Improving the performance of MPEG audio coder layer I using adaptive representation of bit allocation information

In general, all the spectral components of an audio signal are not preceptually significant and so the bit allocation information from the MPEG audio coder layer I often contains a sequence of zeros for all subbands at a higher frequency than a particular subband. Motivated by this, an efficient representation of the bit allocation information and its associated bit allocation algorithm is presented. Computer simulation shows that the modified version of the MPEG audio coder layer I, employing the proposed representation format, performs better in terms of sound quality than the original one