Awkward sleep metrics

[altcmd]

I know the consensus on the forum is to not use the watch for sleep, but here is my recent experience of it.

I slept at 0:06 last night and woke up at about 7am this morning, with the alarm. On wake up the watch showed me a Good Morning message with sleep stats like I slept 6:54 hours and 55% resources regenerated. I stayed in bed until 7:30 and woke up, walked 20 steps and then removed the watch from my wrist. The App showed the metrics, similar to what I saw on the watch, although I see sleep summary minimum bpm to be 61, while in the HR graph, I see 49bpm! Not sure why the difference in what seems like the same metric.

In any case, I picked up the watch again around 9:30 and then at 10:15, I saw the sleep stats on the watch had updated to a duration of 9:32 hours, apparently waking up at 10:13. Same has now synced with the App.

It would seem that the watch expects you to keep it worn for a period of time else assuming I am still asleep? Plus there is a persistent gap in the sleep HR almost every night - not sure why.

[altcmd]

Another wrinkle in this is how the Resources has, or perhaps has not reacted to it. Even if I were to assume that I had fallen asleep and woken up at 10:15, my Resources has supposedly fallen from 55% to 44%. Looks like I was working hard in my sleep as well

[DMytro]

@altcmd I don’t know how to resolve the inconsistency, apart from syncing the watch to SA again, maybe your intermittent sleep result got uploaded instead of the final one.
But I don’t know, why the watch shouldn’t be used for sleep tho - for me it is quite accurate most of the times. And I have an S9 which has arguably much worse HR sensor than s9pp.

[altcmd]

@DMytro The S9PP has been synced multiple times since 10:15 and it continues to be so. I even powered-off and then on and still the same.

[freeheeler]

@altcmd
you might already know, but just to be sure: if your sleep settings are way off your actual sleep (or vice versa) the watch sometimes struggles catching your sleep correctly

[altcmd]

@freeheeler I have been, but was hoping for a revamp with new hardware - not seeing it atleast for now.

[Brad_Olwin]

@altcmd Sleep is a tough metric to follow, movement during sleep has to be interpreted as either sleep or did you wake up? In the morning if I do not trigger the wake up on the Suunto and put my watch down (for example to take a shower) the watch has interpreted this as me going back to sleep. I have had a failure where I wake up and go back to sleep but the watch did not correctly interpret this.
So, it is not perfect and difficult to do. The gaps you are seeing could be due to the watch being too loose or from no movement at all and the watch sleeps.
If not for testing, I would not wear a watch to sleep. The other reason I wear one is for an alarm to wake me with vibration and not wake my wife. My lowest HR during sleep and avg HR are often a good indicator of how hard my prior day was. I find those numbers useful.
What are your reasons for tracking sleep?

[altcmd]

@Brad_Olwin A simple answer to your question would be that I want to track sleep for the same reason as I want to use a watch to track my sports. As I detailed above, the watch did give me the wake-up Good Morning prompt with all the stats, and the same showed on the app too. But what is unclear is why it later assumed I slept until 10 and subsequently did not account for resources. I don’t quite buy the argument that watch was too loose or no movement during sleep - isn’t that why you sleep! I don’t have any concerns on sleep stages as neither Garmin nor Polar or Oura have good consistency on that but here my concern is more with the simple wake up and sleep times which I feel Suunto consistently gets wrong for me. Whether one needs to wear a watch to sleep or not is subjective but a feature has been made available to be used and a consumer can only assume that it works. Else Suunto should just remove them as the science on it is just not conclusive. But here again Suunto has added running power which has its own disagreements.

[el2thek]

@altcmd resources are based on your heart rate. So that would explain that. What are your sleep settings?

[freeheeler]

@altcmd
sleep tracking works really well for me, even with S9B.
the only thing I can imagine why your watch thought you went back to sleep is because wakeup time is set later than you actually got up this morning. but only you can assess that.

I use sleep tracking for the same reason as you, tracking sports why not tracking sleep when we have the possibility. maybe I can see a pattern some day.

how were the recent messurements? you said this one was wrong, how about the others?
I wouldn’t ask for removing a feature because it fails every once in a while

[altcmd]

@freeheeler I wasn’t advocating for removing it but saying that don’t use it defeats the purpose. But the sleep tracking time frame could very well be the culprit here - my time is set from 10pm to 8am, so waking up earlier than 8am and getting a Good Morning report is perhaps not good enough for the watch to understand that the user is awake. If on-wrist detection has been improved on the watch, I am not sure why this cannot be used to further confirm if the user is still asleep or not, rather than just assume that he is until he presses a button after 8am.

[altcmd]

@el2thek I know but just my point that sleep has been assumed without detecting a heart rate, but resources has continued its normal trend - an anomaly in the code, I feel.

[el2thek]

@altcmd Sleep tracking also works without tracking your heart rate. I think you’ll get better results having your sleep times a bit too early, than too late.

Alternatively, wear your watch a little longer before taking it off. The 20 steps you described could also have been going to the toilet and back to bed.

[altcmd]

@el2thek Got it. Thanks!

[freeheeler]

@altcmd
mine is set from 22:30 - 8:00 and my actual bedtime varies between 21:00 to 0:00 and 5:00 to 9:00. the watch gets my sleep really well, unless I watch TV and sit lazy there it happens that the watch starts tracking even before I go to bed. On the other hand, after an exhausting day I go to sleep earlier, the watch tracks, too and that’s what I expect.
On the other side, in the morning, once I’m up the watch recognizes but I am always active then.
I don’t remember that it has ever added more hours e.g. when I got up at 6 but layed lazy down for 1 more hour.

long story short: I suggest you play with the sleep limits and maybe find the setup that fits you best

[altcmd]

@freeheeler Thanks! Will do!

[mikekoski490]

@altcmd Mine is OK generally but I would prefer if there was an option to choose ‘Stop sleep tracking?’, so you can lie in bed with a coffee and not have that tracked as sleep. COROS had that option. I also noticed last night a 1 beat difference between Min. HR values on the same nights SA results and am still getting the odd gap in the sleep HR graph (?).

[isazi]

@mikekoski490 if I’m not wrong the 1bpm difference between watch and SA is rounding. Been there for a while I think.

[Brad_Olwin]

@altcmd mine is set from 22:30 to 6:30 and I typically wake 5:30 to 6:00 with no detection issues. For me Suunto does not recognize a short awake period during the night but does effectively detect sleep and morning wake times.

[altcmd]

@isazi Any thoughts on why I am seeing a 12 bpm difference?