FR235 HRM and Sleep Data Comparison

So I went on my first run with FW 3.20/WHMR 2.30, this time I also disabled smart recording to have more data points.

It was supposed to be a steady run, but in the mist of curiousity about the sensors accuracy I just had to incorporate a few pace changes.

Actually I wanted to try to trigger the crossover problem. I was running steady pace until BPM was about the same as SPM. Then I slowed down but keeping the SPM in the same range, just wanting "real" bpm going down but see if the FRs HR reading would lock onto cadence.

Accidentally, or rather stupidly I tricked myself however. On the page where I had setup Cadence I setup HR for the lap instead of current HR. I didn't notice and thought "Bam there you go crossover" while it actually wasn't locked to cadence. So I even stopped and took off the watch because (avg) HR wouldn't change... Well... :rolleyes:

It was my standard 6k lap, though after 5k I got the "battery low" warning. Since I didn't want to risk losing the run I stopped and saved it. Still had 11% left though I then found out. The last Km I recorded in a seperate run.

Is there a way to merge two activities?

Here's the result of my first 5k.
Blue: Polar H7
Orange: FR235

https://connect.garmin.com/modern/activity/979283710

Is there a way to merge two activities?

Try the file combiner on this page
https://www.fitfiletools.com

another tennis challenge

Here's another challenge for the FR235. This was a two hour tennis session I recorded with the FR235 (red, FW 3.20) and a Tickr X chest strap.

I'm not sure if the timing is exactly on both devices, so this might account for a lag as well. It's hard to say because sometimes the FR235 looks to lag behind and sometimes the peaks are quite spot on compared to the chest strap.

My opinion:

Is the FR235 accurate? No, it isn't. I think there's so much stop and go as well as arm movements involved, furthermore at times holding a tennis ball or supporting the racket with my non-dominant hand (where I wear the FR) might flex the wrist a bit and lead to inaccurate readings.





In the zoom of the first half you can see the FR was actually tracking HR quite good, there are a few drops and spikes, but it is not bad at all.





In the second half (where we slowed down in doubles mode) the FR's HR seems smoothed out vs the constant up and downs recorded by the chest strap.






Overall I'd say you can't really say it can track such a workout accurately, but still gives a very good idea of where you are. Also average and max HR are reported almost the same on both devices. So maybe even calories calculated could be reasonable.

Overall I'd say you can't really say it can track such a workout accurately, but still gives a very good idea of where you are. Also average and max HR are reported almost the same on both devices. So maybe even calories calculated could be reasonable.

I think you show what we largely already know. An optical HRM is generally going to have a noisier result than a chest strap HRM simply because the OHRM is measuring an effect of an effect of the heart's contractions whereas the chest strap HRM is measuring the actual electrical field of the heart. There will a higher signal variance simply by nature of the mechanisms involved.

When we ask if it is "accurate" we need to define terms -- just what exactly does "accurate" mean? Are we looking for medical grade electrocardiogram readouts? Or are we looking for average & max HR in the workout plus the time in different HR zones?

Let's be careful to avoid letting perfect become the enemy of good. We need to define and remember what is "good enough" for the purpose intended.

Example - when you drive a car it's fairly important to keep the edges of the car between the centerline and the side of the road. It is far less important whether you maintain your inside tire edge at 24.5" or 25.0" inches from the centerline.

I think you show what we largely already know. An optical HRM is generally going to have a noisier result than a chest strap HRM simply because the OHRM is measuring an effect of an effect of the heart's contractions whereas the chest strap HRM is measuring the actual electrical field of the heart. There will a higher signal variance simply by nature of the mechanisms involved.

Well, true, we already know that. I do this mostly because I am very curious myself. And while for the most people owning a FR235 this is not relevant I think it's just interesting to actually see how current technology performs in different situations.

I didn't mean to leave the impression I think the sensor is useless, that's why I also said there's so many different factors playing into it, making it quite different from running or cycling. You could consider tennis as a non-structured interval workout (though a runner would probably not train like that). I was just interested of how the sensor/software would cope with a HR quickly changing between low and high.

For instance: in this specific training, I think you can see how the software possibly applies smoothing during the second half of the training. HR was varying within a smaller range and more regular than before. That was probably interpreted as noise or spikes that need to be smoothed out. Could... not sure though.

When we ask if it is "accurate" we need to define terms -- just what exactly does "accurate" mean? Are we looking for medical grade electrocardiogram readouts? Or are we looking for average & max HR in the workout plus the time in different HR zones?

Let's be careful to avoid letting perfect become the enemy of good. We need to define and remember what is "good enough" for the purpose intended.

Absolutely... Thanks for pointing that out.

Interestingly, to date I have not seen a post defining "accuracy" other than things like "except from some spikes it's pretty close" or "the HR is all over the place". But then again, it's probably subjective for different people even for the same activity.

For instance I will never try to be in a specific HR zone while playing tennis, the only purpose here would be to use it as a simple training log, and maybe get some estimates like calories, distance or steps from it. Another good use would be for Training Load/Recovery purposes. Actually those were my ideas for this situation. And for that it's "accurate" enough.


For running I have found the sensor to be mostly "accurate" enough so I could ditch the chest strap. Do i need to define accuracy now? Just thought about it but it's not easy to put in words. :rolleyes: I'd say it should report the correct zones and it should not be lagging behind much so that it can be used for intervall trainings.

Maybe someone could calculate a standard deviation with some data as an example, for me it's been too long since I've done that sort of math :rolleyes:

Yes, "accuracy" is a squiggly thing to nail down in this aspect as what's needed will vary by user and his/her specific training needs. I think it is useful to present the data; and its interpretation into information left to the reader. For me I'm defining the term as whether or not the 235 is giving me sufficient quality information to guide my training during and after runs. So far it's doing okay, albeit with a sample size of four runs in the five days I've owned it.

I'm reminded of years ago I had great success using a Polar running watch with a foot pod and HRM strap. During a run I could view pace / distance / HR / HR zone, and post-run the watch presented only max & average HR info on a per lap basis, plus an overall average zone. Using that watch I trained for and ran two marathons, succeeding in my goal of a sub-3:30.

I also had a Garmin 610 for years and its HRM2 chest strap that got really f'd up in winter and often gave bullsh:t readings early in a run if I had the wrong shirt and/or forgot to use electrode gel. It'd also sometimes wig out with bizarre HR readings. I still managed to train for my goals successfully.

So maybe not having owned "perfection" I'm just not aware of what I'm missing.

Here's a fun question to toss around: barring really bizarre sustained variances (and barring cadence-lock), what does the typical runner or other athlete gain from the higher "precision" of a chest strap HRM? Then, how do those gains translate into better helping that individual achieve his/her goals?

So maybe not having owned "perfection" I'm just not aware of what I'm missing.

Nicely put in a nutshell! I guess owning perfection (as in equipment that give "perfect" information) is reserved to professional athletes or maybe people who have money to toss around.

Here's a fun question to toss around: barring really bizarre sustained variances (and barring cadence-lock), what does the typical runner or other athlete gain from the higher "precision" of a chest strap HRM? Then, how do those gains translate into better helping that individual achieve his/her goals?

I'd say zone based training helps to diversify and structure training to achieve the "optimal" effect. If higher precision means we have better control of our training (i.e. we are more likely to be in the zone that we target) we might get closer to that optimum.

But a good point if you consider the following: I suppose most runners don't know their true HRmax, nor their true lactate treshold, etc... So training zones are based on averages, estimates and assupmtions (we know that already).

Furthermore the concept of training zones could be seen like a model, and models are usually reduced to a few variables (like HRmax for instance), just a simplification of reality. In reality there are a bazillion more variables. So I guess things like weather, climate, time of day, illness, recovery state, emotional state and so on have an influence and would mean different zone definitions for each of us and for any given moment.

Then we could also define 20 zones instead of 5, or every HR number could be their own zone, but this would be too much to grasp of course. Or like many say, we could just run without technology and go by feel :)


To summarize my fortnight experience with the FR235 HRM so far: I'm pretty pleased with it, not really run into cadence locks (even if I tried to trigger it on purpose), average/max metrics are mostly within a few heartbeats and the average "picture" (let me call this "within the same heart rate zone") looks very good when compared to chest strap.

Steps appear to be accurate

Havent checked other like steps. I thought about doing this, having two options: using the iPhone with a step counting app (I had tried some before and all gave very different step count) b) using one of those 1$-step counters I have somewhere laying around.

It's probably also kind of complex as there are many movements of the arm involved. Some magazine is always doing a "tooth-brushing" test to see how many wrong steps are counted during arm movement.

I might just try to compare this on a normal walk... will post here if I do this.

EDIT:

out of curiousity I just walked a bit around the house. Stepcount:
FR235: 313
1$-device: 341
runtastic pedometer app: 370
Argus app: 310
Pacer app: 317

Don't know what to make of this, haha... just that the runtastic app seems more sensitive to movement.

After all I think it just comes down to what kind of regular arm movements you have in your daily life... Like a cook or painter would probably get easily to many steps counted due to arm movements...

For me personally daily activity tracking is a nice to have gadget but probably will become more and more uninteresting after some time (hell I already know without a device that I could move around more ��). For me it's more interesting to have an accelerometer to estimate indoor-runs. Though I still have to see how it will work on the crosstrainer (elliptical) I have down in the basement.


EDIT2: yes, it has a 24/7 HRM. Though it seems to record in intervalls. As the manual says it collects more measurements when inactive. I hope they will include a setting in afuture firmware to disable 24/7 HRM but still turning it on for activities automatically. Until now there's only on or completely off.

My Fitbit Surge and the 235 were pretty accurate.

Fingers crossed it gets better, both for those who've already bought a 235 and because i want one! (one that works)

I'm still struggling to determine why battery is so bad for some but I am starting to suspect bluetooth.

I have left the watch with all the default seeting e.g. 24/7 HR etc BUT the only thing I turn off is bluetooth, and just turn it on when I want to sync to the phone. I am getting about 4 days use per charge (includes 1~2 1 hour runs with GPS).

As to bluetooth, I'm reminded of what many experienced with the Fenix3. The watch boosts the transmit power for the bluetooth radio when a paired phone was out of range, turned off, or the Garmin Connect Mobile app was hung and not responding. This boosted transmit power would really burn through the battery in short order, as in high to low charge overnight.

I've definitely noticed GCM being flaky at times, with me needing to kill / restart the app (iOS) sometimes after the watch has gone out of range and then come back.

At present I'm at 70% battery about 43 hours and one 35 minute GPS run since coming off the charger at 100%. Everything on.