Due to the variability in the sets of color-matching functions, different real observers have associated different color-representation systems. This fact darks the interpretation of the experimental results obtained for different observers. In this work, we propose a transforming operator, which optimizes the mapping of the color-representation system associated with an arbitrary observer into the system spanned by a standard set of color-matching functions. When comparing the experimental results for different observers, this transformation allows rls to obtain more reliable conclusions. If the the data are analyzed in the systems obtained by applying the optimized operator, the differences between the colorimetric behavior of two real observers are less than when, as usual, the results are compared in the systems spanned by the sets of color-matching functions of the individual observers,