With an increasing interest in personal audio systems, the car cabin is regarded as an interesting environment for generating clearly audible sound in one “bright” zone without disturbing people in another “dark” zone. Two algorithms, acoustical contrast control and least squares method, are applied for producing two independent listening zones, one zone including the front passengers and the other one including the rear passengers. The generation of an acceptable level of acoustical contrast between the two zones, using an array of four standard car audio loudspeakers, is limited to relatively low frequencies. In order to increase the frequency range of acoustical contrast and achieve better control performance, an array of loudspeakers on the roof of the car is investigated in this paper. In particular, a selection process for the configuration of the source array is described. Free field simulations are used to calculate the responses of the source array and investigate the performance of two control algorithms. A practical implementation is also planned.