Encoding of 3D Structure in the Visual Scene: A New Conceptualization