G4SiPM Integration

[c-dilks]

There is some interest in the dRICH group to use G4SiPM for our photosensors. Currently Ajit is working on https://eicweb.phy.anl.gov/EIC/detectors/ecce/-/merge_requests/28 (to be moved to Github), which adds new material properties from a model in G4SiPM.

We're wondering if it's possible to instead use this a a DD4hep plugin. If I understand correctly, having access to G4SiPM would be nice for more control, be more realistic, and allow us to do noise studies.

I'm neither familiar with DD4hep plugins, nor this G4SiPM package, but after staring at some ACTS plugin code, I suppose we would have to write some glue code. Alternatively, maybe a JANA2 (EICrecon) plugin, but after some discussion in the last dRICH meeting there seems to be preference for having things like noise models upstream of any reconstruction.

Any thoughts or ideas how we could integrate this package?

[wdconinc]

That's an interesting question. I know the DD4hep devs are open to people posting questions through their issue tracker on github. I am not quite sure how integration would look like, though. What does this look like at a high level in xml to you? I'm having trouble seeing how this integrates into the DD4hep model of readouts, but on the other hand having a modular unit that says 'read this out with SiPMs' seems useful (s/SiPM/ACLGAD/, s/SiPM/LAPPD/) and would make sure different groups don't have to reimplement the same readout approaches...

[c-dilks]

From an XML file, I better let Ajit comment since he's spent some time working with this package. I suppose we would need to produce or use a properties file; see also here for more properties. We also have to choose which SiPM model to use. If this has common value for more than just the dRICH, perhaps it's better to bring this up to the DD4hep devs.

[ajitkmaurya]

I don't get "What does this look like at a high level in xml to you?". I guess we can add properties manually to the XML file. I am not familiar with integrating one package with other. What I did was copy some of the sensor properties from G4SiPM and add them to the XML file.

[c-dilks]

The idea of XML files in DD4hep is to put the main constants/numbers there, basically a configuration file, and leave the actual implementation in C++ files (DD4hep plugins). For G4SiPM, we have these property files for configuration. Actually, I think it would make more sense to add a node in our compact/drich.xml file that points to a particular G4SiPM properties file; that way there's no need to do any additional XML parsing (of SiPM parameters in an XML file), and we configure G4SiPM "as is". Of course we would need to add G4SiPM as a DD4hep plugin first.... (see below)

[c-dilks]

Is there some way to add a new SensitiveDetector via a plugin, following existing DDG4 plugins? I'm not sure where to start, but I see some of the necessary ingredients in G4SiPM such as hits and the SensitiveDetector itself. Perhaps we should ask the DD4hep developers for some advice.