The paper describes a bit-rate scalable coder of the current MPEG-4 Audio working draft. The coder allows to decode useful subsets of the bitstream. This scalability is achieved by combining a core coder which, in general, is not a perceptual coder, with one or more stages based on the perceptual coding modules of the MPEG-4 Audio working draft. Despite the usage of nonperceptual core coders such as ITU-T G.729, G.723.1 or the MPEG-4 CELP coder, the system still maintains complete control over the psychoacoustic model in the combined states, thus reaching the same quality as the un-scaled MPEG Audio coders, although at somewhat higher bit rates. However, the system achieves a better bit-rate efficiency than a simul-cast solution.h
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.