Multi-Dimension Vector Quantization in High-Fidelity Audio Coding
Multi-dimension vector quantization (MD VQ) is proposed which compromises the performance achievable with unconstrained VQ but provides useful and favorable tradeoffs between performance and complexity. MD VQ is introduced in a Musicam (ISO/MPEG-1 Layer II) coding scheme in order to improve its performance and preserve the bandwidth of the audio signal at low bit rates (below 64 kbit/s/ch). Average distortion was traded against codebook storage for a given search complexity and rate in a high-quality audio coder. This technique leads to a substantial reduction in the overall bit rate and codebook storage for the audio coder.
Click to purchase paper 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 $20 for non-members, $5 for AES members and is free for E-Library subscribers.