PEM Face Response

Henri implemented the PEM face response function calculated from the test beam. I just reposited the code as a static function in the Electron package. This can be used both by the simulation to get the response right, and by analysis jobs to correct the electron energy.

The next thing to do is convert the simulation code to use this function. (Now it uses the equivalent code local in the Gflash package.)

• Before any changes, I got the following plots from Henri's script:
  • 4.4.0int2
• Made the following changes:
  • GflashCalResponse.cc was changed to use the new function from the Electron package rather than GflashPEMFaceCorrector.
  • A few bugs in the face response code were fixed, especially to use some new information about which towers have mirror-symmetric geometry relative to the main tower we use for the map.
  • An apparent bug in the extrapolation (passing the Bfield in kG when it was expected in T) was fixed.
• After those changes:
  • 4.4.0int2MOD
  • Looks good. The change from two small red blotches to one big red blotch is expected since this is a type 5 tower and we are now modeling the fact that odd and even towers have mirror-symmetric geometries (in phi). So the neighbor towers have their peaks on the opposite side in phi.
• Mar 26, 2002: Finally, after updating to 4.4.0int6, and adding GflashSim.cc from the head (this includes Yoshiaki's changes for PPR simulation):
  • 4.4.0int6PPR
  • Well, the relative response part looks OK, so I'll reposit that, but as you can see the overall response is much higher (?) with the new version. I expect this has to do with the PPR changes, and I'll have to look at how the PEM tuning is affected.
    • Mar 29, 2002: Actually, this is an obsolescence effect. The accessor for EM energy which used to include just PEM energy now adds in the PPR energy, so it's OK that the scale is higher now. In fact, it's closer to unity in the center of the tower now, so that's good.