On the solution of one classical problem in vibrational spectroscopy