What if elephants talked with very low sounds?

Elephants can make a deep rumble, like a very low hum.

Some rumbles are so low that people cannot hear them at all. Why would an elephant talk so low?

Elephants trumpet, roar and scream, but many of their calls are deep rumbles. Some rumbles are so low that people can barely hear them, or cannot hear part of them at all.

This lab asks what changes when a sound gets very, very low. In the model you can turn a sound's pitch down from an ordinary musical note to a 20 hertz rumble, then compare what happens to the wave.

Many elephant calls are rumbles whose fundamental frequency lies in the infrasonic range, below about 20 hertz, the lower limit of human hearing (Cornell's Elephant Listening Project; U.S. National Park Service).

This lab asks what lowering a sound's frequency does to the wave itself. The sound model lets you drop the pitch from 440 Hz, the note A, to 20 Hz, and read off the speed, the wavelength and the time the sound takes to cross up to 2,000 meters. The what-if is a real one: elephants already call this way, so the lab separates what the physics of the wave shows from what only field recordings can tell us.

People can hear lots of sounds, high and low.

But some sounds are too low for our ears. Those very low sounds are called infrasound.

The National Park Service says people with normal hearing can hear sounds from 20 hertz to 20,000 hertz. Hertz counts how many sound waves pass each second. Sounds below 20 hertz are called infrasound, and we cannot hear them.

Our voices sit well inside that range. Cornell's Elephant Listening Project gives men's speaking voices as about 70 to 200 hertz and women's as about 140 to 400 hertz. In the model, a 440 hertz note crossing 2,000 meters of air at 20 °C travels at 343 meters per second, arrives in 5.83 seconds and has waves 0.78 meters long.

Frequency is the number of pressure cycles per second, measured in hertz. The National Park Service gives normal human hearing as 20 Hz to 20,000 Hz, with infrasound below and ultrasound above. Cornell's Elephant Listening Project notes that people with very good hearing can only just detect 20 Hz, and gives speaking frequencies of about 70–200 Hz for men and 140–400 Hz for women.

In the model's dry air at 20 °C, c = 331.3 √(1 + T/273.15) gives 343 m/s. A 440 Hz tone then has a wavelength λ = c/f of 0.78 m and crosses 2,000 m in 5.83 s.

In the lab, Olly turns the sound down low, like an elephant rumble.

The sound still takes the same time to get there. But each wave gets very, very long!

Now turn the pitch down to 20 hertz, the lowest the model goes and right at the bottom of human hearing. Part of many elephant rumbles is even lower. Two things are worth watching.

First, the sound arrives at exactly the same time as before: 5.83 seconds for 2,000 meters. Low and high sounds travel at the same speed. Second, each wave is now 17.2 meters long instead of 0.78 meters, about 22 times longer. A low sound is a sound with long waves.

Set f = 20 Hz with everything else unchanged. The speed stays 343 m/s and the 2,000 m crossing still takes 5.83 s, because in the model, as in air, speed does not depend on frequency. The wavelength becomes λ = 343/20 = 17.2 m, 22 times the 0.78 m wave of the 440 Hz note.

That is all the model can show: same speed, longer wave. It does not calculate loudness, so it cannot tell you how far away the rumble could still be heard. That question needs field measurements, which the real-world section describes.

All sounds in the air go at the same speed, high or low.

Low sounds have long waves. Scientists say low sounds can travel farther than high sounds.

A sound's speed is set by what it travels through, not by its pitch. OpenStax uses a marching band as proof: if high notes traveled faster than low ones, the music would sound jumbled from far away, but it doesn't. Because every pitch moves at the same speed, a low pitch must have a long wave.

The National Park Service explains that low sounds travel farther than high ones, which makes infrasound good for talking over long distances. Cornell's scientists add that in forests, higher sounds are bounced and soaked up more by plants. Our model leaves all of that out: it shows only speed and wavelength, not how loud a sound stays.

The speed of sound is nearly independent of frequency; OpenStax notes that music from a distant band arrives in step, which would be impossible if pitches traveled at different speeds. Since c is fixed, λ = c/f, so lower frequencies have proportionally longer waves.

Why low frequencies carry farther is a separate question about losses. The National Park Service states that low-frequency sounds travel farther than high-frequency ones, and Cornell's Elephant Listening Project explains that in closed habitats higher frequencies are reflected and absorbed more by vegetation.

The model's limits: it times a single pulse in a straight line, with no loudness, fading, absorption, echoes, wind or ground effects. Any statement about range comes from those outside sources, not from the model.

Scientists listen to elephants with special recorders.

They learned that elephants can hear each other's rumbles from far, far away.

People did not know how much elephants used these low sounds until 1984, when scientist Katy Payne discovered that much of their talk happens below what humans can hear. Later, scientists played recorded calls to wild African savanna elephants, and the elephants responded from 2 kilometers away. Langbauer, Payne and colleagues estimated that real calls could be detected about 4 kilometers away.

In noisy rainforests the distance is shorter. Cornell's Elephant Listening Project estimated about 800 meters for an average forest elephant rumble against the usual background noise.

Cornell's Elephant Listening Project traces the discovery to a hunch Katy Payne had at the Washington Park Zoo in Portland, Oregon, where she sensed a thrumming in the air near Asian elephants (ELP dates the discovery to 1984); work there with William Langbauer Jr. and Elizabeth Thomas confirmed infrasonic calls. Playback experiments in East Africa showed savanna elephants responding over 2 km, and Langbauer, Payne and colleagues estimated a true detection range of about 4 km. On dry-season evenings a temperature inversion can raise the listening area from about 30 to 300 square kilometers, and savanna elephants make most loud low calls in those hours.

In Central African rainforest, ELP estimated about 800 m for an average rumble, limited mostly by background noise. A Stanford team (O'Connell-Rodwell, Günther and Klemperer) also reported rumbles making waves in the ground; O'Connell-Rodwell and colleagues believe elephants sense these through receptors in their feet and trunks, an idea still being studied.