Your giff conversion has anti aliased the part that is most offending, slow the rate to 1/8 and lets see if the giff preserves the noise.It can be even seen in gif form:
When it comes to image quality it does, difference is in performance.
Yeah, that makes it super obvious. Surprisingly not just the grate either, but almost the entire image has some instability. Even the RT reflections have that wonky issue. It's probably something that can be fixed in future.It can be even seen in gif form:
The video is a stacked comparison for the PC and I only caught it because of your gif, the frame-rate on the PC is off the scale and being downsampled as a massive cheat for comparing stability,
Oh that's bullshit and you know it.
Some very obvious differences which somehow DF had failed to recognize.. as usual lol
No it isn't, the light pulsing on the Pro up and down cover 2 or more times the distance on the PC in the same time, even though it is very faint so I only locked in on it for the PC given how it stood out on the Pro, and went to compare thinking it was noise.
This, and we can still see the fizzling even with this lower quality side-by-side. I do like the smoother lines of PSSR, but the instability is much more distracting.different upload:
different upload:
What?
I wasn;t accusing you.
Mate, he literally just captured a few seconds of the video and then converted the clip into a gif. He didn't change anything. How can we have a serious discussion when you're always out there being more paranoid than Bin Laden?
I am guessing the ps5 has lower value sharpness compare to even the pc version at zero. Look at the box at 12:10, the pro almost eliminate most the aliasing but appear to look softer. If you take look from left to right, it seem obvious the setting go from lower to high.
So Alex didn't sync the video perfectly, to demonstrate an effect that had nothing to do with the animation of Ratchet.
So what is the explanation of the animations not synching? And are you saying that you can fairly compare temporal instability if using 2 or 4 times the number of animation frames on PC for the same animation length on Pro or not?
Because doesn't makes sense to compare a Pro with a 4.000$ monster that Alex rocks. You have the same bad habit as Batalla of talking about PC as it where a system.Except you can use much higher settings on PC:
Alex had to limit PC version to PS5 performance mode for fair comparison.
No there is at least twice the time of animation on the PC condensed into the same duration as the Pro animation - is what I saying appears in that clip.
PSSR blurry? hell no
It even beats THE native 4K
The foliage looks absolutely incredible
Damn! We definitely needed as many reasonable people as possible here.
But the instability of the aliasing on the DLSS image is in-effect being fast-forwarded and the 60fps downsampling is blending away the instability that will be present because of the stair step aliasing in the stills of the DLSS footage.
I'm keeping my hopes up that the PS6 will have enough power to change the script. Until then is a hard pass.
Forget the noise in the pro image, lets talk about why this isn't a comparison lock at 60fps on the PC and the animation isn't in-synch and appears twice as fast, making the comparison of temporal instability illegitimate.
Interesting. So the pro is not really taking the performance with PSSR, but most likely the fidelity mode. More grass and post processing from fidelity mode all confirm from this video is in the pro mode.
Give it to me, now!
FF Rebirth screenshots are absolutely hysterical though. Looks good on Pro, but the original is just beyond laughable.
Damn, we definitely needed as many reasonable people as possible here..
The temporal shimmering on fine details is one of the most bothersome elements of TAA and FSR, even DLSS isn't perfect. I hope they clean it up because it's one of the most distracting elements of games now. In 4K minor aliasing artifacts aren't that disturbing when actually playing games, but pixel crawl and fine elements flickering like cables and fencing+ghosting can be really distracting. That said the PSSR seems much, MUCH better at it than FSR.
If you take a 120 FPS animation and throw away half the frames to display in a 60 FPS video, how is the 60 FPS animation any different from one that "natively" rendered at 60 FPS?
I think (guessing) his argument is that at higher framerates the differences between each frame will be smaller and therefore the temporal accumulation will have less errors and end up being more stable. I don't know if that's accurate or not.
different upload:
What?
That's perfectly true, but my question was about the animation. His argument is that because the animation is different, Alex must have run at a different frame rate. Where as, as I see it, if Alex had downsampled from 120 FPS footage, we would expect the animation to be the same, but the image quality to be cleaner due to DLSS having more samples to work with.
I know I said this in another thread, but thanks for making these comparisons. Seeing them like this is so much more informative than just side-by-side.
The most impressive aspect of PSSR is in its ability to remove specular aliasing, giving it almost a prerendered CG look
It shows in the latest Ratchet video all around
Crazy, sounds like there is a lot of graphic potential being left on the table.
"Here Square, let us fix your IQ." - Sony
