Illustration of the movie scoring scheme. From the GRRT data, via a sliding window, we selected a set of images from which a movie was created. From this movie, complex visibilities were generated and the χ² between the synthetic data was computed while allowing the movie to rotate and vary in source size and flux density. The χ² were minimised either using an evolutionary algorithm (EA) or an MCMC scheme. After the optimisation, the sliding window was advanced until the entire GRRT data set was sampled.