r8b::CDSPFIRFilter Class Reference

Calculation and storage class for FIR filters. More...

#include <CDSPFIRFilter.h>

Inheritance diagram for r8b::CDSPFIRFilter:

Public Member Functions
int	getBlockLenBits () const
	Returns filter's block length, expressed as Nth power of 2. The actual length is twice as large due to zero-padding.
const double *	getKernelBlock () const
	Returns pointer to filter's kernel block, in complex-numbered form obtained via the CDSPRealFFT::forward() function call.
int	getKernelLen () const
	Returns filter kernel length, in samples. Not to be confused with the block length.
int	getLatency () const
	Returns filter's latency, in samples (integer part).
double	getLatencyFrac () const
	Returns filter's latency, in samples (fractional part). Always zero for linear-phase filters.
bool	isZeroPhase () const
	Returns true if kernel block of this filter has zero-phase response.
void	unref ()
	Reduces reference count to this object.

Static Public Member Functions
static double	getLPMaxAtten ()
	Returns the maximal allowed low-pass filter's stop-band attenuation, in decibel.
static double	getLPMaxTransBand ()
	Returns the maximal allowed low-pass filter's transition band, in percent.
static double	getLPMinAtten ()
	Returns the minimal allowed low-pass filter's stop-band attenuation, in decibel.
static double	getLPMinTransBand ()
	Returns the minimal allowed low-pass filter's transition band, in percent.

Friends
class	CDSPFIRFilterCache

Detailed Description

Calculation and storage class for FIR filters.

Class that implements calculation and storing of a FIR filter (currently contains low-pass filter calculation routine designed for sample rate conversion). Objects of this class cannot be created directly, but can be obtained via the CDSPFilterCache::getLPFilter() static function.

Member Function Documentation

getKernelBlock()

const double * r8b::CDSPFIRFilter::getKernelBlock

(

)

const

Returns pointer to filter's kernel block, in complex-numbered form obtained via the CDSPRealFFT::forward() function call.

Zero-padded, gain-adjusted with the CDSPRealFFT::getInvMulConst() multiplied by ReqGain constant, immediately suitable for convolution. Kernel block may have "zero-phase" response, depending on the isZeroPhase() function's result.

unref()

void r8b::CDSPFIRFilter::unref

(

)

Reduces reference count to this object.

This function should be called when the filter obtained via the filter cache is no longer needed.