ibb.co
ibb.co
Thanks.Fucking nazi UK government.
FINAL FANTASY VII REBIRTH 10 01 2026 23 48 35 hosted at ImgBB
Image FINAL FANTASY VII REBIRTH 10 01 2026 23 48 35 hosted on ImgBBScreenshot 10 01 2026 23 47 51 hosted at ImgBB
Image Screenshot 10 01 2026 23 47 51 hosted on ImgBB
this is not path tracing. this specific issue does not happen with preset E and high settings (no path tracing). it still happens with high internal resolutions when using preset M. it doesn't even happen with TSR which I show here
Depends on the game and res. Dlaa gives a bit better graphics and that is nice if you have the power. Arc raiders is a good example. It is a luxury option.
I don't think it's worth it either but it's definitely not 0 IQ gain, DLAA 4.0 looks incredible and a step above quality.
So you are saying that those games wouldn't look any different if you displayed them at the resolution fed to PSSR? No, I didn't think so, so the source feeding PSSR is not pristine. That's a slight of hand that PSSR can only be commended when it improves games beyond their native flaws, and be criticised when anything in non-pristine or pristine games has more noise than another AI model regardless of comparison of overall composition?
He is saying that if the games is upscaling from a high resolution, like 1440p-4K then PSSR works really well. Games like Demon Souls, Astro Bot, GT7 or Spider-Man. All upscaling from a really high resolution. So of course PSSR would look great.
That describes any ML temporal upscaler.
Wtf... The asymmetric blame pressure in the UK has gone ape-shit. It's so depressing.
Pretty sure they are lower resolution than that in terms of filled pixel count - because it is a sparse native renderer feeding PSSR according to patents - even if the effective high frequency detail is effectively 1440p, so were still talking around 720p-900p in actual pixel counts being ML reconstructed to 4K, which if comparing those games at 720p-900p as non-sparse renders to the PSSR 4K would you describe that linear scaled image as a pristine source?
It isn't about denoising, and we saw this in the maximum furthest distant sections of the frustum with DLSS on Rachet, where DLSS is filling in stuff with no group of points because its non-sparse native resolution loses all high frequency detail at that projected minified range on small items, unlike PSSR which was reconstructing smaller items all the way to the back plane because of the higher frequency source detail that remained.
I haven't seen those issues in GT7, but will look for them on a tunnel track the next time I play, but is the same as I already said PSSR is doing its job, and RT denoising from lower resolution is arguably not afforded enough runtime - assuming it isn't fixed - or outside its remit in PSSR1, and done inadequately in the lighting pass - like it is done in Lumen.
Thanks! I'll try that out.Overlay probably only shows override via Nvidia app. Use dlss dev info available in dlss swapper.
Final Fantasy XVI:
Bojji
said.
The games have been pixel counted at 1440p or higher. Dragons Dogma 2 and Path of Exile 2 allows you to select the resolution you're upscaling from. Developers like Capcom have also confirmed resolutions. Like MH Wilds upscaled from 1404p.
It's not all games. Play any of the games I mentioned. PSSR struggles with noise. And denoising. Like a lot. Silent Hill 2 Quality mode looks dreadful due to PSSR. Silent Hill F as well. MGS3 has issues. Control with the PS5 Pro patch. Star Wars Outlaws is another example. Visual noise is a major, major problem with PSSR in some titles. Not all though, some genuinely look great. But if PSSR didn't have problems. developers wouldn't be going back to remove it, or make it optional, remove it outright, or why Sony is updating it with the work being done on FSR4.
The bolded literally describes every upscaling solution. You want to distance it from the rest and make it seem different, for whatever reason, but it's not.
Facts. It seems we still have people buying into the pre-release hype of PSSR that has been hit or miss at best. When you take games that were already incredible looking, PSSR looks great, but toss it some actual problem games and the issues crop up significantly compared to DLSS. The more game needs PSSR the worse it performs and that's where this new DLSS shines. Any upscaler can look great when you are running with high resolution assets and stable framerates, it's when the shit hits the fan that they separate themselves and boy did DLSS 4.5 separate itself in that department.
Anyone? I've been out of the country and barely been on the PC.
A 5080 should be able to chew anything up at 1440p.
This applies to most games, but not every single one, especially those with Path Tracing, because they are designed with future GPUs in mind. The RTX 5090 can run PT in Cyberpunk at 1440p with DLAA at 60 fps, whereas the RTX 5080 requires DLSSQ to achieve similar result. With the RTX 4080 Super, I get slightly worse results, but still fully playable.
Christ you're still going.
Just use the Nvidia app? Haven't come across any issues with 4.5 from what I've played and tried.
Anyone who has played both versions of ANY of these games like AW2, Silent Hill, etc. knows how full of shit you sound right now. There is a clear winner, it's just not the one you want. And as for "garbage in garbage out" that's the entire point of upscaling, to turn something unacceptable into something usable, so yes, it's 100% a fault of PSSR if it can't do that, and even Sony knows this which is why they are working to make it better. The fact that a PSSR toggle now exists in these games says it all.
Oh, Jesus. I didn't realize he was one of the inmates or I wouldn't even have bothered.
I'm more than happy to have my view challenged if you can do better than reply with such a low effort drive by post like yours
What played 864p native with DLSS ultra performance like he tested , and used Pro level hardware like RTX 30xx card? No, and I'm sure the most vocal people playing aren't on a card anything like that and probably still using MFG and much higher native.
What played 864p native with DLSS ultra performance like he tested , and used Pro level hardware like RTX 30xx card? No, and I'm sure the most vocal people playing aren't on a card anything like that and probably still using MFG and much higher native.
Anyway I've time stamped a section clearly showing an example of the same instability, strobing, flicker noise in both solutions- best seen shown played back at minimum speed as the full scene A and B shots are brief but easily visible at normal speed too. If it was a straight fullscene A then B video
And provided a screen grab highlighting, where to pay attention for this example's instability on both
Can you really not see that same issue? Or even see the trees transition in an ugly way on both - marginally worse with PSSR - 6secs earlier, etc, etc?
So all the selective pixel peeping that is different from the full scene composition context of what people actually see/experience when they play negates that both are ugly unstable messes in contradiction to what Alex is saying in the video?
When I opened DLSS swapper it indicated that that game and a handful of others just didnt update to 4.5. So I used the swapper to do it. I havent had a chance to try it back at what it was.
So all the selective pixel peeping that is different from the full scene composition context of what people actually see/experience when they play negates that both are ugly unstable messes in contradiction to what Alex is saying in the video?
Or is that last gif because you are so close to your screen that's genuinely how you experience games in pixelpeep vision?
edit: sorry if this hurts your eyes, but this is the full composition of the fence shot, and foreground/background wins and losses look more like a draw to me, especially when the trees are softer more tree like than alpha card look via DLSS, as does the light bleeding in DLSS onto the car that should be shadowed like the PSSR shot, but YMMD
