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

Garmin Venu 2 Scientific Sleep Test: some issues

I have been a subscriber of one channel on YouTube that produces scientific tests of various health trackers. 

The channel released a sleep tracking test of Venu 2. 

Results look so-so - there are issues with REM detection also it is not very precise at detecting start and end of the sleep. I hope Garmin keeps track of these issues and tries to fix them if possible.

Link: https://www.youtube.com/watch?v=2Y5M_N9CrNI

  • There may be some "issues", but honestly, I have owned a Ticwatch Pro, and Samsung Galaxy 3, along with the new Galaxy Watch 4 Classic, and the venu 2 that I had is much better than any of those.  Actually, in my opinion, the Venu 2 is at least as good as the Fitbit Sense in detecting the various stages of sleep. Just one person's opinion,, and of course, as Rainmaker says, "Your mileage may vary"..

  • How would you compare the Venu 2 to Galaxy Watch 4? I've been interested in that watch, but I don't have a Samsung phone (LG). I also would love the blood pressure feature but I'm in the US.

  • Former Member
    0 Former Member over 3 years ago
    Results look so-so - there are issues with REM detection

    That's because your V2, or any other wrist worn device, can detect REM. There are 3 requirements of REM sleep.

    1.  Slightly twitching of the muscles.

    2.  Increased brain activity.

    3. And obviously; rapid eye movement.

    A wrist wearable may be able to detect #1, but cannot detect numbers 2 & 3.  Wearables use educated guesses when someone enters REM based on averages from clinical studies. Some manufacturers guess better than others, but none have the ability to actually detect REM.  If you listen to the gentleman in the video, he even says, "let's see if it can predict...".

    one channel on YouTube that produces scientific tests of various health trackers.

    They are comparing readings the wearables provide to data they have from scientific sleep test.  They cannot conduct scientific sleep test on the watch specifically because wearables aren't capable of detecting all the requirements to determine if the user is in a state of REM.  They're just informing you on how well that manufacturer/device guessed.   Your $400 USD V2, not $400 USD sleep tracker, measures HR, respiratory rate and movement during sleep just like other wearables that have the sleep tracking feature.  But none of these alone, or together, can measure REM.  There is no set standard that says this HR, respiratory rate and/or amount of movement equals REM.  Diet, health and medication can affect those and has nothing to do with what stage of sleep you're in.  Your watch is doing nothing more than taking measurements throughout your sleep and then using a proprietary algorithm; that every manufacturer uses, to say the user was most likely in this state of sleep at this time for this amount of time.  These proprietary algorithms are based on averages taken from clinical studies conduct with equipment that cost thousands and thousands of dollars.  There's a reason why EVERY SINGLE manufacturer states in a disclaimer that their data are estimates.  Because they're nothing more than calculated guesses.  Even the YouTube channel that conducted the "test" has to be taken with a grain of salt.  No scientific results can come from comparing a wearables results once, twice or even 10 times.  The data is speculation unless several subjects were tested multiple times under controlled conditions.  There have been many, many threads in these forums where people claim their REM sleep is wrong compared to their Fitbit, Galaxy, etc.  Nope.  They weren't "wrong".  They were just uninformed/misinformed on how these watches work and they preferred those results over the one's Garmin was giving them.  

  • Interesting review, in my experience in the past couple weeks I get a ton of rem sleep but not much deep sleep. I don’t dwell on the stages too much. My huge frustration with garmin sleep tracking so far has been its failure to detect sleep and wake up times correctly. And then even after editing the correct times, the software doesn’t recalibrate the sleep stages, sleep score, body battery, etc. Really wish they would correct this issue or add in a feature to manually start and stop sleep times like whoop and other devices have. 

  • Hi,

    Thank you for your answer. I think it adds a lot of depth to the discussion. By no means I am a scientist (at least in the right branch of science). I think that the channel did a great job at highlighting issues, despite conducting the test only 4 times. I agree that in reality the sample size should be determined using the expected differences in measurements between the control and two tested devices. 

    When comparing Garmin Venu 2 (my first Garmin device) to my previous Huawei Band 3 Pro I found the latter to be almost perfect in detecting the beginning and end of sleep. However, use of analytics, I feel that in general Garmin does a much better job of predicting your energy levels and performance throughout the day based on the sleep quality. I just wish it was more accurate with detecting transitions between sleep stages. I think that this is something that can be fixed by software updates.

  • You can actually stop the sleep manually by going to the menu, finding the sleep stats widget that would say Sleeping.. and press it. 

    I find it puzzling that my previous Huawei Band 3 Pro, which was much dumber, was so accurate and precise at detecting sleep start and sleep. I think that the device is still in its early stage and that Garmin will improve accuracy.

  • Like the channel, but really don't get the statistics haha. 

    I found the sleep tracking to get much better after a week or two. I had really strange sleep data in the beginning of owning it or also after a full reset. But like Body Battery, Stress etc. I found sleep to get more accurate over the first weeks. And he only tested it for 4 days. So maybe he should wear it a few weeks and then compare it. But I guess that's hard to do if you have a channel which tests so many different devices. 

    Atm I wear the Fenix 6s Pro with the similar or same sleep tracking as on the V2 and found it some days almost perfect, esp awake time. But some days mehhh. Don't know why there are sometimes such big differences. Today I had over 9h of sleep and felt good, a little too long sleep but I really felt good. But score only said 69 and BB 75 and after a not so good sleep I sometimes get 90 score and 95 BB, BB rising high from low value the day before. 

    But I really like the channel, not many of this kind out there. 

    But I have read in studies that no watch, none, can read sleep stages correctly. So you should always take these results with a grain of salt. It is just a hint, esp awake time. But esp deep sleep is quite hard to measure or also differentiate from REM sleep etc. So sleep results on a watch can give you a hint but not more. You need to go to a sleep labor and do a sleep study to get correct results. 

  • Well, I'm still evaluating the GW4, and as a smart watch, it's fine.  It's quick, it's snappy, no choppiness, and I really like that part of it.  The voice commands are good, accurate, and again, quick.  The downfall, and I suppose this is because I've owned Garmins (from early Forerunners, Fenix's (3, 5, 6, 6Pro) Venu and Venu 2, along with Apple watches, is that the health tracking on me, just isn't accurate.  I've found the SpO2 readings are consistently 4% higher, and on me, for some reason, if I walk around, say go to the store, and push a cart around the store, the heart rate is intermittent for at least an hour,  sometimes as much as 2.  No one, not even Samsung can explain why this is.  The ease of exporting Garmin stats to other platforms is also so easy compared to Samsung health, which, by the way is ONLY on the phone, there is no web presence for it, so you have to "analyze" your activities on the phone's screen.  Granted you can export your activities using HealthSync, but not the day to day sleep/hr/steps, etc.  Samsung health gives a lot of metrics, but there's no real way to compare say, what you did no that run last week with this mornings run.  I'd say if you want a smart watch, with the voice message reply, and all the other little things, (No Starbucks app though, what's up with that?) then go for the GW 4.  You won't have BP or EKG if you don't have a Samsung phone though.  If you're interested more in your health data, the by all means, I honestly think the Venu 2 is head and shoulders above the Samsung.  Venu 2 measures HR every 2 minutes, the GW4 every 10, and that's with the "Measure continuously" switch on.  I tried the auto recognition of an activity yesterday, and it didn't show up until this moring, and there was no Hr info, just an entry "You walked this far for this long" and showed average, min, and max speed.  I'm seriously considering returning the GW4 and re-purhcasing the Venu 2, I returned my first one, because I was caught in the (now famous) shutdown in the mornings after syncing" syndrome.  I hope that helps, if you would like more info specific to you, let me know and I'll try to help.

  • Thanks for the reply! Great information. I've read the GW4 battery is pretty bad too. 

    FYI, here is modified software that removes restrictions:

    https://forum.xda-developers.com/t/restrictions-removed-samsung-health-monitor-wearos-1-1-1-153-root-age-country-device-restriction-removed-29th-aug-2021.4322527/

    This will give you ECG and BP. 

    I love the idea of more features, but inaccurate data is worse than no data. My Venu 2 screen shuts off every couple of days, and there's no easy, repeatable way to get it to turn back on. It also has trouble tracking my heart rate when walking (can get into the 200s) and at the start of a run (about 20 beats higher than it should be).

  • A few weeks ago I was watching a movie in the cinema. Movie started 9 pm and the venu thought I fell asleep. Not noticing I was actually awake and drove home (I was the driver) but still Venu 2 thought I still was sleeping.

    Algorithm is very primitive obviously.