is used to get a Fourier size equal to the exact size of the time data window. This may give slow execution if the size of the time data window is not an order of two.