Introduction

Interpolation continues to occupy thought at SEP, and in this issue alone there are contributions from Balog, Claerbout, Ji, Karrenbach, and Zhang. It is a major topic in Claerbout's new book (1991,1992), and Spitz (1991) has recently introduced a new scheme with a useful reference list. Why should I add yet another paper? The answer is that this method is quite different and remarkably simple: Interpolate the unaliased, low-frequency data by classical linear means, and then reconstruct the high frequencies from the low. The computation is straightforward, using one-dimensional modules and well-known techniques. Indeed, the method is simple enough that it seems unlikely that it is new. What may be novel are the pointers to other work that suggest where the method will work, and where it will fail.