Depth information from binocular disparity and familiar size is combined when reaching towards virtual objects.