Work is being done on NARVAL spectra of Arcturus and the Sun with several line lists available in order to define the one most suited for the analysis of AP reference stars already observed with that instrument. This involves much tedious work of visualization to keep or reject lines, and select continuum windows. Definition of good fit to be refined as well as objective criteria to select enough lines for a high quality analysis. Another goal is to select select more lines in Sun and Arcturus which are a priori rejected because they are strongly damped, but which become more relevant in metal poor stars.