Sleep tracking

I'm also fairly new to the Garmin ecosystem. When it comes to sleep analysis, there is only so much that technology that is worn on your wrist can do for you. They can use accelerometer, hr data, hrv data, pulse ox in the 945, but unfortunately, when it comes to sleep tracking, it's all about brainwaves. No matter what Garmin and or FirstBeat claims, there is no direct relationship between above and brain waves. There is a huge difficulty in determining between REM and deep sleep for example, or laying in the bed, but not sleeping and REM sleep etc. The best you can hope for is consistency between measurements and trends that is relevant to you.

To demonstrate my point, please look at my last night's sleep below. I'll explain what really happened:

I most definitely didn't get sleep until about 9:16, so anything between 8:23 and 9:16 is awake.

You see the pink awake line shortly after 1am? I actually woke up, so far so good. But anything between that and the 3am pink line is bogus. I was awake all the time, laying in the bed.

I think what is needed here is the ability to be able to edit the sleep data after the fact, with some AI probably learning about what your patterns mean and improve over time. But as I said above, this will never be as accurate as a sleep study would.

The sleep tracking results happens on Garmin servers. All the watch does is send the data via Connect to their servers.

I agree. Compared to the Fitbit, sleep tracking is sub-par. Makes the rest of the metrics the watch provides questionable. 

I consistently get just one "Deep" sleep at exactly 10 minutes after "bedtime" - even if I am just lying quietly and reading. And that's the only deep sleep the watch thinks I'm getting. Also, it doesn't seem to know that I'm awake. Shows I'm in REM sometimes when I'm actually up walking around and my heart rate rises accordingly.

Doesn't matter where the actual results are determined, they're useless.

I'll add an issue to this and a request that Garmin invest in this area.

I frequently have gaps in the night with no data at all. Not sure why that happens, assuming it will continue, but my issue is with how Garmin handles it. This is subtracted from my total night sleep, making the overnight sleep metric not useful to me. I'd like Garmin to either handle this differently, or to at least let me edit it.

Hello, I've been synchronizing my sleep data every morning for about three months. Looking immediately at the graph in Garmin Connect, everything is perfect (for example, 4 hours of deep sleep, 2.5 hours of light sleep and 15 minutes awake).

After about six hours Garmin Connect re-elaborates the data and they all become wrong!

Every morning when I wake up I synchronize and save the exact page of the sleep in .pdf, making me a personal archive.

Can anyone check if it also happens to him?

I've run into this.  What I do is go to edit, and save once I wake up.  

I have also experienced a similar situation, not sure what to do about it!

I hope the invest a bit more time on this and do the best they can within the constraints they have, it's an interesting metric.

It would also be nice if they used more distinct colors, I don't understand the need for different shades of blue and purple, nothing wrong with using green, yellow, etc for different types of data. Just makes it more difficult to visually parse.

Exactly