Dear dr. Rorden,

Thank you for your fast reply.
The last time we updated from github was a bit more than a month ago, so I'll try the latest version as well.
It's a good idea to see if this particular behavior occurs consistently with a particular vendor/scanner, I will look into that.
I'm still a bit puzzled about the tolerance though, because what I've observed have been inconsistencies in the order of rounding errors. These are below the 0.2mm threshold. The software picks it up though and interpolates the first few slices in the axial direction in my current example.

Thank you for your time and answer, I will update DCM2NIIX and check if the occurrence happens with the same vendor/scanner.

Kind regards,
Martin