[soliMilos]

Ambit1 distance anomalies have been discussed previously with the Ambit1 usually undercutting the raw GPS track log distance by 1-3%. However, most of these reports have been on more linear moves (e.g. running, walking). Does anyone use their Ambit for more dynamic moves (e.g. soccer, tennis)? If so, have you observed any extreme differences in distance? What has your experience been like?

Here are a few of my moves and distances:

Move	Soccer	Soccer	Tennis
Time	1:34:10.3	0:37:9.5	1:28:03.4
Ambit 1 Distance	3.75 mi	1.32 mi	1.09 mi
Raw GPS Distance	5.96 mi	2.10 mi	2.49 mi
Difference	2.21 mi (37%)	0.78 mi (37%)	1.4 mi (56%)

My question: Does this seem normal for this type of move? Should I send the Ambit in for warranty repair?

I contacted Suunto and after asking me to reset my GPS (twice) they suggested sending the Ambit in for warranty repair. However, the exchange I had with them suggested that they didn't know much about their own product. I leave for Switzerland for 3 months in a 2 weeks and I plan to use the Ambit for hiking and skiing. If I send the item in for warranty repair, I most likely will not get it back in time. Unfortunately, I haven't had a chance to more extensively test the Ambit outside of soccer and tennis. Before I send it in, I'd appreciate any feedback or experiences you guys have had with the Ambit.

 

[XCJagge]

I'd say your watch is fine.

To not count all zig zag in and measure officially measured maratons as 44 km device can't calculate raw distance between points. No-one knows what kind of algorithm Suunto uses, but there is penty of ways to do curve fitting to point series.
Curve fitting - Wikipedia, the free encyclopedia

I think Suunto's goal what it comes to measururing distance is to measure marathon style runs right and it does that job pretty well. So all zig zag ans sharp turn you can't find in normal marathon courses are smoothed out. And if you zig zag to pass people those will be smoothed out too and at finish line you watch shows the distance about right. You may know how official measuring is done, cutting all possible corners to measure it as short as possible. But when you play tennis, most of your movements are sort zigzagging is algorithm thinks it's is is an attempt to run a marathon. Also trail runs on a narrow zigzagging foot path gets filtered too in similar way. I have tried measuring my trail runs by clicking the same track with google earth, doing the curve fitting illustated in the page I linked by hand and I usually get pretty close what ambit told. So that's the way it works now and because marathons are officially measured and those who run those care more about it, and trail runs aren't usually that officially measuerd, so I think it works as it should. Except there could be a parameter so users could tell the algorithm how much these filtering is done for each activity, so we could fine tune it about right for each of our activity.