• Competitor rules

    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.

Fidelity Super Resolution in 2021

D.P. said:
The problem is you need AA or you get jaggied and ppwpixel flickering. For most modern gsme engines, the only way to do AA is TAA (not by choice, but this is literally all you can do).

do your question doesn't make sense
Click to expand...
It does make sense though. I would say that I want AA but I have turned it off in many games where TAA/FXAA was the only offering. Here we might go into the realm of subjectivity wherein for you AA is absolutely necessary but for me, and many like me, they would want AA but not if it destroys the rest of the image.
And this is the real point. You will have to have TAA in any spatial-only super resolution technique. This id why DLSS 2 woekd so well, because it is better than TAA.


But as a quick rebuttal. DLSS can run st native resolution and upscale from there, before a final downample before display. This only makes sense on old games when you sre not looking for performance but extra visual fidelity. But non one bothers backporting DLSS to old games
Click to expand...
Funnily enough, that's what I had suggested somewhere but for FSR. Also, my question was genuine and not a statement on the quality of DLSS. So no rebuttal needed.
 
Tinek said:
Right, so you do not have a point.
Click to expand...

You asked what my original point was and I linked it. Here it is again...

Wrinkly said:
The 1060 launched 5 years ago. How long do you want to keep a GPU when tech such as DLSS and raytracing is now almost 3 years old? Do note that Nvidia did add software support for RT on Pascal.
Click to expand...

Tinek said:
Unless you think everyone with Pascal should just go and buy new GPU? So then, tell me - what GPU and where? :)
Click to expand...

I think the 1060/70 has served its purpose, perhaps even the 1080. It's time to move on. We shouldn't be in the position that we get excited about console ports. What the hell happened to PC master race?

As far as where to get a new GPU from, there are plenty of people getting them and then selling on through Ebay, Facebook, CEX, etc. Put some effort in, beat the scalpers.
 
Tinek said:
In general (I assume you know but many people do not) science isn't about ditching/deleting past knowledge, as that would be very silly. Science discards what doesn't work and keeps only what works (scientific theory). And then, over time, science builds upon it, adding more and more understanding, refining math, expanding theories, joining smaller ones into bigger ones etc. In other words, it's building on top of older knowledge instead of replacing it fully, unlike what common person imagines. Hence, Newton was right in what he knew, his equations work just fine still, on Earth, as they always have. And the same will always be with Einstein's equations.
Click to expand...

This concept is nicely illustrated by Isaac Asimov’s “The relativity of wrong”… worth a read if you haven’t seen it.
 
Abhishek Shukla said:
It does make sense though. I would say that I want AA but I have turned it off in many games where TAA/FXAA was the only offering. Here we might go into the realm of subjectivity wherein for you AA is absolutely necessary but for me, and many like me, they would want AA but not if it destroys the rest of the image.
Click to expand...

I'm in this camp. A bit of an AA purist, going back to this idea of in the old days what we did was take pixel subsamples to actually increase the amount of source data that went into the final image, and average across that. It didn't just eliminate jaggies but it actually cleaned up the image very nicely especially if you did things like transparency anti-aliasing on chain link fences and vegetation, which I think was peak AA for me.

I've always looked at it from a kind of information theory point of view that if your method of AA is taking more samples, then a higher source quality produces a better output. But for things like post processing AA like FXAA all you're doing is working with the finished image which cannot add any new information in, all it can do is blur that image to mask details you don't like which if anything is lowering image quality for the sake of hiding jaggies, it actually just destroys loads of the fine detail. SMAA is very marginally better, TAA is a whole other mess blurred over several frames, you had additional temporal data but it's not fine grained subpixel data like you'd want for IQ improvement.

Quite frankly these days if I was say stuck at 1080p then I'd use a 2.0 resolution scale or SSAA to get good AA or I'd go home. I'm running at 4k so that's just not doable on current hardware although it is for much older games, it's got a high enough pixel density that while AA is still noticeable it's no where near as bad back in the day when pixels were huge. I've only really recently made an exception for DLSS 2.0 because it's basically required for RT and I saw that as a valuable trade off.
 
Last edited:
PrincessFrosty said:
like transparency anti-aliasing on chain link fences and vegetation, which I think was peak AA for me.
Click to expand...

Used to drive me up the wall making Quake 3 levels how stuff like chain link fences would mipmap into funny floating shapes and/or disappear at a distance and up close you'd see the mask edges. Sadly don't have any screenshots of it but I found a nice way around it with 2 different stages in a shader of the same texture but slightly different rendering methods, little bit of a performance hit but one level would hide the shimmering at a distance from fudging the LOD bias so it didn't get mipmap'd out of existence at a distance while smoothing the edges a bit close up with extra pixels so the edges weren't quite so hard.
 
Wrinkly said:
I think the 1060/70 has served its purpose, perhaps even the 1080. It's time to move on. We shouldn't be in the position that we get excited about console ports. What the hell happened to PC master race?

As far as where to get a new GPU from, there are plenty of people getting them and then selling on through Ebay, Facebook, CEX, etc. Put some effort in, beat the scalpers.
Click to expand...

The problem in my eyes is that new gpus cost too much money, where is the days where you could buy mid end card for 300? Gone! Now you need to pay 500-600 for 3060, and this is a joke. Not everyone can spend silly money on gpu, i will rather buy ps5 then i will spend 500 on gpu. And i bet there are a lot of people thinking the same. So thanks to AMD for bringing FSR, so you can live through these hard times. And maybe with next gen gpu the prices will fall to reasonable level.
 
linaaslt said:
The problem in my eyes is that new gpus cost too much money, where is the days where you could buy mid end card for 300? Gone! Now you need to pay 500-600 for 3060, and this is a joke. Not everyone can spend silly money on gpu, i will rather buy ps5 then i will spend 500 on gpu. And i bet there are a lot of people thinking the same. So thanks to AMD for bringing FSR, so you can live through these hard times. And maybe with next gen gpu the prices will fall to reasonable level.
Click to expand...

While you thank AMD for FSR you should also realise that GPU prices have been rising for some time due to lack of competition from AMD.
 
Wrinkly said:
While you thank AMD for FSR you should also realise that GPU prices have been rising for some time due to lack of competition from AMD.
Click to expand...

I will be the unpopular one and say that AMD Polaris was the peak price/performance architecture. I pretty much own the two extremes of that architecture, HD7970 (with which my wife still games decently) and RX590 and TBH it still is good enough for 1080/60. AMD's worst mistake was to stop production amid the pandemic, a $300 RX590 would have been like printing money without touching newer production capabilities.
 
People keep saying AMD were not competing but ignore that this was only at very top tiers (2080 Ti, 1080Ti). I get that the halo effect is a a factor but the majority of gamers had ample choice from AMD where they gave similar or better price/performance.

So this time round they decided to price the same because no matter what they do they sit around 18% - 20% market share. So they may as well make money from those who do consider AMD. Had they had ample prodiction they could have gained marketshare but they lied worse than Nvidia for a change on 6800 availablilty.
 
As someone whom was hoping for great things on this tech, at moment I feel it is a complete bust and nothing close to what was hopeful for. Coming from having a 5950x and 6900xt I was hopeful that there was a competing product with FSR but honestly so far of what we have seen (in fairness very little still) it looks worse than when DLSS 1.0 dropped and they really should have been targeting the quality of DLSS 2.0 for a competing solution. This is unfortunately not it at this time. Maybe in another 12 months it might be better but it is hard to understand how they are going to achieve performance parity to the DLSS solution because of the huge difference in how they are completing the process.
 
ICDP said:
People keep saying AMD were not competing but ignore that this was only at very top tiers (2080 Ti, 1080Ti). I get that the halo effect is a a factor but the majority of gamers had ample choice from AMD where they gave similar or better price/performance.

So this time round they decided to price the same because no matter what they do they sit around 18% - 20% market share. So they may as well make money from those who do consider AMD. Had they had ample prodiction they could have gained marketshare but they lied worse than Nvidia for a change on 6800 availablilty.
Click to expand...


AMD's cpu stock has improved but most of its wafers are still going to consoles which are constantly sold out, MS and Sony want more. You can make several times more Xbox/ps5/ryzen5000 from a wafer compared to a rx6000 gpu, so the margins on gpu is poor when factoring in the opportunity cost.

it's unlikely AND gpu stock will improve until RDNA3 in mid to late next year.

At the current rate, AMD's RX6000 GPUs may not make an appear either on the steam hardware survey for another 6 months
 
No great surprise - https://wccftech.com/fidelityfx-super-resolution-coming-to-xbox-microsoft-is-excited-about-it/

Following the recent news that Microsoft's Xbox Series S and X consoles now support AMD's FidelityFX suite, we assumed that would soon include the newest Super Resolution tool, due to launch on June 22nd for PC graphics cards. Confirmation came swiftly through a Microsoft spokesperson, who shared the following statement with IGN:

At Xbox, we’re excited by the potential of AMD’s FidelityFX Super Resolution technology as another great method for developers to increase framerates and resolution. We will have more to share on this soon.
Click to expand...
 
Not gonna lie.. that was pretty funny being blurred :D

(DLSS 1 should have missing detail and smearing on the character though lol). DLSS 2 only looks good because the image isn't moving (perf mode looking at you lol).
 
Tinek said:
Was Newton wrong? Or was his theory of gravity just incomplete and relatively basic, as we simply gained much more understanding over time? We already know there's much more to discover and learn about gravity and quantum physics, but if our current equations work (and they do) they aren't wrong now and won't be wrong in the future (as they'll work just as well). However, they might be at worst incomplete and describe only part of something bigger. It's a big difference.

In general (I assume you know but many people do not) science isn't about ditching/deleting past knowledge, as that would be very silly. Science discards what doesn't work and keeps only what works (scientific theory). And then, over time, science builds upon it, adding more and more understanding, refining math, expanding theories, joining smaller ones into bigger ones etc. In other words, it's building on top of older knowledge instead of replacing it fully, unlike what common person imagines. Hence, Newton was right in what he knew, his equations work just fine still, on Earth, as they always have. And the same will always be with Einstein's equations.
Click to expand...
i take your points and im not saying its wrong and we need to forget the old/past. science is all about improving on, however that all doesn't mean that something cannot be wrong, sometimes it just is due to a miscalculation and can be improved upon look at the calculations for 12j for black holes and how it took a long time to get the calculation to work due to understanding all of the elements involved (wrong in the beginning but got there in the end), anyway i think this is for a another topic :) but i do agree with you.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom