How can I be sure that what is happening is a shift, not just an average range of the peak that is too broad? Should I define 10 cm-1 between peaks as a pattern that recognizes shift or this range would be wrong? Can shifts happen in a smaller scale, like 5 cm-1?