This discussion has been locked.

You can no longer post new replies to this discussion. If you have a question you can start a new discussion

OHR Polling frequency when asleep/not moving (vs visualization on 4hr-graph)

Just wanted to check something I noticed.

When looking at the 4 hour HR-graph on the watch it looks great, frequent sampling and no weird spikes anymore it seems. Also, looking at Garmin Connect the sampling frequency has increased massively when movement is detected (in whichever form). When sleeping however, it decreased even more (as much as 4-5 hrs between measurements) and also yesterday when going to a meeting where I had to sit and listen to a speech for ~45 minutes no measurement seems to be taken in these 45 mins.

Before it was about the same frequency when sleeping vs when awake, easily 30-60 mins between sampling. Now when asleep it's 60-240 mins and while awake 1-5 mins... Again, suits me fine but also looking at the HR graph when waking up it appears as if it has been sampling just as frequently the last few hours while I was asleep and this does not appear to be the case from Garmin Connect.

Is Garmin simply 'pimping' the HR graphs with random noise around the last measurement in these hours? Making it look like it's very active, while in reality it just plots around the last measured value.

While I'm pretty happy with the current way of measuring I wouldn't be fond of hiding the low-frequency sampling this way. It just isn't reality and gave me a complete invalid impression.

Can anyone comment on this, or did someone else notice the same?

What would be the best way to check hr-sampling frequency? I already got the fit files from the monitoring folder and converted them to csv, but have a hard time interpreting the data...

0 ASmugDill over 9 years ago

RHR is the lowest heart rate observed when the subject is awake, no?

When I say it's sampling RHR every 15 minutes, I'm referring to while you're sleeping/at rest. It's not set in stone, but for most people their RHR is while they're asleep.

My understanding is that the RHR is the lowest heart rate when a person is awake, so it is irrelevant what the heart rate monitor does when the user is asleep anyway, especially when the watch does not use heart rate as a factor for detecting sleep.

Of course, some individual users may also have their own reasons for being interested in how their sleeping HR, but that should not be conflated with the idea of RHR. That said, for all I know, the four-hour RHR shown in the Heart Rate widget could well be using an incorrect definition tacitly – I've not have been of the habit to check it first thing in the morning for a while now.

0 Former Member over 9 years ago

My understanding is that the RHR is the lowest heart rate when a person is awake, so it is irrelevant what the heart rate monitor does when the user is asleep anyway, especially when the watch does not use heart rate as a factor for detecting sleep.

Remember, that's wikipedia, and not a medical journal :).

I think there are a number of definitions. Fitbit certainly uses a different one, something closer to (or exactly) what you've said. And for me their RHR value is always about 10BPM higher than what i expect. That's pretty common from what i've seen, including DC Rainmaker who's mentioned the past that their measure of RHR is high.

For most people, the lowest stable HR across the day is a fairly good measure of what RHR is meant to be. And this tends to happen when you're asleep.

Anyway, we're digressing. This thread is about the decreased(?) sampling rate when asleep after the firmware update. Or at least what ends up being displayed in GC.

0 Former Member over 9 years ago

Came here to start exactly this thread. I've unfortunately seen this behaviour every night since updating. Bizarrely, on the watch the itself the sampling looks pretty good, as others have noted. I also see a difference between what the watch says my RHR is and what GC claims.

The following sums up my experience so far...

I'm very happy with the increased sampling rate during the day, but horrified it's gone almost to zero during sleep. One of the two key reasons to 24/7 sample your HR is to find the RHR value. If Garmin is going to degrade the sampling exactly when most people are in this low HR state, what's the point?!?

Can i ask though, i have recording set to "smart", since i find little difference for the accuracy of my runs, but slightly less battery drain. Does this setting also determine the 24/7 sampling rate? What do others have it set to?

I'm going to flick mine over to 1-sec in a few days once i've confirmed the above behaviour is the new norm and not just the new firmware "settling in". I hope i don't need to.

0 ASmugDill over 9 years ago

One of the two key reasons to 24/7 sample your HR is to find the RHR value. If Garmin is going to degrade the sampling exactly when most people are in this low HR state, what's the point?!?

Again, I think it really depends on what the definition of RHR is – and, no, the individual user is not given room to decide for himself or herself what that definition is. If Garmin officially uses a definition that requires the user to be awake, then the minimum HR observed (or observable) at any time of day is not the RHR, and it doesn't matter if the individual product owner/user is also interested in that metric irrespective of whether that is the RHR; the point (to answer your question) would be that the manufacturer has made a design decision to prioritise battery power consumption for gathering what is relevant data for determining RHR, and unilaterally force a compromise onto users about where to cut back on battery power consumption. It's still not about giving users a choice or configurable option to decide for themselves what is important; you get what you're given, and if enough people complain loudly enough for long enough, everyone ends up being given something else (being a different compromise made for the individual as opposed to allowing user an increased level of control).

0 Former Member over 9 years ago

Again, I think it really depends on what the definition of RHR is – and, no, the individual user is not given room to decide for himself or herself what that definition is. If Garmin officially uses a definition that requires the user to be awake, then the minimum HR observed (or observable) at any time of day is not the RHR, and it doesn't matter if the individual product owner/user is also interested in that metric irrespective of whether that is the RHR; the point (to answer your question) would be that the manufacturer has made a design decision to prioritise battery power consumption for gathering what is relevant data for determining RHR, and unilaterally force a compromise onto users about where to cut back on battery power consumption. It's still not about giving users a choice or configurable option to decide for themselves what is important; you get what you're given, and if enough people complain loudly enough for long enough, everyone ends up being given something else (being a different compromise made for the individual as opposed to allowing user an increased level of control).

I guess we all stand corrected then. And thank you for restating what most of us already know and in such authoritarian terms. My my question was actually rhetorical.

0 ASmugDill over 9 years ago

My my question was actually rhetorical.

Hence my “Captain Obvious” answer. You'd probably know by now that I'm a big fan of product manufacturers making design decisions without consulting users or allowing users any illusions that their views make a difference; the more strongly users believe that in the age of social media companies need to bend to consumers, the more I like to point out it doesn't play out that way in market reality. As consumers, we're not all like-minded.

0 Former Member over 9 years ago

I too have the straight line on my graph following the update. I have looked at my .fit files and have confirmed a recorded sample was taken every 15 mins. My HR fluctuated 2-3 bpm while sleeping, so I can't explain the new graph. I wonder if it's just showing my average HR during that period. I'm just speculating though. Have you been able to view your .fit files? I'm curious to know if you'll find the same thing.

0 FlipStone over 9 years ago

I too have the straight line on my graph following the update. I have looked at my .fit files and have confirmed a recorded sample was taken every 15 mins. My HR fluctuated 2-3 bpm while sleeping, so I can't explain the new graph. I wonder if it's just showing my average HR during that period. I'm just speculating though. Have you been able to view your .fit files? I'm curious to know if you'll find the same thing.

I attempted to and it indeed seems there is a polling interval of 15 minutes while asleep (sometimes disturbed by movement or big increase in measured hr or something?).

What I mean by that (and the following I see a few times):

I see a number of readings measured 15 minutes apart, they are in the 50-50-49-51-51 range. Then suddenly 12mins later there is a 57 reading, 1 minute after that a new measurement of 52 and then 15 minutes later a 50 again... (and then agian 15 min intervals of 50bpm).

So I guess there is some smoothing or averaging in the graph in garmin connect.

While awake this seems to be somewhere between 1 and 5 minutes (more often 1 and around that then 5) and every now and then above that (I've seen a few 8 and 9 min intervals as well).

Anyway... this is what I have been able to figure out I think, as far as I understand the fit file format ;)

0 Former Member over 9 years ago

I attempted to and it indeed seems there is a polling interval of 15 minutes while asleep (sometimes disturbed by movement or big increase in measured hr or something?).

What I mean by that (and the following I see a few times):

I see a number of readings measured 15 minutes apart, they are in the 50-50-49-51-51 range. Then suddenly 12mins later there is a 57 reading, 1 minute after that a new measurement of 52 and then 15 minutes later a 50 again... (and then agian 15 min intervals of 50bpm).

So I guess there is some smoothing or averaging in the graph in garmin connect.

While awake this seems to be somewhere between 1 and 5 minutes (more often 1 and around that then 5) and every now and then above that (I've seen a few 8 and 9 min intervals as well).

Anyway... this is what I have been able to figure out I think, as far as I understand the fit file format ;)

Looks as if we're responding to the same questions in two different post. LOL

0 FlipStone over 9 years ago

Looks as if we're responding to the same questions in two different post. LOL

Haha, I noticed yeah ;)

Ah well, spreading knowledge to even more people I guess :)

Question btw... Perhaps you know.

You mentioned the timestamp was in seconds from midnight. Is a certain timezone used? Because the sleep start seemed to be 9~ish hours after midnight for me (which I can only explain if a different timezone is used as reference). Also, the timestamp conversion led to 14th of april in 1996 instead of 2016 (I believe the 14th of april was correct in this case). I live in The Netherlands btw.