It has been a challenging task to deliver high-volume 3D multi-view media with spatial audio to users. The challenge stems from the availability of a large number of user equipment with different capabilities, range of viewing preferences, as well as diverse network characteristics. With the target of facilitating 3D media delivery to multiple users simultaneously and in a synchronised way with consistent and high quality, the ROMEO project [1] focuses on developing new methods for compressing 3D multi-view video and spatial audio, as well as optimising the networking and compression jointly by incorporating content aware adaptation across the Internet. The approach combines the Digital Video Broadcast (DVB) technology together with a Quality of Experience (QoE) aware Peer-to-Peer (P2P) distribution system. This paper introduces the overall system architecture of the project, outlines its several building blocks, and provides some preliminary results.