Optical HR Sensor Modes

Did I just discover something that all of you have already known? 

As an owner of an F6 Pro, F7, and Epix Gen 2, I thought I knew most everything there was to know about these watches.  I use my Garmin watches to primarily record biking, swimming, running, kayaking, and hiking activities.  I do a lot of walking and strength training that I do not record but I certainly monitor my HR on my watch face during these activities.  Today while lifting weights at the gym and at around the half way point of my workout I looked at my HR as displayed on my watch face and knew it was incorrect.  So...I thought I would go to my widget to see if the Garmin stock watch face display just had not been refreshed.  The HR was the same as watch face.  Bummer.  Then, for some reason, I thought I would push the start button while in the HR widget display to see if that would refresh the HR scan.   To my surprise my HR as displayed in the HR widget immediately jumped from 67 to 102 – which is approximately where I thought my actual HR would be – and my HR, as displayed by the watch, remained accurate for the duration of my strength training today. 

Is this functionality something that has always been in the Fenix firmware?  Does pushing start while in the HR widget put the optical HR reader in a more advanced mode that is more accurate but less battery efficient? 

I appreciate all of you

  • Yup, it seems that you have to tell the watch that you are doing an exercise to activate accurate HR tracking while exercising. 

    I noticed the same when using my Edge 1030 plus on my stationary trainer and using the Epix 2 to measure the HR (and recording intensity minutes) without starting an activity on the Epix. HR on Epix 70 while HR on 1030 plus 150 ;). intensity minutes way off because of this.

  • Strange, when I push the start button in the HR widget on my Epix Gen 2 (FW 8.15), a menu comes up (Pan Chart | Broadcast HR | Settings). The HR value does not change.

    Do you get reliable heart rates when recording an activity or do you also get lower values than expected?

  • Yes, when I actually record an activity my watch HR is accurate and also fairly consistent with my chest strap that is paired with my Edge that I use to record when riding. 

    in this case at the gym today I was not recording and saw a HR on my watch face and in the HR widget that was not accurate.  When I pushed the start button in the HR widget screen to see if it would correct the reading, the screen stayed the same but my HR reading immediately increased to my expectation. No other change occurred. 

    I am on 7.35 firmware. 

  • While the readings on the watchface can be accurate too(if i do some stairs quickly i can easily go over 100...after a training session also the HR on the watchface is definitely higher 10-20 bpm than normally, etc) i never tried during some heavy type of workout when i should reach 160/170 or more bpm.

    Could be wrong but opening HR widget is like asking the watch to take a more accurate reading(comparable to the activities probably?)...it's like when you open SPO2 widget and it starts to take the measurement...it means it's at "full power".

    As Kris, i also get some options when i press start but it's nothing related to a measurement of the HR.

  • OK, then it's not the same problem I experience (see https://forums.garmin.com/outdoor-recreation/outdoor-recreation/f/epix-2/290199/unreliable-step-count-and-heart-rate-sensors).

    Did you already have the occasion to test if the readings in the HR widget are consistently too low when you do a high-load exercise?

    The menu in the HR widget seems to be new in the beta FW versions, I didn't know that.

  • I just noticed the same thing doing a bike ride. It was showing around 60% of my actual HR when I was not doing an activity. When I then started an activity in jumped up to the correct HR. I have never noticed this before. I often don't start an activity for things like walks and hikes because it saves battery. In the past I am sure this used to be accurate. Has something changed?

  • Has something changed?

    Nope. Been there since the F6 at least if not before. There’s a Garmin Support note about this that you can find by searching the interweb. 

  • I guess I had always interpreted that as being the frequency of updates varied. For example it updates HR every couple of seconds when out of an activity but updates it a few times a second when doing an activity.

    I don’t understand the use of a HR monitor that records your HR at 60% less than it actually is. It makes all the others stats such as calories, intensity minutes, recovery etc pointless. 

    I’m sure it did not used to be this inaccurate when not doing an activity but maybe something else was going on (for example MoveIQ activities). 

  • I don’t understand the use of a HR monitor that records your HR at 60% less than it actually is.

    That might occur off you choose to do an activity without starting an activity. Under most circumstances when not actually doing an activity where heart rate would be expected to be higher and more changeable than normal, the lower polling rate is sufficient. Easy cycling or easy walking are generally low heart rate activities with very few rapid changes.

    Simply put, if you're going to be active and anticipate a higher than normal heart rate, record an activity. If you're just going for a walk at normal pace or easy bike ride with no efforts you likely need not select an activity.

    I’m sure it did not used to be this inaccurate when not doing an activity

    There have been some quite 'aggressive' discussion about this in the past. It has been this way for a long while. The main purpose is of course to conserve battery.

  • I understand the polling frequency is higher in activity mode compared to normal usage. But could it really explain the inaccuracies reported above?

    Both on the watch face and in the HR widget, I often see HR values changing at a frequency of 1 second, alternating with intervals where it is stable for a few seconds, so I wouldn't expect more than a few seconds of delay.