A held vowel

Musicians describe voices and guitars with the same words: bright, dark, nasal, honky, warm, cutting. It is easy to take that as loose metaphor. It is closer to physics.

A voice starts as a buzz. The vocal folds chop the airflow from the lungs and produce a sound rich in harmonics, and the vocal tract then filters it: its resonances let some harmonics out strongly and hold others back.¹ That is the source-filter model of the voice, as Gunnar Fant described it in 1960.¹ It explains what a vowel is. Moving the tongue, lips and jaw changes the resonances, and the vowels are set largely by the two lowest.² The same resonances shape a singer's timbre. Trained male opera singers show a peak in the 2 to 4 kHz region, called the singer's formant, which can carry more power in that range than a whole orchestra and helps the voice project over it.¹

An electric guitar has the same structure. A plucked string is a source rich in harmonics; everything after it is a filter. Even the cable is part of the filter: a passive pickup and the capacitance of the cable form a tuned circuit, "rather like a mixing desk's mid EQ set to boost", and a cable with less capacitance moves that resonance up while one with more moves it down.³ The amplifier and the loudspeaker shape the sound again, and nobody wants a flat chain.³ "Today we could make a guitar pickup with a 20Hz to 20kHz response," Paul White wrote in Sound On Sound, "But we'd hate it!"³

So the shared vocabulary is earned. "Nasal" or "honky" names a resonance in the middle of the range, in a throat or in a pickup; "dark" names one that sits low; "cutting" names energy where the ear and the room let it through.

The difference is that a voice moves its resonances every fraction of a second, and a guitar's are fixed by its hardware. A guitar tone is a held vowel. Pickups, cable, amplifier and speaker together set one shape of resonance, and the instrument sings everything through it. An overdriven amplifier bends the comparison, because distortion adds harmonics, which no filter can do; in the voice's terms that roughens the source rather than changing the vowel. The long search guitarists call chasing tone is, in these terms, choosing which vowel to hold.

Two devices loosen the hold. A wah pedal moves a single resonance, and with only one it cannot mimic speech.² Its name is what that one moving resonance sounds like. A talk box sends the amplifier's output through a tube into the player's mouth, and a microphone picks up the result: "It is now the guitar sound, rich in harmonics, rather than the voice that is filtered by the vocal tract before it reaches the microphone."² The guitar borrows the whole moving filter, and it talks. Researchers at the University of New South Wales use the same arrangement to measure the vocal tract, injecting a synthesised sound into the mouth through a tube.²

One vocabulary can describe both, then, as far as it describes resonances. It stops where the voice does something no fixed filter can: changing its shape continuously to make words. Much of what makes a voice one person's, its accent and timing, lives in that motion. A guitar, without a talk box, has only its held vowel, and a good part of what players call their sound is the vowel they chose.