Wrist-based heart rate reliability worse since new updates

anic17 22 days ago

Since one month ago or so, my Garmin Forerunner 965 has been tracking my heart rate deceptively bad, considering its price, both in running; have had cadence locks for 20 minutes, normal HR values and when pausing it goes back to cadence lock and surprisingly, for indoor cycling, where there's no arm movement at all.

This seems to happen more when I start the activity with a heart rate more elevated than usual, and then reports around 115-125 when the HR as checked by a chest strap is around 165-170.

I am aware that wrist-based heart rate is not the most accurate thing, but it is honestly too bad for the price I paid for it and considering that HR is the base metric for all other estimations and data (calories, VO2Max, training load, endurance score, RHR, HRV, stress, intensity minutes).

Any fixes, to recalibrate the sensor (apart from using a chest strap, which isn't always available)? Removing the watch and putting it again does not work.

Top Replies

5767025 17 days ago +1

I have also seen very poor heart rate accuracy recently (on my Forerunner 955, but I assume the same HR software as the 965). At first I put this down to the colder weather, but it wasn't so bad in previous…

All Replies

0 flowstate 12 days ago in reply to flocsy🤠

Interesting chart. The real question is what does the "(activity) heart rate" chart look like for the 1st 12 minutes? In other words, we can see that the wrist HR data was bad, but did the watch choose to use the wrist HR or external HR for that period? (I'm not implying that it has to be one or the other for the entire period, of course.)

I ask because you obviously have dynamic source switching enabled (otherwise you wouldn't get "external heart rate" and "wrist heart rate" in your FIT file), and in my experience, the watch will sometimes choose wrist HR for short periods during an interval workout when the wrist HR data is objectively worse than the external HR.

I've never seen such a huge deviation between wrist and external HR though (both in terms of magnitude and length of time).

Maybe the problem is that it was cold outside and you weren't warmed up?

0 flocsy🤠 12 days ago in reply to flowstate

The heart rate displayed and recorded was from the external sensor for the whole time of the activity.

It was a bit cold when we arrived 30 minutes before the start, but it was pretty nice weather when we started, and we did warm up as well.

0 anic17 12 days ago in reply to flocsy🤠

I usually get graphics similar to those, and same thing happens to most of my tempo/threshold runs, it gets stuck at around 130 bpm and then suddenly jumps to 180 bpm as expected. Tried everything: taking the watch out for a couple minutes, switching it to the other wrist, cleaning the HR sensor, swinging my arm, and starts reading actual data when it feels like it.

0 flocsy🤠 11 days ago in reply to anic17

Was it always like that or only in the last few months, since this or the previous firmware update?

0 Woody 11 days ago

interesting to see - almost a year ago (or two years?) there was a similar discussion about worse wrist based hr after an update, that was never fixed. So it becomes even worse with every iteration... or there are only new people recognizing how bad whr ist?

0 flocsy🤠 10 days ago in reply to Woody

I think both are true. Many other things become worse, they add usually more bugs than useful features, so IMHO most people would be better if they wouldn't try to add features and did only bug fixes

0 Lewis 10 days ago

Reboot the watch, go to system menu to about menu click 8 times light button and do "reinstall software" it will reinstall the firmware, and give a try, if this doesn't work, make a backup and reset to default, I have this experience with some Garmin watches and resetting to defaults solves all normally, remember to wear the watch about 2cm wrist bone and tight

0 anic17 10 days ago in reply to 3902835

Get the same issues when doing gym, got a 33 bpm low peak for some reason (I was about 100 bpm). It has gotten so unreliable that it almost seems a random number generator just like Strava's estirmated power

0 RockBastard 9 days ago in reply to anic17

To me, it seems to be a temperature thing. Running indoors on a treadmill is spot on. However, running outdoors shows a slight discrepancy between my HRM Strap and the watch, with a 4-beat lower average heart rate on the watch.
In the pictures, the differences between running at 2 °C and indoors are shown.

0 Woody 9 days ago in reply to RockBastard

it would be easier to compare if you switch "single y-axis" on.

But nevertheless: In your first graph you can see that it took the typical 5 to 10 minutes to get in sync.
When it's cold i am not even able to use the wrh when hiking - it will lock to my cadence most of the time (regardless of the position of the watch on the wrist). Even when i am shopping outside in the winter i know that every calculation the watch is doing is wrong because of the erraneous hr recorded (or in other words: almost every feature of the watch is useless because of its huge deviations from the reality).