Can we simulate a mechanism generating a 2-peaked S1?

[pdeperio]

One of ER outliers from the waveform watching campaign:

Passes lax v0.5.2 cuts, so may be cut by latest lax version to be checked later.

[tunnell]

Two S1s or AP?

[feigaodm]

Looks like a dark count coincidence with a good S1. By looking at the WF, the spike is not an AP following a real S1 signal. Maybe it's only a feature in Rn220 calibration where event rate (and junk/lone hit) is too high.

[feigaodm]

After looking at more and more WFs, I think this can be due to statistical fluctuations.

[JelleAalbers]

The waveform has a lot of dark hits, so a coincidence with a dark hit seems reasonable. Afterpulses are always possible. Finally we have secondary spikes in our PE waveforms at a similar location (https://xe1t-wiki.lngs.infn.it/doku.php?id=xenon:xenon1t:aalbers:double_pe_emission#updatesource_of_the_discrepancy_in_mean_gain), though they are not as strong

In any case, this seems like a research question rather than a pax issue. Would you agree?