The Sound Book: The Science of the Sonic Wonders of the World

by Trevor Cox

  Intriguingly, Arnott’s story of the sailing ship mentions the weather as being a crucial part of the tale. The report might well be correct if a temperature inversion helped direct sound to the concave sail.

  A few years ago I presented two science shows at the Royal Albert Hall in London to thousands of children. Though better known as a music venue, the hall is actually dedicated to the promotion of art and science, and it was built on land purchased with the profits of the Great Exhibition of 1851. For an amateur performer like me, a complex science show is a daunting challenge, made all the worse in this case by the vastness of the arena. Fortunately, the acoustics have been significantly improved since the hall opened 130 years ago. Indeed, the Prince of Wales struggled with his opening speech. According to the Times (1871):

  The address was slowly and distinctly read by his royal Highness, but the reading was somewhat marred by an echo which seemed to be suddenly awoke from the organ or picture gallery, and repeated the words with a mocking emphasis which at another time would have been amusing.9

  The hall’s ubiquitous curved surfaces are probably what caused the mocking echoes. From above, the floor plan appears as an ellipse, and the whole structure is topped with a large dome. The curved surfaces focus sound like Kircher’s elliptical ceiling. But how such reflections are perceived depends on the size of the room. In the vast Royal Albert Hall, the curves cause disastrous echoes. Sound appears to come from several places in the room and not just the stage. In a small room the focused sound arrives quickly; in a larger room the reflections are delayed.

  You can test this phenomenon out with a friend.10 Find a large open area with one big reflecting wall, such as a large building at the side of a park, or the side of a quarry. A quiet place away from noise is best. For the exercise to work, you need to hear the sound bouncing off the wall with few reflections from other surfaces. The wall does not need to be curved if it is large enough. If you and your friend stand some distance apart but each of you is the same distance from the wall, then the effect is more obvious. The test would work particularly well on a snowy day, when the sound reflecting from the ground would be absorbed by the snow and all traffic would have ground to a halt.

  Walk toward the wall while chatting with your friend, and at some point you will become aware of the reflection from the building. As you get closer, you will hear the echo from the wall becoming louder because the reflection has traveled a shorter distance. But then as you approach even nearer, starting about 17 meters (56 feet) away, the sound of the reflection will appear to gradually get quieter until, at about 8 meters (26 feet) from the wall, it appears to disappear altogether. It is still there, but it is no longer being heard separately: your brain has combined it with the sound traveling directly in a straight line from your friend.

  The way the brain combines sounds is important, because otherwise we would rapidly become overwhelmed by the vast number of reflections that accompany us. As I type this sentence, the rattle of the keyboard is being reflected off the desk, the computer monitor, my phone, the ceiling, and so on. Yet my hearing is not overwhelmed by all these different reflections; the sound still appears to be coming straight from the keyboard as it should be.

  The same thing happens in Kircher’s small room. The reflections from the elliptical ceiling arrive quite quickly, and unless the reflections are very loud, the brain does not hear them as separate from sound traveling directly between talker and listener. By contrast, the Royal Albert Hall is so vast that the focused reflections arrive much later, creating “mocking” echoes.

  Acoustic engineers made various attempts to remove the echoes from the Royal Albert Hall. The most successful solution was to add the “mushrooms” hanging from the ceiling. Originally installed in 1968 following an idea by Ken Shearer of the BBC, these large discs hang down at the base of the dome and reflect sound away from it.

  Although it is no longer possible to appreciate (or not) the echo from the hall’s ceiling, there are plenty of other domes to explore. A few miles from my home, for example, is Manchester Central Library, which has a great dome with a focal point that used to be close to the microfiche machines. Every time a glass plate fell to cover the microfiche, a startlingly loud echo resounded from the ceiling.

  The library is currently closed for renovation. One hopes that the building work will not be as sonically unsympathetic as the renovation of the US Capitol in the nineteenth century in Washington, DC, which ruined a fine ceiling echo from a famous whispering dome.11 The Capitol’s dome used to be an almost perfect hemisphere centered at the head height of visitors, and although the ceiling appeared coffered with indented squares, it was actually smooth, with trompe l’oeil painting creating the illusion of structure and texture. Before 1901, this domed space was a great draw for tourists. According to the New York Times in 1894:

  The whispering gallery still holds the palm among the show places of the great marble structure. Once in a while an old resident of Washington is initiated into the mysteries of the echoes and other acoustic phenomena which abound in this old-time chamber, and he feels a little ashamed of his tardiness in seeking this remarkable entertainment.12

  But while it was great fun for the tourists, it was a poor place for the House of Representatives to hold debates. As the Lewiston Daily Sun put it in 1893:

  The orator who was not cautious enough to remain in one spot during the delivery of his address found the acoustics of the hall taking strange liberties with his elocution, transforming his crescendo sentences into comical squeals, or causing his pianissimo phrases, his stage whispers, to shriek and wail as he moved to and fro from one echo point to another.13

  A gas explosion and fire in 1898 elsewhere in the building led to replacement of the wooden dome by a fireproof construction. Real plaster coffers were installed in place of the trompe l’oeil, making the effect of the focus weaker and less remarkable. As the eminent acoustician Lothar Cremer remarked, “To everyone’s dismay, the famous focusing effect was remarkably reduced, since the precise geometric reflection had been supplanted by a fuzzier diffuse reflection.”14

  Changing from a smooth surface to one covered in lumps and bumps is rather like taking a perfect optical mirror and scratching it badly or adding a frosting. The irregularities of the surface cause the light or sound to be scattered away from the focal point. For an optical mirror, the result is a fuzzy image; for the Capitol’s dome this scattering weakens the sound reflections, so whispers are no longer remarkably loud and voices are less distorted.

  Figure 5.3 Diffuser designed for the curved wall of the National Museum of the American Indian.

  The effect of the coffers on the Capitol’s focus reminds me of an engineering project I worked on a few years ago. I designed diffusing surfaces for the large circular Rasmuson Theater at the National Museum of the American Indian in Washington, DC. To keep the curves from causing focusing and creating mocking echoes, I designed a bumpy surface that, like the coffers in the Capitol’s dome, scattered sound in all directions away from the focal points. A cross section through the diffuser looked like a city skyline (Figure 5.3). As sound waves meet the diffuser, the irregular heights of the blocks force the reflections to go off in different directions.

  My innovation was coming up with a way of determining where to place the “skyscrapers” and how high to make them. I used a trial-and-error process in which a computer program tries out many different skylines. For each configuration, the program predicts how sound will reflect from the surface and evaluates whether the focus for the curved surface will be removed. The program keeps rearranging the skyscrapers until a good design is found. This iterative process, known as numerical optimization, has been used in many branches of engineering, including designing parts for the space shuttle. One of the reasons this method is so powerful for acoustic diffusers is that it allows surfaces to be designed that fit the visual appearance of the room. The acoustic treatments do not have to look like ugly add-ons. Curves, s
kylines, pyramids—whatever shape the architect desires—the method can be tasked with finding features that have the best acoustic performance.15

  The great thing about a dome is that if you stand right under the centre and clap your hands, you will be deafened a moment later by the echo of your hands clapping. If you say, in mock horror, “A HANDbag?”, you will hear the echo of Dame Edith Evans a second later, from the heavens.16

  This is the journalist Miles Kington encouraging you to unleash your inner Lady Bracknell from The Importance of Being Earnest by Oscar Wilde. But while domes are fun, a completely spherical room is even better because reflections are amplified even more.

  The Mapparium in Boston is a 9-meter-diameter (30-foot) sphere and was built in 1935 following a suggestion by architect Chester Lindsay Churchill. It is a giant hollow globe of the world, with the seas and continents vividly drawn on stained glass. It took eight months to paint and bake all 608 glass panels, which are mounted on a spherical bronze frame. Visitors traverse a walkway cutting through the center of the Earth linking up two opposite points on the equator. Three hundred lightbulbs illuminate the globe from the outside. Looking at the world from the inside out is an odd experience, but what also strikes visitors are the strange acoustics, which were an accidental by-product of the geometry.

  William Hartmann, from Michigan State University, and colleagues documented the various illusions that can be heard. Usually, moving farther away from a listener makes a talker’s voice quieter, but that’s not always the case in a spherical room. Imagine, Hartmann writes, “you are on the Mapparium bridge two meters [about 7 feet] to the left of dead center. Your friend is exactly at the center and is talking to you. His voice seems rather quiet. Now your friend walks away from you, and his voice gets louder and louder until he is about two meters to the right of dead center.”17

  The sketches in Figure 5.4 show what is happening (to make it easier to see, this drawing uses a circle rather than a complete sphere). When the talker speaks from the center (top diagram), all the reflections are focused back, so the talker appears to be surprisingly quiet to the listener left of center. If the talker moves to the right, then the place where the reflections focus moves closer to the listener. The sound will be loudest when the talker and listener are arranged symmetrically about the center (bottom diagram).

  Figure 5.4 Focusing in the Mapparium.

  Although this effect is particularly strong in the Mapparium, which has curved surfaces above, below, and in front of the speaker, given the right structures it can even be heard outdoors, as reported by José Sánchez-Dehesa from the Polytechnic University of Valencia in Spain.18 The Cempoala (Zempoala) archaeological site near Veracruz in Mexico is one of the most complete surviving examples of an Aztec ceremonial center. It is a large grass plaza dotted with the remains of various structures, including a low, circular enclosure topped with merlons. The guidebook hedges its bets about the purpose of the structure: “associated with gladiator worship of Mexica[n] (Aztec) origin, although it may have served as an intake for rainwater.” In pictures, the structure looks like a stone sheep enclosure made of large, round pebbles. Whatever the purpose of the circle, it displays an acoustic focus. Sánchez-Dehesa reports that if a listener stands in the right place while a companion walks and talks along a diagonal through the circle, the sound gets louder as the speaker moves farther away.

  Hoping to visit a spherical room, I sought out urban explorers, people who like to illicitly explore sewers, abandoned subway stations, and derelict buildings, enjoying the visceral thrill of trespassing in ghostly places where others do not go, and finding hidden histories and traces of previous occupants.19 One member of Subterranea Britannica, a group dedicated to the legitimate exploration of underground spaces in the UK, sent me an e-mail describing a dome in Berlin that was one of the most important listening stations for the West during the Cold War. The enclosed photo showed the dome on top of a derelict tower; this was a place I had to visit.

  The abandoned spy station sits atop Teufelsberg (“Devil’s Mountain”; Figure 5.5), which rises up from the Grünewald forest. As I walked up through the woods on a hot summer’s day toward the listening station, it seemed unbelievable that this large hill was man-made. It was constructed from millions of cubic yards of rubble created by bombing raids and artillery bombardments during World War II.20

  Figure 5.5 Teufelsberg.

  Before entering, I had to sign a liability waver because the derelict buildings have many holes and missing walls with no protection from the shear drops below. My German tour guide was Martin Schaffert, a young history scholar sporting a neat beard, small ponytail, glasses, and a flat cap. As Martin explained the history of the site, I surveyed what remained of the buildings. Doors and walls were missing, with the debris of the crumbling buildings lying on the floor mixed with broken glass from semi-illicit parties. Where walls were intact, they were covered in graffiti. My eye was drawn upward, where on top of the main building were three domes; some were vandalized, with their walls partly in tatters, but the topmost one, sitting on top of a five-story tower rising out of the roof, was intact.

  These were radomes (a contraction of the words radar and domes), used to hide the spying activities from prying eyes as the British and Americans listened in on broadcasts and wireless communications from East Germany, Czechoslovakia, and the Soviet Union. The spherical domes were also used to protect the listening equipment from the elements, especially wind and ice. All that remains now are the concrete plinths to which the antennae were bolted. The domes were constructed of triangular and hexagonal fiberglass panels stretched between a scaffold, looking like giant soccer balls. Fiberglass is transparent to electromagnetic waves and so ideal for radomes; this was one of the reasons the material was developed during World War II.

  The tower supporting the top dome was missing all its walls, but the dome itself was virtually complete because it had been rebuilt and reused for air traffic control over Berlin. Every surface of the dingy stairwell up the center of the tower was coated in graffiti. As I climbed the stairs to the dome, I could hear the reverberated voices of other visitors enjoying the acoustics. With a midfrequency reverberation time of about 8 seconds, the radome sounds a bit like a cathedral. Musicians come to perform music in the space. But there is more to enjoy than just the reverberance.

  Once inside, I stood and watched other visitors; it was wonderful to see their faces light up when they realized how odd the acoustic was. The slightest sound, even a simple footstep, created a ricocheting effect. Some were prompted to explore with gusto (a good hard stomp could be heard to repeat eight times, sounding like distant firecrackers), while most contented themselves to playing more subtly, almost treating the space with the same reverence as they would a place of worship.

  I then climbed onto the old antenna plinth to get to the center of the room. The dome is roughly the top two-thirds of a sphere about 15 meters (50 feet) in diameter, constructed from yellowing hexagons. A band of graffiti 2 meters (about 7 feet) high ran above the floor, interrupted only by a second small opening, through which was an unprotected shear drop to the roof of the building five stories below. I got out my recorder to dictate my impressions of the place, and noticed how every word was doubled up by the reflections from the radome.

  Here in Teufelsberg I wanted to explore an effect that happens in the spherical Mapparium.21 The unusually strong focus makes it possible to experience the strange sensation of whispering into your own ear. Or, as Hartmann puts it:

  As you approach the exact center of the Mapparium sphere you suddenly become aware of strong reflections of your own voice . . . If you sway to the left, you hear yourself in your right ear. If you sway to the right, you hear yourself in your left ear.22

  In Teufelsberg the effect is strongest if you look upward while whispering, because there is a larger concave area to focus the sound. Thus, with my head tilted backward, five stories up in a fiberglass radome in Berlin, I discovere
d an exciting binaural sonic wonder, an effect that reveals how we work out where sound is coming from. Having two ears enables mammals to perceive the locations of sound sources. Hearing evolved to enable animals to sense danger, alerting them to predators sneaking up and trying to turn them into lunch. Humans have good vision, but our eyesight cannot detect threats from behind, so being able to hear and locate danger is crucial.

  There are a couple of main ways in which we sense where sound is coming from. Imagine that someone is talking to you from your left side. The sound arrives first at your left ear because it takes slightly longer to reach your right ear. Your brain also makes fine distinctions in loudness. The sound has to bend around your head to reach the right ear, causing it to be much quieter at high frequency. (The loudness of low frequencies coming from far away is minimally affected by your head itself). Your brain compares the timing in each ear at low frequency and the relative level of high frequencies to decide where the sound is coming from.

  Spherical rooms can mess with both of these cues. The loudness cue can be distorted, creating an unexpected localization. It makes sound appear to come from the wrong direction, as Hartmann describes: “Suppose you are on the Mapparium bridge facing South America. There is a source of noise to your right, but you discover that you hear the noise coming from your left!”23 Strong, focused reflections from the sphere create a loud reflection at the left ear, fooling the brain into localizing the noise on the left side.

  Localization is usually based on the first sound arriving at the ears (the precedence effect). This rule of thumb serves us well because the earliest sound takes the quickest route, which is usually a straight line between talker and listener. You may have sat through a church service in which the sermon appeared to be delivered by the loudspeakers rather than by the preacher. The reason for this impression is that the sound from the loudspeakers reaches the listener first. Adding a little delay electronically in the public address system, so that the first sound wave to reach the listener comes directly from the preacher’s lips, solves this problem.