Reviewed 20 September 2026Quick answer: lower the hardware volume, choose 5 seconds, 30 seconds or continuous playback, then start one noise color. Continuous mode remains active only while this tab is visible and stops immediately when requested. This page generates sound; it does not listen to the room, measure frequency response or calibrate speakers.
Browser noise generatorPlay pink, white or brown noise
Lower the output volume before starting. Choose five seconds, 30 seconds or continuous playback, then select a noise color.
Ready. Start with pink noise at a low hardware volume.
No microphone, recording or spectrum analysis. Continuous mode stops when you press Stop, leave the page or hide the tab. These are not calibrated noise levels or sleep tracks.
How long can the noise generator play?
Five-second mode is the default for a quick output check. Thirty-second mode supports a longer comparison. Continuous mode loops a locally generated ten-second buffer until you press Stop, leave the page or hide the tab. It does not run as an unattended sleep track, collect microphone input or claim a focus, therapy or hearing benefit.
Download original noise-test WAV files
Each file is ten seconds of deterministic mono PCM with short start and end fades. The files contain no speech, tracking, metadata or third-party audio.
Pink noise WAV
Filtered random signal with progressively less energy toward higher frequencies. Mono, 16-bit PCM, 44.1 kHz, approximately 882 KB.
Download pink noiseWhite noise WAV
Random signal with approximately equal power per unit of frequency. Mono, 16-bit PCM, 44.1 kHz, approximately 882 KB.
Download white noiseBrown noise WAV
Integrated random signal with stronger low-frequency emphasis. Mono, 16-bit PCM, 44.1 kHz, approximately 882 KB.
Download brown noiseThe files have a digital peak ceiling of 3% of full scale, but that is not an acoustic loudness measurement. Amplifiers, speakers, headphones and system processing determine the sound level that reaches a listener.
What changes between pink, white and brown noise?
Pink noise
The filter reduces energy toward higher frequencies, producing a more octave-balanced comparison than white noise. This implementation is a practical browser and file signal, not a laboratory reference.
White noise
The random samples have approximately equal power per unit of frequency. It often sounds brighter because wider high-frequency bands contain more energy.
Brown noise
The random signal is integrated so energy is weighted more strongly toward lower frequencies. It can sound much darker and does not mean the output has stronger bass response.
Why the samples differ each time
Noise is not a repeating sine tone. The live generator uses a fixed seed for repeatability within this implementation, while the color filters change the distribution across frequency.
How to use noise for a basic speaker comparison
- Select the intended speaker, headphone or receiver output and lower its hardware volume.
- Play pink noise once. Confirm that audio comes from the intended route and stop if it is harsh, uncomfortable or distorted.
- Keep the same volume and position if comparing another output or noise color. Change only one variable at a time.
- Use the downloadable files when the target device has no suitable browser.
- Confirm suspected problems with familiar audio and the equipment maker's test or service instructions.
What this page can and cannot tell you
A short noise sample can confirm that the selected route plays, provide a repeatable broad-spectrum stimulus and help you compare whether a rattle or interruption repeats. It cannot separate a speaker, room, amplifier, crossover, equalizer, codec, listener or hearing effect.
The output is mono. Use the speaker and sound test for left/right routing, the frequency sweep generator for a changing sine frequency, or the 5.1 surround test for a browser-reported six-channel route. If a phone speaker sounds muffled after recent liquid exposure, noise playback will not help; use the wet-phone guide instead.