The chart design document for i1iSis says scaling of +-8% is accepted, but I didnt try anything less than 6mm.
You meant something like 10% ? so 6.6x6.6 ?
What I mean is:
1. Don't use the minimum (all my targets for customers are larger because there's no benefit to getting too close to a size that can make the unit barf on a target).
2. Some users do not print the targets at exactly 100%, so even with a minimum (which, again, buys nothing useful), if the print is scaled too much, the unit will barf.
3. The 8% fudge factor seems a bit too high based on my experience. I'd say closer to 4-5%. But if the print is scaled too far, the triangles on the sides will be out of whack, and you can't measure them.
4. Unlike an i1 Pro Spectrophotometer, the iSis is very picky about size. Producing a target that produces measurement problems serves no purpose, especially if you are not the person doing the scanning and are sending it back to be scanned by a service. In-house, you can always recreate or reprint a target. If this is a remote operation, scanning issues cost time and money.