Resonance and Formants

Push a swing at the right rhythm and each push adds efficiently to the motion. A string, air column, instrument body or room behaves similarly: it responds more strongly near frequencies supported by its size, shape, material and boundary conditions.

Resonance does not create energy from nowhere. It stores and returns energy from a source. Damping controls how quickly that energy is lost.

High Q. A narrow resonance is selective and can ring for longer. It produces a pronounced peak around a small frequency range.

Low Q. A broad resonance affects a wider range and usually dies away faster. Its color is less sharply pitched.

More damping. Energy is removed more quickly, which lowers and broadens the response. Soft materials and friction often add damping.

These terms describe a response, not a judgment. A high-Q room mode can be troublesome; a high-Q filter can be exactly the effect a synthesizer needs.

A source supplies a spectrum. A filter changes the balance of that spectrum. The result is what reaches the listener.

Voice. The vocal folds provide a pitched, harmonically rich source. The throat and mouth shape it with resonances.

String instrument. The vibrating string provides energy. The bridge and body radiate some frequency regions more efficiently than others.

Subtractive synthesizer. An oscillator provides the source. A filter emphasizes or removes parts of its spectrum.

A formant is a resonance peak or region associated with a vocal tract or another resonant system. In speech, the first few formants carry much of the information that distinguishes vowels.

Raise the sung note while keeping the mouth shape similar. The harmonics move with the fundamental, but the vocal-tract resonances remain in roughly the same frequency regions. Different harmonics pass through those regions, yet the vowel can remain recognizable.

Whispering removes a regular vocal-fold fundamental, but the noisy source still passes through the same filter shape. That is why whispered vowels remain intelligible.

Instrument bodies and air columns create broad spectral regions that contribute to family resemblance. A violin does not apply one perfectly fixed filter at every pitch and dynamic, but its construction repeatedly favors characteristic regions.

Players also change the filter. Bow position, embouchure, mute, pickup position and articulation alter which partials are produced or radiated.

Pitch-shifting a voice without formant correction moves both pitch and resonant envelope. Large upward shifts can produce the familiar small, bright character. Formant-aware processing separates those controls.