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 

[lenny4d]

@tomp2

First link https://medium.com/@pkuhar/comparing-optical-heart-rate-monitor-technical-implementations-450c6b35bd...

[ahefner33]

You can compare them with the times but ths is strap vs ionic

[SunsetRunner]

@ahefner33 I'm not sure if I understand this correctly but it looks like for low-intensity activity ( from your HR, time and distance I presume it's a walk ) the HR is higher than the strap registers. If so, this is exactly what I observed when I took my strap and just walked. Ionic HR is higher compared to the strap. With higher intensity, the Ionic stays behind the strap and shows lower ( sometimes a lot ) readings. I usually don't use walking as a form of activity but did it once or twice. The longest tracking I did was approximately 1hr30m and after analysing the data I found out this:

There are 5301 records of HR for the specified timeframe. It's a lot! I merged data coming from PolarH10 ( strap ) and Ionic and run some simple analysis. The result doesn't really surprise me:

- 92% of all the records HR is higher on Ionic - almost ALL THE TIME considering the walk is 1,5hr!

- On average, the difference is about 15bpm. Median shows a pretty much same result.

This is a totally opposite result to ones I capture during high-intensity training. So far I assumed at least walking can be done right but looks like it can't. Moreover, due to higher readings, it certainly overestimates calories burned. I'm really tired of analysing the data as Fitbit knows better and numbers certainly lie.

[ahefner33]

@SunsetRunner

Correct except yours is much more better organized lol. Yes its a basic walk since I was noticing just walking around, my hr was showing very high numbers. So i took these snaps and screens.

Its not in order but if you look at the duration times that are closest are the comparable heart rates at the time.

I agree also that my ionic will show lower heart rates when running than strap almost up to 10 bpms. 

I have the same issue with charge 3 and even had a running hr bpm ang difference right around 20bpms. Off a 5 mile jog

[Dwalouke]

My heart rate is not being accurately recorded either. It seems that my ionic is not capturing my heart rate as it increases in the beginnings of my workouts. This morning I rowed a very hard 2,000 meters and 5 minutes in to the row, with heart pounding and me gasping, my ionic was reading about 110 and in fat burn zone. It wasn’t until about minute 8 that it actually caught up. It did the same thing about 45 mins later when I got on a spinner and as the sweat rolled off my nose, it took my Ionic well over 5 minutes to recognize my elevated, peak zone heart rate. Thishas happened repeatedly  and started happening within the last couple weeks. What to do?

[FredFornicola]

My HR readings have been wildly inaccurate the last few days when I'm on my bike. I have it set for Spinning to get a more accurate reading. This morning I restarted the Ionic 3 times because the first time I was too low, the second time it spiked like crazy and the third time I just let it go. I monitored my pulse manually every few minutes and I was around 130-135 for most of the workout. The graph below from this mornings session says otherwise

Keep in mind that I never stopped pedaling when I restarted so I was already cycling for about 10 minutes. I have no idea where the spikes and dips came in because I was not doing intervals today. Yesterday's results were worse from doing intervals. 

[tomp2]

I think this is an example of how inconsistent the Ionic can be. I got the opposite yesterday. I did a 10 minute treadmill running session, as a warm up to a gym session. I'd normally expect my HR to rise to about 120 in an short exercise like this. Instead, within about 45 seconds, it had leaped up to 150. For the next few minutes, it carried on rising, eventually peaking at almost 180. My professionally measured HR max is 160. Apart from 180 being impossible for me to reach, I'd only been running for a few minutes and was hardly out of breath. Over the remainder of the 10 minutes, the Ionic reading slowly fell back down to between 120 and 130. I suspect that, when I'd finished, the reading was about right. Results like this are completely useless. My patience with the Ionic is wearing very thin in less than two months of use.

[lenny4d]

@tomp2"My patience with the Ionic is wearing very thin in less than two months of use."

Yeah. Thanks for all the analysis. I have a Versa but it is the same thing, I now started to wear a Versa AND a Charge 2 (my earlier workhorse tracker), and now I am even sadder. I love my Versa, it is so much more comfortable, the screen is great (as is the Ionic), I like the potential of the apps, I like the music, the wallet, but if the most fundamental aspect - the HR sensor - sucks, it is game over.  Now I have a fancy step-counter.

I can see more and more that fitbit most likely had to make some (software) design compromises, to get the battery time up etc (all competing watches use more or less the same hardware sensor with minor differences), but this is not good enough. 

I would have had no problem if they had implemented an "accurate HR" mode with limited battery life (more intense use of the sensor (higher sampling rate), more advanced processing and higher cpu load), if I had the choice to switch. I would neither have a problem if they allow one to link to a bluetooth cheststrap and get hi-accuracy heartrate readings from there.

But they have to do something, otherwise there will be a lot of fitbits in the junkbin after christmas. If the new firmware had solve all issues it would have been rolled out long ago. So this sits a lot deeper. I really like my Versa, but I am starting to doubt now as well...

[Freeco]

---

@tomp2 wrote:

My patience with the Ionic is wearing very thin in less than two months of use.

---

Same here... I've actually started researching if a Garmin Vivoactive 3 wouldn't perform better.
I'd hate leaving the Fitbit ecosystem after 3.5 years, cause I have all my historical data here and also own an Aria to get weight into the picture. But accuracy does go over sentiment...

[FredFornicola]

Another day of HR bouncing every which way. This was a steady paced ride. No change in resistance or RPMsHELP!

[Kael]

I've done a lot of testing with various devices over the past few years, to include MS Band 2, Polar Loop, Scosche Rhythm+ (Optical Armband), Samsung Fit, Samsung Fit2, Fitbit Charge 2HR (iirc), Fitbit Ionic, Wahoo Tickr X (chest strap) and a few different Garmins. Many were worn at the same time (e.g. 1 per wrist, chest, arm).

From what I've seen in the raw data, optical-on-wrist is still not picture perfect. It works decently for resting heart rate, but any extreneous workout and it's best to look at is as a "reflection" of the workout. It's not perfect, and if compared to data from the same device, it'll give you a general idea of whether you went harder or not. The average tends to fall close to other devices, but the min/max can be quite off. Opticals really start to struggle with accuracy around 150ish bpm and above. For best results, I wear mine snug with screen facing inwards on the wrist (opposite of a traditional watch position). This serves to block extra light as well as give a better area for the sensor (no hair).

The most accurate consumer HR devices I've used were chest based electrode devices. Followed by  the Rhythm+ (optical armband), and then the optical-on-wrist. Ionic is pretty accurate, but needs to be worn right for the particular user and is susceptible to displacement errors. It also (like most opticals) lags about 30-45 seconds behind other methods.

Another guy that does good tests on this stuff is DCRainmaker, worth googling that (not sure if I can post a link).

[FredFornicola]

I have to agree with you on the accuracy of the resting heart rate. Two weeks ago I had an EKG and Ionic was only off by one beat.

[lenny4d]

@FredFornicola

>> I have to agree with you on the accuracy of the resting heart rate. Two weeks ago I had an EKG and Ionic was only off by one beat.

Well, EKG measures your pulse (while resting), while fitbit resting heartrate is something derived not in real-time but based on analysis of your heart rate while seemingly at rest, but not at sleep (based on motion sensors) . Some kind of average. That the two match is good, but one has to keep in mind that the underlying process is very different.

@Kael

>> For best results, I wear mine snug with screen facing inwards on the wrist (opposite of a traditional watch position).

Now, that's interesting,You mean in line with the palm of your hand. I never tried that!!

>> The most accurate consumer HR devices I've used were chest based electrode devices.

They are 100% accurate. Of course they have some filtering too, but the electrical pulse measurement is very reliable. They don't need motion compensation/filtering like the optical measurement needs! I use these to benchmark my fitbit all the time (with good results on the CH2 at any time, and generally poor results on the Versa, especially at higher bpm (when there is obviously also a lot of body movement).

See my link that I posted earlier for some good comparison and inner workings of optical heart-rate sensor hardware:

https://medium.com/@pkuhar/comparing-optical-heart-rate-monitor-technical-implementations-450c6b35bd...

[tomp2]

Helpful comments. Thanks.

Tom

[Freeco]

Seems the Garmin watches are experiencing the same kind of problems with oHR. There's also quite some complaints on their forum.

A bit weird this kind of thing doesn't come up in reviews though. Take DC Rainmaker for example. He has a head to head comparison between the Ionic and the Vivoactive 3. I interpret his review that both devices perform quite comparable. It's a personal choice which one suits you better. There's no real negatives in his review.

The difference between his and my training is that he does outdoor running and cycling, while I train indoor on the elliptical. That makes me think it's because sweat will evaporate quicker outdoor, because of wind on the skin. While indoor there's hardly any air movement, except from my arms going back and forth.

Another clue is that my Ionic does seem to measure properly till around 14-15mins in my workout. Below is my measured heart rate during 4mins warmup, then 36mins targeted at 155bpm with resistance controlled by a Polar T31 chest strap, and a 6min cool down:Cardio workout targeted at 155bpm

After the 5min warm up my HR was at 125bpm, then my elliptical started increasing resistance for me to hit the 155bpm marker. My Ionic registered that quite well for a few mins, but started struggling when (I guess) my arm got wet from sweating.

What I would like to try is put a peace of cloth between the watch and my arm (with a hole for the sensors & LED of course), to try and 'pull' some sweat from underneath the watch.

And maybe I'll order some aftermarket sporting band, one with more holes in it so sweat doesn't stay trapped underneath.

[AlejandraFitbit]

Hello guys, I hope you are doing fine.

I appreciate all the screenshots and information about your heart rate information and as recommended, if you are having problems with this data, take a look at this post where you will be able to see some things that might be affecting this reading and some tips to improve it.

See you around. 

[tomp2]

Useful comments, Freeco.

I read a lot of reviews before deciding to go with the Ionic, but you’re right - a basic failing to record HR properly isn’t mentioned in any of them. Makes you think that Fitbit, Garmin, etc probably come the heavy very quickly and ask for critical reviews to be taken down. I just wish I had discovered this forum before I bought my Ionic. I certainly would not have bought it if I had read the saga of the firmware upgrades, read about the failings of the wrist bands, and several other issues. Basically now, like others, I am left making “the best of a bad job” with a watch that I cannot return to the store from which I bought it, because I have no categorical evidence that it's broken. I also wish I had studied more about the physiology of wrist-based pulse readings.

I agree with what you say about mis-readings in the opening part of an exercise session. I’m now having to do a kind of “pre-warm up” with my Ionic, until the readings stop going haywire, before starting any session where HR tracking is important to my record keeping. My Ionic tracks my sleep, my steps, etc adequately well, and the issue is not about the fit of the thing on my wrist, despite the mods now having twice posted about that in this thread.

Tom

[Freeco]

Finally, some progress

One of the suggestions FB support always makes is to put the watch further away from your wrist. I had already tried that in the past, but it's just not practical. Once your arm gets a bit wet, the watch will just slide towards your wrist again. And it doesn't take long to do so.

But I thought of creating a border, with the sweat bands I bought a couple of weeks ago. So the watch would just push against the band, but can't slide down.

For day to day wearing I have 2 notches left on the band. During workouts I'd tighten it up 1 notch. This time I didn't do that, I left it at 2 notches but pulled the watch away from my wrist as much as possible so it's really tight up my arm (after my workout I could see the holes in the band made marks on my skin).Then I put the sweat band on, to hold the watch in place.I did another cardio workout on my elliptical, as explained earlier:

a 4min warmup, with the elliptical on lowest resistance

a 36min workout, while the elliptical would try to get my HR at 155bpm by varying the resistance

a 7min cool down

The result:

There's 2 dips mid-way, but other than that the Ionic measured reasonably well!

If this isn't a fluke, I'm a happy customer again

Next time I'm going to loosen the watch 1 notch, cause it might have been a little too tight.

[lenny4d]

@Freeco"There's 2 dips mid-way, but other than that the Ionic measured reasonably well!"

What are you comparing against? Do you cheststrap or other electrode measurement? You were not ~30mins in peak-zone were you?

[Freeco]

Compared to my post of the 27th, which is the same workout from the day before (1min difference in the cool down period)

During the 36min workout I actually was around 155bpm all the time. That's the goal of the cardio workout program I selected on my elliptical. It'll try to get my at a target HR (I choose 155) by varying its resistance. My elliptical is linked to a Polar T31 chest band for those HR controlled programs.

The elliptical will not get me at 155bpm spot on all the time, but quite close. Typically it can keep me between 150-160bpm quite well.

So if the Ionic also measured a HR around 155bpm during that period, it did a good job at that.