Last night we saw that there is some angular motion of the optics coupling to DIFF.  The first attached screen shot shows our diff noise (calibrated in volts, with a foctor of 10 for the FDD which is not currently in) showing coherence with the OpLevs below 3 Hz.  Most of the RMS comes from frequencies where there is coherence with the OpLevs.  Above about 0.4 Hz, there is coherence with ETM oplevs, but not ITMs, suggesting that our drive to the suspension is causing the coherence. The second screenshot confirms that the ETM opLevs have higher noise when DIFF is locked and we are driving the suspension (solid lines) than when we are not (dashed lines). In both plots ETMY seems to have higher noise than ETMX. 

While we were measuring the ETMY UIM transfer function last night we also measured the transfer function to the OpLevs, screen shots of this and our older (not quite complete) measurement for ETMX are attached.  There are resonances at 0.5Hz, 1 Hz, and 1.3-1.5 Hz that correspond to our noise in DIFF, and the coupling to for ETMX seems to be about a factor of 10 less than the coupling for ETMY.