Problem with distance and pace measurements on Instict

Hi, this is typical in my experience and as I and trux mentioned this is because of sweat getting in the sensor hole.  You didn’t see it on your hike because you didn’t sweat as much (would be my guess?).  Hill repeats will certainly get you sweating.  The tape will prevent this.  Do the same run with some tape over the hole and see for yourself.

Okay, thank you. I will try. 

For the hike I wore the watch over my long sleeve jacket. 

I only turn heart rate on for running.

I realise heart rate is probably the best indicator of effort, but I really don't use it to train. Maybe if I was more serious in the future and the issue is a problem I could wear the watch on top of a wrist band and then get HR from a chest or arm strap.

I will try taping up the hole, then if that doesn't help see if I have a wrist band around and try that next. 

Todays run I used the gaffa tape method and the elevation graph looks good, no spikes or drop outs. 

Thank you for the tip and help. 

Will have to see if it holds up to the hill repeats. I ordered some sweat wristbands so if there's an issue I can forgo the heart rate sensor data and place the watch on it's own sweat band throne.

It appears that this thread has morphed into an elevation/sensor discussion.  I continue to have good luck when using a wrist band between watch and hand when wearing the watch on my left arm.  This permits the wrist OHR sensor to work, though I typically use a strap when running.

Regarding the OP's issue with distance measurement, I always have this problem on trails with switchbacks in the woods.  The Instinct consistently under-measures distance by about 10%.  Back when I was more interested in finding a solution, I noticed similar complaints in the Fenix 5 forums.  As the Instinct and Fenix 5 series use the same GPS, I had intended to find out if a fix was ever made for this on the Fenix.  I lost steam on this and just live with the poor results.  But, it might pay to see if Garmin ever fixed the issue for the Fenix 5.

GeoffR said:

The Instinct consistently under-measures distance by about 10%. 

Could you please share a link to one of such activities? I analyzed one already in the past, but cannot find the concerned activity with the drop-outs anymore, and the couple of ones that I checked now, they all look all right.

GeoffR said:

As the Instinct and Fenix 5 series use the same GPS, I had intended to find out if a fix was ever made for this on the Fenix.

I do not know about F5, but earlier I participated iin a similar thread on the Fenix 6 forum, and the last time I checked (3 weeks ago), several users reported a significant improvement, though there was another one who still wasn't quite happy.

From what I observed on the activities I saw, the problem was not really with the GPS accuracy, but with the distance calculating algorithm. Namely, when the GPS signals drops (which is, up and then, inevitable even with the best hardware), the algorithm, instead of connecting the last known keypoint with the new one (interpolating so the distance), completely skips the distance of the missing signal, which is evidently the wrong way.

As I wrote, currently I cannot find an activity among my own ones, where there would be some GPS dropouts, and where I could test whether the distance algorithm is still wrong, but if you have one, I'd be interested in seeing it.

When I'm in open territory (hiking above treeline, cycling on roads, running on streets, all with minimal switchbacks) I find the GPS derived distance to be pretty accurate.  My hunch is similar to yours: the distance algorithm under measures when there are drops. It seems Garmin could make better use of the accelerometer to estimate the distance when GPS coverage is spotty.

As I recall, this was the problem the Fenix 5 group was chasing (1-2 years ago).  Fenix 6 and all the new Garmin devices use the Sony chipset which I'm sure has different issues and uses a different algorithm.  I guess I lost interest once the new devices came out, as I figure Garmin will be largely focused on dealing with issues on those devices.

The problem with any tracks I provide is they will not have precise actual distance measurements for comparison.  They are in heavy tree cover with "common knowledge" distance derived from trail maps, older GPS tracking results, and phones which may have more power and better sensors to deal with the problem.

Ultimately, I decided that if I wanted more accurate distance tracking while trail running, I'd need to add a foot pod.

GeoffR said:

Fenix 6 and all the new Garmin devices use the Sony chipset which I'm sure has different issues and uses a different algorithm.

I do not believe there is a direct relation to the GPS chipset. The situation may be perhaps better with different chips, because there are then less signal dropouts, but on my mind the track distance is being calculated in the main firmware, and quite likely the same faulty algorithm is used for most Garmin devices.

GeoffR said:

The problem with any tracks I provide is they will not have precise actual distance measurements for comparison. 

I do not need that. I just need an activity with some short GPS drop outs, done with the recent firmware. I do not have any such, only with an older firmware, but if I analyze those, and report it to Garmin, they won't even bother looking at it, hence I need a fresh activity. I can review the raw data, and if the problem is stil there as it was, I can very clearly demonstrate the fault of the algorithm, and send a bug report to Garmin.

Sorry to be off topic again, the tape solution isn't working during sweaty workouts the sensor still gets blocked. 

When you say wrist band, do you mean one of those thin silicone/rubber ones, the type that charities and festivals use?

Looks like I will have to mount the watch on a sweatband if I want the altimeter to work correctly when running, and just forgo the heart rate monitoring or buy a chest strap.