A loudspeaker comprises a region devoid of electroacoustic transducers and at least two groups of electroacoustic transducers at the periphery of the region. Each of the groups is configured to produce a different sound field representative of an audio signal and each of the groups comprise a different set of transducers. This allows a more compact loudspeaker to be provided because the transducers can be located around the periphery of a visual display screen. Sound quality is not compromised because groups of transducers forming a subset of the totality of transducers are used for each sound signal, thereby avoiding alias beams.