DEVICE AND METHOD FOR GRADUATED ENCODING OF A MULTICHANNEL AUDIO SIGNAL BASED ON A PRINCIPAL COMPONENT ANALYSIS

摘要

The invention concerns a system and a method for graduated encoding of a multichannel audio signal including a transformation by principal component analysis (PCA) of at least two channels (L, R) of said audio signal into a principal component and at least one residual sub-component (r) defined by rotation by a transformation parameter (?), including the following steps: forming a residual structure (Sf_r) per frequency sub-band from said at least one residual sub-component (r); and defining an encoded audio signal (SC) comprising said principal component (CP), at least one residual structure (Sf_r) of a frequency sub-band and said transformation parameter (?).