This task will normally report a lot of line candidates. This is because it regards even very faint features as possible lines. The automatic line identification routine works best when it has access to a list of all features and you are therefore encouraged not to tune ech_linloc to reduce the number of candidates. The auto identifier starts by using only the brightest dozen lines so the result will not be biased by the faint features unless they are essential to the identification.
Martin Clayton