Concealing of cuts between two audio time signals in the frequency domain, whereby only frequency components that make the signals audible are modified thus avoiding acoustic holes

摘要

Method for concealing cuts between two audio time signals in the frequency domain, whereby frequency processing is undertaken so that the spectrum of the two signals can be matched to each other on either side of the cut. Accordingly an averaged spectrum is determined on either side of the cut and the spectral signal is adjusted over a finite time before and after the cut using an appropriate weighting function. An Independent claim is made for a method for optimizing matching of audio signals on either side of a cut in which analysis and modification begins directly on either side of the cut and is applied to modification regions of the same time length.