Traditionally, the mean squared error between the current and the previous frames has been considered as an appropriate criterion for selecting the displacement vectors in hybrid video coders. However, with a spatial coding method utilizing the correlation between samples, the error criterion should take both the mean squared error as well as the shape of the spectrum into account. In particular, it can be shown that when the spatial domain coder is based on a DCT and dynamic bit allocation, improved performance can be achieved by using the geometric mean of the DCT coefficient variances as error criterion. Furhermore, it is shown that with the new error criterion, the probability distribution of the motion vectors is changed drastically, allowing for further coding gains as well as reductions in complexity.