Friday 13th February

Tutorial: An innovative approach for game audio: Using modal synthesis to reduce memory consumption and increase sound variations

Francois Thibault, AudioKinetic

Game developers are constantly faced with having to make sound design trade offs when dealing with sound variety to reduce repetition, and the size of run-time memory available for audio data. Modal synthesis provides a high-performance solution to these very common problems while accurately modeling a wide range of resonant impact type sounds. The main advantages of using modal synthesis are:

• Synthesis transformations allows for a wide range of variations to be created.
• Production costs are reduced, as no sound variations need to be recorded.
• Audio memory footprint is highly reduced, since no dedicated audio files are used as variations and modeled audio data can be compressed more effectively than the original sounds.
• Quality and performance of the synthesis can be scaled dynamically at run-time.
• Increase interactivity by controlling synthesis parameters based on context-specific (game engine/physics) run-time parameters.

This session presentation will detail the main components of modal synthesis which consist of the signal analysis phase, transformation and the synthesis. This presentation will also include demonstrations of compelling sound transformation examples while exposing pitfalls and advantages of different implementations. Attendees will leave with a clear understanding of the real practical benefits of using modal synthesis as a tool for game audio production. From a technical perspective, attendees will also leave with a good understanding of the inner workings of real-time modal synthesis techniques and offline modal parameter estimation techniques based on linear predictive coding. This session will demonstrate that modal synthesis is realistic and ready for next generation gaming platforms.