Clear criteria are discussed for the evaluation and design of surround-sound recording and reproduction systems, free from any quadrifontal (four-source) assumptions. Some weaknesses in the quadrifontal approach are discussed, and difficulties in meaningfully testing systems are mentioned. It is observed that the ultimate aim of surround systems is to provide a good illusion of an intended encoded directional effect, and that this aim is not dependent on any particular choice of number of channels, number or position of loudspeakers, or on any particular method of originating surround program material. An appendix describes the similarities and differences between the resultant -kernel- system approach and the current matrix approach.
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.