Next: Discrete Fourier basis
Up: Interpolation with Fourier basis
Previous: Interpolation with Fourier basis
For the continuous Fourier transform, the set of basis functions is
defined by
| |
(18) |
where is the continuous frequency. For a 1-point
sampling interval, the frequency is limited by the Nyquist
condition: . In this case, the interpolation
function W can be computed from formula (13) to be
| |
(19) |
The interpolation function (19) is well-known as the
Shannon sinc interpolator. A known problem with its practical
implementation is the slow decay with (x - n). This problem is
solved in practice with heuristic tapering Hale (1980),
such as Harlan's triangle tapering Harlan (1982). While
the function W from equation (19) automatically
satisfies properties (3) and (17), where
both x and n range from to , its tapered
version may require additional normalization.
Next: Discrete Fourier basis
Up: Interpolation with Fourier basis
Previous: Interpolation with Fourier basis
Stanford Exploration Project
11/11/1997