Tower By Tower Pion Peaks

Calibration

I am trying to calibrate the BEMC using neutral pions.  As of now I am only using the L2 Gamma trigger for the second longitudinal running period, although I plan on incorporating the transverse data as well since the calibration should be the same for the two running periods.  The procedure, so far, is listed below.  Items 1 - 4 have been done for the long2 data set and are not difficult.  Items 5+ I have not done yet, nor am I certain if they are the correct way to proceed.

1.  Find all pion candidates using Frank's finder.
2.  Calculate the mass for each di-photon pair.
3.  Find the tower struck by each photon in the candidate and fill that tower's histogram with the invariant mass of the pion candidate.  Note that if the two photons strike two different towers, both towers are filled with the mass of one pion.
4.  Fit the towers with a gaussian + linear function and extract the location of the peak (see below.)

* * *

5.  Using the mass peaks, calculation a correction factor for each tower, and apply this correction factor to the tower response (gains? I'm not sure.)
6.  Repeat until desired level of precision is reached.

Plots

Good  Tower:

Poor Statistics (Hope to improve w/ Transverse):

Poor Fit:

Spiked Fit  (Chi^2 = 60/32 if you can't read it.)

A .pdf with all 4800 towers, and their fits can be found here.  It is about 50 MB.