Hi, After a couple of chat with Leo Spyracopoulos, I understand better the link between s / rad ans s in spectral density mapping. Leo started with : "The results from the notebook should not be multiplied by 2 pi. Finally, by definition, a radian is a dimensionless quantity with no physical unit associated with it, so there is no error in the 2006 paper. However, it is probably best to quote the spectral densities using just units of time." He also told me about some papers writing their tauc values in ns / rad instead of ns, but keeping the same value (no scaling with 2 pi). Leo ended saying that : "When Jeff Peng wrote the paper he felt it was important to quote the spectral density values in s/rad. If you look at his spectral density values and just use ns instead of ns/rad, you get the right values. However, I think that the spectral densities should really only be quoted in s (or ns)." I thus think that everything is now fine with jw_mapping within relax. The output are in s, which is fine and consistent with SI units. Finally, I think that every place with ' s / rad ' should be changed for 's'. Cheers ! Séb