Flash times not consistent between data/mc

[bear-is-asleep]

I'm submitting a ...

• bug report
• feature request

Checklist

• I've added at least 1 label under Labels
• Assigned a reviewer under Assignees

Expected Behavior

Flash times should be consistent in data/mc

Current Behavior

Flash times are defined wrt the trigger in data, but wrt the beam in MC

• MC sample : MC2025A Spring25 validation - BNB + Cosmics
• Data sample : Spring25 validation - DevSample

^^mc

^^data

Possible Solution (optional)

How to reproduce

[lynnt20]

In order to make the MC more similar to data, we have to implement a timing corrections module which @VCLanNguyen is currently writing. It takes a bit of special consideration since the trigger has different offsets for various configs, such as zero bias vs. light triggered, and off beam vs. on beam configs will want to use different reference frames.

In the meantime, what you choose to do with the reconstructed timing in data will depend on the trigger, but as long as the opt0 time window is defined to be wide enough, you can do the relevant corrections in the analysis stage and ensure that you keep all the flashes that you want.

[lynnt20]

For all currently decoded PDS data, the time reference is the TDC ETRIG. The TDC ETRIG is not necessarily aligned with the beam gate (it matches PTB HLT 2, which I believe is sent whenever these two conditions are fulfilled: the beam gate acceptance + light above threshold). In other words, TDC ETRIG can "move" with light-arrival w.r.t. the beam gate.

For zero-bias data, ETRIG and gate are lined up. For light-triggered data, ETRIG and gate are not lined up.

To perform a correction, you can obtain the TDC RWM - TDC ETRIG time difference, where the RWM is aligned with the beam gate. Simply add that time difference to the OpFlash time during analysis, and you should be able to replicate the top-hat plot. You can then apply a cut to require light during the beam spill only

Below image is from Spring25 Validation Dev Sample after the above correction.