Fenix 6X : critical underestimation of the distance

O.Ret. said:

I had the chance to run with 4 watches today (2 * 2 ) on what I think is a 13km run

First step should be finding out exactly what the actual distance of the run is. I hope you're not just using measurements from other GPS devices as the basis.

As for the accuracy, I agree that those tracks don't look great, but without a scale on the image you posted, we can't tell exactly how far off it is. 2 meters? 10 meters? 100 meters?

Tracks are going to be off by as much as +/-10m just because of the satellite chips we have available for wearables today. All wearables use single-band GPS receivers, because nobody makes a dual-band chip for wearables. Sometimes you'll get lucky and it'll be spot on. Sometimes you'll get less lucky, and it'll drift by 10m. If you need it to be more accurate than that, then a wearable is the wrong answer for you, you'll need a larger dual-band GPS device.

Your F5 happened to match the actual path more closely in the instance in your image you posted, but does it always? I doubt it.

O.Ret. said:

Has anyone been able to get a reply from Garmin on that issue ?

No, because I haven't contacted them, I'm perfectly happy with my distance measurements and my tracks.

FYI, data recording to 1 second just dictates the size of the data file that's uploaded.  Smart recording also pings the satellites every second but if you're going in a straight line it doesn't record the point to the data file.  This is just to reduce the size of the transfer out to your GC activity, it doesn't make any difference in accuracy or lack of accuracy of the distance collection. 

GPS is primarily impacted by reflected signals from buildings and vertical terrain that causes ghost sources.   

FWIW, I just watched three different 6X's measure the last 40 miles for me to within .25 miles, the last 60 miles for one friend to within .4 miles and for a third the entire 100 miles to within .75 miles during a recent 100+ mile endurance event we ran.  Those distances also include side trips off the certified course for bathroom, meal tent and personal tent which added a little extra.  Two of us used Stryd's for distance and pacing till their batteries ran out.   

I'm not saying yours isn't having a problem in your terrain.  For us in our terrain, no high buildings or hills so no secondary reflections and minimal trees, the measurements were pretty spot on for very long distances over up to 48 hours of an activity.   

Excuse me for being dim, but I don't get why you're so bothered about the distance between the GPX and fit files? There are some apparently large differences in the distances between the two watches anyway that I'd be more interested in, especially if that is repeatable. 

On a recent 10km run with a F6 and a 5+ mostly on trail on undulating single track and open path I ended up with this on a F6 Sapphire

and this by comparison on a 5+

I've been running this route for years and the distance is always the same give or take a few metres, and that's going back to a 310XT, 910XT, 920XT, 935, or 945.

Would there be a difference in the .gpx files compared to the .fit files? Very possibly, but why does that matter? We know Garmin post-process the received data. So what? as long as the results are consistent I really cannot understand why I should be bothered. And no, the course has not been measured with a calibrated wheel. 

I also posted in this thread - https://forums.garmin.com/outdoor-recreation/outdoor-recreation/f/fenix-6-series/211485/fenix-6-battery-and-hr-issues?ReplySortBy=CreatedDate&ReplySortOrder=Descending

about multiple runs on the same course over the years - 2012, 2015 and 2019/20

Bonjour,

thanks for your reply.

I may not have been cear enough, the data I'm showing has been recorded in the same time : 6X on the right wrist, 5x onthe left and 5 on my wife's wrist (+6s recently).

Indeed I'm not looking for perfect accuracy on the map, especially that in the city with building it is quite complicated. But at least I'm looking for consistency in the distance. I can afford having a 1-3% error but more is starting to be annoying.

For a few run with both watches, i'm seeing that if the GPX files are pretty the same in terms of distance read through GPX reader, there is consequent difference in the distance that Garmin is reading. the 6X algo is downsizing considerably the distance in my case.

That may be related to my specific watche (I sent anyway an email to Garmin).

If we look at that one : the GPX are issued from Garmin Connect for each of the file, and then the FIT as well.

This is the same run : 6X my left wrist, 5X my left wrist, 6S my wife left wrist, 5 my wife right wrist.

The GPXs are in line to me, small deviations but very close. Largely affordable. Hence the issue is probably not the GPS recording.

F5 & F5x are giving 12.8/12.9 km, again very close. F6S 12.3 & F6X 12. Thta is quite a consequent difference : almost 7%.

On another run, again 5X right wrist, 6X left wrist, 5 on my wife wrist (at this moment no 6S), we have the same kind of reading. The difference between 5/5X and 6 is now 1km for a 10km run.

It is 10%. Far too much to be normal.

I'm really struggling to see ths as normal.

In urban canyon areas, +/- 5% is a pretty good result IMHO. Have you got 3D distance enabled on one watch but not the others? You're asking too much from a tiny GPS receiver that's fastened to a moving object, that's also moving. Also, if you took all these watches on the same run, there's an excellent chance that they interfered with one another (DC Rainmaker has experienced this problem).

Also you still haven't clarified why you think the distance isn't accurate, other than that it doesn't match the GPX file or the fenix 5. Have you measured what the distance actually is using anything other than another GPS device?

Most people report that the distance reported by the F6 is generally spot on, give or take 1-2%. So unless your firmware is broken somehow, it seems unlikely to me that it's off by as much as you think it is.

So perhaps before you spend any more time trying to solve the problem, you should verify that there actually is a problem first.

Bonjour,

measuring using basecamp drawing gives 13.1km. Let's say +/-1% error, so ranging from 12.95 to 13.25.

F6X gives 12.26 and 12.00 while the related GPX indicate13.08 & 13.24.

F5 & F5X give 12.76,12.89, 13.12 (and 13.00 another time), related GPX indicate 13.16, 13.08,13.39.

The fact that there is so much difference for the F6X, 1km & 750m, hence 7.5% & 5.7%, while there isn't that much on F5 &F5X is making me skeptical.

Am I missing a setting on the 6X ?

O.Ret. said:

7.5% & 5.7%,

That error while on the higher side then I've seen is still within what is usually accepted as reasonable for a wrist worn, consumer GPS device. However, given that we have come to expect and to get, better precision, I would be looking at collecting more data points for the same run to get a better handle on the degree of variance.

I do have data for similar runs going back a number of years recorded on different watches. While I've never really looked to see if there is a specific watch that has consistently under or over measured, the results have always been within 50m or so. 

I think you need to move away from concerning yourself about the difference between the .gpx and .fit files and perhaps have a conversation with Garmin about the differences in what is being displayed in Garmin Connect or the watch. After all, it's the .fit file that is being used to produce the output metrics. The .gpx file is produced FROM the .fit file I believe.

O.Ret. said:

Am I missing a setting on the 6X ?

Just verify that you don't have 3D distance or 3D speed enabled on any of the devices (F5 and F6). It is available under the app settings.

I never used the F5 series but used a F3 for four years and my F6x Pro is very close to the distances I got with my F3 and when comparing to friends who still use F3 we get almost identical distances.