Fixing time synchronisation for PhotonBacktracker for ProtoDUNE HD MC.

[vpec0]

I have observed the backtracker did not match OpHits to correct true particles and in general it failed to find corresponding true depositions. Looking at the OpFlash times w.r.t. the true times, the flash times came 250 us earlier. This fix is configuring the backtracker to account for that.

[jroto]

Hi Viktor, the backtracker used to work well... Might this delay be related to the pretrigger in the SPE template (either during the digitization or during the deconvolution?

[vpec0]

Hi Viktor, the backtracker used to work well... Might this delay be related to the pretrigger in the SPE template (either during the digitization or during the deconvolution?

@jroto Hi Jose. If you are referring to the deconvolved waveforms, those were not present in my tests. The offset is present in the PDHD 2024 production, which still has the old waveform sims and no deconvolution applied. Also, I note that the offset is -250 MICROseconds! The original FD configuration was +260 NANOseconds. O(100 us) should not be a result of a pretrigger window.

It may be that the signal sims were misconfigured. I may take a look later. But for now, this configuration works with the official production. See the attached plot for DT between recob::OpFlash::Time() and MCTruth generation time - pairing all true cosmic muons with all reco flashes.

[jroto]

This is very strange...
But many things have changed since the last MC production, so I'm not sure if this fix would work with the current configuration...

[jroto]

I think it should be related with this Pretrigger in the digitizer:
https://github.com/DUNE/duneopdet/blob/794b4fef7ae01302a13cd8683dca7f41693e2472/duneopdet/OpticalDetector/opticaldetectormodules_dune.fcl#L33
But this is not used anymore with the new digitizer since November.

[vpec0]

Here's how opdigi was configured for the production:

opdigi: {
   BackTime: 3.86e-1
   CrossTalk: 2e-1
   DarkNoiseRate: 10
   DefaultSimWindow: true
   FrontTime: 1.3e-2
   FullWaveformOutput: false
   InputModules: [
      "PDFastSim",
      "PDFastSimExternal"
   ]
   LineNoiseRMS: 2.6
   MaxAmplitude: 5.94e-2
   Padding: 100
   PeakTime: 2.8e-2
   Pedestal: 1500
   PreTrigger: 100
   PulseLength: 5.2
   ReadoutWindow: 1000
   SPEDataFile: "SPE_DAPHNE2_FBK_2022.dat"
   SSP_LED_DigiTree: false
   TestbenchSinglePE: false
   TimeBegin: 0
   TimeEnd: 1600
   VoltageToADC: 1.515e2
   algo_threshold: {
      ADCThreshold: 10
      DWindow: 10
      Name: "SSP_LED"
      Pedestal: 1500
      PreTrg: 20
      ReadoutWd: 700
   }
   module_type: "OpDetDigitizerDUNE"
}

TBH I don't see how any of this should introduce -250 us offset.

[vpec0]

Another thought on the matter. This 250 us may be related to the TPC pretrigger configuration.

[jroto]

I think so, should be related to that.

[lpaulucc]

Yes, 250 us is the TPC pre-trigger. This is indeed strange...
As a side note, the config you put above, Viktor, for the opdigi of the last production does not use the testbench SPE so this is not related to this.

[aolivier23]

Any updates on reviewing this PR? I can try to merge it Friday morning Chicago time if we can get it approved in time.

[vpec0]

Any updates on reviewing this PR? I can try to merge it Friday morning Chicago time if we can get it approved in time.
@aolivier23

I have converted this to a draft PR.

We don't have concensus if this fix is needed. It is needed for the current MC PDHD production, however, current simulations might not need it. We would need someone to test if the PhotonBacktracker default configuration worked alright with the latest dunesw. And we need to decide how to implement the fix for the former production only.