Lanning and Johnson's paper presents the technique of Walsh transform domain low-pass filtering as a means of enhancing transitions in well logs prior to boundary picking. They argue that existing fast means of computing the Walsh transform (fast Walsh transform algorithms) make this procedure particularly efficient. In this discussion it is shown that, for all practical cases of interest, the results of their paper can be obtained without any mention of the Walsh transform. In fact, using the Walsh transform unnecessarily increases computational complexity. Specifically, it is shown that the low-pass sequency filtering described for obtaining a stepped version of the original signal with a given step width is equivalent to segmenting simply the signal into equal length segments and replacing all values in each segment by the segment average.