Paul Mitchinson is a part-time writer and a full-time father of two. He writes when he can. » more about me

Bouncing off the Walls

Lingua Franca, April 2001

ON OCTOBER 15, 1900, the Boston Symphony Orchestra performed a gala concert to inaugurate its new home, the now-legendary Symphony Hall. At the time, scientific and technological innovations in acoustics promised an era of unprecedented sonic magnificence, and Symphony Hall stood as a monument to the harmonious collaboration between art and science. Today, a bronze plaque hangs in the building. “Symphony Hall,” it boasts, was “the first auditorium in the world to be built in known conformity with acoustical laws [and] designed in accordance with…mathematical formulae.” Indeed, when the plaque was originally unveiled in 1946, modern science seemed at the height of its powers, capable of mastering everything from atomic power to the invisible, ineffable world of musical sound.

But the science of acoustics has since changed its tune. Though Symphony Hall is still ranked as one of the greatest concert halls in the world (alongside Amsterdam’s Concertgebouw and Vienna’s Grosser Musikvereinssaal), architectural acousticians are now more likely to be burned in effigy than memorialized in bronze. Ever since World War II, cities from Paris to London, from Toronto to New York, have fallen victim to multimillion-dollar concert halls that embody the latest “advances” in acoustic science yet sound little better than transistor radios. When acousticians failed to preserve the once-famed acoustics of New York’s Carnegie Hall during renovations in 1986, the New York Times critic Bernard Holland gave voice to a growing feeling of disenchantment when he declared that “acoustics are not a science, not even an art, but a roll of the dice.”

Did Symphony Hall really owe its success to “known conformity” with acoustic laws? Or was its effective design merely fortuitous? After a hundred years, the discipline of acoustic science has conspicuously failed to answer that question—though not for want of trying. If there is a single figure whose story conveys the soaring ambitions, shattering disappointments, and continuing ambiguities of acoustic science, it is Leo Leroy Beranek, an eighty-six-year-old electrical engineer who has been hailed as the field’s presiding genius and criticized as a technocrat with calculators in place of ears.

When thousands of music lovers make their annual pilgrimage to the Tanglewood Music Shed in Lenox, Massachusetts, they unknowingly appreciate the work of Beranek’s consulting group, Bolt, Beranek, and Newman (BBN), which transformed an unconventional outdoor space into a worthy summer home for the Boston Symphony Orchestra. When Soviet premier Nikita Khrushchev memorably banged his shoe on his desk at the United Nations General Assembly in 1960, it was BBN that ensured the sound would be sharp, precise, and dramatic.

But for all his accomplishments, Beranek will forever be associated with one of the greatest, and most public, acoustic disasters of the twentieth century: the opening of Lincoln Center’s Philharmonic Hall in 1962. Though the hall has since been adjusted and readjusted, gutted, rebuilt, and even renamed, few would consider Avery Fisher Hall (as the hall is now known) an unqualified acoustic success.

The Lincoln Center disaster drove Beranek out of concert hall design for many years. But in the mid-1990s, he made a much-publicized return to acoustic consulting, shaping the soundscape for a concert hall and an opera house in Tokyo Opera City. The acoustics of these halls have earned Beranek enthusiastic reviews from musicians and conductors, despite an experimental—and thus acoustically risky—architectural design that includes, among other anomalies, a pyramid-shaped ceiling. Could architectural acoustics at long last be coming of age? Has Beranek finally discovered, as one of his colleagues has claimed, the “Rosetta stone” of pure sound?

WALLACE SABINE, the Harvard physicist who masterminded Symphony Hall, founded the science of architectural acoustics in 1895. At the time, Harvard professors and students were complaining about the acoustics of a lecture hall in the newly built Fogg Art Museum. Sabine identified the problem immediately: excessive reverberation. “A word spoken in an ordinary tone of voice was audible for five and a half seconds afterwards,” he wrote in his Collected Papers on Acoustics. “During this time even a very deliberate speaker would have uttered the twelve or fifteen succeeding syllables.” The result was an incomprehensible muddle of verbiage.

For two years, Sabine experimented on the room, filling it with an array of cushions, carpets, and student bodies. Before long, he came up with a simple way of calculating reverberation, which helped solve the Fogg’s problem. Sabine’s formula states that reverberation time rises in direct proportion to a room’s cubic volume and in inverse proportion to the amount of sound-absorbing material. The type of material also matters, and he calculated absorption coefficients for a wide variety of objects, including plaster walls, windows, and hair cushions on upholstered chairs. He put his formula to good use when he became a chief consultant to the builders of Symphony Hall.

The Sabine formula remains, with certain modifications, the basis for modern acoustic science. But progress ever since has been frustratingly slow. More than sixty years after Sabine’s work on reverberation, a 1958 textbook on building acoustics admitted that reverberation time remained “the only acoustical quality that can be measured objectively.” In the 1950s and 1960s, it would be Beranek’s turn to advance the discipline.

BERANEK SHARES Sabine’s story with me during a gray January afternoon I spend with him in his three-thousand-square-foot condominium overlooking the Charles River in Cambridge, Massachusetts, just a few blocks away from Harvard Yard. Like Sabine, Beranek has close Harvard connections: In 1938, as a twenty-four-year-old graduate student in electrical engineering, he helped Professor Frederick Hunt invent the first lightweight phonograph pickup, enabling the development of long-playing recordings. In 1953, he became the first president of BBN. (His most celebrated accomplishment, perhaps, came in 1968, when BBN was awarded a million-dollar contract by the U.S. Department of Defense to create a crucial precursor to the Internet.) Regaling me with anecdotes and gossip from the worlds of engineering and music, Beranek takes me in a cab down to Symphony Hall, where we enter the building through a side door. As a former chairman of the Boston Symphony Orchestra, Beranek is a familiar presence here, and choruses of “Hi, Leo!” pursue us as we coast up the narrow staircase and step out onto the stage to admire Sabine’s handiwork in detail.

Beranek still seems in awe of Sabine’s accomplishment. He gestures eagerly around the hall, pointing out the architectural details that help create acoustic magic: the small stage area; the orchestra shell; the sound-diffusing balcony fronts, niches, and coffered ceiling.

How much of Symphony Hall’s magnificence is the result of science, I ask him, and how much of it is luck? “Well, they chose the right model,” says Beranek, alluding to the old Gewandhaus concert hall in Leipzig, “and Sabine’s formula enabled him to compute the reverberation time.” But he admits that serendipity played a part. For instance, the hall’s architects had wanted to create a fireproof building, so they constructed the walls out of concrete and plaster on wire lath. Fortunately, the solid construction allows the hall to reverberate nicely with a strong bass sound; thin wood walls would have caused the hall to swallow low notes. In addition, Beranek adds, “the stage enclosure was a pure guess, done by intuition entirely. But it worked.”

Intuition, as Beranek came to understand, has its place in a scientific approach to acoustics. After all, acoustic qualities are always to some degree a matter of subjective judgments. Musicians and conductors talk about the desirability of spaciousness, intimacy, strength, warmth, and clarity. In response, acoustic scientists use sophisticated microphones, speakers, and mathematical formulas in an effort to isolate independent variables—or “orthogonal parameters,” in the language of mathematics—that they believe correspond to these mysterious qualities.

Take reverberation time, for example. Audiences know that orchestras sound dry, ragged, and ill blended when performing in a space with little or no reverberation, such as a recording studio. With too much reverberation, on the other hand, swift contrapuntal passages sound blurred. But when the reverberation time is “just right”—about two seconds for the best halls, according to extensive polling of musicians and conductors—the strings begin to “sing,” and a good orchestra coalesces into a single instrument.

Following in Sabine’s footsteps, Beranek began to look for other orthogonal parameters that correspond to aspects of acoustic excellence. One of the first acoustic qualities he focused on was so-called intimacy. In an “intimate” hall, listeners feel “connected” with the performers, as though they were listening to the concert in a small room. Beranek suggested that intimacy was directly related to what he dubbed the “initial-time-delay gap” (ITDG). During an orchestral concert, direct sound reaching the audience is quickly followed by reflections from the sidewalls or ceilings. If these early reflections begin much later than twenty to thirty milliseconds after the initial impulse, he discovered, acoustic intimacy suffers. BBN’s design of a new orchestra enclosure and acoustic canopy for the Tanglewood Music Shed in 1959 incorporated the insights of this theory; the Shed was a huge success.

BERANEK SOON BECAME the preeminent figure in the field of acoustic science and was asked by Lincoln Center to be the acoustic consultant for Philharmonic Hall. To prepare himself for such a public commission, Beranek conducted a massive survey of fifty-four concert halls around the world, making detailed technical measurements of reverberation time, loudness, and the ITDG. More ambitiously, he tried to assess the relative contribution of each variable to the psychological perception of acoustic excellence. The results were published in his 1962 book, Music, Acoustics, and Architecture.

“I thought that Beranek’s book had really sewn up the question of concert hall design,” says Harold Marshall, professor emeritus of architecture and founder of the Acoustics Research Centre in Auckland, New Zealand. Even musicians were in agreement. In the book’s foreword, the esteemed conductor Eugene Ormandy gushed that “Lady Luck has finally been supplanted by careful analysis and the painstaking application of new but firmly grounded acoustic principles.” Beranek seemed optimistic about Philharmonic Hall as late as May 1962, when it staged its first “tuning concert,” conducted by a young Seiji Ozawa. “As the strings entered,” Beranek wrote in Music, Acoustics, and Architecture, “it was apparent that the hall would fulfill its designers’ great expectations and that the new home for the Philharmonic would be an acoustical success.”

But four months later, when Philharmonic Hall celebrated its official opening, Beranek’s reputation quickly fell to pieces. The orchestra sounded harsh and unblended, and the bass section was almost inaudible. The musicians were up in arms, unable to hear themselves on the stage. Harold C. Schonberg, the Times’s illustrious music critic, said that the conventional wisdom about Philharmonic Hall was that it was a “great big, yellow, $16,000,000 lemon.” George Szell, the celebrated conductor of the Cleveland Orchestra, was witheringly dismissive.

In hindsight, some of Philharmonic Hall’s failures were the result of ill-advised architectural decisions. Beranek’s original plan had called for a rectangle-shaped hall, but in order to increase the hall’s capacity, architect Max Abramovitz squeezed in additional seating by designing bulging concave sidewalls. Unfortunately, a concave surface is known to “focus” reflected sound on particular areas, causing echoes and other distortions. And Lincoln Center eliminated the sound-diffusing details proposed by Beranek. “They ran out of money,” Beranek explains. “If we’d had good diffusion, then the other problems could have been dealt with.”

But for all these mitigating circumstances, many problems did appear to be the result of Beranek’s recommendations—specifically, his efforts to ensure a short ITDG. Given the hall’s great width, most of these early reflections were produced by a series of reflecting panels hung low from the ceiling. “Leo had been very successful with overhead panels in Tanglewood,” explains Manfred Schroeder, a professor of physics at Germany’s University of G–ttingen. “So from that he took encouragement to introduce similar panels in Philharmonic Hall.” But Philharmonic Hall’s panels were smaller and more dispersed than those in Tanglewood. The long wavelengths associated with low-frequency notes “bent around the panels, got lost, and wouldn’t come down,” says Columbia University’s Cyril M. Harris, the acoustician responsible for Philharmonic Hall’s rebirth as Avery Fisher Hall in the 1970s. “It was a little like listening to a hi-fi set with the bass turned off.” Today, Beranek admits that it was a staggering professional humiliation.

THOUGH disastrous for Lincoln Center, Beranek’s failure provided a rare and valuable opportunity for other acousticians: They could test out their own theories by explaining what had gone wrong. Scores of researchers descended on New York, eager to perform an autopsy on a fresh acoustic corpse. Some acousticians discerned what became known as the “seat-dip effect,” a significant deterioration of bass reverberation as sound passes over rows of seats.

Schroeder’s insight about the failure of Beranek’s ceiling panels to reflect bass notes was one of the first inklings of the importance of what has since become known as “early lateral reflections.” In several academic papers in the 1970s, Harold Marshall and Michael Barron (a Ph.D. student at the time) expanded on this point, concluding that the subjective quality of “spatial impression” in music is directly related to reflected sound from the sidewalls. Such reflections make listeners feel “enveloped” by the sound of an orchestra, which seems to surround the audience rather than emanate from a small fixed point on the stage.

In 1972, Marshall incorporated these discoveries into the arena-shaped Christchurch Town Hall in New Zealand, in which lateral reflections were created by low reflecting walls that separated the audience into blocks. The innovative design proved an acoustic success. The precise relationship between lateral reflections and a listener’s sense of spaciousness has yet to be established with confidence, and several different orthogonal parameters have been proposed. But the effort to create lateral reflections became popular around the world. The architectural implications, for the most part, have been conservative. “One of the justifications for a parallel-sided hall is that if you keep things narrow enough, you’ll get lateral reflections,” explains Barron, who now lectures at the University of Bath. Wide fan-shaped halls were out; standard rectangular halls with narrow sidewalls—often called shoe boxes—were in. As long as one reined in the architects, Marshall and Barron implied, good sound could be guaranteed by good mathematics.

NOT EVERYONE responded to the debacle at Philharmonic Hall with such optimism. While Schroeder and Marshall leaped at an opportunity to improve the science of acoustics, devising new measurements and formulas, others threw in the towel. Russell Johnson, an acoustician who joined BBN in 1954, helped the firm design several acoustically unsatisfactory halls—including Philharmonic Hall. The experiences made him increasingly disenchanted with the scientific attitude in acoustics. “I started to say to myself that the established way isn’t working,” Johnson says over the phone from the offices of Artec, his New York-based consulting firm. “Scientists and engineers devised rules based on mathematics and formulae about the physics of sound. But most of these scientists did not love music, did not understand music. They weren’t really interested in music, and because they knew so little about music, when they opened a new hall and went to listen to it, they immediately declared it perfect.” His conclusion? “I had to teach myself a new way to design these things.”

In fact, as he readily admits, Johnson taught himself a very old way to design halls—a tradition that he says had been lost during the first half of the twentieth century. His philosophy, which is based on close consultation with musicians themselves and on a cautious architectural aesthetic, is shared by Harris, another acoustician known for his conservative approach. Harris thinks that the scientific approach to acoustics has been grossly overrated. “It helps, I suppose, to sell a job. If several consultants are being considered for a job, many architects might be sold on the fact that ‘Oh, I have a computer. I can check all this stuff out ahead of time.’ That gives them a certain comfort.” Johnson has been particularly successful with his more intuitive approach. Artec, which he established in 1970 after leaving BBN, has grown into one of the most well regarded acoustic consulting groups in the world. Its many critically acclaimed halls include the Eugene McDermott Concert Hall in Dallas and Symphony Hall in Birmingham, England.

But Johnson’s conservative approach has also infuriated some architects. Last year, he was commissioned to improve the acoustics of Toronto’s Roy Thomson Hall, a large semisurround auditorium. Johnson’s plan to narrow the upper part of the hall has enraged its original architect, Arthur Erickson. In an Op-Ed in Canada’s National Post in November, Erickson was highly dismissive of Artec’s faith in the shoe-box-shaped hall: “In the intangible pseudo-science of acoustics, where so much is dependent on the subtleties of personal experience, I question the wisdom of proceeding on a course that contradicts the whole basis on which the hall was based.”

BERANEK, FOR HIS PART, assumes that architectural innovation will press onward despite the misgivings of acousticians like Johnson. And so he continues to view architectural challenges as opportunities for new acoustic research. The number of acoustic qualities that he believes can be expressed mathematically has now increased to half a dozen or more, as enumerated in his 1996 book, Concert and Opera Halls: How They Sound. Some variables, such as warmth, can be measured through a simple variant of Sabine’s original equation, using a ratio of reverberation times for lower- and higher-frequency sounds. Others, such as spaciousness, can be calculated by entirely new formulas, such as the “interaural cross- correlation coefficient” (according to Beranek and several Japanese researchers) or the “lateral fraction” (according to Marshall and Barron). Spatial effects, in fact, are now believed to have several different components, measurable by different formulas. Diffusion continues to resist mathematical expression; it has to be estimated visually, by eyeballing fine-scale and large-scale architectural irregularities. But Beranek continues to insist on the primary importance of intimacy, measurable by the ITDG.

Besides reverberation time and the ITDG, none of these variables can be calculated on the basis of architectural plans alone. Even computers provide only crude estimates. (“The computer,” says Beranek, is only “an added means of convincing architects that you’re right.”) Technology has allowed some progress, however. Miniature microphones can be embedded in architectural models, with useful results. It remains difficult to duplicate the absorbing effect of an audience, but this, too, has improved.

Beranek has made use of these recent innovations in his work in Japan. For Tokyo Opera City’s concert hall, he established a range of acceptable values for each parameter and tried to duplicate that range in the model. The architectural hurdles were significant. The architect insisted on covering the interior surfaces with wood, which is known to absorb significant bass; as a result, Beranek decided to eliminate any additional absorbent material, including carpeting, from the hall. A dramatic pyramid-shaped ceiling might have caused any number of acoustic irregularities, but these were compensated for by the use of diffusers and a large over-stage reflecting canopy of precisely calculated dimensions. “What has been accomplished is a miracle!” commented the cellist Yo-Yo Ma in a letter to Beranek. “This hall simply has some of the best acoustics in which I have ever had the privilege to play.” The pianist András Schiff agreed, praising the hall’s “warm, round, and reverberant” sound.

Beranek will face an even greater challenge this summer when he tunes Daiichi-Seimei Hall, a new auditorium near Tokyo. The interior walls are concave, a feature that often leads to catastrophic acoustic imbalances, as in Philharmonic Hall. Beranek seems unfazed. “We had to add acoustical materials in certain places to cut down on the reflections,” he explains. But because those modifications tend to limit reverberation, Beranek had to make certain that the ceiling height in the original design was sufficient to ensure optimal reverberation time. The tinkering seems to have resulted in a hall with acoustics as good as any traditional rectangular hall. At least, that’s what measurements taken in the model suggest. Only in June, during the hall’s tuning concert, will Beranek know for sure. If he succeeds in making a hall with bulging sidewalls sound good, it will be a significant step toward personal redemption from the Philharmonic Hall experience.

BUT ARE BERANEK’S Japanese halls significant steps forward for science? Many authorities are skeptical. “I read the professional papers coming out of that project,” comments Russell Johnson. “I feel that quite a bit of art went into that project, even though they told reporters that it was 99 percent scientific.” Even some pure scientists are wary. “It seems just a bit too good to be true,” says John Bradley, an acoustics expert at the National Research Council in Ottawa, Canada. “Beranek and the Japanese seem to focus on particular measures being the magic single measure. That’s how he ran into trouble with Philharmonic Hall. It’s more complicated than that; concert halls are multidimensional problems. If you focus on just one thing, you might get some of the others wrong.”

Scientific advances always involve risks, experiments, and guesses. And the only way of advancing knowledge, as Marshall explains, is by “confronting a new problem that doesn’t have a precedent.” Most scientists are fortunate enough to confront these problems in dim windowless laboratories, well away from the public eye; acousticians are not so lucky. Despite a century of research, acoustic science remains something of a high-profile crapshoot. Leo Beranek, ever the gambler, may be just a little better than most at loading the dice.