I think there might be some scratches on layers 0 and/or 1 on an ATMR drive, so I'd like to take the platters out and inspect them. But how accurately must they be realigned when I put them back? Is it enough to make a marking on the side and align them carefully visually?

I have transplanted ATMR platters from one HD to another in a different case, and that worked fine. So off-center alignment doesn't seem to be very sensitive, but I have no experience with rotational alignment of platters.

it needs to be microscopically exact. in other words you need to find a way to keep the two platters together as they are removed as one stack.

if this is for a client... i would not do it.

if it is just for you and the data is not critical... have at it, but be prepared for a total loss.

Thanks! But that wouldn't make it much easier to inspect the layers between the platters.

Do you have experience with Hitachi ATMR disks is particular, or is this a general warning? For other drives, I'd avoid this, but I had another ATMR case where I had to move platters from one drive to another, successfully. I just used a basic platter tool, a metal cylinder, that holds the platters together. I would be surprised if the platters didn't move a little bit that time, easily one millimeters relative to each other, and that worked fine. Definitely no micrometer accuracy there. And I didn't even measure off center - I tried on a test drive first, and then moved platters from patient to donor. It worked fine both times, so ATMR seems to be very tolerant. But I'm very happy to hear others' experience on these drives!

Anyway, how does one measure the exact rotation very accurately? I understand how to do the measurements to off center alignments, but it's not clear to me how to measure micrometer differences in rotation in a practical way.