I was reading a paper that described how the force a low-thrust torsion pendulum was measured. In it, the paper states a laser is bounced off a mirror and the displacement is "...based upon the beam reflection time." The paper states that the device can measure sub-micrometer displacements.
Conceptually as I understand it, this measurement device would have 4 major components. A laser emitter, a mirror, a detector, and a controller. The controller would power on the laser, then note how long it takes for the detector to respond to the signal.
However, the time of flight difference for ranges this size are vanishingly small. For instance, it'd take a photon roughly $3.336×10^{-12}$ seconds to travel an additional 1mm along the beam path.
If you flip that over to cycles per second, that would suggest the controller has to operate at around ~300 GHz. Only then could it check the sensor often enough to have the temporal resolution to resolve a 1mm change in the beam path length.
This seems like an absurd clock speed for any sort of computer controller. Is there another component, or concept that I'm missing?