Sleep is not one block: naps, split nights and the midnight problem
Most sleep trackers assume you fall asleep once and wake up once. Plenty of real sleep does not work like that — and the calendar itself makes a night that crosses midnight surprisingly awkward to write down. Here is how to record the shape you actually slept in.
Open almost any sleep app and it will ask you two questions: when did you go to bed, and when did you get up. One number in, one number out, one bar on a chart. It is a clean model and it fits a particular life — asleep by eleven, awake at seven, every day, indefinitely.
It does not fit a nap. It does not fit waking at four and reading for an hour. It does not fit a shift that starts at eight in the evening, a newborn, a long-haul flight, or a Sunday afternoon that quietly turned into ninety minutes on the sofa. For a lot of people it does not fit most weeks. And when the tool cannot express what happened, people do the only thing available: they round. They log the big stretch and drop the rest, or they log bed-to-rise as though the two hours awake in the middle did not occur.
Rounding is not a small compromise. It deletes precisely the part of the record that would have been interesting later.
One block is a convention, not a law of nature
The idea that a healthy adult sleeps in exactly one uninterrupted stretch is more culturally specific than it feels. Historians of sleep — Roger Ekirch’s work is the usual starting point — have collected a large number of pre-industrial European references to a "first sleep" and a "second sleep", separated by a waking interval of an hour or so in which people prayed, talked, tended fires or simply lay there. How widespread the pattern was, and how it should be interpreted, is genuinely debated among historians, and you should treat confident claims in either direction with suspicion. What is not seriously disputed is that the modern expectation of one consolidated block is recent, and is bound up with artificial light and industrial working hours.
Biphasic patterns of one kind or another are also entirely ordinary in many parts of the world today, where an afternoon sleep is a normal part of the day rather than a sign that something has gone wrong the night before.
None of which tells you how you personally should sleep. It is simply a reason not to treat a night that came in two pieces as a failed version of a night that came in one.
The midnight problem
Here is a genuinely awkward thing that every sleep log has to solve and most solve invisibly. You went to bed at 23:00 and got up at 07:00. Which day was that?
Every answer has a cost. Put the whole stretch on the day you went to bed and your Tuesday contains eight hours of sleep that mostly happened on Wednesday. Put it on the day you woke up and Tuesday night vanishes from Tuesday. Invent a "sleep day" that runs noon to noon and you have introduced a private calendar that disagrees with every other date in your life, including the one on the entry you wrote the dream into.
SleepMind takes the option that keeps the calendar honest: the stretch is stored as two, one on each day. 23:00 to 24:00 belongs to Tuesday; 00:00 to 07:00 belongs to Wednesday. The app tells you it is doing this at the moment you enter a range that ends at midnight, rather than quietly reshaping what you typed and letting you discover the discrepancy in a chart three weeks later.
The consequence is worth understanding, because it is the price of the choice: a night like that shows up as seven hours on Wednesday and one on Tuesday, not as eight hours on either. If you are looking at a single day in isolation that can read oddly. Across a week it is exactly right, and it is the only version in which the dates on your entries mean what dates normally mean.
What a 24-hour timeline lets you say
Instead of two time fields, SleepMind gives the day a single 24-hour bar and lets you put as many stretches on it as the day actually had, in half-hour steps. That sounds like a small interface difference. It changes what is expressible:
- A nap. 14:00 to 15:30, sitting on its own, not smuggled into the night before or the night after.
- A night that broke. 23:30 to 03:30, awake, then 04:30 to 07:00 — two stretches, with the gap visible as a gap.
- A late shift. 09:00 to 15:00, in daylight, which is a normal working pattern and not an anomaly to be explained away.
- A day with nothing. No stretches at all, which is also information.
Two stretches cannot overlap: the app refuses the entry rather than warning about it, because two pieces of sleep on top of each other is not a thing that happened and the arithmetic downstream would be wrong for as long as it stood.
Each stretch carries its own score out of five. This matters more than it sounds. A wonderful nap and a wretched night are not one middling sleep, and averaging them into a single daily rating throws away the only interesting thing about the day. When a stretch scored badly you can tag what did it — noise, heat, cold, stress, the phone, stuffy air, light in the room, or simply going to bed too late — eight things, all of which are things you could in principle change.
If you work shifts
Shift workers get the worst of every sleep tool, because the tools assume a fixed night and shift work does not have one. A few things help when you are logging rather than sleeping.
Log the sleep, not the intention. A rota says when you were meant to be off; the log should say when you were actually unconscious. They diverge constantly and the divergence is the interesting part.
Use plans as a template rather than a target. SleepMind lets you build separate plans for weekdays, weekends and naps, each with its own days of the week and its own bedtime and wake reminders, and when a plan matches the day it pre-fills the timeline — dashed and badged, clearly marked as not-yet-confirmed. On a rotating rota, the useful move is often to build the plan you follow most weeks and then drag it around on the days you did not, which is still much less work than entering everything from nothing.
And be sceptical of the score on weeks like that. A number built for a seven-hour night will grade a rota badly no matter how well you handled it. That is a limitation of the number, not a verdict on you — there is a whole guide about what that score can and cannot mean.
What the record is actually for
A week of honest stretches is not a diagnosis and it is not a measurement. It is something less impressive and more usable: a picture of the shape your rest actually has. Whether the naps are a symptom of the nights or a perfectly good adaptation to them. Whether the four-in-the-morning waking is every night or only after certain evenings. Whether "I never sleep" means four hours or means seven hours that arrived in an order you did not like.
Those are questions you can look at once the record stops flattening everything into a single bar. They are also, if it comes to it, far more useful things to bring to a doctor than a vague sense of sleeping badly — but bringing them to a doctor is the point at which an app has done its job and should get out of the way. SleepMind measures nothing. It has no access to a microphone, a motion sensor or a wearable, and everything in it is something you typed. That is the correct basis for a journal and a disqualification for an instrument, and it is worth being clear about which one you are holding.
Frequently asked questions
Do naps count towards my total for the day?
Yes. Every stretch you log on a given date contributes to that date’s total, and each keeps its own score so you can still see which was which. If you want to know whether the naps are helping or wrecking the nights, the answer is in having both recorded separately rather than in leaving the naps out.
Why does a night that crosses midnight get split into two?
Because one calendar day has to own each hour, and the alternatives are worse. Put the whole stretch on the day you went to bed and most of it is filed under a day it did not happen on; invent a noon-to-noon sleep day and your sleep record disagrees with every other date you use. Splitting it keeps the dates meaning what dates mean, and the app tells you it is happening as you enter it.
Is sleeping in two blocks bad for you?
That is not a question a journalling app can answer, and anyone answering it confidently in a marketing page should be ignored. What can be said is that the one-consolidated-block expectation is historically recent, that biphasic patterns are ordinary in many places, and that a broken night is not automatically a medical problem. If you are waking repeatedly and it is affecting your days, that is worth raising with a doctor — not because the pattern is inherently wrong, but because a doctor is who can tell.
What if I genuinely do not know when I fell asleep?
Put in your best estimate and move on. Half-hour resolution is deliberate: it is about as precise as unassisted recall actually is, and a tool that asks for the exact minute is inviting you to invent one. Consistently-estimated data is fine for spotting patterns; precisely-fabricated data is not.
Can I add last night’s sleep later in the day?
Yes. The date strip lets you move to any day and edit it, so the common pattern — dream in the morning while it is still there, hours at lunchtime when you have a minute — works exactly as you would want. Neither half is required in order to save the other.
Write it down before it goes
SleepMind puts the dream note on the first screen, so you can tap a feeling and type a fragment before you have stood up. The sleep log is folded one tap below, on a 24-hour timeline that takes naps, split nights and late shifts. Free, on iPhone, with no sign-up wall.