GetDiff8x16B

Evaluates difference between current bi-predicted and mean of two reference blocks of 8x16 elements.

Syntax

IppStatus ippiGetDiff8x16B_8u16s_C1(const Ipp8u* pSrcCur, Ipp32s srcCurStep, const Ipp8u* pSrcRefF, Ipp32s srcRefStepF, Ipp32s mcTypeF, const Ipp8u* pSrcRefB, Ipp32s srcRefStepB, Ipp32s mcTypeB, Ipp16s* pDstDiff, Ipp32s dstDiffStep, Ipp32s roundControl);

Parameters

pSrcCur	Pointer to the block of size 8x16 in the current plane.
srcCurStep	Step of the current block, specifying width of the plane in bytes.
pSrcRefF	Pointer to the forward block of size 8x16 in the reference plane.
srcRefStepF	Step of the forward reference block, specifying width of the plane in bytes.
mcTypeF	Type of the following forward MC type IPPVC_MC_APX.
pSrcRefB	Pointer to the backward block of size 8x16 in the reference plane.
srcRefStepB	Step of the backward reference block, specifying width of the block in bytes.
mcTypeB	Type of the following backward MC type IPPVC_MC_APX.
pDstDiff	Pointer to the destination block of specified size containing difference between the current and reference blocks.
dstDiffStep	Step of the destination block, specifying width of the block in bytes.
roundControl	Parameter that determines type of rounding for half a pel approximation; may be 0 or 1.

Description

This function is declared in the ippvc.h header file. The function ippiGetDiff8x16B_8u16s_C1 evaluates the difference between the current block and the mean of two reference blocks of specified size. One of the reference blocks is called forward and belongs to the previous frame in accordance with the type of motion compensation. The other block is called backward and belongs to one of the following frames. The result is stored in block pDstDiff. Encoding is performed accurate to half a pel and rounding must be specified.

Return Values

ippStsNoErr	Indicates no error.
ippStsNullPtrErr	Indicates an error when at least one input pointer is NULL.