I have separated the camera originals from the marked copies for our UR5e bezel trial, but the chosen reject annotation changes edges when the bezel is reseated; quality has the physical part, and I want one checked feature location before the camera specialist spends time on that travelling arrow.
They identified a fine scratch beside the locating notch; the big arrow was following a reflection along the opposite edge, added later by the person preparing the slides.
The notch gives you something useful to match, but keep the whole view with each crop so the next person does not have to rediscover which edge is which.
Yes, both physical bezels were checked together; the accepted one produces a stronger bright band than the scratch on the rejected one, which explains the original temptation to mark it.
Ravi, our tutor controls the shared lesson and has withdrawn the guessed arrows; Pavel, the scratch becomes hard to see at one permitted seating position, so the optical comparison is not finished.
Give the fixture designer that awkward position with the optical comparison, because removing it from the pictures would not stop an accepted bezel sitting there.
Designer confirms the position is within the required range; vision is comparing another lighting arrangement with both identified bezels, and the lesson now shows the weak view as an unresolved imaging example.