Why do Young slits projecting via an intermediary mirror to a screen still shows the normal pattern of interference? This happens even though the mirror is so close that not all interference is constructed yet. The more dramatic experiment being to use a mirror for each slit so as to split and then using two other mirrors to recombine the light. It is my understanding from QM that the interference pattern only happens because there is no measurement between the slits and the screen. However, I don't see how the mirror avoids any measurement at all. Do mirrors somehow escape measuring light and if so, what is the justification for this exception? I find it even harder to accept this possibility considering the mechanical effect of radiation pressure on the mirror.