Why rain alone leaves gaps
Rain is mostly middle. Measured on the loop this page plays, 56% of its power sits between 250 Hz and 4 kHz and only 28% below 250 Hz. That is the right shape for covering voices, dishes and a television through a wall. It is the wrong shape for a lorry, a boiler, or the flat downstairs, all of which live under 250 Hz where rain has little to give.
So the preset puts a quarter-strength brown noise under the rain rather than turning the rain up. Brown noise carries 66% of its power below 250 Hz. Twenty-five is enough to fill the floor and low enough that you hear rain, not a hum.
The mistake: turning it up until it works
Rain sounds loud before it sounds effective. Because its energy is bunched in the range your ear is most sensitive to, raising rain past roughly 65% makes the room feel louder without covering any more of what wakes you. People do it anyway, then wake at 3am with rain in their head.
If the noise still gets through at 55%, raise brown noise five points and leave the rain alone. Judge it by whether the disturbance is still distinct, not by how present the rain feels.
Why the timer stops at 45 minutes
Forty-five minutes covers falling asleep, which is the part most people need help with, and stops before the sound becomes something you need in order to stay asleep. The chip is there if your problem runs all night — the all-night option keeps it going to morning.
Set the timer to what your noise actually does. A neighbour who comes home at midnight is a 45-minute problem. A road is an all-night one.
What this preset actually plays
Measured on 2026-09-04 from the exact files this page loads, decoded to 48 kHz mono, with loudness from the EBU R128 meter and the spectrum from an averaged 8192-point FFT.
| Layer | Preset level | Loop | Loudness | RMS | Centre | Low / mid / high |
|---|---|---|---|---|---|---|
| Rain | 55% | 32.02 s | -15.6 LUFS | -19.3 dBFS | 1940 Hz | 28% / 56% / 16% |
| Brown Noise | 25% | generated, no loop | -15.5 LUFS | -13.9 dBFS | 537 Hz | 65.9% / 31.7% / 2.3% |
- Rain loop join: the step between the loop’s last sample and its first measures 0.23× that file’s own 99.9th-percentile sample-to-sample move — inside the material, so a butt-join would not tick. The player overlaps each pass by 192 ms regardless.
- What this preset cannot do: These numbers describe the source, not your room. A phone speaker on a bedside table will not reproduce the brown-noise layer at all, and the preset cannot mask a sound louder than itself.
Play rain sounds for sleeping
- Open this page in any browser.
- The preset is already applied: rain at 55%, brown noise at 25%, timer at 45 minutes.
- Press play, then lower rain in five-point steps until it stops being the thing you notice.
- Leave the phone across the room and let the timer stop it.
Frequently asked questions
How loud should rain sounds be for sleeping?
Low enough that you can still hear a smoke alarm and a voice at the door. This preset starts rain at 55% and brown noise at 25%, which covers ordinary household and street noise on a small speaker. If the sound is the last thing you notice before sleep, it is too loud.
Should rain play all night or stop?
Match the timer to the noise. This page arms 45 minutes, which covers falling asleep. Switch to the all-night chip if traffic or a road runs until morning, since sound stopping halfway can wake a light sleeper more than the noise it was covering.
Why is there brown noise under the rain?
Rain has only 28% of its power below 250 Hz, so it barely touches low rumble. Brown noise has 66% down there. A quarter-strength layer fills the floor without changing what the mix sounds like.
Does the rain loop repeat noticeably?
The loop is 32.02 seconds. The player overlaps each pass with the next by 192 milliseconds, chosen per file so the level holds steady, so there is no click or gap at the join.
Can I change the balance?
Yes. Both faders are live on this page, and the timer chips change the stopping point. Opening the mixer carries the same preset across and adds the rest of the library.