Reduce ETMX L3 counts by factor of 4, change ylim to smarter fractional system, adjust search window
In order to adjust for the ETMX L3 DAC change, I look for relevant channels happening after the change date and reduce their counts values by a factor of 4. To represent this normalized version accurately, I've changed the ylims from being in counts to being normalized to the saturation threshold. In addition, I have expanded the saturation search window to 300 seconds, expanding out to 15 minutes if necessary to get the full extent of SUS saturations. Should fix #40 (closed) and #47 (closed).
If accepted, we should rerun on locklosses after October 30th, when the DAC change occurred.