Audio effects for the latest MPEG surround technology

Abstract

Incorporating with human perceptual modeling techniques, perceptual audio coders achieve low bit-rate audio coding performance while maintaining high fidelity of the reconstructed audio signals. Auditory models form the backbone of such coders and auditory masking is the key point that ensures such low bit-rate coding by removing perceptual redundancy. However, most of those auditory models that are currently used in many medium operate on one-channel monophony or two-channel strereophony. Multi-channel sound is now on its way into people’s daily life as the next step toward a more realistic audio reproduction. On the other hand, though there are multi-channel sound representation formats, such as DTS Digital and AC03, the data rates of these representation formats are so high that they are not possible or suitable for same practical applications. Furthermore, though some compatible multi-channel audio dawn-mixing algorithms are popular and efficient in generating high-quality stereo or mono signal, there is a shortcoming that these algorithms are not able to reconstruct the multi-channel signal. In this thesis, we focus on analyzing some of the auditory models that are currently used in audio coders and compatible multi-channel down-mixing schemes, and towards the objective of achieving a latest multi-channel audio experience called MPEG Surround. This MPEG Surround model incorporates among spatial audio coding (SAC) which focuses on modeling perceptually relevant spatial information in a parametric domain in stead of removal of irrelevant information form the signal (sub-band or transform) domain, offering bit-rate efficient and mono/stereo compatible transmission of high-quality multi-channel audio. It is observed that the audio performance of the MPEG Surround technology is much better compared to that obtained by the compatible multi-channel audio down-mixing coders.