Lately we've been having better range at the beginnings of our locks.  I wondered if that was due to the ADS handing off to the camera servos early in the lock, whereas we'd set the ADS setpoints after the IFO was thermalized. Spoiler alert: I don't think this explains things.

I turned on the ADS lines for a little while, while Vicky and Naoki were working on some sqz things ('Lines on' time in first attachment).  I also turned on the ITMY dither line (even though we'd been setting the ITMY A2L values lately based on CHARD dithers, not ITMY dithers). 

When Vicky and Naoki were done, I turned on the ADS servos to converge their error signals closer to zero, then turned back on the camera servo (which resets the camera setpoints to the current values).

The second attachment shows the camera centroids over the same time span (and the setpoints in white, which change after the camera servo is cycled on).  I cycled the camera servo on twice, which is why the setpoint is changed twice in this screenshot. The ADS servos have been on and converged around the -20min mark on this x-axis, but there's not a very major change in the spot centroid as seen by the cameras. 

I don't think there was a very noticeable change to the sensitivity.  Here are 2 spectra, each 30 avgs.  Yellow is before I turned on the ADS servos (but the ADS lines were on), starting at 22:53:20 UTC.  Red is after the ADS servos have converged, starting at 23:00:29 UTC. 

I'm not really convinced that this has anything to do with the change in sensitivity over the first few minutes of each lock, so I'm not going to make any changes to any locking sequence things based on this.  We just lost lock due to an earthquake, so the IFO will relock with all its normal process next time around.