Voice Pitch & Resonance Trainer

About the Voice Pitch & Resonance Trainer converter

The hard part of voice training is that you cannot hear yourself properly while speaking. Bone conduction makes your own voice sound different to you than it does to everyone else, so people practise for months without knowing whether the thing they are changing is the thing they think they are changing.

This shows two numbers as you talk — pitch and resonance — and keeps a record of each session on your device, so the slow change over weeks becomes something you can actually see.

Pitch, measured properly

Pitch appears as a live trace with your target range shaded behind it. Where the line breaks, the sound had no pitch — breaths and most consonants genuinely do not, and drawing a line through them would be inventing numbers.

The detection deliberately does not work by finding the loudest part of the spectrum. A voice's strongest partial is very often a harmonic rather than the fundamental, so that approach reports an octave too high with total confidence. This uses a time-domain method that asks a different question — at what delay does the sound most resemble itself — which is what makes the reading trustworthy on a real voice rather than only on a tuning fork.

Resonance, which most tools leave out

Pitch is the obvious number and the easy one to chase. Resonance depends on the size and shape of the vocal tract rather than on how fast the vocal folds vibrate, and it does a great deal of the work in how a voice is heard.

This is exactly why pitch-only practice so often disappoints. Someone raises their pitch, notices little change in how they are read, and concludes the work has failed — when the missing piece was resonance the whole time. Here the first two formants are shown alongside the pitch, together with the gap between them, which is a rough shorthand for how bright a vowel sounds.

Those figures are estimates and the page says so. Formant tracking through whatever microphone is to hand is not a precise measurement; it is far more useful as something to watch move while you experiment than as a number to take literally.

A record that means something

A single reading tells you almost nothing. Voice work changes slowly enough to be invisible day to day, and the only real evidence anyone gets is a line across weeks.

Each recorded session keeps a handful of figures: the median pitch, the range, how much of the time you stayed inside your target, and a steadiness measure. Steadiness is usually the number that improves first, well before the pitch itself settles.

The median is used rather than the average on purpose — one octave slip, which any detector produces occasionally, would drag an average badly while leaving a median untouched.

It stays on your device

The audio is examined frame by frame and thrown away. Nothing is recorded, nothing is uploaded, and no account is involved. Only the few numbers from a session you chose to save are kept, in this browser.

That matters more here than almost anywhere else on this site. A recording of your own voice, tied to a goal that may be private and not yet shared with anyone, is close to the last thing anyone should be asked to upload to a stranger's server.

Because the history lives in your browser, clearing your browsing data will delete it — so export is available from the first day rather than added later, and an imported file is merged with what is already there rather than replacing it.

Targets are a starting point, not a rule

The suggested ranges are commonly cited averages for speaking pitch, and they overlap heavily. Plenty of people sit outside them and are heard exactly as they intend.

They exist so the range control starts somewhere sensible, and every one of them is meant to be dragged to wherever suits you.

Frequently asked questions