Thanks Kevin.
My focus is on discovering why the iPhone/Memory-Map combination has started behaving in this way. I can filter/reduce the number in various tools I have after the initial import, such as with GPS Utility or Memory Map itself, on the desktop PC.
I also now have Memory-Map's reply, which is essentially that they see no evidence that the cause is due to their app.
But I will try a test with a different app, probably Pocket Earth.
However, I am coming to the view that it's a hardware or iOS problem with the iPhone 6S+. This thread seems to show that others are having recent problems too.
gps not working properly after iOS 9.3 update | Official Apple Support Communities
Some background detail follows for you and any others using iPhone or iPad for GPS work. All feedback would be appreciated.
Sample of various walks, showing the sudden increae in trackpoint recording rate:
https://dl.dropboxusercontent.com/u/4019461/TrackpointsSummaryTable.jpg
The GPS file (after import to PC and saved as GPX):
https://dl.dropboxusercontent.com/u/4019461/20160826Clapton-B.gpx
Estimates derived from various versions of that file:
Using original 4703 trackpoints:
Start 09:32, end 12:28, dur'n 2:55, 6.36 miles, avg. 2.2 mph, gross asc/desc 700 ft
Reduced in Mem-Map (PC) by a factor of 8:
Start 09:32, end 12:28, dur'n 2:55, 6.03 miles, avg. 2.1 mph, gross asc/desc 430 ft
Google Earth elevation profile (no GPS data): gross asc/desc 383 ft
Converting original to a 299 waypoint route in Mem-Map gave gross asc/desc 403 ft
My best guess, for my finished walk print out:
Start 09:32, end 12:28, dur'n 2:55, 6.2 miles, avg. 2.1 mph, gross asc/desc 400 ft
https://dl.dropboxusercontent.com/u/4019461/20160826Clapton-B.jpg
And finally, filtering by a user in another forum with Garmin Basecamp, set to a minimum of 100 feet between points:
Start 09:32, end 12:28, dur'n 2:55, 5.82 miles, avg. 2.0 mph, gross asc/desc 340 ft
(At more than 15 times fewer than the original, that edit IMO is probably getting a bit
too light on trackpoints, but it's an interestingly arguable issue!)
My conclusion so far is that the original recording for some inexplicable reason captures a greatly excessive number of trackpoints compared to earlier recordings. That results in a significantly exaggerated gross ascent/descent, based on other evidence, personal 'feel', etc. And also a smaller exaggeration in distance. But it's then unclear what reduction is necessary to get the 'most realistic' estimates of distance and ascent/descent.
--
Terry, East Grinstead, UK