Anyone bought a Instinct 3 50mm yet? Can share experience?

I did not use Garmin Pay yet on my I3, so cannot really tell how fast the payment is, but generally, the I3 AMOLED model is waaaay much faster and more responsive than I2. I3 Solar still feels a bit laggy at times, though less than I2. 

And if anyone is interested about the accuracy of the HR sensor, then I can show some data too. Yes, the watch seems to have the older Elevate 4 sensor, but what I saw till now, is rather impressing anyway. For example today I used the Instinct 3 AMOLED and the HRM-Pro strap to record my run. It was a ½marathon over a hilly terrain, and with some sprints inserted during the run, so the HR is really changing wildly most of the time.

Additionally, it was rather cold today, freezing temperatures, and I've been running just in a t-shirt with long sleeves, and without gloves, which is usually not quite optimal, since the cold-induced peripheral vasoconstriction makes the HR detection quite challenging for the optical sensor, due to the lower blood circulation, and often the sensor easily locks on the cadence, in winter. Also, I started the run without any warm-up (warm-up is normally recommended to avoid the cadence lock).

However, when I compare the HR curve from the internal sensor (red), with the data from the strap (blue), using https://www.fitfileviewer.com/, both curves are practically identical - you'll barely see any red parts of the curve, because it is exactly under the blue line:

Only once I zoom to the max, and look only at the first 6 minutes of the run, two minor differences become visible - it takes the first 10 seconds before the signal from both sensors merges - the initial difference was 10 bpm. Not really surprising, since I've been manipulating with the watch, when starting. And then, around the 4th minute, there is an abrupt change of HR coming from the strap, causing some difference to the internal sensor for some 20s. That was in the time I was adjusting the chest strap, since it slipped a bit too much during the initial descent.

Very helpful info - thanks!

BTW, that told, I have to mention also that unlike some other people who used to complain about the accuracy of the internal HR sensor at I2, I rarely had any serious problems with I2 either (though it was not as accurate as now here). So perhaps I am just a lucky person for whom the optical sensor is as good as the HRM.

And also - for any serious training, the HRM is unavoidable, anyway, because you also need accurate HRV, and that's practically impossible with a wrist optical sensor, while moving.

Could you compare HR accuracy between Garmin Instinct 3 and chest strap in passive monitoring mode?

All HR accuracy comparisons between optical sensors in Garmin watches and chest straps I was able to find are made when activity is recorded in sport mode. I am interested in finding out how accurate Gamins are in continuous HR measurement mode. Could you record e.g. 30 minutes of HR data using chest strap and compare it to the HR data from Garmin app but with no sport mode on? E.g. few minutes of sitting, standing up, few minutes of walking around the house doing some chores, maybe walking up the stairs, etc.

Adam Adam said:

Could you compare hr accuracy between garmin instinct 3 and chest strap in passive monitoring mode?

Not easily, since in the passive monitoring mode, the strap is not connected and the watch does not log the HR from the strap. I would have to record the strap data with a 3rd party app. And also, when the watch is not in the activity mode, it records only averages over 2 minutes samples, hence any peeks and lows will be smoothed out, making the comparison also less accurate.

I'll give it a try though, when I find some time.

Here you go, I recorded 1h20 of varying HR on two separate GC accounts, and superposed both graphs on one image below. The thicker dark curve is the all-day data from the internal sensor of Instinct 3 Amoled, and the fine red curve comes from HRM-Pro on the other account.

The curve of the internal HR sensor consists of only 40 data points (2 minute averages), while the curve of the external HRM contains 1600 data points (GC plots 3s averages), hence the red curve is much finer, while the dark curve is very strongly smoothed, removing all rapid changes, and the peeks and drops.

The peak on the right side was 139 bpm on HRM-Pro, and rather short, hence averaged to 130 bpm on the watch.

The steeper rising of the internal HR on the right side is likely caused by the cadence lock. It corresponds to the moment I started to jog lightly. When the watch is not in the Walk or Run activity mode, it does not apply activity-based filtering of the noise signal of the cadence. It means the cadence necessarily disturbs the HR signal stronger than when recording a corresponding activity. Lower sampling rate and lower led power in the non-activity mode contribute to the noise sensibility too.

Thank you. I encourage you to post this results on a Garmin Instinct reddit. I bet many people would be interested to see your work. For example, there are already people showing comparisons of Instinct 3 GPS accuracy. 

Would it be possible to export raw number data to a file and post it somewhere, e.g. Google Sheet?

Thank you. I have found similar test but author somehow was able to record real time HR data from sensor in the non-activity mode to avoid 2 minute averages.

www.reddit.com/.../

Adam Adam said:

I have found similar test but author somehow was able to record real time hr data from sensor in the non-activity mode to avoid 2 minute averages.

He apparently used the HR Broadcasting on all of those devices, hence not in the passive mode, because when HR broadcasting is on, the sensor is in the same mode as during an activity (minus the activity-specific noise filtering).