Abstract: A method of transforming a video image includes sampling a digital video signal at a higher frequency f.sub.2 than a usual frequency f.sub.1 and storing the sampled signals. The stored information is interpolated and addressed so as to create a video signal representing a required transformation, the frequency of the video signal being reduced to the usual sampling frequency f.sub.1. The method can be applied for enlarging and reducing the video image. Where the image is reduced, the reduction is limited to prevent the effect of interference between Fourier spectral harmonic bands and a fundamental band of the components of a reduced video image signal when it is created from the stored information. Preferably, the reduction is limited to provide gaps or guard bands which prevent this interference. The method advantageously provides high resolution horizontal components in the transformed video image.