I personally find the sharpened look of FSR to look really bad to the point where I prefer conventional bilinear upscaling, not to mention that using FSR also uses precious compute time on weaker systems.
FSR1 is pretty bad as it’s just upscaling the static image, I agree.
FSR2/3, XeSS and DLSS are temporal, meaning they use info from the previous frames to construct a higher resolution image that gives much better results. They also need to be implemented in the game engine, meaning not every game supports them.
That might explain why I don’t have the highest opinion of FSR. I’ve only used it in Counter Strike 2 and Godot. I’ve never tried XeSS or DLSS but thanks to your comment I just learned that XeSS can apparently run non intel GPU vendors from their github page
FSR2/XeSS upscaling pretty much acts as free anti aliasing, making it look better. And you get better UI rendering.
I personally find the sharpened look of FSR to look really bad to the point where I prefer conventional bilinear upscaling, not to mention that using FSR also uses precious compute time on weaker systems.
FSR1 is pretty bad as it’s just upscaling the static image, I agree.
FSR2/3, XeSS and DLSS are temporal, meaning they use info from the previous frames to construct a higher resolution image that gives much better results. They also need to be implemented in the game engine, meaning not every game supports them.
That might explain why I don’t have the highest opinion of FSR. I’ve only used it in Counter Strike 2 and Godot. I’ve never tried XeSS or DLSS but thanks to your comment I just learned that XeSS can apparently run non intel GPU vendors from their github page