This conceptual and theoretical progress paralleled significant engineering progress by small and large companies as well as universities. Much of this effort has focused on meeting fidelity, robustness and economic constraints.

While no breakthroughs are expected, a number of design choices are now available. There has also been significant experimentation with a variety of different marking spaces. For example, frequency decompositions such as DCT, FFT, wavelet and Fourier-Mellin transforms.

While there is no clear superiority of one space over another, considerable expertise has been developed for embedding watermarks in MPEG and JPEG encoded content. This work facilitates the design of very inexpensive watermark detectors that are suitable for mass deployment, e.g. for DVD copy control applications.