ref: 18a380a7c209b482f63adeb19f5d398dcb1c10bd
dir: /celt/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.
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 COPYRIGHT OWNER
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 "bands.h"
#include "rate.h"
#include "pitch.h"
#ifndef OVERRIDE_vq_exp_rotation1
static void exp_rotation1(celt_norm *X, int len, int stride, opus_val16 c, opus_val16 s)
{
int i;
opus_val16 ms;
celt_norm *Xptr;
Xptr = X;
ms = NEG16(s);
for (i=0;i<len-stride;i++)
{
celt_norm x1, x2;
x1 = Xptr[0];
x2 = Xptr[stride];
Xptr[stride] = EXTRACT16(PSHR32(MAC16_16(MULT16_16(c, x2), s, x1), 15));
*Xptr++ = EXTRACT16(PSHR32(MAC16_16(MULT16_16(c, x1), ms, 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(PSHR32(MAC16_16(MULT16_16(c, x2), s, x1), 15));
*Xptr-- = EXTRACT16(PSHR32(MAC16_16(MULT16_16(c, x1), ms, x2), 15));
}
}
#endif /* OVERRIDE_vq_exp_rotation1 */
static void exp_rotation(celt_norm *X, int len, int dir, int stride, int K, int spread)
{
static const int SPREAD_FACTOR[3]={15,10,5};
int i;
opus_val16 c, s;
opus_val16 gain, theta;
int stride2=0;
int factor;
if (2*K>=len || spread==SPREAD_NONE)
return;
factor = SPREAD_FACTOR[spread-1];
gain = celt_div((opus_val32)MULT16_16(Q15_ONE,len),(opus_val32)(len+factor*K));
theta = HALF16(MULT16_16_Q15(gain,gain));
c = celt_cos_norm(EXTEND32(theta));
s = celt_cos_norm(EXTEND32(SUB16(Q15ONE,theta))); /* sin(theta) */
if (len>=8*stride)
{
stride2 = 1;
/* This is just a simple (equivalent) way of computing sqrt(len/stride) with rounding.
It's basically incrementing long as (stride2+0.5)^2 < len/stride. */
while ((stride2*stride2+stride2)*stride + (stride>>2) < len)
stride2++;
}
/*NOTE: As a minor optimization, we could be passing around log2(B), not B, for both this and for
extract_collapse_mask().*/
len = celt_udiv(len, stride);
for (i=0;i<stride;i++)
{
if (dir < 0)
{
if (stride2)
exp_rotation1(X+i*len, len, stride2, s, c);
exp_rotation1(X+i*len, len, 1, c, s);
} else {
exp_rotation1(X+i*len, len, 1, c, -s);
if (stride2)
exp_rotation1(X+i*len, len, stride2, s, -c);
}
}
}
/** 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 * OPUS_RESTRICT iy, celt_norm * OPUS_RESTRICT X,
int N, opus_val32 Ryy, opus_val16 gain)
{
int i;
#ifdef FIXED_POINT
int k;
#endif
opus_val32 t;
opus_val16 g;
#ifdef FIXED_POINT
k = celt_ilog2(Ryy)>>1;
#endif
t = VSHR32(Ryy, 2*(k-7));
g = MULT16_16_P15(celt_rsqrt_norm(t),gain);
i=0;
do
X[i] = EXTRACT16(PSHR32(MULT16_16(g, iy[i]), k+1));
while (++i < N);
}
static unsigned extract_collapse_mask(int *iy, int N, int B)
{
unsigned collapse_mask;
int N0;
int i;
if (B<=1)
return 1;
/*NOTE: As a minor optimization, we could be passing around log2(B), not B, for both this and for
exp_rotation().*/
N0 = celt_udiv(N, B);
collapse_mask = 0;
i=0; do {
int j;
unsigned tmp=0;
j=0; do {
tmp |= iy[i*N0+j];
} while (++j<N0);
collapse_mask |= (tmp!=0)<<i;
} while (++i<B);
return collapse_mask;
}
opus_val16 op_pvq_search_c(celt_norm *X, int *iy, int K, int N, int arch)
{
VARDECL(celt_norm, y);
VARDECL(int, signx);
int i, j;
int pulsesLeft;
opus_val32 sum;
opus_val32 xy;
opus_val16 yy;
SAVE_STACK;
(void)arch;
ALLOC(y, N, celt_norm);
ALLOC(signx, N, int);
/* Get rid of the sign */
sum = 0;
j=0; do {
signx[j] = X[j]<0;
/* OPT: Make sure the compiler doesn't use a branch on ABS16(). */
X[j] = ABS16(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))
{
opus_val16 rcp;
j=0; do {
sum += X[j];
} while (++j<N);
/* If X is too small, just replace it with a pulse at 0 */
#ifdef FIXED_POINT
if (sum <= K)
#else
/* Prevents infinities and NaNs from causing too many pulses
to be allocated. 64 is an approximation of infinity here. */
if (!(sum > EPSILON && sum < 64))
#endif
{
X[0] = QCONST16(1.f,14);
j=1; do
X[j]=0;
while (++j<N);
sum = QCONST16(1.f,14);
}
#ifdef FIXED_POINT
rcp = EXTRACT16(MULT16_32_Q16(K, celt_rcp(sum)));
#else
/* Using K+e with e < 1 guarantees we cannot get more than K pulses. */
rcp = EXTRACT16(MULT16_32_Q16(K+0.8, celt_rcp(sum)));
#endif
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] = (int)floor(rcp*X[j]);
#endif
y[j] = (celt_norm)iy[j];
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>=0, "Allocated too many pulses in the quick pass");
/* This should never happen, but just in case it does (e.g. on silence)
we fill the first bin with pulses. */
#ifdef FIXED_POINT_DEBUG
celt_assert2(pulsesLeft<=N+3, "Not enough pulses in the quick pass");
#endif
if (pulsesLeft > N+3)
{
opus_val16 tmp = (opus_val16)pulsesLeft;
yy = MAC16_16(yy, tmp, tmp);
yy = MAC16_16(yy, tmp, y[0]);
iy[0] += pulsesLeft;
pulsesLeft=0;
}
for (i=0;i<pulsesLeft;i++)
{
opus_val16 Rxy, Ryy;
int best_id;
opus_val32 best_num;
opus_val16 best_den;
#ifdef FIXED_POINT
int rshift;
#endif
#ifdef FIXED_POINT
rshift = 1+celt_ilog2(K-pulsesLeft+i+1);
#endif
best_id = 0;
/* The squared magnitude term gets added anyway, so we might as well
add it outside the loop */
yy = ADD16(yy, 1);
/* Calculations for position 0 are out of the loop, in part to reduce
mispredicted branches (since the if condition is usually false)
in the loop. */
/* Temporary sums of the new pulse(s) */
Rxy = EXTRACT16(SHR32(ADD32(xy, EXTEND32(X[0])),rshift));
/* We're multiplying y[j] by two so we don't have to do it here */
Ryy = ADD16(yy, y[0]);
/* 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);
best_den = Ryy;
best_num = Rxy;
j=1;
do {
/* Temporary sums of the new pulse(s) */
Rxy = EXTRACT16(SHR32(ADD32(xy, EXTEND32(X[j])),rshift));
/* We're multiplying y[j] by two so we don't have to do it here */
Ryy = ADD16(yy, y[j]);
/* 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: It's not clear whether a cmov is faster than a branch here
since the condition is more often false than true and using
a cmov introduces data dependencies across iterations. The optimal
choice may be architecture-dependent. */
if (opus_unlikely(MULT16_16(best_den, Rxy) > MULT16_16(Ryy, best_num)))
{
best_den = Ryy;
best_num = Rxy;
best_id = j;
}
} while (++j<N);
/* Updating the sums of the new pulse(s) */
xy = ADD32(xy, EXTEND32(X[best_id]));
/* We're multiplying y[j] by two so we don't have to do it here */
yy = ADD16(yy, y[best_id]);
/* 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[best_id] += 2;
iy[best_id]++;
}
/* Put the original sign back */
j=0;
do {
/*iy[j] = signx[j] ? -iy[j] : iy[j];*/
/* OPT: The is more likely to be compiled without a branch than the code above
but has the same performance otherwise. */
iy[j] = (iy[j]^-signx[j]) + signx[j];
} while (++j<N);
RESTORE_STACK;
return yy;
}
unsigned alg_quant(celt_norm *X, int N, int K, int spread, int B, ec_enc *enc,
opus_val16 gain, int resynth, int arch)
{
VARDECL(int, iy);
opus_val16 yy;
unsigned collapse_mask;
SAVE_STACK;
celt_assert2(K>0, "alg_quant() needs at least one pulse");
celt_assert2(N>1, "alg_quant() needs at least two dimensions");
/* Covers vectorization by up to 4. */
ALLOC(iy, N+3, int);
exp_rotation(X, N, 1, B, K, spread);
yy = op_pvq_search(X, iy, K, N, arch);
encode_pulses(iy, N, K, enc);
if (resynth)
{
normalise_residual(iy, X, N, yy, gain);
exp_rotation(X, N, -1, B, K, spread);
}
collapse_mask = extract_collapse_mask(iy, N, B);
RESTORE_STACK;
return collapse_mask;
}
/** Decode pulse vector and combine the result with the pitch vector to produce
the final normalised signal in the current band. */
unsigned alg_unquant(celt_norm *X, int N, int K, int spread, int B,
ec_dec *dec, opus_val16 gain)
{
opus_val32 Ryy;
unsigned collapse_mask;
VARDECL(int, iy);
SAVE_STACK;
celt_assert2(K>0, "alg_unquant() needs at least one pulse");
celt_assert2(N>1, "alg_unquant() needs at least two dimensions");
ALLOC(iy, N, int);
Ryy = decode_pulses(iy, N, K, dec);
normalise_residual(iy, X, N, Ryy, gain);
exp_rotation(X, N, -1, B, K, spread);
collapse_mask = extract_collapse_mask(iy, N, B);
RESTORE_STACK;
return collapse_mask;
}
#ifndef OVERRIDE_renormalise_vector
void renormalise_vector(celt_norm *X, int N, opus_val16 gain, int arch)
{
int i;
#ifdef FIXED_POINT
int k;
#endif
opus_val32 E;
opus_val16 g;
opus_val32 t;
celt_norm *xptr;
E = EPSILON + celt_inner_prod(X, X, N, arch);
#ifdef FIXED_POINT
k = celt_ilog2(E)>>1;
#endif
t = VSHR32(E, 2*(k-7));
g = MULT16_16_P15(celt_rsqrt_norm(t),gain);
xptr = X;
for (i=0;i<N;i++)
{
*xptr = EXTRACT16(PSHR32(MULT16_16(g, *xptr), k+1));
xptr++;
}
/*return celt_sqrt(E);*/
}
#endif /* OVERRIDE_renormalise_vector */
int stereo_itheta(const celt_norm *X, const celt_norm *Y, int stereo, int N, int arch)
{
int i;
int itheta;
opus_val16 mid, side;
opus_val32 Emid, Eside;
Emid = Eside = EPSILON;
if (stereo)
{
for (i=0;i<N;i++)
{
celt_norm m, s;
m = ADD16(SHR16(X[i],1),SHR16(Y[i],1));
s = SUB16(SHR16(X[i],1),SHR16(Y[i],1));
Emid = MAC16_16(Emid, m, m);
Eside = MAC16_16(Eside, s, s);
}
} else {
Emid += celt_inner_prod(X, X, N, arch);
Eside += celt_inner_prod(Y, Y, N, arch);
}
mid = celt_sqrt(Emid);
side = celt_sqrt(Eside);
#ifdef FIXED_POINT
/* 0.63662 = 2/pi */
itheta = MULT16_16_Q15(QCONST16(0.63662f,15),celt_atan2p(side, mid));
#else
itheta = (int)floor(.5f+16384*0.63662f*fast_atan2f(side,mid));
#endif
return itheta;
}