MityDSP::tcSigProcVector Class Reference

#include <SigProc/SigProcVector.h>

List of all members.

Static Public Member Functions
static int	VectorMultiply (short *restrict InA, float *restrict InB, float *restrict Out, int N, int StrideA, int StrideB, int StrideOut)
static int	VectorMultiply (float *restrict InA, float *restrict InB, float *restrict Out, int N, int StrideA, int StrideB, int StrideOut)
static int	ToFloat (short *restrict InA, float *restrict Out, int N, int StrideA, int StrideOut)
static int	VecMulAdd (float *restrict InX, float *restrict InM, float InB, float *restrict Out, int N, int StrideX, int StrideM, int StrideOut)
static int	VecAddMul (float *restrict InX, float *restrict InM, float InB, float *restrict Out, int N, int StrideX, int StrideM, int StrideOut)

---

Detailed Description

The tcSigProcVector class provides utility methods for handling math vectorized operations in an efficient manner.

See also:

tcSigProcVector Page

---

Member Function Documentation

int tcSigProcVector::VectorMultiply	(	short *restrict	InA,
		float *restrict	InB,
		float *restrict	Out,
		int	N,
		int	StrideA = 1,
		int	StrideB = 1,
		int	StrideOut = 1
	)			[static]

Performs a vector multiply according to:

Out[index] = InA[index]*InB[index]

Parameters:

	InA	vector of N input elements
	StrideA	Word increment needed to get to next item in A
	InB	vector of N input elements
	StrideB	Word increment needed to get to next item in B
	Out	location to write the output buffer
	StrideOut	Word increment needed to get to next item in Out
	N	Number of words in the vectors. This must be a minimum of 20, must be divisible by 4.

return -1 if N is less than 20 (no operations done) or 0 on success

int tcSigProcVector::VectorMultiply	(	float *restrict	InA,
		float *restrict	InB,
		float *restrict	Out,
		int	N,
		int	StrideA = 1,
		int	StrideB = 1,
		int	StrideOut = 1
	)			[static]

Performs a vector multiply according to:

Out[index] = InA[index]*InB[index]

Parameters:

	InA	vector of N input elements
	StrideA	Word increment needed to get to next item in A
	InB	vector of N input elements
	StrideB	Word increment needed to get to next item in B
	Out	location to write the output buffer
	StrideOut	Word increment needed to get to next item in Out
	N	Number of words in the vectors. This must be a minimum of 20, must be divisible by 4.

return -1 if N is less than 20 (no operations done) or 0 on success

int tcSigProcVector::ToFloat	(	short *restrict	InA,
		float *restrict	Out,
		int	N,
		int	StrideA = 1,
		int	StrideOut = 1
	)			[static]

Converts show data vector to floating point data vector:

Out[index] = (float)(InA[index])

Parameters:

	InA	vector of N input elements
	StrideA	Word increment needed to get to next item in A
	Out	location to write the output buffer
	StrideOut	Word increment needed to get to next item in Out
	N	Number of words in the vectors. This must be a minimum of 20, must be divisible by 4.

return -1 if N is less than 20 (no operations done) or 0 on success

int tcSigProcVector::VecMulAdd	(	float *restrict	InX,
		float *restrict	InM,
		float	InB,
		float *restrict	Out,
		int	N,
		int	StrideX = 1,
		int	StrideM = 1,
		int	StrideOut = 1
	)			[static]

Performs a vector multiply then add (to a constant) according to:

Out[index] = InX[index]*InM[index]+InB

Parameters:

	InX	vector of N input elements
	StrideX	Word increment needed to get to next item in X
	InM	vector of N input elements
	StrideM	Word increment needed to get to next item in M
	InB	constant to be added per index
	Out	location to write the output buffer
	StrideOut	Word increment needed to get to next item in Out
	N	Number of words in the vectors. This must be a minimum of 20, must be divisible by 4.

return -1 if N is less than 20 (no operations done) or 0 on success

int tcSigProcVector::VecAddMul	(	float *restrict	InX,
		float *restrict	InM,
		float	InB,
		float *restrict	Out,
		int	N,
		int	StrideX = 1,
		int	StrideM = 1,
		int	StrideOut = 1
	)			[static]

Performs a vector add then multiply according to:

Out[index] = (InX[index]+InB)*InM[index]

Parameters:

	InX	vector of N input elements
	StrideX	Word increment needed to get to next item in X
	InM	vector of N input elements
	StrideM	Word increment needed to get to next item in M
	InB	constant to be added per index
	Out	location to write the output buffer
	StrideOut	Word increment needed to get to next item in Out
	N	Number of words in the vectors. This must be a minimum of 20, must be divisible by 4.

return -1 if N is less than 20 (no operations done) or 0 on success