Voice analysis, explained
What is a voice analyzer?
A voice analyzer is a tool that measures the acoustic properties of a voice recording — most importantly pitch (fundamental frequency), pitch range, and how those values move through a sample. This free online voice analyzer combines those measurements with a spectral-balance proxy and recording-quality checks to produce an explainable acoustic profile.
The analyzer runs completely in your browser. When you press record, your device captures a short sample and a YIN pitch-detection algorithm inspects stable voiced frames locally. The result is a snapshot of this recording, not a determination of who you are.
What this voice analyzer measures
- Pitch — the median fundamental frequency of your recording, shown in Hertz.
- Pitch range — the P10–P90 span, which shows how much your pitch moved through the sample.
- Pitch movement — measured in semitones, describing how expressive the intonation is.
- Spectral balance — a local brightness proxy computed from the recorded signal.
- Recording quality — voiced-seconds, frame count, and a confidence label that tell you how reliable the measurement is.
This page summarizes a recording, so the range you see is the movement inside that one sample. If you want the lowest and highest notes your voice can reach instead of the span of a single reading, the vocal range test glides through your comfortable range and reports the detected notes directly.
How to analyze your voice
- Find a quiet room and keep your microphone about a hand's length away.
- Press the record button and read the fixed passage, or speak naturally for 10–15 seconds.
- Stop, and the analyzer shows your pitch profile, range, confidence, and acoustic cues.
- Repeat with the same passage and setup to compare sessions over time.
How to interpret an acoustic voice profile
Read recording confidence first. It summarizes whether the browser found enough clear voiced material for a stable calculation; it is not confidence about identity. Then consider median pitch alongside the P10–P90 range and pitch movement. A similar median can hide very different contours, while a broad range can reflect natural emphasis rather than poor control.
The presentation reference combines only the cues disclosed in the report. It describes where this recording sits on a broad acoustic map, not who made it. Voices overlap extensively, and language, accent, articulation, listening context, equipment, and cultural expectations remain important. The tool must not be used to assign gender, sex, age, personality, or personal identity.
Make repeated analyses comparable
Use the same passage, room, device, microphone distance, posture, volume, and speaking intention. Record two or three ordinary samples rather than repeating until a preferred score appears. If you want a history, write down the setup and complete metrics yourself—the site deliberately does not save accounts, recordings, or result timelines.
Expect normal variation from different words, sentence stress, fatigue, time of day, emotion, and microphone processing. A difference between sessions is an observation, not proof that a training method caused it. Several controlled measurements plus careful listening provide better context than one headline number.
If you are preparing to sing or present, the guided voice warm up exercises are a reasonable first step before recording, since a gentler start can make the task more consistent than beginning cold.
When you review a sequence, compare like-for-like metrics instead of selecting only the result you prefer. Check whether confidence stayed adequate, whether the median moved in the same direction across ordinary attempts, and whether the central range widened or narrowed with the task. Keep a separate baseline for fixed reading and spontaneous speech because their intonation demands differ. Small day-to-day changes are expected; a repeatable pattern under similar conditions is more informative than an isolated high or low value.
Technical and ethical limits
The analyzer estimates fundamental frequency and a limited spectral-energy proxy. It does not directly measure formants, vocal-tract anatomy, placement, nasality, breathiness, articulation, accent, emotion, or clinical resonance. Noise, clipping, creaky or breathy sound, overlapping speakers, and strong room reflections can reduce accuracy or cause occasional octave errors.
Record only yourself or someone who has clearly agreed. Do not use the report for employment, education, housing, credit, insurance, healthcare, access, authentication, surveillance, or judgments about another person. This is an educational browser measurement, not a diagnostic or eligibility tool. Pain, persistent hoarseness, or a sudden lasting change needs appropriately qualified professional advice.
Illustrative result and troubleshooting
An illustrative fixed reading could show a median pitch of 175 Hz, a typical range of 148–207 Hz, moderate pitch movement, and high recording quality. These values are examples only, not a target or a population norm. The median describes the centre of the accepted frames, the range describes movement in that passage, and the quality label tells you whether the signal was clear enough to interpret.
If the result is missing or implausible, check permission, microphone selection, room noise, echo, clipping, and distance first. Speak naturally for 10–15 seconds and repeat the same passage. Whispering, vocal fry, breathiness, Bluetooth processing, and another speaker can reduce stable frames. This tool is a good fit for personal learning, recording checks, and self-directed practice; it is not a fit for medical diagnosis, speaker authentication, hiring, access decisions, or judging another person.
Why this voice analyzer is different
Every calculation — YIN pitch detection, spectral analysis, statistics — happens in your browser. There is no account, no audio-upload endpoint, and no history database. You can verify this yourself in your browser's network panel while recording: no request carries your audio or derived features.
Looking for real-time pitch while speaking or singing? Try the free online pitch detector for a live Hz and musical-note readout.
Voice Gender Detector
A voice gender detector is another name people use for a recording-level presentation check. Our voice gender detector page explains that quick-check workflow. It uses the same local engine; this analyzer page focuses on interpreting the pitch, movement, spectral proxy, and recording confidence together.
Neither name means that a voice can reveal gender identity or biological sex. The measurements describe one sample, and listener perception depends on cues and context beyond this report.