GPS Accuracy

That's good to hear, seems like you're still keeping an eye on it with the M400 though ;-)

The GPS is decent, but when the watch is covered by a sleeve, it goes nuts, even when it is in GPS+Glonass mode. Anyone else encountered this problem? The 935 has better reception when covered with a sleeve?

No problems with sleeves with my FR935. A sleeve shouldn't really be a problem unless it has some "metallic" content. We'd come to some sort of conclusion 6 months ago that the reason the FR935 was more accurate than the F5 was its lack of a metallic bezel that could be weakening/degrading the signals of the GPS satellites so if you're adding another "layer" it could take that a step further.

You may be right, as I have a metal button on my jacket sleeve, so it could interfere somehow with the GPS signal. I will try to cover the watch with a "non-metallic" sleeve to see if the same thing happens or not...

From Garmin, regarding the possibility of the F5 bezel interfering with GPS strength:


"I believe I have seen similar comments on forums, stating the bezel is affecting the accuracy, however this is unlikely, as the bezel itself, serves a purpose of enhanced antenna, so the result should be quite the opposite.

Once you have made another run, please provide the one from the Vivoactive to compare with.

Thank you for your feedback."

FWIW - Since changing to 1 second recording on my F5, I have found it almost spot on compared to both my Samsung phone with Strava and my Vivoactive HR. There is sometimes 10m difference if that. Pace can freak out under trees but not a huge concern as prefer average pace/lap pace anyway. Worth noting I am only using GPS, not with Glonass.

I also bought the 935 as a test and again the pace/distance was practically the same with 1 second recording. Decided to keep the F5 sapphire as the FR935 felt cheap, buttons were terrible (less clicky and sticking sometimes - known issue), UI felt laggy plus I am likely to scratch it.

Garmin are hardly going to admit there is a problem with the design of the F5 (and of the F3 before that), however they are about to release an update to the F5 to fix the bluetooth connection issues without ever admitting they are there...issues not present on the identical (other than the bezel) FR935 so make what you will of it.

Anyway the real test (and the real problem) was the wide variation in 1km autolaps between the FR935 (or FR235) and the F5, a big problem when you're doing some serious training and also because it's hard to accept a loss of accuracy when upgrading to a new model. Had it not been for that issue I (and others) probably wouldn't have started looking at actual tracks (I couldn't care less honestly) and found that poor accuracy was what was causing the problem with the lap variations. By the way 1s recording has no bearing whatsoever on the internal accuracy of the watch, just on what the track looks like on GC (or Strava).

Appreciated - I wasn't siding with the quote, just wanted to post their reply here as I had recently raised the issue with them. I've since emailed them back asking to clarify further. They also suggested I use 1 second recording to improve logged accuracy - which again, I have replied to quoting the difference to them as their explanation didn't add up at all.

As for that BT update that's coming up - do you have a link as I am struggling to find any material on it and I'm interested. That said, I've been lucky and never had any issues with BT on my F5.

Back to GPS, my autolap is set to 1 mile and again, might have been a good couple of days and tests but both F5 and 935 were within 1-3 seconds of each other logging each lap. Was a 5 mile run and involved trees at various times around the lake.

Can someone confirm that if I am using smart recording and I decide to cut a few small corners then it might not necessarily pick that change up in time and record a slightly greater distance?

Hi I recently run with both my GPS watch and I got different distance ( off by 200~500m in 5km ), my Fenix 5X always lesser distance than my Spartan Sport, I wonder which is the inaccurate one, then I planned to do some test, below are tested results as a reference:

Date, time: 28 Dec 2017 Morning 7.46~8.30am, Thursday

Clear sky, Sunny day

Started recorded from small town area with moderate traffic, driving car with both watches with about 40~70km/h,

No high building, no big tree, no under ground, no under flyover

After 17km drove on bridge all the way to the end about 100~110km/h


Comparison Watch -

• Suunto Spartan Sport Wrist HR
  

• firmware 1.11.56 ( up to date )
  
• Glonass - ON
  
• GPS accuracy - Best ( 1sec signal )
  

2. Garmin Fenix 5X Sapphire

• firmware 7.1 ( up to date )
  
• Glonass - ON
  
• 3D Speed - Off
  
• 3D Distance - Off
  
• Ultratrac - Off
  
• Data Recording - every second
  

Reference to ~

• Car Mileage - 2014 Honda City
  
• Google Map - manual distance measurement
  

Results:


Car = 37.8km

SSSWHR = 37.01km ( -0.79 vs Car, +0.06 vs Google )

F5x = 36.74km ( -1.06 vs Car, -0.21 vs Google )

GoogleMap = 36.95km



Monitoring on the way

• At 5km
  

• SSSWHR was 150m lesser than car
  
• F5x was 300m lesser than car
  

2. At 15km

• SSSWHR was 500m lesser than car
  
• F5x was 700m lesser than car
  

As results, the Suunto Spartan Sport is closer to the car meter as well as Google map, i'm not sure this can say this that the Spartan Sport is more accurate? what do you think guys?

Appreciated - I wasn't siding with the quote, just wanted to post their reply here as I had recently raised the issue with them. I've since emailed them back asking to clarify further. They also suggested I use 1 second recording to improve logged accuracy - which again, I have replied to quoting the difference to them as their explanation didn't add up at all.

As for that BT update that's coming up - do you have a link as I am struggling to find any material on it and I'm interested. That said, I've been lucky and never had any issues with BT on my F5.

Back to GPS, my autolap is set to 1 mile and again, might have been a good couple of days and tests but both F5 and 935 were within 1-3 seconds of each other logging each lap. Was a 5 mile run and involved trees at various times around the lake.

Can someone confirm that if I am using smart recording and I decide to cut a few small corners then it might not necessarily pick that change up in time and record a slightly greater distance?

Sure, we're all just users here sharing our experiences, so to your questions :
1 1s/Smart Recording : again, no impact on internal calculations, it's just what the track looks like on GC, Strava, etc...Now if you're having accuracy problems it is of course required to turn 1s on (storage space is not an issue anymore so need to skimp anyway) to see where the problems might be (next to buildings, when making sharp turns, etc...)
2. The "BT Update" is in fact NEW HARDWARE, both DC Rainmaker and 5Krunner have mentioned a new version of the F5 with these BT connection (and possibly others) problems sorted. This can't be fixed by software apparently.
3. Not sure what to say for your identical autolap between the F935 and F5, that's good news of course but I (and many others) got very irritated because of varying 1k (1 mile would be the same) distances on our usual training courses. Now if you're running in an area with no obstructions (trees, hills, buildings, etc...) the problem will be less noticeable of course.

My F5 is now better than when I bought it, maybe because of software updates.

I've also got a better GPS track during my runs within firmware v7.0 on my F5 Sapphire , really better but not perfect. Still a bit off-track on the map but closer to the way. After updating to version 7.10 it got worse again: jumping offset, peaks going crazy sometimes.

I am still wondering:
- Sometimes tracking is good and sometimes with almost same conditions it is worse again.
- Or vice versa, one run with sunny weather and clear sky the GPS is inaccurate, but under heavy rain it's much better.

I also saw that tracking for "running" and "cycling" on the same route is also a difference: cycling is always closer to the way/street. Corners, crossings, etc. are always hard shaped when I'm changing my direction. My settings in running and cycling are the same on my Fenix.


Does anyone know similar behavior running:cycling on the same route?


PS: I've also opened an issue on Garmin Support due to inaccuracy: At the beginning with 6.0 I've had a measured difference within 300-500m on a short track with 3,5km in total. That's 10-20%. That's absolutely inacceptable (e.g. security risks by tracking in alpine regions). Reply by Garmin Support was like "This is normal and not worse. Please keep firmware uptodate", always a few days after releasing a new firmware version. Ha ha ha.