Smoothing of spectral data in the Fourier domain

Abstract

Smoothing of spectral data using Fourier transforms is described and demonstrated with Lorentzian, sinc(2), and sine smoothing functions. Four parameters are defined and used to study the smoothing operation. It is also concluded that the best smoothing function is a sinc function if we require that the distortions due to the smoothing operation are smaller than the residual noise. Sine smoothing using Fourier transforms is also compared to least square smoothing in the frequency domain, and the advantages of sine smoothing are illustrated.