Analysis and Synthesis of Binaural Parameters for Efficient 3D Audio Rendering in MPEG Surround

This paper describes a novel method to simulate an multichannel audio experience using stereo headphones. In contrast to conventional, convolution-based methods, the current approach is based on parametric representations of spatial audio. An audio scene with multiple virtual sound sources is represented by a mono down-mix signal of all sound source signals, accompanied by certain statistical (spatial) properties. These statistical properties of the sound sources are combined with statistical properties of head-related transfer functions to estimate "binaural parameters" that represent the perceptually-relevant aspects of the auditory scene. Subsequently, a binaural rendering stage re-instates the estimated binaural parameters on the down mix. The advantage of this approach is that the computational complexity of the rendering process is virtually independent of the number of simultaneous sound sources. If combined with parametric multichannel audio coders such as MPEG Surround, the proposed method is advantageous over conventional methods in terms of perceived quality and computational complexity.