we lost spatial overlap on several instances.

It seems that this happens when SASE2 changes number of pulses.

when we have SASE2 changing number of pulses, spatial overlap is lost and this goes with a change in the position read by the xgm in the order of ~20um (e.g. attachment 1).

Yet, the XGM position reading is not really representative of the effective pointing changes: a drift of that order, occurring when there is no change in SASE2, does *NOT* correspond to a loss in the overlap (seen in the N2 signal in our experiment).

in summary: some evidence suggests that SASE2 number of pulses changing affects our FEL beam pointing. Yet, we don't have a diagnostics allowing us to quantify this.