In this case it's light bending around a gravity well which produces multiple "copies" of the star, but the light from each copy arrives at different points in time.
So they could capture the light and attempt to capture the "same" light again later to try to verify or change the result they received.
I did some more reading about it after posting, and the original experiment was intended to test if light "chooses" to be a wave OR a particle in a way they could affect, and that it could only be one or the other at a time. The truth seems to be more that it acts as both at the same time and whatever sensor equipment you use to pick it up is what it acts like.
[citation needed]