18.20/19.18 HR bug still here

They were not adressing similar issues for several years... When i first reported the false measurements during simple hiking activitys with my fenix6 a while ago the support man's first question was, from where did i know that the watch did not measure correct. My answer that i had compared it to a professional heart rate monitor triggered him immediately to the response that i do not have to compare this watch with a medical device (in a way that i got the feeling that i was not the first person telling this and they have some standard responses).
I tried to explain him that nevertheless with what i am comparing to the difference is way too much. He did not even think about helping me with this problem (it would have been so satisfying to just hear that he got the problem and will report it to the development...)

At least he told me that i should throw away the watch if i am not satisfied....
I was hoping that with the switchover to the forerunner 965 and its more modern HR-sensor this problem should be fixed. It was not fixed, nothing changed - ist just became worse.

Vizi said:

However, nothing was done to fix it until now.

What are they doing to fix it?

As I said, nothing.

They are still researching and collecting evidence, which they have been soing for a long time but did not fix the problem. Thus, I am not conviced they actually doing anything. Instead, they are trying to calm down users and buy time until next new watch is released.

They let the "old" watches to their destiny with unsatisfied users to start a new adventure with a new watch and new unsatisfied users 

Ohh you said "until now" so I assumed something was being done. Like I said, put pressure on the reviewers...DC Rainmaker, Desfit, Quantified Scientist, etc... to review the algorithm post initial launch. Quantified Scientist did it with Whoop 4 when they improved algorithm. If he hears enough people saying that Garmin regressed on a lot of their most expensive watches, maybe he will review to see if they actually got worse.

Locking on is normal for most WHR sensors, even Apple which probably has the best algorithm in the business takes time as well, same with Fitbit. As the reviewers state, Garmin, Polar, etc... relies more on a quality sensor, while Apple, Google, etc.. rely more on Machine Learning algorithms/aka AI. I'm more concerned with a regression in their WHR sensing rather than it always being an issue. As many reviewers and users have not seen it as always an issue, but end users have seen regression from something that was working fine. End Users need answers, Reviewers need pressure to review regression instead of the next new watch.

please have a look on my previous postings with graphics. They show repeatedly that they are primarily messing up the data for the first 10 minutes and then they are working mostly as expected. In the graph i posted yesterday the difference at the starting point was already there after just starting the recording without doing any movements - so there was nothing to lock on.

I did comparisons with the FR 965 between my Polar H10 and Verity Sense and both Polar showed up without any problems. Measurement with the Verity Sense had been on the upper arm. I'll make a comparison while wearing on the forearm in the next days, we'll see. :)

Saganist said:

Locking on is normal for most WHR sensors

We are not talking about locking on here, it is normal that it takes up to 30 seconds. But after some software updates, locking on proper HR takes 10, 20 minutes. That is not normal, and that was not the case 12 months ago for same watches (fenix 7, epix, forerunner 255,265,955,965).

Saganist said:

put pressure on the reviewers...DC Rainmaker, Desfit, Quantified Scientist, etc... to review the algorithm post initial launch

Interesting idea to indirectly push Garmin to take this problem seriously, but I do not see it happening since there is no money in reviewing what was already reviewed. That is why there are only reviews of new things, no matter how significant or insignificant they are. 

The more fair comparison would be to another WHR sensor on the wrist. We all know WHR isn't as good as a chest strap in most conditions, nor an optical sensor on the arm without tendons etc.. to deal with. We also know that they always take some time to lock onto HR during activates, but usually a few minutes, not 10. Even past reviews of Garmin watches show they lock on after a few minutes. I haven't seen anything about Garmins algorithm always being broken, only that it's always been slightly above average, and that it's severaly regressed with a certain software or sensor firmware update on A LOT of it's best selling and most popular watches.

Basically if they can't beat out a $100 dollar FitBit, or a $200 Apple Watch SE, then it's a real problem that reviewers need to address.