by Trevor Cox
Most of the large sonic wonders that I have discovered were accidental, but what sounds could be made if we really tried? What shapes could exploit the physics found in accidental sonic wonders to create new aural effects? One could take inspiration from the seventeenth-century Jesuit scholar Athanasius Kircher. In addition to writing about the infamous cat piano, he imagined and drew fantastical acoustic contraptions such as speaking statues and a music-making ark to mechanize music composition. Perhaps modern inventors should come up with contemporary versions of these sketches.
While traveling the world looking for sonic wonders, I have begun dreaming up a few of my own. Investigating the distortions from the spherical radome in Teufelsberg made me think of a fairground years ago when I played with fun-house mirrors. One was curved in a way that turned me into a distorted goblin. Another was bent differently, creating a disfiguring reflection that stretched out my legs so that my torso almost disappeared. Could I design a whispering gallery using a complex curve? Such a design would be new; the whispering galleries and walls I have discovered have all been simple arcs, curves, or domes.
In his paper on whispering galleries, physicist C. V. Raman describes Golghar, the old government granary in Bankipore, India. Built in 1783, it is shaped like a beehive, and at 30 meters (100 feet) high there are great views to be seen from the top. Raman notes of the sound inside, “As a whispering gallery there is perhaps no such building in the world. The faintest whisper at one end is heard most distinctly at the other.”37 But it was photos of the outside of the building that intrigued me. An external spiral staircase hugs the outside like an old-fashioned amusement-park slide. If the outside wall of the staircase were altered to gently curve, sound could spiral up the walls of the staircase. It would be a whispering “waterslide,” to complement the whispering gallery inside.
The computer program that produced animations of St. Paul’s Cathedral enables me to design odd-shaped whispering galleries and test whether they will work. This is how acoustic engineering is done nowadays. Before places are built, computers are used to check whether an actor will be heard clearly by the audience, or whether a public address system in a railway station will produce intelligible announcements. Thus I turned my engineering skills and scientific knowledge on their head. Instead of designing to remove acoustic aberrations caused by curved surfaces, I used the same tools to maximize aural distortions.
Figure 5.11 New whispering walls.
Some of Richard Serra’s artworks in the Guggenheim Museum in Bilbao, Spain, are giant steel walls that behave like whispering walls. Taking inspiration from these works, I thought about the potential design. I wanted whispers to hug an S-shaped curve like a fun-house mirror, but unfortunately, sound does not hug a convex section. I solved this problem by forming the S shape from two arcs (Figure 5.11). With that arrangement the sound passes around the inside of the first curve and jumps the small gap separating the two sheets before skimming around the inside of the second bend to be heard by the listener as surprisingly loud.
The delight in these places comes from hearing a voice carry an unexpected distance, and this effect is more dramatic if the sound is a quiet whisper to begin with. Later mathematical analysis by Lord Rayleigh suggests another reason for whispering: high frequencies, like the sibilant tones in whispers, hug the walls closer than the lower-frequency sounds of normal speech. It also seems that the St. Paul’s gallery is particularly adept at transporting sound. Acousticians have attributed this feature to the slight sloping of the walls. By tilting the walls inward at the top, less sound goes upward and is lost to the top of the dome.
I now had the answer for why the sewer had made my voice spiral. While the analysis by Lord Rayleigh shows that the hugging effect in whispering galleries is more pronounced for larger circles, the theory also demonstrates that the same circulating effect happens in smaller places, even tunnels only a couple of yards across. With my head close to the ceiling of the sewer, my speech simultaneously disappeared down the tunnel and skimmed and circulated around the curved walls as happens in a whispering gallery. I was not hearing an audio illusion in the sewer; the sound was genuinely spiraling.
Singing Sands
A
year after my visit to St. Paul’s Cathedral, I traveled to Kelso Dunes (Figure 6.1) in the Mojave Desert, California, with sound recordist Diane Hope in the hope of hearing a sand dune sing. Kelso is one of about forty documented sites where this happens.1 The English naturalist Charles Darwin recounts tales of the “El Bramador” hill in Chile, which was called the “roarer” or “bellower” by locals.2 Ancient Chinese writings describe festivities at the Mingsha Sand Dunes: “It is customary on the tuan-wu day (the Dragon festival on the fifth of the fifth moon) for men and women . . . [to] rush down . . . in a body, which causes the sand to give forth a loud rumbling sound like thunder.”3
A sand avalanche starts the dune singing. The slope has to be steep, and the sand has to be very dry. But dry sand is by definition the loosest, which meant my feet struggled to find purchase on Kelso’s surface. I had prepared for the extreme temperatures of the summer desert, but I had not realized that searching for musical dunes was going to be an aerobic workout. I wanted to gasp for air as I struggled up the dune, yet I had to hold my breath so as not to ruin the audio recordings.
As I trudged up the sandy slope, my feet made a burping noise in the sand. It reminded me of the first part of Marco Polo’s description. He wrote that dunes “at times fill the air with the sounds of all kinds of musical instruments, and also of drums and the clash of arms.”4 I was getting nothing dramatic like drums, but I was creating something musical. Each labored footstep created a single honk, like a badly played tuba. Toward the top of the slope I got so tired that I resorted to scrambling up on all fours, producing a comical brass quartet.
While the burping was entertaining, I was frustrated because the dunes were not in full voice. What I had traveled to hear was a sustained boom that supposedly reaches 110 decibels, resembling the loudness of a rock group, audible up to a mile away.5 It was getting toward late morning. The wind was making recording difficult and the heat was becoming unbearable, so we retreated down the dune to try again the next day.
Figure 6.1 Kelso Dunes.
Back at the campsite, I listened to a recording of my phone interview with Nathalie Vriend from Cambridge University, who had studied the singing sands for her doctorate. I was looking for hints of how to find the best place on the dunes. Worryingly, in the middle Nathalie mentioned that a friend had visited Kelso in recent times and had been disappointed by the sound. I also reviewed key scientific papers. I hoped that, by understanding the physics, I might have a better chance of getting the dunes to properly sing the next day. While scientists agree that burping sand is a necessary ingredient, there is a heated debate about what actually causes the loud booming. Is a deep layer of the dune vibrating like a giant musical instrument? Or are the sand grains locked in a synchronized avalanche?
During a proper boom, thousands of sand grains sing in a coordinated choir stretched over many yards across the dune. Waterfalls are similarly made from an extended orchestra, but the instrumentalists are tiny bubbles. The loudest waterfall I have ever heard is Dettifoss on the glacial river Jökulsá á Fjöllum in Iceland; it is also the most powerful in Europe. Many years ago, my wife and I cycled there on a bitterly cold and unpleasant morning. The road was more like a track, rough and potholed, and the northerly headwind poured cold air down from the Arctic with such strength that it brought us to a standstill at times.
We made tortuous progress, first through winding moorland and then through the sandur—a desolate landscape of glacial outwash and black volcanic silt where little grows. Leaving the bikes behind, we gingerly walked to the cliff edge overlooking the falls, which are more than 100 meters (325 feet) wide and 44 meters (145 feet) high.6 Standing very close to the precipice, a pang of fear swept through my body: with 180 cubic meters (6,000 cubic fee
t) of water pounding over the edge per second, a slip now would end in certain death. The incessant, thunderous power meant we had to shout to be heard. The noise seemed to cover every frequency at once, from a bass rumble up to a high-pitched hiss. The falls presented an overwhelming and isolating noise, like the sort controversially used by the CIA to achieve sensory deprivation during interrogations.7
Water might be a simple substance, but it can make a vast range of sounds, from babbling brooks to crashing waves, from torrential rain to the plink of a single drip. Describing Yosemite Falls, the American naturalist John Muir wrote that the water “seem[s] to burst forth in irregular spurts from some grand, throbbing mountain heart . . . At the bottom of the fall . . . [i]t is mostly a hissing, clashing, seething, upwhirling mass . . . This noble fall has far the richest, as well as the most powerful, voice of all the falls of the Valley, its tones varying from the sharp hiss and rustle of the wind in the glossy leaves of the live-oak and the soft, sifting, hushing tones of the pines, to the loudest rush and roar of storm winds and thunder among the crags of the summit peaks.”8
For decades, scientists have been interested in how falling water, such as crashing waves, creates sound underwater, because the noise impedes submariners listening for the enemy. But I want to know about what happens above the surface, and fortunately, scientists have now turned their attention to this question.
Laurent Galbrun from Heriot-Watt University in Scotland has been investigating how to make an impressive-sounding fountain or water feature while pumping the minimum amount of water, to reduce energy use. In parallel work, Greg Watts and colleagues at Bradford University in England have been investigating water falling onto different rocks and into pools, in search of the best sounds for hiding traffic noise. After recording a diverse playlist of water features, they had a panel of listeners judge the pleasantness of each sound. This experiment had to be undertaken in an acoustic laboratory, a room hardly conducive to making aesthetic judgments on outdoor water features. So the researchers built a theater set within the laboratory. It was a garden balcony, complete with bamboo screens, potted plants, and garden furniture, to get the subjects into the right frame of mind.
After the subjects had scored how much they liked each sound, Watts came to the conclusion that the worst noises had a booming quality, reminiscent of water flowing down into drains or utilitarian culverts. The most pleasing sounds sploshed and splashed, having a natural randomness as the water fell onto an uneven surface formed from small boulders. In similar tests, Galbrun found that the gentle babbling of a slow-moving natural stream was the most relaxing of all water sounds tested.9
The source of the waterfall sounds surprised me at first. A television crew recently filmed what happens in my university’s anechoic chamber. A high-speed camera captured a single drop falling into a fish tank of water. A slow-motion video from the top looks pretty. The drop causes a narrow column of water to rise up from the surface, creating ripples. But to understand what is heard, you need to look from the side, just underneath the surface of the water. While the ripples are visually impressive, a single tiny air bubble generates most of the sound. As the bottom of the drop penetrates the surface of the water, a bulging meniscus is created from which a tiny air bubble suddenly breaks away. This bubble of trapped air is only a few millimeters in diameter and so is easy to overlook and difficult to film. It may be small, but the air inside the bubble vibrates, resonates, and creates a plink that travels through the water and into the air.
Water falling onto rocks sounds very different because no underwater bubbles can be made (unless a layer of water has built up on the stone). Again, it is easiest to think about what happens when one drop smashes onto a rock and is splattered across the stone. As the falling drop smears itself into a thin layer of water on the stone, it disturbs the air around it, creating the sound.
A couple of months after the television crew filmed a single bubble, I got to learn more about aquatic sounds from artist Lee Patterson. We met up in the English Lake District, where Lee described how, in local ponds and water courses in northern England, he had discovered underwater sounds as rich as tropical rainforests. We chatted about the piece he was going to compose from his recordings in the Lake District. The Laughing Water Dashes Through was going to be a work prompted by the devastating floods that had hit the nearby market town of Cockermouth a few years previously. Lee explained how the work would explore the “different forms of energy embodied by water flow, and the sound that happens as a by-product of the water flow.”10
He was recording in a small, enclosed, flooded quarry on the day I visited. With blazingly hot sunshine and the birds singing around us, it was an idyllic spot (provided we stood with our backs to an ugly concrete shed). Lee had simple homemade hydrophones, constructed from a sliver of shiny piezoelectric material, which makes electricity when it is deformed by underwater sound waves, embedded in the tops of brightly colored plastic screw caps from pop bottles. He cast these into the water, turned up the amplifier, and passed me the headphones.
I heard a malevolent munching and crunching. It was as if an animal was trying to nibble away at my eardrum. The sound came from tadpoles scraping the hydrophones, in the vain hope that there was algae on the bottle caps. The tadpoles were swimming among oxygenating pondweed, and with a careful repositioning of the hydrophones, strange mechanical chirps could be heard, like bacon being deep-fried. These were caused by a fast stream of small bubbles rising up from the pondweed, looking like champagne bubbles rising up in a glass. It turns out that photosynthesizing plants were making the bubble streams.11
A few days later, I spoke to Helen Czerski from the University of Southampton, who studies how sound is made as bubbles are created. Her research shows that as bubbles form at a small nozzle, sound is created because the bubble initially has a teardrop shape while attached to the nozzle, but when it breaks away into the body of the water it forms a sphere. This shape change causes the bubble to vibrate, resonating the air inside and creating sound. Helen was skeptical that this is what happens with pondweed because natural bubbles from photosynthesis form more slowly and therefore probably lack the impulsive kick as they break off. She thought it was more likely that I was hearing the bubbles bumping into each other or into the hydrophones.
The waterfall at Dettifoss in Iceland can be explained by scaling up the effects of a single oscillating bubble, to consider the vast number of bubbles in the white cascade. Each ball of trapped air is a different size and plinks a note at its own particular frequency. In the cascade, the combination of millions of random plinks creates a vast bubble orchestra, which fizzes and roars.
Each waterfall has its own voice. If it tends to have lots of larger bubbles, it will have a bassy rumble. Smaller bubbles result in more hissing, like the Yosemite Falls described by Muir. Surrounding rocks can further alter the sound. Svartifoss in southern Iceland is only about 20 meters (65 feet) high. It has water cascading from a horseshoe of overhanging cliffs made up of hexagonal basalt columns. The name, meaning “black fall,” comes from the color of the rocks, and on the day I visited, the color was strongly emphasized by the overcast, drizzly weather. However, it is worth an hour’s walk even in the rain, because not only do the surrounding rocks make stunning holiday snapshots, but they also amplify the water as it slaps and hisses against them.
Another impressive Icelandic waterfall is Seljalandsfoss, where you can go behind the curtain and be surrounded by noise, as the fizzing from the water hitting the pool is reflected from the cliff behind you. The flow of water is not constant, which means the noise splutters. Close your eyes, and you can imagine a small freight train rumbling past overhead.
While waterfalls are common, the sound of a tidal bore, a single tall wave that sweeps inland up a narrowing estuary, is much rarer. The bore of Rio Araguari in Brazil is named pororoca in the language of the native Tupi, which translates as “mighty noise.”12 Closer to my home is the Severn bore, near Glouce
ster in England. Early on a misty September morning, part of a brief Indian summer, forecasters predicted a four-star bore on the heels of a large tide driven by the autumnal equinox. As I wandered along the banks of the river, I saw a few surfers midstream, clutching their boards ready to catch the wave; this must be a good place to watch, I thought. I first stood close to the water’s edge but then realized that the silt all around me was from the previous night’s tide, so I retreated higher up the bank. You have to be careful when dealing with tidal forces. In China, eighty-six people were swept away by a tidal bore on October 3, 1993.13
Then I waited, waited some more, and then waited even longer. Twenty minutes behind schedule, a rumble started downstream. The bore came into view and broke on the opposite bank, forming a continuously breaking wave right across the river. It resembled a big ocean breaker, but instead of the soothing rhythm of waves crashing on the shore one after another, the bore produced the continuous sound of a breaking wave.
Right behind the Bay of Fundy in Nova Scotia, the Severn Estuary has the second-highest tidal range in the world—as much as 14 meters (45 feet) for spring tides. A map of the River Severn shows its sinuous funnel shape. What the map does not show is that the depth of the river decreases rapidly as you go inland. When the huge tide enters the estuary at the sea, the water is forced up the narrowing channel, which gets shallower and shallower. The excess water can go only one way, upward, thus forming the surge wave.