Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
See this is exactly what I mean from zx128 above. If it is pointed out that the open nature if FSR is a benefit to us as consumers we get some utter trolling drivel as a response.
See this is exactly what I mean from zx128 above. If it is pointed out that the open nature if FSR is a benefit to us as consumers we get some utter trolling drivel as a response.
It's a nice day, get out in the sunshine.
Some people could use the vitamin D i think.It's 27C, no clouds, no wind, likely 35C+ in the sun... I wouldn't call that a nice time to go out, especially around Noon.
Some people could use the vitamin D i think.
Yes you can use SMAA and it will have better IQ but then you will have other artefacts.Just a thought, can FSR be enabled in Avengers with FXAA or SMAA instead of TAA?
I dont quite get why FSR is a big deal. It's just another post processing AA filter.
It doesnt create higher resolution detail.
Clearly shows you don't know them well enough. I suggest you check the base resolutions of both.
It really isn't....
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:
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.
- 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 quality and DLSS quality both used the same internal rendering resolution which is 1440p. If you want to play about with the internal resolution to find advantage, say compare FSR ultra quality to DLSS. Thats FSR 1662p internal resolution vs 1440p internal resolution. If the goal is to compare both at their best. Then I can set DLSS internal resolution to native resolution and FSR ultra quality. DLSS was originally envisaged to run at native and provide better AA than TAA.
No matter what you do, DLSS quality mode is above FSR ultra quality for image quality. Also TAAU is above FSR for quality as well, without the AI network. Its just down to the fact temporal upscaling has better image quality than any spatial upscaling method. Argue however you like, temporal upscaling is just far better for image quality but slower than spatial. Temporal is also not as easy to implement.
The arguments that FSR is easier to implement does not mean much in the real world. Most major 3d engines have support already for DLSS. Its just a few clicks in the engine to enable DLSS.
Here's the thing.
You've decided you like that comparison because FSR quality uses the same internal resolution as DLSS quality. That's a specific angle and an acceptable one but you need to remember what the point of the comparison is.
I saw a comparison quite recently about FSR being able to run ultra quality with equal or better fps than DLSS quality on the same hardware. A 3090 and 5950 to be exact. This implies (small sample size) that for the same amount of effort from the hardware you can run FSR ultra quality and be better off than running DLSS quality.
So if you're the consumer and you're sitting there with your 3090 and 5950 and FSR is running ultra quality and DLSS is running quality and you're getting an excellent 4K image with the SAME hardware usage does the fact that FSR is using a higher resolution source image make it better or worse or is that irrelevant.
I'm seeing a still of "FSR quality" stuck next to a still of "DLSS quality" as if it means everything when it really doesn't.
FSR requires more resolution to be meaningful so working from the same low resolution image is favourable to DLSS.
However if FSR can produce the desired (e.g. 4K) image by working from a higher resolution image than DLSS and get equal or more FPS (small sample size) then it's earned a large gold star for getting the desired result with less effort.
If you blank this feature of FSR and insist that comparisons are only valid if working from the same original image then you're lying to yourself and others about the valid aspects of each method.
I dont quite get why FSR is a big deal. It's just another post processing AA filter.
Edge reconstruction is exactly what AA is. Posting the definition of how this AA works just supports my point further.
So yes, it really is.
What do you think MSAA does for example? It would actually be better than FSR as it works at the model level.
Remember this post by me. This is real evidence that FSR is not anywhere near DLSS for quality. This is what a real argument looks like. Not the sophistry and slander you come off with.
As I said funny how there is such a one sided application of that.
There is no reason why DLSS cannot be discussed in here when comparing the merits or disadvantages compared to FSR in a rational and unbiased way. It's the people who ignore the advantages and focus only on the negatives that are the problem.
The description posted is literally what AA does. Detects edges, uses spatial data (TAA uses temporal data as well) to improve those edges to make them look like a higher resolution.
I dont need evidence, the definition is literally what AA does. Not even sure what you want by evidence. It's like asking me for evidence a duck is a duck.
MSAA would also do a better job.
Which is exactly what I referred in my comparison. You have jet engines, which are literally doing the same thing as rocket engines do - they burn stuff, throw it out with high speed and propel the engine forward. Does that mean jet airplane is a space rocket, though, even though they use exactly same principles?
It's a silly comparison. AA techniques and FSR aren't remotely as complex as jet engines.
They are post processing filters. A better analogy for jet engines is say the entire game engine.