Datuming as proposed is very expensive because you must do calculations for every couple of source-receiver pairs, within a reasonable aperture, because the operator does not separate. Instead of wave equation upward continuation, Schneider 1995 uses a dynamic time shift for efficient 3-D water layer replacement. The dynamic shifting approach has the benefit that it is very fast, but also the drawback that it is really a 1-D operator. Prestack datuming in midpoint in offset coordinates is quite expensive, but can be made less so by selectively limiting the range of offsets and azimuths from each input midpoint which contribute to a given output trace. In the limiting case the datuming operator bears some resemblance to the common azimuth migration operator of Biondi 1994.