The repetition of grains at a constant rate results in an audible “frame rate”. For example, if a new grain is started every 10 ms, a 100 Hz component will be evident in the sound.
If this sound has strong spectral energy near 2 kHz, for example, the result would be a sound with harmonics at integer multiples of 100 Hz and a “formant” in the vicinity of 2 kHz.
“Vibrato” can be created by slowly varying the time spacing between grains.
A technique for synthesizing vocal sounds using short decaying sinusoidal bursts called FOFs (fonctions d'onde formantique or formant wave functions) can be viewed as a form of granular synthesis.
The FOF speech synthesizer of Xavier Rodet makes use of 5 parallel FOF oscillators, each with controls for center frequency, bandwidth, amplitude, and decay.