Dear Garmin, why do you hate singletrack runners?

Former Member over 6 years ago

So let's get something out of the way. I love Garmin products. I've had every Fenix on day one, and I've been a Garmin user since the 310xt with a LOT of devices in between that and the Fenix series. I've got an InReach Mini, an Edge 1000, Varia display and radar, etc, etc, etc.

For the past 7 years or so, I've been a 3 day a week runner and a 3 day a week mountain biker for the most part. On my bikes I use power meters, wheelspeed sensors, etc. I love data. I use a lot of data.

One thing i know is that I can trust MTB's and wheelspeed sensors to know distance data when calibrated well (and I take the time to do that...first time I pair a wheelspeed sensor or move one I always let my watch get good GPS signal and then do a nice easy ride on a straight stretch of road). And I ride and run the same trails a LOT. And my running speed on those trails just doesn't change much.

So I can tell when the GPS recorded distance on a run sucks and when it's decent. For most of the Garmin product life, it's been mediocre. Usually 7-8% short on average, sometimes 10-12%. But I could swear that there were a couple Fenix5x firmware versions that were REALLY good. Like never less than 5% short, and seemingly closer than that. And then right before the Plus line came out, it seemed like the last 5x firmware sent it back to the 10% range again.

And now with the 5x+ I'm still seeing EVERYTHING I run on trail to be about 12% short, sometimes 15%. What gives? Yes, I'm on one-second recording. I've tried it with just GPS and with GPS+Galilleo. No difference. I always go into the activity at least a minute, often 2-3 minutes before I push start so it SHOULD be getting good GPS signal before the activity starts. But trail distance is simply always WOEFULLY short. And like I said, I know it is short because I'm running the same trails I bike and I know the distances. I also know I'm not NEARLY as slow in the woods as the device makes me out to be. On the road I can be VERY consistent at 9:30-10min/mile pace (yes, I'm slow as heck). I know on trail you can often be at least a minute slower than that, sometimes a full two minutes slower than that. But most runs the device has me at high 13's. No way, Garmin. No way.

Come on, get this stuff better. Trail runners deserve better in the flagship device. I know you can because I've seen it. There have been firmware versions that have been better, I swear.

--Donnie

0 dkatchalov over 6 years ago

That's interesting. Have you considered the fact that you have a bad unit & therefore should do an RMA?

Reason I say that is that I'm a trail runner myself. I've used Fenix 3 through to the Fenix 5X+ currently. At some point, the Fenix 5X got REALLY good with distance measurements. Yes the track would sometimes be off vs. the map but the distance & elevation would always be spot on.

An example I've used on these forums before - my last 5 ultras (50km x 2, 56km, 100km x 2) were all recorded on either a 5X or 5X+ and all measured within 400 metres of the actual course distance. My last ultra (100km with 4,500m of climbing) measured 99.4km and that's under heavy tree cover, lots of single-track, lots of switchbacks, etc.

0 Dr Phil over 6 years ago

935 works (usually) on trails for me. I've changed to GPS+Galileo to check it out and have seen some differences just a day apart. I'll give it another run or two before I switch back to GPS only. Look at the start and finish of these two. Less than 24hrs difference.
Good track
Not so good track

0 Former Member over 6 years ago

That's interesting. Have you considered the fact that you have a bad unit & therefore should do an RMA?

Reason I say that is that I'm a trail runner myself. I've used Fenix 3 through to the Fenix 5X+ currently. At some point, the Fenix 5X got REALLY good with distance measurements. Yes the track would sometimes be off vs. the map but the distance & elevation would always be spot on.

An example I've used on these forums before - my last 5 ultras (50km x 2, 56km, 100km x 2) were all recorded on either a 5X or 5X+ and all measured within 400 metres of the actual course distance. My last ultra (100km with 4,500m of climbing) measured 99.4km and that's under heavy tree cover, lots of single-track, lots of switchbacks, etc.

I suppose it's possible, but the main reason I was personally discounting it was because it seemed like to me my 5x got really good and then got bad again. I do remember the really good and being like 'DANG, they finally got this right!' But it didn't last.

And nothing else about the 5x+ I have seems off. I hike singletrack and it's fine. I have done a couple bike rides with no wheelspeed that even seemed closer to right than my trail runs.

What are your settings for your runs? Smart recording or 1s? GPS or GPS+Galilleo? Anything else changed from default?

I'm gonna be really pissed if it's hardware, that much I know. I still have my 5x...maybe I should run with both one day and see what happens.

--Donnie

0 Former Member over 6 years ago

If I wear my Fenix 3 (the only Fenix I have) with the antenna, i.e. the watchface, facing upwards, I get better distance consistency, compared to when I'm letting it sit as a watch normally does when I'm running. Which implies the antenna points all over the place, but far from always straight up.
My old Forerunner 310 had an angled design, allowing the antenna to aim upwards more of the time. Receivers have improved, but nothing beats a good signal from the antenna.
Bad reception in the GPS world usually means distances too long, not too short. But since they at Garmin of course are aware of this issue, my bet is that they have various more or less clever software, intended to detect and remove the oddities in the track. You know, where you go straight but the track suddenly deviates to the side, then back again. The problem with such software filtering is that you can't ever know for sure if the runner is actually not sober, but wobbling all over the place, or if it was a misdirected antenna that caused the wrong signal timing. Hence it's a delicate balance to remove as much detours as you should, but not too much.

Note that this is nothing I know for sure. It's just my interpretation of how the Fenix behaves, and possible as well as plausible reasons for that behavior.

0 dkatchalov over 6 years ago

I suppose it's possible, but the main reason I was personally discounting it was because it seemed like to me my 5x got really good and then got bad again. I do remember the really good and being like 'DANG, they finally got this right!' But it didn't last.

And nothing else about the 5x+ I have seems off. I hike singletrack and it's fine. I have done a couple bike rides with no wheelspeed that even seemed closer to right than my trail runs.

What are your settings for your runs? Smart recording or 1s? GPS or GPS+Galilleo? Anything else changed from default?

I'm gonna be really pissed if it's hardware, that much I know. I still have my 5x...maybe I should run with both one day and see what happens.

--Donnie

My F5X was set to 1s, GPS only. My F5X+ is set to Smart, GPS+Galileo. Results/accuracy pretty even across both watches.

One thing I always did is to wait around 60-90 seconds before starting my activity to let the GPS settle. I usually stand still and switch on the map screen to ensure my position is where I expect it to be and I am not moving (on the map). This indicates to me the GPS lock is good.

10-15% difference in your distances is not normal. I can't imagine getting a 45km distance reading in a 50km race or a 90km reading in a 100km race. I don't think I've ever had that on any Garmin watch I've owned. Definitely an RMA candidate if all other causes are ruled out.

0 nmyeti over 6 years ago

When I run single track with another friend with a 5x I almost never see big differences in distance. Elevation can be different depending on initial calibration.

0 Former Member over 6 years ago

Just hit me...could my RD-POD be causing it? Does it do any calibration with steps?

--Donnie

0 fbomba6 over 6 years ago

The solution to this problem might not be cheap but it's easy: https://www.stryd.com/
Apart from the interesting data about running power, the Stryd footpod is amazingly precise in all conditions, with no calibration whatsoever. I love it!

0 th3oretiker_ over 6 years ago

I think there are two possible reasons why single track distances are so bad:
- Newer watches "augment" GPS distances with wrist movement data. When you power walk uphill instead of running, your wrist barely moves and so the watch shortens the distance
- GPS is shortcutting serpentines

Both factors may even sum up and so you "lose" a lot of distance on single tracks.

For me, the solution was to buy a Stryd, too. If you are not interested in Running Power, you may consider a Stryd Live for half the price. A Garmin footpod however is totally useless for single tracks, as it requires proper calibration for different running parameters (speed, ground surface...).

0 Former Member over 6 years ago

Garmin's RD-POD isn't a footpod. You wear it on the back of your shorts. But it does do some other things and may muck with the overall data. I hadn't thought of the arm-swing thing, but I do think that might have something to do with it. I'm a bigger-than-average dude (for endurance sports, anyway) at 6-2 and 210#.

--Donnie