A currently common method to design high-performance workpieces is to combine two or more materials to one compound. In this way, workpieces can be composed of the most qualified materials according to local loads. When machining high-performance workpiece compounds high quality requirements concerning the accuracy of dimension and shape as well as surface roughness must be fulfilled. However, in case of parallel machining, where the cutting edge moves from one material into the other within one cutting tool revolution, unequal cutting properties have a significant negative influence on tool wear and surface quality. Shape deviations of the surface occur, which are not detected when machining the single materials. The four most significant shape deviations that affect the workpiece quality are the material height deviation, transition deviation at the material joint as well as surface roughness deviation. This paper contains new approaches on the prediction of the surface shape that is generated by a face milling process. The focus is on the transition deviation at the material joint. It arises from a force impulse that is applied on the cutting tool and creates a wavy surface on the workpiece. This shape is predicted via cutting force prediction as well as frequency response analysis of the cutting tool and workpiece in relation to different tool holders. Furthermore, deviations between calculated surface shapes and measured surface shapes subsequent to machining tests are evaluated.