Optical HRM still totally wrong with 21.19 - In fact, worse than ever

Hi 

unfortunately I have bad news. You closed OHR related topics like this
https://forums.garmin.com/sports-fitness/running-multisport/f/forerunner-955-series/384760/heart-rate-values-from-optical-hrm-still-totally-wrong-with-20-26

and this
https://forums.garmin.com/sports-fitness/running-multisport/f/forerunner-955-series/362273/heart-rate-values-totally-wrong-after-update-18-22

But you were asking to give you feedback, if we encountered that bug again.
This is the case - in fact, in my opinion the optical HRM is worse than ever now.

With 20.26 I was quite ok with the measurements of the OHR while walking/hiking and for higher intensity workouts I usually use a HRM strap (beside bike commuting).
With 21.19, the OHR cannot be trusted AT ALL, even for low intensity stuff like walking/hiking.

See this heart rate values, that I recorded during an hours walk today:

The average is a HR of 70!!!
While walking one hour with a pace below 10 minutes/kilometer.

My resting HR is 48, still, this is just random. It is nowhere near the real pulse I had during that walk.
And regarding blood flow: Yes, it is cold now in Germany (4 degrees Celsius), but I was wearing a thick fleece pullover and Polartech gloves and the watch was below it all.

Yes, you may contact me and have a look at my activities and I live in Germany.

  • At least for me it seems to be fixed at the moment.

    Ich compared the wrist heart rate and external via HRM Pro. After one minute the wrist catches up and stays good.

    I'm also located in Germany. Temperatures were around 0°C.

  • OHR was always not good for me, especialy on rough bike rides or interval runs.  But the option of source switching in actual state is also very bad idea. 

    I tought it will change from my external HR to wrist only when external will out of juice,signal or other. But no, it randomly decides what to use, very often it choose OHR for main source of hr. For some time now, OHR gives me to low hr number and it skews a lot. 

    Turning it off helps. But the external hr should have higher priority than OHR when connected with source switching. 

  • Thats a thing I can not understand what the engineers where thinking about. When I'm wearing an extra sensor especially ment to do HR measurements why do you trust the fault-prone OHR sensor.

    Automatic switching would in my opinion be OK if you get clearly erractic data e.g. a straight line.

    Meanwhile i did some more round - one was fine - OHR locks pretty fast:

    But today I saw an behavior which I never had bevore - OHR measures a higher HF than HRM Pro:

  • Hi there.

    I had the same issues, took me some time to figure out a solution. Here is a way, which brings me back my correct heart rate (comparable with my HRM chest band) and the watch faces:

    1. downgraded my Forerunner 955 solar to 15.19 (from here: https://forums.garmin.com/sports-fitness/running-multisport/f/forerunner-955-series/397509/zip-file-firmware-17-xx-or-18-xx)

    2. updated to 16.17 (here still i was not able to install any watch faces, evel with old apks)

    3. updated to 19.18 (here I can install watch faces).

    under 19.18 I have the following sensor FWs

    Update from 16.17 to 19.18.
    WHR 16.01.02  
    ANT 28.15
    TSC 1.14
    Sensor 30.20

    Under 16.17 I had the following Sensor version reports, which was also ok, but I could not install watch faces.

    WHR 13.00.04
    ANT 28.05
    TSC 1.14
    Sensor Hub 25.63

    I explicitly skpped the installation of anything 17. or 18. or 20. or 21.

    Version 19.18 works fine for me, in running, cycling, walking and heart rate absolute values and dynamics are comparable to that what my garmin chest band reports.

    hope that helps anyone,

    73

  • Hi there.

    I had the same issues, took me some time to figure out a solution. Here is a way, which brings me back my correct heart rate (comparable with my HRM chest band) and the watch faces:

    1. downgraded my Forerunner 955 solar to 15.19 (from here: https://forums.garmin.com/sports-fitness/running-multisport/f/forerunner-955-series/397509/zip-file-firmware-17-xx-or-18-xx)

    2. updated to 16.17 (here still i was not able to install any watch faces, evel with old apks)

    3. updated to 19.18 (here I can install watch faces).

    under 19.18 I have the following sensor FWs

    Update from 16.17 to 19.18.
    WHR 16.01.02  
    ANT 28.15
    TSC 1.14
    Sensor 30.20

    Under 16.17 I had the following Sensor version reports, which was also ok, but I could not install watch faces.

    WHR 13.00.04
    ANT 28.05
    TSC 1.14
    Sensor Hub 25.63

    I explicitly skpped the installation of anything 17. or 18. or 20. or 21.

    Version 19.18 works fine for me, in running, cycling, walking and heart rate absolute values and dynamics are comparable to that what my garmin chest band reports.

    hope that helps anyone,

    73

  • Hi , thank you for your input, but downgrading is no option for me.
    Despite the fact, that it is not officially supported and the files for downgrade may be from a not trustworthy source, I clearly want it to be fixed in the official firmware.

    Also my usecase is, that I usually use external HRM for sports (running, biking, indoor rowing), so I am not impacted so very much, but still I want the OHR at least to be reliable for walking, hiking and my seldom commute rides via bike.

  • Hi ,

    thank you for your input and the comparison charts between OHR and external HRM - and good to see, that it seems to be ok for you. Still, for me, I have had some experiences where the OHR was even not picking up my correct HR while walking.

    Like discussed via mail, I will start with examples from walking.

    This is the first walking from 08.12.2024 - the HR average is 70, which is just not correct at that pace below 10 min/km.
    Also please notice the big drop at around 4:00 - clearly the algorithm struggeling to find the right range.

    Second walk from 13.12.2024 - here with the altitude. I walk down a hill and my average pulse *might* be quite alright at 80 (although I think it was higher), but you can see odd jumps in the HR, where the algorithm struggles - see e.g. after 5:00, 8:20, 13:00, 21:30 and so on.

    So here it gets really interesting - it is a walk also from 13.12.2024. It is exactly the same route, but reverse - so walking the hill up. And average HR is *LOWER* than walking down... also again odd jumps and drops, e.g. at 12:00

    Last example for now, from 15.12.2024 - walking at around 10:00 min/km and the average HR is around 80 (too low). also a lot of non plausible jumps up and down when the algorithm struggles.

    Feel free to contact me for more examples and also you can have a look into my account and the activities.

  • I also want to share some examples from bike (commuting) with you - yes, it may be more tricky in general, but still you will see, that we have a problem here, as sometimes the OHR is not only off by a bit but *a lot*.
    And please believe me, I am nearly 45 and an experienced biker - I do *not* grab the handlebars with a lot of force, so my arms muscles are not very tense...

    OK, look at this one - I am commuting really fast and the OHR thinks, my HR is at around 80 for nearly the whole time. At one point the jump up and then the drop, but even 110 is clearly too low.

    Here we see the picture that I have reported before - OHR is way too low in the beginning, at 12:00 it jumps up and might be about right at this point.

    Kind of the same - too low at start, later about right - only that this time, it does not have one big jump up, but two smaller.

    Also kind of the same, too low at start and two jumps up, kind of right in the second half.

    Feel free to contact me for more examples and also you can have a look into my account and the activities.

  • Thanks for your examples. I will try if I can reproduce this using a chest strap during my next walking activities.

    But I want to share a theory I had before the 21.xx software was released. I had bigger trouble with the Wrist HRM Caching up when wearing longsleeve clothing. So just an idea - maybe the additional movement wenn the clothing gets in touch with the watch or the shade has a negative impact in this case.

  • Today I did a walking activity with HRM Pro

    Comparison between the two HR-Sources:

    HR (Chest-Strap) over pace:

    So the HR Readings of the HRM Pro seems OK, also the HR Reading is not very nice but in my case just about acceptable.

    I had some long sleeve cloth since it was cold Wink