Once the cloud sizes and cloud centers are known, the final calculations
are fairly common for this kind of model. The cloud is split between adjacent
pixels assuming gaussian distribution of charge density in the cloud. This
implies a very simple routine for evaluation of the signal amplitude in
pixels adjacent to the cloud center. After that readout noise with a gaussian
distribution is added to each pixel and a gain factor is introduced to
adjust the amplitude of the signal to that of the paticular device that
is being modeled. The last stage is an event finding routine - a procedure
similar to the one used in data analysis software to determine whether
the amplitude of an event is great enough to be included into an eventlist.
The output of the program is a standard eventlist which is entirely compatible
with numerous software tools available for the real data analysis. In Figure
4.131
are shown standard grade histograms for experimental data taken under monochromatic
illumination at 4510 eV and the histograms of simulated data for the same
energy. While there are some small discrepancies, all the important features
of the response, such as escape and fluorescent peaks, low energy peak,
low energy tail, peculiar shape of the low energy tail of the horizontaly
split events, are reflected in the model, and agreement between the model
and the data in general is good.
Figure 4.131: Graded histograms comparing
observed data & simulation at 4510 eV Standard graded histograms comparing
experimental data with simulated data. Data collected for a monochromatic
source at 4510 eV. (top) Data; (bottom) Simulation
