ABOUT

 

The ERC Advanced Grant G3S (Generative Spatial Synthesis of Sound and Music) project proposes an original approach to artificial intelligence by integrating AI with spatial audio. In contrast to proprietary, resource-intensive and language-based AI such as ChatGPT, the project focuses on the development of frugal, local and open-source AI methods for the generation of novel sound spaces. While AI has been increasingly applied to music and sound, spatialization has remained largely overlooked, even as 3D spatial audio technologies have undergone significant industrial development and standardization.

The ERC G3S project, designed by and for musicians and sound creators, will explore generative approaches to spatialization based on machine learning. As such, the project seeks to break away from the standards that shape the way we create and perceive spatiality, opening up to a diversity of spatial-audio models, practices and generative processes.

Over the coming five years, the G3S project will articulate and implement four major research objectives: 1) the design of a unified operational representation of existing spatialization engines, 2) the proposal of a thesaurus and quantitative measures to describe the spatiality of sound, 3) the generation of sound spaces from machine learning, and 4) the design of user interfaces to explore spatial audio.

To explore these objectives, we will produce open source environments compatible with audio standards that will be shared with the computer music community and  evaluated through composer commissions, creations, workshops and concerts.

The G3S project is funded by the European Research Council (ERC), Grant No. 101199875 and hosted at Université Paris 8.