Today I installed an optical lever to sense the motion of the suspended input mirror of the filter cavity. Such lever uses a lens and a pair of PSD respectively placed in the focal plane and in the image plane. In these scheme the PSD in the focal plane is only sensitive to mirror tilts while the one on the image plane can only sense its shift. A picture of the optical lever is shown in the figure attached.

A first injection of white noise has been performend in order to excite mirror resonances. After identifying the degree of freedom (tilt or shift) at which each resonance belongs it could be possible to fine-tune the PSD positions to optimize the signal decoupling.