Ionic's tracking of heart rate

[marcwilliams25]

My first Fitbit was the Surge, bought it for the GPS more than anything (track my hikes) I was so impressed by it that this year I thought I'd upgrade to the Ionic, got to admit that now it seems like a poor decision. The Surge seemed to track my heart rate during workouts quite accurately. It picked up on peaks where I'd push myself during a session. The Ionic is nowhere near as accurate in my opinion.

To give you an example, yesterday I just did a gentle 30 minute run on a treadmill, nothing really that taxing. When I synced my Fitbit apparently, my heart rate was in the peak zone throughout the 30 minutes. Today I did a "3 rounds for time" workout that had a 20 calorie assault bike as an element, after I finished it I couldn't get off the floor for about five minutes, I was totally gassed, when I synced my Ionic my heart rate was just seven minutes in the cardio zone. That can't be accurate

Moderator edit: Updated subject for clarity 

[SilviaFitbit]

A warm welcome to the Community @marcwilliams25! I appreciate the details you've mentioned. I would like you to make sure, you've double check that you're properly wearing your Ionic What should I know about my heart rate data?. 

You can try restarting it and continue monitoring the information you're getting. 

Hope this helps! Keep me posted. 

[lenny4d]

I know what you mean.... I have had a Charge 2 for over a year and it was always pretty accurate (within +/- 5%), but with the Versa it is no-where near. Especially worrisome is that it can give numbers well above my max heart rate, like 180-190. That should simply not be possible 🙂

I use the treadmill and elliptical for calibration, if you have a decent one it should have handles with electrodes and give you a very accurate reading that you can compare with the reading at the same time. I use my phone with the app to look at the reading so I don't have to look at the watch. Anywhere within 5% off I consider acceptable.

"All" of the fitbits seem to have different HB sensor hardware (just looking at it on the bottom of each tracker/watch), and thus different firmware to interpret the signals. That is probably where it sucks because as I wrote, HB readings much higher than the known max HB read are simply not possible. It never ever happened on my Charge 2.

[tomp2]

I am a very new Ionic user at the moment (less than 10 days). This is my first post here.

I happened to be wearing my new watch during a couple of rest days from training and was surprised when I got an on-screen "reward" notification while just out walking. When I looked at the graph of my HR, it seemed that the walk had taken me well into the kind of zone I'd be in at the gym just after warm up for a heavy session. I know very well what that would feel like, and the walk definitely hadn't taken me there, no matter what the Ionic said.

First time I used it at the gym, therefore, I decided to to a straight test between the Ionic and my HR recorded off a Polar wireless HR chest strap (with new battery) communicating with a WattBike which was in turn connected by bluetooth to the WattBike app on my iPhone. I did three repetitions of a short sessions of 10 minutes, starting the "HR Graph" app I've added to my Ionic at the same time that the WattBike app started. What I got at the start of the first rep was the Polar belt saying HR was 100 when the Ionic said 90. I carried on, and within about three minutes, both were reading around 108. They stayed within a beat or so of each other until the end of the 10 mins, having reached around 130. Five mins break then I began again. Virtually the same result - either the HR strap was over reading at the start, or the Ionic was under reading at the start. However, both pretty much agreed after about three minutes. Same again on the third rep.

Puzzled, I began reading up some stuff online about how wrist-based HR measurement worked. I wish I'd done this before buying, because despite good reviews on Techradar etc , this would have put me off buying the Ionic as a replacement for my venerable old Polar S810iHR watch.

It seems that many wrist-reading devices can't properly distinguish between the vibration of walking and a wrist pulse. There is plenty on the internet about this. This typically causes over-reading when walking. On the Wattbike, my hand was held still on the bars, and there were no extraneous vibrations. Whether that meant I got a "truer"reading that way, I don't know. Maybe the Ionic HR algorithm tries to compensate for vibration from walking (and, I assume, from running) by simply holding back the reading by several beats, I can only guess.

Next test will be on a treadmill, again using my Polar HR belt that talks to the treadmill screen, and I'll see what the Ionic claims, by way of comparison.

Tom.

[lenny4d]

@tomp2:

>> and was surprised when I got an on-screen "reward" notification

Yeah, fitbit produces A LOT of 'rewards' 🙂 I ignore them as I don't care at all and don't care to find out if these can be switched off or not

>> I decided to to a straight test between the Ionic and my HR recorded off a Polar wireless HR chest strap

Good, that's the best way to get to know the fitbit HB sensor. The chest strap is 100% accurate, it measures via electrodes. I use the electrode handles on the treadmill and eliptical, it does the same thing. I consider 5% as a the max deviation that should be acceptable. I don't think fitbit gives any accuracy numbers. Generally, low-impact low HB figures are always reliable (say below 100). Whereas many people experience problems at medium to high impact high (true) HB figures, say above 130-140. But it varies. My Charge 2 was very accurate at least up to around 150 bpm.

>> It seems that many wrist-reading devices can't properly distinguish between the vibration of walking and a wrist pulse. There is plenty on the internet about this. This typically causes over-reading when walking.

Allthough all share the same technology concept, there are substantial hardware variations, and probably even more so software (firmware) differences. Which make that some models are a heck more robust than others. Arm motions are generally filtered out (for as much as it works). But there are differences in how many lights are used, and how many sensors are used. More of each can increase accuracy. Smarter software (takes more time to develop) can help. Yes, the software also looks at the intensity of what you are or were doing.

>> Next test will be on a treadmill, again using my Polar HR belt that talks to the treadmill screen, and I'll see what the Ionic claims, by way of comparison.

I suggest strongly to experiment with firmness of wristband and how high above your wrist you have the watch, to see what works well and what does definitively not work well. Once you know this you can use this in daily excercise routines

[SunsetRunner]

A few days ago I did and posted similar comparison of Ionic and PolarH10 tracking during half an hour of outdoor cycling.

https://community.fitbit.com/t5/Ionic/Heart-rate-accuracy-test-cycling/m-p/3016226#M73222

Today, I collected some data for simple walking and what I find interesting, Ionic overestimates my efforts and shows HR significantly higher. Actually, I thought it could be right and Polar wrong but old school method of checking pulse with fingers told me that the wrong one was Ionic.

[tomp2]

Thanks. Helpful advice.

[Freeco]

I have the same frustration with HR accuracy of my Ionic as you, Marc.

I've used a Surge for 3 years. Although not a perfect device, functionally I was quite happy with it. It's biggest flaw were the bubbling band, and low step count on an elliptical.

Fed up with the bubbling bands, since June I've upgraded to an Ionic expecting similar (or better) HR tracking. But I'm a bit disappointed.

First few weeks I didn't really notice an issue. I did notice lower HR measurements than usual.
But I've been doing the same kind of workouts on my elliptical the past few months/years. I've grown stronger/fitter in the meantime, so I thought my heartrate had gradually decreased over time. Still, when the elliptical randomly selected a tougher resistance, I could still go in the peak zone for a short while. The past few weeks I've started increasing the intensity of my workouts trying to hit the peak zone again. But since using the Ionic I've hardly been in de peak zone anymore. Yesterday I've done another intense session, where my elliptical picked up HR's over 170 bpm through its chest strap. My Ionic hasn't gone over 141 bpm...
A few days ago I did a 45min session where my elliptical would try and keep my HR at 158 bpm through varying the resistance. My Ionic was also way below that.

Bottom line, where my HR measured by my Surge was mostly in the higher part of the cardio zone and from time to time in the peak zone (quite in line with the chest strap of my elliptical), after changing to an Ionic it says I'm in the high fat burn zone to mid cardio zone. So ~15-20 bpm lower. While doing the same workouts (or even more intense), wearing it the same way on my wrist.

I'm actually wondering if the Ionic isn't starting to malfunction. Yesterday after my workout I looked at the LED on the back. I noticed a small red blinking dot directly to the right of green LED. When I look again now, I don't see that red dot anymore. Will check again the coming days...

[KeviZa]

When I first purchased my FitBit ionic , I decided to test it's accuracy compared to my Garmin Forerunner 620 using a chest strap to monitor heart rate . I went for a run of about 10km's and was very surprised that what the Ionic was saying in regards to heart rate , distance and average speed were very close to what the Garmin was saying . 

That was the only time that I actually tested the Ionic compared to another device but as far as that test went , the Ionic was pretty accurate . 

[tomp2]

Since my previous post, I've had a chance to do another real-world experiment with my Ionic.

The "Sports" wristband I bought for the watch arrived yesterday. This had a much finer range of adjustment than the one provided with the watch. If wrist movement or a watch that was too loose was an issue, I hoped this would help resolve it.

My gym has a number of elliptical trainer and stepper devices that record HR via the bars you hold while exercising. Several have fixed, static bars plus optional "swing" bars. I chose to exercise on one of these because I'd effectively be getting two sets of comparative figures against the Ionic reading, if I did an exercise set that used both types of bar. I borrowed a Polar HR wristwatch similar to the one I used to own, and exercised with a Polar HR strap on too. This meant I had three simultaneous readings available while exercising (Ionic, Polar, and elliptical trainer console), plus an option to get a reading on the console from a different hand position.

Everything was carefully zeroed before I began. At rest, before exercise, I had a common reading of 88 bpm (+/- 1 bpm) from all three sources. First off, I did an increasingly hard seven and a half minute session holding the static bars on the elliptical trainer. The Polar reading and the HR reading on the trainer's console were within a beat or two of each other throughout this. The Ionic was initially all over the place, before eventually coming to plus or minus a few beats of the other two after five minutes exercise. My hands were held almost completely still throughout, with just my legs and lower body working. I was unimpressed with the Ionic.

I checked that the Ionic was gaining full and consistent contact with my wrist, then for the following seven and a half minutes, I repeated the experiment, but using the "swing" bars on the elliptical trainer, so that my arms were moving as well as my legs. Again, I got consistency via the trainer's handles and the Polar HR strap. This time, however, the Ionic figures were simply a joke, occasionally ranging, almost at random, from 90 beats to 140 beats while the other devices agreed on around 130. 

I took a short break, during which I dried my wrist and the back of the Ionic, and checked the Polar HR strap reading against another piece of equipment in the gym that can take a reading direct from a Polar wireless strap. All fine. Then I repeated the two seven and a half minute repetitions described above. With my hands almost static on the fixed bars of the elliptical trainer, the Ionic gave me slightly less variation than before, but the trainer and the Polar strap again were within a beat or so throughout. The Ionic was almost always about 5% to 10% out by comparison. Switching to the "swing" bars of the trainer again sent the Ionic into fantasy land.

At the end of this set of tests, I stood perfectly still while holding the trainer's swing bars for about three minutes, and let my heart rate settle back to a rested reading. All three devices settled to the same heart rate reading.

My conclusion from this is pretty obvious - the Ionic just can't cope with readings where body an/or arms are moving. Even if I hadn't had the readings from the elliptical trainer itself and from the Polar strap, what the Ionic was showing me during the exercise (via the Heart Rate Graph app I've added to it) would have had no credibility at all to anyone familiar with their body and accustomed to hard exercise.

If I have any charity left in me, I may repeat these test on my next visit to the gym, but change the fit of Ionic's wristband to make it a little looser. It seems perfect at present, but I'd not want these results to be dismissed by anyone suggesting I had it "too tight".

I am beginning to think I should have spent my money on an egg-timer or a sundial, either of which possibly give more accurate readings! Oh, and I've not used the Ionic while _running_ yet. I'm just waiting for a small problem with one foot to heal. I am, of course, not optimistic what the Ionic is going to show me.

Tom

[marcwilliams25]

My mate has an Apple watch and after a hike, my Ionic (and has to be said my old Surge) reports significantly more steps than his Apple, so yes I believe the Fitbit may be overestimating steps as well  

[marcwilliams25]

Ah yes the bubbling strap! Mine eventually fell apart but to their credit Fitbit offered either a replacement or an Ionic at half price. Beginning to regret choosing the Ionic as apart from the strap I was quite happy with the Surge  

[lenny4d]

@tomp2  yes you are doing the right type of tests, I have done many of these (both elliptical and treadmill) with my charge 2 and never had a problem within +/- 5%. I still have to do the same with my Versa (which is more or less the same as the Ionic, the red 'light' (I didn't know it can light up!) is/must be the sensor, the green in the middle is the led light, the charge had the opposite config with two led's and one sensor).

In my experience the best in these type of tests is to keep an as steady pace (HR) and arm movement as possible, gradually increasing and see when things fly off and what you can do about it. Another thing to check (and worries me) is sweat build-up under the watch. The Versa and Ionic have a much larger surface contact area. You won't have that at 100bpm but at 170 that's a whole different story. With me that's a problem.

[tomp2]

Thanks. Yes, I dried my arm and the back of the watch between tests, just as a precaution, although there wasn't a big build up there after the first test, 

I'm going to continue experimenting with the Ionic, while using the Polar strap on various devices in the gym, although I am already fairly firmly of the view that wrist-based measurement is not really very accurate.

Tom

[SunsetRunner]

@tomp2 the type of activity seems to have quite a big impact on the heart rate reading. It may depend on how hard one is flexing muscles of forearms, the motion of wrist etc. It is worth noting that this isn't really HR measured but blood flow. So minor inaccuracies I find acceptable. Today I gave a go to the, in my opinion, Ionic nemesis - the rowing machine. Two sets, each targetted at burning 250kcal with rest in the middle. Second set is with more resistance.

[Freeco]

@lenny4d: "the red 'light' (I didn't know it can light up!) is/must be the sensor, the green in the middle is the led light"

No, it was in the LED section, not in one of the 3 optical sensors. The green LED is a bit off-center, a bit to the left (if you hold the Ionic with the charging contacts on top). That red dot I saw was slightly to the right of the center.

I know, I wasn't expecting a red blinking dot there. I've only seen it once though. Maybe it was a weird reflection of the lights above my mirror or so. Don't know... I'm paying extra attention to it now every time I take off my Ionic.

[tomp2]

Those graphs would be pretty similar to the kind of difference between the strap and the Ionic that I had the other day. 

I have a recurrent back problem that prevents me from using the rower, so I am spared that "nemesis" at least!

Tom

[SunsetRunner]

Now here's an interesting result. I went for a 15k run this afternoon but I put the Ionic under the wrist sweatband:

The wristband was supposed to prevent the watch from moving and getting unwanted light from outside. It did the job. The watch was stable for my whole run. Here's the result:

 It's not perfect however there are 3 major HR drops that occur in both readings. To be honest I was expecting lot more mess on the Ionic side and average HR difference is only 7 beats between strap and Ionic. Of course, running, especially in quite a steady pace seems to be quite easy to monitor for Ionic but I'm quite positive that all major drops have been registered too ( unlike in the cycling test ). This is quite acceptable. I will still run the numbers though as the whole run took almost 1,5 hour so graphs have to fit lots of data in quite narrow display space but I find results maybe even promising. I will try to test it same way during cycling, rowing and weightlifting. The downside is that I could not see the screen for the whole time because it was under the wristband.

[Freeco]

hmm, that's definitely worth a try.

Today I tried loosening my Ionic a little allowing for a bit higher positioning.

Same issue though. The Ionic registered way lower bpms compared to my chest strap.

[lenny4d]

@SunsetRunnergood stuff! Look forward to next results.