Interesting phenomenon what I noticed last days…
One night, after the defined begin of sleep, I left my s9pp from wrist away to take a shower, then later i wear it back and went to bed.
And noticed the next day that the sleep begin, started at the time as I put it away from wrist.
Thought well, that’s wrong, it shouldn’t take it in account.
But after some thinking I thought, if the system detect no signs (no pulses no accelerations), and if it happen by defined the sleep time.
It belong to sleep ? or no sleep ? then defensive thinking, that’s sleep.
I found that not good, but a dev, I would program like this.
An other day, i left the watch 30 min before the sleep begin, during one hour, then wear it again, and was active 1h and went to bed.
The system interprets the sleep began half an hour before the sleep begin, and was near of 9,5h long with 70% quality, what’s fully wrong.
Seems that the algo used the first h with 100% good, then the 1h sleep brake (a brake of 1h? really?) made it worse to 70%… the rest was maybe good.
Thus i think it would be not bad, or better said, better if we have our sleep represented as diagram, and maybe have the possibility to correct it afterwards.
Don’t know how is the experience of the suunto community colleges about the sleep tracking by their side…
How it is ? some more interesting experiences ?