A method of using a computer to generate virtual autostereoscopic images from a three-dimensional digital data set is disclosed. The method includes establishing a first point of view and field of view of a subject volume including a region of interest. The method includes reading at least one scene parameter associated with the field of view of the subject volume. The method includes determining a second point of view offset some distance and along some vector from the first point of view based on a value derived from at least one scene parameter. The second point of view at least partially overlaps the first field of view. The first and second points of view each create a view plane with a view orthogonal to the subject volume.