Towards a User-Friendly System for Environmental Sound Synthesis in Games

The aim of this project is to explore the creation of an intuitive procedural audio software system for synthesizing environmental effects (natural phenomena sounds) that can be implemented into the Unreal games engine. Current systems are dependent on pre-recorded assets and/or are 'in-house' solutions that are created and implemented by programmers. A need for a system that enables the facile creation of audio assets as well as their real-time manipulation has been identified. The system synthesizes the sounds in real time using physically inspired models and eliminates the need for pre-recorded assets. The synthesized natural phenomena sounds that are explored are rain, wind, thunder, and fire. The sound synthesis system will offer dynamic interaction with the real-time game environment without the need for the programmer to access lower-level parameters. Topics of exploration include whether the program might improve game development workflow by abstracting lowlevel controls and having the sound sources intuitively interact with the game environment. The system will present a user-friendly interface thereby minimising the need for knowledge of traditional programming languages. The system's audio processing is achieved using Max-MSP and links to the Unreal Engine 4 (UE4) game engine. Future work will investigate more complex interactions between the sound system and the game engine, as well as developing the system as a plugin or a standalone program.