Notes on the Search for Startling Innovations in 3D Audio

All print journalism now trails an Internet shadow: the digital version, a Platonic reflection consisting of what might have been if you could have elaborated, footnoted, and linked far beyond the margins of the lines on the page.

On paper, the limits of space -- of word counts and ad/edit ratios and the cost of printing and distribution -- exist in an inverse relationship to the extensibility of online prose. In broadcast media, where the limiting factor is time, you encounter a similar kind of extensibility whenever Jon Stewart winds up a Daily Show interview segment with a frustrated unwillingness to stop, announcing that the conversation will continue off the air and "we'll throw the whole thing up on the Internet."

Information in general wants to be free, but online text, with a nominal distribution cost approaching zero, really is free, and every time I commit my professional words to paper these days, I immediately begin scheming to scoop up the words and ideas and connections that were left on the cutting-room floor, and reconstitute them online.

For the March issue of The Atlantic, I contributed a report on startling innovations in 3D audio as developed by Edgar Choueiri, a professor of applied physics at Princeton University. (Read "What Perfection Sounds Like.") In this post, I'd like to unpack the Internet shadow of this particular article of mine by presenting a few notes on useful background information and context that travels in the slipstream of the finished product, along with some deleted scenes from my encounter with Professor Choueiri.

My dispatch on a very specific something new under the hi-fi sun points toward overarching questions about the general state of audio innovation, and this assignment sent me into the heart of those fascinating inquiries. My experience, ears, judgment, and research convince me that Edgar Choueiri's 3D audio algorithms and playback system represent a dramatic improvement in the spatial realism and virtual sound-staging of stereo. It's an achievement whose novelty and pleasurable impact justifies the hyperbole of the article's title. (Ideally, his sound filter requires recordings of an actual soundstage and ambiance; it doesn't work at all with mono recording, although it provides surprising enhancements even for typical stereo pop concoctions, where the spatial location of voices and instrumentals is simulated at a mixing board by a technique called "pan-potted mono.")

It's important, however, to stipulate that Choueiri's Pure Stereo is a culmination of research on crosstalk cancellation conducted by a far-flung community of engineers over many years. Science is never wholly original. One of the trickiest challenges of science and technology journalism is how to accurately characterize innovative achievement in a clear and distinctive light, while giving due consideration to the wider range of work in the field. Putting one guy's beautiful solution in bold relief risks obscuring the surrounding network of colleagues (and competitors) along with the deep bibliography of research that stands behind any truly significant breakthrough.

One crucial predecessor of Choueiri's is Ralph Glasgal, whose earlier work on crosstalk cancellation and ambiance simulation has proceeded under the rubric of Ambiophonics. Glasgal's website is an illuminating resource. Bob Carver, an ingenious and storied pioneer of audio design, made a somewhat Ahab-like stab at an analog solution to 3D audio some 40 years ago, and dubbed his technology Sonic Holography. (Choueiri has a vintage Carver Sonic Holography Generator Model C-9 in his gear rack at his Princeton lab.) Another hotspot in audio science is the Institute of Sound and Vibration Research at the U.K.'s University of Southhampton. Choueiri's decision to focus on 2-channel stereo 3D was based in part on the Southhampton lab's successful implementation of crosstalk cancellation with six speakers.

The recent and increasingly chastened hysteria over visual 3D movies and TV has its counterpart in a parallel gold rush to commercialize 3D audio. Princeton's 3D audio technology will doubtless become available to consumers soon, but it's just one player in a proliferating 3D sound multiverse. The industry sent out an important signal last year with the establishment of the 3D Audio Alliance (3DAA), a trade group devoted to pooling knowledge and creating technical standards. (This recent episode of TWiT's Home Theater Geeks podcast is devoted to the 3DAA launch, featuring Alan Kraemer of SRS Labs, Inc., a leading purveyor of "advanced audio enhancement.") Hearing Choueiri's 3D audio demo was even more exciting for me than the sometimes thrilling cinema 3D of Avatar. It might be worth betting that 3D audio has a better prospect for success in the near future than a thousand James Camerons breaking the fourth wall on screens everywhere.

One of the reasons you don't hear much about genuine audio innovation is that the audiophile press practices a blatant silo journalism, narrowly focused on refinement rather than advancement. There's absolutely nothing wrong with refinement; in the right circumstances it creates a valuable and highly significant species of progress. Gadgets like the brilliant new HRT iStreamer mean that iPhones and iPads can now become uncompromising high-end audio source components. Cheap digital storage removes any practical barrier to playing uncompressed high-definition audio files. The vinyl resurgence is not a purely nostalgic exercise: improvements in cartridges, turntables, and phono amps lets us hear the full sonic potential of old LPs that vintage gear could never reveal. A well-designed 21st century amplifier with mid-20th century vacuum tubes can seduce the ear for perfectly up-to-date reasons.

But refinement is for connoisseurs, not pioneers and creators. Nothing puts the baroque artisanal excesses of hi-fi in perspective faster than spending time with hard-headed audio engineers in the recording industry or researchers at university audio labs, where experiment and the scientific method trump silly gold-plated luxe. Reading about high-quality audio would be much more fun if news from these scruffy studio boffins, pro gear vendors, and bleeding-edge psychoacoustic academicians became a steady part of the conversation.

The relatively primitive level of spatial music reproduction has been a blind spot in audio journalism, despite never-ending hype about hi-fi that sounds like the real thing. Now engineers and physicists like Choueiri, who can harness the mathematics of wave theory and write powerful audio software algorithms, are about to give us all wonderful new sonic gifts. While we wait for Choueiri's Pure Stereo to arrive, a closely related and equally mind-boggling digital signal-processing technology is already available from Smyth Research. Rather than 3D audio through loudspeakers, Smyth's Realiser A8 system provides headphone listeners with sound that is indistinguishable from playback through loudspeakers. The system's ability to emulate exact room and speaker configurations is said to be uncanny, defeating everything that seems unnatural about listening. Stereophile's Kal Rubenstein was suitably agog. "I couldn't believe it," he wrote in a detailed review. "For the first time in my life, headphone listening was not only convincing but enjoyable."

When I left Choueiri in Princeton, he was immersed in the next stage of his research. For true stereo (including his Pure Stereo), you have to listen in a delimited "sweet spot" between the two speakers. Choueiri is now working on an extension of his 3D filter that will work with a multi-speaker array to produce what he calls "the holy grail -- multiple sweet spots." Meaning two, three, four or more listeners in a room could hear the same spatially realistic sound at the same time -- a profound boon for social listening and home theater watching with a group.

UPDATE: Choueri has just announced that on January 25th he successfully produced three distinct sweet spots "using a special non-conventional loudspeaker technology in combination with my optimized XTC filters." He added, "The extension to more sweet spots is relatively trivial."

Finally, a few notes about Choueiri himself, a fascinating character worthy of an in-depth profile.

Born in Lebanon in 1961, Choueiri was an Apollo-age science and audio geek, the only 13-year-old in the country with a quadrophonic sound system. He sketched rooms with walls made of loudspeakers, and tape-recorded a message admonishing his future 30-year-old self to be devoting his life to the science of space exploration. When civil war broke out in 1975, he went abroad to study in France and the U.S. He earned his Ph.D. at Princeton and stayed on to become a full professor and director* of the university's Electric Propulsion and Plasma Dynamics Laboratory. (In 2009, Choueiri wrote an excellent article for Scientific American, "New Dawn for Electric Rockets" [PDF link], about the history, state of the art, and cutting edge of plasma and electric rockets. The magazine also posted an accompanying video produced by Space Channel France.)

Choueiri is a seriously committed audiophile, and his home setup features a vintage Studer master reel-to-reel tape recorder/player*, a monster VPI turntable, and a home-brewed version of his Pure Stereo filter. His ability to produce, in his listening room, a just slightly less amazing quality of 3D audio than his laboratory setup is a testament to the promise of universal accessibility his work offers. (You can even get a serious inkling of that sound via this video posted at the Princeton 3D Audio and Applied Acoustics Laboratory website.).

His remarkable music collection includes hundreds of vintage classical and jazz releases (including the very first commercially available stereo recording of a major performance, a pair of March 1954 sessions with Fritz Reiner and the Chicago Symphony playing Also Sprach Zarathustra and Ein Heldenleben), and a fantastic archive of second-generation master tapes. He played one of those masters, from an unreleased Bob Dylan session in March 1970: it was astonishingly vivid, especially with the help of the Pure Stereo filter.

Choueiri's intensive work on 3D audio has had one slightly melancholy outcome. Audio used to be his hobby, a way to relax, now it's come to be a part of his professional life. So he's now learning magic.

This post originally stated that Professor Choueiri was the founder of the Electric Propulsion and Plasma Dynamics Laboratory and that his home featured an Ampex tape deck. We regret the errors.

Thumbnail image: cliff1066/Flickr.