Hi there, I have some P L2/L3-edge XANES spectra. I am not sure how to remove the background and normalize these data. Due to the fact that the pre edge and the post edge region is not very flat the the edge-step normalization using AUTOBK dose not work very well. CLNORM background subtraction seems to work better in some case. But some spectra get partly negative values after CLNORM (see .prj attached). So my questions are: 1:Is CLNORM the best choice in my case, especially when I plan to do linear combination fitting with these data? 2: Athena needs the knowledge of the absorbing atom an the edge to use CLNNORM, but the spectra contain both L3 an L2. So, which edge do I have to choose L3 or L2? 3: Would it be also correct to use a Victoreen-Fit for the pre edge without any further normalization and than do the linear combination fitting? Or do I have to normalize spectra in general when I am doing linear combination fitting, to eliminate effects of different concentrations of the absorbing atom in the mixtures? thanks, a lot for your help jens