Most people are familiar with white noise, that static sound of an air conditioner that lulls us to sleep by drowning out any background noise.

Except technically, the whirl of a fan or hum of the AC isn’t white noise at all. Many of the sounds we associate with white noise are actually pink noise, or brown, or green, or blue. In audio engineering, there’s a whole rainbow of noise colors, each with its own unique properties, that are used to produce music, help relaxation, and describe natural rhythms like the human heartbeat. If you know what to look for, you can start to notice the colors of the noise that make up the soundscape around us.

If you decompose a sound wave, you can break it down into two fundamental characteristics: frequency, which is how fast the waveform is vibrating per second (one hertz is one vibration per second), and amplitude (sometimes measured as “power”), or the size of the waves. The noise types are named for a loose analogy to the colors of light: White noise, for example, contains all the audible frequencies, just like white light contains all the frequencies in the visible range.

In musical sound waves, the frequencies are spaced at intervals that we find pleasing to the ear, creating a harmonic structure that gives a sound its unique tone quality, or timbre. (This is what makes the same note sound different on a flute than it does on a violin.) The noises we hear every day—boots stomping across the floor, a car honking outside, the jingling of keys—are made up of sporadic waveforms, a random distribution of frequency and amplitude.

And then, in a separate category, there are the colored noises. Unlike the inconsistent bang of a drum or shouting voice, these sounds are a continuous signal, but they aren’t exactly pleasant. The word “noise” actually comes from a Latin word for nausea; in audio engineering, the term describes any unwanted information that interferes with the desired signal, like static on the radio.

Pure white noise sounds like that hissy “shhh” that happens when the TV or radio is tuned to an unused frequency. It’s a mixture of all the frequencies humans can hear (about 20 Hz to 20 kHz), fired off randomly with equal power at each—like 20,000 different tones all playing at the same time, mixed together in a constantly changing, unpredictable sonic stew.

The other colors are similar to white noise, but with more energy concentrated at either the high or low end of the sound spectrum, which subtly changes the nature of the signal. Pink noise, for example, is like white noise with the bass cranked up. It’s a “shhh” sound with a low rumble mixed in, like the soft roar of a rainstorm.

Pink noise sounds less harsh than white noise because humans don’t hear linearly. We hear in octaves, or the doubling of a frequency band, which means we perceive as much sonic space between 30-60 Hz as between 10,000-20,000 Hz. We’re also more sensitive to higher frequencies (one to four kHz, which is about the frequency of a crying baby, sounds the loudest), so white noise, which has the same intensity at even the highest tones, can sound way too bright to our ears. The energy in pink noise drops off by half as the frequency doubles, so every octave has equal power, which sounds more balanced.

Spectral analysis of white and pink noise, with frequency on the horizontal axis and power on the vertical axis (The Physics Hypertextbook)

In recent years, pink noise has become the darling of the noise spectrum, dethroning white as the in-vogue option on sound generators for sleep or concentration. In 2013, a study published in the journal Neuron found that pink noise helped participants achieve deeper sleep; in recent years, various health blogs have touted it as the key to a better night’s rest.

The inverse pattern of pink noise, also called 1/f noise, can also be applied to plenty of systems outside of sound. If you take the rise and fall of the tide, for example, and break it down into waveforms plotted on a graph, it will follow 1/f, which happens to be the exact midpoint between pure randomness and correlated movement. It turns out much of our world operates in this sweet spot between chaos and control: The pink noise pattern has been found in most genres of music, the shot lengths in Hollywood films, the structure of DNA, the rise and fall of the tide, the flow of traffic, and variations in the stock market. The world is basically awash in pink.

Brown or “Brownian” noise, a deeper version of pink, is not actually named after the color; the name comes from the fact that  the signal mimics the “random walk” pattern produced by Brownian motion, or the random movement of particles in liquid. The sound (not to be confused with the mythical “brown note” noise) is a deeper, bassy rumble, kind of like ocean waves or heavy winds.

Blue noise, which has more energy concentrated at the high end of the sound spectrum, is just the opposite: It sounds like the hiss of a water spray, a high-pitched screeching noise, with no bass tones at all. It’s essentially the inverse of pink noise: With blue noise, frequency and power increase at the same rate, so each octave has as much energy as the two octaves below it combined.

Because the high-pitched frequencies of blue noise are harder for the listener to discern, sound engineers use it for a process called audio dithering, which is intentionally adding noise to a signal to minimize any distortions that appear during the production process. Adding noise randomizes the errors, helping to smooth out the rough edges.

Gray noise sounds the same at every frequency; like pink, it’s calibrated to sound more balanced to the human ear. There is no single example of gray noise, because every human has a slightly different hearing curve. In medicine, it’s used to treat hyperacusis, an increased sensitivity to normal sounds, or tinnitus, a ringing in the ear.

White, pink, and blue noise are the only colors to have official definitions in the federal telecommunications standard, while brown and gray have accepted meanings in certain industries. Meanwhile, the other colors of the noise rainbow have only been informally defined. Green noise, for example, has been described as a signal with more energy concentrated in the middle of the sound spectrum; with in a limited frequency range around 500 Hz, it supposedly simulates the ambient noise of nature. Orange noise is sometimes described as a clashing, cacophonous noise like an out-of-tune ensemble. Violet noise is simply a more intense version of blue, with even more energy concentrated in the highest audible frequencies.

And there’s one more color of noise given an official meaning: black. It’s a spectral density of roughly zero power at every frequency. If white is all frequencies at once, black is the color of silence.