DLSS / FSR vs DRS - which do you prefer and why?

[the8thark]

DLSS - Deep Learning Super Sampling
FSR - Fidelity Super Resolution
DRS - Dynamic Resolution Scaling

For many games, there is not enough hardware power or programming optimisation for a game to run 100% of the time at a stable frame rate. So a compromise needs to be made. We have a few different compromises that we can choose from

1. Have a lower native screen resolution for the game and use software to increase the resolution via DLSS and/or FSR.

2. Have a higher native resolution and stable frame rate. However when the game gets a little too intensive, lower the screen resolution using DRS.

3. Use neither and have a higher stable native screen resolution but use a variable frame rate and lower that when needed to keep the game playable.

Which option would you prefer?
Also do you like just one option or having things like performance and graphics modes. Essentially the option to choose option 1/2 or option 3 in a game?

 

[ExLight]

I'm not even sure if I'd notice much of a gap between those.
If the game is good, then hyperealistic graphics (mainly textures) aren't a must for me to enjoy the experience, so I'll just go with whatever interferes with performance the least.

Could you describe what are the main performing characteristics of each for those that don't know the difference very well?

 

[Bowsette Plus-Ultra]

I tend to go with DLSS. Partly, because I have an RTX 3070 and so I go with the native software, but mostly because AMD's FSR just always looks a bit too blurry for me. I experimented with it in Warhammer Vermintide 2 and was surprised by how fuzzy the game looked with it enabled. I also tend to performance mode when I use it, since I've gotta have my frames.

I'm not even sure if I'd notice much of a gap between those.
If the game is good, then hyperealistic graphics (mainly textures) aren't a must for me to enjoy the experience, so I'll just go with whatever interferes with performance the least.

Could you describe what are the main performing characteristics of each for those that don't know the difference very well?

They're methods of rendering games at a lower resolution and using an automatic AI upscaler to kind of pretend render them at your native resolution. Since you're technically rendering at a lower resolution you can use upscaling to eke out a couple more frames (or make a game playable if it's particularly unoptimized). DLSS is Nvidia's proprietary method, while FSR is AMD's. Dynamic resolution is usually the in-built option.