Wrong sleep tracking and no SpO2

Sleep tracking on Garmin devices (and others) is so poor that I largely ignore it. Two nights ago I had REM sleep while I was awake. And I'm routinely credited with light sleep while wide awake. I do find Body Battery tracks how I feel quite well, including the ups and downs.

Sleep quality is an important part of the new training readiness feature. Out of 6 parameters, 2 are linked to sleep detection (sleep and sleep history).

So it became harder to simply ignore it

When discussions like this appear i'm always fascinated on how the watch works differently on different people. Would be really nice to understand the reason.

I never got any recording of the sleep outside of my sleep schedule(+- 1-2 hours as the schedule is not super strict).

I went to search for a nap and i found the following one that was pretty long and recorded well by Oura

In Garmin, there's no trace of any sleep recorded...there's no movement at all and the HR is lower(not like during the night but it's close).

Zoom on the nap period.

The sleep starts pretty fine within the schedule...also with higher HR.

If i go for a nap i put my Garmin in sleep mode too(i don't use DND...i have set only the shortcut for the "sleep" mode if i don't want notifications, etc) but never had any recording of sleep and overwriting of the previous night.

Mladen T said:

fascinated on how the watch works differently on different people.

Sleep data depends on the accuracy of the OHR, which we know is affected by skin tone etc., and how the watch is worn. Factor in the different positions people sleep in then it become easy to see just how difficult it is to get consistency. The Oura ring is on a finger so less prone to some of the possible errors that a OHR device is exposed to.

Recording of naps is a different issue and we know Garmin don't do that currently. But equally, if you look around the interweb not many other watches do naps well either. We can  only keep pushing on Garmin to make the changes and improvements that we'd like to see. Sometimes it just takes a while but if you follow Garmin's history, things do change and do get added.

Probably my point was not really clear...as people get a recording of sleep during the day and outside of their sleep schedule i just wanted to show how that situation is not happening with me(and many others, i know) and i used, as reference, one of the days in which i really took a nap.

Oura recorded the nap so i guess the factors for sleep recording were all there(like shown also by Garmin)...lower HR, no movement, and not really short period.

Would be interesting to try to find a common point among the people that have the issue, but i guess it's not something that can be done easily on the forum 

Mladen T said:

people get a recording of sleep during the day and outside of their sleep schedule

I don't take many naps (actually we call them SCANs now - Senior Citizens Afternoon Nap as it sounds better to say we're going for a SCAN!) but have yet to see a sleep session recorded when I do. I have my sleep schedule set 2130 to 0610 but sleep starts both before and after the start time and finishes before and after depending on the time I actually settle. Similarly, I always read lying in bed for up to an hour before settling to sleep without any effect on my sleep recording. Why is that? That's the $64k question. I have wondered if the problem affects people with high resting heart rates more than those with low. Just a thought but nothing conclusive to support that hypothesis.

A few examples of wrong sleep detection starting when I start watching TV around 9:30pm. I took the screenshots in the morning, before I changed the sleeping time to the actual go-to-bed time (typically right after 11pm). The scheduled sleep period starts at 11pm, but clearly has no effect on sleep detection. There is also clearly no learning process if you adjust your sleeping time subsequently.

The problem is that the larger pink awake period that precedes the going to bed routine taxes the sleep score significantly, which then taxes the training readiness.

As described in the beta forum, the new issue with the latest firmware is that sleep history is not adjusted after correcting the sleeping time manually, although the sleep score of that day is modified, but not in the history...

What a mess...

Etupes25 said:

The scheduled sleep period starts at 11pm, but clearly has no effect on sleep detection.

The sleep detection is more "sensitive" around the sleep schedule with a timeframe of 1-2 hours i would say(can't remember the exact value that i read somewhere on Garmin website/forum) but 3 hours looks a bit too much(i could be wrong of course and maybe the timeframe is really 3 hours).

I have another funny comparison.

My sleep schedule starts at 9:30/10:00pm and if we use a 3 hours timeframe, as per your worst case, the sleep detection should be more sensitive from 7:00pm.

So i have this day where i fell asleep between 6pm and 8pm, then i woke up and went to sleep around 10pm.

For whatever reason, Oura decided to consider it as one sleeping cycle, as you can see.

On Garmin i have no sleep until 10:08, which matches more or less(10-15 minutes of difference) with Oura's 2nd part.

The average HR is around 56 between 6-8pm and just a bit of movement.

So, if we consider again the 3 hours timeframe i would say that 1 hour of sleep(7-8pm) was too short to start the cycle and it was affected by the 2 hours of awake too.

Did you try to set your sleep schedule a bit later? Let's say around midnight or even 1am? It's not the best solution but could be a nice workaround.

Out of curiosity, can you post the HR graph around the sleeping period too?

I have noticed that if sleep is less than 3hrs it doesn't record it.

That's absolutely true but the 3hours don't need to be continuous, at the beginning, to start the recording, otherwise those 20-30 minutes of awake from Etupes would stop the tracking.

In my case, the awake time of 2 hours was, most probably, too long to link the first part of the sleep(i would say only from 7pm to 8pm as the part from 6pm to 7pm was too far from the sleep schedule) with the 2nd part, from 10pm.

If my awake had been, let's say, 30-40 minutes only, i would imagine that the sleep cycle would have started at 7pm with 30-40 minutes of awake, somewhere in between 8 and 10pm and then the real sleep.

The maximum awake that i had without interrupting my sleep(within the sleep schedule of course) was around 1 hour.

I woke up, did breakfast but then decided to go back to sleep