@GregZeger suitably chastised me for my GPS test earlier I the week. Basically I need to (wo)man-up and get out there on the streets with 4 watches on my arms. I need to be proud of what I am.
Well. Something like that.
Anyway, I did. Just for Greg I sort-of repeated the run test wearing the 4 running watches. This time I had the mental agility to DISable ULTRATRAC before I started on the 920.
Being a bit more serious. It was at night (c’mon 4 watches in daylight…no way) and quite hot. Combination of tree cover and hills and open areas. Cloudless sky and I did a Z2/Z3 >90 minute extensive endurance run.
This time I had the Garmin 920XT on the underside of my arm and the V800 on the underside of the other arm. ie Runner 3 and SPARTAN in their proper positions. I mention this as it could make a difference – indeed if the TomTom was in the underside position its GPS WOULD be compromised because of the antennae location.
Let’s start off with the speed/pace, below. Click it and draw your own conclusions. To me the Polar stands out with the red spike at 32 minutes. Nothing on the route was unusual. It has a few spikes earlier as well as might be said to a lesser degree of the TomTom. The nature of the wearing position sometimes caused the watch head to bang on the other watch clasp…could be that?
So the Suunto Spartan ULTRA and the Garmin 920XT seem to track each other speed/pace-wise – the TomTom too joined in the co-tracking fun after 15 minutes or so.
An interesting thing I found out recently is that micro-adjustments are made to satellite positions every 4 hours or so (search: “GPS ephemeris”). If you haven’t sync’d in 4 hours you won’t have accuracy until about 15 or so minutes into your ride/run when that data is effectively downloaded as you go. This applies EVEN IF YOU HAVE ASSISTED-GPS. If you look at the blue SPEED track above for TomTom this could very well explain it took a little while to get more accurate. The TomTom had not been sync’d for about a day…
The Runner 3 recorded the shortest distance, with the others coming in higher at 920XT (+90m), V800 (+190m), Spartan (+260m). So we are talking less than 1% variation to the mean total distance. HOWEVER I looked at some other graphs (same run) and the gained/lost distances seemed to come at low speeds where I stopped for a drink and went through various gates.
Contrary to the previous paragraph, the Suunto and the TomTom autolapped NEAR-EXACTLY at the same time for EVERY 1KM autolap (I can’t explain that, but they did). To the point where I assumed they had the same GPS chip and algorithms.
In terms of following the actual route and in stark contrast to the last test; this time it was the TomTom Runner 3 and the V800 that ‘threw a wobbly’ at various points and were notably off track nearer to the start in places – probably due to the ephemeris point, above. As shown to the right.
That’s not to let the Garmin and Suunto off the hook. They were each guilty of several LESSER misdemeanours en-route.
The Garmin 920XT MAY have performed a little better in open country but it was very hard to spot ANY trend. Garmin and Suunto were mostly in agreement nearer to the start (and correct) whereas TomTom and Garmin were in agreement half way through (and correct). Nearer to the end Polar and Suunto were more in agreement (and correct).
I could post lots more images but it would take quite a while and wouldn’t really say much that is definitive or helpful.
So I don’t know what to say really, other than ALWAYS SYNC YOUR WATCH BEFORE RUNNING and turn OFF ULTRATRAC if you want an accurate GPS track!
Whereas the SUUNTO SPARTAN just won last time, this time maybe the Garmin 920XT just edged it. But it was close and all were good enough FOR ME.
3 hours of running with lots of watches. Not a lot of conclusions. Sorry.