Fig. 17

(a) Shows the reconstructed PSF, following Eq. 40, from the Pearson correlation coefficients depicted in Fig. 13a. The distribution implie that only 18 out of 1000 photons are detected within the intended pixel, which is unlikely given the good contrasts measured with the detector. (b) Shows the Pearson correlation coefficients \(\rho _{m,n}\) as line profiles with \(m=0\) for the horizontal profile and \(n=0\) for the vertical profile. The red and orange dotted lines represent the offsetted Gaussian surface fit, with which the beam correlation is determined. The fit parameters of the surface fit are provided within the graph along with an inset, showing all the Pearson coefficients of the signal on the detector logarithmically scaled for visibility. The red and yellow lines on the inset represent the direction the respective line profiles were taken. Subtracting the beam correlation before reconstructing the PSF leads to (c) the detector PSF. In comparison to (a) the probability of photons being measured in the intended pixel is significantly increased, making it more reasonable. (d) Shows the change in the measured smoothed gain (above) and k-value (below), when cropping the acquired frames according to the diagonal binning Fig. 15d but without the actual binning. It can be seen that the smoothed gain increases for ‘binning values’ that discard most of the detector, contrasting Fig. 15c. Further, it can be seen that independently of the gain, the k-value decreases for the same image section.