Nvidia’s upcoming DLSS 5 technology has received a strong negative response from gamers and the industry following its initial preview, according to a report by Ars Technica.
The initial reaction from developers and players has been largely critical. Observers have raised concerns that the technology alters the intended visual style of games. Characters no longer look like themselves once processed under DLSS 5. The thumbnail and video above are a prime example where Grace Ashcroft from Resident Evil Requiem looks more like actress Amber Heard with DLSS 5 on.
Some critics described the output as overly processed, with reduced contrast and a uniform appearance. Ars Technica noted that feedback has compared the visuals to heavily filtered imagery and said the system can diminish artistic direction.
Game developer Mike Bithell said the technology appears to remove intentional artistic design from games. Jeff Talbot, senior concept artist at Gunfire Games, said the results added unnecessary detail while reducing character in the visuals. What is the point of expensive and sophisticated face capture technology if characters end up looking completely different?
New Blood Interactive CEO Dave Oshry also criticized the technology, expressing concern that future players may accept such visual changes as standard.
How DLSS 5 Works
According to Nvidia, the system relies on a game’s internal color and motion data to enhance scenes with consistent lighting and materials. This allows the model to interpret elements such as characters, hair, fabric, translucent skin, and environmental lighting conditions.
The company said this approach addresses limitations of existing generative video models, which can lack predictability and precise control.
Since its debut with the RTX 2080 in 2018, deep-learning super-sampling (DLSS) has been widely used to improve performance by upscaling images and increasing frame rates. However, DLSS 5 introduces a different approach.
Nvidia describes DLSS 5 as a real-time neural rendering model that combines generative AI with traditional rendering techniques. The company says this method can deliver photorealistic visuals comparable to those seen in film production while maintaining creative control for developers.
Unlike earlier versions that focused on upscaling or frame generation, DLSS 5 applies changes to lighting and textures using generative AI.
Nvidia plans to release DLSS 5 in the autumn.
