Start comfortably low
Use a clear vowel at a sustainable low note, not vocal fry.
Low and high note finder
Explore detected low and high notes during a gentle 15-second glide, without forcing either extreme.
Your lowest note, highest note and range span will appear here.
100% client-side. Audio is never uploaded. Results are stored only when you choose Save.
Use a clear vowel at a sustainable low note, not vocal fry.
Move through connected notes so the detector captures the range between extremes.
The useful result is your comfortable range, not a painful maximum.
Low and high notes are the 2nd and 98th percentile frequencies, rounded to the nearest equal-tempered note (A4 = 440 Hz). They are not proven comfortable singing limits.
The semitone span is calculated logarithmically between the low and high frequencies. This is a technical range estimate, not a final classical voice-type classification.
Explore detected low and high notes during a gentle 15-second glide, without forcing either extreme.
110–220 Hz spans 12 semitones. A brief isolated extreme may be excluded by the percentile endpoints.
Read the measurement guideAutocorrelation analyzes 2,048-sample frames about every 45 ms, within 55–1000 Hz. Frames need RMS ≥ 0.01 and correlation ≥ 0.55. A result requires 20 detected frames and at least 20% detection coverage.
Median, mean and standard deviation use detected frames. Endpoints are the 2nd and 98th percentiles; trimming extremes does not guarantee removal of octave errors.
Microphones, noise and octave errors affect results. Accuracy has not been independently validated. Pitch bands are interface descriptions, not population norms or diagnoses.
These tools share one pitch detector; each page offers a different way to interpret its measurements.
Begin on a comfortable low note, glide upward gradually, then explore your comfortable high notes without forcing.
Low and high notes are the 2nd and 98th percentile frequencies, rounded to the nearest equal-tempered note (A4 = 440 Hz). They are not proven comfortable singing limits.
No. Bass, baritone, tenor, alto and soprano classification also depends on tessitura, tone and trained comfort.
Median, mean and standard deviation use detected frames. Endpoints are the 2nd and 98th percentiles; trimming extremes does not guarantee removal of octave errors.