A New Backward Predictor for MPEG Audio Coding
In MPEG-2 advanced audio codec (AAC), a second-order backward adaptive lattice structure predictor is used for removing the redundancy of each spectral component. Although it enhances the performance of the audio codec significantly, this backward predictor requires a large amount of computations and memory. In this paper, a new backward prediction scheme is proposed to reduce the complexity of the adaptive lattice algorithm. Instead of using traditional adaptive algorithms, the backward predictor is found by block linear prediction (LP) analysis. The distinct advantage of the new scheme is that the performance and computational complexity can be balanced by less frequently updating the predictor coefficients. Our preliminary experimental results show that the new backward predictor gives better or almost the same performance while requiring less complexity than the adaptive lattice algorithm.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is temporarily free for AES members.