Abstract
This paper contains an analysis and discussion of the factors affecting the accuracy with which spectral parameters can be determined in photon-correlation spectroscopy on Gaussian-Lorentzian light using either direct, homodyne or heterodyne detection. Both weak and strong signals are considered; the dominant noise sources being photodetection and intensity-fluctuation noise respectively. Where possible theoretical analysis is compared with computer simulation. Computer simulation also enables the behaviour, under those conditions which are too complex for direct analysis, to be studied. The effects of different normalization methods is considered; existing analysis on the extraction of spectral linewidth is extended; and the accuracy with which frequency shifts can be measured in heterodyne detection is investigated. The findings are summarized in the form of operating guide-lines for obtaining the greatest accuracy in the determination of the spectral parameters.