ffmpeg/libavcodec/aacsbr_fixed.c

586 lines
22 KiB
C

/*
* Copyright (c) 2013
* MIPS Technologies, Inc., California.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the MIPS Technologies, Inc., nor the names of its
* contributors may be used to endorse or promote products derived from
* this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE MIPS TECHNOLOGIES, INC. ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE MIPS TECHNOLOGIES, INC. BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* AAC Spectral Band Replication decoding functions (fixed-point)
* Copyright (c) 2008-2009 Robert Swain ( rob opendot cl )
* Copyright (c) 2009-2010 Alex Converse <alex.converse@gmail.com>
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* @file
* AAC Spectral Band Replication decoding functions (fixed-point)
* Note: Rounding-to-nearest used unless otherwise stated
* @author Robert Swain ( rob opendot cl )
* @author Stanislav Ocovaj ( stanislav.ocovaj imgtec com )
*/
#define USE_FIXED 1
#include "aac.h"
#include "sbr.h"
#include "aacsbr.h"
#include "aacsbrdata.h"
#include "aacsbr_fixed_tablegen.h"
#include "fft.h"
#include "aacps.h"
#include "sbrdsp.h"
#include "libavutil/internal.h"
#include "libavutil/libm.h"
#include "libavutil/avassert.h"
#include <stdint.h>
#include <float.h>
#include <math.h>
static VLC vlc_sbr[10];
static void aacsbr_func_ptr_init(AACSBRContext *c);
static const int CONST_LN2 = Q31(0.6931471806/256); // ln(2)/256
static const int CONST_RECIP_LN2 = Q31(0.7213475204); // 0.5/ln(2)
static const int CONST_076923 = Q31(0.76923076923076923077f);
static const int fixed_log_table[10] =
{
Q31(1.0/2), Q31(1.0/3), Q31(1.0/4), Q31(1.0/5), Q31(1.0/6),
Q31(1.0/7), Q31(1.0/8), Q31(1.0/9), Q31(1.0/10), Q31(1.0/11)
};
static int fixed_log(int x)
{
int i, ret, xpow, tmp;
ret = x;
xpow = x;
for (i=0; i<10; i+=2){
xpow = (int)(((int64_t)xpow * x + 0x40000000) >> 31);
tmp = (int)(((int64_t)xpow * fixed_log_table[i] + 0x40000000) >> 31);
ret -= tmp;
xpow = (int)(((int64_t)xpow * x + 0x40000000) >> 31);
tmp = (int)(((int64_t)xpow * fixed_log_table[i+1] + 0x40000000) >> 31);
ret += tmp;
}
return ret;
}
static const int fixed_exp_table[7] =
{
Q31(1.0/2), Q31(1.0/6), Q31(1.0/24), Q31(1.0/120),
Q31(1.0/720), Q31(1.0/5040), Q31(1.0/40320)
};
static int fixed_exp(int x)
{
int i, ret, xpow, tmp;
ret = 0x800000 + x;
xpow = x;
for (i=0; i<7; i++){
xpow = (int)(((int64_t)xpow * x + 0x400000) >> 23);
tmp = (int)(((int64_t)xpow * fixed_exp_table[i] + 0x40000000) >> 31);
ret += tmp;
}
return ret;
}
static void make_bands(int16_t* bands, int start, int stop, int num_bands)
{
int k, previous, present;
int base, prod, nz = 0;
base = (stop << 23) / start;
while (base < 0x40000000){
base <<= 1;
nz++;
}
base = fixed_log(base - 0x80000000);
base = (((base + 0x80) >> 8) + (8-nz)*CONST_LN2) / num_bands;
base = fixed_exp(base);
previous = start;
prod = start << 23;
for (k = 0; k < num_bands-1; k++) {
prod = (int)(((int64_t)prod * base + 0x400000) >> 23);
present = (prod + 0x400000) >> 23;
bands[k] = present - previous;
previous = present;
}
bands[num_bands-1] = stop - previous;
}
/// Dequantization and stereo decoding (14496-3 sp04 p203)
static void sbr_dequant(SpectralBandReplication *sbr, int id_aac)
{
int k, e;
int ch;
if (id_aac == TYPE_CPE && sbr->bs_coupling) {
int alpha = sbr->data[0].bs_amp_res ? 2 : 1;
int pan_offset = sbr->data[0].bs_amp_res ? 12 : 24;
for (e = 1; e <= sbr->data[0].bs_num_env; e++) {
for (k = 0; k < sbr->n[sbr->data[0].bs_freq_res[e]]; k++) {
SoftFloat temp1, temp2, fac;
temp1.exp = sbr->data[0].env_facs[e][k].mant * alpha + 14;
if (temp1.exp & 1)
temp1.mant = 759250125;
else
temp1.mant = 0x20000000;
temp1.exp = (temp1.exp >> 1) + 1;
temp2.exp = (pan_offset - sbr->data[1].env_facs[e][k].mant) * alpha;
if (temp2.exp & 1)
temp2.mant = 759250125;
else
temp2.mant = 0x20000000;
temp2.exp = (temp2.exp >> 1) + 1;
fac = av_div_sf(temp1, av_add_sf(FLOAT_1, temp2));
sbr->data[0].env_facs[e][k] = fac;
sbr->data[1].env_facs[e][k] = av_mul_sf(fac, temp2);
}
}
for (e = 1; e <= sbr->data[0].bs_num_noise; e++) {
for (k = 0; k < sbr->n_q; k++) {
SoftFloat temp1, temp2, fac;
temp1.exp = NOISE_FLOOR_OFFSET - \
sbr->data[0].noise_facs[e][k].mant + 2;
temp1.mant = 0x20000000;
temp2.exp = 12 - sbr->data[1].noise_facs[e][k].mant + 1;
temp2.mant = 0x20000000;
fac = av_div_sf(temp1, av_add_sf(FLOAT_1, temp2));
sbr->data[0].noise_facs[e][k] = fac;
sbr->data[1].noise_facs[e][k] = av_mul_sf(fac, temp2);
}
}
} else { // SCE or one non-coupled CPE
for (ch = 0; ch < (id_aac == TYPE_CPE) + 1; ch++) {
int alpha = sbr->data[ch].bs_amp_res ? 2 : 1;
for (e = 1; e <= sbr->data[ch].bs_num_env; e++)
for (k = 0; k < sbr->n[sbr->data[ch].bs_freq_res[e]]; k++){
SoftFloat temp1;
temp1.exp = alpha * sbr->data[ch].env_facs[e][k].mant + 12;
if (temp1.exp & 1)
temp1.mant = 759250125;
else
temp1.mant = 0x20000000;
temp1.exp = (temp1.exp >> 1) + 1;
sbr->data[ch].env_facs[e][k] = temp1;
}
for (e = 1; e <= sbr->data[ch].bs_num_noise; e++)
for (k = 0; k < sbr->n_q; k++){
sbr->data[ch].noise_facs[e][k].exp = NOISE_FLOOR_OFFSET - \
sbr->data[ch].noise_facs[e][k].mant + 1;
sbr->data[ch].noise_facs[e][k].mant = 0x20000000;
}
}
}
}
/** High Frequency Generation (14496-3 sp04 p214+) and Inverse Filtering
* (14496-3 sp04 p214)
* Warning: This routine does not seem numerically stable.
*/
static void sbr_hf_inverse_filter(SBRDSPContext *dsp,
int (*alpha0)[2], int (*alpha1)[2],
const int X_low[32][40][2], int k0)
{
int k;
int shift, round;
for (k = 0; k < k0; k++) {
SoftFloat phi[3][2][2];
SoftFloat a00, a01, a10, a11;
SoftFloat dk;
dsp->autocorrelate(X_low[k], phi);
dk = av_sub_sf(av_mul_sf(phi[2][1][0], phi[1][0][0]),
av_mul_sf(av_add_sf(av_mul_sf(phi[1][1][0], phi[1][1][0]),
av_mul_sf(phi[1][1][1], phi[1][1][1])), FLOAT_0999999));
if (!dk.mant) {
a10 = FLOAT_0;
a11 = FLOAT_0;
} else {
SoftFloat temp_real, temp_im;
temp_real = av_sub_sf(av_sub_sf(av_mul_sf(phi[0][0][0], phi[1][1][0]),
av_mul_sf(phi[0][0][1], phi[1][1][1])),
av_mul_sf(phi[0][1][0], phi[1][0][0]));
temp_im = av_sub_sf(av_add_sf(av_mul_sf(phi[0][0][0], phi[1][1][1]),
av_mul_sf(phi[0][0][1], phi[1][1][0])),
av_mul_sf(phi[0][1][1], phi[1][0][0]));
a10 = av_div_sf(temp_real, dk);
a11 = av_div_sf(temp_im, dk);
}
if (!phi[1][0][0].mant) {
a00 = FLOAT_0;
a01 = FLOAT_0;
} else {
SoftFloat temp_real, temp_im;
temp_real = av_add_sf(phi[0][0][0],
av_add_sf(av_mul_sf(a10, phi[1][1][0]),
av_mul_sf(a11, phi[1][1][1])));
temp_im = av_add_sf(phi[0][0][1],
av_sub_sf(av_mul_sf(a11, phi[1][1][0]),
av_mul_sf(a10, phi[1][1][1])));
temp_real.mant = -temp_real.mant;
temp_im.mant = -temp_im.mant;
a00 = av_div_sf(temp_real, phi[1][0][0]);
a01 = av_div_sf(temp_im, phi[1][0][0]);
}
shift = a00.exp;
if (shift >= 3)
alpha0[k][0] = 0x7fffffff;
else {
a00.mant <<= 1;
shift = 2-shift;
if (shift == 0)
alpha0[k][0] = a00.mant;
else {
round = 1 << (shift-1);
alpha0[k][0] = (a00.mant + round) >> shift;
}
}
shift = a01.exp;
if (shift >= 3)
alpha0[k][1] = 0x7fffffff;
else {
a01.mant <<= 1;
shift = 2-shift;
if (shift == 0)
alpha0[k][1] = a01.mant;
else {
round = 1 << (shift-1);
alpha0[k][1] = (a01.mant + round) >> shift;
}
}
shift = a10.exp;
if (shift >= 3)
alpha1[k][0] = 0x7fffffff;
else {
a10.mant <<= 1;
shift = 2-shift;
if (shift == 0)
alpha1[k][0] = a10.mant;
else {
round = 1 << (shift-1);
alpha1[k][0] = (a10.mant + round) >> shift;
}
}
shift = a11.exp;
if (shift >= 3)
alpha1[k][1] = 0x7fffffff;
else {
a11.mant <<= 1;
shift = 2-shift;
if (shift == 0)
alpha1[k][1] = a11.mant;
else {
round = 1 << (shift-1);
alpha1[k][1] = (a11.mant + round) >> shift;
}
}
shift = (int)(((int64_t)(alpha1[k][0]>>1) * (alpha1[k][0]>>1) + \
(int64_t)(alpha1[k][1]>>1) * (alpha1[k][1]>>1) + \
0x40000000) >> 31);
if (shift >= 0x20000000){
alpha1[k][0] = 0;
alpha1[k][1] = 0;
alpha0[k][0] = 0;
alpha0[k][1] = 0;
}
shift = (int)(((int64_t)(alpha0[k][0]>>1) * (alpha0[k][0]>>1) + \
(int64_t)(alpha0[k][1]>>1) * (alpha0[k][1]>>1) + \
0x40000000) >> 31);
if (shift >= 0x20000000){
alpha1[k][0] = 0;
alpha1[k][1] = 0;
alpha0[k][0] = 0;
alpha0[k][1] = 0;
}
}
}
/// Chirp Factors (14496-3 sp04 p214)
static void sbr_chirp(SpectralBandReplication *sbr, SBRData *ch_data)
{
int i;
int new_bw;
static const int bw_tab[] = { 0, 1610612736, 1932735283, 2104533975 };
int64_t accu;
for (i = 0; i < sbr->n_q; i++) {
if (ch_data->bs_invf_mode[0][i] + ch_data->bs_invf_mode[1][i] == 1)
new_bw = 1288490189;
else
new_bw = bw_tab[ch_data->bs_invf_mode[0][i]];
if (new_bw < ch_data->bw_array[i]){
accu = (int64_t)new_bw * 1610612736;
accu += (int64_t)ch_data->bw_array[i] * 0x20000000;
new_bw = (int)((accu + 0x40000000) >> 31);
} else {
accu = (int64_t)new_bw * 1946157056;
accu += (int64_t)ch_data->bw_array[i] * 201326592;
new_bw = (int)((accu + 0x40000000) >> 31);
}
ch_data->bw_array[i] = new_bw < 0x2000000 ? 0 : new_bw;
}
}
/**
* Calculation of levels of additional HF signal components (14496-3 sp04 p219)
* and Calculation of gain (14496-3 sp04 p219)
*/
static void sbr_gain_calc(AACContext *ac, SpectralBandReplication *sbr,
SBRData *ch_data, const int e_a[2])
{
int e, k, m;
// max gain limits : -3dB, 0dB, 3dB, inf dB (limiter off)
static const SoftFloat limgain[4] = { { 760155524, 0 }, { 0x20000000, 1 },
{ 758351638, 1 }, { 625000000, 34 } };
for (e = 0; e < ch_data->bs_num_env; e++) {
int delta = !((e == e_a[1]) || (e == e_a[0]));
for (k = 0; k < sbr->n_lim; k++) {
SoftFloat gain_boost, gain_max;
SoftFloat sum[2] = { { 0, 0}, { 0, 0 } };
for (m = sbr->f_tablelim[k] - sbr->kx[1]; m < sbr->f_tablelim[k + 1] - sbr->kx[1]; m++) {
const SoftFloat temp = av_div_sf(sbr->e_origmapped[e][m],
av_add_sf(FLOAT_1, sbr->q_mapped[e][m]));
sbr->q_m[e][m] = av_sqrt_sf(av_mul_sf(temp, sbr->q_mapped[e][m]));
sbr->s_m[e][m] = av_sqrt_sf(av_mul_sf(temp, av_int2sf(ch_data->s_indexmapped[e + 1][m], 0)));
if (!sbr->s_mapped[e][m]) {
if (delta) {
sbr->gain[e][m] = av_sqrt_sf(av_div_sf(sbr->e_origmapped[e][m],
av_mul_sf(av_add_sf(FLOAT_1, sbr->e_curr[e][m]),
av_add_sf(FLOAT_1, sbr->q_mapped[e][m]))));
} else {
sbr->gain[e][m] = av_sqrt_sf(av_div_sf(sbr->e_origmapped[e][m],
av_add_sf(FLOAT_1, sbr->e_curr[e][m])));
}
} else {
sbr->gain[e][m] = av_sqrt_sf(
av_div_sf(
av_mul_sf(sbr->e_origmapped[e][m], sbr->q_mapped[e][m]),
av_mul_sf(
av_add_sf(FLOAT_1, sbr->e_curr[e][m]),
av_add_sf(FLOAT_1, sbr->q_mapped[e][m]))));
}
}
for (m = sbr->f_tablelim[k] - sbr->kx[1]; m < sbr->f_tablelim[k + 1] - sbr->kx[1]; m++) {
sum[0] = av_add_sf(sum[0], sbr->e_origmapped[e][m]);
sum[1] = av_add_sf(sum[1], sbr->e_curr[e][m]);
}
gain_max = av_mul_sf(limgain[sbr->bs_limiter_gains],
av_sqrt_sf(
av_div_sf(
av_add_sf(FLOAT_EPSILON, sum[0]),
av_add_sf(FLOAT_EPSILON, sum[1]))));
if (av_gt_sf(gain_max, FLOAT_100000))
gain_max = FLOAT_100000;
for (m = sbr->f_tablelim[k] - sbr->kx[1]; m < sbr->f_tablelim[k + 1] - sbr->kx[1]; m++) {
SoftFloat q_m_max = av_div_sf(
av_mul_sf(sbr->q_m[e][m], gain_max),
sbr->gain[e][m]);
if (av_gt_sf(sbr->q_m[e][m], q_m_max))
sbr->q_m[e][m] = q_m_max;
if (av_gt_sf(sbr->gain[e][m], gain_max))
sbr->gain[e][m] = gain_max;
}
sum[0] = sum[1] = FLOAT_0;
for (m = sbr->f_tablelim[k] - sbr->kx[1]; m < sbr->f_tablelim[k + 1] - sbr->kx[1]; m++) {
sum[0] = av_add_sf(sum[0], sbr->e_origmapped[e][m]);
sum[1] = av_add_sf(sum[1],
av_mul_sf(
av_mul_sf(sbr->e_curr[e][m],
sbr->gain[e][m]),
sbr->gain[e][m]));
sum[1] = av_add_sf(sum[1],
av_mul_sf(sbr->s_m[e][m], sbr->s_m[e][m]));
if (delta && !sbr->s_m[e][m].mant)
sum[1] = av_add_sf(sum[1],
av_mul_sf(sbr->q_m[e][m], sbr->q_m[e][m]));
}
gain_boost = av_sqrt_sf(
av_div_sf(
av_add_sf(FLOAT_EPSILON, sum[0]),
av_add_sf(FLOAT_EPSILON, sum[1])));
if (av_gt_sf(gain_boost, FLOAT_1584893192))
gain_boost = FLOAT_1584893192;
for (m = sbr->f_tablelim[k] - sbr->kx[1]; m < sbr->f_tablelim[k + 1] - sbr->kx[1]; m++) {
sbr->gain[e][m] = av_mul_sf(sbr->gain[e][m], gain_boost);
sbr->q_m[e][m] = av_mul_sf(sbr->q_m[e][m], gain_boost);
sbr->s_m[e][m] = av_mul_sf(sbr->s_m[e][m], gain_boost);
}
}
}
}
/// Assembling HF Signals (14496-3 sp04 p220)
static void sbr_hf_assemble(int Y1[38][64][2],
const int X_high[64][40][2],
SpectralBandReplication *sbr, SBRData *ch_data,
const int e_a[2])
{
int e, i, j, m;
const int h_SL = 4 * !sbr->bs_smoothing_mode;
const int kx = sbr->kx[1];
const int m_max = sbr->m[1];
static const SoftFloat h_smooth[5] = {
{ 715827883, -1 },
{ 647472402, -1 },
{ 937030863, -2 },
{ 989249804, -3 },
{ 546843842, -4 },
};
SoftFloat (*g_temp)[48] = ch_data->g_temp, (*q_temp)[48] = ch_data->q_temp;
int indexnoise = ch_data->f_indexnoise;
int indexsine = ch_data->f_indexsine;
if (sbr->reset) {
for (i = 0; i < h_SL; i++) {
memcpy(g_temp[i + 2*ch_data->t_env[0]], sbr->gain[0], m_max * sizeof(sbr->gain[0][0]));
memcpy(q_temp[i + 2*ch_data->t_env[0]], sbr->q_m[0], m_max * sizeof(sbr->q_m[0][0]));
}
} else if (h_SL) {
for (i = 0; i < 4; i++) {
memcpy(g_temp[i + 2 * ch_data->t_env[0]],
g_temp[i + 2 * ch_data->t_env_num_env_old],
sizeof(g_temp[0]));
memcpy(q_temp[i + 2 * ch_data->t_env[0]],
q_temp[i + 2 * ch_data->t_env_num_env_old],
sizeof(q_temp[0]));
}
}
for (e = 0; e < ch_data->bs_num_env; e++) {
for (i = 2 * ch_data->t_env[e]; i < 2 * ch_data->t_env[e + 1]; i++) {
memcpy(g_temp[h_SL + i], sbr->gain[e], m_max * sizeof(sbr->gain[0][0]));
memcpy(q_temp[h_SL + i], sbr->q_m[e], m_max * sizeof(sbr->q_m[0][0]));
}
}
for (e = 0; e < ch_data->bs_num_env; e++) {
for (i = 2 * ch_data->t_env[e]; i < 2 * ch_data->t_env[e + 1]; i++) {
SoftFloat g_filt_tab[48];
SoftFloat q_filt_tab[48];
SoftFloat *g_filt, *q_filt;
if (h_SL && e != e_a[0] && e != e_a[1]) {
g_filt = g_filt_tab;
q_filt = q_filt_tab;
for (m = 0; m < m_max; m++) {
const int idx1 = i + h_SL;
g_filt[m].mant = g_filt[m].exp = 0;
q_filt[m].mant = q_filt[m].exp = 0;
for (j = 0; j <= h_SL; j++) {
g_filt[m] = av_add_sf(g_filt[m],
av_mul_sf(g_temp[idx1 - j][m],
h_smooth[j]));
q_filt[m] = av_add_sf(q_filt[m],
av_mul_sf(q_temp[idx1 - j][m],
h_smooth[j]));
}
}
} else {
g_filt = g_temp[i + h_SL];
q_filt = q_temp[i];
}
sbr->dsp.hf_g_filt(Y1[i] + kx, X_high + kx, g_filt, m_max,
i + ENVELOPE_ADJUSTMENT_OFFSET);
if (e != e_a[0] && e != e_a[1]) {
sbr->dsp.hf_apply_noise[indexsine](Y1[i] + kx, sbr->s_m[e],
q_filt, indexnoise,
kx, m_max);
} else {
int idx = indexsine&1;
int A = (1-((indexsine+(kx & 1))&2));
int B = (A^(-idx)) + idx;
int *out = &Y1[i][kx][idx];
int shift, round;
SoftFloat *in = sbr->s_m[e];
for (m = 0; m+1 < m_max; m+=2) {
shift = 22 - in[m ].exp;
round = 1 << (shift-1);
out[2*m ] += (in[m ].mant * A + round) >> shift;
shift = 22 - in[m+1].exp;
round = 1 << (shift-1);
out[2*m+2] += (in[m+1].mant * B + round) >> shift;
}
if(m_max&1)
{
shift = 22 - in[m ].exp;
round = 1 << (shift-1);
out[2*m ] += (in[m ].mant * A + round) >> shift;
}
}
indexnoise = (indexnoise + m_max) & 0x1ff;
indexsine = (indexsine + 1) & 3;
}
}
ch_data->f_indexnoise = indexnoise;
ch_data->f_indexsine = indexsine;
}
#include "aacsbr_template.c"