1
0
mirror of https://github.com/f4exb/sdrangel.git synced 2024-11-26 17:58:43 -05:00
sdrangel/sdrbase/dsp/decimatorsif.h

1250 lines
40 KiB
C++

///////////////////////////////////////////////////////////////////////////////////
// Copyright (C) 2018 Edouard Griffiths, F4EXB //
// //
// This program is free software; you can redistribute it and/or modify //
// it under the terms of the GNU General Public License as published by //
// the Free Software Foundation as version 3 of the License, or //
// //
// This program 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 General Public License V3 for more details. //
// //
// You should have received a copy of the GNU General Public License //
// along with this program. If not, see <http://www.gnu.org/licenses/>. //
///////////////////////////////////////////////////////////////////////////////////
#ifndef SDRBASE_DSP_DECIMATORSIF_H_
#define SDRBASE_DSP_DECIMATORSIF_H_
#include "dsp/dsptypes.h"
#include "dsp/inthalfbandfiltereof.h"
#define DECIMATORS_IF_FILTER_ORDER 64
template<uint InputBits>
struct decimation_scale
{
static const float scaleIn;
};
template<uint InputBits>
const float decimation_scale<InputBits>::scaleIn = 1.0;
template<>
struct decimation_scale<8>
{
static const float scaleIn;
};
template<>
struct decimation_scale<12>
{
static const float scaleIn;
};
template<>
struct decimation_scale<16>
{
static const float scaleIn;
};
template<typename T, uint InputBits>
class DecimatorsIF {
public:
// interleaved I/Q input buffer
void decimate1(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate2_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate2_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate2_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate4_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate4_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate4_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate8_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate8_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate8_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate16_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate16_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate16_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate32_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate32_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate32_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate64_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate64_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
void decimate64_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ);
IntHalfbandFilterEOF<DECIMATORS_IF_FILTER_ORDER> m_decimator2; // 1st stages
IntHalfbandFilterEOF<DECIMATORS_IF_FILTER_ORDER> m_decimator4; // 2nd stages
IntHalfbandFilterEOF<DECIMATORS_IF_FILTER_ORDER> m_decimator8; // 3rd stages
IntHalfbandFilterEOF<DECIMATORS_IF_FILTER_ORDER> m_decimator16; // 4th stages
IntHalfbandFilterEOF<DECIMATORS_IF_FILTER_ORDER> m_decimator32; // 5th stages
IntHalfbandFilterEOF<DECIMATORS_IF_FILTER_ORDER> m_decimator64; // 6th stages
};
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate1(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal, yimag;
for (int pos = 0; pos < nbIAndQ - 1; pos += 2)
{
xreal = buf[pos+0] * decimation_scale<InputBits>::scaleIn;
yimag = buf[pos+1] * decimation_scale<InputBits>::scaleIn;
(**it).setReal(xreal);
(**it).setImag(yimag);
++(*it); // Valgrind optim (comment not repeated)
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate2_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal, yimag;
for (int pos = 0; pos < nbIAndQ - 7; pos += 8)
{
xreal = (buf[pos+0] - buf[pos+3]) * decimation_scale<InputBits>::scaleIn;
yimag = (buf[pos+1] + buf[pos+2]) * decimation_scale<InputBits>::scaleIn;
(**it).setReal(xreal);
(**it).setImag(yimag);
++(*it);
xreal = (buf[pos+7] - buf[pos+4]) * decimation_scale<InputBits>::scaleIn;
yimag = (- buf[pos+5] - buf[pos+6]) * decimation_scale<InputBits>::scaleIn;
(**it).setReal(xreal);
(**it).setImag(yimag);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate2_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal, yimag;
for (int pos = 0; pos < nbIAndQ - 7; pos += 8)
{
xreal = (buf[pos+1] - buf[pos+2]) * decimation_scale<InputBits>::scaleIn;
yimag = (- buf[pos+0] - buf[pos+3]) * decimation_scale<InputBits>::scaleIn;
(**it).setReal(xreal);
(**it).setImag(yimag);
++(*it);
xreal = (buf[pos+6] - buf[pos+5]) * decimation_scale<InputBits>::scaleIn;
yimag = (buf[pos+4] + buf[pos+7]) * decimation_scale<InputBits>::scaleIn;
(**it).setReal(xreal);
(**it).setImag(yimag);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate2_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float intbuf[2];
for (int pos = 0; pos < nbIAndQ - 3; pos += 4)
{
intbuf[0] = buf[pos+2];
intbuf[1] = buf[pos+3];
m_decimator2.myDecimate(
buf[pos+0],
buf[pos+1],
&intbuf[0],
&intbuf[1]);
(**it).setReal(intbuf[0] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(intbuf[1] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate4_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal, yimag;
for (int pos = 0; pos < nbIAndQ - 7; pos += 8)
{
xreal = (buf[pos+0] - buf[pos+3] + buf[pos+7] - buf[pos+4]) * decimation_scale<InputBits>::scaleIn;
yimag = (buf[pos+1] - buf[pos+5] + buf[pos+2] - buf[pos+6]) * decimation_scale<InputBits>::scaleIn;
(**it).setReal(xreal);
(**it).setImag(yimag);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate4_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal, yimag;
for (int pos = 0; pos < nbIAndQ - 7; pos += 8)
{
xreal = (buf[pos+1] - buf[pos+2] - buf[pos+5] + buf[pos+6]) * decimation_scale<InputBits>::scaleIn;
yimag = (- buf[pos+0] - buf[pos+3] + buf[pos+4] + buf[pos+7]) * decimation_scale<InputBits>::scaleIn;
(**it).setReal(xreal);
(**it).setImag(yimag);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate4_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float intbuf[4];
for (int pos = 0; pos < nbIAndQ - 7; pos += 8)
{
intbuf[0] = buf[pos+2];
intbuf[1] = buf[pos+3];
intbuf[2] = buf[pos+6];
intbuf[3] = buf[pos+7];
m_decimator2.myDecimate(
buf[pos+0],
buf[pos+1],
&intbuf[0],
&intbuf[1]);
m_decimator2.myDecimate(
buf[pos+4],
buf[pos+5],
&intbuf[2],
&intbuf[3]);
m_decimator4.myDecimate(
intbuf[0],
intbuf[1],
&intbuf[2],
&intbuf[3]);
(**it).setReal(intbuf[2] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(intbuf[3] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate8_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal[2], yimag[2];
for (int pos = 0; pos < nbIAndQ - 15; pos += 8)
{
xreal[0] = (buf[pos+0] - buf[pos+3] + buf[pos+7] - buf[pos+4]);
yimag[0] = (buf[pos+1] - buf[pos+5] + buf[pos+2] - buf[pos+6]);
pos += 8;
xreal[1] = (buf[pos+0] - buf[pos+3] + buf[pos+7] - buf[pos+4]);
yimag[1] = (buf[pos+1] - buf[pos+5] + buf[pos+2] - buf[pos+6]);
m_decimator2.myDecimate(xreal[0], yimag[0], &xreal[1], &yimag[1]);
(**it).setReal(xreal[1] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(yimag[1] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate8_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal[2], yimag[2];
for (int pos = 0; pos < nbIAndQ - 15; pos += 8)
{
xreal[0] = (buf[pos+1] - buf[pos+2] - buf[pos+5] + buf[pos+6]);
yimag[0] = (- buf[pos+0] - buf[pos+3] + buf[pos+4] + buf[pos+7]);
pos += 8;
xreal[1] = (buf[pos+1] - buf[pos+2] - buf[pos+5] + buf[pos+6]);
yimag[1] = (- buf[pos+0] - buf[pos+3] + buf[pos+4] + buf[pos+7]);
m_decimator2.myDecimate(xreal[0], yimag[0], &xreal[1], &yimag[1]);
(**it).setReal(xreal[1] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(yimag[1] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate8_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float intbuf[8];
for (int pos = 0; pos < nbIAndQ - 15; pos += 16)
{
intbuf[0] = buf[pos+2];
intbuf[1] = buf[pos+3];
intbuf[2] = buf[pos+6];
intbuf[3] = buf[pos+7];
intbuf[4] = buf[pos+10];
intbuf[5] = buf[pos+11];
intbuf[6] = buf[pos+14];
intbuf[7] = buf[pos+15];
m_decimator2.myDecimate(
buf[pos+0],
buf[pos+1],
&intbuf[0],
&intbuf[1]);
m_decimator2.myDecimate(
buf[pos+4],
buf[pos+5],
&intbuf[2],
&intbuf[3]);
m_decimator2.myDecimate(
buf[pos+8],
buf[pos+9],
&intbuf[4],
&intbuf[5]);
m_decimator2.myDecimate(
buf[pos+12],
buf[pos+13],
&intbuf[6],
&intbuf[7]);
m_decimator4.myDecimate(
intbuf[0],
intbuf[1],
&intbuf[2],
&intbuf[3]);
m_decimator4.myDecimate(
intbuf[4],
intbuf[5],
&intbuf[6],
&intbuf[7]);
m_decimator8.myDecimate(
intbuf[2],
intbuf[3],
&intbuf[6],
&intbuf[7]);
(**it).setReal(intbuf[6] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(intbuf[7] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate16_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal[4], yimag[4];
for (int pos = 0; pos < nbIAndQ - 31; )
{
for (int i = 0; i < 4; i++)
{
xreal[i] = (buf[pos+0] - buf[pos+3] + buf[pos+7] - buf[pos+4]);
yimag[i] = (buf[pos+1] - buf[pos+5] + buf[pos+2] - buf[pos+6]);
pos += 8;
}
m_decimator2.myDecimate(xreal[0], yimag[0], &xreal[1], &yimag[1]);
m_decimator2.myDecimate(xreal[2], yimag[2], &xreal[3], &yimag[3]);
m_decimator4.myDecimate(xreal[1], yimag[1], &xreal[3], &yimag[3]);
(**it).setReal(xreal[3] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(yimag[3] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate16_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal[4], yimag[4];
for (int pos = 0; pos < nbIAndQ - 31; )
{
for (int i = 0; i < 4; i++)
{
xreal[i] = (buf[pos+1] - buf[pos+2] - buf[pos+5] + buf[pos+6]);
yimag[i] = (buf[pos+4] + buf[pos+7] - buf[pos+0] - buf[pos+3]);
pos += 8;
}
m_decimator2.myDecimate(xreal[0], yimag[0], &xreal[1], &yimag[1]);
m_decimator2.myDecimate(xreal[2], yimag[2], &xreal[3], &yimag[3]);
m_decimator4.myDecimate(xreal[1], yimag[1], &xreal[3], &yimag[3]);
(**it).setReal(xreal[3] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(yimag[3] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate16_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float intbuf[16];
for (int pos = 0; pos < nbIAndQ - 31; pos += 32)
{
intbuf[0] = buf[pos+2];
intbuf[1] = buf[pos+3];
intbuf[2] = buf[pos+6];
intbuf[3] = buf[pos+7];
intbuf[4] = buf[pos+10];
intbuf[5] = buf[pos+11];
intbuf[6] = buf[pos+14];
intbuf[7] = buf[pos+15];
intbuf[8] = buf[pos+18];
intbuf[9] = buf[pos+19];
intbuf[10] = buf[pos+22];
intbuf[11] = buf[pos+23];
intbuf[12] = buf[pos+26];
intbuf[13] = buf[pos+27];
intbuf[14] = buf[pos+30];
intbuf[15] = buf[pos+31];
m_decimator2.myDecimate(
buf[pos+0],
buf[pos+1],
&intbuf[0],
&intbuf[1]);
m_decimator2.myDecimate(
buf[pos+4],
buf[pos+5],
&intbuf[2],
&intbuf[3]);
m_decimator2.myDecimate(
buf[pos+8],
buf[pos+9],
&intbuf[4],
&intbuf[5]);
m_decimator2.myDecimate(
buf[pos+12],
buf[pos+13],
&intbuf[6],
&intbuf[7]);
m_decimator2.myDecimate(
buf[pos+16],
buf[pos+17],
&intbuf[8],
&intbuf[9]);
m_decimator2.myDecimate(
buf[pos+20],
buf[pos+21],
&intbuf[10],
&intbuf[11]);
m_decimator2.myDecimate(
buf[pos+24],
buf[pos+25],
&intbuf[12],
&intbuf[13]);
m_decimator2.myDecimate(
buf[pos+28],
buf[pos+29],
&intbuf[14],
&intbuf[15]);
m_decimator4.myDecimate(
intbuf[0],
intbuf[1],
&intbuf[2],
&intbuf[3]);
m_decimator4.myDecimate(
intbuf[4],
intbuf[5],
&intbuf[6],
&intbuf[7]);
m_decimator4.myDecimate(
intbuf[8],
intbuf[9],
&intbuf[10],
&intbuf[11]);
m_decimator4.myDecimate(
intbuf[12],
intbuf[13],
&intbuf[14],
&intbuf[15]);
m_decimator8.myDecimate(
intbuf[2],
intbuf[3],
&intbuf[6],
&intbuf[7]);
m_decimator8.myDecimate(
intbuf[10],
intbuf[11],
&intbuf[14],
&intbuf[15]);
m_decimator16.myDecimate(
intbuf[6],
intbuf[7],
&intbuf[14],
&intbuf[15]);
(**it).setReal(intbuf[14] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(intbuf[15] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate32_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal[8], yimag[8];
for (int pos = 0; pos < nbIAndQ - 63; )
{
for (int i = 0; i < 8; i++)
{
xreal[i] = (buf[pos+0] - buf[pos+3] + buf[pos+7] - buf[pos+4]);
yimag[i] = (buf[pos+1] - buf[pos+5] + buf[pos+2] - buf[pos+6]);
pos += 8;
}
m_decimator2.myDecimate(xreal[0], yimag[0], &xreal[1], &yimag[1]);
m_decimator2.myDecimate(xreal[2], yimag[2], &xreal[3], &yimag[3]);
m_decimator2.myDecimate(xreal[4], yimag[4], &xreal[5], &yimag[5]);
m_decimator2.myDecimate(xreal[6], yimag[6], &xreal[7], &yimag[7]);
m_decimator4.myDecimate(xreal[1], yimag[1], &xreal[3], &yimag[3]);
m_decimator4.myDecimate(xreal[5], yimag[5], &xreal[7], &yimag[7]);
m_decimator8.myDecimate(xreal[3], yimag[3], &xreal[7], &yimag[7]);
(**it).setReal(xreal[7] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(yimag[7] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate32_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal[8], yimag[8];
for (int pos = 0; pos < nbIAndQ - 63; )
{
for (int i = 0; i < 8; i++)
{
xreal[i] = (buf[pos+1] - buf[pos+2] - buf[pos+5] + buf[pos+6]);
yimag[i] = (buf[pos+4] + buf[pos+7] - buf[pos+0] - buf[pos+3]);
pos += 8;
}
m_decimator2.myDecimate(xreal[0], yimag[0], &xreal[1], &yimag[1]);
m_decimator2.myDecimate(xreal[2], yimag[2], &xreal[3], &yimag[3]);
m_decimator2.myDecimate(xreal[4], yimag[4], &xreal[5], &yimag[5]);
m_decimator2.myDecimate(xreal[6], yimag[6], &xreal[7], &yimag[7]);
m_decimator4.myDecimate(xreal[1], yimag[1], &xreal[3], &yimag[3]);
m_decimator4.myDecimate(xreal[5], yimag[5], &xreal[7], &yimag[7]);
m_decimator8.myDecimate(xreal[3], yimag[3], &xreal[7], &yimag[7]);
(**it).setReal(xreal[7] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(yimag[7] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate32_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float intbuf[32];
for (int pos = 0; pos < nbIAndQ - 63; pos += 64)
{
intbuf[0] = buf[pos+2];
intbuf[1] = buf[pos+3];
intbuf[2] = buf[pos+6];
intbuf[3] = buf[pos+7];
intbuf[4] = buf[pos+10];
intbuf[5] = buf[pos+11];
intbuf[6] = buf[pos+14];
intbuf[7] = buf[pos+15];
intbuf[8] = buf[pos+18];
intbuf[9] = buf[pos+19];
intbuf[10] = buf[pos+22];
intbuf[11] = buf[pos+23];
intbuf[12] = buf[pos+26];
intbuf[13] = buf[pos+27];
intbuf[14] = buf[pos+30];
intbuf[15] = buf[pos+31];
intbuf[16] = buf[pos+34];
intbuf[17] = buf[pos+35];
intbuf[18] = buf[pos+38];
intbuf[19] = buf[pos+39];
intbuf[20] = buf[pos+42];
intbuf[21] = buf[pos+43];
intbuf[22] = buf[pos+46];
intbuf[23] = buf[pos+47];
intbuf[24] = buf[pos+50];
intbuf[25] = buf[pos+51];
intbuf[26] = buf[pos+54];
intbuf[27] = buf[pos+55];
intbuf[28] = buf[pos+58];
intbuf[29] = buf[pos+59];
intbuf[30] = buf[pos+62];
intbuf[31] = buf[pos+63];
m_decimator2.myDecimate(
buf[pos+0],
buf[pos+1],
&intbuf[0],
&intbuf[1]);
m_decimator2.myDecimate(
buf[pos+4],
buf[pos+5],
&intbuf[2],
&intbuf[3]);
m_decimator2.myDecimate(
buf[pos+8],
buf[pos+9],
&intbuf[4],
&intbuf[5]);
m_decimator2.myDecimate(
buf[pos+12],
buf[pos+13],
&intbuf[6],
&intbuf[7]);
m_decimator2.myDecimate(
buf[pos+16],
buf[pos+17],
&intbuf[8],
&intbuf[9]);
m_decimator2.myDecimate(
buf[pos+20],
buf[pos+21],
&intbuf[10],
&intbuf[11]);
m_decimator2.myDecimate(
buf[pos+24],
buf[pos+25],
&intbuf[12],
&intbuf[13]);
m_decimator2.myDecimate(
buf[pos+28],
buf[pos+29],
&intbuf[14],
&intbuf[15]);
m_decimator2.myDecimate(
buf[pos+32],
buf[pos+33],
&intbuf[16],
&intbuf[17]);
m_decimator2.myDecimate(
buf[pos+36],
buf[pos+37],
&intbuf[18],
&intbuf[19]);
m_decimator2.myDecimate(
buf[pos+40],
buf[pos+41],
&intbuf[20],
&intbuf[21]);
m_decimator2.myDecimate(
buf[pos+44],
buf[pos+45],
&intbuf[22],
&intbuf[23]);
m_decimator2.myDecimate(
buf[pos+48],
buf[pos+49],
&intbuf[24],
&intbuf[25]);
m_decimator2.myDecimate(
buf[pos+52],
buf[pos+53],
&intbuf[26],
&intbuf[27]);
m_decimator2.myDecimate(
buf[pos+56],
buf[pos+57],
&intbuf[28],
&intbuf[29]);
m_decimator2.myDecimate(
buf[pos+60],
buf[pos+61],
&intbuf[30],
&intbuf[31]);
m_decimator4.myDecimate(
intbuf[0],
intbuf[1],
&intbuf[2],
&intbuf[3]);
m_decimator4.myDecimate(
intbuf[4],
intbuf[5],
&intbuf[6],
&intbuf[7]);
m_decimator4.myDecimate(
intbuf[8],
intbuf[9],
&intbuf[10],
&intbuf[11]);
m_decimator4.myDecimate(
intbuf[12],
intbuf[13],
&intbuf[14],
&intbuf[15]);
m_decimator4.myDecimate(
intbuf[16],
intbuf[17],
&intbuf[18],
&intbuf[19]);
m_decimator4.myDecimate(
intbuf[20],
intbuf[21],
&intbuf[22],
&intbuf[23]);
m_decimator4.myDecimate(
intbuf[24],
intbuf[25],
&intbuf[26],
&intbuf[27]);
m_decimator4.myDecimate(
intbuf[28],
intbuf[29],
&intbuf[30],
&intbuf[31]);
m_decimator8.myDecimate(
intbuf[2],
intbuf[3],
&intbuf[6],
&intbuf[7]);
m_decimator8.myDecimate(
intbuf[10],
intbuf[11],
&intbuf[14],
&intbuf[15]);
m_decimator8.myDecimate(
intbuf[18],
intbuf[19],
&intbuf[22],
&intbuf[23]);
m_decimator8.myDecimate(
intbuf[26],
intbuf[27],
&intbuf[30],
&intbuf[31]);
m_decimator16.myDecimate(
intbuf[6],
intbuf[7],
&intbuf[14],
&intbuf[15]);
m_decimator16.myDecimate(
intbuf[22],
intbuf[23],
&intbuf[30],
&intbuf[31]);
m_decimator32.myDecimate(
intbuf[14],
intbuf[15],
&intbuf[30],
&intbuf[31]);
(**it).setReal(intbuf[30] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(intbuf[31] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate64_inf(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal[16], yimag[16];
for (int pos = 0; pos < nbIAndQ - 127; )
{
for (int i = 0; i < 16; i++)
{
xreal[i] = (buf[pos+0] - buf[pos+3] + buf[pos+7] - buf[pos+4]);
yimag[i] = (buf[pos+1] - buf[pos+5] + buf[pos+2] - buf[pos+6]);
pos += 8;
}
m_decimator2.myDecimate(xreal[0], yimag[0], &xreal[1], &yimag[1]);
m_decimator2.myDecimate(xreal[2], yimag[2], &xreal[3], &yimag[3]);
m_decimator2.myDecimate(xreal[4], yimag[4], &xreal[5], &yimag[5]);
m_decimator2.myDecimate(xreal[6], yimag[6], &xreal[7], &yimag[7]);
m_decimator2.myDecimate(xreal[8], yimag[8], &xreal[9], &yimag[9]);
m_decimator2.myDecimate(xreal[10], yimag[10], &xreal[11], &yimag[11]);
m_decimator2.myDecimate(xreal[12], yimag[12], &xreal[13], &yimag[13]);
m_decimator2.myDecimate(xreal[14], yimag[14], &xreal[15], &yimag[15]);
m_decimator4.myDecimate(xreal[1], yimag[1], &xreal[3], &yimag[3]);
m_decimator4.myDecimate(xreal[5], yimag[5], &xreal[7], &yimag[7]);
m_decimator4.myDecimate(xreal[9], yimag[9], &xreal[11], &yimag[11]);
m_decimator4.myDecimate(xreal[13], yimag[13], &xreal[15], &yimag[15]);
m_decimator8.myDecimate(xreal[3], yimag[3], &xreal[7], &yimag[7]);
m_decimator8.myDecimate(xreal[11], yimag[11], &xreal[15], &yimag[15]);
m_decimator16.myDecimate(xreal[7], yimag[7], &xreal[15], &yimag[15]);
(**it).setReal(xreal[15] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(yimag[15] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate64_sup(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float xreal[16], yimag[16];
for (int pos = 0; pos < nbIAndQ - 127; )
{
for (int i = 0; i < 16; i++)
{
xreal[i] = (buf[pos+1] - buf[pos+2] - buf[pos+5] + buf[pos+6]);
yimag[i] = (buf[pos+4] + buf[pos+7] - buf[pos+0] - buf[pos+3]);
pos += 8;
}
m_decimator2.myDecimate(xreal[0], yimag[0], &xreal[1], &yimag[1]);
m_decimator2.myDecimate(xreal[2], yimag[2], &xreal[3], &yimag[3]);
m_decimator2.myDecimate(xreal[4], yimag[4], &xreal[5], &yimag[5]);
m_decimator2.myDecimate(xreal[6], yimag[6], &xreal[7], &yimag[7]);
m_decimator2.myDecimate(xreal[8], yimag[8], &xreal[9], &yimag[9]);
m_decimator2.myDecimate(xreal[10], yimag[10], &xreal[11], &yimag[11]);
m_decimator2.myDecimate(xreal[12], yimag[12], &xreal[13], &yimag[13]);
m_decimator2.myDecimate(xreal[14], yimag[14], &xreal[15], &yimag[15]);
m_decimator4.myDecimate(xreal[1], yimag[1], &xreal[3], &yimag[3]);
m_decimator4.myDecimate(xreal[5], yimag[5], &xreal[7], &yimag[7]);
m_decimator4.myDecimate(xreal[9], yimag[9], &xreal[11], &yimag[11]);
m_decimator4.myDecimate(xreal[13], yimag[13], &xreal[15], &yimag[15]);
m_decimator8.myDecimate(xreal[3], yimag[3], &xreal[7], &yimag[7]);
m_decimator8.myDecimate(xreal[11], yimag[11], &xreal[15], &yimag[15]);
m_decimator16.myDecimate(xreal[7], yimag[7], &xreal[15], &yimag[15]);
(**it).setReal(xreal[15] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(yimag[15] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
template<typename T, uint InputBits>
void DecimatorsIF<T, InputBits>::decimate64_cen(FSampleVector::iterator* it, const T* buf, qint32 nbIAndQ)
{
float intbuf[64];
for (int pos = 0; pos < nbIAndQ - 127; pos += 128)
{
intbuf[0] = buf[pos+2];
intbuf[1] = buf[pos+3];
intbuf[2] = buf[pos+6];
intbuf[3] = buf[pos+7];
intbuf[4] = buf[pos+10];
intbuf[5] = buf[pos+11];
intbuf[6] = buf[pos+14];
intbuf[7] = buf[pos+15];
intbuf[8] = buf[pos+18];
intbuf[9] = buf[pos+19];
intbuf[10] = buf[pos+22];
intbuf[11] = buf[pos+23];
intbuf[12] = buf[pos+26];
intbuf[13] = buf[pos+27];
intbuf[14] = buf[pos+30];
intbuf[15] = buf[pos+31];
intbuf[16] = buf[pos+34];
intbuf[17] = buf[pos+35];
intbuf[18] = buf[pos+38];
intbuf[19] = buf[pos+39];
intbuf[20] = buf[pos+42];
intbuf[21] = buf[pos+43];
intbuf[22] = buf[pos+46];
intbuf[23] = buf[pos+47];
intbuf[24] = buf[pos+50];
intbuf[25] = buf[pos+51];
intbuf[26] = buf[pos+54];
intbuf[27] = buf[pos+55];
intbuf[28] = buf[pos+58];
intbuf[29] = buf[pos+59];
intbuf[30] = buf[pos+62];
intbuf[31] = buf[pos+63];
intbuf[32] = buf[pos+66];
intbuf[33] = buf[pos+67];
intbuf[34] = buf[pos+70];
intbuf[35] = buf[pos+71];
intbuf[36] = buf[pos+74];
intbuf[37] = buf[pos+75];
intbuf[38] = buf[pos+78];
intbuf[39] = buf[pos+79];
intbuf[40] = buf[pos+82];
intbuf[41] = buf[pos+83];
intbuf[42] = buf[pos+86];
intbuf[43] = buf[pos+87];
intbuf[44] = buf[pos+90];
intbuf[45] = buf[pos+91];
intbuf[46] = buf[pos+94];
intbuf[47] = buf[pos+95];
intbuf[48] = buf[pos+98];
intbuf[49] = buf[pos+99];
intbuf[50] = buf[pos+102];
intbuf[51] = buf[pos+103];
intbuf[52] = buf[pos+106];
intbuf[53] = buf[pos+107];
intbuf[54] = buf[pos+110];
intbuf[55] = buf[pos+111];
intbuf[56] = buf[pos+114];
intbuf[57] = buf[pos+115];
intbuf[58] = buf[pos+118];
intbuf[59] = buf[pos+119];
intbuf[60] = buf[pos+122];
intbuf[61] = buf[pos+123];
intbuf[62] = buf[pos+126];
intbuf[63] = buf[pos+127];
m_decimator2.myDecimate(
buf[pos+0],
buf[pos+1],
&intbuf[0],
&intbuf[1]);
m_decimator2.myDecimate(
buf[pos+4],
buf[pos+5],
&intbuf[2],
&intbuf[3]);
m_decimator2.myDecimate(
buf[pos+8],
buf[pos+9],
&intbuf[4],
&intbuf[5]);
m_decimator2.myDecimate(
buf[pos+12],
buf[pos+13],
&intbuf[6],
&intbuf[7]);
m_decimator2.myDecimate(
buf[pos+16],
buf[pos+17],
&intbuf[8],
&intbuf[9]);
m_decimator2.myDecimate(
buf[pos+20],
buf[pos+21],
&intbuf[10],
&intbuf[11]);
m_decimator2.myDecimate(
buf[pos+24],
buf[pos+25],
&intbuf[12],
&intbuf[13]);
m_decimator2.myDecimate(
buf[pos+28],
buf[pos+29],
&intbuf[14],
&intbuf[15]);
m_decimator2.myDecimate(
buf[pos+32],
buf[pos+33],
&intbuf[16],
&intbuf[17]);
m_decimator2.myDecimate(
buf[pos+36],
buf[pos+37],
&intbuf[18],
&intbuf[19]);
m_decimator2.myDecimate(
buf[pos+40],
buf[pos+41],
&intbuf[20],
&intbuf[21]);
m_decimator2.myDecimate(
buf[pos+44],
buf[pos+45],
&intbuf[22],
&intbuf[23]);
m_decimator2.myDecimate(
buf[pos+48],
buf[pos+49],
&intbuf[24],
&intbuf[25]);
m_decimator2.myDecimate(
buf[pos+52],
buf[pos+53],
&intbuf[26],
&intbuf[27]);
m_decimator2.myDecimate(
buf[pos+56],
buf[pos+57],
&intbuf[28],
&intbuf[29]);
m_decimator2.myDecimate(
buf[pos+60],
buf[pos+61],
&intbuf[30],
&intbuf[31]);
m_decimator2.myDecimate(
buf[pos+64],
buf[pos+65],
&intbuf[32],
&intbuf[33]);
m_decimator2.myDecimate(
buf[pos+68],
buf[pos+69],
&intbuf[34],
&intbuf[35]);
m_decimator2.myDecimate(
buf[pos+72],
buf[pos+73],
&intbuf[36],
&intbuf[37]);
m_decimator2.myDecimate(
buf[pos+76],
buf[pos+77],
&intbuf[38],
&intbuf[39]);
m_decimator2.myDecimate(
buf[pos+80],
buf[pos+81],
&intbuf[40],
&intbuf[41]);
m_decimator2.myDecimate(
buf[pos+84],
buf[pos+85],
&intbuf[42],
&intbuf[43]);
m_decimator2.myDecimate(
buf[pos+88],
buf[pos+89],
&intbuf[44],
&intbuf[45]);
m_decimator2.myDecimate(
buf[pos+92],
buf[pos+93],
&intbuf[46],
&intbuf[47]);
m_decimator2.myDecimate(
buf[pos+96],
buf[pos+97],
&intbuf[48],
&intbuf[49]);
m_decimator2.myDecimate(
buf[pos+100],
buf[pos+101],
&intbuf[50],
&intbuf[51]);
m_decimator2.myDecimate(
buf[pos+104],
buf[pos+105],
&intbuf[52],
&intbuf[53]);
m_decimator2.myDecimate(
buf[pos+108],
buf[pos+109],
&intbuf[54],
&intbuf[55]);
m_decimator2.myDecimate(
buf[pos+112],
buf[pos+113],
&intbuf[56],
&intbuf[57]);
m_decimator2.myDecimate(
buf[pos+116],
buf[pos+117],
&intbuf[58],
&intbuf[59]);
m_decimator2.myDecimate(
buf[pos+120],
buf[pos+121],
&intbuf[60],
&intbuf[61]);
m_decimator2.myDecimate(
buf[pos+124],
buf[pos+125],
&intbuf[62],
&intbuf[63]);
m_decimator4.myDecimate(
intbuf[0],
intbuf[1],
&intbuf[2],
&intbuf[3]);
m_decimator4.myDecimate(
intbuf[4],
intbuf[5],
&intbuf[6],
&intbuf[7]);
m_decimator4.myDecimate(
intbuf[8],
intbuf[9],
&intbuf[10],
&intbuf[11]);
m_decimator4.myDecimate(
intbuf[12],
intbuf[13],
&intbuf[14],
&intbuf[15]);
m_decimator4.myDecimate(
intbuf[16],
intbuf[17],
&intbuf[18],
&intbuf[19]);
m_decimator4.myDecimate(
intbuf[20],
intbuf[21],
&intbuf[22],
&intbuf[23]);
m_decimator4.myDecimate(
intbuf[24],
intbuf[25],
&intbuf[26],
&intbuf[27]);
m_decimator4.myDecimate(
intbuf[28],
intbuf[29],
&intbuf[30],
&intbuf[31]);
m_decimator4.myDecimate(
intbuf[32],
intbuf[33],
&intbuf[34],
&intbuf[35]);
m_decimator4.myDecimate(
intbuf[36],
intbuf[37],
&intbuf[38],
&intbuf[39]);
m_decimator4.myDecimate(
intbuf[40],
intbuf[41],
&intbuf[42],
&intbuf[43]);
m_decimator4.myDecimate(
intbuf[44],
intbuf[45],
&intbuf[46],
&intbuf[47]);
m_decimator4.myDecimate(
intbuf[48],
intbuf[49],
&intbuf[50],
&intbuf[51]);
m_decimator4.myDecimate(
intbuf[52],
intbuf[53],
&intbuf[54],
&intbuf[55]);
m_decimator4.myDecimate(
intbuf[56],
intbuf[57],
&intbuf[58],
&intbuf[59]);
m_decimator4.myDecimate(
intbuf[60],
intbuf[61],
&intbuf[62],
&intbuf[63]);
m_decimator8.myDecimate(
intbuf[2],
intbuf[3],
&intbuf[6],
&intbuf[7]);
m_decimator8.myDecimate(
intbuf[10],
intbuf[11],
&intbuf[14],
&intbuf[15]);
m_decimator8.myDecimate(
intbuf[18],
intbuf[19],
&intbuf[22],
&intbuf[23]);
m_decimator8.myDecimate(
intbuf[26],
intbuf[27],
&intbuf[30],
&intbuf[31]);
m_decimator8.myDecimate(
intbuf[34],
intbuf[35],
&intbuf[38],
&intbuf[39]);
m_decimator8.myDecimate(
intbuf[42],
intbuf[43],
&intbuf[46],
&intbuf[47]);
m_decimator8.myDecimate(
intbuf[50],
intbuf[51],
&intbuf[54],
&intbuf[55]);
m_decimator8.myDecimate(
intbuf[58],
intbuf[59],
&intbuf[62],
&intbuf[63]);
m_decimator16.myDecimate(
intbuf[6],
intbuf[7],
&intbuf[14],
&intbuf[15]);
m_decimator16.myDecimate(
intbuf[22],
intbuf[23],
&intbuf[30],
&intbuf[31]);
m_decimator16.myDecimate(
intbuf[38],
intbuf[39],
&intbuf[46],
&intbuf[47]);
m_decimator16.myDecimate(
intbuf[54],
intbuf[55],
&intbuf[62],
&intbuf[63]);
m_decimator32.myDecimate(
intbuf[14],
intbuf[15],
&intbuf[30],
&intbuf[31]);
m_decimator32.myDecimate(
intbuf[46],
intbuf[47],
&intbuf[62],
&intbuf[63]);
m_decimator64.myDecimate(
intbuf[30],
intbuf[31],
&intbuf[62],
&intbuf[63]);
(**it).setReal(intbuf[62] * decimation_scale<InputBits>::scaleIn);
(**it).setImag(intbuf[63] * decimation_scale<InputBits>::scaleIn);
++(*it);
}
}
#endif /* SDRBASE_DSP_DECIMATORSIF_H_ */