Figure shows an example of a combined display of 3-D seismic data with a 3-D velocity function. The data and the velocity model are independently sliced, the slices are then rendered in a three-dimensional space and displayed in the same view. To facilitate the interpretation of the display, a contour plot of the velocity model is superimposed to the velocity slice, and two different colormaps are used for color-coding the velocity model and the seismic data. Figure 8 shows the CM/AVS network we used to generate the display in Figure . The seismic data are read, gained, and sliced on the CM-5, and then passed to the AVS module field to mesh that generates 3-D geometries that are finally rendered and displayed by the geometry viewer module. Similarly, the velocity model is read, sliced, and gained on the CM-5, before being rendered together with the data by geometry viewer. Although the rendering is limited to 3-D geometries, the capability of our CM/AVS modules to manage datasets with higher dimensionality than three can be exploited by stepping in time along the other axes. For examples, the data shown in Figure are the results of many 3-D poststack migrations merged together along the fourth axes of the dataset. The migrations were obtained using different filters for the Hale-McClellan algorithm Biondi and Palacharla (1993); Palacharla and Biondi (1993). The quality of the results can be easily compared by switching between different migrations by slicing the 4-D dataset along the fourth axis.