this post was submitted on 16 Mar 2026
72 points (98.6% liked)
Games
21262 readers
318 users here now
Tabletop, DnD, board games, and minecraft. Also Animal Crossing.
Rules
- No racism, sexism, ableism, homophobia, or transphobia. Don't care if it's ironic don't post comments or content like that here.
- Mark spoilers
- No bad mouthing sonic games here :no-copyright:
- No gamers allowed :soviet-huff:
- No squabbling or petty arguments here. Remember to disengage and respect others choice to do so when an argument gets too much
- Anti-Edelgard von Hresvelg trolling will result in an immediate ban from c/games and submitted to the site administrators for review. :silly-liberator:
founded 5 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
DLSS has to invent detail to improve an image. Up until now it tried to make the image look like the original version or how it thinks it would look like at higher resolution.
Nvidia decided that's not enough slop, so DLSS 5 will make everything look like a DALL-E image prompt. In this case, it decided to give the character a lip-filler injection and full makeup.
NGL I thought it was just different lighting making the makeup color pop a little more, but now that you mention it the lip filler is just right there.
It kinda makes sense when they just train it on the actual game running at a higher resolution. Since the training data is exactly the target image. This is stupid though.