Read a short passage aloud and VoiceScope maps your pitch in real time — frequency in hertz, musical notes, and how it moves over time. Everything runs in your browser; audio never leaves your device.
Free browser-based pitch analysis
VoiceScope is a free online voice pitch analyzer for speech and singing. It records from your microphone, estimates fundamental frequency (F0), and turns each voiced moment into a frequency, musical note, and readable pitch contour. Processing stays in your browser.
Pitch is the perceptual counterpart of fundamental frequency: how quickly the vocal folds repeat a vibration. VoiceScope reports that estimate in hertz and maps it to the nearest equal-tempered note. Median pitch describes the centre of a recording, while the contour shows changes that a single average cannot reveal.
A short voice test often returns one number. VoiceScope also keeps the underlying time series, highlights trimmed outliers, lets you replay the exact recording, and stores recent sessions locally for comparison. That makes it useful for checking speaking pitch, vocal exercises, intonation, and detector errors rather than treating one number as a verdict.
Microphone analysis and recording history run locally on your device. VoiceScope measures F0, not attractiveness, health, gender, resonance, or the complete character of a voice. Results are a statistical and acoustic reference, not a medical diagnosis.
Speaking F0 varies by person, age, physiology, language, emotion, and context. A range is more useful than a hard category. VoiceScope shows a descriptive statistical reference while keeping the raw recording available for review.
No application server receives the recording. Analysis and history use browser APIs and local IndexedDB storage. An optional anonymous GoatCounter page-view request is separate from microphone processing.
Yes. A steady sung note normally produces a clear note and frequency. The tool is designed around voice F0 rather than instrument tuning, so it does not display cents deviation from concert pitch.
Smart mode looks for short, locally inconsistent points with weak detector confidence. Sustained changes and clear short phrases are retained; isolated low-confidence glitches are marked as trimmed.