It is my understanding that when the voice cracks the vibration of the vocal folds have become asymmetrical. That is, the left and right vocal folds are not opening and closing at the same time. One of the folds will move to close before the other one is ready to do so. This action implies that something of sufficient force has disturbed them from their natural tendency for equal perturbation. That force is initiated by the sudden removal of a standing wave in the vocal tract that is providing a stable oscillating environment through the effect of “inertance” (See Ingo Titae).
The singer causes this by not correctly adjusting the resonance of the vocal tract to maintain the existence of a standing wave. Put another way, the singer is not adjusting his vocal tract formant spaces to tune to one of the harmonics of the sung tone. Adjusting a formant to match a harmonic will create a pneumatic spring,as it were,that provides a maximum efficient transfer of vocal fold energy to the air in the vocal tract. The vocal folds can, as it were ‘lean on’ this stable energy transfer. Titze has created the term “inertance” to identify this phenomenon.
But if that stable energy transfer is suddenly lost, the vocal folds can become asymmetric in their oscillation.
This most commonly occurs when the voice is entering the lower portion of the passaggio and the need to change resonance strategies is required (as outlined in my previous post). It will also occur when the voice is singing near the top of its range, the moneyed high note. In this situation the whole vocal mechanism is in its most tenuous balanced configuration and the vocal folds are more dependent than ever on the positive effects of inertance. The slightest alteration of the stability of the whole vocal mechanism will produce an asymmetric vocal fold function and the terrible sounding ‘crack’ occurs.
All of the above implies that the singer is performing acoustically, without a microphone. Acoustic singing is the only singing that always requires the most efficient adjustments of the vocal tract in order to be heard in a large venue or over a large orchestra. When singing with a microphone such an efficient vocal production is not as required. The voice can be easily heard because it is amplified. Inheritance can be applied but it is not as necessary and, as a result, the potential to have the voice crack is not as great.
Cracking can occur rather frequently during belting even though belting is usually amplified. This is because the belting singer does not make the resonance changes at the passaggio as is required in classical singing. This is done by choice in order to create the “yell like” quality that is desired in present day Broadway and other pop forms and is considered a more direct infusion of emotion into the tone.