Information loss in polarization.

[anorlunda]

A single photon approaches a polarization filter. It's orientation is described by a probability amplitude wave function. It is not reflected or absorbed, but passes through the filter. Now it's orientation immediately after is definitely known.

Question 1: is there no information/enthalpy gain/loss in this event?

Question 2: it seems that this event could not be time reversible. We can not have an event where a photon of orientation known to be the same as the filter, passes through and becomes a photon of unknown orientation.

 

[wotanub]

You have to consider the polarization with respect to different axes noncommuting observables.

When you put the photon on a filter, you are choosing to measure the polarization with respect to some axis. Let's say it passes through a 0 degree polarizer. Now you know it would definitely pass through a second 0 degree polarizer and would surely not pass through a 90 degree. But what about a polarizer oriented at any angle between? You don't know, and you're back in a superposition of amplitudes to either pass through or not. There is one angle where it is deterministic, but all the others have some doubt. I know that wasn't exactly what you had in mind, but that's as close as you get because there's no way you could make a measurement, then end up knowing less about the photon.

You are right that the information about the photon's polarization with respect to axes x-y is lost when you measure the polarization with respect to rotated axis x'-y'.

 

[anorlunda]

I'll try to rephrase my question. (These forums are great. If you don't formulate your question carefully, you don't get answers.)

A quantum state may be observable (such as by passing a photon through a polarizing filter), but there is no such thing as unobservation. The known orientation of the photon after the filter can never be restored to the wave function of probability amplitudes it had before the filter.

Is this not an arrow of time; that we can observe quantum states but never unobserve?

Is it related to the second law of thermodynamics?