February 10, 2009 Physics

Teleconference Time: 11:00 EST/10:00 CST

• Phone:
  • +1-800-377-8846 : US
  • +1-888-276-7715 : Canada
  • +1-302-709-8424 : International
  • then enter participant code: 39527048# (remember the "#")

Items for followup from past meetings

• None

Documents to Review

• See past correspondence on BCAL Reconstruction Work page.

Tentative Agenda

1. Progress on reconstruction effort at Regina (Blake)
2. Code understanding and 'standard release' issues
3. Analysis goals
4. Discussion of priority tasks
5. Any other business

Minutes

Attendees: Matt, Ryan, Mihajlo, Claire (Indiana), David L. (JLab), Blake, Zisis (Regina)

1. Progress: Blake reported on his progress. The 'calibration' of the reconstruction code for the 4x6 segmentation geometry is essentially complete. Thrown over reconstructed quantities now are approximately 1 over the entire range of z (length) with the exception of the last 2-3 cm.
2. Code: There was discussion on 'standard releases' of the code and a comment that most instructions seem aimed at developers rather than users. The most recent standard release is the 20090106 one, but this is overdue for updates. In svn look at repos/tags off the top level for the 'stable' tag. Change logs typically explain changes done for standard releases. Example of module checkout out: release/source/library/fcal. David promised to write HOW-TOs for code checkout aimed at users.
3. Analysis:
   1. We need to focus on the DR+PDE of the SiPM plus pulse shape/size that influence our gate width and the 'contamination' from the DR within the gate. Threshold is also affected. BCALMCResponse can be used to generated hits at a given rate, and if a given channel has a valid hit we add the dark pulse to it, or we create a new hit if there was none. We need to evaluate how 'messy' things are.
4. We need to develop a figure of merit (FoM) as function of PDE and DR. The best reference is the work that Mihajlo and Matt did prior to the Calorimetry Review when they looked at eta-pi0 and f1-pi0, thus comparing 4 and 8 photons in the final state. The issues were multiplicity and photon energy (low, high). Understand the chi-square and have flat confidence-level plots.
5. We will need Pythia background and determine the S:N connection to the FoM. There are technical issues related to CPU and storage (see numbers below). Skimming of Pythia events is desired, for example, to keep only events with total neutral energy above 6 GeV. Tools exist for all this. The pi0 and eta factories in HDGEANT are ok. genr8 should be used for phase-space generation of the four-vectors of particles. (Something else would be needed for PWA -- spins). David mentioned that hddm-merge-events can be used to 'mix' events from different files (ie hadronic and e-m mixing).
6. David suggested that we examine the single photon efficiency with the e-m coherent-brem background and pickout the signal photons; this would affect FCAL the most.
7. We need to evaluate cuts on m_pi0 inv mass to suppress combinatorial background.
8. Priorities: We decided to skip the work on the 3x3 geometry at present as this did not seem a viable alternative for the readout as far as the straight forward solution using FM PMTs.
9. Other Business:
   1. David reported that he is working on a semi-parametrical package for charged tracking and neutral handling, in order to quickly extract resolutions. We would like to interact with Blake to tune his package with Blake's work. Goal is to extract efficiency and resolution versus angle and energy.
   2. Mihajlo and Claire reported the time estimates for generating Pythia events at IU cluster are about 8h to simulate 400K events with ~2G of hddm output.
   3. Soon we will be generating a lot of Monte Carlo and book keeping will become important. Zisis mentioned that Richard had a package called OpenShop geared for such work that at some point was installed at Regina too. We will contact him offline to investigate the feasibility of using it.
10. Next Meeting: We decided to meet weekly at 11 EST/10 CST.