sPHENIX discussion of software

[Anthony Frawley <afrawley AT fsu.edu>]

Hello,

Joseph Bertaux has done a nice job of putting together local to global transforms for the INTT based on the survey data. He and I have been working together on getting the survey alignment parameters in a form that we can use in offline.

But when we compared the transforms from the survey to the transforms used in simulation, we found that the phi positions of the ladders in simulation are different by half of a ladder spacing relative to the actual detector. We identified how this happens on the software side:

In coresoftware, the default ladder phi offsets are correct:

https://github.com/sPHENIX-Collaboration/coresoftware/blob/0d893f314c83936bdcd5ced628997992d80fa946/simulation/g4simulation/g4intt/PHG4InttSubsystem.cc#L191

(notice the nonzero offsets are for layers 3 and 5)

However the values in G4_TrkrVariables.C are incorrect, and they override the correct values from PHG4InttSubsystem.cc:

https://github.com/sPHENIX-Collaboration/macros/blob/7d8c9b53784694a9b95c2200cb2db19c6745494e/common/G4_TrkrVariables.C#L80C5-L80C5

(notice the nonzero offsets are for layers 4 and 6)

This does not cause problems when analyzing simulated data, since the offline code uses the same G4 geometry as the simulation does. However it does mean that:

1)  We will have to change the G4 geometry before we analyze real data, because the offline code currently thinks the INTT ladders are not where they really are.

2) Fixing the INTT geometry will break analyzing the existing simulations.

We should discuss how to deal with this at the next software meeting.

Tony