How accurate is a phone at this, really?
A phone microphone can support specific acoustic tasks. The honest answer requires task-by-task evidence, production boundaries, and more than one metric.
Read article →BLOG · WRITTEN BY THE TEAM, NOT A CONTENT FACTORY
What I've been writing.
Plain-English articles on sleep breathing, phone-audio observations, validation, privacy, and where a wellness record stops. No "10 sleep hacks." Each article is designed to be read in a few minutes.
START WITH YOUR QUESTION
Not sure which article to open first?
Breathing & snoring basics
Start with the terms SomniSense uses and what your nightly patterns can — and cannot — tell you.
Explore this topic →Accuracy & privacy
Understand phone-based tracking, validation, and how your recordings stay under your control.
Explore this topic →Habits & experiments
Learn how tagged and untagged nights can be compared without turning an association into a causal claim.
Explore this topic →Wearables & comparisons
Learn why different sleep products measure different signals and should not be read as equivalents.
Explore this topic →-
-
Why your sleep audio never leaves your phone
Sleep audio is analyzed on your phone and is not uploaded by default. Learn how explicit sharing differs from the account, aggregate and technical data used for the service.
Read article → -
What "validated against PSG" should tell you
A validation claim is only useful when it names the task, dataset, reference, and limits. Here is how to read that evidence in plain language.
Read article → -
Cannabis, CBD, and sleep: what a tracker cannot conclude
A personal tag can organize nights for comparison. It cannot establish safety, efficacy, sleep-stage effects, or a biological mechanism.
Read article → -
Apple Watch sleep stages and a separate sound record
Apple Watch estimates REM, Core and Deep sleep. Learn what those stage estimates show and why a separate phone-audio candidate cannot explain or verify them.
Read article → -
Why nights vary — and why a tracker should not invent the cause
Room conditions and routines can be useful tags. A phone-audio record can compare nights, but it cannot identify temperature, cortisol, or a mechanism.
Read article → -
Eight hours in bed and still tired? Duration is one clue
Time in bed and how you feel in the morning are different observations. A multi-night sound record can add context without identifying the cause.
Read article → -
Weekend sleep looks different? Compare before explaining
Weekday and weekend nights may differ. Repeated observations can show a pattern, but a tracker should not assign the cause.
Read article → -
Apple Watch sleep scores and BRI measure different signals
A watch score and SomniSense candidate acoustic-event rate are not equivalent measures and should not be used to validate each other.
Read article → -
What phone audio adds to wearable breathing data
Wearable sleep, breathing and sound features vary by model and companion-phone setup. See what an independent bedside recording can add without using one result to explain the other.
Read article → -
Wine and sleep: record the observation, not the cause
A wine tag can help compare nights. A phone-audio record cannot establish sleep quality, sleep-stage effects, or causation.
Read article → -
Understanding BRI — and why it is not AHI
BRI is SomniSense's product-defined rate of candidate acoustic patterns per estimated sleep hour. It is not clinical AHI.
Read article → -
Sleep apnea 101 — symptoms, evaluation, and phone-audio limits
A plain-English guide to common symptoms, why they are not specific, when professional evaluation matters, and what SomniSense cannot determine.
Read article →