Introduction

After getting preliminary SMD gains using a slope analysis and preliminary pre and post shower tile gains using the MIP id I went on to try to fit the towers.  The requirements were:

1. MIP signal in SMD planes with 7 empty strips on either side
2. MIP signal in pre and post shower layers with energy [0.5,3.0]*expected from preliminary fits.

I then fit the tower MIP signals using a Gaussian, using the fitted mean to calculate gains. 

Results

Attached is a zip file with the MIP fits for each tower.  Also attached is a pdf which shows the ratio of these gains to those in 2006. 

Issues

I need to double-check the mask files I used for these tower fits and those fits that were bad. 

Right now the gain is coming out systematically lower than ideal as shown in the Figure below. (Points are zero were bad fits or masked out).

To get these gains I used the same MIP dEdx and sampling fraciton used in the past.  If I change these to the values uses in the EEMC slow simulator, this will make up much of the difference, but doesn't solve the problem.  It appears that the gains have dropped more than expected from the high voltage adjustment that was made in 07 or 08.

Should I be using the means of the distributions instead?  Playing with the fit to try to grab the tails that are there (this should push the fit mean up)?