Trellis Based Approach for Joint Optimization of Window Switching Decisions and Bit Resource Allocation
The fact that audio compression for streaming or storage is usually performed offline alleviates traditional constraints on encoding delay. We propose a rate-distortion optimized approach, within the MPEG Advanced Audio Coding framework, to trade delay for optimal window switching and resource allocation across frames. A trellis is constructed where stages correspond to audio frames, nodes represent window choices, and branches implement transition constraints. A suitable cost, comprising bit consumption and psycho-acoustic distortion, is optimized via multiple passes through the trellis until the desired bit-rate is achieved. The procedure offers optimal window switching as well as better bit distribution than conventional bit-reservoir schemes that are restricted to only ``borrow' bits from past frames. Objective and subjective tests show considerable performance gains.
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.