Footnotes and references

1. There was a third problem, too: the word "quadraphony" itself. The word is a combination of Greek and Latin roots, and should really have been called "tetraphony" (Greek) or "quadrasonics" (Latin). Not that it would have sounded any better…

2. Michael Gerzon regrettably passed away in May, 1996, as a result of an illness that had troubled him for many years. Gerzon was responsible for an astonishing number of audio-related patents, rivaling Blumlein in their breadth and scope. With a number of colleagues, Gerzon carried out work in many branches of audio, ranging from surround sound and sound for HDTV with Geoff Barton and others, to digital audio dither techniques (with Peter Craven) and reverberation algorithms for audio software specialists Waves. He will be sadly missed.

3. Current Ambisonic decoders can cope with a varying number of speakers in a number of different configurations. For horizontal surround, the minimum is four, and this is adequate for a great many listening environments from studios to living rooms. In larger playback areas, such as movie theaters or for live sound, six, eight or more speakers can be used. For with-height reproduction a minimum of six speakers is required, but for most applications this is inconvenient as the array does not include speakers in the normal stereo (or horizontal surround) positions, so eight is more common.

4. The fact that Matrix H - and thus eventually UHJ - owed something to Sansui's QS quad matrix system is at least part of the reason why Ambisonic decoders produce excellent results decoding Dolby Surround recordings - which use a matrix also based on QS (as does Circle Surround, incidentally). The absence of "logic decoding" in an Ambisonic decoder (it would be a disaster as Ambisonics relies on all the speakers' contributions to build the image) also makes the imaging a good deal more stable.

5. Each form of UHJ actually has its own name, but nobody ever uses them. Instead, you generally refer to the member of the UHJ hierarchy by mentioning how many channels it has. But theoretically we should apparently call 2-channel UHJ "BHJ"; 2.5-channel UHJ "SHJ"; 3-channel UHJ "THJ" and 4-channel Periphonic UHJ "PHJ". I had never heard of these abbreviations until Martin J Leese mentioned them in his invaluable Ambisonic FAQ, and can only hazard a guess at what the extra letters stand for. Probably "B" is "Basic", "T" is "Three", and "P" is "Periphonic". Your guess is as good as mine as to what "S" stands for ("Squashed"?).

6. The phase relationships in 2-channel UHJ sometimes provide problems for listeners new to Ambisonics, and a few people complain about the speakers being "out of phase" or mixes having "out of phase" components. The simplest way of sorting these out is to listen in a darkened room or with the eyes closed. The trick is to "unlearn" what we are used to in listening to conventional "panpotted mono" two-channel recordings, and to listen to the 2-channel UHJ recording more like the way we listen to sounds in the outside world. Sitting in the dark and visualizing the original recording environment (if there was one) can help in this respect. Very often, after a short time, listeners who previously experienced problems suddenly "get it" and the recording springs to life for them. The experience is analogous to looking at those stereoscopic images available in a number of books today. You have to learn the knack of looking at them - and when you do, the 3D effect suddenly becomes impressively evident.

7. A major player in the exploitation of Ambisonics for multimedia is Lake DSP Pty. Ltd. of Sydney, Australia. This company has developed a powerful DSP-based system which is capable of creating an almost infinite set of "virtual surround" playback systems, for example decoding B-Format or regular 5.1 surround recording for stereo headphones. An impressive demo CD-ROM is available on request.

8. The Soundfield Mic was ultimately manufactured by Calrec Audio Ltd, and a description of it by Howard Smith was published in Studio Sound in October 1979. Descendants of this mic are now manufactured by Soundfield Research in the UK.

9. The article originally appeared in IBA Technical Review 14, Latest Developments in Sound Broadcasting, June 1981, and was republished in Studio Sound in August 1982.

10. A complete discussion of these units and how they are used appeared in the article An Introduction to Ambisonic Mixing, Studio Sound, October 1982.

11. An astonishing Ambisonic Discography is published on the Internet and impeccably maintained by Mark Anderson. It can be found at

12. Visit the ARA Web site.

13. Namely the report of the DVD Consortium's Working Group Four, issued in January 1998.

14. Such as the Surround Sound Internet mailing list.

15. Even so, Logic 7 makes old Dolby Surround matrix recordings sound halfway decent for the first time: a major achievement which only scratches the surface of Logic 7's abilities. Unfortunately, Griesinger is not a great fan of Ambisonics, as he is of the opinion that wavefront reconstruction theory means that the Ambisonic sweet spot will be tiny. We hope that while he is in San Francisco for the upcoming 1998 AES Convention he will have a chance to listen to Jeff Silberman's excellent B-Format recordings which do not seem to suffer from this problem…

16. Don't ask me why. Some of our nomenclature is a bit arcane in Ambisonics.

Return to the article