A least-squares method of transforming interferograms into spectra is compared with the traditional fast Fourier transform (FFT) for varying degree of point-spacing unevenness. A Gaussian distribution is assumed for the deviations from even spacing. FFT gives inferior spectra quality above a critical unevenness, which decreases as more interferograms are averaged.
H. Weidner and R. E. Peale, "Quality Analysis for Least-Squares Transformation of Unevenly Spaced Interferograms," Appl. Spectrosc. 52, 587-590 (1998)