Performs inverse quantization on an inter coded block stored in a one-dimensional buffer.
Syntax
IppStatus ippiQuantInvInter_H263_16s_C1I(Ipp16s* pSrcDst, int indxLastNonZero, int QP, int modQuantFlag);
Parameters
pSrcDst |
Pointer to the coefficient buffer of the block. The parameter contains quantized coefficients on input and dequantized coefficients on output. |
indxLastNonZero |
Index of the last non-zero coefficient. This parameter provides faster operation. If the value is unknown, set to 63. |
QP |
Quantization parameter. |
modQuantFlag |
Flag equal to a non-zero value when Modified Quantization mode is in use, equal to 0 otherwise. |
Description
The function ippiQuantInvInter_H263_16s_C1I is declared in the ippvc.h header file. This function performs inverse quantization on inter coded block. The dequantization process is specified in [ITUH263], subclause 6.2.1. When not in Modified Quantization mode, the output coefficients are clipped to the range [-2048, 2047] ([ITUH263], subclause 6.2.2). The overall procedure is defined in [ITUH263] as:
where i = 0, 1, 2, ..., indxLastNonZero
if (modQuantFlag == 0):
pDst[i] = MIN(MAX(pDst[i],-2048),2047).
Note
The function ippiQuantInvInter_H263_16s_C1I can be applied to a buffer of arbitrary size (indxLastNonZero can be any positive number), and can thus be used, for example, to process multiple blocks in one call. In this case, for any intra block following the first one, the intra DC should be processed separately, if not in Advanced Intra mode.
This function is used in the H.261, H.263, and MPEG-4 encoders and decoders included into Intel IPP Samples. See introduction to H.263.
Return Values
ippStsNoErr |
Indicates no error. |
ippStsNullPtrErr |
Indicates an error condition if pSrcDst is NULL. |
ippStsQPErr |
Indicates an error condition if QP is out of the range [1, 31]. |
ippStsOutOfRangeErr |
Indicates an error condition if indxLastNonZero is negative. |