Adaptive volume

<p><strong>Why this helps:</strong> On-device microphone snore and breathing detection is established in shipping sleep apps; privacy-first implementations run a Core ML / on-device classifier so raw audio never leaves the phone (Snollo product documentation, 2025, describing on-device Core ML classification of snoring and breathing). Smartphone microphones can reliably detect snoring during ordinary sleep (Shin &amp; Cho, *Sensors*/PMC, 2014). Using "you have gone quiet or begun snoring" as a cue to fade audio is a reasonable extension of that detection; the specific benefit of auto-fading on sleep depth is plausible but not separately validated, so we describe it as a convenience, not a clinical effect.</p><h4>Explainer (intro card)</h4><p><strong>Header:</strong> It listens for you to drift off, then bows out.</p><p><strong>Body:</strong><br>You do not always know the moment you fall asleep, so you cannot reliably reach over and turn the sound down at the right time. Adaptive volume does it for you. Using your phone's microphone, it listens for the signs that you have settled, longer, slower breathing, the quiet of a still room, or the soft sound of snoring, and when it is fairly sure you are asleep, it eases the volume down and then off. No abrupt stop. Just a slow fade, the way a good night-light dims itself.</p><p>If you stir and the room gets active again, it can gently bring a low level of sound back. You stay asleep. The app does the reaching-over.</p><p><strong>Button:</strong> Turn on adaptive volume<br><strong>Secondary link:</strong> How your audio is handled</p><h4>Privacy note (must be reachable before enabling, and from settings)</h4><p><strong>Header:</strong> Your audio never leaves this phone.</p><p><strong>Body:</strong><br>This is the part that matters most, so we are putting it plainly. When adaptive volume is on, all listening happens on your device. The microphone audio is analyzed right here, on your phone, by an on-device model that is only deciding one thing: does this sound like sleep yet. No recording is saved. No audio is uploaded. Nothing is sent to us or to anyone else, and there is no audio file for anyone to find, because none is kept.</p><p>You can turn this off at any time, and when it is off the microphone is not used at all. We would rather you trust this feature than use it, so if anything here gives you pause, leave it off and fade the sound manually instead.</p><h4>Settings copy</h4><ul><li><strong>Adaptive volume</strong> — Let the app listen for sleep and fade the sound for me. Default: Off. Turning this on grants microphone access used only for on-device sleep detection.</li><li><strong>Fade length</strong> — How slowly the sound bows out once sleep is detected: 1 / 3 / 5 minutes. Default: 3 minutes.</li><li><strong>Bring sound back if I stir</strong> — Restore a low sound level if the room becomes active again. Default: Off.</li><li><strong>Snore-aware</strong> — Include snoring as a sign of sleep. Default: On.</li><li><strong>Microphone</strong> — Used only while adaptive volume is on, only on-device, never recorded, never uploaded. Manage system permission.</li></ul><h2>Sources</h2><ul><li>Snollo product documentation, 2025 — on-device Core ML classifier for snoring/breathing; raw audio never leaves the device.</li><li>Shin, H. &amp; Cho, J., "Unconstrained snoring detection using a smartphone during ordinary sleep," *BioMedical Engineering OnLine* / PMC, 2014 — smartphone microphone snore detection during normal sleep.</li></ul>