Reed Selection Calculator

When the calculator finds a skip pattern is closer to your target EPI than any single whole-number threading, it reports the two ends-per-dent counts you alternate between — for example, threading 2 ends in one dent, then 3 in the next, repeating across the width. The reported actualEPI for a skip pattern is computed as bestReed × (bestEndsPerDent + 0.5), reflecting that average of the two counts, so you need a fairly even split of both counts across your total ends to actually hit that average sett.

The search only considers a fixed list of standard dent sizes — 4, 5, 6, 8, 10, 12, 15, 16, 20, and 24 dents per inch — and picks whichever one, combined with 1 to 4 ends per dent, gets numerically closest to your target EPI. It has no knowledge of which reeds are actually in your collection, so always cross-check the recommendation against your available reeds before committing to a warp.

For each candidate reed, it first checks whether targetEPI divided by that reed's dent count rounds cleanly to a whole number between 1 and 4, tracking how far off that rounding is. It also tests an alternating skip pattern between the two nearest whole numbers, but only prefers the skip pattern when its error is meaningfully smaller — specifically, less than 80% of the best whole-number error found so far — so a skip pattern only wins when it's a clearly better fit, not just a marginally closer one.

Because reed dent counts are fixed and ends-per-dent must be a whole number (or a two-number skip pattern), it's rarely possible to hit an arbitrary target EPI exactly. The calculator recomputes the actual achievable EPI from the chosen reed and threading, then reports the percentage difference from your target as EPI deviation — a small deviation is normal and expected, not a sign of a miscalculation.