Sense heart rate monitor is not accurate

[blenderbender]

I've been feeling the heart rate monitor is way off on my Sense and just proved it today pitting it against my Polar chest strap monitor. It's way off (low) over 15% on the way up and though the difference is less on the way down it's still significant. Is this usual and is it anything that can be addressed? I should have known, paying more for all the silly bells and whistles and the most important feature isn't even close to being accurate. Who gives a flying ____ about nighttime skin temperature.

Moderator edit: updated subject for clarity.

[hawk.welder]

I might try a cheap one.

[gpk111]

Giampi, Very nice work. Thanks for sharing. I'm doing similar work using 2 of the samne devices to see what the variance is. Some technical questions if you don't mind: 

1. How did you post your graph on this forum? I don't thsee the ability to attch anything when I post. 

2. How did you collect (download) that detailed data? Fitbit dashboards are dashboards that remind me of idiot lights in a car. I use the squash site.

3. How did you syncronize the data sources? My Versa 2's do not sample the heart rate consistently. The sampling period varies from 5 to 15 seconds, so when you compare two devices, there is always a skew that has to be adjusted. I use spreadsheets, so it's labor intensiv.

Thanks!

Gerry 

[Rich_Laue]

 

@gpk111 this forum doesn't let people with only a few posts add photos. Eventually you will see this icon when posting. 

[gpk111]

I have a PhD in math and have an objective evaluation of the accuracy and responsiveness of the versa 2 hr monitoring technology.

What's this censoring and posting limitation all about? This is not China!!

Sent from Samsung Galaxy smartphone.
Get Outlook for Android<>

[Dopovo]

Hello @gpk111 ,

@Giampi71 seems unfortunately no longer active in this forum (he was very helpful to many users) so I'm trying to answer to your questions:

1) as@Rich_Laue said option will be enabled as soon as you reach some post counter

2) it is possible to download the tcx file with GPS and HR points only when you tracked a session. Bare in mind that the tcx is correctly formatted only if GPS is used (it's a many years old bug from Fitbit)

3) when the session is tracked HR point are every second and so it us possible to sync several sources using services like mygpsfile.com or DC Rainmaker Analyzer (that is the one used by Giampi if I'm not wrong)

[gpk111]

Dopovo, thanks for jumping in. 

Before I dive in and research your guidance,, i wonder if I would go down a rabbit hole with a Versa 2. Is detailed session tracking and the ability to download tracking files even available on that model?

Thanks

[DaveDi]

I finally sent my Sense in under the warranty after a couple of hour+ long workouts in a row where it was reading more than 30 BPM low. At several points, my Life Fitness chest strap was (accurately) reporting 130-140 BPM while the Sense was reporting under 100 BPM. The customer support rep's response was that because they use different technology, they might not show the same number.

It's been almost 3 weeks since I sent it in, and I know they have it, but I haven't been contacted by anyone. I'm curious as to what they're eventually going to do.

[gpk111]

Dave,

I've been tracking my Versa 2 HR accuracy for a while. Comparisons are against another Versa 2 and the built in sensors on the commercial treadmill. I hope to post more more detailed observations soon. Bottom line: My Versa 2 is extremely accurate at rest. It is somewhat useful when exercising, but very unreliable in observing high intensity activities. 

- 160 BPM could show 120 BPM. Finger pulse and treadmill sensors could be 160 BPM while watch shows 120 BPM. Not always, but display never goes higher than say 145, while data does.

- Data downloads show higher heart rates, but I believe it rarely recognized anything over 150 BPM. I download sampling every 5-10 seconds, but I'm hoping to get leads from DOPOVO on the possibility of 1 second sampling. That may make a difference.

Gerry     

[Rufous44]

@gpk111 oh yes, it's definitely possibly. TCX file can be downloaded from Fitbit app regardless the tracker used (as i said providing that gps has been used to track the sport session, typically run, biking or hiking)

[Dopovo]

@gpk111 to further clarify the 1 second sampling occurs when you start a session via exercise app on your tracker. Also, only in this case, in the Fitbit app, under the session you recorded, you can download the file TCX to analyze.

The precision of gps doesn't matter but it needs to exist. So if your plan is to check in alternative way the HR track you can start a run session  while doing any activity, technically even reading, resting, doing housework and so on.

[gpk111]

Dopovo...thanks much.

My exercise regimen is centered around treadmills ( knee preservation!) . I don't even know if my versa 2 has gps. Thoughts before i dive in?

Thx again.

Sent from Samsung Galaxy smartphone.
Get Outlook for Android<>

[gpk111]

Here is part 1 (of 2) to describe my accuracy investigation of (in my case) Versa 2. I actually have 2 Versa 2 units, so I wore them both during HIIT. One on each wrist. Not bad, but not very useful to check instantaneous heartrates reliably.

Note gaps in the readings. Sometimes there are no readings. 

The grapph at the borrom shows the % difference of the two units.        

[gpk111]

This is part 2 (of 2) my HR accuracy analysis. Data is based on HR downloads from the squash site, not the session downloads from Fitbit mentioned on this thread. The squash site acctually extracts its data from the fitbit database.

The graph shows that sampling rates vary from 1-3 seconds at very high HRs (say 150-180) to 5-15 seconds at "normal" HRs. The concept and design seems right, but instantaneous accuracy probably suffers because of movement and skin contact.   

[Darrellwrites]

It seems that we have some excuses, but I haven't seen a fix. Whoop has me with 141 hr. Sense has me below zone. My labored breathing tells me that Whoop is a heck of a lot closer. Anybody seen a fix or is that too much to hope for? I would use Whoop exclusively, but that's device can't attach to My Fitness Pal and others.

[Thook]

The data is so bad on the device I stopped using it. If heart rate is so far off how can any data be trusted?

[ScottIowa]

I agree my Sense could not be trusted. I stopped using it and bought an Apple Watch. Much happier and does what I need. Fitbit continues to sell a defective product.

Scott

[MarkMM]

No, they don't continue to sell it. Sense and Versa 3 have been removed from the Fitbit store. 

[FabriceB]

This watch is absolute rubbish for what regards heart measurement accuracy. Perfect example is two running sessions a few days apart
on the same track and distance:

. 7 km at 5'50 / km, average BPM: 153
. 7 km at 5'20 / km, average BPM: 155

And today I ran same track at super easy tempo for me (could have easily spoken during the whole run):

Average pace 6'19 km, crazy high average of 146 BPM and a peak at 171! My
calculated max heart rate is 164...

Useless...

[Cdeme123]

A few things I have noticed with my Sense.  Different exercises produce different levels of HR accuracy.  When on a stair climber it can be fairly accurate when comparing it to the HR monitor on the machine.  Also, I noticed that if I slightly wet my wrist under the tracker and make sure it's fairly snug, that usually helps.

However, exercises like HIIT training where my arms are flailing about, the displayed HR is WAY OFF.  Usually way too low.  What I found interesting was that even though my pulse was 155, the Sense was reading 85 the Sense still said "In Zone."  That implies to me that it is tracking a higher HR but the software is not calculating HR correctly.  That would be good news if Fitbit had any interest in fixing this problem...but they don't.

[Cdeme123]

I see this post is pretty old.  I was struggling with the same issue but then discovered that if I move the Sense up my are an in inch or two, then snug it down...the results are much improved. When not exercising, it seems pretty accurate in a normal position (about 1 finger width up from your wrist bone).  When exercising, I move it about 2-3 finger widths up from the wrist bone.  The difference for me as been night and day.

Cheers,