Things are not as bad as they first appear. Step-scan measurements often involve sample modulation, such as repetitive stretching of a polymer film. A second level of demodulation is then required to extract the spectral response from the detector signal. Much of the noise in the single-beam spectrum is rejected by the second demodulation, resulting in performance more competitive with rapid-scan.

Consider a difference spectrum of a polymer in a stretched vs. relaxed state. To perform this measurement by rapid-scan, it would be necessary to subtract a spectrum of one state from a spectrum of the other. For a 1 hour averaging time for each of these two spectra, the difference spectrum would have a noise level on the order of 10 to 30 microabsorbance units. Any time dependence of the spectral difference would be lost. A 2 hour step-scan measurement will produce difference spectra with a comparable noise level, but retaining the time-dependence of the spectral changes.