A typical data processing step for acquired absorption/transmission data is normalizing, i.e. stretching the curve such that it is bounded between 0 and 100%.
However, the absorption/transmission curve changes with the density and thickness of the sample. Moreover, absorption is a non linear phenomena (for example Beer–Lambert law), so I would expect that the relative height of the various peak would not remain constant across several samples of the same material.
Why, therefore, is it justified to normalize acquired spectroscopic data? Why not present data with original abs./trans. values, while providing details about the sample itself (for example thickness)?
In addition, how can spectroscopy be quantitative if the relative height of the peaks changes with illumination/thickness/density/etc.?