Previous research has demonstrated several advantages accrued by transforming dispersion-based near-infrared spectra from the wavelength domain to the Fourier domain. Those advantages include: (1) smoothing wavelength-domain data without loss of end points, (2) correcting for the particle size anomaly encountered in solid sample analyses by simply omitting the mean term Fourier coefficient from the "retransformation" process, (3) minimizing the multicollinearity problem prevalent in wavelength space, (4) generating wavelength-space derivatives from Fourier space without loss of end points, (5) performing band enhancements via Fourier self-deconvolution, (6) identifying sample type using Fourier vectors, (7) estimating chemical composition using only the first few (≤12) Fourier coefficients, (8) cutting of computer storage requirements by more than 96%, (9) cutting of calibration time by more than 96% and, hence, (10) reducing the drudgery of maintaining calibrations.
R. M. Hoy and W. F. McClure, "Fourier Transform Near-Infrared Spectrometry: Using Interferograms to Determine Chemical Composition," Appl. Spectrosc. 43, 1102-1104 (1989)