As some of you may know, the oplev for IMMT2 is a folded type, so if the injected PEM signal was actually shaking the optical table inside the IFI chamber, that also may be shaking the oplev's folding mirror, and so there might be some responses in the relevant oplev signals.

Fig. 1 is for pitch, and Fig. 2 is for yaw. In Fig. 1 a linear (?) coupling between the PEM injection and pitch is seen clearly. But not for yaw for some reason; maybe I misused the pastavi?? Or, maybe the PEM injection would be done in vertically(?) the stack might be well shaken in pitch but not in yaw (to shaking it in yaw, it should be twisting, which might be difficult to happen...?)

Anyway, the optical table might be shaken lineary with respect to the PEM injection signals. If you have time, please check if, from the IMMT2 oplev spectra (without PEM injection), the optical table's fluctuation could be witnessed or not. Probably some whitening filtering for the relevant QPD would be required.

For your reference the same for MCO oplev (this is also a folded type), which would be the closest to IMMT2, are shown in Figs. 3 and 4; nothing special detected, so MCo might not be shaken due to the PEM injection, and that would mean the optical axis from IMC might not be shaken. (I assume the frequency region in concern would be well out of the ASC loop bandwidth.)