Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Stryker Burks

macrumors newbie
Original poster
Sep 19, 2020
1
0
Just curious, but has anyone tested the GPS Accuracy on a run? (possibly compared to their Garmin or another fitness brand watch). Thanks....Thinking about pulling the trigger as I still have an AW2 but the GPS is horrible.
 
See here... short answer, it's extremely accurate:

This is promising to me. I’ve always felt that my S4 was off, it always tells me I’ve run longer than anyone else I’m with. Excited for the S6.
 
See here... short answer, it's extremely accurate:

No, I do not second that.
Basically what DCR says in its very short test that the AW6 is on par with the Forerunner 745 and that the smoothing of tracks is not there anymore.
The newer Garmins with Sony chipsets have notably worse GPS accuracy than their predecessors and so "extremely accurate" is hyperbole.
With "it is ok or it is good enough" I would agree.

Just my 2c.
 
  • Like
Reactions: James.K.Polk
See here... short answer, it's extremely accurate:

I've long been a fan of DCR.

In reading the article, he doesn't seem to mention whether the change in track smoothing is specific to the S6, or is perhaps a feature of WatchOS7. i.e. do we know that S5 watches won't see the same benefit?

Does the video speak to this?
 
I've long been a fan of DCR.

In reading the article, he doesn't seem to mention whether the change in track smoothing is specific to the S6, or is perhaps a feature of WatchOS7. i.e. do we know that S5 watches won't see the same benefit?

Does the video speak to this?
If you read the comments (yeah I know) you will see that he is going to investigate that but doesn’t know yet.
 
  • Like
Reactions: deeddawg
I've long been a fan of DCR.

In reading the article, he doesn't seem to mention whether the change in track smoothing is specific to the S6, or is perhaps a feature of WatchOS7. i.e. do we know that S5 watches won't see the same benefit?

Does the video speak to this?
Smoothing is definitely purely a software feature
Corelocation api gives a raw position and it’s up to the app developers to implement a smoothing/filtering algorithm
Maybe Apple has tweaked their own app to reduce filtering ?
But there is no reason to only apply that to S6 only
 
  • Like
Reactions: NME42
If you read the comments (yeah I know) you will see that he is going to investigate that but doesn’t know yet.

thanks. I intended to get into those but hadn’t yet done so.

Smoothing is definitely purely a software feature
Corelocation api gives a raw position and it’s up to the app developers to implement a smoothing/filtering algorithm
Maybe Apple has tweaked their own app to reduce filtering ?
But there is no reason to only apply that to S6 only

Good info. It wasn’t clear if the smoothing may’ve occurred within the subsystem or at the surface app level.

It will be interesting to see if there are any S6 specific improvement even if there are also platform-wide improvements.
 
Last edited:
  • Like
Reactions: James.K.Polk
As the SE has the „old“ AW5 chipset, this could be.
Also a factor is GPS antenna placement. What I understood from iFixit teardown, a few things changed from AW5 to AW6. No idea if antenna actually changed.
But from my experience, placing the antenna differently to the upside when running makes a difference on the AW5. So left arm crown left is the better position as the antenna then points upwards.
 
Not sure if posted yet, here's DC Rainmaker's initial take on the SE

Seems the GPS isn't as good as on the S6, suggesting perhaps some of the improvement is definitely in the hardware.

Still all preliminary but worth the read: https://www.dcrainmaker.com/2020/09/apple-watch-se-fitness-impressions-untangled.html

I was looking at this on my Series 5, after reading his review. I still show the very loopy and smoothed out plots with my GPS. Seems that only the Series 6 has the changes.
 
  • Like
Reactions: deeddawg
Another pretty good example of how good the tracking has been on my Series 6. Cellular was on and the run was recorded using the Stryd Watch OS app.

https://strava.app.link/hjzIW6Xxlab
1601935646201.png
 
I’ll cross-post in the other thread too, but I had noticed since getting the S6 that on the HealthFit app, there is now a section for each activity that is “GPS accuracy.” I emailed the HealthFit creator and s/he wrote back:


Thanks for reaching out.

The Apple Watch is reporting the GPS accuracy for each GPS location. This information is stored in Apple Health and this is how HealthFit gets the GPS accuracy. I don’t know exactly how the Apple Watch estimates the accuracy.

The graph shows an histogram. The percentage (Y axis) means the percentage of the population (GPS points). The X axis is the accuracy. (Meter or Foot).”

I’m wondering if there was indeed a change, since Apple now specifically records data on it. Either way...I’m happy with my performance on the S6. Routes definitely seem to appear more precise.
 
I’ll cross-post in the other thread too, but I had noticed since getting the S6 that on the HealthFit app, there is now a section for each activity that is “GPS accuracy.” I emailed the HealthFit creator and s/he wrote back:


Thanks for reaching out.

The Apple Watch is reporting the GPS accuracy for each GPS location. This information is stored in Apple Health and this is how HealthFit gets the GPS accuracy. I don’t know exactly how the Apple Watch estimates the accuracy.

The graph shows an histogram. The percentage (Y axis) means the percentage of the population (GPS points). The X axis is the accuracy. (Meter or Foot).”

I’m wondering if there was indeed a change, since Apple now specifically records data on it. Either way...I’m happy with my performance on the S6. Routes definitely seem to appear more precise.

Hmm that's very curious indeed....Healthfit is reporting GPS accuracy for the run as:

-average: 7.2m
-minimum 5.99m
-maximum 32m

Apple now making this sort of information available and reducing the amount of smoothing when using the stock Activities app certainly shows their confidence in how the S6 watches will perform.
 
My Saturday 10 miler reported:

Average:3.17m
Min: 2.51m
Max: 9.11m

I’m pleased with that, especially for a longer distance, with a number of turns and tree-lined streets!
 
Not sure how accurate those stats are, my location never deviated by 32m as suggested by Healthfit but hey, thanks for pointing this out.

Definitely a very, very interesting data point!
 
Not sure how accurate those stats are, my location never deviated by 32m as suggested by Healthfit but hey, thanks for pointing this out.

Definitely a very, very interesting data point!
It would make a lot of sense that the visible track is smoothed, and that the amount of smoothing may be variable depending on the reported accuracy claim for each datapoint.
 
Which one is better ?
The distance is completely different between Garmin 6X and AW6 !!
It's the same workout ?
 
Which one is better ?
The distance is completely different between Garmin 6X and AW6 !!
It's the same workout ?

I believe they’re different (the top part doesn’t take the same path; the bottom map takes a straight path and not as far up, and the top seems to go through the park) but you can see how close they are (no difference really) on the rest of the route which matches up.
 
  • Like
Reactions: NickYanakiev
Yup! Not to mention that the areas, where the tracked path deviated from the actual path were equally difficult for both watches.

Now, this comparison holds only limited value as the two runs were not recorded simultaneously. GPS accuracy can vary greatly on the same route due to the weather/location and number of available satellites.

Are you guys aware of preventing specific notifications from being displayed during activities?
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.