First: sorry for my bad english, and for my newbiez
I’m using CUDA to realize a gauss filter on a bmp file.
I have 2 float matrix with:
data value from the bmp file
data value of filter matrix
To apply FFT2D :
cufftHandle plan;
cufftComplex *idata;
cufftReal *odata;
/* Create a 2D FFT plan. */
cufftPlan2d(&plan, nx, ny, CUFFT_C2R);
/* Use the CUFFT plan to transform the signal out of place. */
cufftExecC2R(plan, idata, odata);
But my idata is a float matrix, not cufftComplex data.
i use some fft to do big multiplication integer at beginning complex at the end so i use complex
i thinks even you have float you must use complex if you have 5.3 change in (5.3, 0) complex