Abstract
Data from photomultiplier tubes are typically analyzed using either counting or averaging techniques, which are most accurate in the dim and bright signal limits, respectively. A statistical means of adjoining these two techniques is presented by recovering the Poisson parameter from averaged data and relating it to the statistics of binomial counting from Kissick et al. [Anal. Chem. 82, 10129 (2010)]. The point at which binomial photon counting and averaging have equal signal to noise ratios is derived. Adjoining these two techniques generates signal to noise ratios at 87% to approaching 100% of theoretical maximum across the full dynamic range of the photomultiplier tube used. The technique is demonstrated in a second harmonic generation microscope.