Bezel reference views missing the inner lip

Only its outer edge. The mark is farther inside, behind it, so the actual marked face isn't in the image.

Then stop arguing with the brightness slider. Can the inspection plan include a view of the inside face, or is everyone assuming one pretty front picture covers the whole bezel?

Give that marked bezel its own linked views, because a front picture labelled reject without the visible reason can teach someone to reject the wrong feature entirely.

Sofia, an extra view still needs a repeatable presentation. Otherwise Jamie gets another moving reflection puzzle.

Luca, yes, a defined second view. Not someone waving it at the camera until the scratch finally looks photogenic.

Priya, the inner lip is in quality's inspection scope. Luca Barnes, the physical sample and front view already share an ID; I've added a note saying the rejection reason isn't visible in that view.

Good, then the missing view is a real coverage gap; keep the front view's purpose clear rather than changing its label to suggest it can inspect the lip

The integrator has proposed a second presentation with the inner face visible. Quality wants an accepted lip alongside the rejected one, so we aren't only checking whether the new view makes this one bad sample look bad.

Yes, include the accepted one with the awkward finish too if quality has one, since the nicest accepted bezel usually flatters the lighting and tells you very little about the boundary.

Jamie, did they manage both views without changing the holder by hand?

Yes, the checked second presentation uses the same holder. We captured both identified bezels through the proposed sequence; the inner mark is visible, but an accepted surface has a bright band very close to it.

Annoying, but useful annoying. Does the band move when the same accepted piece is presented again, or are you comparing two different finishes?

Sofia's distinction matters; two parts having different appearances is not itself evidence that presentation is unstable, although I cannot judge the band without seeing those paired views.

Same accepted piece, repeated presentation. The band stayed in the same place in those captures. I had mixed that result with our older moving-streak problem when I described it to the team.

Keep that accepted example, then, because the new view needs to cope with its legitimate bright feature rather than have someone quietly remove it from the folder to make tuning easier.

That's a different problem from the old streaks, then. At least you can finally see the surface you're being asked to judge.

Quality has signed off the two-view coverage and the physical judgements on this small reference set. The accepted bright band stays in it. We haven't found a reliable image rule separating it from the rejected lip mark yet.

Thanks for coming back with the accepted example too. Easy one to leave out of an update.