HRM accuracy while resting

[Milanese44]

Hi everyone
I just got my bangle.js 2 and have been figuring out some stuff the last couple days. I love the product and the ability to program personal ideas!

However one thing that confuses me a bit is the HRM. I know there are many discussions about this, but the thing is, for me it seems to be quite accurate a lot of the time. Yesterday I tried it on a run and I never saw an unexpected value. It also seems to adjust well to sitting or walking. What I have noticed however is that while sleeping, the HRM behaves much worse. Im getting about 90bpm for extendet periods of time at night (~50min) when my resting heart rate should be around 50-60 bpm. Does anyone have similar experiences?

I havent done much testing about why this is, but from just running the live HRM it seems to me the way the bangle is oriented influences how well the HR is measured. Does anyone have an idea if this is a known problem?

I was thinking about logging acceleration data, HR and confidence level and see if there is a pattern, but since im new to this that will take me a while probably.

Thank you for any insight you might have and if you know an app that does these logs already let me know please ;)

[Milanese44]

Ok small update: I have gotten the mentioned data but I cant see any correlation. I will still post it here in case anyone wants to look at it

Next step I guess has to be to make sure the HRM ACTUALLY isnt correct by comparing to another device and then I will also try logging the actual HRM measurements to maybe see something but it doesnt seem to be an issue of "health" choosing the wrong measurement but the sensor returning weird measurements.

[gfwilliams]

I'm glad that the HRM sleep issue is sorted. I don't think your HRM is defective - there have been other posts about this as well.

It's a tricky one, there was an Open Source HRM algorithm which we could fiddle with, but unfortunately virtually nobody was willing to even record me test HRM data, let alone make actual improvements. So I moved to a manufacturer-provided HRM sensor library that comes as a binary blob and it seemed to be better - but the downside is there's really very little we can do to debug or improve it.

Recently I did add a setting to cutting edge Bangle.js builds : Bangle.setOptions({hrmStaticSampleTime:1}) and I'd be interested to see if this improves matters for you. It brings the usage of the binary HRM library more inline with the examples and I guess it could help.

[Milanese44]

I see, so I assume I could add this setting to my Bangle and see if it does anything? Where would i have to implement this?
Also, depending on how much data you need I could provide some. I'm not sure I'm skilled enough to make improvements on an HRM algorithm though...

[gfwilliams]

Try just pasting this into the Web IDE's left hand side:

{
let s = require('Storage').readJSON('setting.json',1)||{};
if (!s.options) s.options={};
s.options.hrmStaticSampleTime = true;
require('Storage').writeJSON('setting.json',s);
}
load();

That should enable it in the options (until you change true to false and re-run to disable). But for it to work you'll have to be on a 'cutting edge' build - not just the normal 2v27 release

[Milanese44]

alright, will try it out, thanks!

[Milanese44]

I tried it but it doesn't seem to fix my problem :(