This paper presents a methodology and a set of tools for gesture control of sources in 3D surround sound. The techniques for rendering acoustic events on multi-speaker or headphone-based surround systems have evolved considerably, making it possible to use them in real-time performances on light equipment. Controlling the placement of sound sources is usually done in idiosyncratic ways and has not yet been fully explored and formalized. This issue is addressed here with the proposition of a methodical approach. The mapping of gestures to source motion is implemented by giving the sources physical object properties and manipulating these characteristics with standard geometrical transforms through hierarchical or emergent relationships.