Nighthum

How to use a sleep timer and layered sound without waking yourself

Published · about 8 minutes

A sleep timer has one job and one classic failure mode: it ends. If it ends abruptly, in light sleep, in a room that then goes silent, it can be the thing that wakes you. This is how to set the duration, the fade and the mix so that the sound helps you get to sleep and then leaves without announcing itself.

This guide assumes you have already set the room up — sound source across the room, hard surfaces softened, nothing in the room that clicks. If not, start with the bedroom setup guide and come back.

Step one

Set the level first, and never let the timer change it

Everything downstream depends on the base level, so set it before you touch anything else. Lie down, play the exact sound you will use, and bring it up until the room stops feeling empty — then stop. You should still be able to find your tinnitus if you deliberately listen for it.

That is the ceiling for the whole night. Nothing you add in later steps should push past it: layering sounds adds level, so when you add a second layer you turn the first one down rather than adding on top.

Step two

Cover falling asleep, not the whole night

The sound is there for the part of the night when you are awake in a quiet room with nothing to do. That is the window it needs to cover, plus a margin. It does not need to cover eight hours, and running it all night has real costs: sound at your ears for the entire night, a battery flat by morning, and — if you use earbuds — hours of pressure in an ear canal.

A practical way to pick:

  • Estimate honestly how long you take to fall asleep on an average night, not a good one. Most people underestimate; if you have a diary, use it.
  • Add a real margin on top. A timer that ends while you are still awake is worse than one that runs a bit long, because you then lie there in the silence it left.
  • Somewhere between thirty and ninety minutes is where most people land. Short sessions of ten to fifteen minutes are a different tool — good for settling a spike during the day, not for bedtime.
  • An all-night setting is legitimate if you consistently wake and cannot get back down, but treat it as a deliberate choice rather than a default, keep the level lower than you would for a short timer, and use a speaker rather than earbuds.
Step three

The fade matters more than the duration

This is the part people skip and the part that decides whether the timer wakes you.

You cycle through lighter and deeper sleep all night, and in the lighter stages you are quite responsive to change. A sound that has been steady for forty minutes and then stops is a change, and it leaves a silence in which the ringing is suddenly the loudest thing in the room. So make the ending gradual. A fade of several minutes at the end of the session lets the level come down slowly enough that there is no moment where anything happens. Longer sessions can carry longer fades — a ten-minute fade at the end of an all-night session is not excessive. If your sound source cannot fade, that is a real limitation, and it is worth choosing a different one rather than working around it.

What to avoid

Playlists that end. Tracks that do not loop seamlessly, so there is a click or a gap every few minutes. Hardware timers that cut the power. Any of these produce the same abrupt transition, several times a night in the case of a bad loop.

Step four

Layer by role, and keep it to two

Layering — running rain under brown noise, say, or surf under a notched masking sound — is genuinely useful, and it is also the easiest way to build something too interesting to sleep through. Assign each layer a job:

  • The bed. One continuous, textureless sound that fills the room and never does anything: brown or pink noise, a fan, steady rain, or a notched masking sound shaped around your own pitch. This is most of your level and it is the layer you should stop noticing within a minute.
  • The texture, optional. One quieter layer that gives the bed some variation so it does not feel synthetic — light rain over noise, distant surf, wind. Keep it clearly quieter than the bed. Its job is to make the mix pleasant, not to be listened to.
  • That is all. A third layer almost always turns into something you are actively listening to, and it pushes the total level up without you noticing, because each individual layer still sounds quiet.

If your bed layer is a notched sound, it is worth being sure the notch is in the right place first — an octave error puts the hole in the wrong region and you lose the point of it. How to find the pitch of your tinnitus covers that, and what the difference actually is covers whether notching is the right choice for you at all.

Step five

Cut anything with a pattern

The specification for a bedtime mix is predictable. Not quiet, not pleasant — predictable. Attention is drawn by things that are not predicted, so any layer that produces distinct events is working against you, however nice it sounds while you are awake.

Things to drop:

  • Thunder, and most storm recordings. A roll of thunder is exactly the kind of event that pulls you back up.
  • Birdsong, crickets, anything with discrete calls.
  • Waves with a strong slow rhythm, if you find yourself waiting for the next one. Some people are fine with surf; some track it all night. Test it awake and notice which you are.
  • Anything with speech or melody. Music and podcasts carry information and information is what attention is for. They can be fine for the wind-down before bed; they are poor as a sound floor.
  • Recordings with an audible loop seam. Listen to one full cycle awake, and if you can hear the join, use something else.
Step six

Decide now what happens at three in the morning

You will sometimes wake after the timer has ended, into a silent room, with the ringing right there. What you do in that moment matters, and it is much better decided in advance than improvised at three a.m.

Do not turn it up because the ringing seems louder than at bedtime — it seems louder because you woke into silence, and the level you set is still right. A workable plan: one action, in the dark, that starts a short session at the same low level and nothing else. No unlocking into a bright screen, no scrolling, no deciding which sound to use. Set up whatever your source needs — a lock screen control, a widget, a physical button — so that it is genuinely one action.

Step seven

Change one variable a week

You have three dials: duration, fade length and mix. Change one at a time, keep it for several nights, and judge it against what you have written down rather than against how last night felt.

Read the results this way. Waking as the sound ends means lengthen the fade first, then the duration. Lying awake past the end means lengthen the duration. Waking mid-session means look at the mix for something eventful. Feeling like the sound is not doing anything means check the level was set lying down, late, in a quiet room — that is usually where it went wrong.

All of this reads much better with a couple of weeks of daily numbers behind it: how to keep a tinnitus diary that actually shows change.

Please read

This is general information, not medical advice. Sound approaches do not diagnose, treat or cure tinnitus, and Nighthum is a sound and diary tool rather than a medical device. Keep every sound comfortably quiet, especially with earbuds and especially overnight; louder is not more effective. If your tinnitus is sudden, in one ear only, pulsing with your heartbeat, or comes with hearing loss, dizziness or pain, it is worth getting checked by an audiologist or a doctor.