This is funny but the whole point is completely off the mark.
I think the introduction looked pretty shit between what you said and ps4 tier models/faces but it improve when you are in the open world, the brief part in the sky temple was also good and in the city it's where it looks at its best, but ue5 this is fucking not, the game look the best when the camera is pulled away.
Yes, indeed. This was also clearly visible in the DF video. The PS5 Pro subreddit also mentions a lot of pop in.
Holy moly this game looks great, i wonder if there is any way to remove CA though?
U can see it even on 5090 with max settings(ray reconstruction included):
)
Phantom Liberty was amazing though and the continued fixing of cyberpunk really helped recover their credibility as a studio. I wish there more DLCs for cyberpunk if anything, they put a huge amount of work into creating the world and characters in that game, it's great to see it fleshed out. There's tonnes of other games to play in the meantime, I've got nothing against them taking their time and doing it properly. They're moving engines as well so they really shouldn't rush it.Finally up to 500 devs on Witcher 4. they were around 200 back in 2023 iirc. 7 years on one game and one DLC. I think devs need to stop working on DLCs. Just fix the bugs and move on. Phantom Liberty took 3 years and hundreds of devs that couldve been working on Witcher 4. Who knows maybe we wouldve had the game by now.
U can see it even on 5090 with max settings(ray reconstruction included):
Watch whole 20min and judge for urself if u can somewhat tolerate it or(like unfortunately for myself) its way too distracting, hell some ppl might not even see it(godbless their eyes) so best if u check for urself, doesnt work with some1 else telling u one way or another.
That problem exist coz game is very light on the gpu's vram, even max settings dont change that too much, so personally hoping if not for patch then a mod that can fix it- coz in my layman eyes/terms i would describe it as simply game assets/lod transitions arent using of stronger gpu's vram pool enough.
Max i ever saw was 12,8gigs vram taken on 5090 and trully max settings (native 4k,cinematic, rr, dlaa) like here:
Obviouisly its cranked to 11 if u set game settings to low like here( non remastered oblivion/gothic nostalgia warning
Under those settings pop-ins have their own pop-ins
Alex B : Unless papi Nvidia steps in with some sort of magical Deeplearning (Well pretend_they_are_learning_or_just_faking) Levelofdetail Super Slider then everything should be fine and diddy
HOLY FUCK! The differences are massive here! Groundbreaking, and she reserves her look much better than Grace in the same shot lol, that's weird. But that's OK, the tech is still in its early state & what they've shown apparently is a "snapshot" of what's to come. So excited!The most impressive thing with DLSS5 is what it does for NPC's
I mean fucking WOW
We're looking at a 2 generation difference here
VS
You're all going to love this tech when you see it running in 4K in front of you.
Yeah it's inevitable and I am so enthused by that, finally we get to see major graphical leaps!
So you're suggesting that most of the public hates this look because they've been grifted by youtubers?
It's always been there and it's very distracting.
Pop-in is always disregarded when it's imo the worst visual flaw.
My biggest pet peeve with all video games' graphics is the bending and twisting of supposedly rigid armour and costume elements when a character moves!Crimson Desert on PS5Pro looks pretty awesome. Even without the ray reconstruction from PC, the RTGI, RT reflections on water and all reflective materials, armor pieces reflecting onto itself, RT Shadows for all spot lights like torches, lamps, city lights is impressive. Only the shadows from the sun's direct lighting is raster. It's pretty ridiculous. There's a lot little details everywhere; displacement mapping even on the animals to create short fur which isn't noticeable until you get really up close. Performance mode with pssr2 image quality is pretty great too. The only big drawback is the pop-in.
That, and clipping....feels like progress on these things has slowed to a crawl. Maybe in another decade or so we'll see beards that don't clip into the rest of the model when the head moves.
As I mentioned in the other thread, the guy who responded to Daniel, Jacob Freeman, is basically an evangelist, someone involved in marketing.
All the answers he gave are exactly the same as those in the blog post published by Nvidia.
NVIDIA DLSS 5 Delivers AI-Powered Breakthrough In Visual Fidelity For Games
NVIDIA DLSS 5 infuses pixels with photorealistic lighting and materials to bridge the gap between rendering and reality.www.nvidia.com
Before jump to that conclusion, I would wait for a more detailed document or a word from developers like Bryan Catanzaro. He said that DLSS5 was the result of extensive research and cited a paper they did in 2018 about the initial research.
It wouldn't make sense for it to be just a filter, it would be added to the existing filters and not a feature of DLSS.
The rt reflections on the armor add so much. I've been begging for it since Demon souls. The first guy you arm wrestle has the coolest reflections on his helmet. And he's inside a poorly lit bar.Crimson Desert on PS5Pro looks pretty awesome. Even without the ray reconstruction from PC, the RTGI, RT reflections on water and all reflective materials, armor pieces reflecting onto itself, RT Shadows for all spot lights like torches, lamps, city lights is impressive. Only the shadows from the sun's direct lighting is raster. It's pretty ridiculous. There's a lot little details everywhere; displacement mapping even on the animals to create short fur which isn't noticeable until you get really up close. Performance mode with pssr2 image quality is pretty great too. The only big drawback is the pop-in.
Appeal to Authority now? He says it works with existing lights in the scene, but it's image based. Square that Circle. I work with AI for a living, and have written Game Engines and Rasterisers.Georgian Avasilcutei who worked on Remember Me and Life is Strange at DONTNOD, Dishonored 2 and Dishonored: Death of the Outsider at Arkane, and Hogwarts Legacy at Avalanche
Or funking giblet
Says who?It wouldn't make sense for it to be just a filter, it would be added to the existing filters and not a feature of DLSS.
I already went through this in the other thread, you absolutely can improve the pipeline by adding more maps to the input, but that isn't what it is doing yet, and unless something changes, it will be image / map based (ie: Data generated by what's visible in the framebuffer / encoded in the gbuffer). I agree with that point. But it STILL is generative AI, the same way any Image to Image LORA is, instead it can use pipeline data instead of "prompts" as such.
It doesn't scare anyone, but let's call a spade a spade. It's redrawing the frame, and there are definite technical reasons why while it looks "awesome" it won't be "correct" and "consistent" across an entire game. I use generative AI every day, and it's amazing! I'm even using it right now as we speak to help improve an old game. People have rights to be concerned that it will undermine the artist intention even if you think it looks like shit, the AI will do things that are incongruent. I might play New Vegas with all the mods, and DLSS 5 running at full blast, but this is a fundamental shift in drawing frames, that reduces the need for Artist control in the game itself / rendering tech. There is a reason so many, including yourself have to wishful think your way into saying "Oh it will be geometry aware, it will understand the lights, it will have access to the game". The answer is.. No, it won't, and doesn't need to. So what does that leave us? Do we even need frames...
