How is Sense 2 heart rate accuracy while exercising?

[Rami.Khalawi]

Hi all,

To the owners of the new Fitbit sense 2, how is the heart rate accuracy mainly while exercising? Is there any improvement comparing to the previous model?

Thanks

Moderator Edit: Clarified subject

[crippstar]

Hi - I've noticed an improvement in the Sense 2 over the Sense. In the Sense, if the exercise involved holding anything - a handlebar, a weight, a rowing machine, etc. - the heart rate dropped to a ridiculously low level, like I was laying on a sofa or something! That seems to be fixed in the Sense 2, fortunately. 

[t.parker]

@crippstar I did lots of tests of Sense 2 and chest strap and came to opposite conclusion. Sense 2 is same bad as Sense. You can see rowing and weights here . Since I posted these results nothing changed, same bad HR when rowing and weight training.

[crippstar]

Hi - yeah that makes sense, any kind of tracker like a Fitbit will never be as accurate as a chest strap or proper heart rate monitor, but I did note an improvement over the Sense. It used to absolutely infuriate me when I was on the bike and looked down to see heart rate drop to <70 for no good reason and then jump back if I took my hand off the handlebar! 

[t.parker]

On a bike, I can't see the watch (drop bars so the screen isn't facing sky) but I have never gotten readings even remotely accurate on Sense and Sense 2 (also posted results on previous page). And I'm saying that not even once I had acceptable HR readings on bike. I see no difference between Sense and Sense 2 here.

[slackjaw99]

As a fellow owner of the same Polar chest strap, I was hopeful that, despite the fact it is just not possible to get much in the way of accurate heart activity while doing HIIT with a wrist-only device, the Sense 2 would at least be 'reliably' inaccurate. But, alas, it is not. 

[Tristannyc]

Hey question, how are your END results of the final stats like average BPM and Calories burned. Fitbit has a history of applying some algorith to make up for a lot of the missed spikes of heart rate in the end results to be within margin of error. Would be curious what you think on this generation or if it is just that far gone. The Fitbit Charge 5 for instance and Versa 1(more accurate, ain't that something) compared to an apple watch that locks onto heart rate like a homing device still have very similar end results.

   

[JerryHam]

I've found times where the average HR reported by Sense2 in the FitBit app was quite a bit higher than the MAX HR reported by my iFit Band. This has happened more than once, and generally on low HR activities on the treadmill such as walks. With runs, it is closer, but still thrown off by a rather large spike that almost always happens at the start of a run where it just goes up, up, up and lasts for a couple of minutes.

[t.parker]

@Tristannyc average HR for some activities comes close because it's an average. Average brings lots of data into a single number so even terrible data may get the correct average result. If in one second your HR was 60bpm and then went up to 160bpm then the average from those two readings is 110bpm. The two consecutive readings of 110bpm would result in the same average. If we consider 110bpm to be the correct one, the first readings of 60 and 160 are quite badly incorrect yet the average is correct. It's a lucky coincidence. This is exactly why the average on Fitbit comes often quite close to the correct one. It isn't always the case though. For many activities like cycling even the average is wrong because the data isn't good enough to compensate for bad readings when computing the average. The calories will follow the same pattern as HR (as they are tied together).

[Mr.Light]

Hi

https://www.youtube.com/watch?v=L3pYPvdi1qE

This device without a chest strap is useless. Oh it doesn't support a chest strap.....so it is...

[catz247]

I feel the original sense is better wirh hr then the 2

[slackjaw99]

Yup, my return is already in the mail.

[JerryHam]

Some more data from today. Again, when doing a power walk on the treadmill is where I see it consistently the absolute worst - today at times off by 89%. When I do my normal runs, it is closer but still too incorrect to use. Here's from one of the walks in this month's iFit Walking Challenge (a short, 13 minute, flat, walk on the treadmill). I overrode the super slow speed built into this walk and set it to 4.0 MPH for a nice, comfortable power walk.

During the walk my HR per the iFit Band / Treadmill was mostly in the 80's with a couple of peaks just into the 90's and the max hit 94 for two seconds. That's an important number, since at 55 years old with a 165 max exercise heart rate, FitBit gives me "below zones" up to 93. At 94, that is where it switches to "Fat Burn" and starts adding "Active Zone Minutes" - in Fat Burn zone, it adds one Active Zone Minute per minute of clock time. I should not have received any Active Zone Minutes for this easy walk since my HR only hit the zone for 2 seconds. But since FitBit has no idea what my actual HR is, and overestimated it by so much, it gave me 21 Active Zone Minutes for this walk.

Often I saw this (HR on the Treadmill is from the iFit Band):

 

Real HR at 86, FitBit Sense 2 says 163 - 89% off.

Here's the graph from both iFit and FitBit overlayed on top of each other. The numbers on the left and the green line is my HR per the iFit Band whilst the numbers on the right with the yellow line are from the Sense2:

Hard to believe it can be off by so much, but this is mostly repeatable; I've seen it time and again with power walks like this. Again, it is closer on my runs, but not close enough to use.

In this case, FitBit says my average HR was 140 and the max was 164. But in reality it never even got over 94.

[KenVa]

Oh, so that is why I got such a "deal" this month on the Sense Advanced from Best Buy.  I was replacing a Blaze, an upgrade I thought, until I saw the heart rate when exercising.   Doing the same exercise as always, the heart rate was way low.  I wore both at the same time.  The average heart rate was 94 on the Sense and 131 on the Blaze.  Steps were even more inconsitent- 2,094 on the Sense and 1,115 on the Blaze.  This was a workout on the old style NordicTrak cross-country skier.  I didn't see this thread/issue until after spending a long back and forth with Twitter support today.  No acknowledgement of this issue by the Tech Support person, even after I included a link to this thread.  The entire solution, after I confirmed step accuracy, was a shutdown and restart.  Also, the device wouldn't even restart by holding the button for 10+ seconds.

The Sense is my 3rd Fitbit and I find the handling of this issue very disappointing, to say the least.  This is no way to do business.  I'm pretty pissed.

Anyone have any experience returning a defective product like this to Best Buy?

[mix123]

It would appear the Sense 2 and Sense share the same inaccurate heart rate readings. I think users deserve to know if this can be fixed via firmware/software in the future or if this is just the way it is for these devices. This would let them decide if the Sense 2 will fit their needs or not. Not disclosing this information seems super misleading.

[crippstar]

I had some success with shifting the Sense further up my arm whilst exercising by loosening the strap, but usually the heart rate tracking would only resume accuracy if I took my hand off whatever I was holding and waved it around in the air for a minute. Not ideal! The Sense 2 seems to be much better. I can't comment about Best Buy and returns as they don't exist in the UK, but their nearest equivalent - Currys - would take it back if you said it was defective and not fit for purpose. 

[HR722]

I'm having a similar experience on the fitbit Inspire 2.  Treadmill shows heart rate of 140 while fitbit claims 95 bpm.  There is a serious "lag" between when heart rate initially ramps up to the time the fitbit figures it out.  Will save my time and money by not buying the fitbit Sense or fitbit Sense 2 thanks to information on this thread.

[mix123]

This is crazy. Nordic Skiing. Red is the Sense 2. Green is a Wahoo Chest Strap. Purple is a Scoche Rythm+ on my right wrist. The Sense 2 isn't even remotely close. Having the Scoche on the other wrist gives a good idea that wrist based HR can be fairly accurate. It lines up very close to the chest strap.

[mix123]

It's actually hard to believe it's this bad. I originally just looked at the average and max and thought it was OK. But it's not. 

Check this out. I compared it to 2 other heart rate monitors on a nordic ski. One of the others was wrist based on my right arm.

https://community.fitbit.com/t5/Feature-Suggestions/Fix-the-Sense-2-Heart-Rate-Sensor/idc-p/5281130#...

[CougarCreek]

Return at Best Buy; Depends on how long you have. had the unit. If bought recently for holidays they are offering returns until 15Jan2023.  Worth a try.  I had mine for a week and returned with no issues. 

[FloppyDrive]

This problem has been around for a few generations of Fitbit devices, so I believe they have no plans to fix it. I suspect trying to fix it will foul up their medical approvals for the ECG and Afib notifications and their sleep accuracy. They really need to start from scratch here instead of building onto a bad foundation, because their main competitor is leaving them in the dust.