We examined two flare events. One of the flares saturated the ADC electronics, so we developed a method to extrapolate the intensity of the flare from the diffraction pattern, which produced a result with an uncertainty of the mean of 4.2%. We checked this method on the second flare which did not saturate the CCD, and this flare confirmed the accuracy of the method to within 2.05%. This method effectively increases the intensity range of the camera by at least a factor of 40; even brighter flares than we have observed can most likely also be measured by the same method.
In addition, we conclude that the separation of the peaks is consistent with a dispersion effect.