Fractional-delay finite impulse response (FIR) filters have many uses, such as in communications for timing recovery, or in digital audio for arbitrary resampling.

In this paper, Shyu and Pei propose a reformulated error function least-squares approach for the design of arbitrary fractional-delay filters, and present a numerically stable procedure for the design of such filters. One advantage of the procedure is that it is fast, due to the structure of the matrices used to obtain the coefficients of the filter. In Section 2, the method is explained in a concise manner. In the last section, several detailed examples are presented to illustrate the authors’ method. Crucial details about how the matrices involved are formed and evaluated are only given in the first example and not in the section on the algorithm derivation.

Without being a tutorial on fractional-delay filter design, the paper succeeds in clearly explaining an alternative to the established solution methodology.