The singing body · Chapter 1 of 3
What Actually Vibrates
On a buzz and a tube, which is the whole instrument.
3 min read
The voice is usually taught as a mystery and it is two systems, each of which can be described in a paragraph.
The first is the source. Air comes up from the lungs and meets the vocal folds, two small bands of tissue stretched across the larynx. They open and close rapidly, chopping the continuous airflow into a train of pulses. That train is a buzz, rich in harmonics, and the rate at which the folds close is the pitch.
The second is the filter. Above the folds is a tube: the throat, the mouth, sometimes the nose. Like any tube it has resonances, frequencies at which it responds strongly, and those resonances amplify parts of the buzz and suppress others. Which parts depends on the shape, and the shape is under your control through the tongue, jaw, lips and soft palate.
That is source-filter theory, and the useful thing about it is that the two systems are close to independent. Change the folds and you change the note. Change the tube and you change the vowel. You can sing any vowel at any pitch, which seems trivial until you realise it is a fact about the architecture rather than a skill.
The demonstration costs nothing. Whisper a vowel. The folds are barely vibrating, so there is almost no source, and yet the vowel is perfectly identifiable, because it was never in the folds. It was in the shape of the tube.
Now the part that is routinely got wrong, including by singing teachers.
The folds are not strings. A string is set in motion and vibrates at a frequency determined by its tension, length and mass, and that analogy puts the pitch in the folds where it belongs. But a string's vibration is maintained by its own elasticity, and the folds' is maintained aerodynamically.
Here is the cycle. Air pressure below the closed folds builds and forces them apart. Air rushes through the narrow gap, and fast-moving air has lower pressure, which pulls the folds back towards each other. Elastic recoil completes the closure. Pressure builds again. Repeat.
That happens somewhere between about eighty and a thousand times a second depending on the voice and the note. No nerve conducts at those rates, and no act of will could organise them if it did. Nobody decides each closure.
Which means the thing a singer actually controls is the conditions: the tension and length of the folds, how firmly they approximate, and the air pressure beneath. The oscillation then follows from the physics without being instructed.
This is why the voice can be trained and cannot be operated, and why a great deal of vocal instruction that describes muscular effort directly makes things worse. Telling somebody to tighten something is an instruction about a mechanism they cannot reach. Telling them to change the vowel, the airflow or the imagined placement is an instruction about a condition, and conditions are adjustable.
The instrument is a valve and a tube. Almost everything else said about singing is a way of talking about one of those two without naming it.
The folds decide the note. The tube above them decides what word it is.
Where you'll meet this
This is why you can sing the same pitch on every vowel and speak the same vowel at any pitch. Source and filter are close to independent, which is also why a vocoder and a talk box work at all.