White Noise Generator
A white noise generator builds a steady hiss in the browser instead of playing a recording. This one synthesises six sounds with the Web Audio API — White, Pink and Brown Noise plus Rain, Ocean and Fan — from a two-second buffer set to loop, and adds a 0-100% volume slider and a sleep timer of 15 minutes, 30 minutes, 1 hour or 2 hours.
Sleep Timer
Popular Tools
About White Noise Generator
Nothing here is a recording. Press the round play button in the middle of the visualiser and the page fills a two-second audio buffer with computed samples, then loops it through the Web Audio API, so sound starts instantly and downloads nothing. The bars are the live spectrum of what you hear, tinted to match the sound.
The six buttons differ in how those samples are made. White Noise is raw random values, an even spread across the spectrum. Pink Noise runs the same values through seven stacked filters that roll the level off as frequency rises, which is why it sounds softer. Brown Noise averages each new value into the previous one, which rolls off harder still and sounds deep. Rain boosts about three samples in a hundred to four times the rest; those spikes are the drops. Ocean is brown noise behind a 500 Hz low-pass filter whose cut-off is swung by a very slow oscillator, about one swell every eleven seconds. Fan is white noise through a band-pass centred near 420 Hz, so only a narrow hum survives.
Two controls sit underneath. The slider sets gain directly, so its percentage is amplitude, not perceived loudness: 50% is roughly 6 dB below full scale. Sleep Timer offers Off, 15 min, 30 min, 1 hour and 2 hours, and during playback a line reads "Sound will stop in" with a mm:ss countdown.
Limits: no mixing two sounds, no saving one as a file, no custom timer length, and the stop at the end is abrupt with no fade. Choosing a different sound mid-playback restarts the audio. Nothing is uploaded — the audio is generated on your device. This page is not a medical device and makes no claim about sleep, focus or health; on volume, the NIOSH recommended exposure limit is 85 A-weighted decibels averaged over an eight-hour workday, halving the safe duration for every 3 dBA more.
Use Cases
How to use
Choose one of the six sound buttons: White Noise, Pink Noise, Brown Noise, Rain, Ocean or Fan.
Press the round play button in the middle of the visualiser; the bars start moving in the colour of the sound you picked.
Drag the volume slider — the figure on its right is the gain as a percentage, from 0% to 100%.
Under Sleep Timer pick Off, 15 min, 30 min, 1 hour or 2 hours; while sound plays, the line underneath reads how long is left in mm:ss.
Press the same button again, now showing a pause icon, to stop the sound immediately.
Pro Tips
- Set the volume before choosing a sound: picking a different button while audio plays stops and restarts the audio, so you hear a short gap.
- The sleep timer only counts down while sound is playing, and every press of play restarts it from the full duration — pressing play twice therefore gives you two full sessions, not one.
- The loop is two seconds long, which is easiest to notice on Rain because the drop spikes repeat in the same places; White, Pink and Brown hide the loop point far better.
- The percentage beside the slider is gain, not loudness: 50% is about 6 dB below full scale, so a move from 10% to 20% changes what you hear much more than a move from 80% to 90%.
- There is no fade-out — when the countdown reaches 0:00 the sound cuts. If a hard stop is a problem, use a shorter timer and restart it rather than one long session.
Troubleshooting
I press play and hear nothing.
Browsers only allow audio to start from a real click, so use the round play button in the middle of the visualiser rather than a keyboard shortcut. Then check three things: the volume slider is above 0%, the browser tab is not muted, and your system output is the device you expect.
The sound stops when I switch to another app on my phone.
Mobile browsers suspend audio for tabs that lose focus or when the screen locks, and this page holds no wake lock to prevent it. Keep the tab in the foreground for the whole session; if playback was cut short, press play again, and remember that doing so restarts the sleep timer from the beginning.
The sleep timer restarted on its own.
It restarts from the full duration every time playback starts, and also when you tap a different timer chip while sound is playing. If you want a 15-minute session, set the chip first and press play once — pressing play a second time gives you another 15 minutes.
The visualiser is black or the bars are flat.
The canvas is only drawn while sound is playing, so it stays dark when stopped. If it is flat while audio plays, the volume slider is the cause: the analyser reads the signal after the gain stage, so at 0% there is nothing to draw. Raise the slider and the bars return.
The sound cuts off abruptly when the timer ends.
That is how the timer works — at 0:00 it stops the audio node outright, with no fade. There is no setting to soften it. Either lower the volume before the end or choose a shorter timer and restart it when you want more.
Frequently Asked Questions
It produces a continuous, unchanging sound that spreads energy across the audible range, so quieter irregular sounds are harder to pick out against it. This page generates that sound live rather than streaming a file, gives you six variants to choose from, a volume slider and a sleep timer that stops playback after 15 minutes, 30 minutes, 1 hour or 2 hours.
With the Web Audio API. The page creates a two-second buffer, fills it with values calculated in JavaScript, marks it to loop and routes it through a gain node to your speakers, with an analyser tapping the signal to draw the bars. No audio file is downloaded, which is why playback starts immediately and works the same whatever your connection speed.
They differ in how energy is spread across frequencies. White Noise is raw random samples, an even spread that sounds bright and hissy. Pink Noise passes those samples through seven stacked filters that reduce the level as frequency rises, so it sounds softer. Brown Noise averages each sample into the previous one, rolling off harder still and sounding deep and rumbling.
This page does not rank them, and it makes no health claim about either. The measurable difference is spectral: pink has progressively less energy in the high frequencies than white, which most listeners describe as less harsh. Which one suits you is a listening preference — switch between the two buttons at the same slider position and compare directly.
There is no answer this page can give you: it is a sound generator, not a study, and it takes no position on sleep. What it can tell you is what each button does — White is flat, Pink rolls off with frequency, Brown rolls off hardest, Rain adds spikes, Ocean swells about every eleven seconds and Fan is a narrow hum near 420 Hz. Try them at a low volume.
This page cannot advise you on that, but the volume question has a published reference: NIOSH sets a recommended exposure limit of 85 A-weighted decibels averaged over an eight-hour workday, and halves the allowed duration for every 3 dBA on top. Anything left running for hours is a duration question, which is why the sleep timer here stops at 15 minutes, 30 minutes, 1 hour or 2 hours.
Yes — the page runs in a mobile browser with no install and no account. Two caveats: it does not hold a screen-wake lock, so your display will dim on its normal schedule, and mobile browsers often suspend audio when the tab is backgrounded or the screen locks. Keep the tab in the foreground for an uninterrupted session.
No. White noise is a steady, even spread of energy across frequencies, and its whole point is that nothing in it stands out. A television is speech and music with a level that keeps changing, so it keeps attracting attention instead of blending into the background. Analogue TV static was close to white noise; a modern broadcast is not.
That is a clinical question and this page will not answer it. This is a sound generator with no medical purpose, no diagnosis and no treatment claim. If you have tinnitus, discuss sound therapy and safe listening levels with an audiologist or your doctor, and stop using any sound that makes your symptoms worse.
No. The audio only exists while the page is open, because it is generated on the fly rather than loaded from a file, and there is no export, no recording and no offline package. There is also no way to mix two sounds together or to enter a timer length of your own — the five chips are the only durations.