A couple of visible threads is only a cosmetic problem, not a structural problem. If you use standard 12 mm nipples, and the threads protrude 0.5-1.0 mm below the shank, there will still be around 7 mm thread engagement instead of 9 mm. Nothing to worry about.

A cheap way to fix it is to buy longer nipples. Assuming you are using std. 12 mm nipples, you can just substitute them with standard 14 mm or 16 mm DT Swiss brass nipples to counteract the slightly short spokes. A 14 mm DT Swiss nipple will effectively elongate the spoke by 1 mm, and 16 mm nipple by 2 mm. Notice that because the internal thread on such a nipple is lowered by eg. 1 mm, the spoke shouldn't be flush with the top of the nipple, but should also stop 1 mm below.

Why your spoke calculator gives a different output than other programs can be difficult to say. I always use the "spocalc lite" spreadsheet because it is open source and has been peer reviewed by people who knows their stuff. I won't rely "black box" software that gives out a magic set of numbers.

Some spoke calculators automatically subtract 0.5 mm to compensate for rim compression and spoke elongation under tension, other don't. Some compensate automatically for staggered spoke holes, others don't. DT Swiss's online spoke calculator wrongly subtract 2 mm from spoke length when using 14 mm nipples instead of 1 mm. (hopefully corrected by now). Most rim and hub databases are full of errors, and even manufacturer homepages often give wrong ERD number for rims. ERD, unlike "spoke seat diameter" (Mavic), isn't a number one can derive from the rim alone.

The "ERD" is the point you wish to terminate you spoke*, and that point is usually where the spoke end is flush with the top of the nipple, but that point depends on the length and shape of the nipple. So the ERD changes whether you are using 12 mm or 16 mm DT Swiss nipples. There is some convention, that the ERD is measured with std. 12 mm nipples, but I don't think there exist anything like a fixed definition for what ERD exactly is and how to measure it. I know of at least three competing definitions of ERD, so the ERD that the manufacturer gives out, may be different than the one your program uses.

It is all a bit of a mess, so personally I only count on ERD's I have measured myself. What is really needed is community database project where people can submit hub and rim measurements according to certain, defined standards.


*For one definition of ERD. The correct one as I understand it.