GLAST Calorimeter Software Weekly Report
Week of 17 May 2004

NRL

Sasha:  Last week I was working on 3 topics:
1) calibration algorithms for I&T: moving light asymmetry calibration to calibGenCAL v2r1 package from my private version;
2) corrected pedestal fitting: problem of coarse binning in LEX1 range. After next round of discussion with Pol it looks like there is uncorelated noise in LEX1 ~ 0.5 adc units (the same as correlated part) but correlated noise ratio of LEX8/LEX1 is consistent with 8 (ratio for big signals);
3) simulation of electronics calibration with new Am source (20x36 cm size) - by request of Eric Grove: he plans to buy a new source.

Zach:  Continued to work on translating IDL charge injection analysis routines to Root.

Andrey:  Past week I started working on the light attenuation mapping.  I am in process of getting high-energy muon runs (in fits format) from SLAC. Unfortunately it is a pretty slow thing.  I got so far about 25 files out of 70.  Plus, I found out that not all the files mentioned by Xin in his talk link, are there. I wasn't able to locate about 1/3 of all the wanted files. I should probably ask Xin about where they might be.

I downloaded all the necessary software (like new GlastRelease, calibGenCAL, vcmt v18 and so) on my PC.  Still I haven't installed yet a new compiler. Will do it soon, since I am backing up some of my older stuff, that was under older VC.NET.

Mark:  Continued to look into geometry issues.  Some densities have changed.  Will compute new values.  Succeeded in getting GR v4r2 to work (finally).

LLR

Pol:  no report

Berrie:  Nothing new to report this week

CENBG

Thierry, Benoit, Johan: The GSI-analysis activity was still focused on the quenching effect studied at 1 GeV/nucleon and on investigating the reason for the apparent loss of events between the FRS and the EM in some of the runs.

Two spring students continued to work on the analysis of the CERN experiment and the implementation of a trajectory-selection algorithm in GlastRelease.

GAM

Frederic: Last week I started a systematic study of these GSI events which are tagged with a good Z value by MUSIC but are not seen by the EM (deposited energy below roughly 300 MeV in the first layer). The proportion of good events is essentially stable around 90%, which is reasonable, but it can be lower, for instance for obvious reasons (MiniCal in front of EM, etc...), but also for 2 subsets of consecutive runs (65% for 9 28_Si runs, 60% or below for 3 cocktail runs at 1700MeV/n, including run 46 which was previously extensively studied).   This problem seems not to depend on the EM settings but rather on the beam conditions. I am now looking at the 2 above subsets in much detail to understand better these "lost" events.

SLAC

No report

UCSC

While working on low energy recon, Bill found an apparent inconsistency in how zero suppression is modeled in CALdigi and CALrecon.  In the former, if either end of a crystal is above thold, the crystal is read out.  In the latter, this happens only if both ends are above thold.  [It turns out that this was to make the coding easier in the event of on xtal end producing zero signal;  Sasha will modify CALrecon to make it consistent].