This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Is the sleep tracking that bad?

Everyone on the gamrin subreddit seems to say it's horrible compared to the Fitbit one & other smartwatches. I'd like to buy a vivosmart 4 only for the sleep tracking, HR, pulse ox & stress tracking. Any feedback on these features and how well they're working

Thanks

  • Yes the sleep tracking is that bad. It's not useless, but close to it.

    The rest is really good, esp te stress tracking and body battery feature. 

  • Not knowing what people are expecting or what results they're getting, I'm not entirely sure what the question means. Sleep tracking reliably notes my sleep and wake times automatically, and the movement graph seems to indicate it's tracking my wrist movement throughout the night (and it's definitely tracking heart rate). It does so reliably despite my very erratic sleep patterns - I'm presently up at 5:20 am, for example, while I usually go to sleep between midnight and 2:00, and I fully expect it will accurately determine that I've been awake all this time - and that's really about all I can ask. Of course, I can only speak for my own experience, and I haven't used any other trackers (my mom has some kind of FitBit, and I haven't noted any obvious functional differences when I've looked at her data).

    The objections I've seen on Reddit when I looked just now and comparisons to e.g. FitBit aren't really worth much: they're anecdotal, subjective, and the people complaining don't seem to understand the inherent limitations of sleep tracking using an accelerometer/gyroscope on one wrist and optical heart rate sensor only - all it can do is detect heart rate and wrist movement and attempt to correlate them with known sleep data. It may well be the case that e.g. FitBit's algorithms better reflect how some people move when they sleep, but in doing so they will inevitably reflect other people worse, so there is no way to have one thing that's best for everyone (the best option is to train the algorithms on a per-user basis, and if any given company really does outperform others, this is how they do it; Garmin does do this to some degree, though it's possible that it's not as well implemented as a given competitor).

    The sleep stages thing is BS across the board, for all companies - there is simply no reliable (or even consistent enough to be at all useful) way to detect different stages of sleep based on wrist movement and heart rate. You need an EEG for that, so they're best regarded as being for entertainment purposes only no matter what brand we're talking about.

    The sleep blood oxygenation measurement isn't much better in practice - while it can detect blood oxygen saturation well enough for an optical sensor, I find that it has to be too tight on my wrist to wear all night in order to get consistently reliable readings.

    The stress tracking function (during sleep) is actually much more useful for examining sleep quality than any sleep stages estimates. That compares heart rate to wrist movement to determine high heart rates during inert periods, which is an indicator of poor sleep quality (alcohol consumption close to bed time is a very common cause).

  • Take a look at the The Quantified Scientist channel on youtube. He tests fitness trackers against a professional EEG sleep tracker, and the only company that does okay is fitbit. 

  • Just bought my Venu and last night had about 2 hours sleep and it told me I had 8, I wasn’t even in bed that long…