• State Not Answered
  • Replies 11 replies
  • Subscribers 4 subscribers
  • Views 5091 views
  • Users 0 members are here
Options
Related

GPS Performance Error - Distance Under recording Garmin Instinct

Tom
over 4 years ago

Since the Garmin Instincts release, there have been complaints about its under-recording of distance.  

My own experiences with the Instinct have been that it performs well in good GPS conditions such as open countryside, but struggles as soon as tree cover, hillsides, valleys, walls, hedges etc are introduced.  The more time spent in these conditions the further the Instinct under records distance.

I have compared my Instinct with my Ambit 3 Peak over 60 walks, hikes and trail run where it tends to under record compared to the Ambit by 2-3% sometimes more, sometimes less, depending on terrain. Yes, I am aware of the limitations of GPS technology, but the Instinct underperforms, when tested against other GPS watches in recording distance.

"Garmin Instinct Review THE GPS Summary: After finally completing my formal test I found that the Instinct’s GPS performance in that test was mediocre. In easy GPS reception conditions it was good but when buildings or trees were involved the GPS performance slipped notably and the overall score was 65% and a good Garmin on a good day would normally give up to 79%, with other, non-Garmins scoring even higher. So here we are typically talking about positional accuracy often being more than 5m away from where it should be and some times a lot further away."

the5krunner.com/.../

Until recently I just assumed it was the antenna or chipset, but after running data through a comparison site  https://quantified-self.io/ and Comparing it to a plotted route and the Ambit, I am convinced its an algorithm the Instinct uses that is responsible for under-recording data.

Here is an example from today. Plotted route and the Ambit were extremely close, with the Instinct falling quite a distance behind.  This was a route that had tree cover and went to the bottom of the valley by a river.

However, if you look at the GNSS recorded distance, the Instinct was very close to the actual distance.  Frustrating as when running 10 miles, 3% under-recording means having to run an extra 480 meters which has a large impact on lap times and lap pace.

Is this issue something Garmin can fix as I have read lots of comments from people who returned the Instinct because of under-recording over known distances?

Antanna design/placement issue, GPS Chipset issue, Algorithm problems? 

  • 0 over 4 years ago

    When viewing the GPS tracks overlaid on the same map the Instinct is actually just as good as the Ambit for most of the route.  The underrecording of distance does not reflect how well the Instinct performed, meaning something else is at play.

    instinct in blue, Ambit in Orange.

  • 0 over 4 years ago in reply to Tom

    The Altimeter's performed almost the same, both over a jacket so no sweat to jeopardise the sensors.

    As you can see the lap mile markers get further apart as the distance increases, which really makes no sense as the Instinct was actually on the better side for GPS coming back and should have made the better track and closed the gap not increased the distance it underrecorded. 

  • 0 over 4 years ago in reply to Tom

    Nice posting. Especially the elevation vs time of day comparison plot. I think the watch is over-filtering the GPS data for noise reduction. It would be nice if they gave us back some control over it - I remember on my first 1 or 2 watches "smoothing" was actually a setting. Seems like it should definitely be "fixable" by updating the activity profiles, such as low filtering for suburuban road runs, high filtering for trailrunning, super high filtering for open water swim I also think they use the watch accelerometer as part of distance calculation, but having a hard time isolating it, but you can go start an activity, turn off the GPS, and it will still accumulate distance. If you want to play around with some filtering on your own to see the effect, check out www.gpstrackeditor.com. 

  • 0 over 4 years ago in reply to carleson

    Thank you, I will have a look.  The Instinct definitely does have some genuine struggles in bad GPS areas as well as the GNSS distance being reduced aggressively.

    I went on an 11 mile hike last week and over the 1st Mile which was poor GPS conditions,the Instinct was 7-8% behind the Ambit.  sometimes my lap speeds are 2/10th of a mile slower on the ambit and as much as 1 minute time difference per lap times,pretty poor if you are busting a gut to improve your speed

    .

  • 0 over 4 years ago in reply to Tom

    The Instinct GNSS actually recorded more distance,but then reduced it to less than the Ambit.

    looking at the tracks you can see the Instinct

    cutting corners and not recording turns.

    Instinct in red,ambit light orange,plotted route darker orange.

  • 0 over 4 years ago in reply to carleson

    The Grade graph is also included o. that comparison site for watches that support it.

  • 0 over 4 years ago

    I wish this thread got more attention. I have an instinct since 2019 but only use it when I am in the mountains when on holidays because then I really need its features and I can live with the distance discrepancy. However, during most of the year I am training for long distance running at my athletics club and the Instinct is pretty useless for it because of the under recording of distance. All workouts we do are based on pace, but because the Instinct always measures short my pace is always wrong. For me as a distance runner it does matter whether I am running at 3'50"/km, 4'00"/km or 4'10"/km.  

  • 0 over 3 years ago

    I experience(d) the same with my Instinct, comparing it to a Fenix 6 and an Apple Watch Series 5:

    distance travelled underreported between 0,5% and 6% on the watch and on Garmin Connect, but if you look into the GNSS distance (via the GPX file), it becomes even against those devices. That´s surely some magic that happens there, but it´s not the magic I want from a GPS watch.


  • 0 over 3 years ago

    Definitely agree.  Every race I have run (all of which will have been measured accurately) has come up a bit short on my Instinct.  Never had this problem with my old TomTom.  I have got used to it now but it is a bit poor especially considering your analysis.  I wonder if Garmin do it on purpose to partition the market i.e. you have to buy a Forerunner for running and an Outdoor variant (e.g Instinct) for hiking and so on.  Do the Forerunners have the same issue?

  • 0 over 3 years ago in reply to 3854086

    The older models with the large antenna nub, that were optimised solely for sports seem to perform better such as the Polar V800, Suunto Ambit 3 Series and a few TomTom watches.

    The demand for a smaller watch that could be worn all day means the antenna has to be smaller to fit in the bezel and has to cope with a more complex range of fitness tracking and power management.

    "Regular readers here will know that I am certainly NOT an apologist of the mediocre accuracy of Garmin devices, especially when it comes to their GNSS/GPS accuracy."

    "The old Polar V800 and Suunto Ambit 3 were clearly the best-performing GPS sports devices, admittedly you could almost say they were from a different era."

    "The better-performing GNSS chips of yore were branded as SiRFstar. Maybe they were better-performing because they had better aerials or maybe it was because they used so much power to fix a decent lock onto a good signal."

    "However, that change did not seem to filter through for me with better running GPS performance. Maybe it was sometimes better, maybe not. I’m not quite sure. I just know that I’d kinda given up with my Garmin Forerunner 945 and instead relied on STRYD for speed and distance and an HR strap for HR."
    the5krunner.com/.../

    I don't think the forerunners fair much better.

    Forerunner 745
    " Garmin does seem to have improved GPS/GNSS accuracy in some situations with the Forerunner 745, most notably in Open Water Swimming and to an extent cycling. However, the general GPS experience is variable and sometimes worse than found in competitive products, yet confusingly superior at other times. That said, I rate the Forerunner 745‘s accuracy as ‘broadly acceptable overall‘."

    the5krunner.com/.../

    The Forerunner 645 which comparison to other models.

    "BUT there is NO WAY a 69% would ever get above 80% and beat the Suunto/Polar devices shown there.

    Superficially I would say that I was surprised that the 645 only scored 69%."
    the5krunner.com/.../

    "Every single device has had some problem or other so far on the shorter tests. Yes even the AMBIT3 and V800. Initially TomTom’s Runner 3/Spark 3 appeared as the best performing (although that has since been revised and superceded)"
    the5krunner.com/.../

    A bunch of GPS watches tested for accuracy here.

    "I expected GPS watches to improve with time, but the opposite appears to be happening. With the Garmin devices especially, you can see that the older watches generally do far better than the newer ones. I suspect this is due to compromises to get better battery life and smaller packaging and the cost of GPS accuracy."
    https://fellrnr.com/wiki/GPS_Accuracy

    DcRainmaker commenting that peak GPS accuracy was the Ambit 3 days.