Fan filters for data with variable spatial sampling

We develop a version of the velocity or "fan filter" which is applicable to seismic data with variable spatial sampling. Our formulation of the fan filter involves convolving a multichannel difference operator with the Hilbert transform of the data. We present results of a pre-stack application of this filter to wide aperture variable spaced CDP data. The results show that, in our example, the fanpass filter is able to attenuate multiples because of their residual moveout after NMO.