FT-IR spectra recorded through an absorption cell almost always show an undulating pattern due to interference fringes. This pattern is also sometimes called “channel spectra”. These fringes can arise from multiple-reflections within the cell windows, or possibly even reflections in the mirror ovecoatings. The amplitude of the fringes in a single-beam spectrum is sometimes considerably greater than the amplitude of the noise in the spectrum. When an absorbance spectrum is made from the sample and background spectra, the fringes usually cancel, so that only the noise becomes the limitation on detection sensitivity. However, if there have been physical changes in the apparatus between the recording of the background spectrum and the recording of the sample spectrum, the fringes may be shifted and therefore may not cancel fully. In that case the fringes may interfere with the measurement of small absorbance values. There may be a fringe shift between a cell filled with sample and the same cell evacuated. However there is rarely any shift between the cell filled with sample and the cell filled with nitrogen. To minimize fringe problems when analyzing air, one should make the background spectrum when the cell is filled with nitrogen.