AMD FSR antialiasing discussion

  • Thread starter Deleted member 90741
  • Start date
I find it hard to believe that 4A programmers would somehow suggest that FidelityFX suite as a whole is at odds with their temporal AA/upscaling solution.
What was said initially makes sense, their corrected statement doesn't. Conspiracy or not.
 
curious to see how that works; is it just an API call to FSR so that you don't have to do any of the other setup required? I hope it's swappable with future FSR algorithms so that published titles can just swap the algorithm for the latest one. Though given the state of it, may not be swappable if future FSR goes deep learning, perhaps different inputs are required.
 
Regarding used algorithms, GPUOpen site has been updated recently with this:

How it works
FidelityFX Super Resolution is a spatial upscaler: it works by taking the current anti-aliased frame and upscaling it to display resolution without relying on other data such as frame history or motion vectors.

At the heart of FSR is a cutting-edge algorithm that detects and recreates high-resolution edges from the source image. Those high-resolution edges are a critical element required for turning the current frame into a “super resolution” image.

FSR provides consistent upscaling quality regardless of whether the frame is in movement, which can provide quality advantages compared to other types of upscalers.
FSR is composed of two main passes:

  • An upscaling pass called EASU (Edge-Adaptive Spatial Upsampling) that also performs edge reconstruction. In this pass the input frame is analyzed and the main part of the algorithm detects gradient reversals – essentially looking at how neighboring gradients differ – from a set of input pixels. The intensity of the gradient reversals defines the weights to apply to the reconstructed pixels at display resolution.
  • A sharpening pass called RCAS (Robust Contrast-Adaptive Sharpening) that extracts pixel detail in the upscaled image.
FSR also comes with helper functions for color space conversions, dithering, and tone mapping to assist with integrating it into common rendering pipelines used with today’s games.

fsr-diagram-light-nobjljp0.png

FidelityFX Super Resolution looks for gradient reversals in the source image to reconstruct high-definition edges at upscaled resolution.
https://gpuopen.com/fidelityfx-superresolution/

Let's wait for source code for more details
 
FSR is important because it can be integrated into most games that are on the market, and it can be integrated easily. That's pretty much it. Whether TAAU looks better is not really that important, because TAAU can't be integrated into games like DOTA2 which don't have TAA and are likely lacking the motion vectors etc that would be required. Most likely any game that's Unreal Engine 5 will not need FSR, because they have their own new upscaling to replace TAAU.

FSR will be much much easier to patch into games than TAAU, DLSS, dynamic resolution, variable rate shading and other custom temporal reconstructions. That's pretty much it. You get FSR where the other methods would probably never be patched in anyway.
 
FSR is important because it can be integrated into most games that are on the market, and it can be integrated easily. That's pretty much it. Whether TAAU looks better is not really that important, because TAAU can't be integrated into games like DOTA2 which don't have TAA and are likely lacking the motion vectors etc that would be required. Most likely any game that's Unreal Engine 5 will not need FSR, because they have their own new upscaling to replace TAAU.

FSR will be much much easier to patch into games than TAAU, DLSS, dynamic resolution, variable rate shading and other custom temporal reconstructions. That's pretty much it. You get FSR where the other methods would probably never be patched in anyway.

I would be really surprised if developers go through the trouble of updating older games. Unless Radeon borrows marketing money from its Zen cousin.

The interesting bit for me is to see what developers do going forward with all the options currently on the table. AMD’s best move is to muscle in on consoles where upscaling is standard practice. Will devs embrace a TAA and FSR combo? Or go with a unified solution that does both.
 
I would be really surprised if developers go through the trouble of updating older games. Unless Radeon borrows marketing money from its Zen cousin.

The interesting bit for me is to see what developers do going forward with all the options currently on the table. AMD’s best move is to muscle in on consoles where upscaling is standard practice. Will devs embrace a TAA and FSR combo? Or go with a unified solution that does both.

DOTA2 just got updated today ...
 
FSR is important because it can be integrated into most games that are on the market, and it can be integrated easily. That's pretty much it. Whether TAAU looks better is not really that important, because TAAU can't be integrated into games like DOTA2 which don't have TAA and are likely lacking the motion vectors etc that would be required. Most likely any game that's Unreal Engine 5 will not need FSR, because they have their own new upscaling to replace TAAU.

FSR will be much much easier to patch into games than TAAU, DLSS, dynamic resolution, variable rate shading and other custom temporal reconstructions. That's pretty much it. You get FSR where the other methods would probably never be patched in anyway.


Well for now you don't have a lot anyway...
 
Well for now you don't have a lot anyway...

They said they have 40-45 developers working with it, I think. So at least 45 games. I'd expect a lot of multiplayer games, especially free to play ones, that lack a good reconstruction option to add support. The more people that can play your game, they more money you make. For example I could see a game like Valorant adding support.
 
Back
Top