Violet Noise — What It Is, and a Free Player That Starts Quiet
Violet noise, also sold as purple noise, is white noise tilted upward by 6 dB per octave: the mirror image of brown. Almost all of its energy sits at the top of the hearing range, which is why it sounds thin and sharp, why it tires the ear faster than any other colour, and why it gets suggested for tinnitus. The player below starts low and stops on its own; the text says what the spectrum is, what the ear does with it, and what the published research contains — which, for violet specifically, is nothing.
What this is. Violet noise generated in your browser from a four-second loop, with white, pink and brown beside it so you can hear the tilt by switching. It starts at a quarter of full scale and stops on its own after ten or twenty minutes, which is the cap on every free tool on this site; there is no “off” here because violet is the one colour that should not run unattended. Nothing is streamed, stored or sent anywhere, and no sound plays until you press the button. No headphones needed — noise is not a binaural effect, and a speaker across the room is the gentler way to hear a bright colour. Blue noise, the +3 dB cousin, is not in the player because the shared component does not generate it; the section on the spectrum says where it would sit.
What violet noise is.
Violet noise is white noise whose power rises by 6 dB with every octave. Double the frequency and the energy per hertz quadruples. That single number is the whole definition, and every other property on this page follows from it.
The colours are named by analogy with light: white noise has the same energy at every frequency, the way white light has every wavelength; tilt it toward the low end and it is red or brown, toward the high end and it is blue or violet. The analogy stops there. Violet noise has nothing to do with violet light, and “purple noise” is simply another name for the same signal, not a different colour.
In practice violet noise is made by differentiating white noise: each output sample is the current random value minus the previous one. The player on this page does exactly that, in four lines of code, and scales the result down hard because the raw output is far louder in the treble than anything else the component makes. Differentiation is the mirror of what makes brown noise, which is white noise integrated — a running sum with a slight leak. Brown falls at 6 dB per octave, violet rises at 6, white is the flat line between them, and pink and blue are the half-way slopes at 3 dB either side.
| Colour | Slope per octave | Slope per decade | Made from white by | In the player |
|---|---|---|---|---|
| Brown (red) | −6 dB | −20 dB | Integrating (a leaky running sum) | Yes |
| Pink | −3 dB | −10 dB | A multi-stage low-pass filter | Yes |
| White | 0 dB | 0 dB | Nothing — it is the raw signal | Yes |
| Blue (azure) | +3 dB | +10 dB | A half-order high-pass filter | No |
| Violet (purple) | +6 dB | +20 dB | Differentiating (sample minus previous sample) | Yes — the default |
Two honest footnotes. The slopes describe power density, energy per hertz; since each octave holds twice as many hertz as the one below, flat white already doubles its total energy per octave, and violet multiplies it by eight. And a digital signal cannot rise forever: the slope holds up to half the sample rate, 22 or 24 kHz here, past the top of adult hearing. So most of what makes violet noise violet is energy you cannot hear at all, and what you can hear of it is concentrated in the region hearing loses first. The noise colours guide covers the whole family; white, pink, brown and green each have their own page and player.
How it sounds, and why the player scales it down.
Violet noise sounds like a thin, sharp hiss with no body — the top of a cymbal with the cymbal removed. Where brown noise is a waterfall and pink is steady rain, violet is closer to the sound of air escaping a valve, or the sibilance of a whispered “s” held indefinitely.
Switch between violet and white in the player and the difference is obvious within a second: white has a floor underneath it, a sense of a room; violet has none. Below about 1 kHz violet noise has almost nothing, so it does not fill a space. Everything is happening between roughly 4 and 16 kHz, and because hearing is most sensitive around 2–5 kHz and falls away above that, the colour reads as both piercing and oddly quiet at once.
The player multiplies the differentiated signal by 0.28 before it reaches the volume control, and the slider starts at 25 rather than the 55 used on the other noise pages. Both are deliberate: a raw differentiated signal can reach twice the amplitude of its input, all of it treble, so at an equal slider position violet would be the loudest colour in the set, in the part of the spectrum where loud does the most harm. If violet seems quiet next to brown at the same setting, that is the scaling working; raise it slowly, or not at all.
Age changes this colour more than any other. Sensitivity above 8 kHz declines steadily through adult life, so a listener of sixty hears materially less of it than a listener of twenty. If the player sounds like little more than a faint hiss, that is not a fault in the page.
Hearing safety at high frequencies.
Keep violet noise quiet, keep it short, and do not sleep in it. The reasons are specific to where this colour puts its energy, not generic noise-is-bad caution.
Why it fatigues. Listening fatigue is partly loudness and partly spectrum. The ear is most sensitive between about 2 and 5 kHz, where the ear canal resonates, and a signal that concentrates its energy from there upward is aiming at the most sensitive part of the system. White noise is already the colour people find hardest to tolerate for long; violet is white with the low end removed and the top end pushed up. No study has measured how long people tolerate violet noise; the practical rule is to treat the first flicker of discomfort as the stopping point, not as something to get used to.
Why the ceiling matters more here. Noise-induced hearing damage shows up first as a dip in the audiogram around 4 kHz — a systematic review of 84 studies (Meghji and Phillips, 2021, Ear and Hearing) found the notch reported at 3, 4 or 6 kHz, most often at 4, and warns that its relationship to exposure is less tidy than textbooks suggest. The location is what matters for this page: the damage region is the lower edge of where violet noise lives. In a controlled study of digital music players (Le Prell et al., 2012, Ear and Hearing), an hour or so of music at 93 to 102 dBA produced a temporary threshold shift in young adults with normal hearing, largest at 4 kHz and recovering within hours; the authors note that in animal models some exposures that recover on the audiogram have still cost nerve fibres. A Ghent study (Keppler et al., 2010, Archives of Otolaryngology – Head & Neck Surgery) measured a stock player at full gain through earbuds at around 103 dBA and found temporary threshold and emission shifts after a single hour. A 2024 systematic review of recreational noise in young people (Elmazoska et al., 2024, Journal of Speech, Language, and Hearing Research) found the measurable effects mostly short-term and in the extended high-frequency range, above what a standard hearing test covers — the violet region. None of those studies used violet noise; they establish that the band it occupies is the band where exposure leaves marks first.
In practice. Set your device to its normal level before you touch the slider, and let the slider stay low. Prefer a speaker to earbuds: a speaker a few metres away loses the treble earbuds deliver straight into the canal. Do not run it overnight; the timer stops at ten or twenty minutes and has no “off” on purpose. If you are considering a noise machine for a child, measurements of fourteen infant sleep machines at full volume (Hugh et al., 2014, Pediatrics) found every one above the 50 dBA recommended for hospital nurseries at 30 cm and three above 85 dBA; keep any such device far from the cot, well below maximum, and never on a bright colour. And if a session leaves ringing, fullness or dulled hearing, that is a threshold shift: stop, do not repeat it, and see someone if it has not cleared within a day.
Violet noise and tinnitus.
The idea is simple: tinnitus is usually a high pitch, violet noise is high-frequency energy, so violet should mask it better than the other colours at the same volume. The idea is plausible. It has never been tested.
The premise is sound. Tinnitus pitch really is high for most people. In a clinic series of 192 patients (Shekhawat, Searchfield and Stinear, 2014, European Archives of Oto-Rhino-Laryngology), about three in four matched their tinnitus to 8 kHz or above, and the pitch tracked the frequency where hearing loss reached roughly 50 dB. A Regensburg series of 286 patients (Schecklmann et al., 2012, PLoS One) found pitch following the frequency of maximum hearing loss rather than the audiogram edge. Pitch-matching itself is unreliable: when Moore and colleagues (Moore, Vinay and Sandhya, 2010, Hearing Research) trained eleven subjects to avoid octave errors, their matches dropped by one or two octaves and then correlated tightly with the audiogram edge — so “my tinnitus is at 9 kHz” is a softer number than it sounds, which is why the tinnitus frequency test here calls itself an experiment. Still: the pitch is high, masking works by spectral overlap, and a colour with energy up there will cover a high tone at a lower overall level than brown noise can. That part is physics, not a clinical claim.
What the literature contains for violet noise: nothing. On 2026-10-26 a PubMed search for “violet noise” returned three records — a paper on sensor noise in a motion-capture facility, a mouse hearing-loss model and its corrigendum. “Purple noise” returned three more: a cochlear gene-expression study in rats, a control-systems filter and a paper on noise fluctuations in brain signals. Combined with tinnitus, either term returns zero. Combined with sleep, zero. No trial, no case series, no crossover study has given violet noise to people with tinnitus or anyone else and measured what happened. Anyone selling violet noise as a tinnitus treatment is selling an untested inference.
What the literature contains for broadband sound in general is thin too. The Cochrane review of sound therapy for tinnitus (Sereda et al., 2018) found eight randomised trials and no usable data for any of its main comparisons; its authors concluded there is no evidence that sound generators or hearing aids beat a waiting list, placebo or information alone. One small crossover trial (Mahboubi et al., 2017, Annals of Otology, Rhinology & Laryngology) compared three months of broadband noise with three months of a customised sound and found broadband noise alone changed little beyond an anxiety score and the minimum masking level, in eighteen people. Masking makes tinnitus harder to hear while the masker plays; that is a different thing from changing it.
The one high-frequency approach with trials is the opposite of violet noise. Tailor-made notched music removes a band around the tinnitus frequency rather than adding energy there. The original Münster study (Okamoto et al., 2010, PNAS) reported reduced loudness after twelve months in eight patients against eight placebo controls; the same group’s CONSORT-standard trial of 100 participants (Stein et al., 2016, BMC Neurology) found no effect on its primary outcomes, with a loudness difference appearing only at follow-up; a 2023 trial of 120 people (Tong et al., 2023, Ear and Hearing) found notched music at least as good as a shortened tinnitus retraining protocol; and a 2024 meta-analysis of three studies (Alfonso et al., 2024, American Journal of Otolaryngology) found no significant difference between notched music and simply listening to music. That is the state of the art for doing something targeted at the tinnitus frequency, and it is equivocal. The notched sound therapy page walks through those trials and has a notch-filtered player; binaural beats and tinnitus covers what the rest of the site’s tools honestly can and cannot do.
If you want to try it anyway. Use the player as a masker, not a therapy: low, short, and stopped at the first sign of discomfort. Violet noise at a level that overpowers tinnitus is a loud high-frequency exposure aimed at the part of the ear most tinnitus already comes from. If a quiet setting does not cover the tone, do not chase it with the slider; try pink or white, or stop.
Uses people report — sleep and focus.
Everything in this section is what people say, not what has been measured. The two uses that come up most for violet noise are sleep and concentration, and in both cases the only evidence that exists is for white noise, with violet riding along on the family name.
Sleep. Some people report that a bright hiss helps them drop off, usually because it covers one high-pitched intrusion — a charger whine, a neighbour’s television, their own tinnitus. That is a masking story, consistent with how masking works, and not evidence that violet noise improves sleep. The systematic review that looked hardest at continuous broadband noise and sleep (Riedy et al., 2021, Sleep Medicine Reviews) found 38 studies in wide disagreement and rated the evidence that noise helps sleep as very low, adding that it may also harm sleep and hearing. A 2025 meta-analysis of twelve randomised trials (Ding et al., 2025, Sleep Medicine) was more favourable, reporting better Pittsburgh sleep-quality scores in adults and older adults, with substantial heterogeneity and a plea for better trials; every one of its trials used white noise, many on hospital wards. Nothing in either review involved violet. If you want noise for sleep at all, the lower colours are the safer bet — see brown noise — and the player here deliberately cannot run all night.
Focus. The “noise helps me concentrate” report has one well-known experimental basis: Söderlund and colleagues (Söderlund, Sikström and Smart, 2007, Journal of Child Psychology and Psychiatry) found that white noise during memory tasks improved performance in children with ADHD and worsened it in controls, and proposed stochastic resonance as the mechanism. That is a result about white noise at a moderate level in one population, argued over since, with no equivalent for violet. People who say violet helps them work are usually describing the same thing as the sleep reports: it covers one sharp distraction, a fan bearing or a colleague’s keyboard, at a lower level than a broad colour would need. For a focused twenty minutes at a low level that is a reasonable experiment; for a whole working day it is the wrong colour, and pink or green will be kinder.
Other reports — for infants, for studying, for “clearing the mind” — have no study behind them, and the first is a bad idea for the reasons in the hearing section: an infant cannot tell you the treble is uncomfortable.
Violet vs white, pink and brown.
The same random signal, four tilts. The table is the short version of everything above; the paragraph after it is the one most people actually need.
| Violet | White | Pink | Brown | |
|---|---|---|---|---|
| Spectrum | +6 dB per octave | Flat | −3 dB per octave | −6 dB per octave |
| Where the energy is | Almost all above 4 kHz | Even per hertz; sounds bright | Even per octave; sounds balanced | Mostly below 500 Hz |
| Sounds like | Air from a valve, held sibilance | Untuned radio, a harsh hiss | Steady rain | Waterfall, distant surf |
| Masks best | One high-pitched intrusion | Broadband distraction | General-purpose | Low-frequency: snoring, traffic, bass |
| Fatigue over an hour | Highest | High | Moderate | Lowest |
| Hearing-safety margin | Smallest: energy sits where damage shows first | Small | Moderate | Largest |
| Safe to run overnight? | No; this player will not | Only quietly | Quietly, speaker across the room | Quietly, speaker across the room |
| Human trials of this colour | None | Many, mixed | Some, mostly sleep-pulse studies | Almost none |
Violet vs white is the comparison people search for, and the answer is: white, almost always. White already has all the high-frequency energy violet has, plus a floor underneath it, so it masks everything violet masks and more, at a level the ear tolerates longer. Violet earns its place only when the thing you need to cover is itself high and narrow and you want to add as little total sound as possible. For everything else — sleep, a noisy office, a snoring partner — one of the three lower colours is the better tool, and the noise generator has all five with a full volume range and a longer timer. The main generator puts a white, pink or brown bed under a brainwave beat; it does not carry violet, because violet under a tone is unpleasant.
Web tool and mobile app are different things. The web players, this one included, generate white, pink, brown, green and violet noise in the browser and stop at twenty minutes. The mobile app has 44 presets and 34 ambient sounds, violet noise among them, and no session cap. Neither is a hearing device, a tinnitus treatment or a sleep aid with evidence behind it.
Things people reasonably ask.
What is violet noise?
White noise tilted upward by 6 dB per octave, so that energy per hertz quadruples with every doubling of frequency. It is made by differentiating white noise — each sample minus the one before — and is the exact mirror of brown noise, which is white noise integrated. Nearly all of its energy sits above 4 kHz, so it sounds like a thin, sharp hiss with no body.
Is purple noise the same as violet noise?
Yes. Purple noise is another name for the same +6 dB per octave signal; some apps and product listings use one word, some the other. Blue noise is different: it rises at 3 dB per octave, half the slope, and sits between white and violet.
Does violet noise help tinnitus?
Nobody knows, because no study has tested it. On 2026-10-26 PubMed held no trial, case series or crossover study of violet or purple noise in people with tinnitus, or in people at all. The reasoning is sound as far as it goes — tinnitus pitch is usually high, and masking works by spectral overlap — but that makes violet a plausible masker, not a treatment, and one that concentrates its energy where the ear is most vulnerable. The one high-frequency approach that has been trialled, notched music, removes energy at the tinnitus frequency rather than adding it, and its results are equivocal.
Violet noise vs white noise: which is better?
White, for almost every purpose. White noise already contains the high frequencies violet has, plus everything below them, so it masks more at a level the ear tolerates longer. Violet is only worth choosing when you need to cover one narrow high-pitched sound and want to add as little total sound as possible. For sleep, an office or a snoring partner, pink or brown beat both.
Is violet noise safe to listen to?
At a low level for a short time, yes, which is how this player is set up. The caution is that violet puts its energy from about 4 kHz upward, which is where noise-induced hearing damage appears first and where the measurable effects of recreational noise exposure show up in studies. Keep it quiet, prefer a speaker to earbuds, do not run it overnight, and stop if your ears ring or feel dull afterwards.
Why does violet noise sound quieter than brown at the same volume setting?
Two reasons. The player scales the differentiated signal down hard, because raw violet noise is far louder in the treble than any other colour and all of that loudness lands where the ear is least forgiving. And hearing itself is less sensitive above 8 kHz, increasingly so with age, so a listener hears less of violet than of any other colour at the same electrical level. If it seems faint, that is correct; raise it slowly, or not at all.
Can I use violet noise to sleep?
The player will not let you run it all night, and that is deliberate. The evidence that continuous noise of any colour improves sleep is rated very low by the most careful systematic review, and none of it concerns violet. If a high-pitched intrusion is keeping you awake, a quiet spell of violet or white while you drop off is a masking experiment; for a sound to sleep beside, brown or pink at a low level from a speaker across the room is the kinder choice.
Is anything I do on this page saved?
No. The noise is generated in your browser from a four-second loop; nothing is streamed, written to storage or sent anywhere. No sound plays until you press the button.