Using Formant control

What are formants?

Formants can be described as a sonic “footprint” of an acoustic space. The practical effect of formants could be compared to a multi-peak filter acting on the frequencies in a sound. An acoustic guitar, for example, has a body shape which makes is sound the way it does. The same goes for a human vocal tract (throat and mouth cavity); every human vocal tract has a unique “shape” which gives the voice its character. It is these shapes that produce formants. A big difference between a vocal tract and a guitar body, though, is that the vocal tract changes shape as you sing different vowels. This also means that the formants will change.

When you pitch shift a signal up or down, the formant “multi-peak filter” will move up or down with the signal frequency (just like a traditional synthesizer filter would with full keyboard tracking activated). The result will be a signal that not only is pitched but also changes character. In some situations this might be what you want, but when it comes to pitch shifting vocals you will probably often want the pitch-shifted signal to sound like it’s sung by the same person. Therefore, Neptune features a formant control function.

The Formant section in Neptune lets you control the formants so they don’t move along with the pitched signals. Neptune continuously samples and analyses the input audio and determines both the pitch and the current formants of the signal in real-time. The formants are then automatically applied to the pitch-adjusted output signal in real-time.

The picture below shows schematic examples of a 500 Hz signal, pitch-shifted down and up one octave respectively, without and with formant correction. The formants are the peaks of the gray dotted lines in the graphs:

Pitch-shifting without and with formant correction applied. The left column shows -1 octave pitch-shifting and the right column +1 octave.

Using the Formant function

When the Formant section is not active, the formants will move along with the pitched signal. On vocals this could generate unnatural “gender change” effects, especially on heavily pitched voices. If this is not the desired effect, you should activate the Formant section:

1.	Activate the Formant function by clicking the Formant button.

2.	Adjust the Formant shift with the Shift knob.

The Shift range is ±1 octave.

•	The default 12 o’clock position means that the formants will be “locked” to the input signal and won’t move with the shifted pitches.

Used on vocals this will give the impression of the same person singing at the adjusted pitches. (Refer to the picture in the “What are formants?” section for graphical examples of this particular Shift setting.)

•	Turned further counter-clockwise, the formants will be “locked” to the input signal and displaced downwards relative to the input signal.

Used on vocals this will give the impression of a more deep male character voice.

•	Turned further clockwise, the formants will be “locked” to the input signal and displaced upwards relative to the input signal.

Used on vocals this will give the impression of a more soprano/female/child character voice.

Note that Voice Synth has its own separate “built-in” formant correction and is not affected by the Formant section settings.