INTRODUCTION

The problem of sorting, NMO correction and stacking for PS data has been widely addressed in the past (, ). The solutions presented in these works apply a lateral shift to the trace midpoints, such that the new trace position corresponds with the lateral position of the conversion point. Usually, this correction does not incorporate the effects of the reflector depth and dip.

In order to incorporate the dip effect, () achieves the transformation from CMP-sorted data to CCP-sorted data using a dip moveout operator. In a similar way to the PP-DMO, PS-DMO may reduce the problem of reflection point dispersal due to dip without knowledge of the reflector geometry. Most of the existing PS-DMO operators present errors due to truncation of power series and/or second order approximations (). () show a fast converted dip moveout operator in the f-k domain which partially alleviates approximation errors.

() presents the zero-offset mapping equation for the PS-DMO operator, applying an integral-summation approach in order to implement his PS-DMO operator.

() present a log-stretch f-k PS-DMO operator. His operator correctly handles the kinematics, but doesn't handle amplitudes properly. Here, I present a review of the PS-DMO operator, implement the operator described by () and extend it to handle the amplitudes properly.