For the SR3 P estimator, we had initially developed and installed blend filters (in H1:SUS-SR3_M1_EST_P_FUSION_MEAS_BP and H1:SUS-SR3_M1_EST_P_FUSION_MODL_BP filter banks), named pit_v1, that blended between the OSEMs on-resonance, and the estimator everywhere else (86452). After a bit, Brian Lantz made a pit_v2 that included OSEM damping at two extra frequencies because we were having issues with extra resonances at 0.65 and 0.75 Hz that weren't being damped (86510)(filter comparisons).
Eventually we realized the reason why those two peaks weren't being damped was because we had forgotten to include the model contribution to P from L, and we installed those needed filters(86567). However, we didn't swap the blend filters back to pit_v1 at the time.
So during relocking today (2025/10/13 19:05:00 UTC), I swapped us back to pit_v1. We will run with this until at least tomorrow morning and then verify that we don't see a difference in the damping of those two peaks.
Oli - Thanks for doing this test, I'm looking forward to learning what happens. I think the question here is "how well does it work" and "what do we see" rather than "does it work". I'm hoping this will reduce the RMS a bit, although the analysis is complicated by the cross-coupling w/ length.