"Where's the Real Magazine?" As We See It, February 2001
By John Atkinson
Jonathan Scull told me there'd be trouble when I decided to put the Denon AVR-4800 surround receiver on our December cover. As you can see from this issue's "Letters," he was right. Although reader Bob Laurie and retailer John Weires refer to the AVR-4800 as a "home theater" component, Stereophile's review did examine the Denon's performance in the context of music reproduction. I suspect, therefore, that assuming these writers did read the review, their objection is really to the concept of surround sound, and that they firmly believe that music playback is synonymous with, even demands, a two-channel system.
This is understandable, given that 99.999% of recorded performances you can buy are two-channel, and almost all audiophile-quality playback systems feature two channels. But as Stereophile's founder, J. Gordon Holt, has been pointing out for decades, "stereo" playback doesn't inherently mean "two channels." As the word stereo is derived from the Greek word for solid, any improvement in the perception of stereo image solidity is a legitimate step forward in high-fidelity sound reproduction, regardless of the number of channels. The real question is whether or not someone wants to pay the price to get that benefit in their home—and that, as I point out in response to another reader in this issue's "Letters," is the fundamental issue facing the proponents of DVD-Audio.
At the end of November, I had first-hand experience of this question. At the invitation of audio researcher James D. Johnston ("JJ" to denizens of the Internet newsgroups), I spent a very enjoyable morning at AT&T's Shannon Laboratory in New Jersey, listening to the surround recording/playback system JJ has been involved in developing (footnote 1). Using an array of seven directional microphones to sample the original live soundfield at one point in space gives a five-channel recording that, when played back with five speakers correctly positioned, does an astonishingly impressive job of superimposing that original space on the listening-room acoustic.
I found the experience equally convincing, whether the surround recording was of a string trio, an orchestra, a pipe organ, or a rock band playing live in what sounded like an aircraft hangar. In fact, it was even convincing without any music being played at all. The sounds of organ blower noise, audience coughs and rustles, AC noise—all the live clues that enable your brain to identify the space in which you find yourself and adapt your hearing accordingly—were sufficient to immerse me in the recorded acoustic. I also found the sweet spot more expansive than in a typical two-channel situation; in particular, as I moved away from the prime seat, the perceived acoustic perspective changed much as it would have done in real life. In addition, the sense of perceived space either side of me was better defined than I have experienced from conventional surround recordings.
I drove back to Brooklyn deep in thought. There was no doubt that I had experienced audio playback of considerably higher fidelity than I had ever experienced from a two-channel system. But did I want that experience in my own listening room? As always, it comes down to what you want a recording to do: reproduce the music you love in whatever fidelity is appropriate; or reproduce a very small amount of other people's music in as high a fidelity as possible.
The key to JJ's system lies in capturing precisely the correct spatial information when the original recording is made. This limits the improvement in fidelity to that tiny handful of recordings that have so far been made in this manner. By contrast, I have an enormous collection of conventional two-channel recordings that I bought for their musical value, and the manner in which those recordings have been made doesn't capture enough spatial information for convincing playback in surround sound.
I don't yet know if I want to pay the price in system complexity to get that improvement in fidelity. But the fact that two-channel recordings cannot reproduce the original space was thrown into sharp contrast the evening of my visit to JJ's lab. Courtesy of New York public radio station WNYC, English guitarist Robert Fripp was in residence at the World Financial Center's Winter Garden in Manhattan, performing more in his series of Soundscapes improvisations. Playing solo, Fripp set up cascading layers of slowly shifting chords and melodic fragments, using the current version of his "Frippertronics"—multiple digital delay units and guitar-processing units. As the hall's cavernous acoustic filled with music, and slowly decaying, only vaguely guitar-related sounds approached me from all directions, I was immersed in the experience to an extent that just isn't possible from two-channel playback.
The key to surround-sound reproduction might well depend, therefore, on a synergy between the music and the experience. Or it might well remain a bywater, two-channel playback proving the most parsimonious way of presenting the recorded acoustic with sufficient fidelity to support the music.
But whatever happens, Stereophile will report on it—and put products it thinks important on the cover!—John Atkinson
Footnote 1: See Johnston's paper, "Perceptual Soundfield Reconstruction," presented at the 113th Audio Engineering Society Convention in Los Angeles, September 2000.—JA