From an Australian study in Scientific American:
Although singers can generate very loud sounds, how can they compete with a large and enthusiastic symphony orchestra?
One strategy is to maximize their sound output at frequencies above 2,000 Hz. This is because an orchestra is typically loudest around 500 Hz, with the sound level dropping off quickly at higher frequencies. Furthermore, the ear is most sensitive around 3,000 to 4,000 Hz. To this end, singers often modify the resonances of their tract to produce a characteristic “vocal ring” that considerably boosts the sound output in this frequency range. This is of more value to lower pitched voices than to sopranos.
Ah, but how? Read on here.