That's a rather defective understanding of sound production in music. First, human voice is not the "sound" of the vibrations of the vocal folds (to hear that, use a stethoscope on the larynx): it is the interruption of the air stream by the opening and closing of the folds, an effect similar to a siren where a stream of air is pushed through a rotating disk with holes in it.
A similar effect is used with reed instruments (free reed, single reed, double reed) and brass. Not, however, with woodwinds where the primary oscillator is the air column.
So with the reeds, voice, and similar implements, the primary oscillator (if you want to call a rotating disk even an oscillator) is used for modulation of an air stream, and the main power of the resulting sound comes from the pressure of the air stream rather than the vibrating elements themselves. While the pressure also provides the energy for sustaining the vibration of the primary oscillator, the energy of the oscillator itself tends to be quite smaller than what is put into the air stream.
A lot of other instruments (strings, percussion) however don't use this kind of indirect sound production. Instead, the vibrations of the primary oscillator are "transformed in impedance" by sound conducting elements and emanate from suitable surfaces. In this case, every bit of sound derives its energy from the original oscillation.
With flutes, oscillator, sound transferring medium and transmission to the listener are all air: the medium is the message.
"resonators" shape sound and its overtones but do not generally create it. They can however, like with reed and brass instruments, strongly feed back to the sound source by picking off and amplifying certain frequencies which are then strongly preferred modes of oscillation that get reinforced by the resonator.
Only with woodwinds are oscillator and resonator basically the same.