Reduced heart rate readings during activities

Hi.

I've seen quite a few threads here regarding inaccurate hear rate readings, however, they don't seem to be describing exactly the weirdness that I'm observing.

General description: the watch (Venu 3, primary wearable) is reading way lower heart rate values during activities, when there is no activity being recorded on the watch.

What's happening: I do indoor cycling on Zwift using a smart trainer, I have Garmin HRM-DUAL heart rate chest strap paired with the trainer (this is where Zwift gets heart rate readings from), activity is recorded in Zwift and when finished riding Zwift exports the activity to Garmin Connect. Venu 3 is not used at all for the activity recording, it's just idling (in smartwatch mode). During the ride, the watch will consistently read far lower values than the values I see on Zwift from the chest strap.

For example, here's heart rate chart from an activity, that got imported from Zwift (i.e. readings from the chest strap):

And this is heart rate chart from day summary page on the day of the activity (i.e. readings from the watch):

Starts at ~88bpm, peaks at 111bpm and ends at 97bpm - way-way off from what the strap recorded. With sudden heart rate surge right after the activity...

Now, I can understand, that the watch probably has reduced sampling rate in "resting" mode (i.e. no activity being actively recorded) to save on the battery, so as an experiment I tried few times forcing the watch into "active" mode by recording simple "Cardio" workout at the same time as Zwift is recording the ride, and this results into matching data:

Heart rate from an activity from Zwift:

And from day summary page (obviously, some data aggregation happening here, but it's in the same ballpark):

I'm trying to understand whether there is any issue here with this heart rate discrepancy, could it lead to incorrect calculations of other metrics that rely on heart rate data...

I believe it does? For example, on the weeks when I do not use this "Cardio" workout "hack", I struggle to reach "intensity minutes" target. But doing the "hack" at least once easily fills the circle., and "vigorous minutes" are used in "fitness age" calculations...in fact, ever since I caved in and started Zwifting my fitness age value has returned to my real age (used to be 3-4 years lower). Wouldn't be surprised if VO2 Max estimations are also inadequate because of erroneous heart rate data...

I'm hesitant doing this "hack" all the time (even if immediately discarded when stopped recording), because now it could lead to sort of doubled amount of activity and throw calculations off again but to the other extremum?

Is there some setting I could tweak on the watch to have the watch realize I'm actually doing a workout so it needs to read heart rate more frequently without actually starting a workout on the watch?

Top Replies

All Replies

0 3931661 over 1 year ago

I just received my Venue 3 watch and noticed that when I do my Kettlebell swing exercise it only goes as high as 100 ( i know/feel is really low) compared to where my samsung watch 5 pro would reach 160. Very disappointed as I was under the impression Garmin had the best health tracking.

0 PRK91 over 1 year ago

When an activity is being performed (not wearing a chest strap), all the 6 LEDs on your watch are turned on for better accuracy. During all other non-activity periods, only 2 LEDs are turned on. This is done deliberately to save battery life and also it is not required to have such accuracy 24x7. However, if you want to sample a more accurate reading during the rest period, navigate to the HR widget from the watch face, once inside the HR widget, all 6 LEDs are turned on and you should be getting a more accurate reading.

0 eeyou over 1 year ago in reply to PRK91

PRK91 said:
However, if you want to sample a more accurate reading during the rest period, navigate to the HR widget from the watch face, once inside the HR widget, all 6 LEDs are turned on and you should be getting a more accurate reading.

Indeed, I observe described behavior. Thanks for this tip! Come to think of it, enabling HR broadcasting seems to do the same trick.

I will do a few rides to see if such approach solves the issue.

0 PRK91 over 1 year ago in reply to eeyou

You're welcome

I have not tried the HR broadcasting, but it should probably work. Great :)

0 Benoitgphoto over 1 year ago

I noticed something similar but when I record an activity. Recording a walking activity outside and sometimes the HR reading goes as low as 60-70 while there is no way that this being accurate at all. It can read low for a minute or so and then boum back to 120 ish where the reading makes more sense. Or I shake my wrist a bit and the reading slowly increases. I was thinking that is because of cold temperature outside but I am not sure. I don't know if my Venu 3 is defective or this is software related but indeed there is something wrong with the HR monitor. I did not notice anything similar on resting periods. The HR monitor (v4) on my wife's Vívoactive 5 looks more accurate. Where can we report this to Garmin?

0 eeyou over 1 year ago in reply to Benoitgphoto

You're experiencing phenomenon described here in "False Heart Rate Detection" chapter. I've had this happen many times during outdoor bike rides (with Venu 2 anyway) and indeed cold temperature can be the cause. As an idea, try wearing sweat wrist band over the watch (idea being you're protecting that wrist area from cold air) and see if that helps...

The issue I'm bringing up here is different, though. It's not that sensor occasionally misreads heart rate, instead it fails to recognize heart rate is increasing and it should switch into appropriate accuracy mode...

0 Benoitgphoto over 1 year ago in reply to eeyou

Thanks very much for the tips. I'll try this.

0 eeyou over 1 year ago in reply to PRK91

Broadcasting HR seems to be good workaround for the issue I'm observing.

I'm not positive opening HR widget keeps HR sensor in higher accuracy mode indefinitely until user closes the widget. HR broadcast, however, keeps all sensor lights on until you turn broadcasting off.

Guess I'll be enabling it from now on to have accurate data

Once again, thanks for nudging me into this idea!

Would be ideal, though, if Garmin improved HR change detection and toggled accuracy modes accordingly

0 2679107 over 1 year ago

Welcome to Garmin world!

Unfortunately your experience mimics mine perfectly. The sad truth of the matter is that in order to get HR readings that are anywhere near the actual values your Venu 3 needs to have started an activity (or at least start the HR monitor app). When I cycle without an activity started the venus readings starts off sort of okay only to get completely off after about 20-30 minutes of cycling. For me it normally stabilizes around 80-100bpm when the actual heartrate is 145-165 so it´s not just a little off, it´s off by a country mile and then some.

I have brought this up with Garmin support and after initially supplying them with some videos of the phenomenon (the first line support agreed with me that this couldn´t possibly be normal) I was eventually told that no, this is actually by design. I was also told the usual gibberish about the extra sensors and all and saving battery and that there´s no way they can change this and they actually seem to believe this is not even an issue. I would agree if we were talking 5-10 bpm off the real HR but it´s about 50% off which makes it borderline useless even for casual daily HR monitoring (for calorie counting and such)...

I have thought about trying to make an app that would try to autodetect cycling and indoor cycling activities and start activities when needed so that the watch isn´t actually a burden when training (one more device to fiddle with and turn on) but haven´t gotten around to it yet.

If it makes you feel any better the watch completely ignores cycling VO2max and the FTP is fixed at 200W anyway so it doesn´t really affect those metrics :)

0 eeyou over 1 year ago in reply to 2679107

Funny, yet the watch is perfectly capable of reading HR within like 1-2bpm compared to HRM-Dual HR meter strap, when the watch is forced into "high accuracy" reading state.

It's definitely not the accuracy issue, but recognition that HR is increasing and switching to appropriate accuracy level to continue reading correct values. And it very much sounds like something firmware/software would take care of, which I'd assume Garmin is developing.

"Saving battery" argument is also fun one. If I'm down to trade battery charge to have more accurate numbers - let me do it. Just like you let us choose between GNSS modes (GPS only, GPS+Galileo and other options), which affects battery life, why not let us do that with HR sensor...