What does 'perfect sound' look like? Inside the bizarre science of concert hall design

Zackery Belanger

When Charles Garnier designed the Paris Opera in the 1870’s, he said that architecting the acoustics of a room was a “bizarre science.” Garnier compared the acoustician to an acrobat “who closes his eyes and clings to the ropes of an ascending balloon.”

More than a century later, how sound moves through a room is still sometimes a mystery.

At the spherical Mapparium, a so-called “whispering gallery” in Boston, sound sometimes behaves in strange ways. In some parts of the room sound seems to come from strange directions and it appears to double in volume if two people stand precisely two meters from the room’s center. Nobody is precisely sure why.

The Disney Concert Hall in Los Angeles is considered one of the finest acoustic spaces in the world, its sound architecture hailed as “perfect.” Yet when the acoustician behind it, Yasuhisa Toyota, was asked how he achieved that perfection, he said that after making small models to test sounds and make adjustments, “all we can do is just pray.”

Architectural acoustics as a discipline was founded at the end on the 19th century, when a young Harvard physics professor named Wallace Clement Sabine sought to understand why the students in a new lecture hall could never understand what their professors were saying. Sabine discovered there was too much reverberation. If one syllable was spoken too quickly after another, their sounds would collide as they bounced off the ceiling, the floors, the walls, the furniture and people in the room. He proposed a solution: adding sound-absorbing materials to the walls. Afterwards, at long last, the students could hear.

But today we still don’t really understand all of the interactions between sound and a surrounding room. There are just so many variables. What happens to the behavior of sound if you alter a room’s wallpaper or carpeting or if guests are dressed in winter bundles rather than summer linens? How do you accurately measure the quality of sound, which changes as it moves across both time and space?

With an infinite budget, architects and engineers can build the longest bridge span or the tallest building. But there’s no way to guarantee the best sounding concert hall, because we still don’t really know enough about what makes one good. You can make models and test sounds and control generally for factors like a lot of background noise, but at the end of the day a certain amount of an acoustician’s job is a crapshoot. It’s why the Avery Fisher Hall at Lincoln Center in New York, which was designed to be the pinnacle of acoustics when it opened in 1962, has been gutted again and again in an effort to try and find the right sound.

There are all kinds of theories about what makes a room sound good. One of the leading researchers in the field compares concert hall acoustics to tasting wine: some characteristics are easyily classifiable, but perception and taste are part of the equation, too. Another has suggested that understanding the way sound reflects in a room, not the shape of it, is the critical component to good concert hall design.

But the craziest theory comes from a researcher named Zackery Belanger. He thinks that acoustics is primarily a geometric problem, a theory so radical that he was forced out of his Ph.D. program because his advisor disagreed with it.

“Most people think you just do some computer simulations and out comes the answer,” he told me. “It’s as if we started off with a plane that flew really well and we didn’t understand the science beneath it.”

Right now, the main way places like concert halls are studied is by measuring decays in sound. Researchers play a test sound from a stage and then place microphones around a room to try and understand what happens to the sound when it travels.

Belanger thinks that’s all wrong. He believes that it may be possible to predict almost everything about the way a room will sound by how much its surface area deviates from the surface area of a perfect sphere, like Boston’s Mapparium. The idea here is that the total surface area of a room is more important to the way sound waves bounce around it than any particular characteristic on its own. So you can swap out chairs or carpeting or puffy jackets and it doesn’t really matter so long as it all adds up to the right number.

Belanger is obsessed with figuring out this equation. When I first met him a few years ago, he had just quit his job at a major acoustic design firm and moved across the country to study EMPAC, a renowned performing arts center at the Rensselaer Polytechnic Institute in Upstate New York.

He measured the space’s volume and surface area with a laser scanner to try and figure out how tweaking the surface area might change the sound. His initial data has been promising: in one experiment, he found that changing the room’s shape had the opposite effect that it should have based on current acoustic principles. When he removed a set of panels that were meant to diffuse sound, which decreased the room’s surface area, the sound in the room behaved as though he’d removed panels that were meant to absorb sound, not reflect it. Initially, even Belanger had been skeptical of his crazy hunch, but it seemed like there might actually be something to the idea. Other research he’s done since then, he told me, has backed it up.

“People have been asking for a long time why we don’t have this fully figured out,” he told me. Belanger has since scanned more buildings and plans to eventually publish his results.

Our approach to acoustics, Belanger said, has been too archaeological, looking back at what has worked in the past rather that experimenting with the new. He thinks that maybe we have been looking for answers in the wrong places.

“The most advanced acoustic research right now is just using more microphones,” he told me. “But why do we need to use sound to measure sound?”

Part of the reason we still know so little about all of this is that acoustics is a really, really small field. An architect’s main concern is usually designing buildings that look different and keep the water out and meet code, not spaces that are sonically unique. The Kimmel Center in Philadelphia was designed to mimic the shape of a cello, an idea rooted more in artistic whim than sonic science. Data on what works and what doesn’t is limited, in part because each space acoustic researchers study is architecturally distinct.

Belanger, though, is a bit of a rogue operator in the field. He’d rather publish his work on Medium than in any academic journal.

A leading acoustic researcher Tapio Lokki, the one who compared his work to wine-tasting, told me that he agrees with Belanger that the way the field has typically measured the characteristics of a room is lacking. In fact, at an acoustics conference in Paris, he recently proposed shifting away from this, and focusing instead on what humans hear.

“The primary problem is that we do not understand well enough the human hearing system,” he told me. “My recent thinking is that [a] concert hall is considered good if it supports the music as well as possible. The purpose of music is to evoke emotions and to touch the listener and the role of the hall is to make music as expressive as possible.”

He told me Belanger’s idea is interesting, but also probably too simplistic.

“It assumes that we need a diffuse field to have good sound,” he said.

Another acoustician told me that Belanger’s work simply isn’t grounded in any scientific evidence.

But that’s Belanger’s whole point — that the answer to all of this probably is simple and radically different from the way we currently thing about sound.

Acoustics, he admits, will always be a little bit about art, because the way we each interpret what we hear is different. But he thinks it should be more about science. He thinks we might be missing something fundamental.

“I studied physics in college. Even complicated things tend to have really simple, elegant underlying structures,” he told me. “That’s the way physics works.”

Belanger is the first to admit that his idea is more than just a little out there. When he first started experimenting, he didn’t even really believe it would pan out himself. And even if it doesn’t, he hopes he can move others to search for answers beyond the veil of the obvious.

But if it is true, his theory could be the acoustic equivalent of discovering that the world is round. He imagines a future where you could easily design everything in a concert hall to plug into the right sonic equation, where you actually could tell a computer what you wanted a space to sound like and it would simply spit back out a design.

One day, perhaps we’ll 3D-print perfectly tuned concert halls. Until then, a building’s sound remains one of modern time’s stranger unsolved mysteries.

A version of this piece was presented on November 7, 2015 at the Real Future of Sound at our Real Future Fair in San Francisco.