 |
A theoretical model (Schatzel, K. Quantum Opt. 1990, 2, 287) for the covariance of real quasielastic light-scattering correlation functions has been experimentally verified. The variance calculated from this model was found to be comparable to the variance obtained from a set of experimental data runs while requiring only one or a few runs. The covariance was also verified by direct comparison to the averaged covariance calculated directly from the set of data. The effects of pinhole size upon the covariance were also studied. The nondiagonal covariance was approximately independent of pinhole diameter, while the variance was inversely proportional to the pinhole diameter. Thus, relative noise reduction was verified, and criteria for a suitable compromise between the pinhole diameter and signal-to-background noise ratio was established. This criteria suggests that the optimal pinhole diameter is such that six-ten coherence areas of light are collected. Multiple scattering solutions, with transmissionsŤ40%, were found to have a negligible effect upon this model. The effects of correlation on the error in the fitted parameters were investigated and found to be significant in some cases. Finally, a Mathematica implementation of the covariance into least-squares fitting to account for correlated noise was implemented. The effects of correlated noise on both the fit and the errors in the fit are analyzed. The results are compared to several standard fitting routines.
|
|
For the newest resources, visit Wolfram Repositories and Archives »
