ref: 3b918bac54ceb66a21acaf99845867cb2a8afe02
dir: /libcelt/vq.c/
/* Copyright (c) 2007-2008 CSIRO
Copyright (c) 2007-2009 Xiph.Org Foundation
Written by Jean-Marc Valin */
/*
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions
are met:
- Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
- 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.
- Neither the name of the Xiph.org Foundation 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 COPYRIGHT HOLDERS AND CONTRIBUTORS
``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 FOUNDATION OR
CONTRIBUTORS 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.
*/
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "mathops.h"
#include "cwrs.h"
#include "vq.h"
#include "arch.h"
#include "os_support.h"
#include "rate.h"
#ifndef M_PI
#define M_PI 3.141592653
#endif
#if 0
static void frac_hadamard1(celt_norm *X, int len, int stride, celt_word16 c, celt_word16 s)
{
int j;
celt_norm *x, *y;
celt_norm * end;
j = 0;
x = X;
y = X+stride;
end = X+len;
do
{
celt_norm x1, x2;
x1 = *x;
x2 = *y;
*x++ = EXTRACT16(SHR32(MULT16_16(c,x1) + MULT16_16(s,x2),15));
*y++ = EXTRACT16(SHR32(MULT16_16(s,x1) - MULT16_16(c,x2),15));
j++;
if (j>=stride)
{
j=0;
x+=stride;
y+=stride;
}
} while (y<end);
/* Reverse samples so that the next level starts from the other end */
for (j=0;j<len>>1;j++)
{
celt_norm tmp = X[j];
X[j] = X[len-j-1];
X[len-j-1] = tmp;
}
}
#define MAX_LEVELS 8
static void pseudo_hadamard(celt_norm *X, int len, int dir, int stride, int K)
{
int i, N=0;
int transient;
celt_word16 gain, theta;
int istride[MAX_LEVELS];
celt_word16 c[MAX_LEVELS], s[MAX_LEVELS];
if (K >= len)
return;
transient = stride>1;
/*if (len>=30)
{
for (i=0;i<len;i++)
X[i] = 0;
X[30] = 1;
dir = -1;
transient = 1;
}*/
gain = celt_div((celt_word32)MULT16_16(Q15_ONE,len),(celt_word32)(3+len+4*K));
/* FIXME: Make that HALF16 instead of HALF32 */
theta = MIN16(QCONST16(.25f,15),HALF32(MULT16_16_Q15(gain,gain)));
c[0] = celt_cos_norm(EXTEND32(theta));
s[0] = celt_cos_norm(EXTEND32(SUB16(Q15ONE,theta))); /* sin(theta) */
do {
istride[N] = stride;
stride *= 2;
c[N] = c[0];
s[N] = s[0];
N++;
} while (N<MAX_LEVELS && stride < len);
/* This should help a little bit with the transients */
if (transient)
c[0] = s[0] = QCONST16(.7071068f, 15);
/* Needs to be < 0 to prevent gaps on the side of the spreading */
if (dir < 0)
{
for (i=0;i<N;i++)
frac_hadamard1(X, len, istride[i], c[i], s[i]);
} else {
for (i=N-1;i>=0;i--)
frac_hadamard1(X, len, istride[i], c[i], s[i]);
}
/* Undo last reversal */
for (i=0;i<len>>1;i++)
{
celt_norm tmp = X[i];
X[i] = X[len-i-1];
X[len-i-1] = tmp;
}
/*if (len>=30)
{
for (i=0;i<len;i++)
printf ("%f ", X[i]);
printf ("\n");
exit(0);
}*/
}
#endif
static void exp_rotation1(celt_norm *X, int len, int dir, int stride, celt_word16 c, celt_word16 s)
{
int i;
celt_norm *Xptr;
if (dir>0)
s = -s;
Xptr = X;
for (i=0;i<len-stride;i++)
{
celt_norm x1, x2;
x1 = Xptr[0];
x2 = Xptr[stride];
Xptr[stride] = EXTRACT16(SHR32(MULT16_16(c,x2) + MULT16_16(s,x1), 15));
*Xptr++ = EXTRACT16(SHR32(MULT16_16(c,x1) - MULT16_16(s,x2), 15));
}
Xptr = &X[len-2*stride-1];
for (i=len-2*stride-1;i>=0;i--)
{
celt_norm x1, x2;
x1 = Xptr[0];
x2 = Xptr[stride];
Xptr[stride] = EXTRACT16(SHR32(MULT16_16(c,x2) + MULT16_16(s,x1), 15));
*Xptr-- = EXTRACT16(SHR32(MULT16_16(c,x1) - MULT16_16(s,x2), 15));
}
}
static void exp_rotation(celt_norm *X, int len, int dir, int stride, int K)
{
celt_word16 c, s;
celt_word16 gain, theta;
int stride2=0;
/*int i;
if (len>=30)
{
for (i=0;i<len;i++)
X[i] = 0;
X[14] = 1;
K=5;
}*/
/*if (stride>1)
{
pseudo_hadamard(X, len, dir, stride, K);
return;
}*/
if (2*K>=len)
return;
gain = celt_div((celt_word32)MULT16_16(Q15_ONE,len),(celt_word32)(len+10*K));
/* FIXME: Make that HALF16 instead of HALF32 */
theta = HALF32(MULT16_16_Q15(gain,gain));
c = celt_cos_norm(EXTEND32(theta));
s = celt_cos_norm(EXTEND32(SUB16(Q15ONE,theta))); /* sin(theta) */
#if 0
if (len>=8*stride)
stride2 = stride*floor(.5+sqrt(len*1.f/stride));
#else
if (len>=8*stride)
{
stride2 = 1;
/* This is just a simple way of computing sqrt(len/stride) with rounding.
It's basically incrementing long as (stride2+0.5)^2 < len/stride.
I _think_ it is bit-exact */
while ((stride2*stride2+stride2)*stride + (stride>>2) < len)
stride2++;
stride2 *= stride;
}
#endif
if (dir < 0)
{
if (stride2)
exp_rotation1(X, len, dir, stride2, s, c);
exp_rotation1(X, len, dir, stride, c, s);
} else {
exp_rotation1(X, len, dir, stride, c, s);
if (stride2)
exp_rotation1(X, len, dir, stride2, s, c);
}
/*if (len>=30)
{
for (i=0;i<len;i++)
printf ("%f ", X[i]);
printf ("\n");
exit(0);
}*/
}
/** Takes the pitch vector and the decoded residual vector, computes the gain
that will give ||p+g*y||=1 and mixes the residual with the pitch. */
static void normalise_residual(int * restrict iy, celt_norm * restrict X, int N, int K, celt_word32 Ryy)
{
int i;
#ifdef FIXED_POINT
int k;
#endif
celt_word32 t;
celt_word16 g;
#ifdef FIXED_POINT
k = celt_ilog2(Ryy)>>1;
#endif
t = VSHR32(Ryy, (k-7)<<1);
g = celt_rsqrt_norm(t);
i=0;
do
X[i] = EXTRACT16(PSHR32(MULT16_16(g, iy[i]), k+1));
while (++i < N);
}
void alg_quant(celt_norm *X, int N, int K, int spread, int resynth, ec_enc *enc)
{
VARDECL(celt_norm, y);
VARDECL(int, iy);
VARDECL(celt_word16, signx);
int j, is;
celt_word16 s;
int pulsesLeft;
celt_word32 sum;
celt_word32 xy, yy;
int N_1; /* Inverse of N, in Q14 format (even for float) */
#ifdef FIXED_POINT
int yshift;
#endif
SAVE_STACK;
K = get_pulses(K);
#ifdef FIXED_POINT
yshift = 13-celt_ilog2(K);
#endif
ALLOC(y, N, celt_norm);
ALLOC(iy, N, int);
ALLOC(signx, N, celt_word16);
N_1 = 512/N;
if (spread)
exp_rotation(X, N, 1, spread, K);
sum = 0;
j=0; do {
if (X[j]>0)
signx[j]=1;
else {
signx[j]=-1;
X[j]=-X[j];
}
iy[j] = 0;
y[j] = 0;
} while (++j<N);
xy = yy = 0;
pulsesLeft = K;
/* Do a pre-search by projecting on the pyramid */
if (K > (N>>1))
{
celt_word16 rcp;
j=0; do {
sum += X[j];
} while (++j<N);
#ifdef FIXED_POINT
if (sum <= K)
#else
if (sum <= EPSILON)
#endif
{
X[0] = QCONST16(1.f,14);
j=1; do
X[j]=0;
while (++j<N);
sum = QCONST16(1.f,14);
}
/* Do we have sufficient accuracy here? */
rcp = EXTRACT16(MULT16_32_Q16(K-1, celt_rcp(sum)));
j=0; do {
#ifdef FIXED_POINT
/* It's really important to round *towards zero* here */
iy[j] = MULT16_16_Q15(X[j],rcp);
#else
iy[j] = floor(rcp*X[j]);
#endif
y[j] = SHL16(iy[j],yshift);
yy = MAC16_16(yy, y[j],y[j]);
xy = MAC16_16(xy, X[j],y[j]);
y[j] *= 2;
pulsesLeft -= iy[j];
} while (++j<N);
}
celt_assert2(pulsesLeft>=1, "Allocated too many pulses in the quick pass");
while (pulsesLeft > 0)
{
int pulsesAtOnce=1;
int best_id;
celt_word16 magnitude;
celt_word32 best_num = -VERY_LARGE16;
celt_word16 best_den = 0;
#ifdef FIXED_POINT
int rshift;
#endif
/* Decide on how many pulses to find at once */
pulsesAtOnce = (pulsesLeft*N_1)>>9; /* pulsesLeft/N */
if (pulsesAtOnce<1)
pulsesAtOnce = 1;
#ifdef FIXED_POINT
rshift = yshift+1+celt_ilog2(K-pulsesLeft+pulsesAtOnce);
#endif
magnitude = SHL16(pulsesAtOnce, yshift);
best_id = 0;
/* The squared magnitude term gets added anyway, so we might as well
add it outside the loop */
yy = MAC16_16(yy, magnitude,magnitude);
/* Choose between fast and accurate strategy depending on where we are in the search */
/* This should ensure that anything we can process will have a better score */
j=0;
do {
celt_word16 Rxy, Ryy;
/* Select sign based on X[j] alone */
s = magnitude;
/* Temporary sums of the new pulse(s) */
Rxy = EXTRACT16(SHR32(MAC16_16(xy, s,X[j]),rshift));
/* We're multiplying y[j] by two so we don't have to do it here */
Ryy = EXTRACT16(SHR32(MAC16_16(yy, s,y[j]),rshift));
/* Approximate score: we maximise Rxy/sqrt(Ryy) (we're guaranteed that
Rxy is positive because the sign is pre-computed) */
Rxy = MULT16_16_Q15(Rxy,Rxy);
/* The idea is to check for num/den >= best_num/best_den, but that way
we can do it without any division */
/* OPT: Make sure to use conditional moves here */
if (MULT16_16(best_den, Rxy) > MULT16_16(Ryy, best_num))
{
best_den = Ryy;
best_num = Rxy;
best_id = j;
}
} while (++j<N);
j = best_id;
is = pulsesAtOnce;
s = SHL16(is, yshift);
/* Updating the sums of the new pulse(s) */
xy = xy + MULT16_16(s,X[j]);
/* We're multiplying y[j] by two so we don't have to do it here */
yy = yy + MULT16_16(s,y[j]);
/* Only now that we've made the final choice, update y/iy */
/* Multiplying y[j] by 2 so we don't have to do it everywhere else */
y[j] += 2*s;
iy[j] += is;
pulsesLeft -= pulsesAtOnce;
}
j=0;
do {
X[j] = MULT16_16(signx[j],X[j]);
if (signx[j] < 0)
iy[j] = -iy[j];
} while (++j<N);
encode_pulses(iy, N, K, enc);
/* Recompute the gain in one pass to reduce the encoder-decoder mismatch
due to the recursive computation used in quantisation. */
if (resynth)
{
normalise_residual(iy, X, N, K, EXTRACT16(SHR32(yy,2*yshift)));
if (spread)
exp_rotation(X, N, -1, spread, K);
}
RESTORE_STACK;
}
/** Decode pulse vector and combine the result with the pitch vector to produce
the final normalised signal in the current band. */
void alg_unquant(celt_norm *X, int N, int K, int spread, ec_dec *dec)
{
int i;
celt_word32 Ryy;
VARDECL(int, iy);
SAVE_STACK;
K = get_pulses(K);
ALLOC(iy, N, int);
decode_pulses(iy, N, K, dec);
Ryy = 0;
i=0;
do {
Ryy = MAC16_16(Ryy, iy[i], iy[i]);
} while (++i < N);
normalise_residual(iy, X, N, K, Ryy);
if (spread)
exp_rotation(X, N, -1, spread, K);
RESTORE_STACK;
}
celt_word16 renormalise_vector(celt_norm *X, celt_word16 value, int N, int stride)
{
int i;
celt_word32 E = EPSILON;
celt_word16 g;
celt_word32 t;
celt_norm *xptr = X;
for (i=0;i<N;i++)
{
E = MAC16_16(E, *xptr, *xptr);
xptr += stride;
}
#ifdef FIXED_POINT
int k = celt_ilog2(E)>>1;
#endif
t = VSHR32(E, (k-7)<<1);
g = MULT16_16_Q15(value, celt_rsqrt_norm(t));
xptr = X;
for (i=0;i<N;i++)
{
*xptr = EXTRACT16(PSHR32(MULT16_16(g, *xptr), k+1));
xptr += stride;
}
return celt_sqrt(E);
}
static void fold(const CELTMode *m, int start, int N, const celt_norm * restrict Y, celt_norm * restrict P, int N0, int B, int M)
{
int j;
int id = N0 % B;
while (id < M*m->eBands[start])
id += B;
/* Here, we assume that id will never be greater than N0, i.e. that
no band is wider than N0. In the unlikely case it happens, we set
everything to zero */
/*{
int offset = (N0*C - (id+C*N))/2;
if (offset > C*N0/16)
offset = C*N0/16;
offset -= offset % (C*B);
if (offset < 0)
offset = 0;
//printf ("%d\n", offset);
id += offset;
}*/
if (id+N>N0)
for (j=0;j<N;j++)
P[j] = 0;
else
for (j=0;j<N;j++)
P[j] = Y[id++];
}
void intra_fold(const CELTMode *m, int start, int N, const celt_norm * restrict Y, celt_norm * restrict P, int N0, int B, int M)
{
fold(m, start, N, Y, P, N0, B, M);
renormalise_vector(P, Q15ONE, N, 1);
}