Audio signals, such as a music clip, are inserted with detectable watermarks. The inserted watermark can identify the content producer and provide a signature embedded in the audio signal and cannot be removed. The watermark can survive all typical kinds of processing and malicious attacks. The watermark can be implemented in a watermarking system that employs a covert channel encoder to layer an additional information data message or covert message on top of the watermark. The watermarking system can also include a permutation technique to further hide the watermark and hide the covert message within the watermark.