For a given receiver, the recorded trace is subtracted from all the local slant stacks to which it contributes. The trace is then crosscorrelated against all the stacks to which it contributes, forming a single crosscorrelation function. The lag which gives the largest correlation coefficient is the static value that maximizes the power of the affected stacks. Using this new static value, the trace is stacked back in to all stacks. After a new static value has been estimated for all receivers, the procedure then performs a second iteration, stopping when the stack power has reached a maximum or after a specified number of iterations have taken place.