← Home ← Back to /v/

Thread 725291337

310 posts 68 images /v/
Anonymous No.725291337 [Report] >>725291442 >>725291701 >>725292106 >>725292249 >>725292312 >>725292339 >>725292417 >>725292461 >>725292498 >>725292556 >>725293002 >>725293141 >>725294554 >>725295141 >>725295793 >>725296504 >>725296976 >>725298245 >>725298384 >>725300742 >>725301162 >>725301268 >>725301348 >>725304463 >>725304639 >>725305209 >>725305507 >>725305553 >>725307964 >>725308857 >>725316419 >>725317370 >>725318841 >>725319786 >>725320064 >>725322049 >>725322070 >>725322251 >>725322597 >>725324995 >>725325105 >>725325685 >>725326354 >>725326998 >>725327857 >>725331864
>Unreal Engine is unoptimi-
Anonymous No.725291442 [Report] >>725291639 >>725292017 >>725292874 >>725292882 >>725321672 >>725322871 >>725325685
>>725291337 (OP)
>DLSS ultra performance mode
So in reality the game is running at 1280x720@30FPS
Anonymous No.725291639 [Report] >>725291858 >>725291963 >>725292119 >>725294449 >>725304545 >>725308867 >>725315839 >>725329984
>>725291442
DLSS rivals native, chud.
Anonymous No.725291701 [Report] >>725291915 >>725294443 >>725294581 >>725303012 >>725317541 >>725320990
>>725291337 (OP)
This game's max settings look like a mix of low+medium in other UE5 games. All the usual UE5 features like Nanite Lumen and VSMs have been removed. No shit it runs well lol it's graphically a 2019 game and it'd be embarrassing if it didn't run well.

Not that I'm excusing the piss poor performance of UE5slop like Borderlands 4 or similar technical disasters. But if you turned those games down to low settings they'd perform reasonably too, Arc Raiders just has its settings decreases hardcoded into the game instead of letting the player do it.
Anonymous No.725291802 [Report] >>725294554 >>725295275
The Finals and Arc Raiders are the only optimized UE5 games.
Anonymous No.725291858 [Report]
>>725291639
not on ultra performance mode
Anonymous No.725291915 [Report] >>725292027
>>725291701
>it's graphically a 2019 game
How can you tell the difference between that and a 2025 game other than the latter having low fps?
Anonymous No.725291963 [Report]
>>725291639
Even if the image quality was indistinguishable (it isn't) that's not the point, the point is the game is actually running at a much lower res and framerate and being upscaled because that's how DLSS works, especially in its lowest quality mode which is what we have here. Also we were calling out DLSS being a crutch used by devs to skip on optimization ever since it came out.
Anonymous No.725292017 [Report] >>725297883
>>725291442
it's actually 540p
Anonymous No.725292027 [Report] >>725292403 >>725325389
>>725291915
The 2025 game usually has better reflections, illumination, and distant LOD. All of these are fairly weak in Arc Raiders. And the 2025 game will also run like shit, that's true.
Anonymous No.725292106 [Report]
>>725291337 (OP)
>Doesn't use UE5 specific features
>Game runs well
Who would have thought?!
Anonymous No.725292119 [Report]
>>725291639
DLSS looks amazing for what it is but that's ultra performance mode, literally 33% internal render scale. You'd realistically need 8k for 33% internal render scale to come anywhere close to rivaling native.
Anonymous No.725292249 [Report] >>725292367 >>725292414 >>725292867 >>725300459 >>725305734
>>725291337 (OP)
I am pretty sure ARC doesn't use UE5. I mean, the game is optimized out the wazoo, runs like butter bro
Anonymous No.725292312 [Report] >>725297489
>>725291337 (OP)
>hallway
>standing still
>zero action
>performance mode meaning 800p at best
As soon as you start moving or some action happens the frametime and screen hallucinations are gunna be very apparent.
Anonymous No.725292339 [Report]
>>725291337 (OP)
>DLSS ultra performance mode

This renders in a sub 720p resolution lmao
Anonymous No.725292367 [Report]
>>725292249
It uses a nvidia branch of the ue5 engine which strips away some stuff.
Anonymous No.725292403 [Report]
>>725292027
The 2025 game will also smear and boil every time you walk through a door to a new environment regardless of performance. That's the real constant.
Anonymous No.725292414 [Report]
>>725292249
Same case as with Valorant:
https://technology.riotgames.com/news/valorants-128-tick-servers

They did they homework and rewrite bunch of shitty epic code from scratch
Anonymous No.725292417 [Report] >>725292519
>>725291337 (OP)
wow cool you can explore a nice looking world but do nothing but look containers and some shelves

retards
Anonymous No.725292461 [Report]
>>725291337 (OP)
But can it run on my gtx750!?
Anonymous No.725292498 [Report]
>>725291337 (OP)
>interior locations
lol
lmao
Anonymous No.725292519 [Report]
>>725292417
its a pvp game retard
same value as with any other pvp, container looting just a stalling mechanic that ties you to a place to facilitate situation where other party encounters you
Anonymous No.725292556 [Report]
>>725291337 (OP)
>DLSS Ultra Performance
Yeah I remember some games running really well back on my Intel HD4000, once I turned the graphics down to 720p.
Anonymous No.725292635 [Report] >>725292815
>benchmark
>DLSS/FSR/XeSS/FrameGen
so you didnt benchmark anything? into the trash it goes.
Anonymous No.725292815 [Report] >>725292992
>>725292635
It's a benchmark to stress CPU bottlenecking, as aggressive upscaling actually incurs a CPU cost. You'll stress the CPU more by upscaling 720p to 4k than you would by just rendering raw 720p.
Anonymous No.725292867 [Report] >>725293131
>>725292249
It is UE5, just like The Finals. It's just Embark isn't full of retards, they actually know what they're doing.
Anonymous No.725292874 [Report] >>725300641
>>725291442
it says 219 in the corner. can't you read?
Anonymous No.725292882 [Report] >>725293000
>>725291442
>30fps
It doesn't say there's any frame gen
Anonymous No.725292992 [Report] >>725293509
>>725292815
shhh this is a techlet seethe thread
Anonymous No.725293000 [Report] >>725293169 >>725293219 >>725293269 >>725293528 >>725293537 >>725303839 >>725317145
>>725292882
It says DLSS 4.
Anonymous No.725293002 [Report]
>>725291337 (OP)
>RTX 5090
That's right! You just need a graphics card that pulls 600W!
Anonymous No.725293131 [Report]
>>725292867
>. It's just Embark isn't full of retards, they actually know what they're doing.
in the graphics department, sure
anything else. fuck no
Anonymous No.725293141 [Report] >>725322703
>>725291337 (OP)
>Good results on a 5090
>Literally running on the best graphics card currently available on the general consumer market
I know we like to pretend to be retarded here but seriously, fuck off.
Anonymous No.725293169 [Report]
>>725293000
DLSS4 is a suite of features with the main one being upscaling. When people say DLSS, 90% of the time they're referring to the upscaling. When tech shill channels are talking about frame generation on top of that, they always add a disclaimer saying so.
Anonymous No.725293219 [Report]
>>725293000
DLSS4 is not inherently framegen, framegen is just a feature of dlss4.
Anonymous No.725293269 [Report]
>>725293000
Retard.
Anonymous No.725293509 [Report] >>725293623 >>725293652 >>725294539
>>725292992
>X3D cpu
>gaming stress
There is nothing that even comes close to taxing these CPUs in video games. They are dogshit at anything else because of the hardware limits put on them to stop them destroying the non cache product lines
Anonymous No.725293528 [Report] >>725293772 >>725293921 >>725303839 >>725304743
>>725293000
> pc latency goes down as you enable framegen

Can be shilling more apparent?
Anonymous No.725293537 [Report]
>>725293000
Nice digits retard.
Anonymous No.725293623 [Report]
>>725293509
>There is nothing that even comes close to taxing these CPUs in video games.
lol
enjoy blissful ignorance
Anonymous No.725293652 [Report] >>725295540
>>725293509
>There is nothing that even comes close to taxing these CPUs in video games
The Outer Worlds 2 on max settings.
Anonymous No.725293772 [Report] >>725294945 >>725303839
>>725293528
It's quite simple, some devs are retarded.
They only enable reflex when FG is enabled, this way enabling FG will lower the latency.
Anonymous No.725293921 [Report] >>725296160
>>725293528
The conveniently omitted part is that FG with Reflex (latency reduction measure) can cancel out almost or even all of the added input lag. However you can just use Reflex without FG so it's a dishonest comparison.
Anonymous No.725294443 [Report]
>>725291701
what about epic hardcoding into their engine to not run like complete ass?
Anonymous No.725294449 [Report] >>725294786
>>725291639
Can't wait for the retarded shills who insist DLSS is actually fine because they zoom in on a mostly static scene and look at text in the background.
Anonymous No.725294539 [Report]
>>725293509
>They are dogshit at anything else
Not the 9950x3d.
Anonymous No.725294554 [Report]
>>725291337 (OP)
That's the nvidia version of UE5, which is actually just mostly UE4, or rather a fork of UE5 without the stuff Epic introduced for UE5 that makes it so shitty.
>>725291802
The Final is on that same version of the engine
https://developer.nvidia.com/game-engines/unreal-engine/rtx-branch
Anonymous No.725294581 [Report]
>>725291701
>But if you turned those games down to low settings they'd perform reasonably too
No, BL4 on low settings runs worse than Arc Raiders on max settings, and the former looks like an actual, genuine PS3 game on low settings.
Anonymous No.725294786 [Report] >>725295443
>>725294449
If you have to zoom super far to see the differences, then it's good enough.
Anonymous No.725294945 [Report] >>725295063 >>725295071 >>725303839
>>725293772
Actually I'm starting to believe nvidia is disabling reflex without framegen on purpose to make framegen look and feel good by comparison. You can manually inject Reflex in any games normally using RTSS, but recently I've played a few games where I couldn't do that and there's no separate reflex options in-game
Anonymous No.725294991 [Report] >>725298829 >>725299125
This game has PS4 tier reflections. I don't care how well it runs. Seeing reflections disappear from lakes as you move the camera isn't what was promised after 7 years of RTX marketing. Makes me wonder why I ever upgraded from a 1080 Ti.
Anonymous No.725295063 [Report] >>725303839
>>725294945
Nvidia doesn't have veto over the graphical settings unless it's some direct sponsorship shit like Cyberpunk. In 99% it's the devs being retarded.
Anonymous No.725295071 [Report] >>725295586 >>725300250
>>725294945
If you want you can simply cap the framerate manually so your GPU won't hit ~100% and you don't need reflex at all.
Anonymous No.725295141 [Report] >>725295236
>>725291337 (OP)
lol
Even Quality already looks like shit, ultra performance is absolute coal
>erm I dont notice anything on quality
what way to find out you got shit eyesight/brain
Anonymous No.725295236 [Report] >>725295789
>>725295141
>Even Quality already looks like shit
Maybe on 1080p. But you got to be a turbocontrarian to claim 4K Quality mode DLSS looks like shit.
Anonymous No.725295275 [Report]
>>725291802
And Fortnite. That shit still runs to at 55-60 fps on the OG Xbox One.
Anonymous No.725295443 [Report] >>725295607
>>725294786
They zoom in to look at the few background details it does decently. They ignore how terrible it looks in motion.
Anonymous No.725295540 [Report]
>>725293652
>obsidian
Anonymous No.725295586 [Report]
>>725295071
This. Reflex is merely a safety net for games that aren't behaving well. A properly V-synched game with GPU fencing will feel very responsive.
Anonymous No.725295607 [Report] >>725295742
>>725295443
As opposed to the thing that looks superior in motion which is, uh.. what exactly? TAA? No lol. DLAA sure, but not so much better that you'd want to half the framerate.
Anonymous No.725295742 [Report] >>725296045
>>725295607
>As opposed to the thing that looks superior in motion which is, uh.. what exactly?
Native resolution.
>DLAA sure, but not so much better that you'd want to half the framerate.
>DLAA
>half the framerate
How stupid are you?
Anonymous No.725295789 [Report] >>725296151
>>725295236
It does look like shit compared to ground-truth AA'd 4K. It's only attractive because the only practical alternative is TAA.
Anonymous No.725295793 [Report]
>>725291337 (OP)
>Nvidia shill thread
Anonymous No.725295936 [Report] >>725296001
Engineers will do anything but release a proper anti-aliasing solution.
Anonymous No.725295976 [Report] >>725301172 >>725301334
Why don't you have an X3D CPU, anon? They're not expensive and eliminate 90% of "performance issues" with UE5 and other engines.
Anonymous No.725296001 [Report] >>725296058 >>725296151 >>725296221
>>725295936
We've literally had that for decades. It's called SSAA.
Anonymous No.725296016 [Report]
>we're at a point where devs have to take extra steps and still use band-aids to rein in the retardation of a game engine instead of just having an easily optimized engine to start with
Anonymous No.725296045 [Report] >>725296268
>>725295742
>Native resolution.
With what? Let's hear it. TAA? DLAA?
>How stupid are you?
I just tested 4K DLAA vs 4K DLSS Quality in Horizon. 54fps vs 79fps. "Halve the framerate" is a stretch but it's not worth losing more than 30% of your framerate for such a tiny difference.
Anonymous No.725296058 [Report] >>725323097
>>725296001
If that is true why don't modern games use that instead of TAA?
Anonymous No.725296151 [Report]
>>725295789
>>725296001
SSAA is absurdly expensive to run and nobody sane considers it a realistic solution unless you're playing very old games or your hardware is drastically overpowered for your current resolution. Modern slop already struggles to run acceptably with higher tier hardware and you want an AA method with path tracing tier performance cost? lol
Anonymous No.725296160 [Report]
proofs?
>>725293921
>can cancel out almost or even all of the added input lag
Anonymous No.725296221 [Report]
>>725296001
>hurr durr just get a 4k screen trust me goy you need it
Anonymous No.725296268 [Report] >>725296407
>>725296045
Native resolution TAA is still better than DLSS. The blur TAA adds is nothing compared to upscaling through DLSS. DLAA is better than both.
>I just tested 4K DLAA vs 4K DLSS
We're talking about DLAA vs TAA. All upscaling is shit.
>for such a tiny difference.
Not having the screen look like smeared shit when anything moves isn't a tiny difference, you dumb nigger monkey.
Anonymous No.725296407 [Report] >>725296710 >>725306345
>>725296268
>"Native resolution TAA is still better than DLSS."
Oh you're an insane contrarian, why didn't you just say so?

Anyone actually interested in this topic can test it themselves or look up comparisons on youtube, I recommend you don't listen to rabid contrarians like this one.
Anonymous No.725296504 [Report]
>>725291337 (OP)
Is that an small empty room with a dust particles and a few light sources in it and nothing else? Looka at da frameframe, maama mia bing bing wahoo.
Anonymous No.725296710 [Report] >>725297047 >>725320591 >>725322061
>>725296407
>Oh you're an insane contrarian, why didn't you just say so?
I'm a person with working eyes.
>Anyone actually interested in this topic can test it themselves or look up comparisons on youtube
The last time a retarded nigger faggot piece of shit shill like yourself tried arguing, he posted a youtube video and I actually watched it. Notice how the DLSS completely assrapes the details on the road. Now go ahead and try to post a video with as little actual motion as possible so you can pretend DLSS isn't completely worthless garbage as soon as things start moving. Fucking idiot.
Anonymous No.725296976 [Report] >>725297552
>>725291337 (OP)
That unrionically looks like blurry shit.

What in the absolute fuck has gone wrong with vidya graphics over the last decade.
Anonymous No.725297046 [Report]
>DLSS
Anonymous No.725297047 [Report] >>725297486 >>725307118
>>725296710
>600p youtube video crop of two frames that aren't showing the same place in the game world
You sure proved something here, just not what you think you proved.

https://youtu.be/I4Q87HB6t7Y?si=VbVYVqfCoNM-RAeT&t=212
Anonymous No.725297486 [Report] >>725298761
>>725297047
>600p
It was 4K with max settings to reduce compression artifacts. And compression artifacts wouldn't leave visible gravel and texture on one road and obliterate the other.
>that aren't showing the same place in the game world
You're so fucking stupid, holy shit. The car in the DLSS scene is about 3 feet in front of the other one, but they're literally the same spot. Kill yourself, you retarded shill.
>Posts a game cutscene comparing DLSS 3 and 4
I'd say to post actual gameplay and motion comparing native and DLSS4, but it's better if you just kill yourself for being such a fucking retard.
Anonymous No.725297489 [Report]
>>725292312
show us it happening with a video from your own pc
Anonymous No.725297552 [Report]
>>725296976
>judging graphics by looking at yewtube 1080p screenshots
Please stop.
Anonymous No.725297883 [Report]
>>725292017
>it's actually 540p
Ultra Performance is 540p @ 1080p
In 4k it renders at 720p
Anonymous No.725298245 [Report]
>>725291337 (OP)
>Barely anything in the scene, one tiny enclosed room, no draw distance, single character, blurry mess, hallucinated pixels
>Barely over 200 FPS on top of the line hardware

Nice self-own, nerd
Anonymous No.725298384 [Report]
>>725291337 (OP)
>tranny frames
Anonymous No.725298761 [Report] >>725299118
>>725297486
>>Posts a game cutscene comparing DLSS 3 and 4
There's native TAA vs DLSS4 50% comparisons in there at multiple segments. You'll ignore them because they don't fit your narrative.
Anonymous No.725298829 [Report] >>725298884
>>725294991
>This game has PS4 tier reflections.
Pretty much every current game has PS4 tier reflections, that's not saying much.
Anonymous No.725298884 [Report] >>725299003 >>725299226
>>725298829
No they don't. There's games with good reflections. It's just they're not worth implementing because every time you do that, you get HURRRR WHY IS MY FRAMERATE SO LOOOOOOW except a thousand times louder.
Anonymous No.725299003 [Report] >>725300012 >>725327225
>>725298884
>There's games with good reflections.
Name 10.
Bear in mind that pic related is a PS4 game and has better reflections than 99% of current games, without any form of hardware raytracing.
Anonymous No.725299008 [Report] >>725299282
Love when poor people trash a technology or feature (thats completely free) because they cant afford said tech with said free feature because theyre still stuck on their old RX480 and live in india.
Anonymous No.725299118 [Report] >>725299297 >>725299586
>>725298761
I already posted native vs DLSS4 in motion showing that DLSS raped detail. You proceeded to lie and get mad. You're a retard.
Anonymous No.725299125 [Report]
>>725294991
That's how screen-space reflections work, yes
Anonymous No.725299129 [Report] >>725315152 >>725325782 >>725325909
I don't get the DLSS hate.
>here, we have the american hamburger
>it provides calories and is delicious
>but americans demand more calories
>the meat patties get bigger and bigger
>double patties, triple patties, quadruple patties
>one day, genius chefs invent a way to pack more calories into the patties, so that one single patty is akin to four old ones
>yes, the patties are physically smaller, but they're even more performant than the old ones
>but the american burger enjoyer cries & moans about how his burger tower is only half the size it used to be
What's their problem?
Anonymous No.725299226 [Report]
>>725298884
Unless it's Id Tech.
https://youtu.be/HOL6D7VZ9nk
Anonymous No.725299282 [Report] >>725299385
>>725299008
DLSS is not free
FSR is free and can run on most gpus
XeSS isnt free but can also run on most gpus
Anonymous No.725299297 [Report] >>725299456 >>725300518
>>725299118
Lil buddy here is crying that FRAME GENERATION and DLSS3/2 removes some detail on the asphalt.

Why dont you link to the video that you screenshot btw? Is it because its like 4K at ultra performance or something + Frame Gen?
Anonymous No.725299385 [Report] >>725300179
>>725299282
>Buy Nvidia GPU
>Use DLSS, FREE.
Why is DLSS not free again?
>Buy AMD GPU
>Use FSR, FREE.
Why is DLSS not free again?

XESS isnt free?
You have to pay?
Where?
Anonymous No.725299456 [Report] >>725299509
>>725299297
NTA but you are a disingenuous retard
asphalt was just one of the cases, it will be applicable to any other small details as well

You have to be genuinely retarded or paid shill to argue otherwise, anyone here understands it
Anonymous No.725299509 [Report] >>725300686
>>725299456
im still waiting for you to link that video, why are you so scared to do so?
Anonymous No.725299586 [Report] >>725299961 >>725300518
>>725299118
You posted a 600p image of some youtube thumbnail nigger.
https://youtu.be/I4Q87HB6t7Y?si=eYJO_aTUav-vqQ9g&t=353
Native 4K TAA vs DLSS Performance, that's only 1080p being upscaled to 4k and it looks objectively better. It's sharper, clearer, just better in both macro image and micro image way.
Look at the small particles, they're moving RAPIDLY across the screen so you can't say it's not motion. And it still looks better.

I'm not even trying to change your mind btw I know you're a turbocontrarian, I'm replying more for the benefit of other anons.
Anonymous No.725299961 [Report] >>725300140 >>725300503
>>725299586
>TAA
You might as well render it at 480p. You'll get the same blurry piece of shit image quality as TAA.
Anonymous No.725300012 [Report] >>725307330
>>725299003
The nuHitman engine is genuinely amazing.
Anonymous No.725300083 [Report] >>725300237 >>725300747
>yeah bro you don't need to run it at native, dlss is magic and makes it look perfect even when you upscale from 720p
Anonymous No.725300140 [Report]
>>725299961
Why arent you linking the video? Why are you so scared?
Anonymous No.725300179 [Report] >>725300237
>>725299385
>buying AMD or Intel
Never go full retard.
Anonymous No.725300237 [Report] >>725300420
>>725300083
>i need to see all the visual grain (microscopic dots to "simulate" detail) in everything !!!
you're one of those fags that install that one retards upscaled "4K" textures for skyrim arent you?
>>725300179
You didnt answer my question you said DLSS cost money and that XESS cost money?
Anonymous No.725300250 [Report] >>725301150
>>725295071
This is a weird myth I see all the time that's just not true. Reflex massively decrease latency even if your GPU does fuck all. So I always use it on literally every games, even indies, even emulators.
Anonymous No.725300420 [Report] >>725300552
>>725300237
>goalpost already moved from "better than native" to "actually image quality doesn't matter"
lol cope
Anonymous No.725300459 [Report]
>>725292249
Optimization is what devs are supposed to handle, not the engine itself. Anyway some of those assets in OP's pic are on the fab (UE) storefront. Looks like the devs were too lazy to change parts of that example map around, since you can see wires on the top right clipping into the metal. Would have been a quick manual fix to just move the spline over slightly.
Anonymous No.725300503 [Report] >>725302553
>>725299961
Wait a minute didn't you say "Native resolution TAA is still better than DLSS" what's up? Changed your mind?
Anonymous No.725300518 [Report] >>725300552 >>725300826
>>725299297
>Lil buddy here is crying that FRAME GENERATION
Wrong again. Why are you shill monkeys so fucking stupid? That wasn't frame gen, that was just upscaling.

>>725299586
>You posted a 600p image of some youtube thumbnail nigger.
>thumbnail
No, you stupid fucking lying faggot nigger piece of shit. I just cropped images. It was from a youtube video at max settings at the same spot for comparison. You have no argument. In motion it looks like shit. You are objectively wrong.
>look at this video of slow, controlled smoothed movement with no camera movement
Sometimes I forget you shitskin shills don't actually play video games. The reason I posted a clip from the cars moving is because that's what happens when you actually play a game and stuff starts moving.
Anonymous No.725300552 [Report] >>725300678 >>725300802
>>725300420
Why do you refuse to link your video?
And Im not the guy you were arguing with, Im the guy asking for where your video is? You refuse to post it.
And you said better than native, DLAA isnt native, TAA is native.
>>725300518
How come you refuse to post your video?
Anonymous No.725300561 [Report] >>725300651
Yeah man, I love 645gb games
Anonymous No.725300641 [Report] >>725301307
>>725292874
now show us latency
Anonymous No.725300651 [Report]
>>725300561
>casually generates two thousand shaders for a simple grenade explosion
nothin personnel, kid
Anonymous No.725300678 [Report] >>725300752
>>725300552
>SAAR YOU ARE MUST LINK THE VIDEO
You already admitted that you don't give a shit about image quality. Concession accepted.
Anonymous No.725300686 [Report] >>725300752
>>725299509
fix your detector moron
more than one person here thinks you are a disingenuous lying shill
Anonymous No.725300701 [Report]
a fucking corridor
Anonymous No.725300742 [Report]
>>725291337 (OP)
Excellent, very nice. Now disable DLSS.
Anonymous No.725300747 [Report]
>>725300083
You will pay $1500+tip for an RTX 6080 16GB to play UE5 slop at the recommended settings* and you will be happy

*1080p 60fps on high settings with DLSS 5 ™ Ultra Performance* mode
Anonymous No.725300752 [Report] >>725300846 >>725300932
>>725300678
>>725300686
Still waiting anon. Why are you so scared to post the video?

And again DLAA isnt native you retard.
Anonymous No.725300802 [Report] >>725301013
>>725300552
>How come you refuse to post your video?
https://youtu.be/pSIX4G3QQz4?t=92
See the problem is that you're pretending I'm afraid or hiding something, when in reality I'm just not playing along with your retarded shitskin tech illiterate shill notion that no camera movement and only smoother controlled movements on screen is indicative of gameplay.
Anonymous No.725300826 [Report] >>725300932
>>725300518
Spamming nigger and shitskin will not improve your argument. DLSS looks better than TAA, and in most cases the tradeoff in performance of going to DLAA isn't really worth it. That's the consensus people have reached by trying this stuff out themselves and pixel peeping even the tiniest zoomed details in all kinds of comparisons. If you want to lose your mind over this and sperg out every time someone mentions it, no one is stopping you.
Anonymous No.725300846 [Report] >>725301013
>>725300752
>it's not native
It's running the game at native resolution you fucking retard.
Anonymous No.725300932 [Report] >>725301013 >>725301059
>>725300752
>And again DLAA isnt native you retard.
DLAA is absolutely native resolution, you dumb fucking nigger.

>>725300826
>Spamming nigger and shitskin will not improve your argument.
And crying about it won't make you white.
>DLSS looks better than TAA,
Wrong.
>That's the consensus people have reached
No one cares what people who don't play games and are paid to shill tech say.
Anonymous No.725301013 [Report] >>725301086 >>725301171 >>725305668
>>725300846
DLAA isnt native resolution, do you even know how DLAA works? It even uses AI to "bring out" details.
>>725300802
>1440p, DLSS Quality
No wonder it doesnt look amazing?
The OP is about 4K.
>>725300932
Nope DLAA isnt native.
Anonymous No.725301059 [Report] >>725301249
>>725300932
>No one cares what people who don't play games and are paid to shill tech say.
And everyone cares what some anon on 4chan thinks while sperging out about niggers, of course. You're the authority here, not the tech outlets who do this stuff for a living.
Anonymous No.725301086 [Report] >>725301149
>>725301013
Holy shit are you fucking retarded?
Anonymous No.725301149 [Report] >>725301281
If you want native you have to disable all post process antialiasing techniques such as TAA, DLAA, FSR "Native", XESS "Native", TSR "Native"
>>725301086
I know you are because you think DLAA is native when its not.
Anonymous No.725301150 [Report] >>725301835
>>725300250
It's not a myth it's all that reflex does.
It's dynamic FPS limiter.
Anonymous No.725301162 [Report]
>>725291337 (OP)
game needs 5090 to be optimized
Anonymous No.725301171 [Report] >>725301250
>>725301013
>Nope DLAA isnt native.
You're done arguing now.
Anonymous No.725301172 [Report]
>>725295976
I have and no it doesn't.
even Digital Shiller's videos will often prove this. You can't bruteforce shitty coding on poor engines.
I dropped Wuchang because it was incredibly stuttery no matter what I did.
I guess Ninja Gaiden 2Black and Tokyo Xtreme Racer are both awesome UE5 games and they run perfectly smooth, but these are literally PS3 games actually or made like PS3 games.
Anonymous No.725301249 [Report]
>>725301059
>people who actually play games and can try these things out are the authorities
Yes, that's right.
Anonymous No.725301250 [Report] >>725301378
>>725301171
DLAA is an AI-powered anti-aliasing technique that enhances image quality at native resolution, while "native" simply refers to the game's native, unscaled rendering. The key difference is that DLAA uses the same AI technology as DLSS but without the performance-boosting upscaling component, resulting in higher image quality than traditional methods, but with a greater performance cost than native rendering.

If DLAA is native then TAA is native right so why arent you comparing TAA?
Anonymous No.725301268 [Report]
>>725291337 (OP)
>DLSS 4
>on Ultra Performance mode

So it looks like absolute ass in motion, cool.

I've used DLSS in varios games, and it never looks even close to native.
And it always degrades into a blurry mess during fast paced combat or driving scenes.

I fucking hate contemporary AAA games.
They actually got worse in every possible way, compared to 5 years ago.
Anonymous No.725301281 [Report] >>725301396 >>725301467
>>725301149
>If you want native you have to disable all post process antialiasing techniques such as TAA, DLAA, FSR "Native", XESS "Native", TSR "Native"
Is this what they teach indian students at durgasoft?
Anonymous No.725301307 [Report]
>>725300641
you mean at the bottom right?
Anonymous No.725301334 [Report]
>>725295976
Yeah I've got a 7800X3D.
Will probably upgrade to the next one when it releases.
After two months or so, just to see if they disintegrate themselves once again
Anonymous No.725301348 [Report]
>>725291337 (OP)
>DLSS
Unreal Traversal Stutter Engine 5 only runs smoothly with upscaling and fake frames.
Anonymous No.725301378 [Report]
>>725301250
That's nice that you looked it up, but you're still a tech illiterate retard. DLAA does not use AI generation to increase the resolution. It is literally native resolution. The AA method affecting some pixels you see on screen does not mean you aren't rendering at native resolution. You're confusing terms because you're a tech illiterate monkey, and you're done arguing now.
Anonymous No.725301396 [Report] >>725301454
>>725301281
Do you even know what native games look like? Boot up a game that came out before deferred rendering.
Anonymous No.725301454 [Report] >>725301553
>>725301396
Are you esl or something? What the fuck?
Anonymous No.725301467 [Report]
>>725301281
He's trying to argue semantics to pretend that anything that affects what you see on screen isn't native, even though we're talking about native resolution/scaling. It's a very pathetic shill tactic, even for an indian.
Anonymous No.725301506 [Report]
funny enough the devs talk about how they got it running so good
they basically stripped ue5 of bullshit like light smearing and minor stuff
Anonymous No.725301553 [Report] >>725301687
>>725301454
Funny how you cant answer that yet you understand it perfectly.
Anonymous No.725301593 [Report] >>725301731 >>725301781 >>725302279 >>725304124 >>725306854 >>725307127
why does every thread like this have to be a shit flinging fest between anti dlss schizo and dlss shills
Anonymous No.725301687 [Report]
>>725301553
Do you understand that what you're saying is complete fucking gibberish because you either can't speak english or are extremely tech illiterate?
Anonymous No.725301731 [Report] >>725301951
>>725301593
>anti dlss schizo
There's nothing schizo about knowing what words mean and not wanting shitty smeared visuals.
Anonymous No.725301781 [Report]
>>725301593
Anonymous No.725301835 [Report] >>725302906
>>725301150
No it isn't, it's absolutely a myth. Reflex in fact dynamically delay the presentation of frames so they don't last longer in a queue and instead are presented as late as possible (just in time for refresh) which reduces latency. It's a full algorithm and with RTSS's reflex monitoring you can even see it working and how it progressively reduces latency until it's perfectly balanced. It works even without limiting any FPS at all (the real dev Reflex anyway will).
It will act as an FPS limiter in certain situations but that's not the primary function or the main way it reduces latency.
You can't just replace it with any regular FPS limited, that's why RTSS has the reflex option in the first place.
Anonymous No.725301951 [Report] >>725303332
>>725301731
i was referring to you but that wasn't an invitation for you to speak. pipe down monkey
Anonymous No.725302279 [Report] >>725302550
>>725301593
DLSS has just become a fucking necessary-tech stopgap for performance to justify making people upgrade their hardware, and its completely tainted all semblance of performance talk. People act like frame gen is some genuine quantifier of good tech, like DLSS has to be necessary because games are "just that demanding now", but ultimately it's the perfect shitstorm for double-sided contrarian wars.
Anonymous No.725302550 [Report] >>725302698
>>725302279
and you're adding to the fire by conflating frame generation with upscaling
Anonymous No.725302553 [Report]
>>725300503
No I didn't, that was some other faggot
Anonymous No.725302698 [Report]
>>725302550
They are not the same thing, but they are packaged together in the drivers and basically being billed together by NVIDIA themselves, so retards wouldn't know the difference even despite the phrasing of the sentence separating the two.
Anonymous No.725302906 [Report]
>>725301835
It doesn't matter how Reflex works if you can get nearly the same system latency without it.
It got tested by multiple people.
Simply make sure you have pre-rendered frames set to 0 before using frame cap.
Anonymous No.725303012 [Report] >>725303096
>>725291701
>Arc Raiders just has its settings decreases hardcoded into the game instead of letting the player do it.
But that's a good thing. The art direction has to be made with these lower settings in mind, instead of just telling the player "just play without shadows lol"
Anonymous No.725303096 [Report]
>>725303012
>instead of just telling the player "just play without shadows lol"
I'm still not sure why that's even a thing besides an excuse for shit optimized games to go "WELL TECHNICALLY YOU CAN MAKE IT PLAYABLE"
Anonymous No.725303332 [Report]
>>725301951
>pipe down monkey
Post hand.
Anonymous No.725303839 [Report]
>>725293000
>>725293528
>>725293772
>>725294945
it's this 100%. they're not retarded. they get paid by nvidia specifically to do this. reflex obviously is not connected to DLSS in any way since tons of games offer reflex (with or without boost) and don't support DLSS at all, for example CS2. but in DLSS games you keep seeing that connection. it's nvidia paying devs to tie the latency improvement to using frame-gen, to counter frame-gen's latency and pretend it doesn't exist.

>>725295063
>Nvidia doesn't have veto over the graphical settings unless it's some direct sponsorship shit like Cyberpunk
AS IF nvidia and the fucking games industry were above making deals like this without disclosing them. have you been paying any attention to the satanic shit that goes on in this industry and the Bond villain tier shit nvidia does all the time? ffs
Anonymous No.725304124 [Report] >>725304428
>>725301593
>why do corporate shills and copium huffers keep trying to shout over the people who know what they are talking about?
Because they have sales targets and a quarterly to hit.
If you buy a GPU for upscaling, you are a literal retard. Its supposed to be used on old hardware so you can get more value out of it.
Anonymous No.725304428 [Report] >>725305356
>>725304124
Man, memes aside I think I'm genuinely gonna bite the bullet for an RX 9070 or whatever's solid when it comes to my next computer. NVIDIA is kind of just such a hassle.
Anonymous No.725304463 [Report]
>>725291337 (OP)
>Remove Lumen + Nanite
>Performance is actually decent
Woah...
Anonymous No.725304545 [Report]
>>725291639
until you move the camera because it's temporal gigaslop
Anonymous No.725304639 [Report]
>>725291337 (OP)
The game removed all the resource intensive UE5 features. Of course it'll run better than most UE5 games.
Anonymous No.725304743 [Report] >>725305606
>>725293528
Obviously latency goes down when you enable DLSS 2 as it's rendering the game at a lower resolution -> more framerate --> less latency. Then with DLSS 3.5 and DLSS 4 there are no additional latency gains as the game isn't actually rendering more frames --> no latency decrease.
Anonymous No.725305057 [Report]
Tech Illiterate Retards - The Thread
Anonymous No.725305209 [Report]
>>725291337 (OP)
>DLSS
lol
Anonymous No.725305356 [Report] >>725306158
>>725304428
My main issue with current gpus is how shit the memory bus is. Large textures require better a better bus. Every GPU has 128-256 bus. This is garbage. Upscaling attempts to hide this flaw and it still does not manage to do it as you can see the errors in the image.
Maybe next year someone will put a 384 bus back on a main stream gpu because right now only the fucking 5090 has a 512bus and its not worth $2k.
Anonymous No.725305507 [Report]
>>725291337 (OP)
Runs at 90fps with zero stutters on my 4070 at 2k dlss quality raytracing on epic
Anonymous No.725305553 [Report]
>>725291337 (OP)
Yet it still manages to stutter even if it doesn't show in the frame time graph. They just swept the issues under the rug by basically using UE5 as UE4.
Anonymous No.725305606 [Report]
>>725304743
not quite. there should be a latency increase when activating frame-gen. but nvidia force reflex OFF until you activate frame-gen, then turn it on to hide the latency increase.
Anonymous No.725305668 [Report]
>>725301013
>DLAA isnt native resolution, do you even know how DLAA works? It even uses AI to "bring out" details.
The game is still being rendered at your set resolution. It's just a more accurate, less ghosting and less flickering TAA method which, by the way, is also done using native resolution frames.
Anonymous No.725305734 [Report] >>725322862
>>725292249
Its heavily modified fork of Nvidias already custom UE5. Basically they got already modified UE5 from Nvidia and modified it heavily even further. A lot of stuff like AI and other shit is being done server side like destruction in The Finals for example. At this point that engine is barely Unreal honestly just like Arkham Knights UE3 wasnt since Rocksteady made their own custom rendering pipeline and asset streaming solutions completly from scratch. NRS also used own fork of UE3 up until MK11 in 2019. At some point those custom forks shouldnt be called Unreal but Epic demands them to be. Back in the day studios built own tech using ID Tech and other engines and used custom codenames after modyfing the engine nowadays you can rewrite almost entire engine but you will have to show Unreal logo anyways lmao.
Anonymous No.725306158 [Report]
>>725305356
You don't need a 512 bit bus for larger textures. Texture streaming speeds are well above what is needed, 99.9% of the time you're actually going to be waiting on your engine to stream in the full textures rather than the GPU.
Anonymous No.725306345 [Report]
>>725296407
>youtube
Not to be that guy, but youtube is full of notoriously terrible compression. Any graphics comparison video hosted on youtube is a fucking joke, and is inherently gonna favour AI slop frames which might also have artifcating and issues, which are gonna be masked over by youtubes compression artifacting and stand out less against a native-non generated image, because YT compression will now also introduce compression artifacts to that video. If you want a fair comparison, you need to download the fuckhuge uncompressed video files and compare them.
Anonymous No.725306854 [Report] >>725307478
>>725301593
It is AMD marketers stirring shit up on purpose to make nvidia users seem retarded online. It is guerilla marketing.
Anonymous No.725307118 [Report]
>>725297047
>The main benefit vs TAA is that it doesn't blur the image any more
I know I've always been alone in this, but for years I've just turned AA off in every game, because I'd rather have a crispy as fuck native image, than a blurry image but no jaggies. I don't mind jaggies that much. Honestly, in most games when in motion I don't even notice any jaggies. They can show up in certain situations like cables and wire fences and such, sure. But apart from that? I'd rather take a high res no AA picture.
Anonymous No.725307127 [Report] >>725307216
>>725301593
All TAA is shit, that includes DLSS. Unfortunately many games were designed with a form of TAA to hide flaws in textures, lighting and hair. DLAA is the only real solution in these cases. The issue with DLSS defence force is the common issue of poorfags using it to get workable framerates but instead of accepting that their cards are nearing the end of their lifespans (or that they just bought a DoA card) they start spouting shit like how it's better than native rendering without TAA.
Anonymous No.725307216 [Report] >>725307753
>>725307127
>The issue with DLSS defence force is the common issue of poorfags using it to get workable framerates but instead of accepting that their cards are nearing the end of their lifespans (or that they just bought a DoA card) they start spouting shit like how it's better than native rendering without TAA.
That doesn't make much sense considering that DLSS results get worse the lower your monitor resolution is. It's poorfags that get the least out of it, and it's the people with 4k screens that get the most value out of it.
Anonymous No.725307330 [Report]
>>725300012
It's a shame the higher-ups at the studio are such utter retards, because the games themselves are fantastic
Anonymous No.725307478 [Report] >>725307570
>>725306854
What does AMD have to do with this? FSR4 = DLSS4 and is just as bad it's just that AMD's marketshare is so low it's not in the conversation.
Anonymous No.725307570 [Report] >>725313404
>>725307478
FSR is worse than DLSS because 4.0 is hardware locked to a single generation.
Anonymous No.725307753 [Report] >>725309605
>>725307216
That form of behavior is just natural for TAA which also looks better with higher resolution and framerate. At 4K TAA looks decent enough when not moving, but you'll be using DLSS anyway as your average AAA won't be running native 4k at framerates high enough to make TAA ghosting a non-issue. Not even the 5090 in this case. A more realistic option for a good balance for static and temporal resolution would be 1440p native, high framerate and DLAA so you can feed the system with as much previous frames as possible without any upscaling involved.
Anonymous No.725307964 [Report]
>>725291337 (OP)
You're such a wholesome ironic chungus king, OP! Your creativity knows no bounds!
Anonymous No.725308857 [Report]
>>725291337 (OP)
>all these retarded extra steps to just say "Upscale" or "Downscale"
I'm real tired of this. Just go back to the old options.
>2x
>4x
>8x
>16x
Anonymous No.725308867 [Report]
>>725291639
>signing your own posts

kinda cringe tbhdesu
Anonymous No.725309605 [Report] >>725310000
>>725307753
That's not really what I meant. I meant that as your output resolution increases, the degradation image quality lessens. Quality at 720p will look like shit, quality at 1080p will look mediocre, quality at 1440p will look solid, quality at 4k will look the same at 4k to most - even though these are all just 66% of the total resolution. I don't know exactly how it works but the higher your output resolution, the better the upscaler is at pretending it's the real deal.

TAA doesn't do this, TAAU at 66% res just looks mediocre no matter the resolution from my experience.

>A more realistic option for a good balance for static and temporal resolution would be 1440p native, high framerate and DLAA so you can feed the system with as much previous frames as possible without any upscaling involved.
I get better visual results with 1080p upscaled to 4k than I do with 1440p. My screen is 32 inches and I don't notice many major artifacts (other than volumetrics in some games!) and even when there are artifacts, the overall image quality gains of pretending to be 4K (more stability, seemingly clearer) outweigh the negatives.
Anonymous No.725310000 [Report] >>725310272
>>725309605
>I don't know exactly how it works but the higher your output resolution, the better the upscaler is at pretending it's the real deal.
The more pixels you have the more data the upscaler can pull from the surrounding pixels (namely what colour they are) the more accurate (for what of a better term) the output is. You also somewhat get into the realm of PPI mattering and with each pixel being smaller to the human eye errors become less visible. DLSS, XeSS and FSR were never designed for use below 4k really, the goal was to get good enough image quality without actually brute forcing all 8 million pixels.

This is why people who use 1080p and say upscalers are a meme do so precisely because they ARE seeing an inferior version.
Anonymous No.725310272 [Report] >>725310859
>>725310000
pixels being pushed is just one part of temporal upscalers and induced blur only a minor part of what turns off people from DLSS. The main issue is induced ghosting which varies from game to game and can look absolutely ghastly, made worse if the source is running on low framerates.
Anonymous No.725310859 [Report] >>725311001
>>725310272
Ghosting can't be overcome so when people complain about it as a rule the industry tells users to go fuck themselves as basically everything these days relies on data acquired over several frames. This will not change as too much relies on such methods so games will not have better motion clarity unless you brute force framerates which is a circular problem.
Anonymous No.725311001 [Report] >>725311278
>>725310859
You can just turn DLSS and TAA off to eliminate most if not all ghosting. If the jaggies get to you then DLAA is your best bet, assuming you aren't running the game at fucking 30 fps when your GPU needs to process a game at native resolution for once.
Anonymous No.725311278 [Report] >>725311387
>>725311001
>You can just turn DLSS and TAA off
A lot of modern games won't let you turn off TAA in game and require ini edits to do so and some break when doing so.
Anonymous No.725311387 [Report]
>>725311278
>require ini edits
I do this for most my games anyway or just outright force DLAA from the driver.
Anonymous No.725313404 [Report]
>>725307570
There was a leaked FSR4 for older cards it just runs worse but still better than FSR3
Anonymous No.725315152 [Report]
>>725299129
more like
>used to get full burgers for a good price
>now even though we're paying $10 for a burger instead of $5, the burgers are being shrinkflated
>new burger making method makes burgers appear bigger to justify the price increase
>but when you actually bite into it, a majority of that bigger size is actually air pumped into every individual part of the burger and they faked the patty size so much that you actually get less calories per dollar than the old $5 burgers
>when you complain about the newfound lack of nutrition, the company tries to gaslight you because "Well it's a bigger burger therefore more value" even though it's only cosmetically surface level bigger but this does not hold up to scrutiny when you actually bite into it and get less substance
only a fatass looks at computer stuff and sees food by the way, I'm only humoring your food allegory. Removed from this allegory, DLSS only looks OKAY in screenshots, but there's so much artifacting in motion it feels like you're seeing through the eyes of a drunk person.
Anonymous No.725315743 [Report] >>725316735
UE5 is pretty good for making animations (albeit these are for renders, not gameplay).
Anonymous No.725315839 [Report]
>>725291639
indian post
Anonymous No.725316419 [Report]
>>725291337 (OP)
>DLSS
yeeeeeeah no.
Anonymous No.725316735 [Report]
>>725315743
>UE5 is pretty good for making animations
Cinema4D renders miles better, any actual rendered does. UE is just cheaper than buying licenses for Cinema4D or different renderes for blender
Anonymous No.725317145 [Report]
>>725293000
ch-ch-ch-checked!!!
this is one of my favorite exclusively /v/ memes
Anonymous No.725317370 [Report]
>>725291337 (OP)
>5090
>dlls ultra performance
>4k = 720p
rope now.
Anonymous No.725317372 [Report] >>725320496
You guys really just don't understand anything despite watching all of these videos and talking about how unoptimized unreal engine is 24/7. ARC Raiders does not use ray tracing or nanite geometry, so the performance hit of those that you'd see in other games is not there. That's all there is to it. The "optimization" buzzword you guys throw around needs to be bannable.
Anonymous No.725317541 [Report]
>>725291701
Oh, fourth post best post apparently. It seems my job was already done here.
Anonymous No.725317542 [Report] >>725319646
anyone playing this on base ps5? Is the framerate a steady 60? Asking for a friend.
Anonymous No.725318841 [Report]
>>725291337 (OP)
>5090
>4K DLSS4
>ULTRA PERF
>Small space
lmfao
Anonymous No.725319291 [Report] >>725319527 >>725319657 >>725319836 >>725320325
Why does /v/ have this irrational hatred towards dlss?
I enable it on every game i play and literally can't notice it outside the fact my gpu temps and load drop
Anonymous No.725319527 [Report] >>725320191
>>725319291
if you cant notice the difference between dlss and dlaa you are blind
Anonymous No.725319646 [Report] >>725319831
>>725317542
Fake frames, anon.
Devs instead of properly optimize the games, they use this extremely expensive "tech" that previously was unnecesary, while still doing a bad job at using the engine instead of actually doing something good.
Devs became lazier, greedier and talentless all while hardware developers created a necessity that previously didn't existed
Anonymous No.725319657 [Report] >>725320191
>>725319291
Much like blacks and their ceiling birds, ghosting and blur is for DLSS+LCDfags just "that thing the game does". If you're accustomed to it don't think about it.
Anonymous No.725319786 [Report]
>>725291337 (OP)
>DLSS
>Ultra-performance with 4k, meaning that the internal resolution is 720p
Opinion discarded. Anything involving lowering internal resolutions are not optimizations.

In modern games, some developers are secretly reducing the internal resolution for things like shadows, for the fake optimization.
Anonymous No.725319831 [Report] >>725326717
>>725319646
PS5 doesn’t have frame gen, idiot.
Anonymous No.725319836 [Report] >>725320191 >>725321784
>>725319291
>Why does /v/ have this irrational hatred towards dlss?
Why do retards think that not wanting blurred visuals is irrational?
Anonymous No.725319985 [Report] >>725320103 >>725320892 >>725321578
This is why it’s got better performance, they cut out basic lighting features and hope retards don’t mind.
https://streamable.com/4bq5be
Anonymous No.725320064 [Report]
>>725291337 (OP)
>DLSS Ultra Performance on a 5090
bruh
Anonymous No.725320103 [Report] >>725321186
>>725319985
>and hope retards don’t mind.
Only retards like bad GI.
Anonymous No.725320191 [Report] >>725320498 >>725320502 >>725320591
>>725319836
>>725319657
>>725319527
But the visuals aren't blurred
Have you just not used it since dlss 1?
Anonymous No.725320325 [Report]
>>725319291
You don't play competitively then.

I didn't know anything about it, but as soon as I enabled it, I noticed how strangely delayed it was, in less than 5 seconds.
I tried to turn off and on more times, and I immediately could tell it was a huge problem of input lag, too big to ignore; and I wasn't even playing, but I was in a practice range.

Then I made a research which was just confirming what I knew.
Anonymous No.725320496 [Report]
>>725317372
>ARC Raiders does not use ray tracing
yes it does, but only for global illumination and the game still looks pretty good when it's turned off.
Anonymous No.725320498 [Report]
>>725320191
It's a temporal upscaling method sweetie, it will always add blur. In fact it adds blur even when not upscaling in DLAA mode. It's supposed to do that.
Anonymous No.725320502 [Report]
>>725320191
You're used to it so you can't really tell, but for people who don't upscale their games there is a significant difference between upscaling and no upscaling.
Anonymous No.725320591 [Report] >>725321518
>>725320191
>But the visuals aren't blurred
Yes they are. See
>>725296710
and stop shitposting.
Anonymous No.725320892 [Report] >>725321186
I'll take DLSS Quality mode upscaling over native TAA any day. Fuck TAA.
>>725319985
>retard discovers what screen space reflections are
Anonymous No.725320990 [Report]
>>725291701
>But if you turned those games down to low settings they'd perform reasonably too,
they absolutely do not
Anonymous No.725321186 [Report] >>725321407
>>725320892
>>725320103
>/v/ defending lightning that would’ve been embarrassing in 1998
It’s not even 6 AM in Mumbai, the workday hasn’t begun. Are you guys really doing it for free
Anonymous No.725321407 [Report]
>>725321186
That's cool. Call me when modern games start using deferred rendering with prepass lighting and shadowcasting from multiple light sources that isn't global so it doesn't look like dogshit.
Anonymous No.725321518 [Report] >>725321692
>>725320591
what gpu and resolution are you running?
Anonymous No.725321578 [Report]
>>725319985
That's basic lighting in modern games. Try it out in a non-RT title and it's SSR+shit cubemaps.
Anonymous No.725321672 [Report]
>>725291442
/thread
Anonymous No.725321692 [Report] >>725321754 >>725323553
>>725321518
Not getting into this argument with another retard who's going to talk about how it totally doesn't happen at 4K because he watched a youtube video shill it. I have a 5080 and I have 1440p and 4K displays. DLSS is blurry shit. Nothing changes that. No, I don't care about zoomed in mostly still videos or your malfunctioning eyes.
Anonymous No.725321754 [Report] >>725321909
>>725321692
it's not blurry shit. It's your dogshit 1440p screen
Anonymous No.725321784 [Report] >>725322061
>>725319836
DLSS in its latest form doesn't really blur, that's not the issue. The artifacts you can see with DLSS can manifest through volumetric effects lingering too densely, ghosting trails left by dark objects when slowly moving in front of very light matte backgrounds, moire patterns when some textures are viewed from certain angles, and an increase in disocclusion compared the previous version of DLSS.

But even at 50% it's barely any blurrier than DLAA. It's just more artifact-prone. If you think the main problem of DLSS is blur then you haven't used it in a long time, since the main strength of the latest DLSS form is having the same clarity and stability in motion as it does in still scenes.
Anonymous No.725321909 [Report] >>725322240 >>725322474
>>725321754
>It's your dogshit 1440p screen
Kill yourself. I sincerely mean that. And my 1440p monitor is better than whatever dogshit 4K display you're using.
Anonymous No.725322049 [Report]
>>725291337 (OP)
>if you just disable everything that makes it UE5 the game can run fine with UE4 optimizations
Anonymous No.725322061 [Report] >>725322518 >>725322848 >>725326251
>>725321784
>DLSS in its latest form doesn't really blur
Wrong. See this actual comparison video made to show off how good DLSS4 is and stop shitposting. It's blur and an obliteration of fine detail when things actually move. If you're not going to post a video comparison of DLSS4 vs native of a game where things (including the camera) are actually moving around during normal gameplay, then don't bother replying. You simply stating over and over again that it doesn't do that is irrelevant and I won't reply again.
>>725296710
Anonymous No.725322070 [Report]
>>725291337 (OP)
>DLSS ultra performance
That's like 720p rendering resolution
lol
Anonymous No.725322240 [Report] >>725322584
>>725321909
>mine is better
you're admitting your image quality is blurry. Mine is sharp. Your screen is dogshit if it's blurry
Anonymous No.725322251 [Report] >>725322404
>>725291337 (OP)
it seems the stuttering comes from how the assets are loaded, that shit rapes frametimes
Anonymous No.725322358 [Report] >>725322848
>OLEDs finally getting high VHz to offset sample and hold blur
>games start getting mandatory TAA/DLSS/DLAA blur
people under 30 have never seen a proper moving image
Anonymous No.725322404 [Report]
>>725322251
How unreal engine 5 does streaming is definitely heavily fucked up if even arc raiders has problems with frametimes.
Anonymous No.725322474 [Report] >>725322584
>>725321909
How come you censored your image?
Anonymous No.725322518 [Report] >>725322695
>>725322061
>It's blur and an obliteration of fine detail when things actually move.
That's TAA, which nearly all modern games use and rely on. DLSS cleans up the image, and the new Transformer model is able to maintain said cleaning even in motion, albeit with some artifacts.
>just don't use TAA
Then you don't get to play 99% of modern games, or you're fine with a flickering mess of an image and assets looking wrong.
Anonymous No.725322584 [Report] >>725322692 >>725322776
>>725322240
Post your monitor right now. I sure as shit hope it's some LCD.

>>725322474
I just posted the resolution of one display because retarded tech illiterate faggots keep insisting that resolution is the only thing that matters and that at 4K, DLSS is great. They're idiots and it's not.
Anonymous No.725322597 [Report] >>725322830
>>725291337 (OP)
>1 out of 100 UE5 games runs well
>that means UE5 isn't steaming pile of garbage now
Anonymous No.725322606 [Report]
yeah just get a $2500 gpu and use dlss
Anonymous No.725322692 [Report] >>725322807
>>725322584
It's a 4k oled.
Your image is shit because you're a retard doing something wrong or you have a clown monitor
Anonymous No.725322695 [Report]
>>725322518
>no video posted showing native resolution vs DLSS4 in motion during normal gameplay
And you're done. Dumb fucking shill.
Anonymous No.725322703 [Report]
>>725293141
Nutted to NITR-401, thanks anon
Anonymous No.725322776 [Report] >>725322807
>>725322584
You censored it
Anonymous No.725322807 [Report] >>725322935 >>725322984
>>725322692
>Your image is shit because you're a retard doing something wrong
That's nice, so is my TV. DLSS is still shit and no amount of you being a buttmad, tech illiterate monkey will change that.

>>725322776
Learn what words mean. Cropping and censoring aren't synonyms.
Anonymous No.725322830 [Report]
>>725322597
dont you know? the exception makes the rule!
Anonymous No.725322848 [Report] >>725323090 >>725324027 >>725324701
>>725322358
>>OLEDs finally getting high VHz to offset sample and hold blur
>>games start getting mandatory TAA/DLSS/DLAA blur
TAA blur has been an issue in games for a decade now. I swear you people don't even notice any graphical problems until some jewtubers makes a ragebait video about them.
>>725322061
So how do you propose someone records a video comparison of DLSS4 vs native where
>the camera is moving
>objects on screen are moving rapidly
>the same exact sequence has to be filmed twice for the comparison without any differences
Seems to me that you know this is basically unfeasible so you use it as your argument. I also don't understand why you think only camera motion counts - there's no rendering difference in whether new pixels are being revealed due to camera panning or new pixels being revealed due to objects on the screen moving thus pixels on the screen shifting. There isn't going to be a difference in terms of artifacting and image degradation there - which brings me to my earlier point, you just probably want to discard comparisons you know don't favour your argument.
Anonymous No.725322862 [Report]
>>725305734
That's the one positive thing I can speak about UE5, you shouldn't have to go to these depths just to get a game that doesn't run like shit, but getting access to the source code on Unity is a living nightmare so you gotta do what you gotta do.
Anonymous No.725322871 [Report]
>>725291442
kek timcvcks instantly btfo
Anonymous No.725322935 [Report] >>725324130 >>725324265
>>725322807
My PC is better than yours though so why would I be buttmad? You're the one with the fucking blurry image who can't seem to fix it. Stupid fucker lol
Anonymous No.725322984 [Report]
>>725322807
>I didn't censor it. I just cut parts of the image off so nobody can see what it says
Anonymous No.725323090 [Report]
>>725322848
>TAA blur has been an issue in games for a decade now
I've been aware of TAA being forced into games since Reach. The issue being that while TAA could usually be forced off, AAA games are leaning more and more into DLSS/FSR to get playable performance. Simply turning DLSS/FSR off might not be feasible if you want to play the game.
Anonymous No.725323097 [Report]
>>725296058
because 4k screens, you have to render over 4k. It's great when you are rendering at 4k and sampling down to 1080
Anonymous No.725323553 [Report] >>725324130
>>725321692
>I have a 5080 and I have 1440p and 4K displays.
a 5080 can't do 4k though
Anonymous No.725324027 [Report]
>>725322848
>So how do you propose someone records a video comparison of DLSS4 vs native where
What the fuck are you talking about? Start a game. Make a save. Start recording. Do something. Reload. Switch to DLSS4. Do the same thing. Stop recording.
Anonymous No.725324130 [Report]
>>725322935
Post specs and monitor. Can't wait to see your 5090 paired with mediocre parts and a shitty monitor that you're pretending looks great.

>>725323553
Neither can a 5090 without using DLSS.
Anonymous No.725324265 [Report]
>>725322935
Are you the retard from London who's completely and comically tech illiterate that thinks bigger screen size and resolution is all that matters for monitors?
Anonymous No.725324701 [Report] >>725325360 >>725325385
>>725322848
Here's the best I could do. The Finals 4K, three scenarios:
1) native without antialiasing (pay no mind to its image quality issues, this game is built around temporalslop and this example is only here so we can see what the ground truth clarity is)
2) DLAA
3) DLSS Perf
All settings equal, motion blur and all that shit off.

imgsli.com/NDI3NjA5/0/1

I moved diagonally and flicked the camera with my mouse to the right side, that's why the images are a tad misaligned. The game won't let me set up any keyboard bindings to shift the camera so winging it with the mouse was the best I could do, but there's motion here. I am moving both my character and the camera - you can't say this is somehow not a representation of what a game looks like.

I'm not seeing this insane blurriness you were talking about. No AA is definitely the sharpest and clearest because that's how unantialiased images will always be compared to temporalslop, but DLAA doesn't look much blurrier and there isn't any major quality loss by going down to DLSS Performance - if you want to find artifacts compared to DLAA you have to really zoom hard and pixel peep or watch it on a 55 inch screen.
Anonymous No.725324954 [Report]
*furious typing*
Anonymous No.725324995 [Report]
>>725291337 (OP)
>tiny room
>nothing happening
zed
Anonymous No.725325105 [Report] >>725325292 >>725325608
>>725291337 (OP)
>Looks like 2013's Metro Last Light
>Running on hardware like 15x more powerful
>Needs performance fake frames to reach 200 FPS
LOL
Anonymous No.725325292 [Report] >>725325352
>>725325105
>>Looks like 2013's Metro Last Light
that's a bit of a reach, also MLL is using tiny maps, it looks more like Exodus
Anonymous No.725325352 [Report] >>725325653
>>725325292
OP's screenshot is straight out of Metro Last Light.
And proper culling techniques and LOD's make map size mostly irrelevant.
Anonymous No.725325360 [Report]
>>725324701
And if you want video evidence for proof so you can put a pause frame from the video under the looking glass, I can upload the 2.1GB video somewhere if you like.
Anonymous No.725325385 [Report] >>725326059 >>725328626
>>725324701
Stuff moving on screen AND the camera moving. You know? That stuff that happens when you play games.
Anonymous No.725325389 [Report]
>>725292027
>usually has better reflections, illumination, and distant LOD
No, it has more intensive ones. "better" depends on how it's actually used and I've seen games from the early 2000s with better looking lighting or reflections than new shit.
Anonymous No.725325608 [Report] >>725325780
>>725325105
>>Looks like 2013's Metro Last Light
Mmm, no, it's nothing amazing, but it's not a bad-looking game.
I'd say it's more akin to the raytracing-enhanced version of Metro Exodus from like 2020.
Anonymous No.725325653 [Report] >>725325780
>>725325352
>OP's screenshot
is an image of a compressed youtube video with all the blur options turned on
Anonymous No.725325685 [Report]
>>725291337 (OP)
>>725291442
And inside a building, and that game doesn't use the worst parts, Lumen and nanite
Anonymous No.725325780 [Report]
>>725325608
>>725325653
Wow...Another generic looking brown video game. Yawn.
Anonymous No.725325782 [Report]
>>725299129
BECAUSE YOU DON'T EAT BURGERS FOR THE CALORIES YOU RETARD. YOU EAT THEM BECAUSE THEY TASTE GOOD. MORE PATTIES IS MORE FLAVOR, ADDING CALORIES TO EXISTING ONES AND THEN REDUCING THE AMOUNT YOU GET IS AN OBJECTIVE DOWNGRADE.
Anonymous No.725325909 [Report]
>>725299129
You need glasses or use a va monitor if you think ultra performance dlss looks good
Anonymous No.725326059 [Report] >>725326251
>>725325385
I am panning the camera in those images with 3200DPI, and I am moving as quickly as the game lets me. That is a valid representation of what people do when they play games, and the pixels on the screen are shifting.

The pixels will shift when still objects change their position on screen due to your character's movements or the camera's movements, and the pixels will shift the same way when moving objects change their position on screen even when your character is completely still. The same thing is occurring with the pixels, there is no distinction and the opportunity for artifacts is the same. The only rendering difference can be caused by a variation in speed, because temporalslop is always worse at handling fast moving objects.
Anonymous No.725326251 [Report] >>725327369
>>725326059
>I am panning the camera in those images with 3200DPI, and I am moving as quickly as the game lets me. That is a valid representation of what people do when they play games, and the pixels on the screen are shifting.
There's nothing moving around in the scene. That looks like the tutorial area for The Finals, which I've played and know that it doesn't have stuff moving unless you blow something up. From the very start, here's what I said
>>725322061
>video comparison of DLSS4 vs native of a game where things (including the camera) are actually moving around during normal gameplay
If you're going to actually do that, great. If you're not going to do it, then stop replying. Don't do a half assed version and then pretend you did what I said and it was a valid comparison.
Anonymous No.725326262 [Report] >>725326435
>2025 and people are still falling for FOTM multiplayerslop
Anonymous No.725326354 [Report]
>>725291337 (OP)
>dlss ultra performance
Anonymous No.725326435 [Report] >>725326570
>>725326262
>just dont play games dude!
dont you have some BF2142 servers to go fill with bots or something?
Anonymous No.725326570 [Report] >>725326648 >>725326786
>>725326435
No I play singleplayer games because they last forever and are actually fun.
Anonymous No.725326648 [Report]
>>725326570
>and are actually fun
I feel this way when I play multiplayer.
Anonymous No.725326717 [Report]
>>725319831
ps5 used fsr frame gen in wukong
Anonymous No.725326786 [Report] >>725327030
>>725326570
>i dont have friends
we already knew
Anonymous No.725326998 [Report]
>>725291337 (OP)
>DLSS ultra performance
lol
lmao
Anonymous No.725327030 [Report]
>>725326786
What's the point of friends if split screen is dead? You don't actually think retards you play online with are your friends, right? Do zoomies really?
Anonymous No.725327225 [Report] >>725333067
>>725299003
>No recursion
Anonymous No.725327369 [Report] >>725327716
>>725326251
>There's nothing moving around in the scene.
You're either pretending not to understand how this works, or you just don't understand how this works.

When you are looking at a completely static environment and you move forward, the pixels on the screen are moving. It doesn't matter that the environment is static, in terms of rendering you ARE seeing motion because the pixels are shifting. Are you standing still as a traffic sign sways in the wind? Are you sprinting forward as the sign stays completely still? Why do you think there's going to be a difference in terms of how difficult that is to render, or a difference in terms of how the pixels change color? The raster grid or renderer doesn't care about internal game logic, it doesn't care whether the pixel shifting occured because someone threw a grenade on the screen or you just moved closed to a planted tree.

This is why temporal blur doesn't just blur rapidly moving objects, it blurs the whole image when you move or pan the camera. Because all of those are on-screen motion, both of those make the raster grid change pixel colors. The raster grid does not give a shit whether the change in pixel color was introduced by objects being hurled around or you sprinting forward - it doesn't "know" the difference.

When you keep repeating "this does not count because you moving doesn't matter, objects moving are what matters" - you are trying to make a distinction where there is none. It's a motion scenario.
Anonymous No.725327716 [Report] >>725328290 >>725328351
>>725327369
>You're either pretending not to understand how this works, or you just don't understand how this works.
No, I'm just not playing along with your notion that "any pixels moving on the screen at all have the same propensity to cause ghosting or visual artifacts because motion is motion". You're not a software engineer at Nvidia. You don't get to try to explain it like that. You also don't get to group all temporal blur together. You got asked to post proof. You pretended I made an unreasonable demand. You get a real answer (which upset you) about how to post it, and then you proceeded not to post it while pretending you did.

>When you keep repeating "this does not count because you moving doesn't matter, objects moving are what matters" - you are trying to make a distinction where there is none.
Post a technical explanation about how DLSS works specifically. A published paper or a GDC talk will you. You making generalizations will absolute not, and your headcanon is irrelevant.
Anonymous No.725327857 [Report] >>725327937
>>725291337 (OP)
>ultra performance dlss
>looks like dead space 3 (2013)
Anonymous No.725327937 [Report]
>>725327857
It's almost like graphics haven't actually improved in a decade, just gotten worse optimized.
Anonymous No.725328290 [Report] >>725328351
>>725327716

I'll explain it again in case you're slow.

There's a traffic sign in the distance. Scenario 1) you stand still, and the wind animates the sign. The pixels flicker and shift to visualize the sign's movement due to wind displacement, the sign's edges dance. Scenario 2) the sign is rigid, no wind, so you sprint forward - pixels flicker and shift again, the sign's position warps across the pixel grid to accomodate the fact it's now closer to you.

Now are you suggesting that the renderer somehow knows the difference between these two scenarios? That it produces wildly different image quality and artifacts based on.. what knowledge, exactly? There is NO rendering difference in how the pixels are shifting here. It's the same. The renderer doesn't recognize or know whether the shift occurs due to animation or physics or camera movement or sprinting, it is utterly blind to why those shifts occur and any game logic. All it sees are pixels shifting. This is common sense.

I posted a proper comparison and offered to upload uncompressed video as well. You avoided addressing the main points and just started making dumb distinctions where there are none. And now you're asking me to post software engineering credentials or GDC papers, the goalposts have shifted into orbit. I'll respectfully just call you a dumb faggot and move on, I'm wasting my time with you.
Anonymous No.725328351 [Report] >>725328463
>>725328290
>I'll explain it again in case you're slow.
Your rudimentary explanation of perspective is irrelevant. See
>>725327716
>Post a technical explanation about how DLSS works specifically. A published paper or a GDC talk will you. You making generalizations will absolute not, and your headcanon is irrelevant.
Anonymous No.725328463 [Report] >>725328990 >>725329120
>>725328351
Explain to me first how the renderer knows the difference between the two scenarios I described. Why it would produce wildly different image quality and what rendering justification there would be for such a difference. You have to, your assertion that it knows the difference is so ridiculous that I just need to hear it.
Anonymous No.725328626 [Report] >>725328990
>>725325385
always an eternal game of musical chairs with you shitskins isnt it
Anonymous No.725328990 [Report] >>725329259
>>725328463
>Explain to me first how the renderer knows the difference between the two scenarios I described.
You don't get to answer a question with a question to try to sidestep the fact that you're making shit up.. Do you actually know anything about how DLSS works on a technical level? If the answer is no, then you're a fucking retard and you need to shut the fuck up.Your assumption that all motion is motion is based on literally nothing. Does the deep learning algorithm only use 2 frames for every guess? How do objects moving at different speeds versus consistent ones across the screen affect its guesses? What about objects moving in different directions? Do different shapes or patterns do better or worse? You don't know any of these things and it's pretty clear you've never done any sort of image analysis work at all. You're just another redditor pseud who wants to pretend he's smarter than he actually is and that pulling shit out of his ass is "common sense" and automatically wins internet fights.
>You have to, your assertion that it knows the difference is so ridiculous that I just need to hear it.
No, you dumb fucking my cunt. My assertion is that neither of us know how the algorithm works internally so testing it under various scenarios is necessary. Where do you fucking retards get the gall to act like you're scientist lawyer debator pros when you know less than nothing about anything?


>>725328626
You're the only shitskin monkey here. I explicitly said in the first post I replied to that he should post video of things moving on screen in the game AND the camera moving. Now I know English is hard for third worlders who aren't paid much to shill here, but read my posts slowly.
Anonymous No.725328997 [Report]
Yeah that's what I thought. Have a nice one, dumb faggot.
Anonymous No.725329120 [Report]
>>725328463
>Why it would produce wildly different image quality and what rendering justification there would be for such a difference.
It's not a matter of rendering, it's a matter of generation. If you're going to pretend you know enough to have a technical argument, then you need to get the basic terminology right.
Anonymous No.725329259 [Report] >>725329401
>>725328990
So you spent the last near-hour telling me exactly which comparisons are valid and which ones are invalid (conveniently, all the ones that don't align with your argument are invalid) and now you just directly shout SHIT I DON'T KNOW AND YOU DON'T KNOW EITHER. Renders this whole back and forth pointless, doesn't it you dumbass?

Still waiting on your GDC paper on how the renderer knows the difference between the shifting pixels of a static object as you approach it, and the shifting pixels of a moving object as you stand still. Can't wait for CES and your groundbreaking thesis there.
Anonymous No.725329401 [Report] >>725330165
>>725329259
>So you spent the last near-hour telling me exactly which comparisons are valid and which ones are invalid
No, I said one comparison that shows everything is needed, and you refused to post that with "justification" based on ignorance and you making shit up.
>Still waiting on your GDC paper
I'm not the one that made the claim about how it doesn't work. You did. Now post proof or shut the fuck up. No more replies if you don't back up your claim that all motion is interpreted the same.
Anonymous No.725329984 [Report]
>>725291639
Only at 4K, and only at Quality/Balanced.
Anonymous No.725330165 [Report] >>725330360
>>725329401
>I'm not the one that made the claim about how it doesn't work. You did.
Because my claim aligns with the most basic premises of how game rendering works. The raster grid DOES NOT UNDERSTAND in-game logic. It doesn't know the difference between the shifting pixels of a static object as you approach it, and the shifting pixels of a moving object as you stand still. The render or raster grid do NOT interact with the in-game logic of the videogames. That's completely absurd and anyone who'd even suggest this is a total retard.

It's your claim that the raster grid does understand in-game logic. It's your claim that it knows the difference between the shifting of pixels of a static object as you approach it, and the shifting of pixels of a moving object as you stand still. It's your claim that depending on this knowledge, the raster grid will output different levels of image quality simply because it KNOWS. It's your claim that goes directly against the most basic precepts of what rendering is (it is visualization of graphics, not visualization of in-game logic or in-game physics), and therefore it is up to you to provide evidence because your claim is ludicrous.

I already posted proof, anyone in this thread can see the imgsli link. I'll post it again, here:
imgsli.com/NDI3NjA5/0/1

Have a good night, faggot. See you in some other thread where you continue to bury your head in the sand, I'll be there to remind you of this conversation.
Anonymous No.725330360 [Report]
>>725330165
>Because my claim aligns with the most basic premises of how game rendering works.
We're not talking about rendering, we're talking about generation. They're literally not the same thing on any level. You are ignorant.
Anonymous No.725331864 [Report] >>725333234
>>725291337 (OP)
The fact I can run this game on a 580 with an average of 50ish FPS at 1080p max settings is insane. When are other devs with UE5 games going to step it up?
Anonymous No.725333067 [Report]
>>725327225
holy fuck, would you like the dust particles in the air too while youre at it?
Anonymous No.725333234 [Report] >>725333537
>>725331864
>run this game on a 580 with an average of 50ish FPS at 1080p max settings is
That feels wrong, isn't a 5090 10 times faster than a 580?
Anonymous No.725333537 [Report]
>>725333234
idk dude