kernel void CrossFade( float ratio, int width, int height, global uchar* pInput0, int inputStride0, global uchar* pInput1, int inputStride1, global uchar* pOutput, int outputStride ) { size_t x = get_global_id(0); size_t y = get_global_id(1); global uchar4* pI0 = (global uchar4*)(pInput0+y*inputStride0); global uchar4* pI1 = (global uchar4*)(pInput1+y*inputStride1); global uchar4* pO = (global uchar4*)(pOutput+y*outputStride); uchar4 mixed; mixed = convert_uchar4_sat_rte( mix( convert_float4(pI0[x]), convert_float4(pI1[x]), ratio ) ); pO[x] = mixed; }