Fitbit Sense heart rate is way off

[MarkLiberman]

Just got a new Sense. When I exercise to raise my heart rate, 3 independent monitoring methods all show (for example) a rate of 140-145 -- the sensor on the elliptical, the readout from my Orange Theory "Burn" device, and a fingertip blood oxygenation monitor. But the (apparently worthless) Sense says "89".

I've tried multiple times, different wrists, different placements -- essentially the same result.

In contrast, my Ionic was pretty good at this -- the only problem was that the battery basically died, would only last a few hours.  Overall, Fitbit seems to be making crappy devices these days.

Moderator edit: format

[BlueMountain]

David, I called support on Friday. The gentleman who I spoke to kept telling me that FitBit is confident in the algorithm and that FitBit has undergone several heart studies, and that yes it might be harder to achieve higher heart rates when working out, but that this was technically more accurate and correct than everything else. I feel like this is the worst possible answer that can be given. I have compared with two polar devices with and without chest strap, my FitBit iconic, and my treadmill which has built in sensors. More telling than anything is that I am sweating and breathing hard. I know the feeling. I know when my heart rate is elevated, and by about how much. But the FitBit seems to indicate very limited activity with my heart rate in the 80's. Given that your UI now stresses Heart Rate Zones for a measure of activity, I think this has become even more irritating and pronounced.

I suspect that your support people are correct that your algorithm is technically more correct and accurate than everything else, but that in some specific cases is is measuring incorrectly because of the pattern/form of the heart beat, or because of the dryness of skin, or some human form factor. I suspect that your team should consider making some of this configurable so that things can be tweaked to accommodate all these variations.  In any case, other than you here, I feel like FitBit is pretending there is no problem, and there very clearly is. This is bad faith business and turns off customers. Can you please pass along this message to the appropriate parties at FitBit/Google. I do believe this needs to be made the highest priority at your company since it seems to be damaging the brand and reputation of FitBit as a reliable trustable brand.

[t.parker]

@BlueMountain when I had Ionic, I experienced same issues. HR was always reported lower (20-30bpm) for high intensity activities, however, it was too high for low intensity (like walking). Even showing raw HR data and comparing it with three other devices for which the results aligned would not convince Fitbit that there is something wrong with their software or hardware. If you have three other monitors (including chest strap) showing similar results and only Fitbit shows something totally different then in the eyes of the Fitbit those three devices must be wrong 🙂

[Craigside]

Ok, I’ve been wearing my replacement Sense since last week, I’ve done 2 gym sessions with it Friday and Today.

I did my usual hour weight training, then HIIT cardio session on the elliptical trainer.

Disappointed but not surprised! The HR monitor is way off, indeed my HR didn’t go over 100 on the Sense when it should have shown over 140.

My interpretation of this problem is that;

1. The Sense HR sensor doesn’t work properly.

2. Support are not allowed to acknowledge the fact.

3. It’s hardware and can’t be fixed with a firmware upgrade.

This issue only affects the users who are looking for meaningful HR, and the range of other features look ok to me.

I’m going to return this Sense to the retailer and buy a Versa2, which should give meaningful HR, etc., as my Blaze does.

[Craigside]

Absolutely. I agree with you.

I am returning my 2nd new Sense to the retailer tomorrow for refund.  Unless I read In future that the issue has been fixed I won’t be buying another.

The poor support, aka “denial of the problem” compounds the issue and frankly is just damaging the reputation of Fitbit.

----------------

Absolutely, my experience as well. “Fitbit device is perfect and everything that doesn’t agree is therefore incorrect”...

[DavideFitbit]

Thank you very much for taking the time to share your feedback about the accuracy of the heart rate and your experience with the Customer Support team @BlueMountain @t.parker @Craigside. The team of developers has been informed and is aware of the reports from some users about the accuracy of this information during high intensity workouts and they're constantly working to improve the quality of products and services.  

Thanks again for sharing your experiences. 

[Craigside]

Well, I’ve just returned my latest new Sense to the retailer for full refund.

The HR recording doesn’t provide meaningful data, I strongly suspect it’s all Sense devices are the same as Support explained that this is ‘normal’ and the device is working properly.

The ongoing vague non-acknowledgement of this actual problem is not helping Fitbit’s reputation or customer confidence one iota. It’s a joke in my opinion.

Kind regards.

[pbrb]

I've noticed a similar issue with my sense while doing HIIT workouts. The sense will be relatively accurate (=/- 5-10 bpm) up until about 150-160 bpm, then tends to get wildly inaccurate, especially if I cross into the 170s. Once my HR gets into that range, the HR on my sense plummets to typically the 100-110s and is highly variable for a few minutes, then sorts itself out again usually once my HR falls back into the 150s. Its hugely frustrating that my sense fails to capture any of my peak HR intervals, especially when the cheap HR strap that came with my spin bike manages to do it flawlessly; this combined with the random freezes and restarts has my regretting switching from my whoop band.

[KCl94]

@pbrb It's interesting to see that even though the HR count is awful, the calorie count is pretty close to what we can see from your other device. 

@Craigside dang I am still waiting on my replacement Fitbit and your experience is making me doubt even bothering on giving it a shot. Did you ever compare the calorie count between devices?

[Craigside]

Hi

I don’t watch my calorie burn closely, but it reflects what I’m doing. And looking at my figures during Sense and Blaze, my calorie burn is higher on Blaze than Sense when working out, probably due to the HR being part of the calculation.

ie. Higher HR working out = more calorie burn?

Also, because the Sense HR data is only way off when doing intense workout (in my experience), most of the time it’s ok.

And, the time I spend going flat out doing HIIT is only around 20-30 mins of my 24hr day, which is only around 2%, therefore it has a minor effect on total calorie burn per day?

If you see what I mean.

Good luck.

[Mont-Sly]

I had the same problem when using my cross-country ski machine.  While exercising my heart rate never went over 100, however, when I was doing my recovery - no longer using my arms - my heart rate went over 140.

I contacted support yesterday, and one of the responses referred to movement of the watch during my workout.  Today, I placed the watch face under my wrist, 3 fingers below my wrist bone, and made the band snug.  Seemed to work great today throughout my workout.  It recorded a smooth upward trend to my heart rate, and appears to be very accurate.

[GeorgeK47]

I purchased this Fitbit Sense several weeks ago, and it has been hit and miss every day.  So disappointing.

The Sleep Score shows up, maybe or maybe not.  Some days it's all there with all data and the chards for Deep and REM, Light sleep and etc.  Other days nope, sorry, minimal data, no Score either.  SO FRUSTRATING.  

The Heart Rate is off, doesn't really work.  Just read this forum about all the inaccurate heart rates.  Tells you that it is not accurate.

And I'll tell you, FitBit top management:

After paying the high price for this Sense Fitbit, I'm really disappointed in its accuracy ... it appears to be mainly just a bunch of B.S. marketing hype on your part.

This Sense SHOULD JUST WORK!  Firmware updates should be automatic.  The Fitbit App should be updated automatic.

Why are we the customer trying to troubleshoot the problems with it?  Huh?  Give us an answer, if you can, okay?

We should not have to log out then log in to fix whatever seems to be wrong (according to some of your tech so-called "Help"  ..   

Ugh!  Makes me want to take a sledge hammer to this device!

[GeorgeK47]

Based on both the Heart Rate being inaccurate,

and the Sleep Score / statistics showing up half the time,

with no fault of mine (hear that Tech Support?),

I've decided to return my Fitbit back to Amazon for a refund.

Then I'll be looking into an Apple Watch series 6, 

which I've seen does work as advertised, from people who have it, 

having an accurate Heart Rate at higher levels,

and measuring your oxygen level anytime on demand,

unlike the Fitbit oxygen measurement, working only while sleeping, maybe every other day.

[Craigside]

Hi

I’ve found the Sense HR monitor is rubbish.

l’ve now got the Versa2. It has the previous version excellent working HR monitor.

It also does the blood oxygen while you sleep (with a certain clock face).

Unfortunately the Apple Watch has short battery life, which doesn’t work for me if you have to charge it everyday (yes literally).

Everything works great on the Versa2, it’s how a Fitbit should work!

[KCl94]

@GeorgeK47 @Craigside you guys may want to try the Samsung Active 2? I bought it about two weeks ago to challenge the Sense.

After a few days of testing and comparing the performance of the two, I went with the Active 2. I liked the Sense a lot (despite its flaws) but the Active 2 is just better! The HR was always consistent with my phone, StairMaster, and treadmills HR readings. The only thing I hate about Samsung is their lame community. It is so dead! 

I like the Fitbit environment/app so I may "downgrade" to a Charge 4 or an Inspire 2.

[slowdive]

Do you think it's s hardware or software problem?

I've been testing a few devices (smartwatches and smartbands) and by far sense is the worst for HR. Even mi band 5 (less that 30 euros) was more accurate than sense.

Today I tried sense and polar h10 strap in  a indor cycling class. Sense was terrible. 

Of course polar h10 will is always be more accurate. But comparing polar with other smartwatches/smartbands I get decent results not far away from polar h10. But with sense, jesus... It's really bad.

[Mont-Sly]

I was curious and went to look at the topics in the Versa 3 forum, and low and behold, the same problem appears to exist with the Versa 3 heart rate readings during exercise for some users.  As I've stated before, my personal view, is that it is a problem with the physical movement of the watch during exercise.  Could be the weight and thickness of the actual watch.  When I strap it securely higher on my arm it appears to give me better (seemingly accurate) results.  Though, again, I'm just speculating and it just my personal experience.  Though I would think this sort of thing would have been thoroughly tested before the new models were released.

[t.parker]

Accidentally, I became an owner of Fitbit Sense yesterday (by sharing Amazon account with my wife, long story). Well, I have some time to return it so decided to give it a go (who knows, maybe I will decide to keep it). I have done just one workout of 30 minutes. Spinning on the turbo trainer. It took me a while to extract the HR data from non-GPS activity but here it goes:

Red - Fitbit Sense

Green - Garmin Fenix 6 Pro + PolarH10 (I will test it against builtin wrist sensor later)

The session structure:

0-10min - trying to stay in the zone between 110-129bpm

10-15min - trying to stay in the zone between 129-147bpm

15-20min - trying to stay in the zone 147bpm-166bpm

Then wait to drop HR below 120bpm and pedal as fast as I can to raise HR above 160bpm-170bpm (each time slightly harder). Rest between intervals.

My conclusion:

Good:

• This session wasn't as bad as it could be. Moreover, the "big picture" result is kinda acceptable as for wrist based optical sensor. The avg HR is very close to the PolarH10. Good enough for things like estimating energy output.
• The response to more intense intervals is there, visible spikes where they should be.

Bad:

• Using Sense for training with HR is rather pointless, response to changes is way too slow
• The HR seems to be always slightly below the chest strap
• When getting into higher HR (high aerobic, anaerobic) Sense cannot get it right. Seems like the higher real HR, the lower Sense HR.

	Fitbit Sense	Garmin + PolarH10 (Chest strap)
AvgHR	134BPM	138BPM
MaxHR	164BPM	173BPM

The MaxHR difference is 9BPM. However, what is important is that in this particular workout 7% of data points (captured every second) stands for HR higher than 164BPM. Over two minutes of highest intensity has been capped. Looking at the graph, it is easy to see which peaks have been capped.

This result isn't terrible, though. It approximates calories pretty well (I believe, most of Fitbit users take special care about the calories) and shows areas of higher and lower intensity. However, the result isn't good enough to use it for training. If one uses HR zones in training then it's rather bad choice to use Sense. Moreover, so much advertised new PurePulse doesn't seem to be any better than old one. I must admit, I was expecting a lot worse result.

Today, I'm planning to test Sense running and elevate my HR above 170BPM.

[arturmeireles]

The Sense is much more stable during steady cardio but during a weights workout is totally useless.

Get Outlook for iOS

[t.parker]

@arturmeireles I'm gonna test that, too. The problem with weights training is that most of the wrist-worn sensors are useless. There are number of reasons for it but the major ones are:

- unstable placement of the watch (bending wrist affects readings a lot)

- muscle pump, especially when forearms are involved.

Why watch placement can affect readings is pretty easy to understand but muscle pump is a different story (and I don't think it's been mentioned here). Basically, when weight lifting, muscle pump effect is something that probably we all are looking forward to. After all, it pumped muscles - better pictures for social media, right? 😉 The watch however will struggle with reading accurate HR. When muscle fibers are relaxed, the blood can flow easily between them but when the fibers expand they start pinching on veins. Blood is pumped faster from your heart than it's able to go back. It may cause discomfort or even pain (rock climbers know it ;)). The problem is that it affects also HR sensor which is designed to measure the blood flow (it doesn't measure HR, you want HR - get chest strap). If the blood is forced to slow down and struggles to pass through veins this will cause not only lower HR readings but very often utter rubbish readings. The blood flow may be different in various parts of your body, too. That's why majority (or almost all?) watches struggle with HR during weight lifting. This is the reason why most of sports watch manufacturers recommend to use external HR monitors. Even my Garmin Fenix 6 Pro struggles during weight training or calisthenics (it just hates press-ups). I have never seen a watch (Suunto, Polar, Garmin and Fitbit) that could get HR right when I do rock climbing/bouldering - lots of movement, pumped forearms. That's why I use either PolarH10 or optical PolarOH1 but worn on my bicep (above forearm), less likely to be pumped during such activity. I will test lot more activities but I don't expect accuracy from wrist OHR for weights lifting, it's just barely possible.

[Mateos]

Hi T.Sparker, thank you for posting your results. I saw a similar pattern during a recent run were I was wearing both the Sense and an OH1 arm strap. Fitbit came back with an average of 123hbpm and the Polar strap reported 121. Max HR with the Sense 158 vs 168 with the Polar strap. 

I saw much bigger differences during weight training sessions. Where previously my Ionic managed to catch up with the OH1 strap, the Sense is simply  too laggy to keep pace with the OH1. In that respect the new HR technology of the Sense seems to be less reliable than the “old” sensor in the Ionic. Curious to see the results of your next training sessions.