MityDSP::tcSigProcNoise Class Reference

#include <SigProc/SigProcNoise.h>

List of all members.

Static Public Member Functions
static float	Gaussian (void)
static float	Uniform (void)
static void	Seed (int seed)
static void	zigNorInit (void)
static double	DRanNormalZig (void)
Static Protected Member Functions
static double	DRanNormalTail (double dMin, int iNegative)
Static Protected Attributes
static bool	mbSeeded = false
	when true, srand() has been called
static double	m_adZigX [ZIGNOR_C+1]
static double	m_adZigR [ZIGNOR_C]

---

Detailed Description

The tcSigProcNoise class provides utility methods for generating noise functions.

See also:

tcSigProcNoise Page

---

Member Function Documentation

float tcSigProcNoise::Gaussian

(

void

)

[static]

This routine will generate a gaussian distributed number with standard deviation equal to 1.0 and mean of zero using the Box-Muller method.

Warning:

this method is not thread safe.

Return values:

a

Gaussian random number with mean 0 and std=1.0

Note:

this function taken from Numerical Recipes in C, chapter 7.2

float tcSigProcNoise::Uniform

(

void

)

[static]

This routine will generate a uniform distributed number from 0.0-1.0 exclusive of the endpoints (0.0 and 1.0 will not be returned).

Return values:

a

uniform number between 0.0 and 1.0 (exclusive)

void tcSigProcNoise::Seed

(

int

seed = -1

)

[static]

This routine will seed the base random number generater provided by the BIOS.

Parameters:

seed

if less than zero, the system local tick timer is used.

void tcSigProcNoise::zigNorInit

(

void

)

[static]

This method must be called prior to calling DRanNormalZig()

double tcSigProcNoise::DRanNormalZig

(

void

)

[static]

Compute a gaussian random number for std = 1.0 and mean = 0.0 using the Ziggurat algorithm. Note: zigNorInit() must be called prior to using this routine the first time.

Returns:

gaussian randome number

double tcSigProcNoise::DRanNormalTail	(	double	dMin,
		int	iNegative
	)			[static, protected]

This method is used exclusively by DRanNormalZig.

Parameters:

	dMin	minimum point on curve
	iNegative	0 if positive side of curve, non-zero otherwise

Returns:

gaussian number on the tail of the PDF curve

---

Member Data Documentation

bool tcSigProcNoise::mbSeeded = false [static, protected]

when true, srand() has been called

double tcSigProcNoise::m_adZigX [static, protected]

m_adZigX holds coordinates, such that each rectangle has same area; m_adZigR holds m_adZigX[i + 1] / m_adZigX[i]

double tcSigProcNoise::m_adZigR [static, protected]