Page 1 of 1

Outlier filtering with a real power meter

PostPosted: Mon Apr 13, 2015 3:30 am
by dannyst
I recently got a Stages power meter, and one thing I have learned about real power meters is that they produce a much rougher power curve than you get with calculated power. Overall, I find it a very invaluable tool though because I have an accurate read of my effort that is consistent regardless of if I was on the road or on the trainer. It is nice to be able to compare the two types of workouts for example. However, PP doesn't seem to handle some of the erroneous samples that come from the power meter as well as Strava does for example. PP tells me that my 5 second power is 2370 W. I know that is over double my real capability being that I actually participated in a study at a local university that measured instantaneous power. I wish that I was that strong though :). My 10 second power is also ridiculously high in my history.

It would be nice if something can be done to weed out those kinds of outliers that come with the true power meter. Even if it was just some way to go into my history and remove the particularly high reading or something like that. It's not a huge deal, but I thought I would just put it out there to see. I do love what PP makes possible, especially in the winter months,

Re: Outlier filtering with a real power meter

PostPosted: Tue Apr 14, 2015 12:01 am
by Adam
Can you either upload or email me a .peripedal file from your power meter? I would like to get a better idea of the erroneous samples you are referring to.


Re: Outlier filtering with a real power meter

PostPosted: Wed May 13, 2015 1:08 am
by dannyst
Sorry it took me so long to get around to sending you a file with the outlier data in it, or I am really capable of doing 2300 watts for 5 seconds... I wish :D

The file is too big for an attachment, but here is a link to a zip file with the .peripedal and .fit files for the workout that has me supposedly seeting a PR over 2000 watts. ...