Formants, vowels and timbre
A voice can sing the same note and still say "ee" or "ah". The note is the pitch, which is set by the vocal folds. The vowel is set by the shape of the vocal tract, which shapes the spectrum. The peaks of that shaped spectrum are the formants.
Source and filter
Voice production works like a source and a filter:
- The source is the buzz of the vocal folds, which is rich in harmonics at multiples of the pitch.
- The filter is the throat, mouth and nose cavities. They resonate at certain frequencies and boost the harmonics near them.
Change the position of tongue, jaw and lips and the resonances move. The formants stay at their frequencies even when the pitch changes, which is why we recognise a vowel on any note.
First two formants
The lowest two formants carry most of the vowel identity.
- F1 rises with how far the jaw opens (how low the tongue is).
- F2 rises as the tongue moves forward.
Typical values for adult male speakers, rounded:
| Vowel | Example | F1 | F2 |
|---|---|---|---|
| ee | heed | 270 Hz | 2300 Hz |
| eh | head | 530 Hz | 1800 Hz |
| ah | father | 730 Hz | 1100 Hz |
| aw | thought | 570 Hz | 840 Hz |
| oo | who'd | 300 Hz | 870 Hz |
Women's values are typically about 15 to 20 % higher, children's more, since vocal tracts are shorter. Plotted with F1 against F2, the vowels form the vowel triangle, with ee, ah and oo at the corners.
The singer's formant
Trained opera singers cluster the third, fourth and fifth formants so that a strong peak appears around 2.5 to 3.5 kHz. That peak sits at a frequency where the orchestra has less energy and the ear is sensitive, which helps the voice to project over the orchestra.
Other instruments
Formant-like resonances appear in many instruments. A violin body has resonances that shape its sound regardless of the note. Brass and woodwind instruments have bore resonances. In all of these, the spectral envelope tied to the instrument gives the characteristic timbre, see Harmonics, partials and overtones.
Why it matters for recording
- Equalising a voice moves its formants and can make it sound as if the mouth were a different shape.
- The 2 to 5 kHz region carries consonants and presence; take care when boosting it.
- Strong nasal or throat resonances can be tamed with narrow cuts, see Equalisation basics.
Draft: written by an AI assistant, awaiting technical review.
Found a mistake or have an idea for this article?
Suggest a change