Helium voice

Why does breathing helium change the sound of a voice when the vocal cords are still vibrating at the same rate?

Everyone knows what helium does to a voice, and almost everyone describes it the same way: it makes you sound higher. But the question already contains a problem. The vocal folds are strips of tissue whose vibration rate is set by their mass, their length, their tension and the pressure driving them — and helium changes none of those. So if the thing generating the sound is running at the same rate, what exactly went up?

Let us separate two things that ordinary speech keeps welded together. There is a source — the vocal folds chopping the airflow into a train of puffs, a few hundred times a second — and there is a filter, the tube of throat, mouth and nose the puffs must travel through before reaching the air. The folds decide how often; the tube decides which frequencies come out loud.

Why should a tube have a preference at all? Because a column of air in a pipe of a given length resonates: at certain frequencies the reflections from its ends reinforce, and at others they cancel. Those favoured frequencies are the vocal tract's resonances, called formants, and they are what distinguish one vowel from another. Change your mouth from "ee" to "ah" and you have moved the formants without touching the folds at all — which is why you can sing a single held note through every vowel in the language.

So what sets where a resonance sits? Two things only: the length of the tube, and how fast sound travels inside it. Helium does not change the length of your throat. What it does change is the second one, and dramatically. Sound moves faster in a gas whose molecules are lighter, and helium's are far lighter than the nitrogen and oxygen normally in there. At room temperature the figures are about 343 metres per second in air and about 1007 in helium — roughly three times as fast.

Now put the two halves back together. The folds still buzz at, say, two hundred times a second, so the sound still contains a series of components at two hundred, four hundred, six hundred and so on — unchanged. But the tube, with sound racing through it, now resonates at much higher frequencies than before. The same set of components is present; a different part of the set is being amplified. The high members of the series are boosted, the low ones no longer favoured.

What does that sound like? Thin, reedy, oddly duck-like — and, to most listeners, "higher", because we habitually read a bright, top-heavy sound as a high one. But push on that word. If the perceived pitch really rose, then a sung note would come out sharp. It does not. Sing a note into a microphone, breathe helium, sing it again: the note is the same note, measurably. What changed is the tone colour, not the pitch.

Is there a way to check the reasoning rather than just the claim? There is: the reverse experiment. Sulphur hexafluoride has very heavy molecules, and sound travels through it at only about 135 metres per second — slower than in air. Breathe that, and the same argument predicts the formants should fall, giving a deep, cavernous voice with the fundamental unchanged. That is exactly what happens. Neither gas is "the funny one": the two sit either side of air on a single axis, and the axis is the speed of sound.

Worth saying plainly at this point: breathing either gas from a source that has no oxygen in it can and does kill people by asphyxiation, and the heavy gas is the harder of the two to clear from the lungs afterwards.

The misconception here is unusually clean, so it is worth stating the correction directly. Helium does not raise your pitch. It raises the resonant frequencies of the cavity your voice passes through, which shifts the emphasis onto higher components of a sound whose fundamental rate never moved. Pitch is a property of the source; the helium effect is a property of the filter. Confusing the two is the single most common error about it — and it survives because "sounds higher" is an accurate description of the impression, just not of the mechanism.

A second refinement follows. Because vowels are carried by formant positions, and helium moves all the formants upward together, vowels become harder to identify — speech in helium is not merely comic but genuinely less intelligible. Deep-sea divers breathing helium-oxygen mixtures are hard to understand over an intercom, and their communications are routinely passed through electronics that shift the formants back down.

Two honest qualifications. The factor of roughly three applies to pure helium, and you never have pure helium in your vocal tract — residual air, carbon dioxide and water vapour dilute it, so a real shift is smaller and fades over a few seconds. And the fundamental is not quite perfectly unchanged either; the altered gas density slightly affects how the folds are loaded. Small, though — nothing like the threefold change in the resonances, which is what you are actually hearing.

• How the lowest two formant frequencies alone are enough to identify most vowels.
• What an electronic descrambler has to do to divers' helium speech to make it intelligible.

Every term the collection defines is gathered in the glossary.