;case 8:ctx->simd_f=cpy8;break;} } if(HAVE_X86ASM &&HAVE_MMX) swri_audio_convert_init_x86(ctx, out_fmt, in_fmt, channels);if(ARCH_ARM) swri_audio_convert_init_arm(ctx, out_fmt, in_fmt, channels);if(ARCH_AARCH64) swri_audio_convert_init_aarch64(ctx, out_fmt, in_fmt, channels);return ctx;} void swri_audio_convert_free(AudioConvert **ctx) { av_freep(ctx);} int swri_audio_convert(AudioConvert *ctx, AudioData *out, AudioData *in, int len) { int ch;int off=0;const int os=(out->planar ? 1 :out->ch_count) *out->bps;unsigned misaligned=0;av_assert0(ctx->channels==out->ch_count);if(ctx->in_simd_align_mask) { int planes=in->planar ? in->ch_count :1;unsigned m=0;for(ch=0;ch< planes;ch++) m|=(intptr_t) in->ch[ch];misaligned|=m &ctx->in_simd_align_mask;} if(ctx->out_simd_align_mask) { int planes=out->planar ? out->ch_count :1;unsigned m=0;for(ch=0;ch< planes;ch++) m|=(intptr_t) out->ch[ch];misaligned|=m &ctx->out_simd_align_mask;} if(ctx->simd_f &&!ctx->ch_map &&!misaligned){ off=len &~15;av_assert1(off >=0);av_assert1(off<=len);av_assert2(ctx->channels==SWR_CH_MAX||!in->ch[ctx->channels]);if(off >0){ if(out->planar==in->planar){ int planes=out->planar ? out->ch_count :1;for(ch=0;ch< planes;ch++){ ctx->simd_f(out-> ch ch
static float calc_reduction_3gpp(float a, float desired_pe, float pe, float active_lines)
float ath
absolute threshold of hearing per bands
#define PSY_3GPP_EN_SPREAD_HI_L1
static av_cold float ath(float f, float add)
Calculate ATH value for given frequency.
float prev_energy_subshort[AAC_NUM_BLOCKS_SHORT *PSY_LAME_NUM_SUBBLOCKS]
enum WindowSequence next_window_seq
window sequence to be used in the next frame
#define AAC_BLOCK_SIZE_LONG
long block size
int * num_bands
number of scalefactor bands for possible frame sizes
Macro definitions for various function/variable attributes.
LAME psy model preset struct.
float thr
energy threshold
float correction
PE correction factor.
static av_cold void psy_3gpp_end(FFPsyContext *apc)
float attack_threshold
attack threshold for this channel
#define PSY_3GPP_EN_SPREAD_LOW_L
float nz_lines
number of non-zero spectral lines
psychoacoustic model frame type-dependent coefficients
int size
size of the bitresevoir in bits
static float calc_reduced_thr_3gpp(AacPsyBand *band, float min_snr, float reduction)
#define PSY_LAME_FIR_LEN
LAME psy model FIR order.
#define PSY_3GPP_CLIP_LO_L
#define PSY_3GPP_SPEND_SLOPE_S
#define u(width, name, range_min, range_max)
#define PSY_3GPP_THR_SPREAD_LOW
context used by psychoacoustic model
int flags
Flags modifying the (de)muxer behaviour.
single band psychoacoustic information
static float lame_calc_attack_threshold(int bitrate)
Calculate the ABR attack threshold from the above LAME psymodel table.
uint8_t next_grouping
stored grouping scheme for the next frame (in case of 8 short window sequence)
#define PSY_3GPP_SAVE_ADD_L
static av_cold float calc_bark(float f)
Calculate Bark value for given line.
static av_always_inline double ff_exp10(double x)
Compute 10^x for floating point values.
#define PSY_3GPP_SPEND_ADD_S
3GPP TS26.403-inspired psychoacoustic model specific data
single/pair channel context for psychoacoustic model
static const float psy_fir_coeffs[]
LAME psy model FIR coefficient table.
float barks
Bark value for each spectral band in long frame.
int flags
AV_CODEC_FLAG_*.
float pe_const
constant part of the PE calculation
void * av_mallocz(size_t size)
Allocate a memory block with alignment suitable for all memory accesses (including vectors if availab...
int num_windows
number of windows in a frame
static FFPsyWindowInfo psy_lame_window(FFPsyContext *ctx, const float *audio, const float *la, int channel, int prev_type)
#define PSY_3GPP_SPEND_SLOPE_L
#define PSY_3GPP_THR_SPREAD_HI
constants for 3GPP AAC psychoacoustic model
codec-specific psychoacoustic model implementation
struct AacPsyContext::@31 pe
float thr_quiet
threshold in quiet
static void psy_3gpp_analyze(FFPsyContext *ctx, int channel, const float **coeffs, const FFPsyWindowInfo *wi)
#define AV_CODEC_FLAG_QSCALE
Use fixed qscale.
int prev_attack
attack value for the last short block in the previous sequence
#define PSY_3GPP_SAVE_SLOPE_S
uint8_t num_ch
number of channels in this group
int frame_bits
average bits per frame
int fill_level
bit reservoir fill level
static void lame_apply_block_type(AacPsyChannel *ctx, FFPsyWindowInfo *wi, int uselongblock)
#define PSY_3GPP_SAVE_SLOPE_L
Reference: libavcodec/aacpsy.c.
#define PSY_LAME_NUM_SUBBLOCKS
Number of sub-blocks in each short block.
const FFPsyModel ff_aac_psy_model
static void psy_3gpp_analyze_channel(FFPsyContext *ctx, int channel, const float *coefs, const FFPsyWindowInfo *wi)
Calculate band thresholds as suggested in 3GPP TS26.403.
float st_lrm
short threshold for L, R, and M channels
#define PSY_3GPP_EN_SPREAD_LOW_S
Libavcodec external API header.
struct FFPsyContext::@106 bitres
int sample_rate
samples per second
FFPsyChannelGroup * ff_psy_find_group(FFPsyContext *ctx, int channel)
Determine what group a channel belongs to.
main external API structure.
float win_energy
sliding average of channel energy
void * model_priv_data
psychoacoustic model implementation private data
float active_lines
number of active spectral lines
static const float bands[]
static float iir_filter(int in, float state[2])
IIR filter used in block switching decision.
int avoid_holes
hole avoidance flag
AacPsyBand band[128]
bands information
uint8_t pi<< 24) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_U8, uint8_t,(*(const uint8_t *) pi - 0x80) *(1.0f/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_U8, uint8_t,(*(const uint8_t *) pi - 0x80) *(1.0/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S16, int16_t,(*(const int16_t *) pi >> 8)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S16, int16_t, *(const int16_t *) pi *(1.0f/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S16, int16_t, *(const int16_t *) pi *(1.0/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S32, int32_t,(*(const int32_t *) pi >> 24)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S32, int32_t, *(const int32_t *) pi *(1.0f/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S32, int32_t, *(const int32_t *) pi *(1.0/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_FLT, float, av_clip_uint8(lrintf(*(const float *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_FLT, float, av_clip_int16(lrintf(*(const float *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_FLT, float, av_clipl_int32(llrintf(*(const float *) pi *(1U<< 31)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_DBL, double, av_clip_uint8(lrint(*(const double *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_DBL, double, av_clip_int16(lrint(*(const double *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_DBL, double, av_clipl_int32(llrint(*(const double *) pi *(1U<< 31)))) #define SET_CONV_FUNC_GROUP(ofmt, ifmt) static void set_generic_function(AudioConvert *ac) { } void ff_audio_convert_free(AudioConvert **ac) { if(! *ac) return;ff_dither_free(&(*ac) ->dc);av_freep(ac);} AudioConvert *ff_audio_convert_alloc(AVAudioResampleContext *avr, enum AVSampleFormat out_fmt, enum AVSampleFormat in_fmt, int channels, int sample_rate, int apply_map) { AudioConvert *ac;int in_planar, out_planar;ac=av_mallocz(sizeof(*ac));if(!ac) return NULL;ac->avr=avr;ac->out_fmt=out_fmt;ac->in_fmt=in_fmt;ac->channels=channels;ac->apply_map=apply_map;if(avr->dither_method !=AV_RESAMPLE_DITHER_NONE &&av_get_packed_sample_fmt(out_fmt)==AV_SAMPLE_FMT_S16 &&av_get_bytes_per_sample(in_fmt) > 2) { ac->dc=ff_dither_alloc(avr, out_fmt, in_fmt, channels, sample_rate, apply_map);if(!ac->dc) { av_free(ac);return NULL;} return ac;} in_planar=ff_sample_fmt_is_planar(in_fmt, channels);out_planar=ff_sample_fmt_is_planar(out_fmt, channels);if(in_planar==out_planar) { ac->func_type=CONV_FUNC_TYPE_FLAT;ac->planes=in_planar ? ac->channels :1;} else if(in_planar) ac->func_type=CONV_FUNC_TYPE_INTERLEAVE;else ac->func_type=CONV_FUNC_TYPE_DEINTERLEAVE;set_generic_function(ac);if(ARCH_AARCH64) ff_audio_convert_init_aarch64(ac);if(ARCH_ARM) ff_audio_convert_init_arm(ac);if(ARCH_X86) ff_audio_convert_init_x86(ac);return ac;} int ff_audio_convert(AudioConvert *ac, AudioData *out, AudioData *in) { int use_generic=1;int len=in->nb_samples;int p;if(ac->dc) { av_log(ac->avr, AV_LOG_TRACE, "%d samples - audio_convert: %s to %s (dithered)\", len, av_get_sample_fmt_name(ac->in_fmt), av_get_sample_fmt_name(ac->out_fmt));return ff_convert_dither(ac-> in
#define PSY_3GPP_CLIP_HI_S
static const PsyLamePreset psy_abr_map[]
LAME psy model preset table for ABR.
int window_shape
window shape (sine/KBD/whatever)
#define PSY_PE_FORGET_SLOPE
#define PSY_3GPP_PE_TO_BITS(bits)
int cutoff
lowpass frequency cutoff for analysis
float max
maximum allowed PE for bit factor calculation
static void calc_thr_3gpp(const FFPsyWindowInfo *wi, const int num_bands, AacPsyChannel *pch, const uint8_t *band_sizes, const float *coefs, const int cutoff)
float previous
allowed PE of the previous frame
AacPsyCoeffs psy_coef[2][64]
float min
minimum allowed PE for bit factor calculation
int global_quality
Global quality for codecs which cannot change it per frame.
static av_cold int psy_3gpp_init(FFPsyContext *ctx)
static void psy_hp_filter(const float *firbuf, float *hpfsmpl, const float *psy_fir_coeffs)
float spread_hi[2]
spreading factor for high-to-low threshold spreading in long frame
internal math functions header
static av_unused FFPsyWindowInfo psy_3gpp_window(FFPsyContext *ctx, const int16_t *audio, const int16_t *la, int channel, int prev_type)
Tell encoder which window types to use.
static float calc_pe_3gpp(AacPsyBand *band)
windowing related information
channel
Use these values when setting the channel map with ebur128_set_channel().
#define PSY_3GPP_BITS_TO_PE(bits)
float norm_fac
normalization factor for linearization
int chan_bitrate
bitrate per channel
#define PSY_3GPP_CLIP_LO_S
#define PSY_3GPP_AH_THR_LONG
static const int16_t coeffs[]
int channels
number of audio channels
float pe
perceptual entropy
#define PSY_3GPP_EN_SPREAD_HI_S
static const double coeff[2][5]
#define FF_QP2LAMBDA
factor to convert from H.263 QP to lambda
#define PSY_3GPP_SAVE_ADD_S
information for single band used by 3GPP TS26.403-inspired psychoacoustic model
AVCodecContext * avctx
encoder context
float spread_low[2]
spreading factor for low-to-high threshold spreading in long frame
#define PSY_3GPP_CLIP_HI_L
int window_type[3]
window type (short/long/transitional, etc.) - current, previous and next
#define AAC_NUM_BLOCKS_SHORT
number of blocks in a short sequence
#define PSY_3GPP_SPEND_ADD_L
void * av_mallocz_array(size_t nmemb, size_t size)
Allocate a memory block for an array with av_mallocz().
static av_cold void lame_window_init(AacPsyContext *ctx, AVCodecContext *avctx)
LAME psy model specific initialization.