A full implementation of a Wiener filter requires detailed knowledge of the signal and noise power spectra. Many times this information is simply unavailable or is very poorly known. In such situations, it is convenient to assign some constant value to the SNR and use that in the Wiener filter implentation. This is a somewhat extreme restriction of the Wiener filter, but it makes the filter very easy to use and can produce better results than the pseudoinverse filter in kinverse. (Better in the sense of much less ringing).
Note that the input data objects will be treated simply as a field of complex data, presumably obtained by application of an FFT, although the data could be generated by many other means. The output data will be either KCOMPLEX if the input data is any type other than KDCOMPLEX. Otherwise, the output will be of type KDCOMPLEX. Data objects with maps attached are not supported. Mask data is ignored, if present.