Evidence-led voice reference

Vocal Range Statistics

Modeled reference only. Verified first-party sample: N=0; no “real tester” claims until the collection pipeline is abuse-resistant.

Voice-type references

Sustained-note measurement

Measure the range you can hold

Audio stays on this device

Current note
Microphone is offClarity gate ready
ReadyYour audio stays on this device.Microphone is off
Confirmed low Confirmed high

The curve below is a clearly labeled normal model centered at 22 semitones (about 1.8 octaves), not observed site data. It explains the current percentile placeholder without pretending a dataset exists.

Modeled distribution · not observed tester data

Range-span model

Normal model: mean 22 semitones, standard deviation 6.

Verified first-party sampleN=0
  1. Under 1 octave6%
  2. 1–1.5 octaves25%
  3. 1.5–2 octaves38%
  4. 2–2.5 octaves24%
  5. Over 2.5 octaves7%

Result submission is disabled until D1 storage, server validation, single-use tokens, rate limiting, consent, and robust aggregation are configured and verified.

Questions and limitations

Are these results from real vocalrange.org testers?

No. The verified first-party sample is currently zero and result submission is disabled. The displayed distribution is a rough model based on commonly cited adult-range norms; it will not be relabeled as observed data until abuse controls, consent, sample size, and update cadence are operational.

Why not collect anonymous results immediately?

A public write endpoint can be spammed and make the dataset worthless. Collection requires server-side plausibility checks, short-lived single-use completion tokens, rate limiting without retaining IP addresses, browser deduplication, robust quantiles, consent, and transparent sample counts.

Before the browser asks

Your microphone stays local

The browser sends short audio buffers to this page’s pitch detector. Audio is not recorded, uploaded, or used for statistics.