Deutsch
Español
Français
Italiano
Nederlands
Polski
Português
Русский
Türkçe
Svenska
Chinese
MagyarRTX Remix: Videos show impressive "AI remakes" of Tomb Raider and Need for Speed Underground 2
A few days ago, Nvidia launched the open beta test of RTX Remix, publishing a video (embedded below) showing how the toolset works. Put simply, Nvidia records the rendering process of a game, giving it access to the game's textures, level geometry and lighting.
In the RTX Remix app, games can then be manually reworked, as each object can be selected and adjusted individually. The aim is to minimize effort by, for example, allowing new textures to be created based on the original textures with just one click. Because the AI can recognize what material a texture is supposed to be, light reflections can be made more realisitic, especially when RTX Remix ray tracing is enabled. The aim is to provide missing shadow effects and create a more atmospheric mood.
It is even possible to manually adjust individual objects and create new light sources, giving modders plenty of creative license to give classic games a fresh look. In the videos embedded below, YouTuber Cycu1 shows what games can look like after customization with RTX Remix.
In Tomb Raider, an early 3D game from 1996, the redesigned lighting makes a massive difference and significantly contributes to the mood of the game. However, the texture looks extremely pixelated in places; Nvidia's AI is obviously unable to work wonders if the source material is not detailed enough.
A second video shows an RTX remix of Need for Speed Underground 2, a racing game from 2004, which looks even worse than the original in some parts - glass is barely transparent, tracks are simply too dark in many places. But the damp asphalt with new reflections certainly looks impressive for a game of that time.
This example makes it clear that, depending on the game, it may require lots of fine-tuning to achieve the best results with RTX Remix. Nevertheless, it will be exciting to see what remixes appear in the coming months and years using Nvidia's new tools.
Source(s)
