Heart rate malfunction - still there up to 19.41

emporii said:

latter has always provided the right HR, approximately

How can it be right approximately? It’s either right or it’s not. 

emporii said:

The many comments about these troubles being inherent to optical HR sensors are merely fog around the fact something seriously went wrong with these new watches.

It is not. There are complaints about optical hr issues across all watches with OHR and all manufacturers since day one of OHR being available on wrist worn watches. To deny this is to trivialise all the issues that others have had in the past on other watches, still have on current watches, across all software and sensor updates. There are inherent issues with trying to determine heart rate from blood pulsing through small blood vessels below the skin and covered in varying amounts of subcutaneous tissue using a light source fitted to a watch that may or may not be worn correctly. 

OP clearly stated that he did not have these HR problems with forerunner watch before, so this is not OHR limitation.

You trying to dismiss all HR issues as OHR limitation is trivialising actual problems, especially after so many people reported exactly same issues in 2024 and those users DID NOT have those issues before.

You and Garmin are yet to explain why same users with same watches started having problems this year. What changed? User did not, watch did not. Software did.

The fact that those users did not have those problems before clearly shows that issues are not due to OHR limitation, otherwise they would have those problems since they bought the watch.

Also, Garmin acknowledged that there are issues with HR tracking after users kept reporting problems, and even replaced some watches because of it, which shows that some issues are even due to hardware fault.

Alright, maybe to clarify it to make it better understandable:

Each measurement carried out with sensors have, basicly, 3 types of uncertainties, meaning -every- measurement is by definition an approximation:
A. Its random error, e.g. each individual measurement sample can by coincidence deviate from the true heart rate, for reasons found all over the internet with other watches as well (e.g. sudden arm movement, light falling in, etc)
B. Systematic errors, causing the whole set of measurements to deviate from its true reference. E.g. simply because of the method itself (optical sensors) or systematic interferences that consistently influence the sampling (e.g. type of skin, position of body, etc). You can nicely see this illustrated when for example comparing measurements from your left and right arm.
c. Representative uncertainty: are you actually measuring something thats representative for the parameter you -think- you are measuring. An uncertainty you can not simply determine by statistics only. E.g. are the surroundings valid for a representative measurement, is the sensor  not (accidently) measuring cadens or looking at some (flawed) algoritm, etc.

Now look at the watches themselves: for years i've been running with the Forerunner 245. Its random error (A) is always within a 5 BPM, in any given weather or time of year, except for some rare individual outliers (>95%), the latter something statistically you'd expect with a signal-noise random distribution. The forerunner systematic error is also about 5 BPM, consistently with a negative (lower) offset compared to (hospital) reference measurements. Last, C, there has been no indication the measurement is flawed, given the high performance seen in the random & systematic errors without any deviating or weird subsets. Alltogether you therefor see a measurement uncertainty of give or take less than 4% for an average workout.

Now look at the new line of watches such as the Fenix Pro and Epix Pro. During the last few running cycles, the measurement remained almost staticly at around 120 or 150 for most part of the workout. Even when doing interval training. And sometimes, it just went up 30 BPM within a second or 2, e.g. after 20 minutes of workout, regardless of warming ups. Meaning that the Random error (A) is enormous: during intervals its error deviation (SQR2) is varying per workout but on average well over 25%. And worse, its systematic uncertainty/offset is huge too when you would plot a regression: there is a major offset because even after sudden rises, it rarely ever comes near the true HR. Again, causing well over 20 BPM offset. But the worse part, any comparison with reality and determination of uncertainty seems to be flawed anyway due to C: because of the weird sudden behavior of sudden drops or rises that afterwards remain almost static, it seems the watch is not really showing a representative measurement inc some kind of uncertainty, but likely something else: a weird algoritm, a cadence lock, a hardware glitch, or whatever. But not something that statistically meets the necceasry assumptions for correlation with the HR.

Consumer/business result: so now you have a 200 dollar watch with a < 5% HR measurement uncertainty, and a new line of watches over a 1000 dollar incorporating the best new technology with > 25% measurement uncertainty, if you can even call it an uncertainty. One way or the other, no matter how you look this, it makes it a very troublesome product, rendering most functionalities useless. Given that Garmin already acknowledged that this issue is raised extremely often all over the world, I'd be be very, very worried if I'd be management at Garmin, for a complete product line being returned world-wide and your once happy customers heading to competitors.

Last: Garmin support pointed me out to this website so I could keep the issue visible and track its dev process, before deciding to return the watch. I'd say that instead of practicing debating skills here, all efforts go to finding a solution of this very serious flaw in this product line. I'm now awaiting a Garmin response and new beta release so to test what errors remain.

F8 Solar 47 (red) v Epix Pro 47 (blue) showing WHR in an open water swim. Epix is fine (as has usually been the case) while the F8 is not (which is unusual). So I’m not immune to issues. 

I have few issues with WHR but happy to see the Epix (for me anyway) is still ok, particularly as tracking heart rate from the wrist in the swim poses additional challenges.

Then for a run this afternoon it could not have been closer between the Epix and the F8. BTW, the dips are something physiological. 

I did a long term test downgrading the firmware on my Epix pro 51mm from 19.41 to every subsequent official firmware. On each one I had a problem with incorrect OHR, falling to 35 beats in the middle of the training, at the beginning a much lower heart rate and then a sudden jump ... In the end, when I lowered the firmware to 15.77, the problems disappeared, at least on my watch. There was no herat rate crash.... I've been on that firmware for almost a month now and have no problems. Now, I don't have many new functions brought by the newer versions of the firmware, also it's hard to find an application, watchface, data field from the market that you can install because Garmin has a limitation that you need to have the latest version of the software, at least that's how it throws out the message. But at least I can train and have the correct metrics, and not have to constantly look at the cwatch to notice a drop in heart rate or something like that and have to stop and restart the exercise. That's my experience, unfortunately after 15.77 Garmin messed something up seriously

It's not just the Pro models. Epix Gen 2 watches also have faulty HR sensors. I no longer use my HR sensor at all anymore since the data is unreliable.

Are you just a troll? You obviously are unfamiliar with the issue and have not read the descriptions of the problem. It's not the usual uncertainties with wrist HR sensors (fit, movement). It's a bug where the sensor sometimes works, then doesn't, then does, with no change on the user's part. For example, last time I used mine (I  no longer do), it was while rowing on a machine. The sensor lagged my HR for several minutes, then it caught up for a few minutes, then it dropped down (from about 135 to under 100 and stayed there) with no change in the workout, on a workout where my heart rate was continuously above 130 after warmup. 

The patterns people experience are the sensor doesn't pick up the increasing heart rate after starting a workout, then after 10-15 minutes, it will suddenly jump up 30-40 bpm. That may be followed by sudden jumps to well above actual bpm, all with no change on the part of the wearer. This does not happen on all optical HR sensors. My wife's Apple Watch sensor works perfectly, with no lags and jumps. Nor did it do so on my Fenix 6 Pro or Vivoactive 3. Only on my Epix Gen 2, and if you bothered to check other model forums, you'll find similar complaints on many of Garmin's watches.

How do you downgrade software?

its happening since months after new updates, i have forerunner 255 1 year ago, first months was perfect the readings of the HR sensor, but after new firmwares, something messed up, yesterday i did a race and i didnt start very fast, but the sensor didnt take the values of the chest strap till 5 minutes or more... one year ago it was working really good with short sprints, not the same accuracy of the chest strap but near, now... only working in steady easy runs...

7017726 said:

You obviously are unfamiliar with the issue and have not read the descriptions of the problem

FYI, he’s well aware of the HR issues discussed on Garmin forums but ignores them. He repeats generic replies to sidetrack discussion or downplay concerns and goes silent when people insist he answers to actual question/argument about these HR issues.

Gluki said:

unfortunately after 15.77 Garmin messed something up seriously

Lewis said:

its happening since months after new updates, i have forerunner 255 1 year ago, first months was perfect the readings of the HR sensor, but after new firmwares, something messed up

Everyone who has been on these forums since last year knows that except Garmin and couple users on this forum.