• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

DLSS 5 - Generative AI

kram182 said:
1) Still does nothing for the lighting/physics aspect of it, all tied to things like color and perception in the real world which is why we also have PBR and raytracing/pathtracing etc

2) you just previously claimed you can't see well and don't pixel peep, so how would you know to be frank?

3) People claimed uncanny valley with Playstation 2 and Playstation 3 level graphics etc as well, yet here we are well beyond that and what was then an uncanny valley is quaint and dated today, nor did it hold us back from even more realistic graphics and the strive for yet even more still
Click to expand...
Correct me if I am wrong but I thought that the uncanny valley gets worse the closer you get to almost human but not human. If true then it should be worse now. A 2D stick figure will not produce the effect but a 3D almost human figure would.
 
TangledThornz said:
Just look at that FPS drop tho.


View: https://www.youtube.com/watch?v=tTowPWp3cQc
Click to expand...

Can clearly see @1:28 how it misunderstands the make-up and thinks it's lighting on the eyelids. Also seems to love adding excessive eye shadow, even for characters where there was none in that spot.

I expect if CDPR were to incorporate this they'd set parameters to exclude or reduce it being applied to skin areas.
 
rinaldo00 said:
Correct me if I am wrong but I thought that the uncanny valley gets worse the closer you get to almost human but not human. If true then it should be worse now. A 2D stick figure will not produce the effect but a 3D almost human figure would.
Click to expand...

It 'gets summoned' anytime there's technological/graphics improvements - as shown with people talking about it with PS2 - and something never before seen (whether then, now, or tomorrow) is understandable to be 'unsettling' - but as I said, by the time 'the next uncanny valley' comes along, the old one is quaint and dated and you're used to it by then, so not an uncanny valley anymore - it's a shifting goal post just as technology and photorealism are
 
kram182 said:
It 'gets summoned' anytime there's technological/graphics improvements - as shown with people talking about it with PS2 - and something never before seen (whether then, now, or tomorrow) is understandable to be 'unsettling' - but as I said, by the time 'the next uncanny valley' comes along, the old one is quaint and dated and you're used to it by then, so not an uncanny valley anymore - it's a shifting goal post just as technology and photorealism are
Click to expand...
But the effect does get worse the closer you get to the real thing until you get out of the valley and to the real thing. So right now you would be deeper in the valley than during the PS2 days.

https://en.wikipedia.org/wiki/Uncanny_valley#/media/File:Mori_Uncanny_Valley.svg
 
rinaldo00 said:
But the effect does get worse the closer you get to the real thing until you get out of the valley and to the real thing.
Click to expand...

But my point is there's never going to be a 'peak' uncanny valley, every single time it occurs is the the worst it's ever been to people, until the next time it occurs - or until it just doesn't occur any longer because whatever has become indistinguishable at first glance and under scrutiny - in which case you could say 'the last time' it ever occurred in history was/would be 'peak' - but you also won't know what/when that is until you've crossed the Rubicon
 
kram182 said:
But my point is there's never going to be a 'peak' uncanny valley, every single time it occurs is the the worst it's ever been to people, until the next time it occurs - or until it just doesn't occur any longer because whatever has become indistinguishable at first glance and under scrutiny - in which case you could say 'the last time' it ever occurred in history was/would be 'peak' - but you also won't know what/when that is until you've crossed the Rubicon
Click to expand...
Look at the graph, there very much is a peak, or rather a valley.
 
M76 said:
If that was true the line would be straight up from the uncanny valley to an actual human. Game graphics has passed the depths of the uncanny valley in the 2000s, we are inching ever closer to 100% realism.
Click to expand...
Look at the graph it is almost a straight line for humanoid robot, to bad games were not tested.
 
rinaldo00 said:
It is not my graph, they tested people.
https://en.wikipedia.org/wiki/Uncanny_valley#
Click to expand...

And it's a conceptual hypothesis - everything tested, every baseline and plot on the chart is relative to the technology and criteria shown/of the time - people freaked out seeing The Lost World (1925) in movie theaters cause they thought the dinosaurs were real - are you saying that would that happen to people today in 2026 watching the same 1925 movie?
 
rinaldo00 said:
Look at the graph it is almost a straight line for humanoid robot, to bad games were not tested.
Click to expand...
I said straight up, not straight. This would be straight up. The bottom of the valley is at around 75-80% There is a whopping 20% to go from there at least. Game graphics has been well past that. La Noire is about 85%, heavy rain 90%, Detroit become human and death stranding are at least 95% there.

1787950359923.png
 
kram182 said:
And it's a conceptual hypothesis - everything tested, every baseline and plot on the chart is relative to the technology and criteria shown/of the time - people freaked out seeing The Lost World (1925) in movie theaters cause they thought the dinosaurs were real - are you saying that would that happen to people today in 2026 watching the same 1925 movie?
Click to expand...
No, but the PS2/3 is not over 100 years old ;) The same generation of people have gone through the timeline from the PS2 until now so that you can ask the same people where in the valley they are.

edit: you could certainly test the same people based on before and after pure raster until now.
 
Last edited:
M76 said:
I said straight up, not straight. This would be straight up. The bottom of the valley is at around 75-80% There is a whopping 20% to go from there at least. Game graphics has been well past that. La Noire is about 85%, heavy rain 90%, Detroit become human and death stranding are at least 95% there.

View attachment 822591
Click to expand...
Damn I miss typed, the line for humanoid robot out of the valley is almost straight up. That is what I meant
 
rinaldo00 said:
the PS2/3 is not over 100 years old ;) The same generation of people have gone through the timeline from the PS2 until now so that you can ask the same people where in the valley they are.
Click to expand...
Doesn't need to be 100 years old - they only need to have things supersede them and their 'uncanny valley-ness' to slide them down the graph/chart and prove it's all relative - which this very discussion on/of DLSS5 and claims of the uncanny valley with it that you're participating in prove and do - to the PS3, PS2, The Lost World movie, etc etc
 
kram182 said:
Doesn't need to be 100 years old - they only need to have things supersede them and their 'uncanny valley-ness' to slide them down the graph/chart and prove it's all relative - which this very discussion on/of DLSS5 and claims of the uncanny valley with it that you're participating in prove and do - to the PS3, PS2, The Lost World movie, etc etc
Click to expand...
Yes it is relative as people adjust but everyone has their own reaction to the the effect, some stronger than others, and it is real. My only objection is that when you use terms like moving the goal posts you make people that are truly bothered by it sound as if they just have a grudge and are faking their dislike for the look.
 
rinaldo00 said:
Yes it is relative as people adjust but everyone has their own reaction to the the effect, some stronger than others, and it is real. My only objection is that when you use terms like moving the goal posts you make people that are truly bothered by it sound as if they just have a grudge and are faking their dislike for the look.
Click to expand...

No I don't think they're faking their dislike for the look - I think they're movie goers watching the Lost World (1925) while I'm sitting here telling you Jurassic Park (1993) is still coming and you ain't seen nothing yet - so buckle up, like it or not
 
What a lot of people are calling "uncanny valley" is the mismatch of realism. The faces look real in a still screenshot, but the animations and movement don't.

IMO the animation and movement in games is so far from realistic that I don't find it uncanny at all. I'm not wondering if it's a real person or wondering what's wrong with it. I know it's not real because it's not moving anywhere close to real.
 
kram182 said:
No I don't think they're faking their dislike for the look - I think they're movie goers watching the Lost World (1925) while I'm sitting here telling you Jurassic Park (1993) is still coming and you ain't seen nothing yet - so buckle up, like it or not
Click to expand...
Well that will certainly calm debate!
 
rinaldo00 said:
Damn I miss typed, the line for humanoid robot out of the valley is almost straight up. That is what I meant
Click to expand...
That's exactly the line I highlighted, it takes at least 20% on the horizontal axis for it to go from the uncanny valley to real human.
 

“DLSS 5 Patched to Work on RTX 4000 "Ada Lovelace" GPUs Despite No Official Support​

by DrUsmanSaleem Today, 11:22 Discuss (13 Comments)
The modding community has been hard at work ever since the untimely release of DLSS-NR, a precursor to DLSS 5. Now, modders have gotten NVIDIA's leaked DLSS 5 tool working on RTX 4000 series cards, even though the tool was only built for RTX 5000 Blackwell hardware. The DLL leaked earlier this week via an early-access build of NBA 2K27, and people on the RenoDX Discord quickly got it running in other games using ReShade and RenoDX.

As we noted in our DLSS-NR breakdown, the AI model runs on FP8, which Ada Lovelace GPUs support just fine, but the compiled binaries inside the file only targeted Blackwell's sm_120 architecture, with nothing built for Ampere or Ada. A modder going by "Uncle Burrito" has apparently fixed that by hunting down the incompatible binaries and swapping in ones built for Ada. Tom's Hardware confirmed the patch works on an RTX 4080, and said other developers in the same Discord are working on their own versions of the fix.“
 
You must log in or register to reply here.
Back
Top