Optical Heart Sensor accuracy - why is so bad?

Optical HRM sensors are notorious for inaccuracy in high intensity workouts. Try wearing it tighter and if this does not help then there's not much you can do.

My 645 correlates well with the electronic pulse meter in my elliptical trainer (PreCor) for values from 45 to 155 (I'm a bit older so this is at the 90th percentile of HR). I wear it above my wrist bone and reasonably tight on the wrist. FYI only. YMMV.

TomekL_ I'm afraid that this at present is a limitation of Wrist based HR monitoring and I'm not sure that it will ever reach parity with the accuracy of the strap. You may get better results by changing position, tightness etc but this seems to vary enormously from person to person. I find my 935 (same sensor etc) is OK for a normal run but still less accurate than Chest strap - I have just resigned myself to remembering to put the strap on particularly as I swim where the OHR doesn't work anyway. Hope that adjusting position helps your problem

I tend to get very good results from optical HRM when running and indoor cycling, and better than expected when outdoor cycling. I might get better cycling results that a lot of others because I'm mostly riding a hybrid bike to commute so my position is more upright and not putting as much strain on my wrists.

There's an IQ datafield called Auxiliary HR which allows you to record the data from a HR strap to a secondary channel and save it to the FIT file so you can compare it with the optical HRM data during the activity and afterward.

I used it with both the 235 and 735 for awhile when I first got them and they correlated quite well. Almost identical on steady state runs and with a bit of lag on intervals. The shorter and faster the intervals the worse it was but that's to be expected. I have it installed on my 645 but haven't played with it yet since I got the watch a couple of days before a race and I wore a HRM-Run strap with it there for Running Dynamics reasons and I'm taking 2 full weeks off running because I developed a slight muscle tear in my lower abdomen in the race. I might use it when I go for a longish bike ride on Sunday though.

Wow, wish I had heard of the Auxiliary HR earlier. Now I just need to dig out my chest strap and give it a go next week. Thanks for the pointer. To save others the trouble of searching: https://apps.garmin.com/en-US/apps/88ce4547-7d84-4289-b2c7-3e15ca00185f#0

TomekL_ I'm afraid that this at present is a limitation of Wrist based HR monitoring and I'm not sure that it will ever reach parity with the accuracy of the strap. You may get better results by changing position, tightness etc but this seems to vary enormously from person to person.

This above.

I use an Scosche optical HRM in my arm and it performs almost as well as strap HRMs. Almost, because an strap HRM measures directly heart electric activity, while optical HRMs have to infer pulse through very complicated image processing and calculations. And unfortunately the wrist is not the best place to get good optical HR measurements. When I get my 645 I will use the built-in wrist HRM for normal runs (where wrist HRM perform quite well), but probably not for intervals.

kurtie_bcn yes I was mainly referring to wrist based rather than the arm based but even these have very variable results for different people

Optical is an inferior technology compared to the strap. the trade off is unobtrusiveness. I use the optical to keep the easy runs easy but for threshold work I always use the chest strap.

One great aspect of the watch is you can connect to a strap and bypass the optical any time

Steady state it works well for me, but for any type of interval / varying intensity training I just use a chest strap.

… i was full o hope that the optical sensor can replace my HRM strap. It looks that i was wrong.

An optical heart rate monitor, especially one that becomes part of the watch and worn around your wrist – so that you don't have to buy, charge, put on, clean and maintain yet another device – may be a substitute for a chest strap HRM functionally, but not in terms of HR data quality where the trade-off between user convenience (and/or comfort) and accuracy (and/or reliability) is being offered. You won't get the same or better outcome qualitatively while significantly improving in another area, such as by lowering the burden of using a HRM in practice, but at least now you have a choice – which you can make separately for each timed activity you want to record, and the best compromise is entirely in your court to decide.