This page explains the technical methodology behind all 20 tools on VoiceRangeTest.com — how each tool works, what it measures, how results are calculated, and what technology is used. The tools on this site span three distinct technical categories: voice and microphone input tools, audio output and frequency tools, and music theory and ear training tools. Each category uses a different approach, and each is documented separately here.
Understanding the methodology helps you use the tools correctly and interpret results with accurate expectations. Written and maintained by Harlow, founder of VoiceRangeTest.com.
Category 1 — Voice and Microphone Input Tools
These tools require access to your device microphone to capture and analyse your voice in real time. All audio processing occurs locally within your browser session using the Web Audio API. No audio is recorded, transmitted to any server, or retained after you close the page.
The Core Pitch Detection Pipeline
All microphone-based tools on this site share the same underlying pitch detection pipeline:
Step 1 — Microphone capture. Your browser requests microphone permission via the standard browser permission prompt. Audio is captured as a continuous stream using the Web Audio API’s MediaStream interface. Permission is granted by you explicitly — the microphone is never accessed without your approval.
Step 2 — Signal windowing. The continuous audio stream is divided into overlapping short time windows — typically 20 to 50 milliseconds per window. This allows the algorithm to analyse pitch at frequent intervals and respond to changes in real time without waiting for long sections of audio to complete.
Step 3 — Fast Fourier Transform (FFT) analysis. Each time window is processed using an FFT algorithm, which converts the raw waveform data from the time domain into the frequency domain. The output is a frequency spectrum — a map showing which frequencies are present in the audio signal and at what amplitude. This is the mathematical foundation of all pitch detection on this site.
Step 4 — Fundamental frequency identification. From the frequency spectrum, the algorithm identifies the fundamental frequency — the lowest and strongest frequency component in the signal. This corresponds to the pitch you are actually producing. For example: singing A4 produces a fundamental frequency of approximately 440 Hz. Singing C4 (middle C) produces approximately 261.63 Hz.
Step 5 — Stability filtering. Transient noise, breath sounds, and unstable pitch transitions are filtered out. A frequency must be detected consistently across multiple consecutive analysis windows before it is logged as a confirmed pitch. This prevents background noise, room acoustics, or note transitions from registering as false detections.
Step 6 — Note mapping. The confirmed fundamental frequency is mapped to the nearest musical note using the standard equal temperament tuning system, where A4 = 440 Hz. Each semitone is separated by a frequency ratio of the 12th root of 2 (approximately 1.0595). The result is expressed as a note name and octave number — for example, G3, B4, or C5.
Tool-by-Tool Methodology — Microphone Input
Voice Range Test and Vocal Range Calculator Both tools record the lowest and highest stable pitches detected during your session. As you sing from your lowest to highest notes, the pitch detection pipeline logs every confirmed stable pitch. The lowest and highest values define your measured range. Results are expressed as two note names (e.g. E2 to C5) and the span between them in semitones and octaves.
Voice Type Test Your measured lowest and highest notes are compared against the standard pitch ranges for each of the six classical voice types — bass (E2–E4), baritone (A2–A4), tenor (C3–C5), contralto (F3–F5), mezzo-soprano (A3–A5), and soprano (C4–C6). The voice type whose range best matches your measured result is returned as the classification. This is an acoustic estimate based on pitch range only — timbre, tessitura, and register transition characteristics are outside the scope of automated pitch detection.
Deep Voice Test This tool focuses specifically on the lower boundary of your range. It applies stricter stability filtering at low frequencies, where detection is more susceptible to room resonance and microphone low-frequency roll-off. The result is your lowest reliably detected note.
Pitch Accuracy Test This tool displays a target note and measures how closely and consistently your voice matches it. Your detected pitch is compared against the target frequency at each analysis window. The percentage of time your pitch falls within an acceptable tuning window (approximately ±10–15 cents of the target) is calculated as your accuracy score. A score of 70–80% is typical for untrained singers on sustained notes; trained singers typically score 85–95%.
Pitch Detector Displays your current detected note name, frequency in Hz, and tuning position (flat, in tune, or sharp) in real time, updating at the rate of the pitch detection cycle. A note is displayed as “in tune” when the detected frequency falls within approximately ±10 cents of the target pitch centre.
Note Identifier Identifies the closest musical note to the frequency you are producing and displays the note name and octave. Differs from the Pitch Detector in that it identifies what note you are closest to rather than showing your tuning deviation from a target.
Frequency Finder Displays the exact frequency in Hz of the pitch you are producing in real time. Where the Pitch Detector shows a note name, the Frequency Finder shows the raw acoustic frequency — useful for users working with audio software, acoustic calibration, or frequency-sensitive musical contexts.
Perfect Pitch Test A target note is played through the speaker and you identify it by name without a reference tone. The tool presents notes randomly from across the chromatic scale. Your identification accuracy across multiple trials produces a perfect pitch assessment score. This tool measures aural note recognition — a perceptual skill — not microphone input accuracy.
Vibrato Analyzer Analyses the cyclical pitch variation in your sustained singing tone. Vibrato is detected by tracking the rate (oscillations per second, measured in Hz) and depth (pitch deviation from centre, measured in cents) of periodic pitch fluctuation. Standard healthy vibrato oscillates at approximately 5–7 Hz with a depth of ±50–100 cents. The tool displays rate and depth in real time as you sustain a note.
Category 2 — Audio Output and Frequency Tools
These tools generate or measure audio rather than detecting vocal input. They do not require microphone access for their core function. They use the Web Audio API’s audio output capabilities — the OscillatorNode, AnalyserNode, and related interfaces — to synthesise, play, and measure sound.
Tone Generator — Waveform Synthesis Methodology
The Tone Generator produces pure audio tones at any frequency using four selectable waveform types:
Sine wave — a pure tone with only the fundamental frequency and no harmonic overtones. This is the acoustically simplest waveform and is used for tuning reference, audio calibration, and hearing tests. Musical instruments produce sine-like tones when played very softly.
Square wave — contains the fundamental frequency plus all odd harmonics (3rd, 5th, 7th, etc.) at decreasing amplitudes. Produces a hollow, buzzy timbre. Used in electronic music and as a reference signal for audio testing.
Sawtooth (Saw) wave — contains the fundamental plus all harmonics (both odd and even) at decreasing amplitudes proportional to 1/n. Produces a bright, sharp timbre. Often used to simulate brass and string instrument timbres in synthesis.
Triangle wave — contains the fundamental plus odd harmonics at amplitudes proportional to 1/n². Softer than square wave, closer to sine wave in timbre. Used for musical synthesis and audio testing.
Frequency is adjustable from 20 Hz (the approximate lower limit of human hearing) to 20,000 Hz (the approximate upper limit). The ±1 Hz and ±10 Hz fine-tune controls allow precise frequency adjustment. Octave Up and Octave Down buttons shift the frequency by exactly a factor of 2. Volume control adjusts output amplitude via the Web Audio API’s GainNode.
Audio Frequency Test Plays a series of tones across the audible frequency spectrum and records which you can hear. Tests the upper and lower limits of your hearing range. Results indicate the frequency range your hearing is responsive to under the test conditions. Accuracy depends on your speaker or headphone quality and ambient noise level.
Sound Decibel Meter Uses microphone input to measure the amplitude of ambient sound in decibels (dB SPL). The Web Audio API’s AnalyserNode captures RMS (root mean square) amplitude and converts it to an approximate dB reading. Useful for checking practice room volume levels and understanding how loud your singing voice is. Note: browser-based dB measurement is an approximation — it depends on microphone sensitivity and gain settings. Results should be interpreted as relative measurements rather than calibrated absolute readings.
Microphone Test Verifies that your microphone is functioning and accessible to the browser before you use other voice tools. Displays a live audio level meter and confirms that audio signal is reaching the Web Audio API. No pitch detection is performed — this is purely a hardware verification tool.
Category 3 — Music Theory and Ear Training Tools
These tools are built on music theory logic, interval databases, and structured listening exercises. They do not use the Web Audio API for input, though several use it for audio output (playing reference notes and intervals).
Ear Training Test Plays a musical note or interval and asks you to identify it by name. The test draws from a database of chromatic notes across multiple octaves, presents them in random order, and scores your identification accuracy. Feedback is immediate after each response. The test methodology follows established ear training pedagogy — moving from basic note identification to interval discrimination as proficiency develops.
Interval Ear Training Plays two notes sequentially and asks you to name the interval between them. The 12 intervals of the chromatic scale are tested — minor 2nd through major 7th and octave. Each interval is associated with a characteristic sound pattern that trained musicians learn to recognise. The tool presents intervals in random root-note and ascending/descending configurations to train recognition independent of absolute pitch.
Song Key Finder Identifies the most likely musical key of a song based on input from the user about the notes or chords it contains. The key finder uses tonal centre analysis — comparing the submitted note or chord set against the 24 major and minor keys to find the best match by the number of notes that belong to each key’s scale. Results are ranked by probability of match.
Vocal Scale Finder Takes your measured vocal range as input and identifies which musical scales and keys fall comfortably within it. The tool maps your range against the note sets of major, minor, pentatonic, and modal scales to suggest the most singable options for your voice.
Online Metronome Generates a rhythmic click track at a user-selected tempo in beats per minute (BPM). The timing is implemented using the Web Audio API’s AudioContext clock — the most precise timing source available in a browser, accurate to sub-millisecond precision. The metronome supports time signatures from simple to compound and allows adjustment from 20 to 300 BPM.
Vocal Warm-Up Generator Generates a customised warm-up sequence based on your voice type and range. The exercise selection draws from an established set of vocal warm-up exercise types — lip trills, humming scales, vowel exercises, sirens, arpeggios — and orders them according to standard vocal pedagogy practice: gentle onset first, progressive range extension second, register bridging third.
Singer Range Comparison Methodology
The singer comparison features across the tools draw from a database of documented singer ranges maintained by Harlow. All singer range figures are cross-referenced from multiple recorded sources. The database distinguishes between working range and documented extreme range where data is available. The full research methodology for singer range data is in the Editorial Guidelines.
Privacy and Audio Handling
All microphone audio is processed locally in your browser. No audio is recorded, transmitted, or stored. When you close a tool page, all audio data is cleared. Full details are in the Privacy Policy.
Related Pages
- Accuracy and Limitations — what affects result accuracy and what the tools cannot measure
- Editorial Guidelines — how singer range research is conducted
- About the Author — Harlow’s research background and expertise
- FAQ — common questions about results and tools
- Contact — report a technical issue or inaccuracy
This Testing Methodology page is written and maintained by Harlow, founder of VoiceRangeTest.com. Last updated: June 2026.