Eight hours in bed and eight hours of sleep are different things, sometimes by more than an hour. NOYU shows two metrics to keep them apart.
Sleep duration
Sleep duration is the time you actually spent asleep. It excludes the minutes you spent falling asleep, the times you surfaced during the night and the stretch between waking and getting up. So it will usually read shorter than your time in bed, and shorter than your own sense of the night.
Most adults need somewhere between 7 and 9 hours. Where you sit in that range is individual, and the honest test is how you feel on the days after: whether you wake restored, or need an alarm to drag you out.
A night belongs to the day you woke up. Sleep from Tuesday night into Wednesday morning appears under Wednesday, the same convention Apple Health uses, so the night you're looking at is always the one that set up your day.
Sleep efficiency
Sleep efficiency is the share of your time in bed that you spent asleep. Seven hours of sleep across eight hours in bed is 88 percent.
It's the fragmentation detector. Duration alone can hide a broken night: seven hours of sleep might have been one continuous stretch or a night of repeated waking that happened to add up to the same total. Efficiency separates the two, which is why a night can leave you tired even when the duration looks respectable.
Most healthy sleepers land between 85 and 95 percent. Ordinary brief awakenings, most of which you won't remember, are part of normal sleep - efficiency drops when you take a long time to fall asleep, lie awake in the small hours or spend your mornings in bed scrolling. Near-100 percent isn't the goal either: falling asleep the instant your head touches the pillow usually means you're running short.
How your wearable estimates sleep
No wrist or finger device observes sleep directly. It infers sleep from stillness, heart rate and its rhythms, and skin temperature. For the headline metrics - when you slept and how long - this inference is good, and better still on a device you wear comfortably all night.
Two things follow. First, devices disagree: a ring and a watch on the same body will report slightly different nights, because they're running different estimates on different data. If you wear more than one, NOYU picks one device for sleep rather than blending them into a night neither recorded. Second, edge cases exist: a very still hour of reading can occasionally register as sleep, and a wild night's tossing can register as wake.
Naps are recorded and appear in your data, but your sleep metrics and sleep score describe your main night.
How NOYU uses these metrics
Both feed your sleep score. Duration is the largest part - it's what wearables measure best and what matters most - scored as how close you came to a full night's need rather than against a leaderboard. Efficiency drives the restfulness part, alongside how often you woke.
The stages your device reports (deep, REM) carry a small weight - wrist devices only estimate them - and count only against your own typical levels. The full breakdown lives in Your sleep score.
Read the week, not one night
One short night is Tuesday. It costs you the next day, you sleep a little longer that evening, and the ledger closes.
The patterns worth acting on are weekly ones: an average that's drifted below what you need, efficiency sliding across a fortnight, a routine where the shortfall lands on the same nights each week. NOYU's trend views smooth over a rolling week for this reason - and if the numbers say one thing while your body says another, take your body's side and treat the data as a starting point for working out why.
