Brown noise vs white noise: which one to actually work to
It is 15:20 and two people four desks away are discussing a wedding.
You are not listening. You could, however, produce the seating plan, both sets of parents, and a strong opinion about the venue.
Nobody is being loud. Nobody is doing anything wrong. You simply cannot decide not to hear speech — the decision is made somewhere you have no access to.
The only thing you can change is what else is in the room.
The rule: you are not making it quieter, you are making interruptions harder to find
A room is never silent. It has a floor — ventilation, traffic, a fridge, the building.
What breaks your concentration is not the floor. It is an event poking up above it: a door, a laugh, your name, one clear sentence about a venue.
Masking raises the floor until events stop poking through. You are trading a jagged sound picture for a flat one, and paying for it with more total sound, not less.
Which is why cranking the volume until you hear nothing is the wrong instinct. You are not winning a war of loudness. You are closing the gap between the floor and the peaks.
The three colours, without the maths
All three are random sound. The only difference is spectral tilt — how the energy sits between low and high.
| Colour | Tilt | Sounds like | Best against |
|---|---|---|---|
| White | Flat: as much energy up high as down low | Radio static, a hairdryer, TV snow | Higher-pitched intrusions; harsh to sit under for hours |
| Pink | Energy falls as pitch rises, about 3 dB per octave | Steady rain on a roof, wind in trees | Broad coverage; the least annoying of the three for most people |
| Brown | Falls twice as fast, about 6 dB per octave | A waterfall, an aeroplane cabin, distant surf | Low rumble: traffic, footsteps, bass through a wall |
In plain words: white keeps all its treble and sounds bright and hissy. Brown has thrown most of the treble away and sounds like weather. Pink sits between them, which is why it is the one people call neither annoying nor boring.
If someone tells you brown noise is scientifically superior, they are describing a preference in a lab coat. Brown is better at masking low frequencies and gentler to sit under for three hours. That is a property, not a benefit.
Being honest about the evidence
Most articles on this topic keep this bit for the footnotes. It belongs in the middle.
Two things are well supported. Intelligible speech disrupts verbal work, and masking reduces how detectable an intrusion is. The irrelevant speech effect has been studied since Colle and Welsh (1976) and Salamé and Baddeley (1982): unattended speech impairs short-term serial recall even when you are told to ignore it, even in a language you do not speak. Those are lab tasks — recalling strings of digits — so do not over-map them onto writing a report. The direction has held for four decades all the same.
Lyrics are the most reliable practical finding here. Perham and Currie (2014, Applied Cognitive Psychology) found music with lyrics impaired reading comprehension whether or not participants liked the music. Liking it did not protect them. Small student sample, one lab reading task, and still the most useful sentence in the field: if the task is made of words, do not add words.
What is not well supported is the part everyone quotes. There is no solid basis for the claim that any of these noises makes you work better in general. Results are mixed, samples are small, and the effects that do appear are often modest and do not always replicate.
The most-cited exception is Söderlund, Sikström and Smart (2007, Journal of Child Psychology and Psychiatry), who found white noise improved performance in inattentive schoolchildren and worsened it in attentive ones. Striking, and some later work points the same way. It is also one population and one kind of task, and it does not license the claim that brown noise treats ADHD.
So: sound is a tool for controlling your acoustic environment, not a drug for your attention. Everything below assumes that.
Which sound for which room
People talking near you — pink or brown, set so speech stops being intelligible rather than stops existing. You should still be able to tell someone is speaking; you should not be able to follow it. That target is lower than you think.
Traffic, thin walls, a neighbour with a subwoofer — brown, without hesitation. This is the case it was built for.
A room that is too quiet — rain. Not for masking, but for the same reason some people cannot sleep in a silent hotel. Rain sits roughly in pink territory with a slow irregularity that most people find easier to forget about than a flat hiss.
Café noise deserves its own paragraph, because it is speech, and speech is what the research keeps flagging. It works well for many people on non-verbal work — spreadsheets, sketching, sorting, admin — and quietly costs them on reading and writing. If you love it and you are drafting an essay, that is a real trade rather than a superstition. Run it against rain on a writing day and see which draft exists at the end.
Lo-fi is instrumental, narrow in dynamic range and repetitive by design, which is a sensible specification for background. The trap is not the music, it is the four minutes of choosing, skipping and adjusting — the same reach that eats your afternoon wearing a different hat.
Should I listen to music with words while studying?
For reading, writing, translation, or code with a lot of naming in it: no. That is the one place the evidence is consistent.
For repetitive non-verbal work: fine, and often better than silence, because it keeps a tedious hour tolerable and tolerable is what gets it finished. The cost lands on words specifically. Split your tasks along that line and you have most of the benefit.
Volume, headphones, and not wrecking your ears
The test is your keyboard. You should be able to hear yourself typing. If the keys have vanished, the sound has stopped masking and started competing.
Watch the drift. If you are nudging it up hour by hour, that is adaptation, not need. Turn it down instead.
Dose is what matters. Safe-listening guidance is built on level multiplied by hours rather than a single scary number, and background sound is the unusual case where you run it for four hours straight. That is exactly where a couple of notches quieter pays for itself.
Closed-back over-ear headphones block sound passively, which lets you run everything lower. Quieter masking through better isolation beats louder masking through earbuds.
Active noise cancelling is excellent at steady low-frequency sound — engine hum, ventilation, a train — and poor at speech, the thing you most want gone. ANC plus a modest amount of pink noise handles an open-plan office. ANC alone does not.
Take them off in the break. Your ears want the five minutes as much as your legs do.
Three days, one variable
Everyone argues about this in the abstract. It takes three days to settle for yourself.
| Day | Sound | When | Note at the end |
|---|---|---|---|
| Mon | Brown, low | The hardest 50 minutes of the day | Two numbers, 1–5 |
| Tue | Rain | Same slot, same kind of work | Two numbers |
| Wed | Silence, or earplugs | Same slot, same kind of work | Two numbers |
The two numbers: how easy it was to start, and how often you surfaced. Write them within a minute of finishing. By bedtime you are remembering your mood, not the session.
Same time of day, same kind of work, same volume. One variable, or the test tells you nothing.
Most people find one obvious loser and two roughly equal winners. Take whichever winner you reach for without thinking. The exercise is not there to find the optimum — it is there to stop you re-deciding this every morning, which is four minutes a day spent on a question you have already answered.
When sound is the wrong lever
When it is the phone, not the room. If your interruptions arrive as notifications, no amount of brown noise will touch them.
When more input is the problem. For people with misophonia or broader sensory sensitivity, added sound is another stressor rather than a shield. Earplugs and a quieter room, not a better playlist.
When it is your hearing. Masking helps some people with tinnitus, and there are audiologists who specialise in exactly this. Clinic conversation, not a blog one.
When you cannot work without it. If the sound has become load-bearing, notice it, because it will fail you on a train and in an exam hall. Work one session a week in silence to keep the option open.
And if you reach for sound because you cannot start at all, sound is not the missing piece. That is a problem of session length and a first minute that costs too much, which is where the timer earns its keep — or, if your work needs a long run-up before it goes anywhere, an argument for longer blocks.
Today
Put on pink or brown, low enough to hear your own keyboard, and work fifty minutes.
Then take the headphones off and notice how loud the room actually was.
Fomodo carries seven background sounds — rain, a café and five others — plus a separate switch for the ticking that you set per session rather than once and forever. Seven is the honest limit here: seven is a shelf, not a library, and if the exact sound that works for you is not among them, it is not among them.
The per-session ticking switch exists because the same person wants it on for a pile of admin at 11:00 and very much off at 22:00 with a headache, and asking them to go into Settings twice a day was never going to survive contact with a real week.
Pick one sound. Use it all week. The choosing is what has been costing you.