Changes

The Spartan has the worst [[GPS Accuracy]] of any watch I've tested, but to back up the numbers and to visualize things in a more qualitative way I've included the images below of the GPS tracks from one part of the testing course. I've added Below are the tracks from the latest firmware (1.4.6), along with the earlier firmware I tested (1.2.4) and from the [[Suunto Ambit3]] which is one of the most accurate watches I've tested for comparison. You can see that Suunto have improved the accuracy with the 1.4.6 firmware, but the Spartan is still the least accurate watch I've tested. The inaccuracy should be immediately obviousfirmware release says "A fix to GPS signal filtering, reducing the number of cases with large location offsets in GPS tracking experienced by some users" and you can see that the new firmware doesn't have the problems with offset tracks widely spaced rather than tightly following in the same way that 1.2.4 did. However, though the path shown in blackaccuracy has improved it's still appalling. The blue lap markers are widely scattered, suggesting the Spartan is not simply applying an excessive level of smoothing to compensate for accuracy problemshighlighting positional inaccuracy, but is actually unable to get a position though that's close to realityimproved with this firmware (my analysis focuses on the distance measurement. ) Most watches have a problem where the path curves in the middle section, and the Spartan does especially badly here, though there are a few green lines indicating that occasionally the Spartan is close. The Spartan does badly in the sharp turn on the right side of the picture, an area where most watches do surprisingly well. The section under the bridge doesn't show any obvious signs that the Spartan is failing to reacquire the GPS signal, though you'll notice some lap markers on there the shape of the tracks to the wrong side right of the bridgeis a little odd. There are a few tracks Overall, it's good to see that Suunto have improved the right shapeGPS accuracy, but are offset some distance from the pathand it's a statically significant improvement (p<0. This is most obvious in the sharp turn on the right, 05) but you can see it in other placesthey have a very long way to go. There were similar issues with offset The tracks in look better, so if the [[Garmin Fenix 2]] which were mostly remediated with problems won't be quite so obvious to a firmware upgrade. Howevercasual user, but the improvements in overall accuracy were not as dramatic as you might have expecteddistance measured can be wildly off. (I will continue to test as firmware updates become available.)

|[[File:BridgeSuunto Spartan Ultra1.4.6.jpg| None | Thumb | x300px |This is the 1.4.6 version of the firmware. (This diagram has tracks color coded with green indicating good accuracy through to red indicating poor accuracy, and the lap markers as blue dots.)]]|- valign="top"|[[File:BridgeSuunto Spartan Ultra 1.2.4.jpg| None | Thumb | x300px |For comparison, here's the earlier 1.2.4 Firmware.]]