Tone Generator

Generate precise audio tones from 20 Hz to 20,000 Hz directly in your browser. Choose from sine, square, triangle, and sawtooth waveforms. Use dual-tone mode to layer two frequencies, or sweep across a range for speaker and hearing tests. Real-time oscilloscope and spectrum visualization included — all processed locally, no data uploaded.

Tone Generator Tool

🎹 Interactive Piano

Click keys to play notes — uses current waveform & volume settings

🧪 Hearing Range Test

Test your hearing range across 17 frequencies from 125 Hz to 20 kHz

Understanding Your Output

Circular Gauge & Musical Note

The large green number inside the circular gauge shows the current frequency in Hz (Hertz — cycles per second). The glowing arc fills proportionally on a logarithmic scale from 20 Hz to 20 kHz. Below, the nearest musical note is displayed (e.g., A4, C5, G#3), making the tool useful for tuning instruments or generating reference tones.

Waveform Types

Each waveform has different harmonic content:

• Sine — pure tone, single frequency, no harmonics. The fundamental building block of all sound.
• Square — contains odd harmonics (3rd, 5th, 7th), producing a bright, buzzy, retro sound.
• Triangle — also odd harmonics but with much lower amplitude, creating a softer, mellower tone.
• Sawtooth — contains all harmonics (odd and even), producing the richest, most complex sound.

ADSR Envelope Shape

The ADSR canvas shows a visual representation of how your tone’s volume changes over time. Attack is the ramp from silence to peak volume. Decay is the drop from peak to the sustain level. Sustain is the steady-state level held while the tone plays. Release is the fade-out after you stop. This shapes the “character” of the sound — a long attack creates a swell, while a short attack creates a percussive hit.

Stereo Panning

The pan slider controls audio placement between your left and right speakers or headphone channels. Center sends equal volume to both. Moving toward L or R progressively reduces volume in the opposite channel. This is useful for stereo speaker testing, verifying headphone balance, or creating spatial audio effects.

Noise Types

Three noise colors are available, each with distinct spectral characteristics:

• White Noise — equal energy at every frequency. Sounds like TV static. Used for acoustic measurements and masking.
• Pink Noise — energy decreases at 3 dB per octave. Sounds balanced across frequencies. Used for speaker calibration and sleep.
• Brown Noise — energy decreases at 6 dB per octave. Deep, rumbling sound. Popular for focus and relaxation.

dB Level Meter

The vertical VU meter shows the real-time output level in decibels (dB). Green indicates safe levels, yellow indicates moderate levels, and red indicates high levels. The red peak hold line marks the highest level reached in the last 2 seconds, helping you monitor peaks without constantly watching the meter.

Interactive Piano

The 3-octave piano keyboard (C3–B5) lets you click keys to play notes using the current waveform and volume settings. White keys play natural notes, black keys play sharps/flats. Each key shows its note name, and when pressed, the circular gauge and frequency display update to show the exact Hz value.

Hearing Test Results

After completing the hearing test, a color-coded bar chart displays your results. Green bars indicate frequencies you heard; red bars indicate frequencies you could not hear. The highest green bar represents your upper hearing limit. Typical results: young adults hear up to 16–18 kHz; adults over 40 typically reach 12–14 kHz.

WAV Export

The WAV Export renders your tone offline using an OfflineAudioContext, capturing all current settings (frequency, waveform, volume, and ADSR envelope). The resulting file is a standard 44.1 kHz, 16-bit PCM WAV — universally compatible with audio editors, DAWs, video editors, and media players.

How the Tone Generator Works

Web Audio API Architecture

This tool uses the Web Audio API — the browser’s built-in audio processing engine. The full signal path is: OscillatorNode → GainNode (volume + ADSR control) → StereoPannerNode (L/R panning) → AnalyserNode (oscilloscope + spectrum + dB) → AudioContext.destination (your speakers). Noise generators use AudioBufferSourceNode with pre-computed buffers connected to the same chain.

ADSR Implementation

Envelope shaping uses the GainNode’s linearRampToValueAtTime() method to schedule precise amplitude changes. Attack ramps from 0 to peak, Decay ramps from peak to sustain level, Sustain holds the level, and Release ramps from sustain to 0. All transitions are sample-accurate, scheduled on the audio thread for glitch-free playback.

Noise Generation

White noise is generated by filling an AudioBuffer with uniform random values. Pink noise uses the Voss-McCartney algorithm with 7 rows of filtered random values to achieve the characteristic -3 dB/octave rolloff. Brown noise (Brownian motion) accumulates random values with a leaky integrator, producing the -6 dB/octave spectrum. All buffers are 2 seconds long and loop seamlessly.

WAV Export Process

The WAV download uses an OfflineAudioContext to render audio without real-time playback. The oscillator and gain nodes are configured identically to the live signal path, including ADSR envelope scheduling. The rendered AudioBuffer is then converted to a 16-bit PCM WAV file using manual DataView byte writing with proper RIFF/WAVE headers.

Frequency Accuracy

The Web Audio API generates frequencies with sub-Hz accuracy at the system’s native sample rate (typically 44,100 or 48,000 Hz). IEEE 754 double-precision floating-point ensures scientific-grade precision. Smooth transitions use exponentialRampToValueAtTime() for logarithmic sweeps that sound natural to human ears.