A formant is a resonant spectral feature associated with the filtering of speech by the vocal tract. Its centre frequency and bandwidth describe different aspects of the feature; the first few formants are conventionally labeled . Praat represents estimated formant structure as time-sampled centre frequencies and bandwidths.[ref]
Formant Analysis in the Praat Manual by Paul Boersma. Formant object reference: frame and formant attributes
A simple physical model treats the vocal tract as a uniform tube, closed at one end and open at the other. Under those ideal boundary conditions its resonance frequencies arewhere is the speed of sound and the tube length. For and , the first three model resonances are 500, 1500 and 2500 hertz. A real vocal tract has a varying cross-section, losses and moving articulators, so this illustrates the mechanism rather than predicting every vowel.
These resonances are distinct from the fundamental frequency of vocal fold vibration. The source-filter model separates the repeating excitation from the filter's spectral envelope, explaining how a speaker can change vowel quality without making the same change in pitch.
Articles by others on the same topic
A formant is a concentration of acoustic energy around a particular frequency in the sound spectrum of speech. In phonetics, formants are crucial for characterizing vowel sounds as they represent the resonant frequencies of the vocal tract. When a person speaks, the shape and configuration of the vocal tract (which includes the throat, mouth, and nasal passages) filter the sound produced by the vocal cords, creating these resonant peaks.