make sure you set data recording to “every second” in the system settings menu. I bought one because I wanted a watch that gave me temperature, which I think is an important variable for runs that I didn’t get with my 220. I took it for a couple runs and accuracy seemed fine. But then I took it for a technical trail run, and it was awful. 500m short over 10km. Unacceptably short. I did a quick google, and it seems that EVERYONE who has the Fenix 3 finds that it measures short. So I thought, “I wonder if it’s the smoothing algorithm”? So I turned off dumb aka “Smart” recording and set it to 1sec. Now it’s superb.
I checked in with Ray Maker, and his suggestions mirrored my own intuition: - make sure GLONASS is on and (biggie) make sure data recording is set to 1s.
My guess is that the Fenix collects a bunch of additional data so the files would be a lot bigger if the default recording was more often, so Garmin just, foolishly, focused on file size rather than accuracy. I really think that 1sec recording, especially for something billed as a trail watch, ought to be the norm. Garmin has run into countless problems with their stupid smoothing algorithms, so you’d think they’d just figure out that they should get rid of them and just make 1s the only option along with the “long battery life” option that’s specific to the Fenix when being used for ultras where you need extra long battery life.
Anyway, I was ready to return mine. But now it’s great.
I have no affiliation with Garmin. I’m just basically stuck with them after using there stuff for so long…
Lots of discussion on this on the Garmin forums. The general consensus seems to be that even with 1 s recording, it loses accuracy when running in heavy tree cover or around tall buildings. Works fine mounted on a bike or in open space. I suspect the antenna has a hemispherical pattern and thus degraded performance when worn on the wrist. Since I live in SoCal, it doesn’t really impact me and I like the watch. But my test runs on roads in forested areas left much to be desired.
They did just add the feature that you can use a footpod for pace while simultaneously using GPS for distance and track generation. That makes a big difference for me as the instantaneous pace is now a useful field.
I don’t know what you see, but I found that the one thing that is less accurate than the GPS is the temperate sensor. It always reads warm and I suspect it is influence by my body heat (works great in the water though). I suggest looking into the Garmin’s temperature sensor (tempe) if temperature tracking is critical to you. Plus putting a tempe on one shoe will balance out the footpod on the other.
Anyway to anyone interested in trying the Fenix, I highly suggest REI or somewhere that has a similar return policy.
Wonder if the 920 suffers from the same issue. Went for a long run today and while the course I had plotted previously measured 25 miles, the watch said 22 miles when I was done. Weird thing is the distance was correct when I checked the Garmin Connect website.
Does this also help with pace when under tree cover? I noticed during Mt Tremblant that the real time pace sucked when on the trail in the woods.
Spot
I found that the pace on my 220 was off as well under heavy trees in MT. But I also use a footpod, and am using it with the Fenix 3 also, and that does seem to help. No wrist-mounted GPS is going to be great under heavy trees… Sadly.
Lots of discussion on this on the Garmin forums. The general consensus seems to be that even with 1 s recording, it loses accuracy when running in heavy tree cover or around tall buildings. Works fine mounted on a bike or in open space. I suspect the antenna has a hemispherical pattern and thus degraded performance when worn on the wrist. Since I live in SoCal, it doesn’t really impact me and I like the watch. But my test runs on roads in forested areas left much to be desired.
They did just add the feature that you can use a footpod for pace while simultaneously using GPS for distance and track generation. That makes a big difference for me as the instantaneous pace is now a useful field.
I don’t know what you see, but I found that the one thing that is less accurate than the GPS is the temperate sensor. It always reads warm and I suspect it is influence by my body heat (works great in the water though). I suggest looking into the Garmin’s temperature sensor (tempe) if temperature tracking is critical to you. Plus putting a tempe on one shoe will balance out the footpod on the other.
Anyway to anyone interested in trying the Fenix, I highly suggest REI or somewhere that has a similar return policy.
I found the same with my 220 when I run near my parents in NY. I just sort of accept that. I do use a footpod, though, which seems to help. And I am using a footpod with the Fenix also.
The big issue with the temp sensor seems to be airflow:
sitting around the house, it always reads WAY high
riding the bike, it’s bang on with the sensor on my Edge, as accurate as that is.
running it’s much better than standing around and, in my experience, is very, very close to as accurate as it is with cycling. But I picked up a tempe unit and will compare it when running to the watch by itself. I suspect it is affected here by what wrist you wear it on. If you run with it on your left wrist, the air (temperature/altitude/barometer) sensor is towards your hand. That means that as you run, you are pushing “clean” air over the sensor regularly. If you wear it on your right wrist, then the sensor is in a dead zone a bit more with regards to airflow…
I’ve got a question for my fellow fenix 3 users, not about GPS accuracy however, but speed and cadence accuracy.
I was playing around with my bike on the trainer to compare speed and cadence accuracy against my old Polar RS800CX. The new Garmin sensors seem to randomly cut out for anywhere from a few second up to maybe 30 seconds then sync right back up. While sync’d the reading between the two were virtually identical, but for the life of me I cannot figure out why the Garmin sensors keep dropping out. Has anyone had this or a similar issue?
I haven’t noticed the sensors dropping out on the road.
if the sensors are not dropping on the road, my guess is that you don’t have them mounted close enough. On the trainer, all of that stress of riding goes into flexing the frame. On the road, the bike just shifts back and forth on the road. So the frame is certainly flexing more on the trainer. Perhaps it’s flexing enough to move the speed/cadence sensor out of proximity of the magnet?