I did a quick check of the jitter cleaning, and it looks like we indeed have more jitter peaks right now than we used to (before the vent).  Since it looked like some could be subtracted out, I updated the jitter cleaning coefficients (training on last nights nice long lock).

In the attached figure, the bottom panel shows the cleaning improvement (red is lower than blue, which is good) that we had during last night's long lock (and we've had since the vent).  These are the same cleaning coefficients that we've been using for a long time.  The data in this lower panel is from 10+ hours into a lock/

The top panel shows the cleaning improvement (red lower than blue is good) after I updated the cleaning coefficients.  This data is from about 1.5 hours into a lock, but the coefficents were trained on the data from the lower panel, so its possible it will get a bit better with time.

It didn't do as much as I'd hoped, which maybe means that some of the jitter is not well witnessed by the one IMC WFS diode (pit and yaw of WFS A) that we have in the subtraction system.