METHOD OF COMBINING PARAMETRIC MULTICHANNEL AUDIO ENCODING WITH MATRIX MULTICHANNEL ENCODING THREE-DIMENSIONAL SOUND

摘要

FIELD: physics; acoustics.;SUBSTANCE: invention relates to multichannel audio encoding and transmission and, particularly to methods of encoding a multichannel audio signal, completely backward compatible with stereo devices and formats. The invention involves combination of parametric multichannel audio encoding with matrix multichannel encoding three-dimensional sound through gradual improvement of sound of a direct mixing signal while increasing rate of transmitting data used for transmitting side information, starting from 0 to rate of transmission of data used for parametric methods. More specifically, the invention provides a method for flexible selection of the "operation point" anywhere between matrix reconstruction of three-dimensional sound (absence of side information, limited audio quality) and complete parametric reconstruction (complete required rate of transmitting side information, good quality). This operation point can be selected dynamically (i.e. varying with time) and in response to permissible rate of transmitting side information when determined by individual application.;EFFECT: efficient encoding multichannel audio signals and backward compatibility with solutions for matrix based encoding.;13 cl, 10 dwg