Tamara
has found some runs for which the beam straddles the modules, and she
now sees an asymmetry effect in the module response as well.
She showed a couple of plots (150 GeV and 200 GeV) pions at y=0, x=-160.
They were not exactly the same plots as MPI had produced, but
they demonstrate a pronounced asymmetry.
Tayfun
had contacted Naoko about her new weights and created some
Ntuples following her instructions. He was very impressed by
Naoko's work and her instructions. While running the code,
he found that some cells had a `max. floating point number'.
Naoko said that there are 3 cases for which this could happen:
- The time samples might not be valid
- The range of time samples for the fit might not be correct
- global timing might not have been calculated due to low signal.
Particular problem cells are:
271, 496 (no response)
497, 498 (odd pulse-shape)
465, 490 (large time-shift - maybe wrong)
A huge THANK YOU! goes to Naoko for all her work and her help!!!
Tayfun then showed some plots that compared results obtained
with the old weights and the new weights.
He plotted: (beam position in beam chamber)-(beam position in 2nd EMEC layer)
and finds that the distribution has a shoulder when using the old weights,
and no shoulder with the new weights. The width of the peak
of the distributions is approximately the same for both cases (about 0.083cm).
He also produced scatter plots of the energy in EMEC layer1 vs. layers two
and finds that these distributions look a lot cleaner with the new weights.
The same can be said for the distributions when he plots the
energy in the HEC vs. the energy in the EMEC.
The plots be found at this
web page.
Margret
mentioned that she contacted Pavol about the G4 Ntuples and received
valuable replies and help.
All the G4 Ntuple energies are indeed in MeV. The fact that the EMEC
absorber energy wasn't filled was due to a small bug in the program
and has now been fixed. There is an option to turn on charge collection
in the EMEC, but Pavol strongly advises against it, as that has not
been tested yet and is not really supported.
Tayfun has tested the converted G4 Ntuples and gave Margret some feedback
on it. The essential things are accessible in TBRootAna now. Testbeam Ntuple
entries that make no sense in G4 are filled with zeros.
In addition to the usual Ntuple trees there is an additional Ntuple tree (hg4)
in the converted G4 Ntuple file that contains G4 specific information
with information that the regular testbeam Ntuple does not carry or carries
in a different format. (Like number of hits in the F1,F2 counters, x and y hits
in the beam chambers, total energies in EMEC and HEC LAr and absorbers). This
tree will likely be expanded in the future.
Margret thinks she can do a quick-fix to simulate the HV problems in Geant.
Tamara is to supply Margret with a little wish-list (Ebeam, particle-type, xcryo, ytable
and list of cells with reduced signal due to HV problems)
so Margret can generate some events for her and then we can see whether
we can see an asymmetry in the Geant events.
Richard
reminded everybody that computer problems might be looming ahead, as all Linux
systems will have to undergo security upgrades.