MuRaVes Technical Meeting

Name: MuRaVes Technical Meeting
Start: 2026-04-14T14:00:00+02:00
End: 2026-04-14T15:35:00+02:00
Location: No location set

Tuesday 14 Apr 2026, 14:00 → 15:35 Europe/Rome

Description

The videoconference link for the ZOOM connection is:

ID: 938 3887 9891

passcode: 639525

General Matters:

Hide

Alice at VUB on fridays.
Meeting to split future tasks soon.
TO DO (Alice):
- From 19 to 21 of May, Gabor is in Naples.
- Ask Giovanni his availability for a visit to the INGV.
- Send a reminder to G&G about the visit to Naples and how to organise MURAVES visit (Gabor doesn't have the car this time).
- Set a meeting to discuss things to do on site.

Adithyan and Samip on digitalisation:

From simulation (Dora's) we see that normally muons intercept two bars. This is why currently the ADCtoPHE conversion is computed only using a subset of data with selects clusters of size 2 (2 bars get hit).
Slides show that the mean energy deposit between strips and cluster is reasonable: for 2 bars clusters the cluster energy is about double the strips energy.
For now they will keep fitting both Strip Energy and Cluster Energy to get the conversion.
The Landau in the fit on the cluster energy doesn't work very well, maybe try a Gaussian around the peak or Gaussian with asymmetric tails? Adithyan: Gauss seems to fit better.
Presentation at the general meeting.

Final remark: 4th plane has much more single bar clusters with respect to the plane 1,2 and 3. Alice: Could this mean that the most of the multistrip cluster are low energetic muons that don't pass the led block? If simulation tells that muon mainly generate clusters of 2 bars, are those single bar cluster mainly noise/bkg? Hopefully simulation will help in understand this better.

Gabor: Working on pedestal and 1phe value (Using Gaussian fits didn't work out -> back to manual)

Dora: Needs a final input from G&G for the rest simulation's ready.

Alice:

pythonisation of reconstruction script is finished. V0.3.0 of reconstruction software available and encouraged to be used. Found another bug in the cpp code (tracking.cc) that was the reason of the differences in the displacement variables. Previous version of the repository affected by these bug are noted.
Run time of the python version is longer than cpp version ( ~50s vs ~30s) due to writing the output as root file. Same run time if using pkl or json outputs instead.
NB: The reason is due to the fact that there are branches that are list of list, making conversion particularly slow in python. I believe this is not even necessary.

There are minutes attached to this event. Show them.