hashtags/lmms
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases 1 Wiki Activity

Merge pull request #895 from diizy/filters

Browse Source 
basic_filters: optimize RC highpass & bandpass filters
This commit is contained in:
Tobias Doerffel
2014-06-28 11:55:47 +02:00
parent e6ffb6cd35 27383f91b7
commit 0447e0fe44
1 changed files with 195 additions and 214 deletions
Download Patch File Download Diff File Expand all files Collapse all files

409
include/basic_filters.h
View File

@@ -1,11 +1,11 @@
/*
* basic_filters.h - simple but powerful filter-class with most used filters
*
* original file by ???
* original file by ???
* modified and enhanced by Tobias Doerffel
*
* Copyright (c) 2004-2009 Tobias Doerffel <tobydox/at/users.sourceforge.net>
*
*
* This file is part of Linux MultiMedia Studio - http://lmms.sourceforge.net
*
* This program is free software; you can redistribute it and/or
@@ -26,8 +26,8 @@
*/
#ifndef _BASIC_FILTERS_H
#define _BASIC_FILTERS_H
#ifndef BASIC_FILTERS_H
#define BASIC_FILTERS_H
#ifndef __USE_XOPEN
#define __USE_XOPEN
@@ -39,6 +39,7 @@
#include "Mixer.h"
#include "templates.h"
#include "lmms_constants.h"
#include "interpolation.h"
//#include <iostream>
//#include <cstdlib>
@@ -127,16 +128,16 @@ public:
// reset in/out history for simple filters
m_ou1[_chnl] = m_ou2[_chnl] = m_in1[_chnl] =
m_in2[_chnl] = 0.0f;
// reset in/out history for moog-filter
m_y1[_chnl] = m_y2[_chnl] = m_y3[_chnl] = m_y4[_chnl] =
m_oldx[_chnl] = m_oldy1[_chnl] =
m_oldy2[_chnl] = m_oldy3[_chnl] = 0.0f;
// reset in/out history for RC-filters
m_rclp0[_chnl] = m_rcbp0[_chnl] = m_rchp0[_chnl] = m_rclast0[_chnl] = 0.0f;
m_rclp1[_chnl] = m_rcbp1[_chnl] = m_rchp1[_chnl] = m_rclast1[_chnl] = 0.0f;
for(int i=0; i<6; i++)
m_vflp[i][_chnl] = m_vfbp[i][_chnl] = m_vfhp[i][_chnl] = m_vflast[i][_chnl] = 0.0f;
}
@@ -186,130 +187,135 @@ public:
// (C) 1998 ... 2009 S.Fendt. Released under the GPL v2.0 or any later version.
case Lowpass_RC12:
case Bandpass_RC12:
case Highpass_RC12:
{
sample_t lp, hp, bp;
sample_t in;
// 4-times oversampled... (even the moog-filter would benefit from this)
sample_t lp, bp, hp, in;
for( int n = 4; n != 0; --n )
{
in = _in0 + m_rcbp0[_chnl] * m_rcq;
in = (in > +1.f) ? +1.f : in;
in = (in < -1.f) ? -1.f : in;
in = qBound( -1.0f, in, 1.0f );
lp = in * m_rcb + m_rclp0[_chnl] * m_rca;
lp = (lp > +1.f) ? +1.f : lp;
lp = (lp < -1.f) ? -1.f : lp;
lp = qBound( -1.0f, lp, 1.0f );
hp = m_rcc * ( m_rchp0[_chnl] + in - m_rclast0[_chnl] );
hp = (hp > +1.f) ? +1.f : hp;
hp = (hp < -1.f) ? -1.f : hp;
hp = qBound( -1.0f, hp, 1.0f );
bp = hp * m_rcb + m_rcbp0[_chnl] * m_rca;
bp = (bp > +1.f) ? +1.f : bp;
bp = (bp < -1.f) ? -1.f : bp;
bp = qBound( -1.0f, bp, 1.0f );
m_rclast0[_chnl] = in;
m_rclp0[_chnl] = lp;
m_rchp0[_chnl] = hp;
m_rcbp0[_chnl] = bp;
}
return lp;
break;
}
case Highpass_RC12:
case Bandpass_RC12:
{
sample_t hp, bp, in;
for( int n = 4; n != 0; --n )
{
in = _in0 + m_rcbp0[_chnl] * m_rcq;
in = qBound( -1.0f, in, 1.0f );
if( m_type == Lowpass_RC12 )
out = lp;
else if( m_type == Bandpass_RC12 )
out = bp;
else
out = hp;
return( out );
hp = m_rcc * ( m_rchp0[_chnl] + in - m_rclast0[_chnl] );
hp = qBound( -1.0f, hp, 1.0f );
bp = hp * m_rcb + m_rcbp0[_chnl] * m_rca;
bp = qBound( -1.0f, bp, 1.0f );
m_rclast0[_chnl] = in;
m_rchp0[_chnl] = hp;
m_rcbp0[_chnl] = bp;
}
return m_type == Highpass_RC12 ? hp : bp;
break;
}
case Lowpass_RC24:
case Bandpass_RC24:
case Highpass_RC24:
{
sample_t lp, hp, bp;
sample_t in;
sample_t lp, bp, hp, in;
for( int n = 4; n != 0; --n )
{
// first stage is as for the 12dB case...
in = _in0 + m_rcbp0[_chnl] * m_rcq;
in = (in > +1.f) ? +1.f : in;
in = (in < -1.f) ? -1.f : in;
in = qBound( -1.0f, in, 1.0f );
lp = in * m_rcb + m_rclp0[_chnl] * m_rca;
lp = (lp > +1.f) ? +1.f : lp;
lp = (lp < -1.f) ? -1.f : lp;
lp = qBound( -1.0f, lp, 1.0f );
hp = m_rcc * ( m_rchp0[_chnl] + in - m_rclast0[_chnl] );
hp = (hp > +1.f) ? +1.f : hp;
hp = (hp < -1.f) ? -1.f : hp;
hp = qBound( -1.0f, hp, 1.0f );
bp = hp * m_rcb + m_rcbp0[_chnl] * m_rca;
bp = (bp > +1.f) ? +1.f : bp;
bp = (bp < -1.f) ? -1.f : bp;
bp = qBound( -1.0f, bp, 1.0f );
m_rclast0[_chnl] = in;
m_rclp0[_chnl] = lp;
m_rcbp0[_chnl] = bp;
m_rchp0[_chnl] = hp;
// second stage gets the output of the first stage as input...
in = lp + m_rcbp1[_chnl] * m_rcq;
in = qBound( -1.0f, in, 1.0f );
lp = in * m_rcb + m_rclp1[_chnl] * m_rca;
lp = qBound( -1.0f, lp, 1.0f );
hp = m_rcc * ( m_rchp1[_chnl] + in - m_rclast1[_chnl] );
hp = qBound( -1.0f, hp, 1.0f );
bp = hp * m_rcb + m_rcbp1[_chnl] * m_rca;
bp = qBound( -1.0f, bp, 1.0f );
m_rclast1[_chnl] = in;
m_rclp1[_chnl] = lp;
m_rcbp1[_chnl] = bp;
m_rchp1[_chnl] = hp;
}
return lp;
break;
}
case Highpass_RC24:
case Bandpass_RC24:
{
sample_t hp, bp, in;
for( int n = 4; n != 0; --n )
{
// first stage is as for the 12dB case...
in = _in0 + m_rcbp0[_chnl] * m_rcq;
in = qBound( -1.0f, in, 1.0f );
hp = m_rcc * ( m_rchp0[_chnl] + in - m_rclast0[_chnl] );
hp = qBound( -1.0f, hp, 1.0f );
bp = hp * m_rcb + m_rcbp0[_chnl] * m_rca;
bp = qBound( -1.0f, bp, 1.0f );
m_rclast0[_chnl] = in;
m_rchp0[_chnl] = hp;
m_rcbp0[_chnl] = bp;
// second stage gets the output of the first stage as input...
if( m_type == Lowpass_RC24 )
{
in = lp + m_rcbp1[_chnl] * m_rcq;
}
else if( m_type == Bandpass_RC24 )
{
in = bp + m_rcbp1[_chnl] * m_rcq;
}
else
{
in = hp + m_rcbp1[_chnl] * m_rcq;
}
in = (in > +1.f) ? +1.f : in;
in = (in < -1.f) ? -1.f : in;
in = m_type == Highpass_RC24
? hp + m_rcbp1[_chnl] * m_rcq
: bp + m_rcbp1[_chnl] * m_rcq;
lp = in * m_rcb + m_rclp1[_chnl] * m_rca;
lp = (lp > +1.f) ? +1.f : lp;
lp = (lp < -1.f) ? -1.f : lp;
in = qBound( -1.0f, in, 1.0f );
hp = m_rcc * ( m_rchp1[_chnl] + in - m_rclast1[_chnl] );
hp = (hp > +1.f) ? +1.f : hp;
hp = (hp < -1.f) ? -1.f : hp;
hp = qBound( -1.0f, hp, 1.0f );
bp = hp * m_rcb + m_rcbp1[_chnl] * m_rca;
bp = (bp > +1.f) ? +1.f : bp;
bp = (bp < -1.f) ? -1.f : bp;
bp = qBound( -1.0f, bp, 1.0f );
m_rclast1[_chnl] = in;
m_rclp1[_chnl] = lp;
m_rchp1[_chnl] = hp;
m_rcbp1[_chnl] = bp;
}
// output is second stage-lowpass...
if( m_type == Lowpass_RC24 )
{
out = lp;
}
else if( m_type == Bandpass_RC24 )
{
out = bp;
}
else
{
out = hp;
}
return out;
return m_type == Highpass_RC24 ? hp : bp;
break;
}
@@ -317,146 +323,122 @@ public:
{
sample_t lp, hp, bp, in;
out = 0;
for(int o=0; o<4; o++)
{
// first formant
in = _in0 + m_vfbp[0][_chnl] * m_vfq;
in = (in > +1.f) ? +1.f : in;
in = (in < -1.f) ? -1.f : in;
out = 0;
for(int o=0; o<4; o++)
{
// first formant
in = _in0 + m_vfbp[0][_chnl] * m_vfq;
in = qBound( -1.0f, in, 1.0f );
lp = in * m_vfb[0] + m_vflp[0][_chnl] * m_vfa[0];
lp = (lp > +1.f) ? +1.f : lp;
lp = (lp < -1.f) ? -1.f : lp;
lp = in * m_vfb[0] + m_vflp[0][_chnl] * m_vfa[0];
lp = qBound( -1.0f, lp, 1.0f );
hp = m_vfc[0] * ( m_vfhp[0][_chnl] + in - m_vflast[0][_chnl] );
hp = (hp > +1.f) ? +1.f : hp;
hp = (hp < -1.f) ? -1.f : hp;
hp = m_vfc[0] * ( m_vfhp[0][_chnl] + in - m_vflast[0][_chnl] );
hp = qBound( -1.0f, hp, 1.0f );
bp = hp * m_vfb[0] + m_vfbp[0][_chnl] * m_vfa[0];
bp = (bp > +1.f) ? +1.f : bp;
bp = (bp < -1.f) ? -1.f : bp;
bp = hp * m_vfb[0] + m_vfbp[0][_chnl] * m_vfa[0];
bp = qBound( -1.0f, bp, 1.0f );
m_vflast[0][_chnl] = in;
m_vflp[0][_chnl] = lp;
m_vfhp[0][_chnl] = hp;
m_vfbp[0][_chnl] = bp;
m_vflast[0][_chnl] = in;
m_vflp[0][_chnl] = lp;
m_vfhp[0][_chnl] = hp;
m_vfbp[0][_chnl] = bp;
in = bp + m_vfbp[2][_chnl] * m_vfq;
in = (in > +1.f) ? +1.f : in;
in = (in < -1.f) ? -1.f : in;
in = bp + m_vfbp[2][_chnl] * m_vfq;
in = qBound( -1.0f, in, 1.0f );
lp = in * m_vfb[0] + m_vflp[2][_chnl] * m_vfa[0];
lp = (lp > +1.f) ? +1.f : lp;
lp = (lp < -1.f) ? -1.f : lp;
lp = in * m_vfb[0] + m_vflp[2][_chnl] * m_vfa[0];
lp = qBound( -1.0f, lp, 1.0f );
hp = m_vfc[0] * ( m_vfhp[2][_chnl] + in - m_vflast[2][_chnl] );
hp = (hp > +1.f) ? +1.f : hp;
hp = (hp < -1.f) ? -1.f : hp;
hp = m_vfc[0] * ( m_vfhp[2][_chnl] + in - m_vflast[2][_chnl] );
hp = qBound( -1.0f, hp, 1.0f );
bp = hp * m_vfb[0] + m_vfbp[2][_chnl] * m_vfa[0];
bp = (bp > +1.f) ? +1.f : bp;
bp = (bp < -1.f) ? -1.f : bp;
bp = hp * m_vfb[0] + m_vfbp[2][_chnl] * m_vfa[0];
bp = qBound( -1.0f, bp, 1.0f );
m_vflast[2][_chnl] = in;
m_vflp[2][_chnl] = lp;
m_vfhp[2][_chnl] = hp;
m_vfbp[2][_chnl] = bp;
in = bp + m_vfbp[4][_chnl] * m_vfq;
in = (in > +1.f) ? +1.f : in;
in = (in < -1.f) ? -1.f : in;
m_vflast[2][_chnl] = in;
m_vflp[2][_chnl] = lp;
m_vfhp[2][_chnl] = hp;
m_vfbp[2][_chnl] = bp;
lp = in * m_vfb[0] + m_vflp[4][_chnl] * m_vfa[0];
lp = (lp > +1.f) ? +1.f : lp;
lp = (lp < -1.f) ? -1.f : lp;
in = bp + m_vfbp[4][_chnl] * m_vfq;
in = qBound( -1.0f, in, 1.0f );
hp = m_vfc[0] * ( m_vfhp[4][_chnl] + in - m_vflast[4][_chnl] );
hp = (hp > +1.f) ? +1.f : hp;
hp = (hp < -1.f) ? -1.f : hp;
lp = in * m_vfb[0] + m_vflp[4][_chnl] * m_vfa[0];
lp = qBound( -1.0f, lp, 1.0f );
bp = hp * m_vfb[0] + m_vfbp[4][_chnl] * m_vfa[0];
bp = (bp > +1.f) ? +1.f : bp;
bp = (bp < -1.f) ? -1.f : bp;
hp = m_vfc[0] * ( m_vfhp[4][_chnl] + in - m_vflast[4][_chnl] );
hp = qBound( -1.0f, hp, 1.0f );
m_vflast[4][_chnl] = in;
m_vflp[4][_chnl] = lp;
m_vfhp[4][_chnl] = hp;
m_vfbp[4][_chnl] = bp;
bp = hp * m_vfb[0] + m_vfbp[4][_chnl] * m_vfa[0];
bp = qBound( -1.0f, bp, 1.0f );
out += bp;
m_vflast[4][_chnl] = in;
m_vflp[4][_chnl] = lp;
m_vfhp[4][_chnl] = hp;
m_vfbp[4][_chnl] = bp;
// second formant
in = _in0 + m_vfbp[0][_chnl] * m_vfq;
in = (in > +1.f) ? +1.f : in;
in = (in < -1.f) ? -1.f : in;
out += bp;
lp = in * m_vfb[1] + m_vflp[1][_chnl] * m_vfa[1];
lp = (lp > +1.f) ? +1.f : lp;
lp = (lp < -1.f) ? -1.f : lp;
// second formant
in = _in0 + m_vfbp[0][_chnl] * m_vfq;
in = qBound( -1.0f, in, 1.0f );
hp = m_vfc[1] * ( m_vfhp[1][_chnl] + in - m_vflast[1][_chnl] );
hp = (hp > +1.f) ? +1.f : hp;
hp = (hp < -1.f) ? -1.f : hp;
lp = in * m_vfb[1] + m_vflp[1][_chnl] * m_vfa[1];
lp = qBound( -1.0f, lp, 1.0f );
bp = hp * m_vfb[1] + m_vfbp[1][_chnl] * m_vfa[1];
bp = (bp > +1.f) ? +1.f : bp;
bp = (bp < -1.f) ? -1.f : bp;
hp = m_vfc[1] * ( m_vfhp[1][_chnl] + in - m_vflast[1][_chnl] );
hp = qBound( -1.0f, hp, 1.0f );
m_vflast[1][_chnl] = in;
m_vflp[1][_chnl] = lp;
m_vfhp[1][_chnl] = hp;
m_vfbp[1][_chnl] = bp;
bp = hp * m_vfb[1] + m_vfbp[1][_chnl] * m_vfa[1];
bp = qBound( -1.0f, bp, 1.0f );
in = bp + m_vfbp[3][_chnl] * m_vfq;
in = (in > +1.f) ? +1.f : in;
in = (in < -1.f) ? -1.f : in;
m_vflast[1][_chnl] = in;
m_vflp[1][_chnl] = lp;
m_vfhp[1][_chnl] = hp;
m_vfbp[1][_chnl] = bp;
lp = in * m_vfb[1] + m_vflp[3][_chnl] * m_vfa[1];
lp = (lp > +1.f) ? +1.f : lp;
lp = (lp < -1.f) ? -1.f : lp;
in = bp + m_vfbp[3][_chnl] * m_vfq;
in = qBound( -1.0f, in, 1.0f );
hp = m_vfc[1] * ( m_vfhp[3][_chnl] + in - m_vflast[3][_chnl] );
hp = (hp > +1.f) ? +1.f : hp;
hp = (hp < -1.f) ? -1.f : hp;
lp = in * m_vfb[1] + m_vflp[3][_chnl] * m_vfa[1];
lp = qBound( -1.0f, lp, 1.0f );
bp = hp * m_vfb[1] + m_vfbp[3][_chnl] * m_vfa[1];
bp = (bp > +1.f) ? +1.f : bp;
bp = (bp < -1.f) ? -1.f : bp;
hp = m_vfc[1] * ( m_vfhp[3][_chnl] + in - m_vflast[3][_chnl] );
hp = qBound( -1.0f, hp, 1.0f );
m_vflast[3][_chnl] = in;
m_vflp[3][_chnl] = lp;
m_vfhp[3][_chnl] = hp;
m_vfbp[3][_chnl] = bp;
bp = hp * m_vfb[1] + m_vfbp[3][_chnl] * m_vfa[1];
bp = qBound( -1.0f, bp, 1.0f );
in = bp + m_vfbp[5][_chnl] * m_vfq;
in = (in > +1.f) ? +1.f : in;
in = (in < -1.f) ? -1.f : in;
m_vflast[3][_chnl] = in;
m_vflp[3][_chnl] = lp;
m_vfhp[3][_chnl] = hp;
m_vfbp[3][_chnl] = bp;
lp = in * m_vfb[1] + m_vflp[5][_chnl] * m_vfa[1];
lp = (lp > +1.f) ? +1.f : lp;
lp = (lp < -1.f) ? -1.f : lp;
in = bp + m_vfbp[5][_chnl] * m_vfq;
in = qBound( -1.0f, in, 1.0f );
hp = m_vfc[1] * ( m_vfhp[5][_chnl] + in - m_vflast[5][_chnl] );
hp = (hp > +1.f) ? +1.f : hp;
hp = (hp < -1.f) ? -1.f : hp;
lp = in * m_vfb[1] + m_vflp[5][_chnl] * m_vfa[1];
lp = qBound( -1.0f, lp, 1.0f );
bp = hp * m_vfb[1] + m_vfbp[5][_chnl] * m_vfa[1];
bp = (bp > +1.f) ? +1.f : bp;
bp = (bp < -1.f) ? -1.f : bp;
hp = m_vfc[1] * ( m_vfhp[5][_chnl] + in - m_vflast[5][_chnl] );
hp = qBound( -1.0f, hp, 1.0f );
m_vflast[5][_chnl] = in;
m_vflp[5][_chnl] = lp;
m_vfhp[5][_chnl] = hp;
m_vfbp[5][_chnl] = bp;
bp = hp * m_vfb[1] + m_vfbp[5][_chnl] * m_vfa[1];
bp = qBound( -1.0f, bp, 1.0f );
m_vflast[5][_chnl] = in;
m_vflp[5][_chnl] = lp;
m_vfhp[5][_chnl] = hp;
m_vfbp[5][_chnl] = bp;
out += bp;
}
out += bp;
}
return( out/2.0f );
break;
}
default:
// filter
out = m_b0a0*_in0 +
@@ -488,9 +470,6 @@ public:
/*, const bool _q_is_bandwidth = false*/ )
{
// temp coef vars
_freq = qBound(minFreq(), _freq, 20000.0f); // limit freq and q for not getting
// bad noise out of the filter...
_q = qMax( _q, minQ() );
if( m_type == Lowpass_RC12 ||
@@ -500,41 +479,39 @@ public:
m_type == Bandpass_RC24 ||
m_type == Highpass_RC24 )
{
if( _freq < 50.f )
{
_freq = 50.f;
}
_freq = qBound( 50.0f, _freq, 20000.0f );
m_rca = 1.0f - (1.0f/(m_sampleRate*4)) / ( (1.0f/(_freq*2.0f*M_PI)) + (1.0f/(m_sampleRate*4)) );
m_rcb = 1.0f - m_rca;
m_rcc = (1.0f/(_freq*2.0f*M_PI)) / ( (1.0f/(_freq*2.0f*M_PI)) + (1.0f/(m_sampleRate*4)) );
// Stretch Q/resonance, as self-oscillation reliably starts at a q of ~2.5 - ~2.6
m_rcq = _q/4.f;
m_rcq = _q * 0.25f;
return;
}
if( m_type == Formantfilter )
{
_freq = qBound( minFreq(), _freq, 20000.0f ); // limit freq and q for not getting bad noise out of the filter...
// formats for a, e, i, o, u, a
const float _f[5][2] = { { 1000, 1400 }, { 500, 2300 },
static const float _f[5][2] = { { 1000, 1400 }, { 500, 2300 },
{ 320, 3200 },
{ 500, 1000 },
{ 320, 800 } };
static const float freqRatio = 4.0f / 14000.0f;
// Stretch Q/resonance
m_vfq = _q/4.f;
m_vfq = _q * 0.25f;
// frequency in lmms ranges from 1Hz to 14000Hz
const int vowel = (int)( floor( _freq/14000.f * 4.f ) );
const float fract = ( _freq/14000.f * 4.f ) -
(float)vowel;
const float vowelf = _freq * freqRatio;
const int vowel = static_cast<int>( vowelf );
const float fract = vowelf - vowel;
// interpolate between formant frequencies
const float f0 = _f[vowel+0][0] * ( 1.0f - fract ) +
_f[vowel+1][0] * ( fract );
const float f1 = _f[vowel+0][1] * ( 1.0f - fract ) +
_f[vowel+1][1] * ( fract );
// interpolate between formant frequencies
const float f0 = linearInterpolate( _f[vowel+0][0], _f[vowel+1][0], fract );
const float f1 = linearInterpolate( _f[vowel+0][1], _f[vowel+1][1], fract );
m_vfa[0] = 1.0f - (1.0f/(m_sampleRate*4)) /
( (1.0f/(f0*2.0f*M_PI)) +
@@ -551,10 +528,13 @@ public:
m_vfc[1] = (1.0f/(f1*2.0f*M_PI)) /
( (1.0f/(f1*2.0f*M_PI)) +
(1.0f/(m_sampleRate*4)) );
return;
}
if( m_type == Moog )
{
_freq = qBound( minFreq(), _freq, 20000.0f );
// [ 0 - 0.5 ]
const float f = _freq / m_sampleRate;
// (Empirical tunning)
@@ -572,6 +552,7 @@ public:
}
// other filters
_freq = qBound( minFreq(), _freq, 20000.0f );
const float omega = F_2PI * _freq / m_sampleRate;
const float tsin = sinf( omega );
const float tcos = cosf( omega );
@@ -647,7 +628,7 @@ private:
// coeffs for formant-filters
float m_vfa[4], m_vfb[4], m_vfc[4], m_vfq;
typedef sample_t frame[CHANNELS];
// in/out history
@@ -655,14 +636,14 @@ private:
// in/out history for moog-filter
frame m_y1, m_y2, m_y3, m_y4, m_oldx, m_oldy1, m_oldy2, m_oldy3;
// in/out history for RC-type-filters
frame m_rcbp0, m_rclp0, m_rchp0, m_rclast0;
frame m_rcbp1, m_rclp1, m_rchp1, m_rclast1;
// in/out history for Formant-filters
frame m_vfbp[6], m_vflp[6], m_vfhp[6], m_vflast[6];
FilterTypes m_type;
bool m_doubleFilter;