Polarization in laterally moving reference frame

[Swamp Thing]

Imagine that you are sending photons that are polarized at $45^\circ$, towards a detector that you are facing (e.g. the detectors's input face is parallel to your device/computer screen right now). The detector's polarization axis is also oriented at $45^\circ$, so all the photons are registered.

An observer is moving from your left to your right. Would the $45^\circ$ orientation of the detector be something different in this observer's frame? Would the photons' polarization be different? (Obviously, both would have to change in such a way that the observer's physics would predict 100% detection -- how would that work out?)

 

[andrewkirk]

Because of the Lorentz transformation, the horizontal dimension will be compressed in the observer's frame, while the vertical dimension will be unchanged.

That will increase the angle to the horizontal of the polarisation and detector directions, as measured in the observer's frame of reference, from 45 to something steeper, say d degrees with d>45.

But since the direction of the photon polarisation remains equal to that of the polariser - although both have changed from 45 to d degrees - the observer will still predict that all photons will be registered.