VE7UWU/sdrangel
1
0
Fork 0
mirror of https://github.com/f4exb/sdrangel.git synced 2024-11-15 12:51:49 -05:00
Code Issues Projects Releases Wiki Activity

Fix and refactor fir filter

Browse Source
This commit is contained in:
f4exb 2020-10-31 21:30:45 +01:00
parent b9adbfb1d8
commit d7f8208814
17 changed files with 217 additions and 387 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
plugins/channelrx/demodam/amdemodsink.h
View File

@ -22,8 +22,7 @@
#include "dsp/nco.h"
#include "dsp/interpolator.h"
#include "dsp/agc.h"
#include "dsp/bandpass.h"
#include "dsp/lowpass.h"
#include "dsp/firfilter.h"
#include "dsp/phaselockcomplex.h"
#include "audio/audiofifo.h"
#include "util/movingaverage.h"

2
plugins/channelrx/demodbfm/bfmdemodsink.h
View File

@ -23,7 +23,7 @@
#include "dsp/channelsamplesink.h"
#include "dsp/nco.h"
#include "dsp/interpolator.h"
#include "dsp/lowpass.h"
#include "dsp/firfilter.h"
#include "dsp/movingaverage.h"
#include "dsp/fftfilt.h"
#include "dsp/phaselock.h"

3
plugins/channelrx/demoddsd/dsddemodsink.h
View File

@ -22,8 +22,7 @@
#include "dsp/phasediscri.h"
#include "dsp/nco.h"
#include "dsp/interpolator.h"
#include "dsp/lowpass.h"
#include "dsp/bandpass.h"
#include "dsp/firfilter.h"
#include "dsp/afsquelch.h"
#include "dsp/afsquelch.h"
#include "audio/audiofifo.h"

3
plugins/channelrx/demodnfm/nfmdemodsink.h
View File

@ -24,8 +24,7 @@
#include "dsp/phasediscri.h"
#include "dsp/nco.h"
#include "dsp/interpolator.h"
#include "dsp/lowpass.h"
#include "dsp/bandpass.h"
#include "dsp/firfilter.h"
#include "dsp/afsquelch.h"
#include "dsp/agc.h"
#include "dsp/ctcssdetector.h"

2
plugins/channelrx/demodwfm/wfmdemodsink.h
View File

@ -23,7 +23,7 @@
#include "dsp/channelsamplesink.h"
#include "dsp/nco.h"
#include "dsp/interpolator.h"
#include "dsp/lowpass.h"
#include "dsp/firfilter.h"
#include "util/movingaverage.h"
#include "dsp/fftfilt.h"
#include "dsp/phasediscri.h"

3
plugins/channelrx/freqtracker/freqtrackersink.h
View File

@ -27,8 +27,7 @@
#include "dsp/interpolator.h"
#include "util/movingaverage.h"
#include "dsp/agc.h"
#include "dsp/bandpass.h"
#include "dsp/lowpass.h"
#include "dsp/firfilter.h"
#include "dsp/phaselockcomplex.h"
#include "dsp/freqlockcomplex.h"
#include "util/message.h"

2
plugins/channelrx/udpsink/udpsinksink.h
View File

@ -27,7 +27,7 @@
#include "dsp/phasediscri.h"
#include "dsp/movingaverage.h"
#include "dsp/agc.h"
#include "dsp/bandpass.h"
#include "dsp/firfilter.h"
#include "util/udpsinkutil.h"
#include "audio/audiofifo.h"

4
plugins/channeltx/mod802.15.4/ieee_802_15_4_modsource.h
View File

@ -29,9 +29,7 @@
#include "dsp/nco.h"
#include "dsp/ncof.h"
#include "dsp/interpolator.h"
#include "dsp/lowpass.h"
#include "dsp/bandpass.h"
#include "dsp/highpass.h"
#include "dsp/firfilter.h"
#include "dsp/raisedcosine.h"
#include "dsp/fmpreemphasis.h"
#include "util/lfsr.h"

3
plugins/channeltx/modnfm/nfmmodsource.h
View File

@ -27,8 +27,7 @@
#include "dsp/nco.h"
#include "dsp/ncof.h"
#include "dsp/interpolator.h"
#include "dsp/lowpass.h"
#include "dsp/bandpass.h"
#include "dsp/firfilter.h"
#include "dsp/filterrc.h"
#include "util/movingaverage.h"
#include "dsp/cwkeyer.h"

4
plugins/channeltx/modpacket/packetmodsource.h
View File

@ -29,9 +29,7 @@
#include "dsp/nco.h"
#include "dsp/ncof.h"
#include "dsp/interpolator.h"
#include "dsp/lowpass.h"
#include "dsp/bandpass.h"
#include "dsp/highpass.h"
#include "dsp/firfilter.h"
#include "dsp/raisedcosine.h"
#include "dsp/fmpreemphasis.h"
#include "util/lfsr.h"

4
sdrbase/CMakeLists.txt
View File

@ -101,6 +101,7 @@ set(sdrbase_SOURCES
dsp/filtermbe.cpp
dsp/filerecord.cpp
dsp/filerecordinterface.cpp
dsp/firfilter.cpp
dsp/fmpreemphasis.cpp
dsp/freqlockcomplex.cpp
dsp/interpolator.cpp
@ -108,7 +109,6 @@ set(sdrbase_SOURCES
dsp/glspectrumsettings.cpp
dsp/hbfilterchainconverter.cpp
dsp/hbfiltertraits.cpp
dsp/lowpass.cpp
dsp/mimochannel.cpp
dsp/nco.cpp
dsp/ncof.cpp
@ -245,6 +245,7 @@ set(sdrbase_HEADERS
dsp/filtermbe.h
dsp/filerecord.h
dsp/filerecordinterface.h
dsp/firfilter.h
dsp/fmpreemphasis.h
dsp/freqlockcomplex.h
dsp/gfft.h
@ -266,7 +267,6 @@ set(sdrbase_HEADERS
dsp/inthalfbandfiltersti.h
dsp/kissfft.h
dsp/kissengine.h
dsp/lowpass.h
dsp/mimochannel.h
dsp/misc.h
dsp/movingaverage.h

134
sdrbase/dsp/bandpass.h
View File

@ -1,134 +0,0 @@
#ifndef INCLUDE_BANDPASS_H
#define INCLUDE_BANDPASS_H
#define _USE_MATH_DEFINES
#include <math.h>
#include "dsp/dsptypes.h"
// #undef M_PI
// #define M_PI 3.14159265358979323846
template <class Type> class Bandpass {
public:
Bandpass() : m_ptr(0) { }
void create(int nTaps, double sampleRate, double lowCutoff, double highCutoff)
{
std::vector<Real> taps_lp;
std::vector<Real> taps_hp;
double wcl = 2.0 * M_PI * lowCutoff;
double Wcl = wcl / sampleRate;
double wch = 2.0 * M_PI * highCutoff;
double Wch = wch / sampleRate;
int i;
// check constraints
if (!(nTaps & 1))
{
qDebug("Bandpass filter has to have an odd number of taps");
nTaps++;
}
// make room
m_samples.resize(nTaps);
for (int i = 0; i < nTaps; i++) {
m_samples[i] = 0;
}
m_ptr = 0;
m_taps.resize(nTaps / 2 + 1);
taps_lp.resize(nTaps / 2 + 1);
taps_hp.resize(nTaps / 2 + 1);
// generate Sinc filter core
for (i = 0; i < nTaps / 2 + 1; i++)
{
if (i == (nTaps - 1) / 2)
{
taps_lp[i] = Wch / M_PI;
taps_hp[i] = -(Wcl / M_PI);
}
else
{
taps_lp[i] = sin(((double)i - ((double)nTaps - 1.0) / 2.0) * Wch) / (((double)i - ((double)nTaps - 1.0) / 2.0) * M_PI);
taps_hp[i] = -sin(((double)i - ((double)nTaps - 1.0) / 2.0) * Wcl) / (((double)i - ((double)nTaps - 1.0) / 2.0) * M_PI);
}
}
taps_hp[(nTaps - 1) / 2] += 1;
// apply Hamming window and combine lowpass and highpass
for (i = 0; i < nTaps / 2 + 1; i++)
{
taps_lp[i] *= 0.54 + 0.46 * cos((2.0 * M_PI * ((double)i - ((double)nTaps - 1.0) / 2.0)) / (double)nTaps);
taps_hp[i] *= 0.54 + 0.46 * cos((2.0 * M_PI * ((double)i - ((double)nTaps - 1.0) / 2.0)) / (double)nTaps);
m_taps[i] = -(taps_lp[i]+taps_hp[i]);
}
m_taps[(nTaps - 1) / 2] += 1;
// normalize
Real sum = 0;
for (i = 0; i < (int)m_taps.size() - 1; i++) {
sum += m_taps[i] * 2;
}
sum += m_taps[i] - 1;
for (i = 0; i < (int)m_taps.size(); i++) {
m_taps[i] /= sum;
}
}
Type filter(Type sample)
{
Type acc = 0;
int a = m_ptr;
int b = a - 1;
int i, n_taps, size;
m_samples[m_ptr] = sample;
size = m_samples.size(); // Valgrind optim (2)
while (b < 0) {
b += size;
}
n_taps = m_taps.size() - 1; // Valgrind optim
for(i = 0; i < n_taps; i++)
{
acc += (m_samples[a] + m_samples[b]) * m_taps[i];
a++;
while (a >= size) {
a -= size;
}
b--;
while(b < 0) {
b += size;
}
}
acc += m_samples[a] * m_taps[i];
m_ptr++;
while (m_ptr >= size) {
m_ptr -= size;
}
return acc;
}
private:
std::vector<Real> m_taps;
std::vector<Type> m_samples;
int m_ptr;
};
#endif // INCLUDE_BANDPASS_H

52
sdrbase/dsp/firfilter.cpp Normal file
View File

@ -0,0 +1,52 @@
///////////////////////////////////////////////////////////////////////////////////
// Copyright (C) 2020 kasper93 //
// written by Kacper Michajłow and Edouard Griffiths //
// //
// 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 //
// (at your option) any later version. //
// //
// 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/>. //
///////////////////////////////////////////////////////////////////////////////////
#include "firfilter.h"
void FirFilterGenerators::generateLowPassFilter(int nTaps, double sampleRate, double cutoff, std::vector<Real> &taps)
{
if (!(nTaps & 1))
{
qDebug("Filter has to have an odd number of taps");
nTaps++;
}
double Wc = (2.0 * M_PI * cutoff) / sampleRate;
int halfTaps = nTaps / 2 + 1;
taps.resize(halfTaps);
for (int i = 0; i < halfTaps; ++i)
{
if (i == halfTaps - 1)
{
taps[i] = Wc / M_PI;
}
else
{
int n = i - (nTaps - 1) / 2;
taps[i] = sin(n * Wc) / (n * M_PI);
}
}
// Blackman window
for (int i = 0; i < halfTaps; i++)
{
int n = i - (nTaps - 1) / 2;
taps[i] *= 0.42 + 0.5 * cos((2.0 * M_PI * n) / nTaps) + 0.08 * cos((4.0 * M_PI * n) / nTaps);
}
}

153
sdrbase/dsp/firfilter.h Normal file
View File

@ -0,0 +1,153 @@
///////////////////////////////////////////////////////////////////////////////////
// Copyright (C) 2020 kasper93 //
// written by Kacper Michajłow and Edouard Griffiths //
// //
// 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 //
// (at your option) any later version. //
// //
// 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/>. //
///////////////////////////////////////////////////////////////////////////////////
#pragma once
#define _USE_MATH_DEFINES
#include <math.h>
#include "dsp/dsptypes.h"
#include "export.h"
class FirFilterGenerators
{
public:
static void generateLowPassFilter(int nTaps, double sampleRate, double cutoff, std::vector<Real> &taps);
};
template <class Type>
class FirFilter
{
public:
Type filter(Type sample)
{
Type acc = 0;
int n_samples = m_samples.size();
int n_taps = m_taps.size() - 1;
int a = m_ptr;
int b = a == 0 ? n_samples - 1 : a - 1;
m_samples[m_ptr] = sample;
for (size_t i = 0; i < n_taps; ++i)
{
acc += (m_samples[a++] + m_samples[b--]) * m_taps[i];
if (a == n_samples) {
a = 0;
}
if (b == -1) {
b = n_samples - 1;
}
}
acc += m_samples[a] * m_taps[n_taps];
if (++m_ptr == n_samples) {
m_ptr = 0;
}
return acc;
}
protected:
void init(int nTaps)
{
m_ptr = 0;
m_samples.resize(nTaps);
for (int i = 0; i < nTaps; i++) {
m_samples[i] = 0;
}
}
void normalize(Real sum_fix = 0.0)
{
Real sum = 0;
size_t i;
for (i = 0; i < m_taps.size() - 1; ++i) {
sum += m_taps[i] * 2.0;
}
sum += m_taps[i] + sum_fix;
for (i = 0; i < m_taps.size(); ++i) {
m_taps[i] /= sum;
}
}
protected:
std::vector<Real> m_taps;
std::vector<Type> m_samples;
size_t m_ptr;
};
template <class T>
struct Lowpass : public FirFilter<T>
{
public:
void create(int nTaps, double sampleRate, double cutoff)
{
this->init(nTaps);
FirFilterGenerators::generateLowPassFilter(nTaps, sampleRate, cutoff, this->m_taps);
this->normalize();
}
};
template <class T>
struct Bandpass : public FirFilter<T>
{
void create(int nTaps, double sampleRate, double lowCutoff, double highCutoff)
{
this->init(nTaps);
FirFilterGenerators::generateLowPassFilter(nTaps, sampleRate, highCutoff, this->m_taps);
std::vector<Real> highPass;
FirFilterGenerators::generateLowPassFilter(nTaps, sampleRate, lowCutoff, highPass);
for (size_t i = 0; i < highPass.size(); ++i) {
highPass[i] = -highPass[i];
}
highPass[highPass.size() - 1] += 1;
for (size_t i = 0; i < this->m_taps.size(); ++i) {
this->m_taps[i] = -(this->m_taps[i] + highPass[i]);
}
this->m_taps[this->m_taps.size() - 1] += 1;
this->normalize(-1.0);
}
};
template <class T>
struct Highpass : public FirFilter<T>
{
void create(int nTaps, double sampleRate, double cutoff)
{
this->init(nTaps);
FirFilterGenerators::generateLowPassFilter(nTaps, sampleRate, cutoff, this->m_taps);
for (size_t i = 0; i < this->m_taps.size(); ++i) {
this->m_taps[i] = -this->m_taps[i];
}
this->m_taps[this->m_taps.size() - 1] += 1;
this->normalize(-1.0);
}
};

114
sdrbase/dsp/highpass.h
View File

@ -1,114 +0,0 @@
#ifndef INCLUDE_HIGHPASS_H
#define INCLUDE_HIGHPASS_H
#define _USE_MATH_DEFINES
#include <math.h>
#include "dsp/dsptypes.h"
template <class Type> class Highpass {
public:
Highpass() { }
void create(int nTaps, double sampleRate, double cutoff)
{
double wc = 2.0 * M_PI * cutoff;
double Wc = wc / sampleRate;
int i;
// check constraints
if (!(nTaps & 1))
{
qDebug("Highpass filter has to have an odd number of taps");
nTaps++;
}
// make room
m_samples.resize(nTaps);
for (int i = 0; i < nTaps; i++) {
m_samples[i] = 0;
}
m_ptr = 0;
m_taps.resize(nTaps / 2 + 1);
// generate Sinc filter core for lowpass but inverting every other tap for highpass keeping center tap
for (i = 0; i < nTaps / 2 + 1; i++)
{
if (i == (nTaps - 1) / 2) {
m_taps[i] = -(Wc / M_PI);
} else {
m_taps[i] = -sin(((double)i - ((double)nTaps - 1.0) / 2.0) * Wc) / (((double)i - ((double)nTaps - 1.0) / 2.0) * M_PI);
}
}
m_taps[(nTaps - 1) / 2] += 1;
// apply Hamming window
for (i = 0; i < nTaps / 2 + 1; i++) {
m_taps[i] *= 0.54 + 0.46 * cos((2.0 * M_PI * ((double)i - ((double)nTaps - 1.0) / 2.0)) / (double)nTaps);
}
// normalize
Real sum = 0;
for (i = 0; i < (int)m_taps.size() - 1; i++) {
sum += m_taps[i] * 2;
}
sum += m_taps[i] - 1;
for (i = 0; i < (int)m_taps.size(); i++) {
m_taps[i] /= sum;
}
}
Type filter(Type sample)
{
Type acc = 0;
int a = m_ptr;
int b = a - 1;
int i, n_taps, size;
m_samples[m_ptr] = sample;
size = m_samples.size(); // Valgrind optim (2)
while (b < 0) {
b += size;
}
n_taps = m_taps.size() - 1; // Valgrind optim
for (i = 0; i < n_taps; i++)
{
acc += (m_samples[a] + m_samples[b]) * m_taps[i];
a++;
while (a >= size) {
a -= size;
}
b--;
while (b < 0) {
b += size;
}
}
acc += m_samples[a] * m_taps[i];
m_ptr++;
while (m_ptr >= size) {
m_ptr -= size;
}
return acc;
}
private:
std::vector<Real> m_taps;
std::vector<Type> m_samples;
int m_ptr;
};
#endif // INCLUDE_HIGHPASS_H

6
sdrbase/dsp/lowpass.cpp
View File

@ -1,6 +0,0 @@
#include <stdio.h>
#include <QtGlobal>
#define _USE_MATH_DEFINES
#include <math.h>
#include <vector>
#include "dsp/lowpass.h"

112
sdrbase/dsp/lowpass.h
View File

@ -1,112 +0,0 @@
#ifndef INCLUDE_LOWPASS_H
#define INCLUDE_LOWPASS_H
#define _USE_MATH_DEFINES
#include <math.h>
#include "dsp/dsptypes.h"
template <class Type> class Lowpass {
public:
Lowpass() : m_ptr(0) { }
void create(int nTaps, double sampleRate, double cutoff)
{
double wc = 2.0 * M_PI * cutoff;
double Wc = wc / sampleRate;
int i;
// check constraints
if (!(nTaps & 1))
{
qDebug("Lowpass filter has to have an odd number of taps");
nTaps++;
}
// make room
m_samples.resize(nTaps);
for (int i = 0; i < nTaps; i++) {
m_samples[i] = 0;
}
m_ptr = 0;
m_taps.resize(nTaps / 2 + 1);
// generate Sinc filter core
for (i = 0; i < nTaps / 2 + 1; i++)
{
if(i == (nTaps - 1) / 2) {
m_taps[i] = Wc / M_PI;
} else {
m_taps[i] = sin(((double)i - ((double)nTaps - 1.0) / 2.0) * Wc) / (((double)i - ((double)nTaps - 1.0) / 2.0) * M_PI);
}
}
// apply Hamming window
for (i = 0; i < nTaps / 2 + 1; i++) {
m_taps[i] *= 0.54 + 0.46 * cos((2.0 * M_PI * ((double)i - ((double)nTaps - 1.0) / 2.0)) / (double)nTaps);
}
// normalize
Real sum = 0;
for (i = 0; i < (int)m_taps.size() - 1; i++) {
sum += m_taps[i] * 2;
}
sum += m_taps[i];
for (i = 0; i < (int)m_taps.size(); i++) {
m_taps[i] /= sum;
}
}
Type filter(Type sample)
{
Type acc = 0;
int a = m_ptr;
int b = a - 1;
int i, n_taps, size;
m_samples[m_ptr] = sample;
size = m_samples.size(); // Valgrind optim (2)
while (b < 0) {
b += size;
}
n_taps = m_taps.size() - 1; // Valgrind optim
for (i = 0; i < n_taps; i++)
{
acc += (m_samples[a] + m_samples[b]) * m_taps[i];
a++;
while (a >= size) {
a -= size;
}
b--;
while(b < 0) {
b += size;
}
}
acc += m_samples[a] * m_taps[i];
m_ptr++;
while(m_ptr >= size) {
m_ptr -= size;
}
return acc;
}
private:
std::vector<Real> m_taps;
std::vector<Type> m_samples;
int m_ptr;
};
#endif // INCLUDE_LOWPASS_H