Studies on the recovery and processing of surface waves from ambient‐noise interferometry are now standard practice in seismology. It has recently been shown that body waves propagating within an array of seismic stations can be recovered under appropriate conditions. In this article, we investigate and develop optimal data processing procedures necessary to recover Green’s functions for P waves that are propagating through a network of borehole seismometers near Parkfield, California. In particular, we discuss the effects that common preprocessing operations have on data recorded by these stations, such as 1‐bit normalization, event muting, and decimation. Considering these operations and their effects, we have come to favor a simple whitening procedure that is based on a phase reconstruction of signals. In the context of these stations, we have shown that whitening seismic data in this manner is sufficient to recover body waves via interferometry techniques. Applying these procedures, we have detected P waves propagating between these stations, from which we have produced a first‐order crustal velocity model; the specifications of this model are in general agreement with those of previous velocity models that have been proposed for the Parkfield region.