A project to prototype audio descriptor driven granular sound mosaicing. Written in Max/MSP and was created as a prototype to test the effects of matching audio grains using different descriptor analysis. The full project will now be reimplemented in the SPPySound project with a number of improvments needed to produce the final product