The Mueller optical calculus is used to model common spectrometric experimental schemes. The effects of source and preamplifier noise on the precision of the measurement are investigated, and the effect of gate width variation for gated integration of Gaussian signal peaks in the presence of white or 1/f type noise is evaluated. The results indicate that, for integration of a transient absorbance signal, the choice of gate width depends upon the type of noise, but not its source. For a transmittance (or emission) type of experiment, both the type and the source of the noise determine the gate width to be used for integration of the signal.
