Also, when filtering the imagery for "glints" was the only data used the consecutive imagery taken back then? I.e. are we certain it's not the other way around: a real object (star, comet,...) being obscured? I.e. there are 3 types of imagery to compare: the 2 consecutive images displaying a putative glint, and for example current imagery showing an actual star for example.