Crosstalk correction

Since a signal in one detector segment may induce a (much smaller) signal in adjacent segments, the first step is to remove this error by applying a crosstalk matrix:

$$\vec S_{\mbox{corrected}} = \mathsf{C} \cdot \vec S$$ (10.3)

where $\vec S$ is the measured signal vector (signal of all detector elements for one scan pixel), $\vec S_{\mbox{corrected}}$ is the signal with the crosstalk contribution removed, and $\mathsf{C}$ is a $n \times n$ crosstalk matrix ($n$ being the number of segments). The crosstalk matrix is stored in sm_par as sidet_crosstalk_matrix. It has as many elements as the number of ADC channels squared, but only the first $n^2$ are significant.