how does digital audio watermarking work?

IFPI listed several criteria for an effective watermarking solution, including -

  1. Transparency - The code should in no way affect or degrade the quality of the protected recording;

  2. Robustness - The embedded data must be recoverable, even after several copies have been made (e.g. digital -analogue, digital-analogue-digital;

  3. Security - The watermark should be tamper-proof, and should not be removed without damaging the sound quality of the recording;

  4. Compatibility - The watermark should be capable of processing monophonic, stereophonic and multi-channel (such as the seven-channel audio of surround sound).

Though many companies are working on several differing versions of watermarking technology, the key stages of any watermarking system are similar in theory, as illustrated by British-based technology company CRL's ICE watermarking solution.

Stage 1. Encoded master creation

The first stage of the process is to use an ICE Encoder to produce a stereo encoded master which is then used to create the published end product, such as a CD, a satellite television transmission, or promotions for forthcoming programme schedules. The analogue or digital audio is fed into the ICE Encoder in real-time and an ICE encoded master is created.

Stage 2: The material is broadcast by either a radio or TV station, or distributed for sale in CD /MiniDisc /video etc. form.

Stage 3: Simple decoding stage

ICE decoders are distributed throughout a territory and are designed for remote operation from some central site. If a radio or TV channel requires monitoring, an appropriate receiver is connected to the input of the ICE Decoder. The Decoder can simultaneously monitor up to five separate channels and log into its memory any ICE codes that are detected in the broadcast material. The Decoders are designed to be located centrally and accessed by modem, but local download is also available. ICE Decoders can also detect ICE codes in taped broadcasts. [43]