Hi Davi,
The file source of my workouts to Golden Cheetah and Intervals are the same that comes from my Garmin. Just that Intervals sync from Strava and I feed Golden Cheetah with a original Garmin file download from Garmin Connect.
Nonetheless despite the power zones to be same on all tree, Garmin Connect, Golden Cheetah and Intervals I see a considerable diference in time in zones from Intervals.icu being Golden Cheetah and Garmin connect pretty much the same.
Why is that?
Tks,
Intervals.icu calculates time in zones for power using a 5 second moving average of the power as it tends to be very spikey. Short drop outs are also ignored.
The results are very close except for Z5 (Intervals higher) and Z6 (Intervals lower). So I suspect that this is Z6 spikes being smoothed out into Z5.
That said I don’t know exactly how Garmin Connect and Golden Cheetah do their calculations.
Hi,
Noticed this today after a short (37 minutes) XCO race on an MX track (steep, very physical and quite a lot of stop and go). On GC I have most time spent in Z7 followed by Z6. Z5 through Z2 being cca the same. On Intervals it’s Z6 followed by Z7 with these accumulating 34 seconds less time together.
The question is whether using the moving average is actually good for this application. I would say that for things like dropouts and erratic readings it’s good, but when you’re having valid data but they just happen to be mostly short and powerful bursts your time in zone gets artificially lowered and the ride looks easier than it was.
Food for thought.
