Data Reduction Techniques for High-Quality Digitized Audio
Two techniques for reducing the amount of data in high-quality digital audio were examined: minimum-length (Huffman) coding and differential floating-point coding. It was found that although minimum-length coding produces no distortion, it cannot uniformly reduce all different kinds of audio. On the other hand, differential coding uniformly reduces all kinds of sounds but can cause audible distortion on some signals.
This paper costs $20 for non-members, $5 for AES members and is free for E-Library subscribers.