HDGeant4 Meeting, August 28, 2018

From GlueXWiki
Jump to: navigation, search

HDGeant4 Meeting
Tuesday, August 28, 2018
2:00 pm
JLab: CEBAF Center, A110
BlueJeans: 968592007


  1. Review of notes from last time
  2. Issues on GitHub (all)
  3. pi0/eta -> 2g masses (Sean)



  • FSU: Sean Dobbs
  • JLab: Thomas Britton, Mark Ito (chair), Simon Taylor, Beni Zihlmann
  • UConn: Richard Jones
  • W&M: Justin Stevens

There is a recording of this meeting on the BlueJeans site. Use your JLab credentials.


Thomas reported a bug in mcsmear that has recently been identified and fixed. See Pull Request #9 for halld_sim. The BCAL smearing parameters were being used uninitialized. That resulted in random behavior, e. g., varying widths in energy resolution from run to run of the mcsmear executable. Recent simulations generated for comparison between HDG3 and HDG4 may have the disease.

Review of notes from last time

We went over the notes.

  • The JANA_CALIB_CONTEXT-not-set warning that was recently added appears to be warning whether JANA_CALIB_CONTEXT is set or not.
  • The change in the treatment of the time-to-distance relation discussed last time has been implemented on a branch of halld_sim. Thomas and Simon are testing it. Once it has been merged to the master, Richard will have to re-implement the scheme in HDG4.

Issues on GitHub

We reviewed the list of issues. Justin entered a new one last Friday where optical photons have inconsistent propagation distance vs. propagation time depending on whether or not they cross certain material boundaries. Richard will be looking into this one.

pi0/eta -> 2g masses

Sean showed plots showing a slight shift (5 MeV) in the location of the two photon invariant mass for neutral pions and eta's when comparing HDG3 with HDG4. The origin is unclear at present.

We agreed that, in principle, problems like these should not be solved by smearing parameters in mcsmear that depend on whether the simulated data is generated by HDG3 or by HDG4.