Bad HR readings

The one thing that I have picked up is that you need to wait for the little HR heart shaped symbol to stop flashing before pressing Start button to begin recording the Run (shows up in the sensors list along with the green GPS acquired ring). When the HR symbol stops flashing it indicates it's got a WHR lock.

Ref here http://www8.garmin.com/manuals/webhelp/fenix5/EN-US/GUID-08BC6CE7-EB8F-4392-9B7D-714B54D19499.html

I continue to get bad readings, no matter what I try (higher up the wrist etc.).
I contacted local Garmin Support last week but still have gotten no reply...

I wished the watch would just display "--" in the HR field when it loses its heart rate lock during runs instead of using some algorithm which often ends up being 30bpm or more too high.
All the derived data like VO2 max etc. is wrong because of it too.

Here's another example (most visible at 1h40):

I wished the watch would just display "--" in the HR field when it loses its heart rate lock during runs instead of using some algorithm which often ends up being 30bpm or more too high.

That section looks like cadence lock to me. Can you overlay the heart rate and cadence graphs in Garmin Connect and compare them?

That section looks like cadence lock to me. Can you overlay the heart rate and cadence graphs in Garmin Connect and compare them?

Clearly this is not cadence lock.

My guess is that the algorithm you see working in the beginning, which monitors both frequency and consistency of the raw oHR data and tells the watch either a "lock" or no lock is established keeps running throughout the workout (which would explain the lag of oHR vs. chest strap).
If there's no lock, I'm guessing the watch reverts to a formula, or a table of values based on pace, to return a "best approximation". These values might be based on historical average data collected by Garmin. If my theory is correct, people who are either very fit or not fit at all should see most erratic oHR readings.

So my suggestion to Garmin would be the following:
- Either make the oHR processing even more complex and adjust fallback values of oHR "no lock" to VO2, activity class, ... adjusted or even user historical based heart rate approximations
- Or display "--" when there is no lock (or at least give the option), no shame in that. This way, at least the pool of data used for VO2 calculation etc. is not tainted.

As to why the 935 supposedly performs better, it might be that the fallback values are calibrated different to suit/fool generally fitter triathlon people, it might be that the lighter 935 doesn't move around as much on the wrist so there's more "oHR lock", it might be that the better performance is based on anecdotes by just a few people but gets repeated over and over until it becomes an alternative fact, who knows.

Coming from a Pebble 2 optical heart rate and a Polar H7 heart rate chest strap- my thoughts are to always opt for a chest strap for actual activities. Optical heart rate is nice and can be spot on accurate if the connection is good- even in that $100 Pebble 2. But, when exercising, I'd be on a stair machine and my heart rate on the handlebars was clearly 180bpm and my watch would show 60, 80, or whatever it wanted. I'd take it off, move it around, lick my wrist (I have no idea what to do 0_o) and eventually the reading would match the same thing as the handlebars on the stair machine. I even shaved my wrist right where the back of the watch sits.

Optical heart rate is accurate when it can be. It's even used in medical-grade vitals machines ...but at the fingertip. There are even smart phone apps that can do the same thing by putting your finger over your camera's light and it'll read your heart rate. Measuring at the wrist seems like a marketing gamble because it's a lot more skin the light needs to pass through. If anything, I'd see if wearing the watch face on the other side of the wrist does better as people are designed for protection- with most blood vessels on the inner sides of limbs.

But I'd go for a chest strap for accuracy and reliability.

Like, I'd say there's little difference between a $50 optical heart rate sensor and a $700 optical heart rate sensor if they're both worn at the wrist. But a $50 fingertip sensor would be more reliable than either.

oHR works for me while inactive, but as soon as I go out on a run it becomes useless.
Garmin support finally got back to me but the after sales assistant basically just blamed the victim ("some people are not receptive to oHR", "I have the 5X myself and oHR works perfectly for me" yada yada). He did say I could send it back to the dealer for a replacement but to me there is no indication the sensor is faulty. The problem is how the watch is programmed to handle temporary periods of bad or no readings during workouts, i.e. Garmin tries to be clever about it with some best guess values instead of just showing and recording nothing ("--").

I can understand giving the option to only record oHR data when the oHR readings are good would probably expose just how many bad/no data lapses occur...
It would therefore be nice if we could at least have an option to auto disable oHR during activities (even when not wearing a chest strap) but keep it enabled otherwise.

Well, unfortunately the remark 'it just doesn't work for some' indeed is the case...

If you are darker skinned for example it is likely to not work as good for you. I have a friend for whom the tomtom watch just didn't work, we have about the same skintype and while for me it was perfect for him it was crap (I also had the tomom, I tried his and he tried mine).

Anyway, try wearing it a bit higher up the arm and tighten it even more.

The Fenix5 is quite a bit heavier compared to the 935 for example and more prone to moving around while running, even the slightest movement can have an influence, and being heavier it will be easier to slide a bit I assume.

Not sure if people are wearing them right, not letting it lock in before starting the workout, swinging their arms in such a way to throw it off, have a defective unit, or expecting it to work with exercise that requires wrist movement or flexion.

I use mine for power walking, running, and power walk/run intervals. So far it has worked better than any other OHR sensor I have ever used. I did 2 workouts comparing it to my Suunto Ambit 3 Peak with a chest strap, and it was never more than 1 BPM off and most of the time it was exactly the same.

I snug it 1 notch, give the OHR and GPS time to lock in (while I warm up before actually starting the activity) and go.

My F3HR gave me problems with drop outs and spikes, but this F5 has given me 0 issues so far and has been extremely accurate.

The Fenix5 is quite a bit heavier compared to the 935 for example and more prone to moving around while running, even the slightest movement can have an influence, and being heavier it will be easier to slide a bit I assume.

I have very little fat on my arms and need to strap it very tight to keep it in place, I would have thought more fat would make it easier as one can strap a little more into the skin?

Anyway, it's not about who gets how many bad readings but about how the watch handles it.
Garmin has an algorithm to determine whether the readings are good or bad (the blinking vs. solid heart at the start), why not use it during workouts to filter out bad readings?