The signal processing for the speaker is implemented via a Linux based cluster consisting of 15 single nodes and two control computers. From there, audio signals are routed over (optical fibre) MADI lines to MADI/ADAT-converters located next to the individual, eight-channel loudspeaker modules, resulting in a total of 832 channels of the array. In addition, all components are linked via Gigabit-Ethernet allowing two-way communication for updates, (real time) changes in loudspeaker equalization and for state control. For the sound field synthesis, the software sWonder - developed at the Technical University Berlin and released under GPL License - has been extended and separated in individual software modules for the graphical user interface, a score player and recorder, a control unit, and individual online and offline rendering units. The individual modules communicate via the OSC-protocol for defining dynamic, temporal/spatial scenes.