INTERPOLATION OF MINIMALLY SAMPLED SPECTRA
MetadataShow full item record
Publisher:Ohio State University
In fourier transform spectroscopy the information in the numerical representation of the spectrum is complete when the number of samples in the spectrum is equal to the number of samples in the interferogram. However, for purposes of display and for such analyses as line position and shape determinations, it is desirable to have a more continuous coverage of points along the spectrum. It is well known that additional points can be generated in three equivalent ways; discrete fourier transformation with a unequal sided cosine matrix, zero filling, and convolution in the spectral domain. For reasons of efficiency in data manipulation we have investigated the possibility of interpolating spectral results only when needed. There is presently no mathematical basis for justifying interpolation of combined spectral results such as transmittance of absorbance spectra; but the procedure may nevertheless be justified when the limitations are understood.
Items in Knowledge Bank are protected by copyright, with all rights reserved, unless otherwise indicated.