Soundex Calculator — Phonetic Name Coding

Soundex is a phonetic algorithm that indexes names by sound rather than spelling, so that homophones — names that sound alike but are spelled differently — receive the same code. It was patented by Robert C. Russell in 1918 and later adopted by the U.S. National Archives to index the Federal Census; today the same rules power the SOUNDEX() function in SQL databases. Every code is exactly four characters: one letter followed by three digits.

The tool applies the National Archives rules in order:

1. Retain the first letterof the name as the code's leading character.
2. Map the remaining letters to digits using the fixed table:
   • 1 → B, F, P, V
   • 2 → C, G, J, K, Q, S, X, Z
   • 3 → D, T
   • 4 → L
   • 5 → M, N
   • 6 → R
   • not coded → A, E, I, O, U, Y, H, W
3. Collapse adjacent equal codes.Two or more letters in a row that map to the same digit are coded only once. This also applies to the first letter: if the second letter shares the first letter's code, it is not coded again (this is why the f in Pfister is dropped).
4. H and W are transparent. Two letters with the same code separated only by an H or W are treated as adjacent and coded once — the rule that collapses the s and c in Ashcraft into a single 2.
5. Vowels separate. Two letters with the same code separated by a vowel (A, E, I, O, U or Y) are coded twice, because the vowel breaks the run — the reason both n and m survive in Honeyman.
6. Fix the length to four. If fewer than three digits were produced, pad with trailing zeros; if more than three, keep only the first three.

To keep the result trustworthy, the calculator computes each code with two independent implementations — a single-pass state machine that also produces the derivation table, and a separate marker-and-collapse method — and only displays a code when both agree. The letter-by-letter derivation shown on-screen is the actual decision log from the coder, not a re-explanation.

• U.S. National Archives (NARA) — The Soundex Indexing System
• U.S. Census Bureau — Soundex coding description

The letter-to-digit table and the H/W and vowel separator rules were last cross-checked against the NARA and Census sources on 2026-07-16. Soundex is a fixed historical standard, so these rules do not change; the calculator's worked examples (Smith/Smyth → S530, Robert/Rupert → R163, Ashcraft/Ashcroft → A261) match the codes published in those sources.