Noise Generator

Pick a color to start. The noise loops seamlessly, and you can set a sleep timer or download it as a WAV file.

Equal energy per octave. Sounds balanced, like steady rain.

Volume
-10 dB

How to use the noise generator

  1. Click a color to start it playing. Click another color to crossfade smoothly to it.
  2. Set the volume with the slider. Your device’s volume still applies, so for sleep, turn the device down and use the slider for fine control.
  3. Choose stereo or mono. Stereo sounds wider. Mono focuses the sound in the center, which is useful for speaker testing.
  4. Optional: set a sleep timer: pick a preset or choose Custom and type any length, like 45 (minutes) or 1:30 (hours:minutes). The noise fades out gently over the last minute instead of cutting off abruptly.
Noise generator playing pink noise with the color tiles, play button, stereo option, live spectrum, volume, sleep timer and download labeled
  1. Noise colors. The small line shows how each one’s energy changes from low to high frequencies.
  2. Play or stop (or press Space).
  3. Stereo (different noise in each ear) or mono (identical).
  4. The live spectrum shows the shape of the current noise, low frequencies on the left.
  5. Sleep timer with a gentle fade-out, and WAV download.

Noise colors explained

A noise’s “color” describes how its energy is distributed across frequencies. The names borrow loosely from light: white light contains all colors equally, pink and brown lean toward the low (red) end, and blue and violet lean toward the high end.

ColorSpectrumSounds likeCommon uses
FlatTV static, a hissing radiatorMasking speech, audio and electronics testing
−3 dB/octaveSteady rain, wind in treesSpeaker calibration, sleep, focus
−6 dB/octaveWaterfall, ocean surf, airplane cabinSleep, masking low rumble, focus
+3 dB/octaveHissy spray, sharper than whiteAudio dithering, tinnitus therapy research
+6 dB/octaveVery thin, high-pitched hissHigh-frequency tinnitus masking, tweeter checks
Inverse equal-loudnessEven and “flat” to the earHearing research, perceptually balanced masking

White noise

White noise has the same energy at every frequency, so a spectrum analyzer shows a flat line. Because each octave contains twice as many frequencies as the one below it, most of white noise’s energy is in the treble, and that’s why it sounds bright and hissy. It’s the best signal for testing electronics and measuring frequency response, and it masks speech well in open offices. Many people find it too harsh for sleep.

Pink noise

Pink noise falls by 3 dB per octave, so every octave (20–40 Hz, 40–80 Hz, … 10–20 kHz) carries the same energy. That matches how we perceive pitch, so pink noise sounds balanced, like steady rainfall. It’s the standard signal for calibrating speakers and rooms, because its spectrum resembles music. Sleep studies have used it too: some found that pink noise timed to deep sleep improved memory consolidation.

Brown noise

Brown noise (also called red or Brownian noise, after the random motion it’s based on) drops 6 dB per octave. The bass dominates, so it sounds deep and soft, like a waterfall heard from a distance or the inside of an airplane. It’s popular for sleep and for focusing because it has almost no hiss, and it masks low-frequency sounds like traffic and footsteps better than white noise.

Blue noise

Blue noise rises 3 dB per octave, the mirror image of pink. It sounds sharper and thinner than white noise. Its main technical use is in dithering, both for audio and images: noise shaped toward high frequencies is less noticeable. It’s rarely used for relaxation.

Violet (purple) noise

Violet noise, sometimes called purple noise, rises 6 dB per octave, the mirror image of brown. It’s almost entirely high-frequency hiss. Some people with high-pitched tinnitus find it a useful masker because it covers the same frequency region as their ringing. It’s also a quick way to check whether a tweeter is working.

Grey noise

Grey noise is shaped to sound equally loud at all frequencies to human ears. Our hearing is most sensitive around 2–5 kHz and much less sensitive to bass, so grey noise boosts the lows and highs and dips in the middle. Here it’s based on the inverse of the standard A-weighting curve. It’s used in psychoacoustic research and in some tinnitus sound therapy.

Using pink noise to level-match speakers

If one speaker sounds louder than the other, or your center or surround speakers seem too quiet, pink noise and a sound level meter will fix it in minutes:

  1. Set the noise generator to pink and mono.
  2. Sit in your usual listening position and point a sound level meter (or a phone running the decibel meter) at the ceiling, at ear height.
  3. Play the noise through one speaker at a time. On a stereo setup, disconnect one speaker or use the balance control. On a surround receiver, use its built-in test-noise mode, which works the same way.
  4. Adjust each speaker’s level until all read the same, typically 75 dB with C-weighting for home theater.

Exact calibration files (pink noise at −20 dBFS, the level used by most calibration standards) are on the test tones page.

Noise for sleep and focus

  • Keep it quiet. Aim for the level of a soft shower: you should be able to talk over it easily. For babies, keep it below about 50 dB (the hospital nursery limit) and place the speaker well away from the crib, at least 2 meters (about 7 feet), as researchers recommend.
  • Prefer speakers to earbuds overnight, and if you use headphones, keep the volume well below half.
  • Use the timer. Many people only need noise to fall asleep. The fade-out avoids waking you with a sudden silence.
  • Experiment with color. Brown masks rumble, pink masks voices and household sounds, and white masks high-pitched sounds best.

Tinnitus: sound therapy works best with a level just below your tinnitus, so it blends in rather than covering it completely. Try the color whose pitch is closest to your ringing (violet or blue for high-pitched tinnitus). If you have new or one-sided tinnitus, see a doctor or audiologist.

Downloading noise files

The Download WAV button saves the current color and channel mode as a 48 kHz, 16-bit WAV file of 30 seconds, 60 seconds or 5 minutes. The files are built to loop perfectly, so set any music player to repeat for endless noise without a gap. They’re free to use for any purpose, including commercial projects.

Frequently asked questions

What’s the difference between white, pink and brown noise?

They differ in how energy is spread across frequencies. White noise has equal energy at every frequency, so it sounds bright and hissy. Pink noise loses 3 dB per octave, giving equal energy in every octave, and sounds balanced, like rain. Brown noise loses 6 dB per octave and sounds deep, like a waterfall or distant thunder. Press each color above and watch the spectrum to see (and hear) the difference.

Which noise color is best for sleep?

There’s no single best one, so try them. Many people find pink or brown noise more pleasant than white because they have less harsh treble. Brown noise is popular for masking low rumbles like traffic and neighbors. Pink noise is the most studied in sleep research. Whatever you choose, keep it quiet (around 50 dB or less at your pillow) and use the sleep timer so it fades out.

Does the noise loop seamlessly?

Yes. The noise is synthesized in the frequency domain, so each loop is mathematically periodic: the end joins the start with no click, gap or change in sound. The downloaded WAV files are built the same way, so you can loop them in any player or editor without a seam.

Is the noise generated live or streamed?

It’s generated in your browser. Nothing is streamed, so it works offline once the page has loaded and uses almost no data. A loop is about 11 seconds of independent random noise, which is far too long and too random for your ear to detect the repetition.

What’s the difference between stereo and mono noise?
Stereo sends different random noise to each channel. It sounds wide and envelops you, and most people prefer it for sleep and focus. Mono sends identical noise to both channels. On correctly wired speakers it forms a tight “phantom center” between them, which makes it useful for checking speaker balance and polarity.
Can I use pink noise to calibrate my speakers?

Yes. Pink noise is the standard signal for level-matching speakers because it has equal energy per octave, much like music. Play mono pink noise through one speaker at a time, measure each at your listening position with a sound level meter (or our decibel meter), and adjust until they match. For files at a calibrated level, see our test tones.

Can white noise damage hearing?

Only if it’s too loud for too long. Hearing-safety guidance puts the risk threshold at around 85 dB for 8 hours, and many people play sleep noise much quieter than that. For babies, a 2014 study in Pediatrics found some infant sleep machines can exceed safe levels at full volume, and recommended placing them as far from the crib as possible, at low volume. Hospital nursery guidelines use 50 dB as the limit. If you need to raise your voice to talk over it, it’s too loud.

Tone GeneratorPlay any frequency from 1 Hz to 22 kHz in four waveforms.Test TonesDownload WAV test tones, sweeps and noise for calibration.Decibel MeterMeasure noise levels in dB with your microphone.Speaker TestChannels, full-range sweep, rattle check and wiring polarity.