Use a Spatial FFT () dataset to compute the spatial Fourier transform of a variable or expression. You can add a Spatial FFT dataset by right-clicking the Datasets node and choose Spatial FFT from the More Datasets submenu.

From the Dataset list, choose the dataset to use a the input for the spatial FFT.

Click the Preview Spatial Samples button to get a plot of the spatial resolution in the Graphics window.

From the Resolution list, choose Automatic (the default) to have the spatial sampling resolution set up automatically, or choose Manual to specify the sampling resolution using integers larger than 2 in the Nx, Ny, and (in 3D) Nz fields under Sampling resolution.

From the Layout list under Spatial layout, choose Normal (the default), Periodic, or Use zero padding. For the last layout option, specify paddings as integer values (zero or larger) in the x padding, y padding, and (in 3D) z padding fields.

Under Fourier space variables, the default variable names for the fx, fy, and (in 3D) fz Fourier space variables are sfft1fx, sffft1fy, and sfft1z for a Spatial FFT 1 dataset. If desired, enter other variable names. From the Domain list, choose Frequency (the default), corresponding to the spatial frequencies, or Index, having the axes values as the 0-based DFT indices.

The Mask DC check box is selected by default to set the DC value (zero frequency component) to zero. This is usually necessary because this value is too high and overwhelms the rest of the spectrum. If you want to include the DC value, clear this check box.

From the Resolution list, choose Extra fine, Finer, Fine, Normal (the default), Coarse, No refinement, or Custom. A higher resolution means that elements are split into smaller patches during rendering. If you chose Custom, enter a refinement as a positive integer pin the Element refinement field.