Related
Options

Inaccurate pulsoximeter

Misiek
over 3 years ago

So I replaced 3 watches just to learn that the pulsoximeter sensor in Gamin Tactix Delta (and most likely any other Fenix 6 series watch) is totally inaccurate. Completely out of reality when compared to professional devices. Now, the funny thing is, I have two Xiaomi Mi Band 6 devices and guess what - most of the time the readings are exactly the same as professional devices. How is it possible we get such crappy sensors in such expensive devices when competition can make accurate cheap devices?

I read other threads on this topics, some were closed (not sure why), I literally checked everything (including Garmin support articles), measured on different people, different conditions (bright, dark), different units, the same results. I think it is unacceptable in Covid times to release such inaccurate product for such price.

Any other ideas ? Do you think there is a chance for software upgrade solving the issue or is it simply a problem of inaccurate sensor used by Garmin ?

  • 0 over 3 years ago

    Without trying to start any arguments (as these pulseOX threads tend to do, hence why they keep getting locked), basically here's the answer:

    Wrist-based Pulse Oximetry is an inexact/imprecise technology... for EVERY watch. Your cheapo Xiaomi watches are perceived by you to be more accurate, but that does not mean they are. As the saying goes, "even a broken clock is correct twice a day". Your fingertip-based sensors and your Xiaomi watches might both show 97-99% or so most of the time, give or take 1%, and so you've deduced that the Xiaomi watches are accurate. But if you try to test yourself when your O2 % is outside of the "normal" range, (say, 85%), the fingertip oximeter will likely correctly show 84-86% or so, while the Xiaomi will probably still say 97-99% because they likely programmed it with a bias skew towards the "normal" range to give the perception of accuracy to most people most of the time. Whereas, Garmin doesn't bias/skew their readings to create a false sense of accuracy, they just give us the raw readings (accurate or not).

    The reason the PulseOX exists on this watch is not the reason you (and many others) think. It's there for recording changes to your long-term TRENDS of your bood oxygen saturation. For example, if you significantly go up in elevation, you can watch the trend go down 10%, and then slowly rise back up as you get acclimated. The precise % at any given moment is not important for that.

    TLDR; The pulse oximeters in your Xiaomi watches aren't as good as you think they are, and Garmin PulseOX is not designed or intended for accurately spot-checking your blood O2 saturation like a fingertip oximeter.

  • 0 over 3 years ago in reply to C.sco
    C.sco said:
    The reason the PulseOX exists on this watch is not the reason you (and many others) think. It's there for recording changes to your long-term TRENDS of your bood oxygen saturation

    It seems Garmin happily applies the same logic to everything. 

    For example, the reason BodyFat% exists in my Gamin Index 2 scale is not the reason I (and many others) think. I thought it was for measuring my BF%, in it is there merely for recording changes in my long-term BF% trends. 

    Also the reason my Garmin Fenix has pace field is not the reason I think. I thought it was there to tell my pace, but in reality it is there to show me the trends of my pace during entire run.

    And so on...

  • 0 over 3 years ago in reply to C.sco

    It is not the answer I expected, but I think it makes sense what you wrote.

    On the other hand just to be clear. The results are not very different, it is something like 1-3% difference from professional device. On some occasions the results are exactly the same, sometimes it is 1-3% difference. I do not know why, but with Xiaomi Mi Band 6 it is almost always the same. So my professional device shows 97%, the same result on Mi Band or my old Galaxy Note 8 smartphone, but on Garmin it can be 97%, but also 94% or 100%. Then if I see 95% the same result is on Xiaomi Mi Band 6 (or Note 8), sometimes there is +/- 1% difference, but not more. So this is just little bit strange that such cheap piece of equipment can give quite accurate results, but for Garmin it is floating. 

  • 0 over 3 years ago in reply to C.sco

    Do you have data to back up your assertion the cheap Xiaomi watch skews its readings to give the perception of accuracy?