There is a reason people travel to the coast for rest. Ocean surf is loud in absolute terms yet feels calming to many listeners. Part of that paradox comes from statistics: natural surf approximates pink noise, sometimes called one-over-f noise because its power decreases predictably with frequency. Pink noise has equal energy per octave, which means low rumble carries weight without the harsh uniformity of white noise where every frequency band receives the same power. Estua synthesizes similar statistical textures in real time rather than replaying an eight-minute recording that your brain will detect by minute three.

We are an app studio, not a sleep clinic. We do not run EEG trials or publish medical claims. We read the psychoacoustics literature, dogfood nightly builds, and ship audio that avoids the perceptual traps loop-based apps fall into. This post explains why surf-like spectra matter for sleep onset and how Estua implements them without seams.

What pink noise is and why loops fail

White noise sounds like static on an untuned television — every frequency band contributes equally, which the ear experiences as bright and fatiguing over long exposure. Pink noise rolls off the highs so that each doubling of frequency carries less energy. Real rain on a roof and breaking waves share that gentle slope plus occasional transients: a louder swash, a gust against a window. Those transients are not evenly spaced; they arrive with rough randomness that keeps the soundscape alive.

Recorded loops fail because human hearing is exquisitely sensitive to periodicity. Autocorrelation — a measure of how much a signal resembles a delayed copy of itself — spikes at the loop period even when crossfades hide the audible click. Evolution favored detecting repeating patterns in rustling grass because repetition sometimes meant predators circling. Synthesis breaks that predictability by drifting phase relationships through uncorrelated low-frequency oscillators modulating filter cutoffs and gains.

How Estua shapes surf-like scenes

Typical ocean parameters combine amplitude modulation on a low band between eighty and four hundred hertz at rates around zero point zero five to zero point twelve hertz, a mid-frequency hiss band between one and four kilohertz filtered to pink and gently modulated, and optional high sparkle below eight kilohertz kept thirty decibels or more below full scale so it suggests air rather than needles. Rain scenes add band-passed noise bursts triggered with Poisson timing so drop clusters do not fall on a metronome. Each seed initializes independent random generators so two nights with the same scene name still diverge in detail unless you deliberately reuse a seed you liked.

All of this runs on the audio thread at forty-eight kilohertz in five-hundred-twelve-sample blocks, obeying the no-allocation rules documented in life on the audio thread. Implementation specifics of the synthesis chain — Paul Kellet pink filters, limiter placement, platform audio engines — live in sound that never repeats.

Domestic details that matter at two AM

Sleep audio is a shared environment problem. Estua's sleep timer begins fading volume twenty minutes before off, with the final three minutes ramping down roughly thirty decibels total so a partner is not startled when silence arrives. The wake alarm uses a separate synthesized scene rather than the same noise you slept to — a small choice that prevents the brain from associating the sleep sound with morning urgency.

If you want to explore noise colors interactively, Phyzix includes generators and meters that visualize white, pink, and brown spectra side by side. Sonarish is the wrong tool for sleep — it measures rather than soothes — but its FFT view is useful when verifying that Estua output truly lacks periodic peaks in autocorrelation.