mirror of
https://github.com/thestk/stk
synced 2026-01-12 20:41:52 +00:00
118 lines
3.1 KiB
C++
118 lines
3.1 KiB
C++
/***************************************************/
|
|
/*! \class DelayA
|
|
\brief STK allpass interpolating delay line class.
|
|
|
|
This Delay subclass implements a fractional-
|
|
length digital delay-line using a first-order
|
|
allpass filter. A fixed maximum length
|
|
of 4095 and a delay of 0.5 is set using the
|
|
default constructor. Alternatively, the
|
|
delay and maximum length can be set during
|
|
instantiation with an overloaded constructor.
|
|
|
|
An allpass filter has unity magnitude gain but
|
|
variable phase delay properties, making it useful
|
|
in achieving fractional delays without affecting
|
|
a signal's frequency magnitude response. In
|
|
order to achieve a maximally flat phase delay
|
|
response, the minimum delay possible in this
|
|
implementation is limited to a value of 0.5.
|
|
|
|
by Perry R. Cook and Gary P. Scavone, 1995 - 2002.
|
|
*/
|
|
/***************************************************/
|
|
|
|
#include "DelayA.h"
|
|
#include <iostream.h>
|
|
|
|
DelayA :: DelayA()
|
|
{
|
|
this->setDelay( 0.5 );
|
|
apInput = 0.0;
|
|
}
|
|
|
|
DelayA :: DelayA(MY_FLOAT theDelay, long maxDelay)
|
|
{
|
|
// Writing before reading allows delays from 0 to length-1.
|
|
length = maxDelay+1;
|
|
|
|
if ( length > 4096 ) {
|
|
// We need to delete the previously allocated inputs.
|
|
delete [] inputs;
|
|
inputs = new MY_FLOAT[length];
|
|
this->clear();
|
|
}
|
|
|
|
inPoint = 0;
|
|
this->setDelay(theDelay);
|
|
}
|
|
|
|
DelayA :: ~DelayA()
|
|
{
|
|
}
|
|
|
|
void DelayA :: clear()
|
|
{
|
|
Delay::clear();
|
|
apInput = 0.0;
|
|
}
|
|
|
|
void DelayA :: setDelay(MY_FLOAT theDelay)
|
|
{
|
|
MY_FLOAT outPointer;
|
|
|
|
if (theDelay > length-1) {
|
|
cerr << "DelayA: setDelay(" << theDelay << ") too big!" << endl;
|
|
// Force delay to maxLength
|
|
outPointer = inPoint + 1.0;
|
|
delay = length - 1;
|
|
}
|
|
else if (theDelay < 0.5) {
|
|
cerr << "DelayA: setDelay(" << theDelay << ") less than 0.5 not possible!" << endl;
|
|
outPointer = inPoint + 0.4999999999;
|
|
delay = 0.5;
|
|
}
|
|
else {
|
|
outPointer = inPoint - theDelay + 1.0; // outPoint chases inpoint
|
|
delay = theDelay;
|
|
}
|
|
|
|
if (outPointer < 0)
|
|
outPointer += length; // modulo maximum length
|
|
|
|
outPoint = (long) outPointer; // integer part
|
|
alpha = 1.0 + outPoint - outPointer; // fractional part
|
|
|
|
if (alpha < 0.5) {
|
|
// The optimal range for alpha is about 0.5 - 1.5 in order to
|
|
// achieve the flattest phase delay response.
|
|
outPoint += 1;
|
|
if (outPoint >= length) outPoint -= length;
|
|
alpha += (MY_FLOAT) 1.0;
|
|
}
|
|
|
|
coeff = ((MY_FLOAT) 1.0 - alpha) /
|
|
((MY_FLOAT) 1.0 + alpha); // coefficient for all pass
|
|
}
|
|
|
|
MY_FLOAT DelayA :: tick(MY_FLOAT sample)
|
|
{
|
|
inputs[inPoint++] = sample;
|
|
|
|
// Increment input pointer modulo length.
|
|
if (inPoint == length)
|
|
inPoint -= length;
|
|
|
|
// Take delay-line output and increment modulo length.
|
|
MY_FLOAT temp = inputs[outPoint++];
|
|
if (outPoint == length)
|
|
outPoint -= length;
|
|
|
|
// Do allpass interpolation delay.
|
|
outputs[0] = -coeff * outputs[0];
|
|
outputs[0] += apInput + (coeff * temp);
|
|
apInput = temp;
|
|
|
|
return outputs[0];
|
|
}
|