Method of compressing and decompressing an audio signal

Abstract

A method of compressing an audio signal comprising a series of values is disclosed, the method comprising: utilising a predetermined maximum number of bits B for representing values; dividing the audio signal into a series of blocks of a predetermined length N; and for each of the blocks: computing the maximum number of bits b required to represent the values within a block; packing an indicator of b bits in an output stream; and for each value within the block: where the value exceeds 2B writing only the b most significant bits to the output stream; and where the value does not exceed 2B writing the value to the output stream. The aforementioned methods are well suited to compressing impulse response functions. There is also disclosed a method of decoding an audio signal encoded by the aforementioned methods.

Dependencies are inferred from the claim text.