Following up on my blog post from last year (Looking into event-by-event T0 in FXT), which looked at Run 19 3.85 GeV FXT, I calibrated here Run 19 4.59 GeV FXT. I again found that Irakli's method had the same flaws:

• For roughly ~10% of events, the EPD west had a higher Max TAC than EPD east and was used for the T0
• For another ~10% of events, the EPD east had a value of 0 that was considered valid and was used for the T0

So I calibrated my EbyE T0 using 20k events from run 20181004. Here are plots of the calibrations using the TPC prompt hit times from only the east side as "ptime":

	EPD east	VPD east	BBC east
Distribution of ptime vs. Max TAC for non-zero ptime and Max TAC
Spread profile with fit (errors on the mean are much smaller)
Fit of ptime peak
Fit of ptime peak after slope correction

The EPD and VPD show the narrowest peaks after correction, while the BBC shows the most complete coverage of having a real signal in each event. So the BBC is used as the final fallback option after EPD and VPD are tried first:

Option 1: 0.01683 - 1.664e-05*epde (covers ~80% of events)
Option 2: 0.02853 - 1.711e-05*vpde (brings up to 92%)
Option 3: 0.02298 - 1.620e-05*bbce (brings up to 99.9%)

-Gene