Free Background Noise Generator

What "color" of noise actually means

Noise has "colors" by analogy with light. White noise has equal energy at every frequency across the audible band, flat power spectral density (PSD), just as white light contains every visible wavelength. Pink noise drops 3 dB per octave: it has equal energy per octave rather than per Hz, so lower frequencies sound louder. The brain treats octave-equal energy as more natural because biological hearing perceives sound in octaves, not Hz. Brown noise (also called Brownian or red noise) drops 6 dB per octave: deeper, bassier, with very little high-frequency content. Brown is mathematically the integral of white noise (like a slow random walk), which is also why it's named after Robert Brown's Brownian motion of particles in a fluid.

These three exhaust the popular cases but they're not the only colors. Blue noise rises 3 dB per octave (the opposite of pink) and sounds harsh. Violet rises 6 dB per octave (the opposite of brown) and is the derivative of white noise, used in dithering for digital audio. Grey noise is shaped to sound equally loud at all frequencies given the human ear's frequency-dependent sensitivity (the inverse of the A-weighting curve). Most consumer-facing tools focus on white, pink, and brown because those map cleanly to "fan-like", "rain-like", and "thunder-like" subjective impressions and cover the spectrum from sharp to bassy.

Noise helps with sleep, focus, and tinnitus by a single mechanism: auditory masking. A loud broadband sound raises the threshold at which softer sounds become noticeable. A door slam or a phone notification that would otherwise jolt you awake gets covered by the steady noise floor. The brain quickly habituates to constant noise (it gets filtered as background) but still benefits from the masking effect on intermittent disruptive sounds. This is why noise machines have been a sleep aid since the 1962 Marpac Dohm became the household standard, and why the same principle works in browser tools today.

How this tool works under the hood

Every sample of audio is a number between -1 and 1. The generator runs an algorithm 44,100 times per second (the standard sample rate) to produce a continuous stream, feeds it to the Web Audio API, and the browser sends it to your speakers or headphones. Live synthesis means no audio files are downloaded and no loops are audible: each second of noise is freshly generated.

The algorithms per color: white is the simplest, each sample is just (Math.random() * 2) - 1, drawing a fresh random number per sample. Pink uses either the Voss-McCartney algorithm (sums multiple white sources at different update rates) or Paul Kellet's IIR filter from 1995 (5 to 7 filter coefficients tuned to give a 1/f spectrum within tenths of a dB). Brown uses a leaky integrator: each sample is approximately 0.998 * previous + 0.02 * white, which integrates the white noise input while preventing infinite drift. The leaky integrator is mathematically the discretization of Brownian motion's defining differential equation.

Mixing happens via the Web Audio API's GainNode per source plus a final master GainNode. Each volume slider sets a per-source gain (0 to 1); the master slider sets the overall gain. The auto-stop timer schedules a gainNode.gain.exponentialRampToValueAtTime(0.001, ...) at the chosen end time, fading the noise out over the last second rather than cutting abruptly. The tool never persists audio, never uploads anything, and the synthesis stops the moment you close the tab.

Brief history of noise as a tool

• Karl Pearson coins "white noise", 1932. The British statistician introduces the term by analogy with white light containing all wavelengths. Used initially in signal processing and statistics, not yet for sleep or focus.
• Marpac Dohm sleep machine, 1962. The first mass-market sleep-noise machine, a small box with a fan whose speed varies the spectral content. Becomes a household staple. The same product is still sold in 2026, six decades later.
• Mandelbrot and 1/f noise formalized, 1968. Benoit Mandelbrot and J.W. Van Ness publish the formal mathematics of fractional Brownian motion, giving 1/f^α noise (pink and brown sit at α=1 and α=2) a rigorous statistical footing.
• Voss recursion for 1/f, 1976. Richard Voss publishes a simple recursive algorithm for generating pink noise. Becomes the standard reference for software implementations through the 1980s and 1990s.
• Web Audio API, 2014. The W3C ratifies the Web Audio API as a Candidate Recommendation. Browsers can now synthesize audio in real time without plugins, opening the door to browser-based noise generators with mixing, timers, and effects.
• Brown noise viral on TikTok, 2022 onwards. The ADHD community on TikTok popularizes brown noise as a focus aid. Searches spike; brown noise becomes the breakout color of the early 2020s alongside the established white and pink. Premium soundscape apps (Endel, Brain.fm) layer adaptive variations on top.

About Background Noise

Background noise, also called ambient noise or white noise, is a consistent sound that masks disruptive noises and creates a soothing environment. It works by providing your brain with a neutral stimulus to focus on, helping you sleep better, concentrate during work, or relax during stressful situations. Different types of noise have different effects depending on personal preference.

Types of Noise

• White Noise · Consistent, hiss-like sound with equal power across all frequencies. Like static on a radio or a fan running. Great for sleep and blocking office noise.
• Pink Noise · Sounds like rain, wind, or rustling leaves. Lower frequencies are louder than higher frequencies, making it feel more natural and soothing.
• Brown Noise · Deep, rumbling sound like distant thunder or an air conditioner. Very low frequency, deepest of the three. Ideal for deep relaxation.

Frequently Asked Questions

Is background noise good for sleep?

Yes. Background noise masks disruptive sounds like traffic, neighbors, or pets, creating a consistent auditory environment. Many people find white or pink noise helps them fall asleep and stay asleep. Start with 50% volume and adjust to your comfort level.

Which noise type is best for focus?

This varies by person. Many find pink noise (rain-like) or brown noise (deep rumble) more natural and less distracting than white noise. White noise works well for some. Try each type at different volumes to see what helps your concentration most.

Can I mix multiple noise types?

Yes. You can turn on multiple noise types and adjust their individual volumes to create a custom mix. Use the Master Volume control to adjust the overall volume of your mix without changing individual ratios.

Real-world workflows

• Sleep onset and maintenance. Start with pink or brown noise at 40 to 50% volume. The 2017 Frontiers in Human Neuroscience study found pink noise during slow-wave sleep enhanced memory consolidation. The masking effect smooths over a noisy bedroom (traffic, snoring partner, HVAC). Volume should be just above the ambient noise floor, not loud enough to itself disturb sleep.
• Deep focus during work or study. Many users (especially with ADHD) find brown noise particularly helpful for sustained focus. Pink noise is the runner-up. Avoid white noise during long focus sessions; the high-frequency content can feel fatiguing after an hour. Try 25 to 50% volume on speakers or in-ear headphones.
• Soothing a crying baby. White noise mimics the in-utero environment (the steady wash of maternal blood flow + bowel sounds) and can calm newborns who are overstimulated. Pediatric guidance: place the speaker at least one meter from the crib at 50 dB SPL or below to avoid hearing damage from prolonged exposure.
• Tinnitus masking. For mild to moderate tinnitus, broadband noise at the threshold of audibility (just barely louder than your tinnitus) provides relief during the day or while falling asleep. For ongoing tinnitus management, consult an audiologist about hearing-profile-specific masking; this tool is a starting experiment, not a medical treatment.
• Speech privacy in shared spaces. Open-plan offices and shared apartments leak conversations through walls and doors. Pink or white noise at moderate volume in the room (not on headphones) masks the intelligibility of overheard speech, even when total loudness is similar. The masking effect is especially strong at the 1 to 4 kHz speech range.
• Meditation and breathwork backdrop. Brown or pink noise at very low volume provides a continuous auditory anchor for meditation without the cognitive distraction of music or guided narration. Useful for solo silent practice in a noisy environment where complete silence isn't available.

Common pitfalls and what they mean

• Too loud causes hearing damage. NIOSH recommends a maximum continuous exposure of 85 dB SPL over 8 hours. Sleeping with noise at 75 to 85 dB through the night is over this threshold. Aim for 50 to 60 dB at the ear: loud enough to mask, quiet enough to be safe long-term. Use a phone decibel meter app to check.
• Sleeping with earbuds is risky. Earbuds in the ear canal overnight increase risk of ear infections, wax compaction, and pressure-related discomfort. They also concentrate sound at higher SPL than the same volume from a speaker. Pillow speakers or under-pillow audio strips are safer alternatives. If you must use buds, choose flat over-ear sleep headphones.
• Psychological dependency. Long-term sleep masking can lead to conditioned dependence: difficulty sleeping without it when traveling or in new environments. Take occasional nights off to avoid building the association. For children, taper exposure as they age rather than continuing nightly noise indefinitely.
• Browser tab throttling. Some browsers throttle Web Audio in inactive (backgrounded) tabs to save battery. If you're using the tool overnight, keep the tab in the foreground or use a desktop noise app for guaranteed continuous playback. Mobile browsers especially can pause audio when the screen is locked.
• Looped MP3 vs live synthesis. Many "noise" sites and apps loop a short audio clip, which produces a faint repetitive artifact at the loop point that some sensitive listeners notice (perceived as a click, tick, or rhythm). This tool synthesizes fresh samples in real time, so there's no loop point and no repetition artifact.
• Noise as tinnitus treatment is not medical advice. Generic broadband noise can provide subjective tinnitus relief, but proper Tinnitus Retraining Therapy (TRT) involves audiologist-prescribed noise shaped to your hearing profile plus counseling. If tinnitus is significantly affecting your quality of life, see an audiologist rather than self-treating with browser tools.

Privacy: synthesized in your browser, nothing leaves

Most noise apps and websites stream pre-recorded audio loops from their servers, which means every play sends a request that the operator can log: when you started, how long you played, which color you chose, and the IP address you connected from. Over weeks of nightly use this accumulates into a detailed sleep-pattern profile. The audio itself is harmless, but the access pattern reveals when you sleep, when you wake up at night, and how often you use the service.

This tool synthesizes noise in real time inside your browser using the Web Audio API. No audio is streamed from any server. The only network traffic is the initial page load (HTML, CSS, JavaScript); after that, the browser produces the noise locally and nothing is sent or logged. You can verify by opening browser dev tools on the Network tab: no requests fire while the noise plays. Put the browser in airplane mode after page load and the tool continues to work indefinitely.

When another tool is the right pick

• Dedicated sleep noise machines. The Marpac Dohm (mechanical fan, 1962 design still sold) and modern smart machines (Hatch Rest, LectroFan) are purpose-built for nightly long sessions with low-distortion drivers and battery backup. If you sleep with noise every night, a hardware machine is more reliable than a browser tab.
• Adaptive AI-generated soundscapes. Premium apps like Endel and Brain.fm generate adaptive soundscapes that vary subtly with time of day, your heart rate, or your activity. Useful for users who want richer variety than stationary white/pink/brown noise. These are paid services with biometric integration; this free tool is a starting point if you're not sure whether noise helps you at all.
• Clinical tinnitus management. If tinnitus significantly affects your life, an audiologist can prescribe sound therapy specific to your hearing profile, often with hearing aids that deliver masking continuously throughout the day. This tool is fine for casual experimentation but isn't a substitute for professional evaluation.
• Nature soundscape preference. Some users prefer non-stationary nature sounds (forest, ocean, rain, fire) over stationary colored noise. YouTube and dedicated apps like Calm or Headspace specialize in these. The downside is the same as cloud noise apps: streaming-server access patterns can profile your sleep.

Other frequently asked questions

What's the safe volume for overnight noise?

NIOSH recommends a maximum of 85 dB SPL over 8 hours of exposure. For overnight noise, target 50 to 60 dB at the ear: loud enough to mask intermittent disruptions, quiet enough to be safe long-term. A useful rule of thumb: if you can comfortably hold a normal-volume conversation over the noise, you're around 50 to 60 dB. Use a phone SPL meter app to confirm.

Why does brown noise sound "deeper" than white?

Because brown noise has a 1/f² spectrum: low frequencies dominate by an enormous margin. White noise has flat energy per Hz, so high frequencies dominate perceptually (since the audible band is much wider above 1 kHz than below). Brown emphasizes the bass; white emphasizes the treble. Pink sits in the middle with equal energy per octave.

Why doesn't the noise loop or repeat?

Because this tool synthesizes fresh samples in real time using the Web Audio API. There's no pre-recorded audio file being looped, so there's no loop point and no repetition. Each second is freshly generated from the random algorithm for that color. Many cheaper apps and websites loop a short recording, which produces a faint repetitive artifact at the loop boundary that some sensitive listeners notice.

Will the noise drain my phone or laptop battery?

Yes, but modestly. Audio synthesis is light on CPU compared to video. Expect a battery drain of around 5 to 10% per hour on a phone with the screen on, less with the screen dimmed or off. For overnight use, plug the device in. Browser tab throttling on mobile may pause audio when the screen locks; if so, leave the screen on dim or use a dedicated noise machine.

Can I save my custom noise mix?

Not as a file. Your last-used mix settings (per-source volumes, master volume, timer) are stored in browser localStorage for the same browser on the same device, so they reload when you return to the page. Crossing devices or clearing site data resets the mix. For a portable preset, write down your favorite combination (e.g., "Brown 60%, Pink 30%, Master 70%") and reapply it.

Is there a desktop or mobile equivalent?

Yes, many. Free apps include White Noise Lite (iOS/Android), Sleepa, and Apple's built-in Background Sounds (Settings -> Accessibility -> Audio/Visual). Paid premium options include Endel (adaptive soundscapes), Brain.fm (focus-tuned), and the classic Marpac Dohm hardware. This browser tool is a zero-install starting point; if you find noise helps you significantly, a dedicated device or app is more convenient for nightly use.

Related Tools