Provides finite impulse response (FIR) filters More...

Static Public Member Functions
static double []	BandPass (double fc_Low, double fc_High, double fs, int samples, Window.Type w)
	Returns a normalized band pass filter kernel using the specified window. More...

static double []	BandStop (double fc_Low, double fc_High, double fs, int samples, Window.Type w)
	Returns a normalized band stop filter kernel using the specified window. More...

static double []	DC_Unity (double[] A)
	DC unity gain More...

static double []	Filter (double[] A, double[] kernel)
	Filter operation used to filter data using the specified filter kernel. More...

static double []	HBandH (int samples, Window.Type w)
	Returns a normalized high pass half band filter kernel using the specified window More...

static double []	HBandL (int samples, Window.Type w)
	Returns a normalized low pass half band filter kernel using the specified window More...

static double []	HighPass (double fc, double fs, int samples, Window.Type w)
	Returns a normalized high pass filter kernel using the specified window. More...

static double []	LowPass (double fc, double fs, int samples, Window.Type w)
	Returns a normalized low pass filter kernel using the specified window. More...

static double []	Sinc (double f_fs, int samples)
	Returns a rectangular sinc filter kernel, that is a truncated and shifted version of the ideal sinc function More...

static double []	SpectralInverse (double[] kernel)
	Invert the frequency spectrum of the specified filter kernel. More...

Detailed Description

Provides finite impulse response (FIR) filters

Member Function Documentation

◆ BandPass()

static double [] dsp.FIR.BandPass	(	double	fc_Low,
		double	fc_High,
		double	fs,
		int	samples,
		Window.Type	w
	)

static

Returns a normalized band pass filter kernel using the specified window.

The impulse response has unity gain at DC and left-right symmetry.

Parameters

fc_Low	low cut-off frequency (Hz)
fc_High	high cut-off frequency (Hz)
fs	sampling rate (Hz)
samples	kernel size, should be an odd number
w	Window type

Example

double[] kernel = FIR.BandPass(25, 75, 200, 11, Window.Type.Blackman);
kernel = NArray.Round(kernel, 8);
Console.WriteLine(NArray.ToString(kernel));

0 0 -0.00320435 -0.17950679 -0.04066185 0.44674597 -0.04066185 -0.17950679 -0.00320435 0 0

◆ BandStop()

static double [] dsp.FIR.BandStop	(	double	fc_Low,
		double	fc_High,
		double	fs,
		int	samples,
		Window.Type	w
	)

static

Returns a normalized band stop filter kernel using the specified window.

The impulse response has unity gain at DC and left-right symmetry.

Parameters

fc_Low	low cut-off frequency (Hz)
fc_High	high cut-off frequency (Hz)
fs	sampling rate (Hz)
samples	kernel size, should be an odd number
w	Window type

Example

double[] kernel = FIR.BandStop(25, 75, 200, 11, Window.Type.Blackman);
kernel = NArray.Round(kernel, 8);
Console.WriteLine(NArray.ToString(kernel));

0 0 0.00320435 0.17950679 0.04066185 0.55325403 0.04066185 0.17950679 0.00320435 0 0

◆ DC_Unity()

static double [] dsp.FIR.DC_Unity ( double [] A )

static

DC unity gain

Parameters

A	real array

Example

double[] kernel = FIR.Sinc(0.25, 11);
kernel = FIR.DC_Unity(kernel);
kernel = NArray.Round(kernel, 4);
Console.WriteLine(NArray.ToString(kernel));

0.0605 0 -0.1009 0 0.3027 0.4754 0.3027 0 -0.1009 0 0.0605

◆ Filter()

static double [] dsp.FIR.Filter	(	double []	A,
		double []	kernel
	)

static

Filter operation used to filter data using the specified filter kernel.

FIR filter function runs the convolution function Signal.Convolution and removes the extra samples from the end.
The output signal has the same size as the input signal.
Note: If the same filter kernel is to be used by many signals, run the kernel function just once, and apply the filter function per signal.

Parameters

A	real array
kernel	filter kernel

Example
The following example shows how to filter an input signal (A) using a low pass filter.

double[] A = new double[] { 3, 7, -2, 1, -5, 3, 9, 4 };
double[] kernel = FIR.LowPass(5, 20, 10, Window.Type.Rectangular);
double[] result = FIR.Filter( A, kernel);
result = NArray.Round(result, 2);
Console.WriteLine(NArray.ToString(result));

0 -0.32 -0.75 1.18 3.67 4.4 1.24 -3.04

◆ HBandH()

static double [] dsp.FIR.HBandH	(	int	samples,
		Window.Type	w
	)

static

Returns a normalized high pass half band filter kernel using the specified window

The cut off frequency is set to 0.25 as a fraction of the sampling rate. The impulse response has unity gain at DC and left-right symmetry.

Parameters

samples	kernel size, should be an odd number
w	Window type

Example

double[] kernel = FIR.HBandH(11, Window.Type.Blackman);
kernel = NArray.Round(kernel, 8);
Console.WriteLine(NArray.ToString(kernel));

0 0 0.02134438 0 -0.27085136 0.49901394 -0.27085136 0 0.02134438 0 0

◆ HBandL()

static double [] dsp.FIR.HBandL	(	int	samples,
		Window.Type	w
	)

static

Returns a normalized low pass half band filter kernel using the specified window

The cut off frequency is set to 0.25 as a fraction of the sampling rate. The impulse response has unity gain at DC and left-right symmetry.

Parameters

samples	kernel size, should be an odd number
w	Window type

Example

double[] kernel = FIR.HBandL(11, Window.Type.Blackman);
kernel = NArray.Round(kernel, 8);
Console.WriteLine(NArray.ToString(kernel));

0 0 -0.02134438 0 0.27085136 0.50098606 0.27085136 0 -0.02134438 0 0

◆ HighPass()

static double [] dsp.FIR.HighPass	(	double	fc,
		double	fs,
		int	samples,
		Window.Type	w
	)

static

Returns a normalized high pass filter kernel using the specified window.

The impulse response has unity gain at DC and left-right symmetry.

Parameters

fc	cut-off frequency (Hz)
fs	sampling rate (Hz)
samples	kernel size, should be an odd number
w	Window type

Example

double[] kernel = FIR.HighPass(25, 200, 11, Window.Type.Blackman);
kernel = NArray.Round(kernel, 8);
Console.WriteLine(NArray.ToString(kernel));

0 0 -0.01826524 -0.09838329 -0.23177828 0.69685363 -0.23177828 -0.09838329 -0.01826524 0 0

◆ LowPass()

static double [] dsp.FIR.LowPass	(	double	fc,
		double	fs,
		int	samples,
		Window.Type	w
	)

static

Returns a normalized low pass filter kernel using the specified window.

The impulse response has unity gain at DC and left-right symmetry.

Parameters

fc	cut-off frequency (Hz)
fs	sampling rate (Hz)
samples	kernel size, should be an odd number
w	Window type

Example

double[] kernel = FIR.LowPass(25, 200, 11, Window.Type.Blackman);
kernel = NArray.Round(kernel, 8);
Console.WriteLine(NArray.ToString(kernel));

0 0 0.01826524 0.09838329 0.23177828 0.30314637 0.23177828 0.09838329 0.01826524 0 0

◆ Sinc()

static double [] dsp.FIR.Sinc	(	double	f_fs,
		int	samples
	)

static

Returns a rectangular sinc filter kernel, that is a truncated and shifted version of the ideal sinc function

Parameters

f_fs	frequency as a fraction of the sampling rate (0..0.5).
samples	kernel size, should be an odd number

Example

double[] kernel = FIR.Sinc(0.25, 11);
kernel = NArray.Round(kernel, 8);
Console.WriteLine(NArray.ToString(kernel));

0.2 0 -0.33333333 0 1 1.57079633 1 0 -0.33333333 0 0.2

◆ SpectralInverse()

static double [] dsp.FIR.SpectralInverse ( double [] kernel )

static

Invert the frequency spectrum of the specified filter kernel.

Spectral inversion is performed in time domain and results the frequency response of the impulse to being flipped top-for-bottom.

Parameters

kernel filter kernel with unity gain at DC

Example

double[] kernel = FIR.Sinc(0.125, 11);
kernel = FIR.DC_Unity(kernel);
kernel = NArray.Round(kernel, 4);
Console.WriteLine(NArray.ToString(kernel));
kernel = FIR.SpectralInverse(kernel);
Console.WriteLine(NArray.ToString(kernel));

-0.0417 0 0.0696 0.1476 0.2087 0.2318 0.2087 0.1476 0.0696 0 -0.0417

0.0417 0 -0.0696 -0.1476 -0.2087 0.7682 -0.2087 -0.1476 -0.0696 0 0.0417

Static Public Member Functions

Detailed Description

Member Function Documentation

◆ BandPass()

◆ BandStop()

◆ DC_Unity()

◆ Filter()

◆ HBandH()

◆ HBandL()

◆ HighPass()

◆ LowPass()

◆ Sinc()

◆ SpectralInverse()