L. Dartez, E. Goetz, J. Kissel, T. Shaffer

After we killed the lock stretch by turning on the extra 16 kHz anti-aliasing filter (LHO:82412) -- we found that the OMC wouldn't lock after 4 to 5 attempts during the next lock acquisition attempt. We turned on the 16 kHz AA filter early in the lock acquisition sequence "it would just work" and really -- thinking the cause of the lock loss was the loss of phase margin in the DARM loop. 

However after seeing the abnormally large number of failed OMC attempts, we suspected the new filter again. 
Then we realized -- the extra AA filter will cause phase loss at the 4.19 kHz OMC LSC dither line, and thus the OMC LSC DEMOD needs to be rephased to account for it.

The magnitude / phase of the 65k to 16kHz AA filter at the 4.19e3 Hz dither line frequency is 1.0837 [ct/ct] / -77.2 [deg].

As such, we adjusted the H1:OMC-LSC_PHASEROT from 56 deg to 56 [deg] - 77 [deg] = -21 [deg], and the OMC "locked right up."
(We assume that an 8% increase in the OMC LSC dither line signal would only be *good* for the OMC locking, but this will be confirmed with others later.)

We are now in Nominal Low Noise with the extra AA filter on, and have been for 8 minutes (and the many minutes prior that it takes to get from transitioning to DC READOUT and getting to Nominal Low Noise).

I've accepted the two changes in the safe and OBSERVE SDFs for the h1omc and h1iopomc0 front-end models.