Pasting polarized lenses to a PBS ?

[Zafa Pi]

If a PBS (polarizing beam splitter/polarization analyzer) that emits horizontal and vertical polarized photons when observed has a horizontal polarized lens pasted over the horizontal outlet and vertical one over the vertical outlet does this change its behavior? I'm particularly concerned when the PBS is used with another to make an interferometer. If two PBS's are put back to back and a 45° photon is shot in a 45° comes out (interference). Is it still true for the case when the 1st one has the pasted lenses?

 

[DrChinese]

If a PBS (polarizing beam splitter/polarization analyzer) that emits horizontal and vertical polarized photons when observed has a horizontal polarized lens pasted over the horizontal outlet and vertical one over the vertical outlet does this change its behavior?

Generally, there is no effect in this situation. The additional lenses do not affect the photons. I don't follow your other setup where PBSs are placed "back to back".

 

[Zafa Pi]

Dr. Chinese, Thanks for your reply. Call the the 1st PBS I described PBS(0°) And when the lenses are pasted on PBS*(0°). Now shoot a |45°⟩ photon into PBS(0°) and without detecting what leaves use mirrors (and phase adjusters) to direct both exit ports into a PBS(45°). Now measure what leaves PBS(45°). It will be a |45°⟩ photon (rather than |135°⟩) - interference.
Now in a non-reviewed paper by David Ellerman (Delayed Choice found at ellerman.org that was recommended to me by a physics prof) he says that if PBS(0°) is replaced by PBS*(0°) there is no interference, i.e. what will leave PBS(45°) is either |45°⟩ or |135°⟩ with probability 1/2 each.

I see no reason to believe that, he doesn't cite an experiment, I'm suspicious and you seem to confirm my suspicions.