Original Review date: June 2016. Tested with original Garmin Run HR Strap and 4.40 firmware
The fēnix 2, this is the model that actually kicked off our reviews in early 2015. While it promised much and got great reviews from the gadget and unboxing review sites, it is probably the worst GPS for trail running ever. Which is precisely why we designed tests to see if what’s on the box is actually delivered – because it seemed nobody else bothered to find out if these things actually work or not.
That said if you don’t run trail, it may suit your needs fine. We actually know someone who is very happy with his fēnix 2, he doesn’t run trail. We don’t test for non trail runners so can’t really comment there.
In short this isn’t worthy of a full review – the fēnix2 feel like a half baked project that someone gave up on but got delivered anyway.
What kind of trail runner is it good for?
Not a trail runner. Period. Don’t even accept one as a gift.
What should you expect as a long term running companion?
It’ll let you down, again and again. And again. It’ll look nice on your wrist, but break your heart.
Uuumm. We like the red backlight? The charging clip is better than most of it’s competitors.
The interface is unbearably slow. Accuracy is generally marginal, and pretty poor regarding elevation. It reboots randomly during runs. Normal battery life is bad and the endurance UltraTrac mode doesn’t work at all in the trail conditions where you’d likely use it. If you do get UltraTrac working you won’t be able to record heart rate or from any ANT+ device. You couldn’t connect to ANT+ and Bluetooth simultaneously, and the Bluetooth smartphone connect would eat the battery for breakfast.
Plea to Garmin
Please don’t ever release anything this bad again, you had to be aware of the issues before you put it into production. If this absolutely had to be released to market, at least don’t pitch it to a market where it simply couldn’t perform.
fēnix 2 testing covered in excess of 160km on the surveyed courses over 19 separate days (repeated testing is vital to capture variation). A number of the activities were also specifically planned to assess the impact of poor satellite coverage (via tools that give geometric dilution of precision or GDoP). Potential GDoP was logged for every activity to better make comparisons across models.
Testing settings were generally undertaken with the best data recording setting
The GPS pool against which the fēnix3 was tested included the following models: Suunto Ambit2, Suunto Ambit3 Peak, Polar V800, Garmin FR310XT, Garmin fēnix 3.
During testing multiple GPS devices are carried as a means of validation, if all units have a bad GPS day then a tin foil hat is donned. Though some days with poor GDoP potential saw all units a little degraded, no testing days saw GPS chaos across all models.
So the fēnix 2 wasn’t horrific all the time, but it was certainly the worst performing of all the running GPS watches we’ve tested to date (and hopefully for all time).
Under moderately difficult GPS conditions the median positional error under these conditions at 3.0 metres (average 4.1m) And 6% of points were 10 or more metres off track where better performing models had <1% greater than 10 metres off track during testing.
When the fēnix 2 was taken into genuinely difficult conditions (heavy tree cover, steep terrain, gorges), a slaloming repeat track was the outcome with a median track offset of 7.7m (3metres more than the best performer).
Variance in accuracy was also unsurprisingly high, beaten only by the early fēnix 3 firmware and FR310/910XT in limited sky conditions.
Assessing true distance accuracy is not easy in trail conditions. GPS measured distance is dependent on positional accuracy and sampling rate. Unfortunately errors in positional accuracy can both increase and decrease distance via corner cutting and random scatter, and both are common in trail conditions. As these can cancel each other out an inaccurate track can still give spot on distances. A third type of error, shadowing, where the track is parallel generally has little effect on distance (unless it’s well off and converges).
The fēnix 2 is the only model we’ve tested to date that has a long term average in excess of 100% of true surveyed distance. As the plot shows, the most common outcome is actually 101% of the actual distance. That said, on any given day it was actually pretty respectable with 67% of recorded distances within 1% of the true distance. Nice, but not good enough to save it.
Tracks under reasonable GPS conditions were pretty varied, some rather nice, following features and changes in direction closesly. Others not so much.
Trackpoint cloud image would be here if we made more of an effort (and we would have if it was worth it).
We only tried UltraTrac a couple of times in moderate to difficult GPS conditions, it didn’t work. By which I mean trackpoints dried up after we started it. As it prevents heart rate being recorded we don’t see the point from a runner’s perspective.
Capture of total climb at 77% is down on all other devices save the forerunners.
Battery Saving Data Recording
As mentioned the UltraTrac didn’t work for us at all.
Data Syncing and Sharing
The sync/power cable design is great, as is the Garmin Connect service from a sync and share perspective (see other Garmin reviews for more details).
Though syncing from the fēnix 2 is pretty slow though and some recorded runs seemed to get eaten and vanish before uploading. We didn’t even try to upload via Bluetooth.
We’d pretty much had enough by now. The box says it does stuff. Maybe it does. Maybe it doesn’t.
Does the constant pace thing, tells you if you are ahead or behind on a pretty screen.
Battery Run Down
You’d be real lucky to get 13hrs from this in real world trail conditions. Often it’s less. Perhaps that’s why it likes shutting itself down after a couple of hours?
Standard Feature Set
General Trail Running
- GPS accuracy under canopy : Not really
- Consistent GPS accuracy : Not really
- Rapid GPS Acquisition : Poor
- HRM : Yes
- Cadence option : Yes
- Battery 8hr with HRM and high accuracy setting : Yes
- Barometer : Yes
- Breadcrumb with waypoint navigation : Don’t know
- Vibration alerts : Yes
- Trail legible display: Yes
- Open data access : Yes
- Fully ANT+ or BLE Compliant : Yes
Ultra Feature Set (as per trail running plus)
- Battery 14hr+ with HRM and high accuracy recording : No
- Battery 24hr+ with HRM and down-sampling : No
- Electronic compass : Yes
Nice to Have Features
- Mobile uploads : Maybe
- Cadence (without footpod) : Yes
- HRV (R-R) recording with recovery estimate/test : Yes
- Footpod GPS override : No
- Basic interval workout ability : Yes
- Pacing function : Constant pace only
- Position/waypoint autolapping : Don’t know
- Custom timer reminders : Yes
- Everyday watch : Yes
- Activity Tracking : No
- Mobile Notifications : No worth the battery drain
Non-Tested Core Features
While functions like heart rate and cadence are core, either we’ve got no means of objectively testing them. So we didn’t for the fēnix 2 (you might be getting the impression we have generally given up by now).
Unit itself has a nice of chunky look and feel. We luck the textured buttons. Screen is legible, beep audible, vibrate good enough for us. The menus are slow, saving a run is very slow. When paused you can’t scroll through to read the time. It’s not a joy to use.
Long Term Verdict
So many freezes and shut downs whilst out. Not all runs can be successfully saved and uploaded. Battery lasts well short of stated specifications. UltraTrac doesn’t allow heart rate and doesn’t work in trail conditions anyway. Enough said.
Not sure what Garmin were thinking in releasing this as a trail capable device, that people wouldn’t notice these things? Then the early fēnix 3 got off to a rocky start – initially was even worse thant the fēnix 2 from an accuracy perspective (though it improved enormously the fēnix 3 still won’t win any accuracy competitions).
Looks like Garmin didn’t have too many qualms about releasing imature products to market in years past. The feature list is certainly more marketable than accuracy and performance. Let’s hope they’ve changed their approach on current models.