Hi there, podcast (#2) pointed me to the “advanced mpa” tab. previous/current is clear, but when I hit the “Extract” button it gives higher PP and TP. In fact, the BT is not on second half of the activity but to an earlier point.
Why and when to use this?
Tnx, still leaves the question: keep the auto generated profile or the one after extraction? And why do they differ, assuming the automatic analysis was done after the ride with the same data available?
Short answer, you should go with the second/bottom signature.
Long answer, advanced MPA allows you to “seed” the Signature Values, while Extract actually runs through and calculates MPA second by second using the seeded signature, and then making small adjustments until the proper signature is extracted. This is generally only needed if certain signature parameters are dramatically off (like if Peak Power is 250+ Watts low, or if threshold is significantly underestimated).
Mathematically, it’s possible to fit almost an infinite combination of signature variables to the data, which is why it’s important to have accurate Peak Power/HIE numbers. However, once an accurate PP/HIE value is known, it dramatically narrows down what the proper signature should be.
OK, short answer: will go with the second, you’re the expert, thanks for your fast response! 
Long answer/new question: so with the extract option, another algoritm is followed compared to the one executed automatically? Since data is the same and outcome is different
