SELF-EXPERIMENT AUDIO ENGINE

Somatic Audio Lab

Subtle, audible, gradually adapting sound for opt-in nervous-system self-experiments.

Idle

Use only on yourself or consenting listeners. Keep speaker volume comfortable. This prototype does not diagnose or treat any condition, and it does not use ultrasonic or intentionally imperceptible stimulation.

Phone-lock playback Checking this browser…

Start the session before locking your phone. When the browser supports it, the lab requests a dedicated playback audio session and lock-screen media controls.

Calls, another app taking over audio, browser power-saving behavior, or older mobile browsers can still interrupt web audio. If the browser suspends audio, return to the page and resume the session.

1. Route audio

Pair Bluetooth/AirPlay at the operating-system level first. On iPhone/iPad, the lab uses the system media route because Safari does not expose a direct in-page speaker picker.

System default output

2. Set experiment

Starting arousal 65/100 Your best estimate of how activated you feel now. 0 = very settled; 100 = extremely keyed up or activated. Target arousal 30/100 Where you want the engine to gently aim by the end. It is a direction, not a promise or a clinical measurement. Session length 20 min How long the engine has to make the change. Longer sessions spread the same shift out more gradually. Subtlety 80% How noticeable the pulsing and movement are. Higher = gentler and less obvious; lower = more pronounced.

3. Run adaptive session

The engine begins near your stated arousal, then slowly changes pulse speed, spectral brightness, harmonic density, and movement toward the target.

Model state -- /100

Slow pulse -- cycles/min

Brightness -- Hz

Time 00:00 remaining --

What do these live numbers mean?

Model stateThe engine's current internal point between your starting and target ratings. It is not a sensor reading.

Slow pulseHow many gentle loudness cycles happen each minute. It slows as the model moves toward a lower target.

BrightnessThe audio filter cutoff. Higher numbers sound brighter; lower numbers sound softer and darker.

TimeHow far you are through the selected session.

Engine level 12% The lab's own output level. Start low. Your phone or speaker volume still controls the final loudness.

Live self-rating

Every few minutes, rate your actual arousal. The engine gently re-anchors its trajectory instead of assuming the sound is working.

Right now 50/100 This is your real felt state, not the model's guess. Logging it blends your rating into the remaining path and reschedules the audio.

No check-ins logged.

Optional heart-rate sensor

On browsers with Web Bluetooth, you can connect a standard Bluetooth Heart Rate Service device. HR is logged as context; it is not treated as a direct measure of emotion.

-- bpm

Not connected

Experiment log

Stored only in this browser unless you export it.

Time	Type	Model	Self-rating	HR	Note