GPS: Completely unacceptable

JacekZet said:

Suunto manual you can find that "To get better GPS tracking results when exercising in difficult terrains like e.g. in a dense forest, ravine or city with tall buildings, it may help to turn on additional satellite navigation systems such as GLONASS or Galileo if available in your watch."

and in the Garmin manual

JacekZet said:

"The GPS + GLONASS or GPS + GALILEO options offer increased performance in challenging environments..."

They are saying essentially the same it's just that Suunto is a bit more cautious in the choice of words.

My understanding is that Garmin are 'tuning' watches with the Sony chipset for best performance with GPS + GLONASS, which is why that's default out of the box. Of course I might be wrong and have misunderstood something.

JacekZet said:

Why then is the possiblity to set the watch on two groups of satellites if it doesn't work? This was supposed to improve signal reception.

We are led to believe that it will improve reception by the choice of words that Garmin use when as many of us have found out, that is not necessarily the case. It's also why it is often recommended that the other combinations are tested, over the same route of course, to see what the effect of the different combinations is.

Certainly in my part of the world (NZ) I don't see much difference between any of the combinations. However, notwithstanding the differences between combinations or the Mediatek chipset I see little variation in distances over the same course that I have ran for quite a number of years. I am aware that the examples I have posted might not be considered as challenging enough to satisfy some, but they do serve to illustrate what I see. I know if I look hard enough through my history I do have other comparison runs in much more challenging conditions that have yielded results with similarly small differences.

JacekZet said:

The same route, which according to both GPS files is 7.35 km. This is a bit too much, because it should be about 7.2-7.3 km

Sorry to say this JacekZet, but for me this is an example of expectations being too high. You have 7.35 km but expect to be about 7.2 - 7.3km. You're looking at a difference of about 150m at worst to 50m at best. I really cannot see what there is to complain about from a wrist worn GPS watch. You say the outcome is too high without knowing for certain what the actual distance would be if measured by a calibrated wheel. Is it realistic to expect exact same results time after time when there are so many variables that affect the result - sattelite positions, running track over the ground, overhead cover (tree cover changes over the seasons) etc, etc?

In the last days I have been doing trainings with different devices on two of my favorite running tracks. Today I finished my comparison on the 10,3km Track.

The first training I did with Forerunner 235 and Forerunner 245M.

The second (today) with Fenix 6X Pro and Forerunner 245M.

Today's training has been a mess again. While the FR245 calculated accurate paces, the Fenix sucked nearly each kilometer. The highest difference in actual pace was 3 minutes 45 seconds. Strangely the results at the end were a bit better (e.g. Fenix shows pace of 8:55 across the full kilometer and calculated the final pace to be 5:10) . But in most parts the results are still way off reality.
Six out of 10 lap-times have a deviation of more than 10 seconds. Which is - for me - not acceptable. It is not possible to do an accurate pace-run with the Fenix 6X Pro.

Here are my results:
www.dropbox.com/.../10km_Compare.pdf

Both devices were on defaults. That means data recording is "intelligent" which does obviously not have a negative effect on the FR245M.

As this Fenix 6X is already the second one, I do not have to repeat those runs with the device set to an every second data recording. Because I know the results. I ran this tracks also with the first Fenix which was configured to do so. The results are the same mess.

I will complete my investigations in this manner with my other tracks. Just for completion and for me to know that I have really tried everything I could to become comfortable with this device. A Farewell in stages.

I will publish the results here.

The Fenix 6XP is a 750€ comedy-show.

Could be Interesting to do another test using only GPS on Fenix 6 :) 

I can understand if you do not want to do it but cold be great. 

Looking at both traces, it's hard to say which one is the 945 and which one if the Fenix. I was expecting the Fenix to be less accurate, but it's really very similar. Both are struggling at times, and are doing good at others. The distance is really close, 20m on 12km, it's nothing.

So the GPS data quality, at least with this example, is not significant. Why is the 6X counting less distance, that's the real question. If the quality of the GPS data of the Fenix was really less good, I would understand why they have a different algorithm. But it's not the case. Did you notice significant differences in the quality of the GPS traces ? I mean the GPS points, not the displayed distance.

Dennis.stevens1982 said:

Both devices were on defaults. That means data recording is "intelligent" which does obviously not have a negative effect on the FR245M.

Just to clarify here. Data is tracked at 1s intervals irrespective of whether you have 'Smart' or '1s' recording set. All that setting affects is how the data is saved and the quality of the track displayed on the map. Smart recording will usually result in smaller files, which was more of a concern with earlier watches where space was more limited. 

Does that setting affect post activity calculations? I can't answer that and not going to hazard a guess.

It's really down to the signal rather than the watch. Sometimes I get very good results sometimes not so good by running on the same track. Depends really on the satellites, weather, coverage/blockage of the signal. 

The  fenix 6 is excellent but the precision of the gps tracking is down to different factors. The watch is just one of them.

I did. And suprisingly the result was perfect. The distance was the same as on my Forerunners. And the pace was accurat. Even on those parts of the track where the terrain is really tough for these devices. I will make a summary of that run and publish it here. And I will do a second run with the same setup. To make sure that the Fenix was not just lucky...

Thank's! 

This is what I was trying to point out in our earlier discussion.  In theory, the more satellites available the better the reception and accuracy.  For some reason this isn't always the case as Dennis.stevens1982 and others have found.  GLONASS is said to work much better in the northern hemisphere's based on orbit.  I live on the northern coast of Washington state not too far from Mother Russia and I found I get better tracks using GPS only.  You think based on my location that GLONASS would improve my signal and provide better accuracy, but it isn't the case.  Why this is the case, I don't know and I'll leave that for others to debate.  If people are having accuracy issue, it's best to test all three settings than assume one is better than the other two.