Possible intensity scaling issue for two-dataset comparison

[magnuspalmblad]

I noticed that the dot products are much lower when scaling the spectra (i.e. using a value for Scaling far from 1), even when removing low-intensity (noise) peaks. Is this expected? In absence of noise, I would not expect scaling spectra to have such a dramatic effect. Two spectra that are very similar before scaling, should also be very similar after scaling.

[magnuspalmblad]

The behavior reported above was for Scaling <<1. For Scaling > 1, the dot products are higher.