¹

Pulse heights can differ, for example, between different particle types. The timing dependence on pulse heights will also change if discriminator thresholds are changed.

²

Replacement of a digital delay box can affect timing by tens of nanoseconds.

³

If you plan to modify the code, please take a look at how the Makefile works. You can use make commands to copy new HMS code into SOS code and vice versa.

4 An appropriate run is one that has focal plane distributions that are representative of your data, and that is reasonably large, say about 100k events (but the arrays

in the code max out at 199,999; analyze fewer events than that, or modify the code)
5   The analyzer assumes that you have a run of pure beta =1  particles. If the spectrometer also sees slower particles, you need to put appropriate
cuts in the analyzer routine h_dump_tof.f. If you are doing a calibration on particles with ( beta <1 ), you'll also need to edit that routine so
that it dumps out a value of  beta  calculated from the spectrometer momentum for each event.
 

6  To check quality of tracking, run the kumacs hdist and hres

on the histogram file [run].hbook; the first should show fairly flat drift distance distributions from 0 to 0.5 cm,
and the second should show narrow, single-peaked residuals. If this is not the case, redo the tracking.

7  Reasonable values for the fit to beta  are sigma _beta ~ 0.02  (0.02) for the HMS (SOS)