Vocal formants are the resonance frequencies of the vocal tract (pharynx, mouth, nasal cavities) that characterize each vowel. They appear as dark horizontal bands on a spectrogram and allow distinguishing an “a” from an “i” regardless of the voice pitch.
Discover FizziQ
How to measure it in class
With the FizziQ app, it is possible to visualize formants using the spectrogram.
Steps:
- Open FizziQ and select the “Spectrogram” instrument
- Pronounce a sustained vowel (for example “aaaa”) for 2-3 seconds
- Observe the horizontal bands that appear: these are the formants
- Repeat with other vowels (e, i, o, u) and compare the formant positions
- Note that the first two formants (F1 and F2) are sufficient to distinguish vowels
Scientific activities on this topic
Possible extensions with FizziQ: comparing the formants of different voices (man, woman, child) and studying the effect of accent on formants.
Learn more
Origin of formants:
When the vocal cords vibrate, they produce a sound rich in harmonics. The vocal tract (throat, mouth, lips) acts as a resonator that amplifies certain frequencies and attenuates others. The amplified frequencies are the formants.
The main formants:
- F1 (first formant): related to mouth opening (200-800 Hz). Open vowels (a) = high F1, closed vowels (i, u) = low F1.
- F2 (second formant): related to the front/back position of the tongue (800-2500 Hz). Front vowels (i, e) = high F2, back vowels (o, u) = low F2.
Vowel triangle:
By plotting F1 on the y-axis and F2 on the x-axis, the vowels are distributed in a triangle-shaped space, whose vertices correspond to the extreme vowels (i, a, u). Note: this triangle is not identical in all languages. Vowel systems vary from one language to another, both in the number of vowels and in the exact position of their formants. The values below are orders of magnitude for French vowels.
| Vowel | F1 (Hz) | F2 (Hz) |
|---|---|---|
| i | 270 | 2300 |
| e (as in French “é”) | 400 | 2000 |
| a | 700 | 1200 |
| o | 500 | 900 |
| u (as in French “ou”) | 300 | 750 |
Formula
There is no single formula because formants depend on the geometry of the vocal tract. The vocal tract can be modeled as a resonant tube of length L:
f_n ≈ (2n-1) × c / (4L)
where:
- f_n: frequency of the nth formant (Hz)
- c: speed of sound (≈ 340 m/s)
- L: length of the vocal tract (≈ 17 cm in adults)
Application examples
- Voice recognition uses formants to identify words
- Helium raises the formants because sound travels faster in it, hence the “duck voice”
- Opera singers develop a “singer’s formant” around 3000 Hz that allows them to cut through the orchestra
- Speech therapy uses formant analysis to correct speech disorders
FAQ
Q: Why do a man and a woman pronouncing the same “a” have different voices but the same recognizable vowel? A: The fundamental frequency (pitch) differs, but the ratios between the formants remain similar. Our brain identifies these ratios, not the absolute values.
Q: How do you see formants with FizziQ? A: On the spectrogram, formants appear as more intense horizontal bands. F1 is the lowest band, F2 is the next one.
Q: Do formants change with age? A: Yes, because the vocal tract grows. Children have higher formants than adults.
Related concepts
Spectrogram - Timbre - Resonance Frequency - Fundamental Frequency - Vocal Cords