We set out to design an electromechanical performer able to read a score of commands and to be extremely accurate in its work. [re]BO[u]NDS does not renounce composition merely to show a useless precision mechanism, albeit a complex one. Its structure is a well-defined composition form: a mechanism made human by the composer. In the center of the room a vertical structure is set up, composed of 3 long glass cylinders filled with water, placed below the drippers. When a drop of water falls into a cylinder, the sound of the drop will be captured and processed in real time by a complex computer architecture which also oversees the spatialization. The original sound will bounce, within the space through all emitters. The multiphonic system will make all sound directions audible, highlighting several dialogues between sound points. This “bounced structure” will be created by a sound patch designed with a range of delays and electronic elaborations that are characterized by parameters evolving in the time domain. These rebounds include macro-sounds (groups of autonomous cells) and micro-sounds (small portions of the drop of water). Imaginary forces join speakers as cells of the same structure, drawing BONDS within the space. The original sound, like a domino effect, will REBOUND in all directions through all the sound sources. From the “crystalline” sound of a drop of water to the “crystallization” of chaos due to its fall to the ground.