Thanks for the info Chris. I’ll have a chat to the coder implementing this effect and see which of the two suggested approaches suits us best. We are only interested in a small number of meshes, but do need to use the actual scene depth, so the second approach sounds more appropriate.
I think the presumed interaction of meshes not included in that pass with meshes included in that pass was a misunderstanding by me and can safely be ignored.
Thanks,
Mike.