Modeling audio fingerprints by P.J.O. Doets