The Watermarking of Digital Sound by Band Division and Spectral Spreading

Abstract

This paper presents a technique for the digital watermarking of musical sound. It consists of two main steps : the first is the division of the audio-frequency band into several sub-bands and the second is the embedding of a watermark in one of these bands by the spectral spreading method. The proposed scheme produces a watermark that is robust against data compression and ordinary attacks, while retaining high quality for the sound because we randomly select the specific band for the embedding of the watermark among the many sub-bands. For example, our experimental results have shown that the quality of watermarked music is good in about 45dB and that the watermark remains detectable after MPEG1 Audio LayerIII compression and jitter attacks.