Bioware used AI upscaling to help remaster the Mass Effect trilogy

Reading time icon 2 min. read


Readers help support MSPoweruser. When you make a purchase using links on our site, we may earn an affiliate commission. Tooltip Icon

Read the affiliate disclosure page to find out how can you help MSPoweruser effortlessly and without spending any money. Read more

Mass Effect

Mass Effect developer Bioware has been using every tool at their disposal to bring the beloved sci-fi trilogy back in style, including AI.

With three massive RPGs to bring back in a single collection for Mass Effect: Legendary Edition, the development team utilised AI upscaling to help decrease the workload on reviving every asset in the trilogy.

Reported by GamesBeat, all of the games’ original uncompressed artwork was fed through an AI up-res program alongside some custom tools to help get the remasters rolling.

“We knew we wanted to increase the resolution on all the textures across the trilogy,” explained environment and character director at Bioware Kevin Meek. “So that means all of the visual effects, the user interface, the environment art, the character art — every single texture we hit right away with two main changes. First, we increased cap that the engine places on the texture sizes. And secondly, we ran all the original, uncompressed source art through an AI up-res program along with some other custom batch tools.”

“Across the entire trilogy, we’re talking tens of thousands of textures with a resolution increased at least four-times — if not 16-times.”

https://twitter.com/Dreamboum/status/1333249708823891968?s=20

Mass Effect: Legendary Edition isn’t the only remaster to use AI upscaling to help bring old artwork into the future. The upcoming remaster of SaGa Frontier was the first big remaster announced to use the benefits of AI upscaling. However, Bioware’s efforts to use this method of remastering is certainly the largest.

More about the topics: ai, BioWare, ea, machine learning, Mass Effect, Mass Effect Legendary Edition, next-gen

Leave a Reply

Your email address will not be published. Required fields are marked *