Since Dave needed us to go out of Observe for a few minutes to clear some memory so the CW hardware injections wouldn't stop overnight, I thought about changing the *way* in which the NonSENS noise estimate is turned off.  However, I then realized that I wasn't 100% sure that it would work, so I backed out my changes.  However, I didn't get the guardian code reverted and reloaded before we went back to Observing.  So, I have asked the operators (currently Camilla and Oli) to reload the NOISE_CLEAN guardian next time we're out of Observing. 

I had been thinking about whether there was a nice way to have the output of the nonsens noise estimate be saved, without actually sending a non-zero signal to the calibration pipeline, since currently the noise subtraction is turned off (and has been since we lowered our laser power).  I had thought about doing this by turning off the output of the filter bank, and having the gain be non-zero.  But, then I realized that I'm not 100% sure where the NOISE_EST_DBL_DQ is saved from, and if doing that would have made the NOISE_EST_DBL_DQ become non-zero, which would then effectively have the subtraction be on (which I don't want).  Since I was trying to be speedy, I just backed out my change, and will spend some more time thinking about how I want to do this. 

However, since I had set the NOISE_CLEAN guardian to "turn off" the noise estimate in the way I wasn't sure would work and loaded it, we now need to reload the old version of the NOISE_CLEAN guardian.  The guardian is back and ready, but Dave was done and we went back to Observe before I loaded the NOISE_CLEAN guardian.  If we do not, then next time we lock the operator will see SDF diffs on the OAF model.  Please do not accept those OAF SDF diffs. 

If we get to NomLowNoise and there is an SDF diff of OAF-NOISE_WHITENING_GAIN = 1, then reload the NOISE_CLEAN guardian, request it to DOWN, then request it to SUBTRACTING_NOISE.  This should have the output switch be ON, and the GAIN = 0.