Cees Taal | Smoothing your data with polynomial fitting: a signal processing perspective

[@bhaskartripathi]

wonderful explaination!

[@xarran]

you have no idea what's happening at 25 hz? it's a harmonic distortion of 50 hz, lol

[@papanyanz]

Actually you should NOT have 50hz signal, but a 100hz one, since power is proportional to current and or voltage squared! Therefore negative side of sine turns positive "reflecting" by x axis and 2x times more frequent. I don't know how tjey came up with 50hz, i think they anticipated that ambient light somwhere there and then readjusted their scale from that, but that introduces 2x frequency error also to the useful signal.