Figure 1 -- Isolation ratio for single-thrown Monte Carlo events (black single gammas, red single photons) with realistic pT and vertex distributions.  The three figures correspond to gammas and pi0's with 5<pT<6 GeV, 6<pT<8 GeV and 8<pT<12 GeV.  The fit s to a double gaussian w/ constrained mean plus a constant and is only there to give some estimate of the resolution of the photon energy.

Figure 2 -- Rejection vs efficiency curves derived from the above figures.

Observations:

The rejection power at a given efficiency does not degrade as fast vs pT as I would have expected.  Next step is to feed this cut realistic (pythia) events and see what we get.

Figure 3 -- Isolation ratio for single-thrown Monte Carlo events (black single gammas, red single photons) with realistic pT and vertex distributions.  The three figures correspond to gammas and pi0's which reconstruct within eta bins 1-4, 6-8, and 9-12 respectively.

Figure 4 -- Rejection vs efficiency for the above isolation ratios.

Observations...

Well... somewhat unexpected, but energy resolution at large eta leaves alot to be desired... eta bins 5-8 seem to be pretty good.  Perhaps 2 tower clusters would work reasonablywell at large eta and would improve upon the energy sharing situation.

07/27/2008

Probably not surprising that we have problems at large eta.  If we have issues w/ the leakage correction, this is where they should show up. Next try a different correction model.