Cross talk simulation

[awhitbeck]

This should be done on top of the new digi code.

Need a way to generate hits from optical cross talk in the scintillators/simps. This can be done using an NxN (where N is the number of bars in an array) matrix that tells how much light from 1 bar generally leads into another. I propose that we do this on a per hit level so that the time structure of these signals are maintained.

[awhitbeck]

first attempt at an implementation is here. The code compiles and runs. Testings now....

[awhitbeck]

I did manage to test this code; it seems to work.

I am still debating how the user-interface for this code should work, specifically, how the should cross-talk matrix be configured. The two extreme options are:

• Option 1: to assume that there are only 2 free parameters, one which determines the cross talk for nearest neighbors, another for next-to-nearest neighbors. I think this is sufficient for now and certainly routed a physically intuitive hypothesis: optical cross-talk is largest for bars that are physically near each other.
• Option 2: users pass the full NxN matrix, where N is the number of channels. This is likely better suited to a database setup -- which is a development we'll need down the line. This would allow for a full calibration of these parameters to be adopted in simulation and potentially used in reconstruction.