smoothing

Smoothing may be applied to spectroscopic data for a several purposes—for example, to improve the quality of the spectral appearance, to smooth out or remove apparent noise, to decrease spectral resolution to match bandwidth between two spectra measured at two resolutions, and to remove apparent random and structured variation in signals. Smoothing allows enhanced manual viewing and multivariate processing by eliminating spectral variation caused by random measurement noise. The most common applications of smoothing within spectroscopy include boxcar smoothing, Fourier-domain (fast Fourier transform [FFT]) smoothing, and Savitzky-Golay smoothing.