Heart rate inaccurate on Fitbit Sense

[Cameronoates]

Hey 

is anyone else experiencing huge heart rate inaccuracy during exercise with the sense? 

During exercise with the sense it won’t ever get above 120bpm for me 

but if I use my charge 4 during the EXACT same exercise my heart rate is between 145-170bpm in the cardio zone 

It’s extremely disappointing and I don’t want to have to switch to my charge 4 everytime I exercise to ensure accurate heart rate 

defeats the purpose of having the sense 

seriously one of the most frustrating experiences I have ever had.

Moderator Edit: Clarified subject

[Guy_]

Hi @MikeyP132 - This seems a common problem - have a look on the forum for other similar reports. With so many cases it is unlikely to be a specific watch problem.

[opiatedsherpa]

@SunsetRunner I, like many people in this forum, have been having issues with the Sense heart rate sensor.  When I tried using the Fitbit Customer Support bot, all it did was suggest to me was that reset method of holding down the button for 10 seconds to reset it.  I tried it a number of times and had no success, the inconsistency was still with my device.  Seeing as though I needed to charge my device the other night, I was compelled to give it a reset before the charge thinking that it certainly wouldn't hurt.  Lo and behold, I was gaining Active Zone minutes with my exercise and the heart rate monitor wasn't all over the place.

The role of the SimpleClockDuo was that it provides a handy battery charging screen, which actually tells you to a more precise time, when the battery is fully charged.  I've had my Sense since November 2020 and I would say that I have been guilty of taking the Sense off the charger too early, likely hurting the battery's efficiency.  

This was not a joke.  I wanted to pass my find out to everyone, hoping that it can help more people.

[opiatedsherpa]

@SunsetRunner I haven't tested it too much yet, only the one day, so my hesitance to say it's an absolute fix comes from that.  I did it two nights ago, tested it out yesterday, saw some good results, so I thought I would share.

[SunsetRunner]

@UnguruBulan, @NazH here is something which may be intersting, but only one data point.

1. This is a ride taken last week where I experienced the HR problems I have described, that is, I sprint and my HR drops. This ride (like all but one of my rides) was done with GPS and auto figure out exercise on for biking. This pattern of HR dropping (and all the HR dependant calcs) is typical:

GPS, auto detect on

Yesterday I took the same ride/route but without GPS and I manually turned off the auto detect setting as per Naz's post:

This is more in line with what I would expect, HR rises and stays high, does not dip to 'sitting at desk' rate. Yesterday's ride was in the rain with 15kn winds gusting 25. I also denied Strava the data (I carried my cell phone with me, which I do not want to do and is a major factor in purchasing a wearable). The Strava data was in line with my pre-fitbit data (where as my Strava/Fitbit data consistently shows I'm a slacker, not pushing myself). Strava knows I am on a 34lb MTB, fit assumes I am on a 16lb road bike (for calories burned, it is close to my strava data if I take a MTB ride, and afterwards edit the bike from MTB to Road bike).

About an hour after I finished I got a notification stating fitbit has figured out my activity was biking and logged it as such.

I am still a little circumspect of the accuracy as I was pushing the highest gear pretty hard due to the weather and sprinted in highest gear (HR dips each time on second graph) I would expect a few peak minutes... but without carrying a pulse ox (which I could do) there is no way to know.

One data point, but, encouraging. I will continue to do my 30-50 minutes bikes this way (don't have to wait for GPS).

Good call Naz, thanks,

Mattie

[SunsetRunner]

Thank you for your detailed explanation, @opiatedsherpa .

One or two tries certainly fail to be convincing enough. It could be just coincidence. Do it successfully 20 times in a row and it will be meaningful. But thanks for sharing your first result with us.

For the past year I had many days when the heart rates looked about right and, as a consequence of that, the Active Zone Minutes too. But also many other days when the heart rates and AZM were clearly rubbish.

There is "looks like it is right" and "it is right". For months I could only say the first phrase, always subjective, sometimes unknowingly deceptive. Until I got a Polar H10 chest strap, which has already proven to provide very accurate heart rate readings. After taking readings with both devices at the same time for several weeks, it became clear that my Sense heart rate inaccuracies were even worse than I thought.

If you haven't done so yet, read the several @t.parker 's great posts in this thread. His graphs speak for themselves

In the end, it all depends on how much accuracy each one requires his/her Sense to provide, on how lucky you are with the quality of the Sense you bought and, I believe, also on how well your Sense fits your wrist and the way your heart works or you make it work. It may work fine enough for some, not for others. But reading long time Fitbit users posts, it seems clear that Sense (and maybe Versa 3 too, they share the same HR system) seems the worst, or one of the worst, Fitbit device on heart rate monitoring.

[opiatedsherpa]

@SunsetRunner I agree.  It does need more testing to consider it a success.  I just thought I would share my experience.

[Tgisele]

@SunsetRunner Nice to see you back.

[Irish57]

👏👏

[Irish57]

Hoping Miguel you are feeling stronger each day! I’ve had some medical
issues as well, COVID despite vaccination and booster and serious cervical
spine issues. Recovered from COVID!👏 Under going treatments for my spine.
Hope to be stronger soon!
You take care, the Sense is a benefit in many ways, I just wish while
exercising I did not look at in dismay! I’m sure we’ll chat more!
Prayers!
Beth

[UnguruBulan]

@SunsetRunner glad you found some hopeful recipe to the HR issues, I admit neither yours nor the one posted earlier by @opiatedsherpa make any sense to me, but I understand that out of never ending frustration we are willing to try everything 🙂

I think I am on that "looks like it is right" category for now per @SunsetRunner  definitions, with only owning a Sense for 2 weeks I'd expect things may unfortunately "go south" in the future. I didn't mess with any setting yet, I kept all the defaults and considering I don't take all those provided stats too serious and I didn't see crazy unexpected HR jumps/dips, I can say for now that what my Sense reports to me is "alright".

Today I compared the HR provided by Sense with the one reported by my several years old Livestrong LS 15.0T Treadmill, right after my 40 mins running exercise, it was during the "cooldown" brisk walk time when I could hold the treadmill's handles to get a HR reading; and they did match (the pulse was a bit over 120).

Edit: speaking of which (my treadmill) I realized now it includes a wireless Polar heart rate monitor (a chest belt, Polar T34), which I never used. And they say Polar wireless products are 99.9% accurate I should probably see if I can find where I stored it, and give it a try if its battery is still ok (which I doubt it is, as I have this treadmill for about 8 years or so)

[SunsetRunner]

Thank you, @Irish57 . I wish you get well soon. It's great that you recovered from Covid!!! Concerning Sense, I agree with you. Maybe you get a Polar chest or arm band for Christmas, for your workouts. You would like it, believe me. I love mine, it leaves me much more relaxed by knowing exactly how my heart is doing at every moment. It's a pity that my Sense, supposedly the top fitness and priced Fitbit product, is so far from that. Compared to Sense, the Polar bands have a much more reasonable price and you can follow the heart rate on your smartphone, not having to buy another watch just for that.

@UnguruBulan , it seems to me that you have the right attitude towards your Sense heart rate: don't take it too seriously.* And maybe you are one of the lucky ones whose Sense works fine for you. The most unhappy users are the ones (like myself) who take seriously the values given by their Sense, either on workouts to keep fit, or for health issues, or both. And, of course, the high expectations of anyone who buys the most expensive Fitbit watch, taking into account Fitbit's previous great record with fitness devices. Good luck with your Sense!

* Edit: that was my general attitude with fitness before starting having some heart issues, which make me be more careful.

[CYM914]

My inspire 2 massively undercounts strength training too. Today my polar chest strap was reading 130-150 BPM for my 20 minute workout and my Fitbit didn’t get over 80. This was after a 20 min cycling session where the Fitbit and the polar were pretty well aligned. 

[SunsetRunner]

There is a huge (20+ page) thread on this topic and it is one of the most active. However, FitBit lists it as 'solved'.

I think we should all move over to this thread and stop posting in the solved thread.

@Jirvine, I am in a similar boat to you. I received a VA 4 yesterday, just starting to play with it (did not have correct settings for first night's sleep and can't seem to download clock faces). I am still in return window for Sense (but I have been wearing it for 3 weeks, feel a little guilty, but, see my last post in other thread which shows my graphs).

Long story short. I turned off my auto-detect biking and did not use activity (no GPS) and it seems more accurate.

I will be wearing both watches (one on each wrist) and when I learn a little more about watchB (today) I will report back.... having biked with both (no GPS on Sense, but, if I see consistency across 3+ rides will try the GPS again). I think this is a software issue (long story).

Good Luck,

Mattie.

[UnguruBulan]

@SunsetRunner that's what I was thinking too, when finally been convinced to go for a tracker: get the "latest and greatest" Fitbit, which at this moment is Sense. It can't go wrong, right? 🙂 I was thinking, especially that I was already familiar with the company's good reputation.

I read about all those forum complaints especially the terrible HR capability, only after I got the device. I was like others, unpleasantly surprised to read all that.

I'll keep the Sense though, I could already see some funny stats it reports, a notable one for me being the elevation tracking, which has significant variations across exercise instances using the exact same outdoor route.

Like you said, I should start being more considerate with the heart reports, so far I never really cared. And I'm 50 now! I should start giving that the attention it needs.

[SunsetRunner]

My theory of not turning on exercise (GPS off) and changing the setting to not auto-detect biking appears at best to be flawed, at worst doo doo.

I rode today with two wrist worn devices (one fitbit, one not) and while the fitbit did more accurately track my HR (I did not drop to sitting at desk) it was, for large portions of the ride far away from the other device... appears to rise and fall more slowly than the other device.

However, on today's 40 minute ride, both devices showed similar range (within a few BPM)... it just took fitbit much longer to rise to 130 (like 10 minutes longer) and when it finally hit 130 the other device was at 142 and had been up there for several minutes. Both topped out in the low 150s (which I believe to be relatively accurate). Only second data point and first with two devices. Will post better detail/graphs when I have more. I screwed up a setting on new device which shows accurate zones + minutes but doesn't credit them due to the 'automatically determine my zones' as opposed to using the HR zones calculated (like FitBit). So, I can see a graph corresponding to the GPS data (and I carried my phone again, so I have the Strava data). But despite it showing 30+ minutes in 2x zones and 7 in 1x zone it only credited 20 minutes of 1x.

I chatted with a live human and she quickly saw the error of ways and guided me through the fix. Entire round trip, 10 minutes.

Interestingly, Strava and other device had similar calorie burn numbers (447 vs 453) but Strava's was overwritten by FitBit (which thinks I am on a 16lb road bike and not a 34 lb MTB) and came in at 220 something. My independent calcs using Men's Health, Cycling etc. guidelines put at me at 45-60 per mile while FitBit has me at 28-35... up to 40 if I hit a high HR.

I suspect the higher correlation is related to the 40 minute vs. 30 minute ride.... but again, I did not dip into the 'sitting at desk' HR range for the second time.

mattie

PS - I think we should move our discussion to the other thread which had some activity today because it is not 'solved'. I posted there earlier today.

[UnguruBulan]

Reading this thread makes me agitated and causes my heart to race .... and my Sense promptly reports me that. Looks like not accurately enough though. Comparing results with the mighty Polar usually ends in disappointment. And unfortunately Polar seems the right one, as its accuracy is 99.9%

I hope Sense just has some sort of a software bug that will be addressed sooner rather than later.

[SunsetRunner]

---

@UnguruBulan wrote:

I hope Sense just has some sort of a software bug that will be addressed sooner rather than later.

Agreed, given that in my two 'no GPS/Auto Detect Biking' the results are more accurate (and I still think they are using data smoothing logic/linear regression, as evidenced by the multitude of awesome graphs provided by @t.parker  in the other thread, I too think it is a software issue). I'm guessing this adversely affects the results (HR's can change quickly under HIIT 'like' conditions and I do not think FB captures that). But their target market may prefer that.

Given that I've been wearing mine for three weeks I would prefer to give it to my daughter than return for reasons of ethics. I am certainly not moving forward with it (for me) for a multitude of reasons even if the HR is corrected. There are things I will miss, but, as a three week user, I am not addicted to. My daughter is a multi-year FitBit user and wouldn't have it any other way. She has a Versa Lite now and would like an upgrade. She is extremely active, X-Country, dances 5-6 days per week.

Gotta run, or, walk, just got the wrist vibration, and, gotta drop someone off at dance....

Mattie

---

[IgorR11]

I've been doing tests and I leave my smartphone with the bluetooth turned off while I'm with it on a walk, and unfortunately I disable the GPS most of the time, this seems to improve the accuracy a little but I don't think it's enough.
What I've been thinking is that it's too many functions for a device limited in ram memory or whatever they use for applications to run. It seems that the device cannot process and there are bugs in these reads.

[t.parker]

@SunsetRunner whether GPS is on or off, whether the activity is tracked manually or automatically, and whether you going to stand on one leg in full moon starting tracking the exercise - that doesn't really matter. I know that it's human nature to look for a solution but the truth is there is no solution or at least nothing that users can do.

Here's another analysis.

Fitbit Sense

PolarOH1 (optical)

PolarH10 (chest strap)

Now all together:

PolarH10 and PolarOH1 are pretty much the same (I expect slightly less accuracy from PolarOH1, worn on my forearm). Sense has three areas of difficulty, however, if this was the kind of result I'm getting every time I would be more than happy. Sense for most of the time goes head to head with two other monitors. It has some difficulties during uphill sprints but that I can easily forgive. I'm impressed with this result.

Now, guess what I have done differently. Did I turn off GPS? Or worn it higher? Turned off notifications? Tighter? None of it. I just slapped it on my wrist, as usual, started Exercise->Run, and went running. No special treatment. It just worked brilliantly! This shows, that there are times Sense does a great job and the user doesn't have to do anything special. If that's the case, however, then it also means that doing anything special is not a cure. This is probably the reason why it cannot be fixed. Probably, even Fitbit has no idea what's going on. It could be millions of things that cause the issue, maybe it's the environmental thing, maybe it's skin, skin color, sweat, maybe body temperature - could be anything that changes the way the sensor works. We may be guessing and the watch anyway will do whatever it wants. The solution, if there is any, is way beyond users' abilities and looks like it's way beyond the abilities of Fitbit engineers.

[TheBugPeach]

I'm having the same issue. Ive had charge 2,3,4 in the past and changed to the sense and the workout calories,  HR are no where near what used to be recorded.