by Trevor Cox
Maybe I should have taken Daisy to Lake Königssee in Bavaria, Germany’s highest lake, where rock faces rise steeply out of the water. Boat captains there play short phrases on trumpets so that tourists can hear the last three notes repeat, delayed by one or two seconds, after bouncing off the surrounding alps. Or perhaps I should have taken Daisy to the place where seventeenth-century French theologian, natural philosopher, and mathematician Marin Mersenne had carried out his echo experiments. He used a polysyllabic echo to make the first accurate measurements of the speed of sound in air. Nowadays, Mersenne is probably best known in mathematics for his work on prime numbers, but he was also passionate about a wide variety of subjects and devoted to the need for experimentation and observation.5
Unsurprisingly, Mersenne did not use wildfowl in his speed-of-sound experiments. Instead, he stood facing a large reflecting surface, saying the words “benedicam dominum,” and using a pendulum to time the sound. Mersenne must have been a quick talker, because he said this seven-syllable phrase in one second. When he stood 485 royal feet (159 meters) from a large reflecting surface,6 the echo immediately followed the end of the original phrase—“benedicam dominum, benedicam dominum.” This is a polysyllabic echo because many syllables can be said before the echo returns. The echoed words had covered a round-trip of two times 485 royal feet (a total of 319 meters), allowing Mersenne to deduce that 319 meters per second is the speed of sound. This is remarkably close to the correct value of about 340 meters per second (761 miles per hour).7
Now, if Mersenne had used a duck, he could have stood nearer the wall and still heard a distinct “quack, quack” because a duck’s call is but one syllable. In fact, the distance to hear a monosyllabic echo like a quack is about 33 meters (36 yards) from a reflecting surface (660 duck feet?8) because at that distance an echo takes just long enough to bounce back that it can be heard separately from the original sound. To hear a quack echo, I would need to find a stretch of water with a large building or cliff about 30–40 meters (or 30–40 yards) away. But even this would not work, because a duck’s quack is too quiet. Sound becomes quieter the farther you are from a source, by 6 decibels for every doubling of distance, so if a quack measures 60 decibels 1 meter (3¼ feet) from the beak, 2 meters (6½ feet) away it will have dropped to 54 decibels, 4 meters (13 feet) away it will be at 48 decibels, and so on. By the time the reflected quack has undertaken its round-trip of 66 meters (72 yards), the echo will be about 24 decibels. In a completely silent place a human could hear this, but more often other noises, such as the distant rumble of traffic or wind moving through trees, are louder, making the quack inaudible.9 Sadly, even in a silent place Daisy would not be able to hear the echo, because her hearing is less sensitive than a human’s. So the reason the echo from a duck’s call is not heard is pure physics: a quack is not loud enough to be heard after it makes the return trip from the required distance.
Marin Mersenne’s acoustic work extended beyond the speed of sound; he also debunked fanciful tales about 400 years before myth busting became a popular pastime on television. One of the more extravagant acoustic claims in classical literature is a supposed heterophonic echo that, when spoken to in French, replied in Spanish. Marsenne knew this could not be true, but as Professor Fredrick Vinton Hunt wrote in his seminal book Origins in Acoustics, Marsenne “almost convinced himself that one could devise a special series of sounds whose echo might lead a listener to think he had heard the response in a different language.”10 The term heterophony comes from musicology and denotes a melody being simultaneously played with an elaborated variant, so I can only imagine that a heterophonic echo might augment the French words to make them Spanish. Unfortunately, no one knows for sure what was meant by the term, and there are no examples of heterophonic echoes. Luckily, there are other literary games that can be more successfully played with echoes, as I found out in France.
One hot, sunny day in 2011, I was cycling with my family in the Loire valley, and we arrived at the Château de Chinon. The heart of the castle was built by Henri Plantagenêt, who would later become King Henry II of England. But I was more interested in a very unusual road sign just outside the castle walls. It points up a small lane and simply says, “Écho.” How could a collector of sonic wonders resist this invitation? A few hundred yards up the lane was a small, raised turnout and a sign indicating that this was the place to test the acoustics. I yelled and yodeled and appreciated the fine echo.11 What made the experience very satisfying was that the side of the château, which was reflecting the sound, was partly hidden by an orchard, so the clarity of the echo was surprising. I could not resist attempting a traditional piece of echo humor from my guidebook to the Loire:12
Me: “Les femmes de Chinon sont-elles fidèles”
Echo: “Elles?”
Me: “Oui, Les femmes de Chinon”
Echo: “Non!”
This translates into English as:
Me: “Are the women of Chinon faithful?”
Echo: “Them?”
Me: “Yes, the women of Chinon”
Echo: “No!”
Given the right enunciation, with an unnatural stress placed on the last syllable in each phrase, such as the “non” in Chinon, the rhyme worked. By that I mean that the partial words echoed back from the north side of the château and were clearly audible. The sentiments expressed in the poem could not be readily verified.
There are other echo stories. Here is a nineteenth-century account from Wonders of Acoustics by Rodolphe Radau (with a Latin translation in square brackets):
Cardan tells a story of a man who, wishing to cross a river, could not find the ford. In his disappointment he heaved a sigh. “Oh!” replied the echo. He thought himself no longer alone, and began the following dialogue:
Onde devo passar? [Hence I have to pass?]
Passa. [pass]
Qui ? [here?]
Qui. [here]
However, seeing he had a dangerous whirlpool to pass, he asked again
Devo passar qui? [I have to pass here?]
Passa qui. [pass here]
The man was frightened, thinking himself the sport of some mocking demon, and returned home without daring to cross the water.13
Wonders of Acoustics includes much about Athanasius Kircher, a seventeenth-century Jesuit scholar based in Rome who wrote extensively about theater acoustics and other marvels. He was intrigued by manifold echoes—echoes that produce multiple distinct reflections. Included in this category are repeating echoes caused by elaborate structures that turn one word into a whole sentence. For his two-volume masterpiece Musurgia Universalis, from 1650, Kircher produced a drawing of large upright panels, spaced at various distances from a talker, to generate a series of reflections arriving one after another. One such edifice had five panels and was designed to take the word clamore and break it down to echo clamore from the first panel and amore from the second, followed by more, ore, re from the third, fourth, and fifth surfaces, respectively. So, if you shouted the question “Tibi vero gratias agam, quo clamore?” (“How shall I cry out my thanks to thee?”), the echoes from the last word would reply with a Latin phrase “clamore, amore, more, ore, re,” which roughly translates as “with thy love, thy wont, thy words, thy deeds.”14
I thought this seemed very unlikely to work, but the idea was intriguing enough to inspire a quick test. Not having five large panels lying around, I decided to try out the idea by simulating the situation on my computer. I recorded myself saying “clamore” and then, using a piece of prediction software, estimated the reflections returning from each of the panels shown in Kircher’s picture. I played around with how far the panels were from the speaker and the loudness of the reflections in an attempt to produce the echo pattern.15 Much to my surprise, the echo phrase actually worked, but perhaps only because my brain was fooled into hearing patterns I wanted to hear.
I once saw a great demonstration of a similar effect by author Simon Singh, based on the accusation that Led Zep
pelin hid satanic messages in “Stairway to Heaven.” If you play the track backward, you supposedly hear, “Oh here’s to my sweet Satan. The one whose little path would make me sad, whose power is Satan. He’ll give those with him 666, there was a little toolshed where he made us suffer, sad Satan.” So concerned were some religious groups that various US states introduced legislation requiring records to carry warning labels.16 These claims further implied that even when records were listened to normally, with the sound playing forward, the listener would subconsciously decipher the meaning of the backward satanic messages.17
Several groups of psychologists have tested the claims using proper scientific methods. Experiments showed that if you listen to “Stairway to Heaven” backward with your eyes closed, what you actually hear is gibberish. These satanic lyrics are heard only if you have a printed version in front of you. (You can try this out yourself; plenty of websites are dedicated to backward masking with sound samples.) The brain has to make sense of incomplete information all the time, so it is very adept at finding patterns and fitting together different sources of information. But sometimes the brain gets it wrong, in this case matching the written lyrics to the otherwise incomprehensible backward murmurings.
The same thing happens with the “clamore, amore, more, ore, re” echo. When I listened very carefully for this pattern of words, I could pick out the phrase. The effect was especially strong if the echoes were faint and I was forced to strain to hear them. But if I closed my eyes and listened more holistically and analytically, the dominant effect I could hear was many repetitions of “re.” The clever wordplay disappeared.
A multiple echo, or tautological echo, is almost the same as a manifold echo, except the same words or syllables are repeated many times. An episode of the TV show The Simpsons features this phenomenon as a piece of aural slapstick. Marge is in church and yet again is being embarrassed by Homer. She cries out, “Homer, your behavior is heinous” and a tautological echo replies “anus, anus, anus, . . .”18
Athanasius Kircher was also interested in echo pranks. He describes being amused at a friend’s expense in the Campagna at Rome, a lowland plain surrounding the city. His friend cried out, “Quod tibi nomen?” (“What is your name?”), and the echo impossibly replied, “Constantinus.” The conceit was achieved using an accomplice hiding near a cliff where normally there was no echo. The accomplice would shout out the reply after hearing the question, impersonating the improbable sound reflection.19
A more impressive practical joker is Bob Perry, who has taught himself to impersonate an echo. He does an impressive performance of John F. Kennedy’s inaugural address, complete with multiple renditions of each word as though caused by a public address system. With a little practice, you could teach yourself to do this. Select a spoken passage in which the space between syllables is a little longer than normal, as in JFK’s speech, because of the slow delivery, and then say every syllable twice: “Ask ask not not what what you you can can do do . . .”. To make it convincing, the echo word should be a bit quieter.
It is not usually the architecture that renders big speeches and announcements in train stations unintelligible; often electronics are to blame. Bad public address systems send out sound too loudly from too many sources. You hear words from two or more loudspeakers arriving separately because these sources of sound are different distances from you. One engineering solution is to change the position and orientation of each loudspeaker to make sure you hear only one at a time. Engineers can also use loudspeakers that illuminate defined areas rather than speakers that radiate in all directions—the aural equivalent of using a spotlight instead of a general-purpose lightbulb. But targeted illumination is not always possible, in which case engineers add electronic delays to each speaker to ensure that you hear all versions of the speech arriving at roughly the same time. Your brain will then lump the speech from the different speakers into one louder sound, minimizing the confusing cacophony of repetition.
On the TV show Candid Camera, echo prankster Bob Perry stood at Coit Tower, which affords great views of San Francisco, next to a false sign saying, “echo point.” Standing alongside his unsuspecting victim, Bob shouted to create the illusion of sound bouncing off the tower, imitating an echo delayed by about one-fifth of a second. The joke was that whenever the victim tried shouting, there was no echo.
Bob Perry is impersonating what music producers would call a slapback echo, which is a single loud and delayed repetition. This effect was popularized in rock ’n’ roll recordings from the 1950s and helped create the characteristic sound of famous singers like Elvis Presley. The audio engineers used two tape recorders to produce electronic echoes. A single big loop of the magnetic tape was fed through both recorders; the first machine would record the music onto the tape, and the second would pick up the sound from the tape a short time later, thus producing a delayed, slapback echo. The time between the passing of the tape under the record head on one machine to its reaching the pickup on the other machine determined the delay of the echo. On tracks such as “Boogie Disease” by Doctor Ross, the echo delay is about 0.15 second, creating the impression that the electric guitar on this blues recording is playing at double speed, as every strum is repeated.
The same effect gave Elvis’s vocals a distinctive sound on his recordings with Sun Records, such as “Blue Moon.” When Elvis switched to the RCA record label and achieved global hits with songs like “Heartbreak Hotel,” the sound engineers could not work out how to reproduce the slapback echo, and they resorted to adding heavy reverberation from a hallway outside the studio.20 Nowadays, it would be simple to reproduce this effect digitally, as delay is one of the cornerstones of modern pop production. To create the effect without electronics, RCA’s engineers would have had to record Elvis in a studio next to a long tunnel or a tall room with a domed roof that had a slapback echo (remember, one of the dimensions would have to be at least 33 meters (110 feet), which would make it a rather large recording studio).
The Imam Mosque in Isfahan, Iran, might have worked for Elvis’s voice as, according to the old writings on echoes, it is a centrum phonocampticum, the object of an echo. Constructed in the seventeenth century, the building is visually stunning, with dazzling blue Islamic tiles. A huge dome rises to an exterior height of 52 meters (170 feet) and, as one travel guide states, “replicat[es] individual sounds in a series of clear echoes.”21 Tour guides delight in standing underneath the dome and snapping or flicking a piece of paper, which creates a short, sharp “clack, clack, clack, . . .” The room immediately responds with about seven quick-fire echoes.22 Sound bounces back and forth between the floor and ceiling, with the curved dome focusing the sound, forcing it to keep moving vertically up and down in a regimented fashion. Without a dome, the echo from the ceiling would be lost among all the other sound reflections in the mosque.
Luke Jerram is an artist who often uses sound as an art medium. His work Aeolus was inspired by a visit to Iran, where he heard the echoes in the Imam Mosque. I first met Luke about seven years ago, when we both reached the finals of FameLab, a pop idol–style competition to find science presenters for the media. I caught up with Luke when Aeolus, or what he called “my ten-ton musical instrument,” was being installed outside my university’s building at MediaCityUK in 2011.
Aeolus looks like a section through a giant steel hedgehog, an arch 4–5 meters (13–16 feet) high, with 300 long, hollow steel pipes sticking out of the top and sides (Figure 4.1). The shape was inspired by the twelve stuttering echoes that Luke had heard when he clicked his fingers in the mosque. Stand in the right place under Aeolus, and you can hear the focus from the arch subtly amplifying your voice. The light coming through the mirror-lined steel pipes creates geometric patterns echoing the decoration of the mosque.
While the arch is the most obvious visual feature of the sculpture, the main sound effect is created by long wires that stretch almost invisibly from supporting poles to the hedgehog. Each of the wires is dri
ven to vibrate by the wind. Pieces of wood act like violin bridges, transferring the string vibrations to membranes stretched across the ends of the pipes. The membranes then cause the air in the pipes to resonate. Overall, the result is an eerie, pulsing sound like a minimalist piece of music by American composer Steve Reich in which tones come and go depending on the changing wind.
Figure 4.1 Aeolus.
The work is named after the ruler of the four winds in Greek mythology. Luke’s intention is to use “sound to paint pictures in people’s imaginations,” allowing visitors to “visualize the changing landscape of wind around the artwork.”23 The sound is hard to locate, appearing to be flowing vaguely from above. The lengths of the pipe are carefully selected to form a musical scale. Appropriately, the Aeolian mode is used, which, as a minor scale, lends a malevolent, spooky character to the sound.24 If I shut my eyes, I could imagine I was in a B movie during a Martian invasion.
Luke decided to construct Aeolus after meeting a master digger in Iran who had described the construction of underground irrigation canals called qanats. The digging is wet, claustrophobic, and dangerous. The worst task is probably “devil digging,” in which they mine up to a well of water from underneath. Just imagine being in a cramped passageway at the moment the digger breaks through and the water comes cascading down on top of you. What inspired Luke to create a singing building was the howling of a qanat’s air vents in the wind.
Like the Iranian mosque, many grand buildings feature domes, but only rarely do they have the right curvature to achieve distinct echoes. The room diagrammed on the left in Figure 4.2 has a focal point that is too high; the building on the right brings amplified sound back to the listener at ground level and produces a pattern of repeated echoes. By measuring the time between echoes in a recording, I estimate that the interior height in the Imam Mosque is 36 meters (120 feet). The place to stand in the mosque is marked on the floor; this spot is called the centrum phonicum, according to old writings on echoes.