A few observations on footpods and treadmills...seems like a popular topic these days
There have been a lot of posts recently on footpood calibration and use with treadmills. I have struggled in the past with never being able to get a good and consistent calibration, and it didn't help that everytime it looked wrong when using on a treadmill, people would tell me that "the treadmill is probably calibrated wrong, the footpod is great right out of the box".
I think I've proved that pretty wrong at this point. After the 4th treadmill run in 3 weeks, run at yet another random hotel on a random treadmill, I have concluded four things:
1. The footpod is damn consistent, but simply CANNOT be calibrated at multiple paces. You need to pick the pace at which you want it most accurate, and live with the fact that it will underreport or overreport at other paces.
2. Calibration outdoors is irrelevant for the treadmill. My stride is different inside and clearly a different "calibration number".
3. Treadmills throughout the world are pretty consistent. If they are grossly inaccurate I'd be surprised , because at a constant footbpod calibration and footpod reported speed, four totally random mills reported really consistent speeds.
4. Do not say that the footpod is pretty close "out of the box". In my case I need a calibration factor of ~95% at z5 and ~90% at z1/2, which I don’t' consider very accurate at all.
How I came to this conclusion was to take 4 runs, each was 30' with 5' warmup then 6x2'(2')@z5. Each used the same shoes, footpod and 1.5% incline. I documented in my Garmin Connect blurb what treadmill speed setting I used for the z5 and the recovery. The z5 speed was always set 10-15" prior to the interval start, and the interval was always stopped before the speed was ramped down. So the z5 was all "at the pace that the treadmill was giving", and the recoveries always had about 20" total of acceleration or deceleration above the speed that was set.
Looking at the data (after doing all the mph and kph to min/mile conversions…thanks Google), the difference between treadmill-reported z5 and watch-reported was incredibly consistent – the footpod was reporting 0:22-0:23 faster per mile – for every interval, and on every treadmill. The calibration factor implied by the intervals ranged from 0.933 and 0.937. Averaging the implied calibration factor across all intervals in each run yielded an implied calibration factor range that was even tighter. So I set the factor on my FR60 to 93.5%.
Done…but only for z5…
…because the pace difference during the recoveries (which were at a footpod-reported 7:30 vs. z5's at 5:55) was more in the 0:40 range. The implied calibration factors were again in a very tight range, but were clustered around 0.905.
So FINALLY I have something more than someone's say-so and a general "feel" that a calibration is only valid for a certain pace range.
And based on the tightness of the how the results clustered, I strongly suspect that the blanket statement that treadmills are typically quite inaccurate is bogus. These treadmills were located in Chicago, Shangahai, Seoul, and Milan. All different brands. The one in China felt like it was about to fall apart. All different. But all reporting the same differential vs. a footpod on my indoor training shoes.
My "outdoor" calibration has been very consistent and around 96% at the same pace range, so another myth (calibration on a track will give you a good number to use on the mill) gone...
Eventually I'll repeat this experiment at z4, but for the moment (pre-OS) I use treadmills for quick 30' runs when I can't get outside, and so z5 is the most important factor to get right.
0
Comments
End Septemner I drove across the country from Southern California to Ohio and calibrated my footpod at easy pace on the treadmill in the hotel in Grand Jonction, CO. It was a LifeFitness treadmill (very nice to run on). Ran on the same treadmill next day, with consistent results.
Than ran on a treadmill in Kansas City, and for the same treadmill speed the footpod speed was definitely different. Same in St Louis, same in Mentor OH when running on the hotel treadmill versus the local gym treadmill. BUT I do not have real data recorded so can't tell how different they were.
That said, based on your findings, how should I calibrate my footpod on the brand new treadmill I got this week? It sounds like you are saying to use the treadmill speed as your driver, and use footpod more for recording and record keeping (or is this not a correct interpretation)?
Thanks for sharing your thoughts and observations!
Good Stuff Matt as I was trying find an approach at Testing this as well...so the obvious question to you...
If the treadmills aint that far off...why do we bother with the footpod?...I mean I only purchased the footpod to improve the accuracy of my indoor running paces...not so much distance............and while I kind of like having the cadence data...........it's getting a little burdensome for pace....
Maybe...my thinking goes...if I'm pretty much going to be running on TM at only one or two clubs...I should just test the calibration of those mills (ie. string around the belt/measure/mark/count revs) ...be done with it...
....and not worry about it the 5-10 times a year I'm traveling and have a different mill
All this agrees with my anecdotal observations about footpods and treadmills, although I dont have the breadth of Cap't Aaronson's earth-girdling data set.
That's pretty much my reason...makes us geeks, but hey
Lecture about primary and secondary standards in measurements kept in pocket for now. Your data are probably as good as it gets without going to extremes.
Interesting find, at both speeds my footpod and treadmill speed were almost identical! There was only about a 5 seconds / mile difference. (Not a steady difference, but the speed on the garmin would float between treadmill speed + and - 5 seconds / mile. That makes my training a lot easier, since it indicates I can rely on the treadmill to set the speed, but also know I should be getting good data into WKO+ now.