The human voice as a musical instrument combines a non-uniform acoustic tube resonator (the vocal tract) with a forced, ``lip-like'' excitation mechanism (the vocal folds). Unlike the brass player's lips, however, the vocal folds are little affected by acoustic feedback from the vocal tract. A singer trains to control his/her vocal cord mechanism for accurate pitch specification as well as to control his/her vocal tract shape to modify formant structure for better projection.
Singers change some vowel sounds to improve musical tone.
The second formant is typically lower in frequency in the sung vowel.
Trained singers show a strong formant around 2500 - 3000 Hz, referred to as the ``singer's formant''. This formant, which seems to be independent of the particular vowel and pitch, adds brilliance and carrying power to the voice.
The singer's formant is attributed to a lowered larynx and widened pharynx, which forms an additional resonance cavity.
Voice registers correspond to differences in tone caused by different adjustments of the larynx. Two (``heavy'' and ``light'') or three (``chest'', ``middle'', and ``head'') registers are commonly identified.
In the heavy or chest voice, the thyroarytenoid muscles are active and hence shortened. They thicken the vocal folds, which results in their remaining closed over an appreciable part of each cycle of vibration.
In the light or head voice, the thyroarytenoid muscles are passive. In this state they offer little resistance to the cricothyroids, which apply substantial longitudinal tension to the vocal folds, lengthening and thinning them. The glottis closes only briefly, or not at all, and the resulting sound has fewer harmonics than the chest voice.