Let's say we have light contained in a cavity (of initial length $L$) and then instantaneously remove the second mirror, thus removing the barrier between the first and a third mirror which is some further distance away (say $L$ again). If the light was on resonance before the change, do I think about it as two counter propagating wave pulses (trains) which then move out into the unexplored part of the larger cavity over a time equal to $2*L$ (in units where $c=1$)? If initially its off resonance can I still use the same conceptual picture?
Does this imply that the light pulse will remain localised to a length $2L$ and that if I reinsert the second mirror at exactly the right time (waiting for the pulse to 'bounce back') I can 'catch' the light back in the initial smaller cavity?
Can I think about a complimentary photon picture where the same 'catch' can be made as the photons all depart the initial cavity region over a time $2L$ and then return in concert? The reasoning behind the $2L$ being that a left moving photon could be at the second mirror (just after a reflecting collision) and takes a time $2L$ to return to the now vacant second mirror locale.
Is there anything in the statistics of the 'catch' that makes a distinction between photons versus waves?
For clarity, I'm thinking of the medium being a vacuum where there is no dispersion.