EA May Soon Generate Variations Of In-Game Sound Effects Using AI

Video game audio and sound effects can make or break a narrative experience or any title for any matter. But creating each sound effect and then its various versions used throughout an entry can be quite straining for devs. A patent recently published by EA wants to make variations of in-game audio using AI. The devs would just need to produce the base sound effect and let AI handle the output of the many other forms needed for the title.

The patent dubbed “NEURAL SYNTHESIS OF SOUND EFFECTS USING DEEP GENERATIVE MODELS” talks about automating the creation of sound effect variations in a game using AI. The audio could sound different to reflect different emotions or intensity of attacks. For example, an in-game battle in an open field will sound vastly different from a battle in a cave. The AI could consider all these conditions and automatically create the effects.

This specification relates to generating variations of in-game sound effects using machine-learned models. According to a first aspect of this specification, there is described a computer implemented method of training a machine-learned generative model to generate sound effect variations,” mentions the EA patent in detail.

EA argues that creating various versions of sound effects can currently exhaust a lot of dev time and resources. The source material has to be used and processed manually before the result is acquired after multiple attempts. The increasing size of games also magnifies the complexity of the task for the devs.

Creating variations of sound effects for video games is a time-consuming task that grows with the size and complexity of the games themselves. The process usually comprises recording source material and mixing different layers of sounds to create sound effects that are perceived as diverse during gameplay.”

The company also acknowledges that tools already exist that can create different sound effect forms. Nevertheless, synthesizing the audio with given input can be vague and lead to a flawed quality. Compared to hand-crafted sound effects, these synthesized ones are often lacking to be used without reiterations and changes. EA touts this new system as the better alternative to current choices for its intelligent approach.

Related Content:

• EA Patents To Potentially Ban Players For Partnering Up With Enemy Team Members.
• EA Patents To Let Players Play As A Videographer In-Game To Record Gameplay.
• EA Patents To Reenact Playable Moments From Recorded Gameplay Videos.

A common way to create variations is to use procedural tools, which synthesize sound effects in real-time based on input parameters that define them. Determining the proper input parameters is not straightforward for designers and the quality of the sounds is inferior to hand-crafted sound variations.”

All in all, this EA system also comes with its limitations. The legal doc clarifies that only the audio containing no speech is the goal here.