The invention relates to a method for the scalable improvement of the quality of an encoding method according to IT-U Recommendation G.722, including the following steps: —a digital error signal (E) derived from an input signal to be encoded and a prognosis signal is compared in sections to a number of M*LN different reference signals in an iterative process having a number of repeated steps depending on the scope of the expansion, and the reference signal having a minimum error signal of a prescribed error criteria is derived therefrom, —the reference signals are each made up of equidistant Dirac impulses δ(n) according to (I), wherein off=[0 . . . M−1], indicates the distance of the first impulse from a zero time point, αε{α, α, . . . , α} indicates the amplitude value, M the distance between the individual pulses, N the number of pulses, and L the number of different levels, —the information about the reference signal having the minimum error signal is transmitted. c ⁡ ( n ) = ∑ p = 0 N - 1 ⁢ α p · δ ⁡ ( n - off - M · p ) ( I )