Wrist heart rate sensor 23.00 behaviour

Just to be clear, 16 pulses per second should have enough time domain resolution for heart rate measurement during the day. I don't want my watch doing something more fancy then that. I want better battery life instead. Is there some setting which I'm missing?

Whenever I take the watch off, it will briefly enter the high frequency mode, before it detects that there is no movement. Maybe it does this to try to get a reading in difficult circumstances (water under the watch, light leakage).

Yes, when watch is removed I have same behaviour like you describe. That seems normal. Although now, after update if I peek under the watch, I see that it sometimes does cycles of exactly 2.5 minutes slow, 1.5 minutes fast. Other times it seems to enter high frequency for longer periods. I really believe that it was not like that before. And I really notice increased battery drain. I've seen other people talking about battery drain and 23.00. I wonder if HR sensor has anything to with that.

Does it try to track HRV also during daytime? I mean they added all this new stuff, maybe there is a bug or something? Does not look like coincidence to me. Or if it is a feature, then how to turn it off?

More info: opening heart rate widget always triggers fast mode. Heart rate widget glance on a screen - no. Whould be interesting if someone with older firmware version could comment if they have same behaviour or not.

The thing is, my watchface does not even have heart rate visible. And yet watch all the time periodically goes into high frequency mode for no apparent reason (at least this is my understanding).

3928202 said:

Does it try to track HRV also during daytime?

When looking at the files on the device (SW 23.00), it appears that HRV data is logged 24/7 and it has been that way for quite a while. Movement and heart rate data have always been logged 24/7. Perhaps HRV needs a higher sampling rate. I do feel that at night, there is more light leakage from the watch than there was a few months ago. But without more knowledge about the internal workings of the software, I can't infer anything when it comes to driving the LEDs or increased computational workload.

Thank you for your insights. 24/7 HRV logging is new to me. Is it not controlled by the setting "Log HRV"? I have it off. HRV tracking most certainly needs higher pulse rate, it is supposed to measure differences in the order of milliseconds.

Anyway, I've also noticed, that if I really keep still for some time, heart rate sensor stays calm (blinking slowly). It gets triggered by opening heart rate widget or moving, walking through the room. So it seems to kick in when watch registers some movement. By the way, I have MoveIQ disabled, auto max HR, FTP, etc. detection disabled.

Also tried now to turn off Activity Tracking. Still same. I'm really suspecting that this could be the reason why me and some other users suddenly noticed decreased battery life.

3928202 said:

24/7 HRV logging is new to me

For me as well. I only discovered it by accident, the data is logged in the \Garmin\Sleep folder and I expected only data around normal sleep hours in there, but likely because they enabled detecting naps during the day, this has to be logged 24/7. According to a FirstBeat paper, detecting sleep stages needs HRV and breathing rate, so it makes sense to log HRV and breathing rate all day. What might be happening, every time there is a transition between sitting still/moving around, it might be a fall asleep or wake up event, and higher frequency samples are needed, which may be the reason that you see extra LED activity during the day. Because the log entries are 1 per minute, maybe they do extra sampling every minute. But that's all guessing! No guarantee any of this is true - Garmin did not document the implementation, and Firstbeat only discusses the theory.

3928202 said:

Is it not controlled by the setting "Log HRV"?

No, that only pertains to activities. If you have it set on, unprocessed interbeat intervals are stored in arrays in the FIT file, for use by software later on. For example Runalyze shows it. I don't think there is an extra cost to setting this to On. Only a little bit of storage, it's not that much.

3928202 said:

Also tried now to turn off Activity Tracking.

That totally disables wrist HR, doesn't it? And no difference? Then something else must be going on.

Data said:

That totally disables wrist HR, doesn't it? And no difference? Then something else must be going on.

According to Garmin it disables step count, sleep, floors climbed, etc. detection. As I understand, that's like advanced analysis based on accelerometer and HR sensor, enabling watch to guess what the user is currently doing. The HR sensor itself is working and tracking heart rate with that disabled, just no fancy stuff. Again, as I said, disabling Activity Tracking does not make HR sensor calm. So logical conclusion is that it is not caused/related to that. I really thought it was.

Your information is interesting. It think now that it must be related to sleep / HRV. I will dig a bit more when I have time. If we presume that this is a new feature and if we presume that it really causes increased battery drain, then technically speaking this update causes watch to be out of advertised specs. I thought they can't do that. Can they?

Again basically based on a guess; but to investigate, you could consider to enter the beta program to receive the current beta 23.02. It includes a Device Usage widget that displays -amongst other things- time spent on OHR scan, and I assume that the time spent in OHR scan has a relation with the power used by it, so you may be able to compare different settings.

I can't tell if there is a battery usage difference myself, I have GPS on for so many hours that other power usage pales in comparison.

Data said:

the data is logged in the \Garmin\Sleep folder

 Oops, that was WRONG! There is no 24/7 logging in the Sleep folder anymore in version 23.x. I was looking at files from version 20.50. I was so intrigued by those files, just couldn't let it go, didn't realize I was working on months old data. It doesn't mean that the 24/7 HRV data is not collected, if it is, it's just not in the place I thought it was.