Updates to EEMC Simulator

Fast Simulator:

a) Sampling fraction update from 5% -> 4.8% based on Ilya's study

b) Change default mUseFullTower=true to create all ETOW hits (even if ADC=0)

Slow Simulator:

Recently there have been discussions on a few lists about needs for adding the EEMC slow simulator to the BFC, when using timestamps from real running conditions.  The two main reasons are:

1) Add real pedestals to ETOW (needed for PPV to accurately match tracks to endcap towers)   :  HN thread
2) Use realistic gains to convert GEANT energy -> ADC (needed for trigger filtering in the BFC) :  HN thread

In order to set specific slow simulator options when used in the BFC a flag was created (mIsBFC) to set the specific options desired when the slow simulator is used in the BFC, namely:

a) Add pedestals from DB to all channels

b) Smear pedestal by gaussian width of the pedestal from the DB for all channels

c) Set ADC = 0 for channels with: fail bit  != 0 || ped < 0 || pedSigma < 0 || gain <=0  (sanity check on DB tables)

A few other minor modifications that were made are:

- Truncate pedestal smearing at 3 sigma by default

- Apply ADC saturation and assure ADC is non-negative for ESMD/PRS (already implemented for ETOW)

- Based on findings by the geometry "bug catchers", set the preshower layer thickness to 4.75 mm instead of 5 mm.

- Default for MuDst analysis chain is to disable ETOW slow simulator since energy deposits are not stored in MuDst for towers.  To 'redo' the slow simulator for the towers one needs to use StEvent files from simulation production.

The modified source code is checked into CVS here.

Finally a BFC option ("EEss") was created so that the slow simulator can now be included in the default BFC.   To verify that the slow simulator is indeed in the BFC check the log file for this line

StEEmcSlowMaker:INFO  - mIsEmbeddingMode=0, mIsBFC=1

If you want to run the fast simulator only in the BFC the old chain option ("EEfs") can still be used. 

Here are a few things to keep in mind when using simulation produced with the EEMC slow simulator in the BFC:

• Pedestals and Gains from the DB are used to convert GEANT energy -> ADC in the MuDst, thus the same gains must be used in the analysis chain to yield the correct reconstructed enegy.
  • In order to assure that the same DB tables are used in the analysis chain one should use a line like dbMk -> SetMaxEntryTime(YYYYMMDD.HHMMSS) where the timedstamp is set to the DBV used in the BFC options.  For example, if this chain option "DBV20100420" were used in the BFC, in order to gaurantee the same DB tables are used in the analysis chain, one should set dbMk -> SetMaxEntryTime(20100420) in the analysis macro.  The "DBV" value can be found in the BFC log file.
  • Since only the ADC is stored in the MuDst for ETOW, this essentially fixes the gains to be used in the analysis chain (unless StEvent is used to get GEANT energy deposit). 
  • Since both ADC and GEANT energy are stored in the MuDst for ESMD/PRS they have more flexibility, ie. the slow simulator can be used to simulate those hits with any DB tables in the MuDst analysis chain.  
• When DB timestamps are selected to be used in the simulation production, check to make sure that the pedestal, status and gain tables are "reasonable" for those timestamps.