To achieve audio signal robustness, this paper proposed a blind watermarking algorithm that embeds watermark in coefficient bits of the low frequency band and extracts it without original audio signal. It spreads the frequency band for hardware based watermarking to protect audio contents rights, generates watermark sequence on the watermark key using block code, and embeds and extracts watermark in the audio signal. It is implemented on the Blackfin processor using embedded module based LabView. Implementation in the high level language such as the graphical language over the low level architecture demonstrates the balanced method in high level and low level programming. By this test environment, it is possible to easily modularize and expand. In addition, it is possible to prototype promptly, and the Blackfin processor used in this experiment is suitable for real time audio signal processing. Therefore, equipments combined with digital watermarking facilitate insertion and extraction of watermark for variety contents and guarantee an intellectual ownership right of digital contents.