Since it's been such a struggle to get jitter subtraction and laser noise subtraction to work simultaneously, I regenerated the c-code and simulink model that goes into the h1oaf model and Dave rebooted it on Tuesday.

During a brief out-of-observing segment this morning, I tried implementing both jitter and laser noise subtraction, and was successful. I'll keep working on this using some new offline testing infrastructure that I'm working on using the CDS group's librts, but I'll likely bring this to the subtraction review committee for their consideration soon. 

Attached is a spectra of the NOLINES channel (blue) versus CLEAN (red), along with the noise estimates (green + yellow = black).  Over almost all of the frequency band (really, everywhere but the 60 Hz line) the subtraction has either had no effect or has made an improvement.

The second attachment is the range plot from the summary pages, showing that the subtraction is indeed making a small improvement to the range (b;ue is above pink, rather than being identical to pink). 

I accepted the SDF diffs for updated subtraction coeffecients, but we are not applying any subtraction in observing (and won't until given the all-clear from the review committee).