finally you can get gen ai faces in your vidya
Every Assmold follower is furiously jacking it right now, this is their dream come true
I’m still using a 730m on my 2011 vintage dual Xeon workstation. I dunno wtf DLSS even is.
Wow that looks like shit, seems ok on environments I guess but it’s fucking uncanny on faces.
It does have a bit of an uncanny look on faces, but it could do wonders on environments. My main question is how temporal stability will be, all previous DLSS models and the AMD and Intel equivalents have shimmering and poor looking artifacts in some situations and resolutions
I don’t know with environments. It still has an uncanny ability to suck artistic style out of a shot in favour of realism because it’s trained on photographic data.
Faces look uncanny because it’s trained on photography and movies. It’s adding makeup to women that don’t have it and the lighting is photography studio lighting
Good points, and yeah, even if it looks great in environments, it’s probably not something I would play with most of the time, but it would be cool to play once or twice just to see how it all looks, although I don’t like Nvidia and don’t plan on ever owning an Nvidia card again.
Pretty disappointed this is where they are taking dlss, hopefully nobody else in the industry (amd/Intel) put resources into copying this.
it feels like they dipped everything in oil
