ref: a291bbdeddfd41a2f0907ecbd7b819f0eedffdaf
dir: /sys/src/cmd/audio/mp3enc/quantize_pvt.c/
/*
* quantize_pvt source file
*
* Copyright (c) 1999 Takehiro TOMINAGA
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Library General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This library 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
* Library General Public License for more details.
*
* You should have received a copy of the GNU Library General Public
* License along with this library; if not, write to the
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 02111-1307, USA.
*/
/* $Id: quantize_pvt.c,v 1.55 2001/03/05 20:29:24 markt Exp $ */
#ifdef HAVE_CONFIG_H
# include <config.h>
#endif
#include <assert.h>
#include "util.h"
#include "lame-analysis.h"
#include "tables.h"
#include "reservoir.h"
#include "quantize_pvt.h"
#ifdef WITH_DMALLOC
#include <dmalloc.h>
#endif
#define NSATHSCALE 100 // Assuming dynamic range=96dB, this value should be 92
const char slen1_tab [16] = { 0, 0, 0, 0, 3, 1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4 };
const char slen2_tab [16] = { 0, 1, 2, 3, 0, 1, 2, 3, 1, 2, 3, 1, 2, 3, 2, 3 };
/*
The following table is used to implement the scalefactor
partitioning for MPEG2 as described in section
2.4.3.2 of the IS. The indexing corresponds to the
way the tables are presented in the IS:
[table_number][row_in_table][column of nr_of_sfb]
*/
const int nr_of_sfb_block [6] [3] [4] =
{
{
{6, 5, 5, 5},
{9, 9, 9, 9},
{6, 9, 9, 9}
},
{
{6, 5, 7, 3},
{9, 9, 12, 6},
{6, 9, 12, 6}
},
{
{11, 10, 0, 0},
{18, 18, 0, 0},
{15,18,0,0}
},
{
{7, 7, 7, 0},
{12, 12, 12, 0},
{6, 15, 12, 0}
},
{
{6, 6, 6, 3},
{12, 9, 9, 6},
{6, 12, 9, 6}
},
{
{8, 8, 5, 0},
{15,12,9,0},
{6,18,9,0}
}
};
/* Table B.6: layer3 preemphasis */
const char pretab [SBMAX_l] =
{
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 1, 1, 1, 2, 2, 3, 3, 3, 2, 0
};
/*
Here are MPEG1 Table B.8 and MPEG2 Table B.1
-- Layer III scalefactor bands.
Index into this using a method such as:
idx = fr_ps->header->sampling_frequency
+ (fr_ps->header->version * 3)
*/
const scalefac_struct sfBandIndex[9] =
{
{ /* Table B.2.b: 22.05 kHz */
{0,6,12,18,24,30,36,44,54,66,80,96,116,140,168,200,238,284,336,396,464,522,576},
{0,4,8,12,18,24,32,42,56,74,100,132,174,192}
},
{ /* Table B.2.c: 24 kHz */ /* docs: 332. mpg123(broken): 330 */
{0,6,12,18,24,30,36,44,54,66,80,96,114,136,162,194,232,278, 332, 394,464,540,576},
{0,4,8,12,18,26,36,48,62,80,104,136,180,192}
},
{ /* Table B.2.a: 16 kHz */
{0,6,12,18,24,30,36,44,54,66,80,96,116,140,168,200,238,284,336,396,464,522,576},
{0,4,8,12,18,26,36,48,62,80,104,134,174,192}
},
{ /* Table B.8.b: 44.1 kHz */
{0,4,8,12,16,20,24,30,36,44,52,62,74,90,110,134,162,196,238,288,342,418,576},
{0,4,8,12,16,22,30,40,52,66,84,106,136,192}
},
{ /* Table B.8.c: 48 kHz */
{0,4,8,12,16,20,24,30,36,42,50,60,72,88,106,128,156,190,230,276,330,384,576},
{0,4,8,12,16,22,28,38,50,64,80,100,126,192}
},
{ /* Table B.8.a: 32 kHz */
{0,4,8,12,16,20,24,30,36,44,54,66,82,102,126,156,194,240,296,364,448,550,576},
{0,4,8,12,16,22,30,42,58,78,104,138,180,192}
},
{ /* MPEG-2.5 11.025 kHz */
{0,6,12,18,24,30,36,44,54,66,80,96,116,140,168,200,238,284,336,396,464,522,576},
{0/3,12/3,24/3,36/3,54/3,78/3,108/3,144/3,186/3,240/3,312/3,402/3,522/3,576/3}
},
{ /* MPEG-2.5 12 kHz */
{0,6,12,18,24,30,36,44,54,66,80,96,116,140,168,200,238,284,336,396,464,522,576},
{0/3,12/3,24/3,36/3,54/3,78/3,108/3,144/3,186/3,240/3,312/3,402/3,522/3,576/3}
},
{ /* MPEG-2.5 8 kHz */
{0,12,24,36,48,60,72,88,108,132,160,192,232,280,336,400,476,566,568,570,572,574,576},
{0/3,24/3,48/3,72/3,108/3,156/3,216/3,288/3,372/3,480/3,486/3,492/3,498/3,576/3}
}
};
FLOAT8 pow20[Q_MAX];
FLOAT8 ipow20[Q_MAX];
FLOAT8 pow43[PRECALC_SIZE];
/* initialized in first call to iteration_init */
FLOAT8 adj43asm[PRECALC_SIZE];
FLOAT8 adj43[PRECALC_SIZE];
/************************************************************************/
/* initialization for iteration_loop */
/************************************************************************/
void
iteration_init( lame_global_flags *gfp)
{
lame_internal_flags *gfc=gfp->internal_flags;
III_side_info_t * const l3_side = &gfc->l3_side;
int i;
if ( gfc->iteration_init_init==0 ) {
gfc->iteration_init_init=1;
l3_side->main_data_begin = 0;
compute_ath(gfp,gfc->ATH->l,gfc->ATH->s);
pow43[0] = 0.0;
for(i=1;i<PRECALC_SIZE;i++)
pow43[i] = pow((FLOAT8)i, 4.0/3.0);
adj43asm[0] = 0.0;
for (i = 1; i < PRECALC_SIZE; i++)
adj43asm[i] = i - 0.5 - pow(0.5 * (pow43[i - 1] + pow43[i]),0.75);
for (i = 0; i < PRECALC_SIZE-1; i++)
adj43[i] = (i + 1) - pow(0.5 * (pow43[i] + pow43[i + 1]), 0.75);
adj43[i] = 0.5;
for (i = 0; i < Q_MAX; i++) {
ipow20[i] = pow(2.0, (double)(i - 210) * -0.1875);
pow20[i] = pow(2.0, (double)(i - 210) * 0.25);
}
huffman_init(gfc);
}
}
/*
compute the ATH for each scalefactor band
cd range: 0..96db
Input: 3.3kHz signal 32767 amplitude (3.3kHz is where ATH is smallest = -5db)
longblocks: sfb=12 en0/bw=-11db max_en0 = 1.3db
shortblocks: sfb=5 -9db 0db
Input: 1 1 1 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 (repeated)
longblocks: amp=1 sfb=12 en0/bw=-103 db max_en0 = -92db
amp=32767 sfb=12 -12 db -1.4db
Input: 1 1 1 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 (repeated)
shortblocks: amp=1 sfb=5 en0/bw= -99 -86
amp=32767 sfb=5 -9 db 4db
MAX energy of largest wave at 3.3kHz = 1db
AVE energy of largest wave at 3.3kHz = -11db
Let's take AVE: -11db = maximum signal in sfb=12.
Dynamic range of CD: 96db. Therefor energy of smallest audible wave
in sfb=12 = -11 - 96 = -107db = ATH at 3.3kHz.
ATH formula for this wave: -5db. To adjust to LAME scaling, we need
ATH = ATH_formula - 103 (db)
ATH = ATH * 2.5e-10 (ener)
*/
FLOAT8 ATHmdct( lame_global_flags *gfp, FLOAT8 f )
{
lame_internal_flags *gfc = gfp->internal_flags;
FLOAT8 ath;
ath = ATHformula( f , gfp );
if (gfc->nsPsy.use) {
ath -= NSATHSCALE;
} else {
ath -= 114;
}
/* modify the MDCT scaling for the ATH */
ath -= gfp->ATHlower;
/* convert to energy */
ath = pow( 10.0, ath/10.0 );
return ath;
}
void compute_ath( lame_global_flags *gfp, FLOAT8 ATH_l[], FLOAT8 ATH_s[] )
{
lame_internal_flags *gfc = gfp->internal_flags;
int sfb, i, start, end;
FLOAT8 ATH_f;
FLOAT8 samp_freq = gfp->out_samplerate;
for (sfb = 0; sfb < SBMAX_l; sfb++) {
start = gfc->scalefac_band.l[ sfb ];
end = gfc->scalefac_band.l[ sfb+1 ];
ATH_l[sfb]=1e99;
for (i = start ; i < end; i++) {
FLOAT8 freq = i*samp_freq/(2*576);
ATH_f = ATHmdct( gfp, freq ); /* freq in kHz */
ATH_l[sfb] = Min( ATH_l[sfb], ATH_f );
}
}
for (sfb = 0; sfb < SBMAX_s; sfb++){
start = gfc->scalefac_band.s[ sfb ];
end = gfc->scalefac_band.s[ sfb+1 ];
ATH_s[sfb] = 1e99;
for (i = start ; i < end; i++) {
FLOAT8 freq = i*samp_freq/(2*192);
ATH_f = ATHmdct( gfp, freq ); /* freq in kHz */
ATH_s[sfb] = Min( ATH_s[sfb], ATH_f );
}
}
/* no-ATH mode:
* reduce ATH to -200 dB
*/
if (gfp->noATH) {
for (sfb = 0; sfb < SBMAX_l; sfb++) {
ATH_l[sfb] = 1E-37;
}
for (sfb = 0; sfb < SBMAX_s; sfb++) {
ATH_s[sfb] = 1E-37;
}
}
}
/* convert from L/R <-> Mid/Side, src == dst allowed */
void ms_convert(FLOAT8 dst[2][576], FLOAT8 src[2][576])
{
FLOAT8 l;
FLOAT8 r;
int i;
for (i = 0; i < 576; ++i) {
l = src[0][i];
r = src[1][i];
dst[0][i] = (l+r) * (FLOAT8)(SQRT2*0.5);
dst[1][i] = (l-r) * (FLOAT8)(SQRT2*0.5);
}
}
/************************************************************************
* allocate bits among 2 channels based on PE
* mt 6/99
************************************************************************/
int on_pe(lame_global_flags *gfp,FLOAT8 pe[2][2],III_side_info_t *l3_side,
int targ_bits[2],int mean_bits, int gr)
{
lame_internal_flags *gfc=gfp->internal_flags;
gr_info *cod_info;
int extra_bits,tbits,bits;
int add_bits[2];
int ch;
int max_bits; /* maximum allowed bits for this granule */
/* allocate targ_bits for granule */
ResvMaxBits (gfp, mean_bits, &tbits, &extra_bits);
max_bits=tbits+extra_bits;
bits=0;
for (ch=0 ; ch < gfc->channels_out ; ch ++) {
/******************************************************************
* allocate bits for each channel
******************************************************************/
cod_info = &l3_side->gr[gr].ch[ch].tt;
targ_bits[ch]=Min(MAX_BITS, tbits/gfc->channels_out);
if (gfc->nsPsy.use) {
add_bits[ch] = targ_bits[ch]*pe[gr][ch]/700.0-targ_bits[ch];
} else {
add_bits[ch]=(pe[gr][ch]-750)/1.4;
/* short blocks us a little extra, no matter what the pe */
if (cod_info->block_type==SHORT_TYPE) {
if (add_bits[ch]<mean_bits/4) add_bits[ch]=mean_bits/4;
}
/* at most increase bits by 1.5*average */
if (add_bits[ch] > .75*mean_bits) add_bits[ch]=mean_bits*.75;
if (add_bits[ch] < 0) add_bits[ch]=0;
if ((targ_bits[ch]+add_bits[ch]) > MAX_BITS)
add_bits[ch]=Max(0, MAX_BITS-targ_bits[ch]);
}
bits += add_bits[ch];
}
if (bits > extra_bits)
for (ch=0 ; ch < gfc->channels_out ; ch ++) {
add_bits[ch] = (extra_bits*add_bits[ch])/bits;
}
for (ch=0 ; ch < gfc->channels_out ; ch ++) {
targ_bits[ch] = targ_bits[ch] + add_bits[ch];
extra_bits -= add_bits[ch];
}
return max_bits;
}
void reduce_side(int targ_bits[2],FLOAT8 ms_ener_ratio,int mean_bits,int max_bits)
{
int move_bits;
FLOAT fac;
/* ms_ener_ratio = 0: allocate 66/33 mid/side fac=.33
* ms_ener_ratio =.5: allocate 50/50 mid/side fac= 0 */
/* 75/25 split is fac=.5 */
/* float fac = .50*(.5-ms_ener_ratio[gr])/.5;*/
fac = .33*(.5-ms_ener_ratio)/.5;
if (fac<0) fac=0;
if (fac>.5) fac=.5;
/* number of bits to move from side channel to mid channel */
/* move_bits = fac*targ_bits[1]; */
move_bits = fac*.5*(targ_bits[0]+targ_bits[1]);
if (move_bits > MAX_BITS - targ_bits[0]) {
move_bits = MAX_BITS - targ_bits[0];
}
if (move_bits<0) move_bits=0;
if (targ_bits[1] >= 125) {
/* dont reduce side channel below 125 bits */
if (targ_bits[1]-move_bits > 125) {
/* if mid channel already has 2x more than average, dont bother */
/* mean_bits = bits per granule (for both channels) */
if (targ_bits[0] < mean_bits)
targ_bits[0] += move_bits;
targ_bits[1] -= move_bits;
} else {
targ_bits[0] += targ_bits[1] - 125;
targ_bits[1] = 125;
}
}
move_bits=targ_bits[0]+targ_bits[1];
if (move_bits > max_bits) {
targ_bits[0]=(max_bits*targ_bits[0])/move_bits;
targ_bits[1]=(max_bits*targ_bits[1])/move_bits;
}
}
#if 0
FLOAT8 dreinorm (FLOAT8 a, FLOAT8 b, FLOAT8 c)
{
return pow(pow(a,3.)+pow(b,3.)+pow(c,3.),1./3.);
}
#endif
/*************************************************************************/
/* calc_xmin */
/*************************************************************************/
/*
Calculate the allowed distortion for each scalefactor band,
as determined by the psychoacoustic model.
xmin(sb) = ratio(sb) * en(sb) / bw(sb)
returns number of sfb's with energy > ATH
*/
int calc_xmin(
lame_global_flags *gfp,
const FLOAT8 xr [576],
const III_psy_ratio * const ratio,
const gr_info * const cod_info,
III_psy_xmin * const l3_xmin )
{
lame_internal_flags *gfc=gfp->internal_flags;
int sfb,j,start, end, bw,l, b, ath_over=0;
FLOAT8 en0, xmin, ener;
if (cod_info->block_type==SHORT_TYPE) {
for ( j=0, sfb = 0; sfb < SBMAX_s; sfb++ ) {
start = gfc->scalefac_band.s[ sfb ];
end = gfc->scalefac_band.s[ sfb + 1 ];
bw = end - start;
for ( b = 0; b < 3; b++ ) {
for (en0 = 0.0, l = start; l < end; l++) {
ener = xr[j++];
ener = ener * ener;
en0 += ener;
}
en0 /= bw;
if (gfp->ATHonly || gfp->ATHshort) {
xmin = gfc->ATH->adjust * gfc->ATH->s[sfb];
} else {
xmin = ratio->en.s[sfb][b];
if (xmin > 0.0)
xmin = en0 * ratio->thm.s[sfb][b] * gfc->masking_lower / xmin;
xmin = Max(gfc->ATH->adjust * gfc->ATH->s[sfb], xmin);
}
l3_xmin->s[sfb][b] = xmin * bw;
if (gfc->nsPsy.use) {
if (sfb <= 5) {
l3_xmin->s[sfb][b] *= gfc->nsPsy.bass;
} else if (sfb <= 10) {
l3_xmin->s[sfb][b] *= gfc->nsPsy.alto;
} else {
l3_xmin->s[sfb][b] *= gfc->nsPsy.treble;
}
}
if (en0 > gfc->ATH->adjust * gfc->ATH->s[sfb]) ath_over++;
if (gfc->nsPsy.use && (gfp->VBR == vbr_off || gfp->VBR == vbr_abr) && gfp->quality <= 1)
l3_xmin->s[sfb][b] *= 0.001;
}
}
if (gfp->useTemporal) {
for (sfb = 0; sfb < SBMAX_s; sfb++ ) {
for ( b = 1; b < 3; b++ ) {
xmin = l3_xmin->s[sfb][b] * (1.0 - gfc->decay)
+ l3_xmin->s[sfb][b-1] * gfc->decay;
if (l3_xmin->s[sfb][b] < xmin)
l3_xmin->s[sfb][b] = xmin;
}
}
}
}else{
if (gfc->nsPsy.use) {
for ( sfb = 0; sfb < SBMAX_l; sfb++ ){
start = gfc->scalefac_band.l[ sfb ];
end = gfc->scalefac_band.l[ sfb+1 ];
for (en0 = 0.0, l = start; l < end; l++ ) {
ener = xr[l] * xr[l];
en0 += ener;
}
if (gfp->ATHonly) {
xmin=gfc->ATH->adjust * gfc->ATH->l[sfb];
} else {
xmin = ratio->en.l[sfb];
if (xmin > 0.0)
xmin = en0 * ratio->thm.l[sfb] * gfc->masking_lower / xmin;
xmin=Max(gfc->ATH->adjust * gfc->ATH->l[sfb], xmin);
}
l3_xmin->l[sfb]=xmin;
if (sfb <= 6) {
l3_xmin->l[sfb] *= gfc->nsPsy.bass;
} else if (sfb <= 13) {
l3_xmin->l[sfb] *= gfc->nsPsy.alto;
} else {
l3_xmin->l[sfb] *= gfc->nsPsy.treble;
}
if (en0 > gfc->ATH->adjust * gfc->ATH->l[sfb]) ath_over++;
if ((gfp->VBR == vbr_off || gfp->VBR == vbr_abr) && gfp->quality <= 1)
l3_xmin->l[sfb] *= 0.001;
}
} else {
for ( sfb = 0; sfb < SBMAX_l; sfb++ ){
start = gfc->scalefac_band.l[ sfb ];
end = gfc->scalefac_band.l[ sfb+1 ];
bw = end - start;
for (en0 = 0.0, l = start; l < end; l++ ) {
ener = xr[l] * xr[l];
en0 += ener;
}
en0 /= bw;
if (gfp->ATHonly) {
xmin=gfc->ATH->adjust * gfc->ATH->l[sfb];
} else {
xmin = ratio->en.l[sfb];
if (xmin > 0.0)
xmin = en0 * ratio->thm.l[sfb] * gfc->masking_lower / xmin;
xmin=Max(gfc->ATH->adjust * gfc->ATH->l[sfb], xmin);
}
l3_xmin->l[sfb]=xmin*bw;
if (en0 > gfc->ATH->adjust * gfc->ATH->l[sfb]) ath_over++;
}
}
}
return ath_over;
}
/*************************************************************************/
/* calc_noise */
/*************************************************************************/
// -oo dB => -1.00
// - 6 dB => -0.97
// - 3 dB => -0.80
// - 2 dB => -0.64
// - 1 dB => -0.38
// 0 dB => 0.00
// + 1 dB => +0.49
// + 2 dB => +1.06
// + 3 dB => +1.68
// + 6 dB => +3.69
// +10 dB => +6.45
double penalties ( double noise )
{
return log ( 0.368 + 0.632 * noise * noise * noise );
}
/* mt 5/99: Function: Improved calc_noise for a single channel */
int calc_noise(
const lame_internal_flags * const gfc,
const FLOAT8 xr [576],
const int ix [576],
const gr_info * const cod_info,
const III_psy_xmin * const l3_xmin,
const III_scalefac_t * const scalefac,
III_psy_xmin * xfsf,
calc_noise_result * const res )
{
int sfb,start, end, j,l, i, over=0;
FLOAT8 sum;
int count=0;
FLOAT8 noise,noise_db;
FLOAT8 over_noise = 1; /* 0 dB relative to masking */
FLOAT8 over_noise_db = 0;
FLOAT8 tot_noise = 1; /* 0 dB relative to masking */
FLOAT8 tot_noise_db = 0; /* 0 dB relative to masking */
FLOAT8 max_noise = 1E-20; /* -200 dB relative to masking */
double klemm_noise = 1E-37;
if (cod_info->block_type == SHORT_TYPE) {
int max_index = gfc->sfb21_extra ? SBMAX_s : SBPSY_s;
for ( j=0, sfb = 0; sfb < max_index; sfb++ ) {
start = gfc->scalefac_band.s[ sfb ];
end = gfc->scalefac_band.s[ sfb+1 ];
for ( i = 0; i < 3; i++ ) {
FLOAT8 step;
int s;
s = (scalefac->s[sfb][i] << (cod_info->scalefac_scale + 1))
+ cod_info->subblock_gain[i] * 8;
s = cod_info->global_gain - s;
assert(s<Q_MAX);
assert(s>=0);
step = POW20(s);
sum = 0.0;
l = start;
do {
FLOAT8 temp;
temp = pow43[ix[j]];
temp *= step;
temp -= fabs(xr[j]);
++j;
sum += temp * temp;
l++;
} while (l < end);
noise = xfsf->s[sfb][i] = sum / l3_xmin->s[sfb][i];
max_noise = Max(max_noise,noise);
klemm_noise += penalties (noise);
noise_db=10*log10(Max(noise,1E-20));
/* multiplying here is adding in dB, but will overflow */
//tot_noise *= Max(noise, 1E-20);
tot_noise_db += noise_db;
if (noise > 1) {
over++;
/* multiplying here is adding in dB, but can overflow */
//over_noise *= noise;
over_noise_db += noise_db;
}
count++;
}
}
}else{ /* cod_info->block_type == SHORT_TYPE */
int max_index = gfc->sfb21_extra ? SBMAX_l : SBPSY_l;
for ( sfb = 0; sfb < max_index; sfb++ ) {
FLOAT8 step;
int s = scalefac->l[sfb];
if (cod_info->preflag)
s += pretab[sfb];
s = cod_info->global_gain - (s << (cod_info->scalefac_scale + 1));
assert(s<Q_MAX);
assert(s>=0);
step = POW20(s);
start = gfc->scalefac_band.l[ sfb ];
end = gfc->scalefac_band.l[ sfb+1 ];
for ( sum = 0.0, l = start; l < end; l++ ) {
FLOAT8 temp;
temp = fabs(xr[l]) - pow43[ix[l]] * step;
sum += temp * temp;
}
noise = xfsf->l[sfb] = sum / l3_xmin->l[sfb];
max_noise=Max(max_noise,noise);
klemm_noise += penalties (noise);
noise_db=10*log10(Max(noise,1E-20));
/* multiplying here is adding in dB, but can overflow */
//tot_noise *= Max(noise, 1E-20);
tot_noise_db += noise_db;
if (noise > 1) {
over++;
/* multiplying here is adding in dB -but can overflow */
//over_noise *= noise;
over_noise_db += noise_db;
}
count++;
}
} /* cod_info->block_type == SHORT_TYPE */
/* normalization at this point by "count" is not necessary, since
* the values are only used to compare with previous values */
res->tot_count = count;
res->over_count = over;
/* convert to db. DO NOT CHANGE THESE */
/* tot_noise = is really the average over each sfb of:
[noise(db) - allowed_noise(db)]
and over_noise is the same average, only over only the
bands with noise > allowed_noise.
*/
//res->tot_noise = 10.*log10(Max(1e-20,tot_noise ));
//res->over_noise = 10.*log10(Max(1e+00,over_noise));
res->tot_noise = tot_noise_db;
res->over_noise = over_noise_db;
res->max_noise = 10.*log10(Max(1e-20,max_noise ));
res->klemm_noise = 10.*log10(Max(1e-20,klemm_noise));
return over;
}
/************************************************************************
*
* set_pinfo()
*
* updates plotting data
*
* Mark Taylor 2000-??-??
*
* Robert Hegemann: moved noise/distortion calc into it
*
************************************************************************/
static
void set_pinfo (
lame_global_flags *gfp,
const gr_info * const cod_info,
const III_psy_ratio * const ratio,
const III_scalefac_t * const scalefac,
const FLOAT8 xr[576],
const int l3_enc[576],
const int gr,
const int ch )
{
lame_internal_flags *gfc=gfp->internal_flags;
int sfb;
int j,i,l,start,end,bw;
FLOAT8 en0,en1;
FLOAT ifqstep = ( cod_info->scalefac_scale == 0 ) ? .5 : 1.0;
III_psy_xmin l3_xmin;
calc_noise_result noise;
III_psy_xmin xfsf;
calc_xmin (gfp,xr, ratio, cod_info, &l3_xmin);
calc_noise (gfc, xr, l3_enc, cod_info, &l3_xmin, scalefac, &xfsf, &noise);
if (cod_info->block_type == SHORT_TYPE) {
for (j=0, sfb = 0; sfb < SBMAX_s; sfb++ ) {
start = gfc->scalefac_band.s[ sfb ];
end = gfc->scalefac_band.s[ sfb + 1 ];
bw = end - start;
for ( i = 0; i < 3; i++ ) {
for ( en0 = 0.0, l = start; l < end; l++ ) {
en0 += xr[j] * xr[j];
++j;
}
en0=Max(en0/bw,1e-20);
#if 0
{
double tot1,tot2;
if (sfb<SBMAX_s-1) {
if (sfb==0) {
tot1=0;
tot2=0;
}
tot1 += en0;
tot2 += ratio->en.s[sfb][i];
DEBUGF("%i %i sfb=%i mdct=%f fft=%f fft-mdct=%f db \n",
gr,ch,sfb,
10*log10(Max(1e-25,ratio->en.s[sfb][i])),
10*log10(Max(1e-25,en0)),
10*log10((Max(1e-25,en0)/Max(1e-25,ratio->en.s[sfb][i]))));
if (sfb==SBMAX_s-2) {
DEBUGF("%i %i toti %f %f ratio=%f db \n",gr,ch,
10*log10(Max(1e-25,tot2)),
10*log(Max(1e-25,tot1)),
10*log10(Max(1e-25,tot1)/(Max(1e-25,tot2))));
}
}
/*
masking: multiplied by en0/fft_energy
average seems to be about -135db.
*/
}
#endif
/* convert to MDCT units */
en1=1e15; /* scaling so it shows up on FFT plot */
gfc->pinfo->xfsf_s[gr][ch][3*sfb+i]
= en1*xfsf.s[sfb][i]*l3_xmin.s[sfb][i]/bw;
gfc->pinfo->en_s[gr][ch][3*sfb+i] = en1*en0;
if (ratio->en.s[sfb][i]>0)
en0 = en0/ratio->en.s[sfb][i];
else
en0=0;
if (gfp->ATHonly || gfp->ATHshort)
en0=0;
gfc->pinfo->thr_s[gr][ch][3*sfb+i] =
en1*Max(en0*ratio->thm.s[sfb][i],gfc->ATH->s[sfb]);
/* there is no scalefactor bands >= SBPSY_s */
if (sfb < SBPSY_s) {
gfc->pinfo->LAMEsfb_s[gr][ch][3*sfb+i]=
-ifqstep*scalefac->s[sfb][i];
} else {
gfc->pinfo->LAMEsfb_s[gr][ch][3*sfb+i]=0;
}
gfc->pinfo->LAMEsfb_s[gr][ch][3*sfb+i] -=
2*cod_info->subblock_gain[i];
}
}
} else {
for ( sfb = 0; sfb < SBMAX_l; sfb++ ) {
start = gfc->scalefac_band.l[ sfb ];
end = gfc->scalefac_band.l[ sfb+1 ];
bw = end - start;
for ( en0 = 0.0, l = start; l < end; l++ )
en0 += xr[l] * xr[l];
if (!gfc->nsPsy.use) en0/=bw;
/*
DEBUGF("diff = %f \n",10*log10(Max(ratio[gr][ch].en.l[sfb],1e-20))
-(10*log10(en0)+150));
*/
#if 0
{
double tot1,tot2;
if (sfb==0) {
tot1=0;
tot2=0;
}
tot1 += en0;
tot2 += ratio->en.l[sfb];
DEBUGF("%i %i sfb=%i mdct=%f fft=%f fft-mdct=%f db \n",
gr,ch,sfb,
10*log10(Max(1e-25,ratio->en.l[sfb])),
10*log10(Max(1e-25,en0)),
10*log10((Max(1e-25,en0)/Max(1e-25,ratio->en.l[sfb]))));
if (sfb==SBMAX_l-1) {
DEBUGF("%i %i toti %f %f ratio=%f db \n",
gr,ch,
10*log10(Max(1e-25,tot2)),
10*log(Max(1e-25,tot1)),
10*log10(Max(1e-25,tot1)/(Max(1e-25,tot2))));
}
/*
masking: multiplied by en0/fft_energy
average seems to be about -147db.
*/
}
#endif
/* convert to MDCT units */
en1=1e15; /* scaling so it shows up on FFT plot */
gfc->pinfo->xfsf[gr][ch][sfb] = en1*xfsf.l[sfb]*l3_xmin.l[sfb]/bw;
gfc->pinfo->en[gr][ch][sfb] = en1*en0;
if (ratio->en.l[sfb]>0)
en0 = en0/ratio->en.l[sfb];
else
en0=0;
if (gfp->ATHonly)
en0=0;
gfc->pinfo->thr[gr][ch][sfb] =
en1*Max(en0*ratio->thm.l[sfb],gfc->ATH->l[sfb]);
/* there is no scalefactor bands >= SBPSY_l */
if (sfb<SBPSY_l) {
if (scalefac->l[sfb]<0) /* scfsi! */
gfc->pinfo->LAMEsfb[gr][ch][sfb] =
gfc->pinfo->LAMEsfb[0][ch][sfb];
else
gfc->pinfo->LAMEsfb[gr][ch][sfb] = -ifqstep*scalefac->l[sfb];
}else{
gfc->pinfo->LAMEsfb[gr][ch][sfb] = 0;
}
if (cod_info->preflag && sfb>=11)
gfc->pinfo->LAMEsfb[gr][ch][sfb] -= ifqstep*pretab[sfb];
} /* for sfb */
} /* block type long */
gfc->pinfo->LAMEqss [gr][ch] = cod_info->global_gain;
gfc->pinfo->LAMEmainbits[gr][ch] = cod_info->part2_3_length;
gfc->pinfo->LAMEsfbits [gr][ch] = cod_info->part2_length;
gfc->pinfo->over [gr][ch] = noise.over_count;
gfc->pinfo->max_noise [gr][ch] = noise.max_noise;
gfc->pinfo->over_noise[gr][ch] = noise.over_noise;
gfc->pinfo->tot_noise [gr][ch] = noise.tot_noise;
}
/************************************************************************
*
* set_frame_pinfo()
*
* updates plotting data for a whole frame
*
* Robert Hegemann 2000-10-21
*
************************************************************************/
void set_frame_pinfo(
lame_global_flags *gfp,
FLOAT8 xr [2][2][576],
III_psy_ratio ratio [2][2],
int l3_enc [2][2][576],
III_scalefac_t scalefac [2][2] )
{
lame_internal_flags *gfc=gfp->internal_flags;
unsigned int gr, ch, sfb;
int act_l3enc[576];
III_scalefac_t act_scalefac [2];
int scsfi[2] = {0,0};
gfc->masking_lower = 1.0;
/* reconstruct the scalefactors in case SCSFI was used
*/
for (ch = 0; ch < gfc->channels_out; ch ++) {
for (sfb = 0; sfb < SBMAX_l; sfb ++) {
if (scalefac[1][ch].l[sfb] == -1) {/* scfsi */
act_scalefac[ch].l[sfb] = scalefac[0][ch].l[sfb];
scsfi[ch] = 1;
} else {
act_scalefac[ch].l[sfb] = scalefac[1][ch].l[sfb];
}
}
}
/* for every granule and channel patch l3_enc and set info
*/
for (gr = 0; gr < gfc->mode_gr; gr ++) {
for (ch = 0; ch < gfc->channels_out; ch ++) {
int i;
gr_info *cod_info = &gfc->l3_side.gr[gr].ch[ch].tt;
/* revert back the sign of l3enc */
for ( i = 0; i < 576; i++) {
if (xr[gr][ch][i] < 0)
act_l3enc[i] = -l3_enc[gr][ch][i];
else
act_l3enc[i] = +l3_enc[gr][ch][i];
}
if (gr == 1 && scsfi[ch])
set_pinfo (gfp, cod_info, &ratio[gr][ch], &act_scalefac[ch],
xr[gr][ch], act_l3enc, gr, ch);
else
set_pinfo (gfp, cod_info, &ratio[gr][ch], &scalefac[gr][ch],
xr[gr][ch], act_l3enc, gr, ch);
} /* for ch */
} /* for gr */
}
/*********************************************************************
* nonlinear quantization of xr
* More accurate formula than the ISO formula. Takes into account
* the fact that we are quantizing xr -> ix, but we want ix^4/3 to be
* as close as possible to x^4/3. (taking the nearest int would mean
* ix is as close as possible to xr, which is different.)
* From Segher Boessenkool <[email protected]> 11/1999
* ASM optimization from
* Mathew Hendry <[email protected]> 11/1999
* Acy Stapp <[email protected]> 11/1999
* Takehiro Tominaga <[email protected]> 11/1999
* 9/00: ASM code removed in favor of IEEE754 hack. If you need
* the ASM code, check CVS circa Aug 2000.
*********************************************************************/
#ifdef TAKEHIRO_IEEE754_HACK
typedef union {
float f;
int i;
} fi_union;
#define MAGIC_FLOAT (65536*(128))
#define MAGIC_INT 0x4b000000
void quantize_xrpow(const FLOAT8 xp[576], int pi[576], FLOAT8 istep)
{
/* quantize on xr^(3/4) instead of xr */
int j;
fi_union *fi;
fi = (fi_union *)pi;
for (j = 576 / 4 - 1; j >= 0; --j) {
double x0 = istep * xp[0];
double x1 = istep * xp[1];
double x2 = istep * xp[2];
double x3 = istep * xp[3];
x0 += MAGIC_FLOAT; fi[0].f = x0;
x1 += MAGIC_FLOAT; fi[1].f = x1;
x2 += MAGIC_FLOAT; fi[2].f = x2;
x3 += MAGIC_FLOAT; fi[3].f = x3;
fi[0].f = x0 + (adj43asm - MAGIC_INT)[fi[0].i];
fi[1].f = x1 + (adj43asm - MAGIC_INT)[fi[1].i];
fi[2].f = x2 + (adj43asm - MAGIC_INT)[fi[2].i];
fi[3].f = x3 + (adj43asm - MAGIC_INT)[fi[3].i];
fi[0].i -= MAGIC_INT;
fi[1].i -= MAGIC_INT;
fi[2].i -= MAGIC_INT;
fi[3].i -= MAGIC_INT;
fi += 4;
xp += 4;
}
}
# define ROUNDFAC -0.0946
void quantize_xrpow_ISO(const FLOAT8 xp[576], int pi[576], FLOAT8 istep)
{
/* quantize on xr^(3/4) instead of xr */
int j;
fi_union *fi;
fi = (fi_union *)pi;
for (j=576/4 - 1;j>=0;j--) {
fi[0].f = istep * xp[0] + (ROUNDFAC + MAGIC_FLOAT);
fi[1].f = istep * xp[1] + (ROUNDFAC + MAGIC_FLOAT);
fi[2].f = istep * xp[2] + (ROUNDFAC + MAGIC_FLOAT);
fi[3].f = istep * xp[3] + (ROUNDFAC + MAGIC_FLOAT);
fi[0].i -= MAGIC_INT;
fi[1].i -= MAGIC_INT;
fi[2].i -= MAGIC_INT;
fi[3].i -= MAGIC_INT;
fi+=4;
xp+=4;
}
}
#else
/*********************************************************************
* XRPOW_FTOI is a macro to convert floats to ints.
* if XRPOW_FTOI(x) = nearest_int(x), then QUANTFAC(x)=adj43asm[x]
* ROUNDFAC= -0.0946
*
* if XRPOW_FTOI(x) = floor(x), then QUANTFAC(x)=asj43[x]
* ROUNDFAC=0.4054
*
* Note: using floor() or (int) is extermely slow. On machines where
* the TAKEHIRO_IEEE754_HACK code above does not work, it is worthwile
* to write some ASM for XRPOW_FTOI().
*********************************************************************/
#define XRPOW_FTOI(src,dest) ((dest) = (int)(src))
#define QUANTFAC(rx) adj43[rx]
#define ROUNDFAC 0.4054
void quantize_xrpow(const FLOAT8 xr[576], int ix[576], FLOAT8 istep) {
/* quantize on xr^(3/4) instead of xr */
/* from [email protected]. Reported to be 2x faster than
the above code (when not using ASM) on PowerPC */
int j;
for ( j = 576/8; j > 0; --j) {
FLOAT8 x1, x2, x3, x4, x5, x6, x7, x8;
int rx1, rx2, rx3, rx4, rx5, rx6, rx7, rx8;
x1 = *xr++ * istep;
x2 = *xr++ * istep;
XRPOW_FTOI(x1, rx1);
x3 = *xr++ * istep;
XRPOW_FTOI(x2, rx2);
x4 = *xr++ * istep;
XRPOW_FTOI(x3, rx3);
x5 = *xr++ * istep;
XRPOW_FTOI(x4, rx4);
x6 = *xr++ * istep;
XRPOW_FTOI(x5, rx5);
x7 = *xr++ * istep;
XRPOW_FTOI(x6, rx6);
x8 = *xr++ * istep;
XRPOW_FTOI(x7, rx7);
x1 += QUANTFAC(rx1);
XRPOW_FTOI(x8, rx8);
x2 += QUANTFAC(rx2);
XRPOW_FTOI(x1,*ix++);
x3 += QUANTFAC(rx3);
XRPOW_FTOI(x2,*ix++);
x4 += QUANTFAC(rx4);
XRPOW_FTOI(x3,*ix++);
x5 += QUANTFAC(rx5);
XRPOW_FTOI(x4,*ix++);
x6 += QUANTFAC(rx6);
XRPOW_FTOI(x5,*ix++);
x7 += QUANTFAC(rx7);
XRPOW_FTOI(x6,*ix++);
x8 += QUANTFAC(rx8);
XRPOW_FTOI(x7,*ix++);
XRPOW_FTOI(x8,*ix++);
}
}
void quantize_xrpow_ISO( const FLOAT8 xr[576], int ix[576], FLOAT8 istep )
{
/* quantize on xr^(3/4) instead of xr */
const FLOAT8 compareval0 = (1.0 - 0.4054)/istep;
int j;
/* depending on architecture, it may be worth calculating a few more
compareval's.
eg. compareval1 = (2.0 - 0.4054/istep);
.. and then after the first compare do this ...
if compareval1>*xr then ix = 1;
On a pentium166, it's only worth doing the one compare (as done here),
as the second compare becomes more expensive than just calculating
the value. Architectures with slow FP operations may want to add some
more comparevals. try it and send your diffs statistically speaking
73% of all xr*istep values give ix=0
16% will give 1
4% will give 2
*/
for (j=576;j>0;j--) {
if (compareval0 > *xr) {
*(ix++) = 0;
xr++;
} else {
/* *(ix++) = (int)( istep*(*(xr++)) + 0.4054); */
XRPOW_FTOI( istep*(*(xr++)) + ROUNDFAC , *(ix++) );
}
}
}
#endif