Two bezel samples behind one scratch label

Yes. Both are retained, but only one has the sample label attached. The other was in the same bag. I have stopped treating that as sufficient identification.

Have quality examine them as two samples. A shared bag is not a shared identity. Keep the old image links marked uncertain until someone can match the actual features.

I am dealing with photographs whose rejected parts are gone. You have a useful advantage here. Ask quality to point to the judged feature on each retained bezel, then photograph enough of the whole part to find that feature again.

Keep an unmarked copy too. Arrows can cover a faint mark.

Quality found a real scratch on one bezel and no matching scratch on the other. The bright stripe in the second image moves with presentation. We now have separate identities and unmarked photographs. The old shared label is withdrawn.

What is the lighting trial supposed to improve now? The mistaken label is sorted, but I would still want the faint real scratch included when someone proposes a cleaner-looking image.

Reliable distinction between those two cases. I have asked for a comparison using both retained bezels and the intended holder, with repeat presentations. No new lamp has been selected.

That gives the trial a much clearer purpose. I would let the person doing routine inspection see the comparison too; they may recognise a presentation that looks tidy in a demonstration but is awkward to repeat.