Sound therapy for actual sleep onset.
Pink noise plus a delta-range binaural carrier (0.5–4 Hz). Pink-noise enhancement of slow-wave sleep is supported by Papalambros et al (Frontiers, 2017); delta-band entrainment is the more preliminary part of the recipe. Built for actual sleep onset, not for background ambience.
Who this is for
People who lie down at 11pm and are still awake at 1am. People whose brains won't slow down even when their bodies are exhausted. People who've tried melatonin, sleep stories, magnesium, weighted blankets — and still need help with the last 20 minutes of getting under. Not for people in acute insomnia crisis (please see a sleep clinician); for the chronic, manageable kind that needs a nightly tool.
What plays, and why
Layer 1 — Crystal singing bowls
Low-frequency Tibetan and crystal bowls layered slowly. The tradition matters here: the timbre is recognizable, the resonance carries, the harmonic stack is what your nervous system reads as 'safe room.' Free tier doesn't include this.
Layer 2 — Ambient drone pad
An F-minor pentatonic drone, generated fresh each session and woven under the bowls — never the same audio twice. The point isn't novelty; it's that the brain doesn't tune out a repeating loop.
Layer 3 — Pink noise + 2.5 Hz binaural beat
Two pure tones, one in each ear, detuned by 2.5 Hz. Your brain perceives a phantom 2.5 Hz beat — squarely in the delta band associated with deep sleep. Pink noise (1/f spectrum, warmer than white) sits at −24 dB underneath. This is the part with the strongest research backing.
Pacing — 30 or 60 minutes, fading to silence
Arrival → descent → ground → return. The curve peaks around minute 8, holds the deepest entrainment from minute 12 to 22, then fades to silence over the final 8 minutes. You wake to nothing, not to a cutoff.
The research
Wellness, not medical treatment. AmberRoom is not a substitute for clinical care.
vs. the alternatives
Common questions
Can sound therapy actually help me fall asleep?
There's research support, with caveats. Pink-noise enhancement of slow-wave sleep is supported by Papalambros et al (Frontiers in Human Neuroscience, 2017) — they found auditory pulses phase-locked to slow waves increased slow-wave activity and improved next-morning memory. The broader sleep medicine literature also supports continuous pink and brown noise for sleep-onset masking. Direct delta-band auditory entrainment for sleep is more preliminary. Effect sizes across the body of evidence are modest but reproducible. It's not a cure for insomnia, but it can help people fall asleep faster than silence.
How long should I listen for?
Most users fall asleep within 12–25 minutes. AmberRoom's sleep recipe runs 30 or 60 minutes (Pro). The audio fades to silence around minute 24, so you don't wake to a sudden cutoff. Free tier caps at 15 minutes, which is on the short side for most insomniacs — long enough to test the experience, not long enough for reliable sleep onset.
Do I need headphones?
For binaural beats: yes, the illusion only works when each ear hears a different tone in isolation. For pink noise alone: no — speakers are fine. If you sleep on your side, a single in-ear headphone or sleep headband works. AmberRoom detects speakers vs headphones and falls back to a mono recipe (isochronic tones replace binaural) if headphones aren't present.
Will the audio wake me up at the end?
No. The pacing curve resolves to silence around minute 24 of a 30-min session, then fades the noise floor over the final 6 minutes. You should be asleep before it stops.
Is this safe to use every night?
Yes. The audio is non-pharmacological with no known dependence or tolerance effects. Some users develop a sleep-association with the room — Pavlovian, not chemical — which is generally helpful (it cues sleep onset on its own over time) but means switching to silence cold-turkey can feel less effective for a few nights.
What if I have tinnitus?
Don't use the standard sleep recipe — pink noise can make some tinnitus presentations worse. Use the dedicated tinnitus mode (Pro) which uses notch-filtered audio matched to your tinnitus frequency. Calibration is non-negotiable for clinical effect.