Abstract

Computers are common tools in spectroscopy and they offer new possibilities for extracting useful information from experimental data. One of these tools is resolution enhancement of infrared spectra with Fourier self-deconvolution. This procedure is often used for the analysis of the amide I band in proteins. It is part of most commercial spectroscopic software packages, especially in the FT-IR software in which Fourier transformation is the basis of the experimental procedure. Thus Fourier self-deconvolution is becoming a powerful tool in spectroscopy. Since the number of users is expected to increase rapidly, a warning against the problems and pitfalls of the technique is appropriate. Usually one starts the software with a certain set of deconvolution parameters and varies them by a visual check of the result. The resolution enhancement is increased until the spectrum looks unrealistic. Two kinds of artifacts can occur during the process: over-deconvolution, which results in the appearance of side lobes, and the uncontrolled increase of the noise.

Noise in Fourier self-deconvolution

Jyrki K. Kauppinen, Douglas J. Moffatt, David G. Cameron, and Henry H. Mantsch
Appl. Opt. 20(10) 1866-1879 (1981)

Artifact-free deconvolution in light field microscopy

Anca Stefanoiu, Josue Page, Panagiotis Symvoulidis, Gil G. Westmeyer, and Tobias Lasser
Opt. Express 27(22) 31644-31666 (2019)

Physically constrained Fourier transform deconvolution method

Francis A. Flaherty
J. Opt. Soc. Am. A 26(5) 1191-1194 (2009)

Ghosts and artifacts in Fourier-transform spectrometry

R. C. M. Learner, A. P. Thorne, and J. W. Brault
Appl. Opt. 35(16) 2947-2954 (1996)

Parametric blind-deconvolution algorithm to remove image artifacts in hybrid imaging systems

Mads Demenikov and Andrew R. Harvey
Opt. Express 18(17) 18035-18040 (2010)