Preprocess: Study + Simplify Noise fitting #1524
Description
In order to get the big improvements with the counter 1/f filtering in ISO-v4 we ported over the fitting from Junna + Tomoki's script. Now there's different methods (non-linear least sqr fitting vs minimization) and how the model is implemented matters. However at the same time as adding the new algorithm we also changed the nperseg to be larger which increased the number of low frequency points. We didn't study which effect was more important (the algorithm or the points used). A study should be done to find what is the most essential ways to get reliable fits and then we should document this and simplify the function.