Exercise heartrate way off

[dopeydwerg]

Hi,

Recently I switched from the Ionic to the new Sense. But the heartrate during exercises is not very accurate. Actually it's way off comparing with the records of the same exercises of the ionic.

My average heart rate during Weights was normally around 109bpm but now it's 74bpm. for sure this can not be right? And the other day I played squash and I used the tennis exercise to log it and it said my hr never went above 109 while normally it peaks around 165. Is there something wrong with my fitbit sense or is this a common bug?

[Beezy1978]

I didn’t do a bunch of anything intentionally to try to get my Fitbit to work. I just happened to put it on the day before I was going to return it and it works fantastically. I would just like to save a lot of people the hassle of returning a perfectly good product because the heart rate sensor needs to adjust. But whatever.

[Beezy1978]

It sounds like you’re just sour on Fitbit because of your experience with the versa. 
smh

[SunsetRunner]

---

@Beezy1978 wrote:

I’m sorry but if you won’t open the box and try it out yourself and determine whether the heart rate monitor is accurate yourself then yes you are unintelligent. But whatever, do you. SMH...

---

Thank you for being such a pleasant person. Your exemplary example of how to treat people should be placed on a pedestal for all to see and marvel.

---

@Beezy1978 wrote:

It sounds like you’re just sour on Fitbit because of your experience with the versa. 
smh

---

Generally, when a product that I drop $200+ (Versa 2) on doesn't work as advertised, then, yes, I do tend to sour on that product. I returned the Versa 2 in exchange for another one, and the same problem occurred. My Versa 2 experience was a year ago. I did go back to my Ionic last year, so your assertion that I was soured on Fitbit from that experience may be misplaced.

Having returned the Versa 2 twice due to the heart rate accuracy issues, I simply didn't want to go through that mess again. Fool me once, shame on you, fool me twice... applies here.

[DeeNewSenseUser]

Exact same problem here. Wore my Charge for over 18 months and routinely hit my peak heart rate during my running workouts. I haven’t been able to hit peak heart rate yet with the Sense during similar workouts over the last several weeks. Definitely an issue as Fitbit’s brand is really about the variety and accuracy of all the cool health data. 

[mikchaos]

I'm a Versa 2 user and have been using the Sense for roughly two weeks now. I've done lots of workouts, walks, bike rides and hikes with both watches on and wanted to share some of my findings. 

1. The Sense's HR sensor seems to be quite sensitive to changes in the heartrate when you're close to resting heart rate. I've noticed that when I was sitting at my desk and get up to grab some water, the Sense HR shoots up. The Versa 2 takes a little longer to recognize the increased heart rate. After a few moments, they both align again around the same heart rate. Resting heart rate overall is very similar with both watches
2. During workouts and exercises, the Sense and Versa 2 react completely differently. I have constantly observed that the Sense is really bad at picking up high heart rates and often times is roughly 15-20 bpms lower than what the Versa 2 is showing. The biggest difference was noticeable during high intensity workouts where the heart rate jumps up and down quickly. The Sense seemed to "smoothen" it out a bit too much. During hikes and walks where the heart rate stays relatively constant, the difference wasn't that significant

I don't want to say that something is wrong with the new heart rate sensor but it definitely is calibrated differently than previous versions. Since Fitbit has chosen to stay silent on this matter leaving us in the dark whether they consider this an issue that they should fix, I'll be returning my Sense for now. Hopefully this can be adjusted with future Firmware updates and I'll be watching the forums here. Apart from the HR Issue I really like the Sense, new Fitbit OS is better and the additional sensors have potential.

Here's a few workouts comparing the Versa 2 and the Sense:

Workout

Hike

Workout

Walk

[sportionary]

So my latest update with this situation with fitbit.  I've chatted them every single day this week with more information on my issue.  Difference with my issue is I have more proof than most of you do that mine is not working since I purchased 2 sense's one for myself and one for my wife on the same day.  I have her sense to compare my sense too.  My sense I can remove it from my body set it on for example the bed, and it still tracks a heart rate and will fluctuate up and down like it is on a person.  My wife takes hers off sets it next to mine and the heart rate goes to zero and remains zero till she puts it back on and it can start tracking her heart rate.  I've sent them videos of the difference's of the heart rate of the 2.  I've worn my sense which tracks on average 30bpm lower during exercise, switch off to her sense and up jumps the heart rate to a range it should be in for the exercise I am doing.  They have 'escalated' my case a couple times this week.  I got no response till last night when I chatted them yet again which finally prompted a response.  Here is the response for that department:

Hi Kris,

Thanks for your effort and patience in regards of this situation, we appreciate the fact that you have followed our instructions already.

We're contacting you from a specialized department at Fitbit. Let us say, in regards of the discrepancies between the heart rate readings of your Fitbit Sense that we are working on calibrating the sensors and the way the heart rate information is processed so we can provide you with more accurate information as soon as possible so we want to ask you to wait for upcoming updates for your Fitbit Sense. You will see further firmware updates availability at the Fitbit app on your phone.

Hopefully this will resolve your questions.

In the mean time, in case you come across any other inquiries.

Sincerely,

Sheri and the Fitbit Team

So basically sit and wait for who knows how long.  No mention on an explanation or even acknowledgment of how one is working properly and one is not, and how a firmware update across the board will not backtrack the one working and fix the one that isn't.  What is frustrating is I have no idea if the update is coming out in days, weeks or months.  I've asked this question and have yet to get an answer.

[NeilFinny]

Suppose we should have stuck with the old rule of don't dive right in to new tech

[supercad79]

After 5 factory resets and swapping watches with my wife twice, it's apparently tracking correctly. For the past few days I've been getting all of my active minutes and my heart rate has been what I expected it to be. So hopefully my issue is resolved.

One truly odd thing to note is when I swapped my watch for my wife's watch (we both have a Sense), I was getting my active minutes and proper heart rate with her watch, but she literally didn't get any active minutes with my watch because her heart rate never reached her zone level. Thinking that my watch was defective, I swapped them back and that's when my watch started tracking correctly. So now both of the watches work.

So now I'll wait to see if an update keeps the fix or screws it up again.

[D-Mit]

Interesting that Fitbit talked about a firmware update that was supposed to address this. Your post is dated 10-24-2020 and the new firmware update was released a few weeks ago in November.

I just posted a thread in the Versa 3 board: https://community.fitbit.com/t5/Versa-3/Heart-Rate-off-on-rapid-increases/m-p/4610234#M3873

about this exact same issue and how it doesn't seem to be resolved, at least on my watch, so I'm interested to see if the people in this thread have had their HR sensor issues resolved on their Sense. It seems that the Sense and Versa 3 have the same HR issues.

[sportionary]

No change in my issue with the update. It still tracking on average 30 bpm lower during exercise than my ionic did. About ready to change to Apple Watch honestly. 

[D-Mit]

Yeah, I figured as much. 

This may be strike three for this watch for me. I can tolerate the no music thing, and the call notification issue seems to be working for now. But this HR thing is a complete deal breaker. What's the point of a fitness tracker if it can't track the most basic metric of your fitness while exercising? Especially if it is something that it did once relatively accurately.

What I'm upset the most with Fitbit about is if they had some new HR algo they wanted to test they should have tested it before production. They aren't a startup that can put out pre-production level software and explain it away as a small company that is growing with lack of resources. People shouldn't be paying them for beta testing.

Today I was looking at the Galaxy Watch 3, since I have an Android. I might have enough discounts and I think I can get the price down below $250, I'm probably going to get it. 

It's sad, the Versa 3 has everything I want and nothing I don't. It just doesn't work. 

[supercad79]

I was able to resolve it by doing five factory resets. The heart rate tracks really well now, even before the update.

The only issue I have now is that my watch needs to be covered and kept warm to track properly. When it's 20-30° F outside it will usually be wrong due to the cold. When I cover it up and wear gloves, it tracks just fine, but this is a similar problem with many trackers out there.

I'm happy with the watch now. Just took a bit to get there.

[Lazywalker]

Mine did work for a month or so after several resets but now it’s back to incorrect heart rate.

It’s also restarting itself all the time



Sent from my iPhone

[D-Mit]

5 factory resets? 

That seems excessive. Did Fitbit tell you to do that? And if so, why? Is it something that fixes the watch somehow? 

[supercad79]

No, I didn't get any direction from Fitbit support. It was really a result of switching watches with my wife to try and see if it was just my watch or if it happened to hers as well. After switching back a couple of times it started to work properly. I had no other options since the return period at Bestbuy had already expired (only 2 weeks) and Fitbit wouldn't do anything since I didn't buy it directly from their website. I didn't want a $330 refurbished watch.

[steveo57]

This is a very detailed analysis from @Micsb  (with pictures showing the elapsed times) and it tracks along with what I have found.  For exercises that are extended (not short burst) Sense seems to be pretty accurate but there is a lag of a minute or two.  I have checked it on a treadmill which I can get the heartrate up to 105 to 120 when I set the rate of speed at 3.8 to 4 and it maintains that giving active minutes.  The treadmill itself  has a heartrate sensor and (after the lag) is almost dead on when compared after the Sense catches up.  At the end of the exercise I just wait and watch the heart rate and a minute or two later it drops below active minute range (so it has caught up) then I end the exercise.  Works the same way when I walk, run, or play racquetball.  This lag time is definitely a problem and I hope they are working on it but I think the sensors are accurate after the lag time just as @Micsb says.  I think the old watches may have read a bit high and did not have the lag time the Sense does causing confusion and frustration.  Also I think some folks are confused with the switch to active minutes only starting give credit when you get to an elevated heart rate.  For weights or interval training the lag time is a problem because by the time you get your heart rate up with the lag you have already shut down the exercise or did not spend enough time for the Sense to catch up therefore the heart rate seems off and never catches up to give you any accuracy or active minutes.  Bottom line the problem seems not to be the accuracy of the sensors but the lag time of sensing and that needs to be worked out.  By the way I have the new update and the lag is still there.  Since it seems to be algorithm related not accuracy of sensors hopefully they are working on it and will be fixed soon.   

[D-Mit]

---

@steveo57 wrote:

This is a very detailed analysis from @Micsb  (with pictures showing the elapsed times) and it tracks along with what I have found.  For exercises that are extended (not short burst) Sense seems to be pretty accurate but there is a lag of a minute or two.  I have checked it on a treadmill which I can get the heartrate up to 105 to 120 when I set the rate of speed at 3.8 to 4 and it maintains that giving active minutes.  The treadmill itself  has a heartrate sensor and (after the lag) is almost dead on when compared after the Sense catches up.  At the end of the exercise I just wait and watch the heart rate and a minute or two later it drops below active minute range (so it has caught up) then I end the exercise.  Works the same way when I walk, run, or play racquetball.  This lag time is definitely a problem and I hope they are working on it but I think the sensors are accurate after the lag time just as @Micsb says.  I think the old watches may have read a bit high and did not have the lag time the Sense does causing confusion and frustration.  Also I think some folks are confused with the switch to active minutes only starting give credit when you get to an elevated heart rate.  For weights or interval training the lag time is a problem because by the time you get your heart rate up with the lag you have already shut down the exercise or did not spend enough time for the Sense to catch up therefore the heart rate seems off and never catches up to give you any accuracy or active minutes.  Bottom line the problem seems not to be the accuracy of the sensors but the lag time of sensing and that needs to be worked out.  By the way I have the new update and the lag is still there.  Since it seems to be algorithm related not accuracy of sensors hopefully they are working on it and will be fixed soon.   

---

Yeah, this seems to be the assessment that most people are getting. That the problem lies in the algorithm, but this is something that should have been dealt with prior to production, which is really annoying on Fitbit's part. I don't particularly like that we're just hoping for a solution.

[Micsb]

Hi Steve, I have since done some more testing to figure out how to get the most of the Sense and decrease the impact on the HR readings. I think this might have something to do with the Fitbit marketed AI algoritm to be more accurate at estimating the HR. 

1. When walking slowly towards a steep hill, then walking briskly up the hill. My HR start at the bottom around 90 and end at 140-150 on my Polar. The Sense will sometimes stay around 100, with the hill climb too quickly for Sense to figure out what is going on.

2. When walking briskly towards the same hill and up the hill. My HR start at around 115 and this time quickly increase up to the Polar chest strap measured HR.

3. When doing Kettlebell training. If I take a long break between each routine, the HR never catch up before the routine is over, however if I continue with next routine with only a short break before HR drop below 110 on the Polar chest strap, the Sense will much quicker increase and show the correct HR.

My personally conclusion is the AI, when walking slowly or resting, it will go into powersave mode and average HR over a longer period.  A quick rise in HR might initially be flagged as false reading. To be as accurate as possible to avoid false alarms if you use this to actively warn you if the HR goes over a certain threshold. 

On the other hand while exercising the AI will anticipate a change in HR and be more responsive. 

This was a month ago, havent tested this behavior with the new firmware. I found it good enough, and it makes the kettlebell exercises much more challenging switching between exercises without breaks in between. I dont do interval running due to running injuries years ago, I limit myself to 5km daily at a descent pace. 

Could you give it a try with an interval session, where you keep the break short and start next cycle before getting the HR all the way down? I know this is not correct, however I am curious to see if others see the same change in the responsiveness of the Sense HR readings.

[D-Mit]

This all would be fine if the watch was holding off on spiking heart rate when it thought the user wasn't exercising to avoid potential false spikes, although I'd say, IMO, measure what you see, not what you think you see. But I'm willing to deal. 

However, if the user purposefully puts the watch in exercise mode thereby telling the watch to look for heart rate changes then it should be extremely sensitive to spikes. It doesn't do that and still lags.

Again, I'm hoping this gets fixed pretty soon but honestly if the Galaxy Watch drops below $300 again I don't think I'm going to be able to hold back. 

[Micsb]

I agree, activating exercise mode should definitely put priority on responsiveness. Dunno what is going on with the variation of reported problems with the HR readings. Personally I am happy with the performance however I am not weight lifting and do very little hit training. Dont care about its accuracy during kettlebell sessions, dont do kettlebell for cardio training.