Sampled-data audio signal compression with Huffman coding

This paper proposes a new digital audio coding method. It combines subband-coding with an interpolator that recovers the high frequency spectrum according to an analog audio signal model. The interpolator can be designed via sampled-data control theory. Efficient compression can be achieved by the Huffman coding at low bit-rate transmission. The proposed method is seen to possess a better frequency characteristic and a simpler processing algorithm than MPEG-1 audio.