Loop-based sleep applications fail for a reason that has little to do with audio engineering and everything to do with human perception. An eight-minute rain recording, no matter how professionally captured, contains repeating micro-structure: the same cluster of droplet impacts, the same phase relationship between low rumble and high hiss. By the third minute your auditory cortex tags the signal as "recording" rather than "environment," and sleep grows lighter precisely when you needed darkness and continuity. Estua takes a different approach: it synthesizes ambient audio in real time at forty-eight kilohertz on your phone, producing output with no fixed period and no seam to detect.

The signal chain begins with a master clock running at forty-eight kilohertz with a fixed block size of five hundred twelve samples, which equals approximately ten point seven milliseconds per callback. This audio thread is sacred territory — no heap allocation, no logging, no network calls — a discipline shared with Sonarish's capture path and documented in our audio thread post. White noise enters a pink-noise shaping stage using Paul Kellet's economical recursive filter or an equivalent biquad stack, producing the characteristic one-over-f frequency spectrum where energy per octave stays constant. An optional brown-noise layer adds low-frequency rumble for ocean-like scenes. Multiple slow low-frequency oscillators, running between zero point zero two and zero point fifteen hertz with uncorrelated phases, modulate filter cutoff and gain so that no periodicity emerges at thirty or sixty second intervals — the timescales where loop-based apps betray themselves.

Scenes, seeds, and the statistics of surf

Each scene — Rain, Ocean, Wind, and others — layers additional processing on the pink foundation. Rain adds band-passed noise bursts triggered with Poisson-distributed inter-arrival times, mimicking the random spacing of droplets on a roof. Ocean mixes amplitude modulation on a low band between eighty and four hundred hertz with mid-frequency hiss, creating the slow breathing quality of surf without any recorded waveform to repeat. A soft limiter clips at minus one decibel full scale, preventing digital full-scale spikes that would startle a sleeping listener at three AM.

Each night you choose a scene and a seed, or let the application pick randomly. The seed initializes pseudorandom number generators governing noise samples and LFO phase offsets. The same seed produces the same soundscape for that session — useful when you liked last night's rain — but we optionally stir the seed with the calendar date so that "Rain plus seed forty-two" is not identical every night for years. This balances reproducibility within a session against long-term variety across weeks.

Why crossfading two loops is not enough

A common industry workaround crossfades between two loop segments to hide the seam. Crossfading hides the audible click at the splice point, but it does not change the underlying statistics. Autocorrelation of looped rain still spikes at the loop period T, and the human auditory system detects periodicity at remarkably subtle levels — evolution favored spotting patterned threats in rustling grass. Real ocean surf approximates pink noise with occasional transient swash events whose timing follows no fixed schedule. Synthesis costs three to eight percent CPU on modern phones for our stack, but buys uncorrelated output for hours of continuous playback.

On iOS, Estua drives an AVAudioEngine manual source node callback with playback session category; on Android, AudioTrack runs at PCM float encoding with low-latency mode when the OEM supports it. Both platforms enforce the no-allocation rule in the hot path. Estua is not binaural beats, not AI-generated music, and not a streaming wrapper. There are no accounts and no analytics on listening habits. A timer fades volume over twenty to forty-five minutes; a separate alarm uses a gentle synthesized tone from a different scene so the sound that woke you is not the sound you slept to. For the psychoacoustic background on one-over-f noise and sleep, see sleep, waves, and why your brain likes surf; for why we keep all of this local, read why everything runs on your phone.