IlGialloMondadori
Member
No one in their right mind says, "Starfield looks insane."
Stop and think about what you're doing here. You're letting AI slop tech convince you to play fucking Starfield of all games.
-
Hey Guest. Check out your NeoGAF Wrapped 2025 results here!
Stop and think about what you're doing here. You're letting AI slop tech convince you to play fucking Starfield of all games.
This 1000% same.The cognitive dissonance from gamers and "independent" enthusiast channels on social media today has shown me how much these people are grifting their audience of followers. It's truly been something to behold. Etched forever in the halls of retardation and bad historical takes I didn't think was possible.Truly one of, if not the most embarrassing moments ever in videogame history.
funking giblet
Member
No, this DLSS5 is completely different, and is just a lighting change, the original model and paths are calculated using shaders to rasterise everything, and it's just pretending it's super raytraced but it's 100% the artists intention, really, its like they have a PS7. I swear, it's not a new painted in AI representation that is just stable because it has access to some additional data an average LORA wouldn't, no it's a super tech that is reading the triangles and that makes it better somehow. Please believe me, it looks great because it's NOT overpainted slop. It's not overpainted slop. Overpainted slop. Slop. Slop. Slop. Also, Unreal Engine 5 recreations of Mario are great and Nintendo should give that man a job.. Super Eagle mode on ZSNES was the best filter. Motion Smoothing on TVs was best in the mid 2000s, and soap operas look amazing and realistic.
Go into options....you see that little dlss5 switch....change it to dlss4.5....there....all better?
IlGialloMondadori
Member
Perfect for the new nVidia GPU boxes with AI slop models on the cover:
"When we asked 9 year old daughters which shots they preferred, 7/10 chose the DLSS ones."
IlGialloMondadori
Member
It always makes me laugh when people sign off with, "We're done here." like they just mic dropped and everyone clapped.Same people who crap on DLSS5 , were praising this basic, trash ass update, Pissr 2, just hours ago.
How can you say this is amazing
And this is terrible?
We are done here. Some folks need new eyes. Or are just trolling.
Yes, that IS terrible. It looks like someone cranked the sharpness to max and added a massive spotlight above the scene. DLSS5 is an AI filter along with a bunch of shit to mimic the garbage "realistic" mods out there.
Last edited:
cripterion
Member
Kids just be speaking the truth man, no investments whatsoever, don't care or know if AI's behind it
Watch Black Myth : Zong Khui using this tech and everyone saying it looks bonkers come release.
Last edited:
funking giblet
Member
IlGialloMondadori
Member
Thanks for posting this, couldn't figure out a way to post the damn image.The middle one is absolute slop, no artist ever signed off on that. Uggh. The right though? Artisanal, refined, amazing, just a dab of lighting just like the original Artist wanted if he had the power. No to the middle. Literal Sex to the right please, thanks.
It's so obvious this is an AI bullshit filter. Total garbage.
funking giblet
Member
I was going to ask, I feel it was skipped over.
Radical_3d
Member
The Venn diagram of people that likes DLSS 5 and people that find Black Myth's "art direction" acceptable is a circle, indeed. I expect many 6090 sold by when the game releases.
Yeah, I saw that one too but I usually don't engage in serious conversation in GAF and it is too early for me to think of a funny way to insert that image. Congrats to your coffee brand.I was going to ask, I feel it was skipped over.
Last edited:
midnightAI
Member
I think the issue here is really that this is all part of DLSS, so it's done as part of the output rather than part of the scene generation? What I mean by that is, using AI to accelerate and improve lighting is a REALLY good idea, the improvements there could be enormous, but by doing that they have full artistic control. With this it's screwing around with the final image in realtime, impressive but ultimately fake/uncanny valley looking. There will also be question marks over consistency, is the output the same every time you play it? Do the faces change if you come back to them later (if not that is impressive in itself to be fair)
KungFucius
King Snowflake
Maybe a 6095 or bring back SLI or something. No way a 6090 is 2 5090s when a 5090 is just a slight upgrade to a 4090 with more memory. And your looking at $6090 minimum to achieve this too.
He's a moron, that's your answer.
People who fall for this shit should have their internet turned off forever.The tech is absolutely incredible. Just because the face doesn't look exactly like the source material, it doesn't mean you should throw the baby out with the bathwater. You get real-life photorealism at relatively little cost. The technology is mind-blowing.
peish
Member
Why do you all keep blaming Nvidia for this?
If the original high quality artwork was provided by capcom bethedsa
All Nvidia is providing is their ai trainer that compresses the original artwork and ask their tensor cores to reapply the character skin in high quality than what capcom needs to manually do using traditional cuda and texturing?
If the original high quality artwork was provided by capcom bethedsa
All Nvidia is providing is their ai trainer that compresses the original artwork and ask their tensor cores to reapply the character skin in high quality than what capcom needs to manually do using traditional cuda and texturing?
This is the fundamental problem that stupid people are misdiagnosing. It's not a face filter. It's a lighting enhancer - to the point of high end CG. And all of a sudden game models designed, or at least consistent with, low quality lighting don't look quite up to scratch with photorealistic lighting. It's just slightly dissonant. But underlying all this the real story is that they've just hacked lighting rendering to skip like 15 years of RT progress. It's fucking magic.
cormack12
Gold Member
I feel this is another case of 'I'm going to be disingenuously stupid and reductive by getting caught up in the histrionics of an anti-AI movement'. All this crap about artistic intent, yet people had no qualms about running upscaling with shimmer, artifacts, aliasing, ray tracing ruined etc. Everyone wanking over RR in Crimson Desert while the rain disappears etc.
Lighting in games is a terrible problem to solve. RT/GI was meant to be the cure to this but it introduces a whole host of other problems. It was why guerilla used 'hero lighting' as a compromise. So we didn't get mawkish shots of the protagonist looking like she ate lego. It's not an uncommon problem because GI will do its own thing and you lose the authorship of the scene. What you see on screen is not what the author/artist wanted you to see. It's what they compromised on because of the reductions they needed to make. There are countless examples of how lighting trashes scenes.
DLSS 5 adds in the authorship of what the scene should look like at the end. It covers up compromises made when building the game or getting it to run - you can see it in the eyes of Grace, getting more highlights and micro contrast added instead of just a texture and reflection. Instead of having flat shading with a narrow spectrum of variation, the clothes are lit differently depending on the fabric and where it's worn. I need to see an example of brushed material like suede but I'm guessing that will also behave the same way. Or in Forza Horizon/GT7 the cars have that simple but bright, strong paintwork - yet how many have that speckle dust or pearlescent detail in real time? And the compromises on skin, well you can see them in any open world game and in a lot of non AAA studio linear games.
I'm not an advocate of all this DLSS stuff. I don't particularly like upscaling, and I really don't like framegen. But it's here to stay and honestly, this has by far the biggest impact in making games look 'better' and more consistent in how they appear. But let's be honest this is just retrospectively being applied to some games at the moment - Nvidia have already said the tools are there for devs to author how the technology is used and help them hit the goal of making the game look like THEY want it to, without the compromises of raw power. I think a lot of people will be eating crow once devs actually start using this and aligning those two things because the artsitic intent will become very clear, so I'd be cautious of going down this route as an argument.
The meme's are great. I don't know why we can't poke fun at the reveal as well as discuss the technology stacks and the benefits of it once realised. It's derangement levels of response right now. And this feeds the youtubers who want clicks from everyone who wants to feel they are right - so the demand drives the supply while common sense goes out the window.
Lighting in games is a terrible problem to solve. RT/GI was meant to be the cure to this but it introduces a whole host of other problems. It was why guerilla used 'hero lighting' as a compromise. So we didn't get mawkish shots of the protagonist looking like she ate lego. It's not an uncommon problem because GI will do its own thing and you lose the authorship of the scene. What you see on screen is not what the author/artist wanted you to see. It's what they compromised on because of the reductions they needed to make. There are countless examples of how lighting trashes scenes.
DLSS 5 adds in the authorship of what the scene should look like at the end. It covers up compromises made when building the game or getting it to run - you can see it in the eyes of Grace, getting more highlights and micro contrast added instead of just a texture and reflection. Instead of having flat shading with a narrow spectrum of variation, the clothes are lit differently depending on the fabric and where it's worn. I need to see an example of brushed material like suede but I'm guessing that will also behave the same way. Or in Forza Horizon/GT7 the cars have that simple but bright, strong paintwork - yet how many have that speckle dust or pearlescent detail in real time? And the compromises on skin, well you can see them in any open world game and in a lot of non AAA studio linear games.
I'm not an advocate of all this DLSS stuff. I don't particularly like upscaling, and I really don't like framegen. But it's here to stay and honestly, this has by far the biggest impact in making games look 'better' and more consistent in how they appear. But let's be honest this is just retrospectively being applied to some games at the moment - Nvidia have already said the tools are there for devs to author how the technology is used and help them hit the goal of making the game look like THEY want it to, without the compromises of raw power. I think a lot of people will be eating crow once devs actually start using this and aligning those two things because the artsitic intent will become very clear, so I'd be cautious of going down this route as an argument.
The meme's are great. I don't know why we can't poke fun at the reveal as well as discuss the technology stacks and the benefits of it once realised. It's derangement levels of response right now. And this feeds the youtubers who want clicks from everyone who wants to feel they are right - so the demand drives the supply while common sense goes out the window.
IlGialloMondadori
Member
This is nonsense.
Superposition
Member
Hero of Spielberg
Member
geary
Member
Hey that's me told great argument bro
daninthemix
Member
This is a far too sensible take. Your kind aren't welcome here.I feel this is another case of 'I'm going to be disingenuously stupid and reductive by getting caught up in the histrionics of an anti-AI movement'. All this crap about artistic intent, yet people had no qualms about running upscaling with shimmer, artifacts, aliasing, ray tracing ruined etc. Everyone wanking over RR in Crimson Desert while the rain disappears etc.
Lighting in games is a terrible problem to solve. RT/GI was meant to be the cure to this but it introduces a whole host of other problems. It was why guerilla used 'hero lighting' as a compromise. So we didn't get mawkish shots of the protagonist looking like she ate lego. It's not an uncommon problem because GI will do its own thing and you lose the authorship of the scene. What you see on screen is not what the author/artist wanted you to see. It's what they compromised on because of the reductions they needed to make. There are countless examples of how lighting trashes scenes.
DLSS 5 adds in the authorship of what the scene should look like at the end. It covers up compromises made when building the game or getting it to run - you can see it in the eyes of Grace, getting more highlights and micro contrast added instead of just a texture and reflection. Instead of having flat shading with a narrow spectrum of variation, the clothes are lit differently depending on the fabric and where it's worn. I need to see an example of brushed material like suede but I'm guessing that will also behave the same way. Or in Forza Horizon/GT7 the cars have that simple but bright, strong paintwork - yet how many have that speckle dust or pearlescent detail in real time? And the compromises on skin, well you can see them in any open world game and in a lot of non AAA studio linear games.
I'm not an advocate of all this DLSS stuff. I don't particularly like upscaling, and I really don't like framegen. But it's here to stay and honestly, this has by far the biggest impact in making games look 'better' and more consistent in how they appear. But let's be honest this is just retrospectively being applied to some games at the moment - Nvidia have already said the tools are there for devs to author how the technology is used and help them hit the goal of making the game look like THEY want it to, without the compromises of raw power. I think a lot of people will be eating crow once devs actually start using this and aligning those two things because the artsitic intent will become very clear, so I'd be cautious of going down this route as an argument.
The meme's are great. I don't know why we can't poke fun at the reveal as well as discuss the technology stacks and the benefits of it once realised. It's derangement levels of response right now. And this feeds the youtubers who want clicks from everyone who wants to feel they are right - so the demand drives the supply while common sense goes out the window.
IlGialloMondadori
Member
More examples of the AI bullshit.
The ball looks absolutely fucked in motion
When his head is obscured slightly, he looks like fucking megamind
Not an AI filter my ass.
The ball looks absolutely fucked in motion
When his head is obscured slightly, he looks like fucking megamind
Not an AI filter my ass.
Superposition
Member
LMAO
funking giblet
Member
I feel this is another case of 'I'm going to be disingenuously stupid and reductive by getting caught up in the histrionics of an anti-AI movement'. All this crap about artistic intent, yet people had no qualms about running upscaling with shimmer, artifacts, aliasing, ray tracing ruined etc. Everyone wanking over RR in Crimson Desert while the rain disappears etc.
Lighting in games is a terrible problem to solve. RT/GI was meant to be the cure to this but it introduces a whole host of other problems. It was why guerilla used 'hero lighting' as a compromise. So we didn't get mawkish shots of the protagonist looking like she ate lego. It's not an uncommon problem because GI will do its own thing and you lose the authorship of the scene. What you see on screen is not what the author/artist wanted you to see. It's what they compromised on because of the reductions they needed to make. There are countless examples of how lighting trashes scenes.
DLSS 5 adds in the authorship of what the scene should look like at the end. It covers up compromises made when building the game or getting it to run - you can see it in the eyes of Grace, getting more highlights and micro contrast added instead of just a texture and reflection. Instead of having flat shading with a narrow spectrum of variation, the clothes are lit differently depending on the fabric and where it's worn. I need to see an example of brushed material like suede but I'm guessing that will also behave the same way. Or in Forza Horizon/GT7 the cars have that simple but bright, strong paintwork - yet how many have that speckle dust or pearlescent detail in real time? And the compromises on skin, well you can see them in any open world game and in a lot of non AAA studio linear games.
I'm not an advocate of all this DLSS stuff. I don't particularly like upscaling, and I really don't like framegen. But it's here to stay and honestly, this has by far the biggest impact in making games look 'better' and more consistent in how they appear. But let's be honest this is just retrospectively being applied to some games at the moment - Nvidia have already said the tools are there for devs to author how the technology is used and help them hit the goal of making the game look like THEY want it to, without the compromises of raw power. I think a lot of people will be eating crow once devs actually start using this and aligning those two things because the artsitic intent will become very clear, so I'd be cautious of going down this route as an argument.
The meme's are great. I don't know why we can't poke fun at the reveal as well as discuss the technology stacks and the benefits of it once realised. It's derangement levels of response right now. And this feeds the youtubers who want clicks from everyone who wants to feel they are right - so the demand drives the supply while common sense goes out the window.
peish
Member
Im not AI savvy but this showcase is just the Neural texture compression that nvidia introduced with 5000 gpu last year. Only that they have progressed beyond tech demos and used in actual games
Think of NTC as the future of texturing, skinning of human and object materials
Imagine their excitement being soul crushed by internet meme patrol…
Afaik, traditional cuda texture rendering need the developers to manually set up and define each and every skins on your polygons, the wrinkles, bumps, subsurfaces, lighting. Developer also needs to create and store large material textures in vram
With NTC, developers just need to send in a high quality 3D scan of their character art into nvidia ai trainer which than pops up a smaller ai file that contains a small model into vram for the tensor cores to call out and skin the objects in highest quality
Think of NTC as the future of texturing, skinning of human and object materials
Imagine their excitement being soul crushed by internet meme patrol…
Afaik, traditional cuda texture rendering need the developers to manually set up and define each and every skins on your polygons, the wrinkles, bumps, subsurfaces, lighting. Developer also needs to create and store large material textures in vram
With NTC, developers just need to send in a high quality 3D scan of their character art into nvidia ai trainer which than pops up a smaller ai file that contains a small model into vram for the tensor cores to call out and skin the objects in highest quality
funking giblet
Member
Should have stopped there.
Lukaku's First Touch
Member
I like it actually.
John Bilbo
Member
cripterion
Member
More examples of the AI bullshit.
The ball looks absolutely fucked in motion
When his head is obscured slightly, he looks like fucking megamind
Not an AI filter my ass.
Analyse frame by frame on non DLSS 5.0 footage and see the results.
You work for AMD or something?
Last edited:
peish
Member
But i was correct in the summarising whats happening here with NTC. At least i verified with chatgpt before posting it
ShaiKhulud1989
Gold Member
Proves my point. Datasets are just based on more or less stolen stock model assets.The middle one is absolute slop, no artist ever signed off on that. Uggh. The right though? Artisanal, refined, amazing, just a dab of lighting just like the original Artist wanted if he had the power. No to the middle. Literal Sex to the right please, thanks.
funking giblet
Member
No, NTC is just that, a compression method that upscales smaller textures use at the various layers of scene generation. Albedo, Normal Maps, Roughness maps etc and returns a higher quality texture using a NN upsampler trained on similar data across thousands of games. This is not what is happening here. This is using the output Framebuffer, after all those processes, to generate an image on a pre-trained model, with additional scene data being provided, such as motion vectors, lighting directions. Some of these parameters will be per pixel (motion vectors), some will be nearly prompt like, or a simple vector direction for lighting, but ultimately the work on the composited output scene, after shading and post processing has run to paint into it. It's not part of the graphics pipeline in a traditional sense, it's a post process. I imagine they can also create stencils and material maps which can be used, per pixel, to inform areas that require different treatments, or are even skipped altogether, but ultimately... It's an Image to Image LORA.
Last edited:
Ivan
Member
I think presenting this by showing existing games was the wrong way...
Showing something that people haven't seen with unfamiliar, but crazy looking characters would be a different story. And that is what this technology can bring if used correctly, like any AI.
Something fine tuned from the start could look mindblowing if done right.
Showing something that people haven't seen with unfamiliar, but crazy looking characters would be a different story. And that is what this technology can bring if used correctly, like any AI.
Something fine tuned from the start could look mindblowing if done right.
Last edited:
IlGialloMondadori
Member
funking giblet
Member
I think they should apply it to famous movie scenes, and many people will start to freak out and it will be obvious why...
Wildebeest
Member
ShaiKhulud1989
Gold Member
IlGialloMondadori
Member
Some Event Horizon shit going on. His moustache got sicker, though.
Last edited:
Its not a product its a free "option" for those who paid for there $3000 gpu.
Last edited:
PeteBull
Member
I know its silly to assume we all are logical human beings here, but just to somewhat make bit of sense lets look at this tech revealed/tested in the past:
First iteration of dlss, it was terrible and gpu's were too weak to handle it even back then on lastgen games.
Independed reviewers called it out, rightfully so and told us the truth- it wasnt worth it to get 20xx series card for dlss(or rt for that matter).
This tech needed time, adjustments and improvements of all kinds, and gpu's to actually get much stronger to use it properly- aka both software and hardware needed tons of work, and guess what- we reached time when the tech is mature enough and hardware is strong enough so we can actually use it, mainstream.
DLSS5 is more of a proof of concept as of now, future tech in its infant stage, mass market adoption will only come around ps7 launch, 8000 or even 9000 nvidia series of cards so 2036 or so
Every1 should call out its inadequacies, coz it has many(from actual results to hardware required to use it) but i wouldnt just scrap whole idea from the get go.
In the future on our rtx 9060ti and ps7 that tech will look nothing like it does now, by the time avg gamer can afford it it will have many more iterations and likely proper overhaul or three
First iteration of dlss, it was terrible and gpu's were too weak to handle it even back then on lastgen games.
Independed reviewers called it out, rightfully so and told us the truth- it wasnt worth it to get 20xx series card for dlss(or rt for that matter).
This tech needed time, adjustments and improvements of all kinds, and gpu's to actually get much stronger to use it properly- aka both software and hardware needed tons of work, and guess what- we reached time when the tech is mature enough and hardware is strong enough so we can actually use it, mainstream.
DLSS5 is more of a proof of concept as of now, future tech in its infant stage, mass market adoption will only come around ps7 launch, 8000 or even 9000 nvidia series of cards so 2036 or so
Every1 should call out its inadequacies, coz it has many(from actual results to hardware required to use it) but i wouldnt just scrap whole idea from the get go.
In the future on our rtx 9060ti and ps7 that tech will look nothing like it does now, by the time avg gamer can afford it it will have many more iterations and likely proper overhaul or three
Last edited:
IlGialloMondadori
Member
"Painting has been held back for years with challenges around lighting and making images photorealistic. As you can see, DLSS5 tremendously benefits all oil paintings, including all-time classics."
Last edited:
leizzra
Member
This is so bad. It literally looks like all that thumbnails for "Consoles vs PC" videos, where PC image is generated for bigger contrast and is a clickbait. Now we can have this in games, great!
It looks bad, it makes everything look the same (I mean game by game), it changes the look because it's exaggerating everything (everyone becomes older). And it breaks the photorealistic feel when the animation starts because it still is an in-game animation. But probably soon they will 'fix' even this (but this will be harder to sell as 'game assets').
It was ok-ish when it was only for those funny pictures, but now…
And you are really cheering for that?
I don't like the direction in which corporations are pushing the mankind. Nvidia is forcing everything to sell it's AI dominance. They forced RT (that it was one of holy grails of graphics but it was and actually still is too early for that). They are forcing PT (which is even more problematic for hardware). One of the reasons we need upscaling/reconstructing from lower resolutions is that they are pushing too much. So now we need AI to do that. And even though that DLSS and FSR (now even PSSR) are becoming good at it, it still degrades the image. I was playing RE9 on PS5 Pro and the image has many instabilities, noises, strange behaviors. Pragmata's demo on PC (5090) also had issues. Now Nvidia will be forcing more 'realistic' character.
I see that at least some influencers are seeing the problems (at least for now
). On the other hand seeing other people cheering for that, it is already over, like the whole AI market. I don't think that betting everything on AI (like hardware and rendering development) will end well. Just like Game Pass wasn't good for Xbox but some people were thrilled about it.
So yeah, no for this deepfake, soft generative slop. And if characters in Starfield look bad, blame the developers and don't give them easy/lazy/cheap 'fix'.
It looks bad, it makes everything look the same (I mean game by game), it changes the look because it's exaggerating everything (everyone becomes older). And it breaks the photorealistic feel when the animation starts because it still is an in-game animation. But probably soon they will 'fix' even this (but this will be harder to sell as 'game assets').
It was ok-ish when it was only for those funny pictures, but now…
And you are really cheering for that?
I don't like the direction in which corporations are pushing the mankind. Nvidia is forcing everything to sell it's AI dominance. They forced RT (that it was one of holy grails of graphics but it was and actually still is too early for that). They are forcing PT (which is even more problematic for hardware). One of the reasons we need upscaling/reconstructing from lower resolutions is that they are pushing too much. So now we need AI to do that. And even though that DLSS and FSR (now even PSSR) are becoming good at it, it still degrades the image. I was playing RE9 on PS5 Pro and the image has many instabilities, noises, strange behaviors. Pragmata's demo on PC (5090) also had issues. Now Nvidia will be forcing more 'realistic' character.
I see that at least some influencers are seeing the problems (at least for now
). On the other hand seeing other people cheering for that, it is already over, like the whole AI market. I don't think that betting everything on AI (like hardware and rendering development) will end well. Just like Game Pass wasn't good for Xbox but some people were thrilled about it.So yeah, no for this deepfake, soft generative slop. And if characters in Starfield look bad, blame the developers and don't give them easy/lazy/cheap 'fix'.
Last edited:
CrustyBritches
Gold Member
Oh I love this movie. The end fight scene is the best.Some Event Horizon shit going on. His moustache got sicker, though.