A Plugin for Neural Audio Synthesis of Impact Sound Effects
Simpact (simulated impact)
I worked on this project as part of my undergraduate dissertation, focusing on using generative learning to create various impact sound effects. The model was trained using RAVE (A variational autoencoder for fast and high-quality neural audio synthesis) and implemented as a plugin for use in Digital Audio Workstation. This functions similarly to a sampler and controls for tonal qualities are provided. The adapted short paper of this project was recently published at the 2023 Audio Mostly conference, where it received the Best Short Paper award.
TIGA Graduate of the Year: Audio
As part of a group project, this video demonstrates my individual contribution to the completion of the audio implementation of a game. This included creating spatialised, randomly spawnable sound objects on top of ambience layers; integrating in-game parameters into background music generation; implementing the GPT API into character dialogue with text-to-speech to create dynamic interactions; and more!
Game Audio Show Reel
Work in progress...
I'm actively working on putting together exciting content, please stay tuned!