the adapter disappears behind the tray in every clip

Their email offers a close-up of the tool on a bench. Useful for identifying it, but I don't think that answers the handling bit.

It doesn't. Take the close-up for identification, but keep the supplied assembly handling your sample as the outstanding request.

They say the tray screen is why the camera cannot see the pickup. That's fair. I can see the screen in the opening shot.

We used two views for a hidden contact once. Neither was impressive alone; the shared trial identity made them useful together. Ask what view they can provide without changing the handling setup.

Omar, did both cameras run at once or did you repeat the trial?

Both at once in our case. A repeat could be useful, but I wouldn't describe it as the same pickup.

Don't ask them to remove a protective screen for a prettier video.

Toby, just the purchase reference. Nora, agreed, I don't want the screen removed. I want them to propose a view that works with the setup as supplied.

The bench close-up could identify the assembly revision, provided they connect it to what is fitted during the handling trial. The order card alone leaves that open.

And a matching painted colour won't do that.

Funny you mention colour. Their reply literally says 'same blue adapter'. There is no readable part marking in the clip. I'll ask for the drawing reference they're using.

Colour was how we mixed up two otherwise similar holders. The mounting faces differed. I would keep that question separate from whether the samples were handled acceptably.

Omar, exactly. One answers which assembly, the other answers what it did.

I've sent two questions: which supplied assembly is fitted, and how can they show the hidden pickup with that assembly? Much shorter than the message I was about to send. Thanks.

Do the samples have identifiers too, Nathan?

Yes. The packet labels match our sample list and the destination pieces stay visible. The pickup contact is the missing view, not the sample count.

That narrows it nicely. Keep the sample-identification part as received, so the next request doesn't sound as though every part of their recording failed.

Post their proposed view before agreeing to another generic demonstration. They've had enough practice at those.