Polar are possibly correct but how about the quality of the data from the OH1? 

Hi, yes, I am trying to narrow down where does this come from too.

This is bike activity with OH1 and Edge 530, AVG HR 146 - exercise  load 161:

Precisely the same route (6x loop, 800m climbing, 50km) with 945 WHR, just faster by about 3 minutes, AVG HR 153 - exercise load 448 (!):

I could see similar before with 945 + OH1 like with Edge 530 + OH1. Will try to record the same route/pace next time with 945+OH1. But the difference is vast.

What I can see OH1 does more smoothing and catches HR drops during decline in a more real way, but still I consider the difference somehow way too much. That HR spike is not real, that's likely WHR inaccuracy.

Also one observation is that training load and training effects go quite steeply up with OH1 at the beginning of the activity but then the things slow down. Like the 1st loop on 530+OH1 gives exercise load 50, but remaining 5 like 20-25 each.

Does someone know if the training load algorithm is taking into account HRV? OH1 isn't supposed to measure that but in OH1 tracked activities respiration rate can be seen so it's possibly is. That's another thing besides smoothing which could contribute.

I did also this evening check on 945 WHR vs OH1 (using VA3) during light running and it's spot on (OH1 is smoother from both lines):

Any hints are welcomed! I am thinking to record similar with some Garmin chest strap too.

I understand all this stuff are guesstimates, heavily impacted by entered HR zones but in this shape it's quite useless even for some rough directions how to do trainings.

Thx!

153 v 146 in AVERAGE heart rate is a significant difference. Garmin/Firstbeat's TL is driven by EPOC, which in turn is quite sensitive to the duration in higher HR zones. It's a little difficult from the graphs presented because they have a different scale, but just eyeballing it, it seems you spent considerably more time above 150 in the second ride. I'd suggest perhaps looking at the Time in Zone for a little more data on how different the two rides might have been?

Hi there thank you for helping. Please check my attachments just posted and I've got very similar HR but still resulting in significant load difference (also training effect as a result of different EPOC) when using OH1 vs. Wrist HRM.

I am slowly collecting some data, so far nothing perfect, might take some time. This is Edge 530 + OH1 (appears to me the same like 945+OH1 although I don't have the same activity yet collected on both, counts twice training stats...) activity (66km on MTB, mostly gravel, 1200m climbing), 163 AVG HR, Exercise load 356. Appears to be compared to the 945+WHR activity in previous post with exercise load 448 bit low. Right, not sure what is algorithm after (not really intervals here)

HR zones are consistent on 945/530.

How do they guess the EPOC when external HR measurement is switched off after exercise? With the internal optical one which proved not to be the most reliable? And for sports where you don't wear the watch, there would be no EPOC data available at all, as belt is switched off and OHR not possible? I long gave up on intransparent TL calculations both from Polar and Garmin. I still use excel which works fine and I know where TL comes from.

Same experience with 

Scosche RHYTHM 24.

Scosche Rhythm+ is optical HRM like OH1. Anyone seeing similar behavior with non-Garmin chest-strap?

I am hoping that there is some sophistication behind the models, treating that as some rough directions, hopefully at least useful for hobby level. Thing is that 2-2,5x difference in EPOC estimates depending on HRM in use is just making it totally useless.

Former Member Not sure if Garmin can take a look at this issue? I've contacted Garmin regarding this issue but no response so far.