Justin9mm
Member
You've made this whole thing into something it was not.. You're an idiot
Ubisoft on PSSR: "We're now confident that the image quality with PSSR will always be better than with TAA".
- Thread starter Lunatic_Gamer
- Start date
- News Game Dev Platform
The Pleasure
Gold Member
Ashamam
Member
less a joke more shots fired imho. No wonder you got some replies, I mean its a pretty shit joke when you already have six months of improvements literally right in front of you and you suggest it will be years before any more improvements. Thats not a joke, thats straight up trolling.
Last edited:
Mibu no ookami
Demoted Member® Pro™
Right, I love these guys who get caught with their hand in the cookie jar and try to fall back on, oh it was a joke... It wasn't a joke. Rob Schneider was better and that's saying something.
Justin9mm
Member
The fact that people like you and others take such a small comment not directed at anyone, not really critical of anything specific so serious just goes to show the morons we have on this forum.
As an example look at PSSR, teething issues galore, it's still going to be who knows how long before devs come to grips/take full advantage of it. All I was eluding to is that when FSR4 releases in 2026, it will take some time before we see devs releasing future games using it to it's full potential. But the mere fact I have to explain my very simple comment, which was a slight exaggeration (that's why it was a passing joke) and people getting uppity about it shows how bad social media and the internet has become.
I mean one of the replies said I convinced myself the PS6 is going to be releasing soon and implied that I think there is no need for the PS5 Pro? Can you explain how you get that from my comment? And you're defending these people.. Fucking insane
Last edited:
Justin9mm
Member
What do you mean caught? I wasn't caught out with anything. FSR4 releases in 2026, you're telling me it won't take quite some time for future games/devs to use it to its full potential? PS6 is releasing what maybe 2028, maybe late 2027.. Please tell me how I'm wrong?
Ashamam
Member
Just can't help yourself can you? Every response calling someone an idiot or a moron.
Justin9mm
Member
is that all you've got, no actual defense or proof that I'm worng when I say FSR4 is going to take some time to take full advantage in future games?
The thing is, you can't argue with stupid so I don't know what else to tell you. I'm not going to enage in any more commentary, you can respond to this so you can have the last word because I know how ppl like you love that. Go for it, have your last say, I just won't be responding to it because you're just wasting my time over something so trivial. To attack someone for defending themself from someone attacking them for no reason is a new all time low.
BlackTron
Member
If the new version made it better than DLSS, nobody would complain that that news doesn't belong here. It would obviously be relevant to the thread, because it's about AI.
This kind of topic gatekeeping betrays the actual nuanced conversation some people could have about the actual differences and progresses in tech between the solutions. But those only become relevant to the topic the day PSSR overtakes DLSS.
Edit: in other words some people actually care about the upscaling race and it's deeper to them than who is #1 at this exact time and who therefore gets to comment.
Also imagine a thread new footage of Xbox game. Someone makes video directly comparing scenes to PS version. Is that allowed? Buuuut OP was about specific version of game THERES NO REASON FOR THAT STOP NO WAIT MY FEELINGS (unless imaginary thread is from Xbox One X era, then redact the last part)
Last edited:
PaintTinJr
Member
I agree with most of this, but do you honestly believe for one second that with the wealth of Nvidia and commercial flexibility of reviewers and dev/publishers to follow the money, there will ever be a conclusion that anything is better than DLSS? It already had real issues in comparison in game like R&C where PSSR had superior handling in areas, and DF effectively made it all about stability, while using a xx90 card to render at 3x the frame rate to claim a victory for DLSS despite invalidating any fair comparison of the algorithms.
BlackTron
Member
DF's behavior doesn't really affect the relevancy of DLSS in the conversation. Right now I think any attempt to artificially make DLSS seem better is more professional lapse of integrity than terrible sin because it actually is better. I didn't see that video, but I'm not willing to cede defeat on performance in a PS5 exclusive. An more honest assessment may have admitted PS5 ports won't all run as well on PC as their home console instead of pretending DLSS never loses.
By the time PSSR can catch up to DLSS in most normal real world situations (not cherry picking PS5 games) it will be a different time, could be different guy at DF, a new context of conversation. Don't be bringing up this stuff then, move on eventually.
Mibu no ookami
Demoted Member® Pro™
So it was a joke but you're doubling down?
Use it to its full potential? Games are getting FSR4 now, do you think they're going to wait until 2026 to use it on PC? Beyond that, devs already know very well how to use upscaling at this point. I'm not sure where you got this idea of "full potential" from.
And again, you're misinterpreting a lot of things here. Cerny is saying that they're targeting to get PSSR in line with FSR4 for 2026 titles. Not that they're waiting until 2026 to do anything. The improvements are happening from now and through then as it relates to FSR4 specifically.
I think DF did people a disservice to how they presented this and people's reading comprehension didn't help either. At no point in time did he say we're brining FSR4 to PS5, he said that they've already begun implementing the neural networks from FSR4 into PSSR and their target is to get PSSR in its next evolution to the relative same quality of input to output as FSR4 for 2026 titles.
FalsettoVibe
Member
You make good points. I agree that comparing different solutions like DLSS, FSR, and PSSR is valuable in a specific context. But the original post and the initial discussion are specifically about the PS5's TAA vs. PSSR. Adding DLSS into this specific conversation kinda creates a distraction from the topic at hand. Especially when people are being tangential with their posts and adding nothing valuable to the discussion by saying things like:
Now that I think more about it, I don't think that all conversation about DLSS in this thread should be 100% off limits. But you have to agree that there are a certain group that want to turn this into an argument about DLSS and PSSR and which is better. (Obviously DLSS at this point).
Also, I don't think it's "gatekeeping" to ask that the conversation remain on the topic that was originally posted. Especially considering the history of how this comparison has gone on this forum in the past. If someone would like to start a thread about the overall upscaling race that includes all platforms, then that would be a very good topic of conversation and comparisons would fit fine there. Doing it here just deviates from the topic entirely.
BlackTron
Member
Sure there's always a "certain group". Low hanging war bait will always find a way...we won't change that by saying how much closer PSSR is getting to *redacted*
mckmas8808
Mckmaster uses MasterCard to buy Slave drives
That image looks incredible!
PaintTinJr
Member
DF was just one example of a reviewer inside the wider internet with reams of paid influences with Nvidia cards, so my main point - which you actually neglected to answer - and I'll ask the question again : do you honestly believe for one second that anything but DLSS will ever be held up as the best - by the wider internet - when they are better than DLSS in every way - considering there are already things that are worse in DLSS than in PSSR?
Bojji
Member
"Paid Nvidia shills" talking about AMD FSR 4:
Without DLSS4 launch in february AMD would have the best ML reconstruction right now. DLSS2/3 was the king 2020-2025.
PaintTinJr
Member
You know that is a strawman and I don't think you understand how quoting works. If you are going to quote, add whom is being quoted or are you trying to imply I said those words, as an intentional misquote?"Paid Nvidia shills" talking about AMD FSR 4:
Without DLSS4 launch in february AMD would have the best ML reconstruction right now. DLSS2/3 was the king 2020-2025.
Bojji
Member
This is what some people think here, I wasn't directly quoting you (but this is what you suggested in your post).
BlackTron
Member
My opinion is that anything legitimately better than DLSS will become known as better. I'm not willing to say this won't happen because there are DLSS shills today. The fact that DLSS is better anyway makes it a giant nothingburger.
By the time PSSR might be as good or better, the conversation will have taken on a new context and all this "evidence" of bias will be meaningless old news.
PaintTinJr
Member
In what ways specifically, say if hypothetically we took a 5070ti and downclocked until the SMs gave 600 4bit TOPs and we turn off FG/MFG and run it on a game that has a PS5 Pro version with PSSR support, what ways is the holistic use of CNNs in DLSS4 doing a better job than PSSR?
I ask this because you've just stated - an unsubstantiated opinion, in the context of your comment- as a fact, based on ...? Reading the comment it felt like it was based on aggregate internet opinion, rather than based in discussing the algorithms. Which is exactly why I asked the question.
BlackTron
Member
So a back and forth of what specifically might hypothetically happen if something were changed.
No thanks I'm good
PaintTinJr
Member
What??
It was an easy question if the answer was easy to make your point. Instead your lack of answer plays exactly to the point I am making.
Even now while "DLSS is better anyway" we have no point of reference what that means to draw such a conclusion fairly, so how in the future can any of the other techniques by an point of reference surpass DLSS? Clearly they can't because the whole coverage is entirely infected by influencers using +£1500 GPU gifts to influence right now and unless you tell me that's going to change by some miracle, how is the opinion of DLSS versus the rest going to change fairly?
Gaiff
SBI’s Resident Gaslighter
God could come down on Earth and say DLSS is better and you’d argue. Why pretend that you’re interested in what he has to say rather than admit you just want to fight him over what he said?
PaintTinJr
Member
Not at all, I actually understand a lot of the technologies and maths involved, so if he can actually substantiate his point, then that's far more interesting, than just the accepting the PR fluff about how every DLSS is the best thing since sliced bread and everything else is now trash, but without any real technology or knowledge to back that up or any reference metrics and just a huge cloud of internet PR opinion.
yamaci17
Gold Member
all in all I'm happy dlss pushed for much higher quality temporal reconstruction. people act like it created its own problem while disregarding games like uncharted 4 last of us part 2 and rdr 2 look incredibly blurry on PS4 due to TAA. today dlss 4 makes 1080p games look as sharp as they looked back in 2010s (almost)
without dlss, there's no pssr or fsr 4 anyways. look at ps5 pro and its performance mode in ac shadows. look amazingly blurry due to TAAU. PSSR will fix it massively and people then should see the value in it more
i don't think pssr or fsr 4 would exist without dlss being a thing. so I think DLSS always has a place in PSSR or FSR 4 related discussions
Tqaulity
Member
People keep saying this but I disagree. The game has a rough and noisy look to the entire frame that really doesn't look good at all. It's evident indoors as well as outdoors when playing on Quality mode. Not all of it is attributed to PSSR but just saying that the IQ is really poor in this game throughout.
HeWhoWalks
Gold Member
It has a noisy look on PC as well, just not as pronounced. Just how their game meshes with the engine. Rough spots, sure, but a good looking game nonetheless.
Last edited:
PaintTinJr
Member
I haven't played the latter 2 games and thought UC4 was very-mid if not rubbish - I struggled to enjoy it after the Scotland stages that showed no one from ND's art team had ever been to Scotland - however picture quality on base PS4 on the reference ZD9 looked sharp and clean when I played, so you must be letting YT clips fool you - or have a really shitty screen for PS4 play throughs - if you are using the words blurry in a real viewing experience and not doing some 800% zoom pixel peep to mistakenly draw that conclusion.
Panajev2001a
GAF's Pleasant Genius
You are stretching it really…
SKYF@ll
Member
High-quality upscaling technology from low resolution has been around since the PS3 era (used in PlayStation and Bravia). *not for game
SONY's excellent upscaling algorithms will be used in games through collaboration with AMD.
As ML upscaling continues to evolve, in the future it will be difficult for our eyes to tell the difference between DLSS, FSR4 and PSSR.
These are images upscaled from DVD (480i).
killatopak
Member
I’m just glad we have this tech early. Imagine if we had to start from scratch when the PS6 releases.
I don't think you need any real understanding of the technologies or maths involved to do an image quality comparison. Do PS5 Pro users need to get PhD in image reconstruction before they can appreciate PSSR? Do you need a PhD in physics to appreciate the light propagation in a game using ray tracing?
Last edited:
PaintTinJr
Member
Years on we still have the false quote of best version of multiplatform games on 360, without exponent fog, broken sRGB gamma correction, no HDR, less geometry and desynched physics simulations because of screen tearing, etc all because those having the strongest opinions for the richest shilling company had "any real understanding".
Liking TV shop mode of oversaturated whites and crushed blacks on most 360 in the first few years - as was preferred by the exact same crowed - doesn't make human perception right, or better than technologically or scientifically correct observations. So, yeah uninformed opinion carry far less weight in conversations about technological things, as is the evidence based guidance for users commenting on this forum, last time I checked.
1.) Preferences can't be "right" or "wrong". If someone prefers an oversaturated image over one with natural colours, there is nothing more to be said about that. Equally if someone isn't bothered by screen tearing, there is no argument you can make to "prove" that they should be bothered.
2.) My argument is not that it's fine to be uninformed, but rather that being "informed" in a given context, does not necessarily depend on any in depth technical knowledge. If I try different modes on my TV and learn what it looks like when properly calibrated, I can form an informed preference about what settings to use. To do this, do I need to go to university to learn about the mathematical representation of colours? No. Equally, being informed about DLSS or PSSR, can mean observing the image quality in various games and at various quality settings.
(It's not essential to my point but I would even make the stronger argument that it's not possible to be truly informed about how PSSR or DLSS really work, since they are essentially black box solutions).
PaintTinJr
Member
How does preference factor into this?
This sub-point of the discussion is about the internet accepted quasi-truth that DLSS is the best - in reproducing the developer's intended visuals at a technical level (not the user's informed or uninformed choice).
Signal processing is a very technical subject so it really shouldn't be difficult to make a technical argument for DLSS if the evidence supports such an annihilation of PSSR superiority, no?
1.) Preference factors in when assessing whether one solution is "closer" than another to the ground truth image, or involves more acceptable compromises.
2.) My claim is simply that such an assessment can be made without knowing anything about the technical details of how the upscaling algorithm works. (Not that we have access to such details anyway).
Mibu no ookami
Demoted Member® Pro™
High-quality upscaling technology from low resolution has been around since the PS3 era (used in PlayStation and Bravia). *not for game
SONY's excellent upscaling algorithms will be used in games through collaboration with AMD.
As ML upscaling continues to evolve, in the future it will be difficult for our eyes to tell the difference between DLSS, FSR4 and PSSR.
These are images upscaled from DVD (480i).
I agree in terms of upscaling with some caveats.
The ability to upscale at lower and lower resolutions will be a differentiator as well as how heavy a load each upscaler takes on performance.
Then I think you have to account for other elements that are not just upscaling like frame gen, frame extrapolation e.t.c.
PaintTinJr
Member
Your two statements contradict one another. You can't make a proper assessment of whether one solution is closer than another to the ground truth image without having expert knowledge, because image balance is a tradeoff and where a random influencer might focus on one element, an expert in the field will look at things in balance.
We've already seen how flawed your premise is with the historical false claims about Xbox 360 multiplats, and just to show that isn't an outlier, in pre-PSSR release we had so-called technical comparisons of FSR3 to DLSS3 in which FSR3 holistically handles some scenes better, but the influencer quasi-truth of the internet still said DLSS was closer, so history again supports my prediction that there are no circumstances where the wider internet will be influenced to prefer or accept anything other than DLSS as the best.
Yes, but there is no objectively "correct" trade-off. That's my entire point. If solution A extracts more detail, at the expense of temporal stability, and solution B is more temporally stable, but with less clarity, then some people may prefer solution A, and some people may prefer solution B. There is no fact of the matter as to which person is "right", since they are expressing a preference for a particular element of the reconstruction that is closer to the ground truth.
And moreover expressing such a preference does not require any special knowledge, just like you can prefer a particular colour setting on your TV, without being an expert in colour theory.
Last edited:
Corporal.Hicks
Member
TAA can look soft on a PC monitor, especially if you don't add a sharpening mask. I remember seeing PS4 screenshots of Uncharted 4 on my PC and thinking the image looked a little soft because of the TAA, but when I finally played the game on my PS4Pro and Sony FullHD TV, I was blown away with the image quality in this game. The image was clean (no aliasing), yet every edge was razor sharp and I could easily see fine details on the textures. Sony's Bravia TVs have amazing image processing, so this game may not look as good on all TVs, especially 4K TV (upscaling will ruin the image), but on my tv I was blown away and I would never say the image was blurry, not even a little bit.
yamaci17
Gold Member
I don't really care about sharpening or artificial post processing effects that some TVs have. I don't like them.
nvidia and reshade has high quality image processing and sharpening filters that brings out "fine" details on textures. I don't use them either.
also uncharted 4 runs at 1440p on ps4 pro (downsampling on a 1080p screen) which will greatly reduce TAA blur by itself already
I played last of us part 2 on my 1080p screen with a ps4 and ps5. ps4 runs the game at native 1080p/30 FPS and it looks blurry. ps5 runs it at 1440p/60 FPS. it does not look blurry. it is not that difficult to understand. obviously same for uncharted 4 as it runs at 1440p on ps4 pro and 1080p on ps4
Last edited:
PaintTinJr
Member
There doesn't need to be an objectively correct trade-off in the expert field because everyone is equally observant and over time people coalesce around the same ideas, objectives and means to achieve them.
Like we are seeing in new miniLED TVs Hisense, Sony, Samsung, etc are all moving to additive(RGB backlighting) to supplement the common solution of subtractive colour (filter) techniques to improve picture quality, and regardless of TV reviews the R&D teams themselves will have a ground truth on how they measured up to their peer's solutions and that will be true of upscaling and the DLSS team will have know where they've been deficient to other techniques over time, even if consumers blindly back their DLSS at any given version.
I never said the lack of an objectively correct trade-off was an impediment to progress. My claim is simply that the preference of a technology expert is no more valid than the preference of an ordinary user when determining the impact of an upscaling technology to them. So a user can use DLSS and assess the trade-offs it offers without needing to get that opinion validated by such an expert.
And in fact, such technologies are not designed for a select few upscaling connoisseurs, who undergo years of training to truly appreciate the majesty of the algorithm. They are designed for ordinary people who want better performance in their games with less obvious compromises (to them).
Last edited:
Corporal.Hicks
Member
I played Uncharted 4 at standard 1080p, because 1440p doesnt scale evenly to 1080p display and I dont like moire artefacts and blured fine details. The picture on my TV was sharp without looking oversharpened.
Even DLSS relay on sharpening mask because temporal AA reduce edge contrast, and without edge contrast the human eye perceives a blurred picture. Most developers use sharpening filters in their TAA / DLAA / FSR games, but they either use too much or too little sharpening mask, that's why I prefer to turn off ingame sharpening settings and use my own.
DLSS3
Defualt ingame sharpening
Without sharpening mask. Sharpness look very similar to Uncharted 4 TAA screenshots
Sharpening with my own reshade settings (sony bravia TV's have even better sharpening mask)
DLSS4 (preset K)
Default ingame sharpening
Without sharpening mask
Sharpening with my own reshade settings
DLSS3 without shapening mask look equally blurry as TAA. DLSS4 look much sharper, but still I like to add a small amount of sharpening mask, just to get that razor sharp look (MSAA / SMAA like).
You say that you don't use sharpening? I saw your DLSS comparison screenshots and immediately noticed contrast sharpening mask. Like I said, most games use sharperning filters, and it seems you didnt noticed it.
When high contrast details starts to look ebmossed I know sharpening mask was used.
What's more ray reconstruction adds even more sharpening (and what's wore DNR on top of that).
Without RR
Ray Reconstruction
Yet you dont mind using ray reconstruction despite excessive sharpening mask
And that's your option about HL2 RTX. Your screenshots looks like a painting because of that DNR and sharpening filter, but you were happy with the image quality. Dude, Uncharted 4 on PS4 looks much better than HL2 RTX on your PC,
Last edited:
yamaci17
Gold Member
you rely on painterly sharpening on a basic rasterized game. I'm fine with sharpening look in a PATH TRACED game (in a demo where you run path tracing for the sake of path tracing). they're not the same thing, nor will they ever be. you think you're so smart? think again
for avowed, if the game uses sharpening by itself, I have no idea. I don't like them. I don't use them. it doesn't mean I'm actively looking out if a game has sharpening or not. if it is on by default and the game does not have a setting for it, I'm not going to care, since I'd assume it is how devs intended the game to look.
there's a big difference there but I don't think you would understand
and lastly, I DON'T LIKE sharpening itself. I LIKE what ray reconstruction is doing. I don't like that it also uses sharpening alongside with it. and that means I can be fine with sharpened looked, if the game overall looks great to me. that would even be true for uncharted . it's just that I just won't go out of my way to use that. because I also can enjoy a game if it is blurry. I can see past through sharpening or blurriness and still be fine with the image quality. still not understanding? then it is not my problem
if I only saw uncharted 4 with sharpening, I would also say "I'm quite pleased with uncharted 4 graphics". I wouldn't say "oh this looks sharpened and bad". no. if it was an in game option? I'd turn off the sharpening. then say "hey this looks worse to me with sharpening". get it?
I'm being pleased by PATH TRACING graphics, not "sharpened look" of the game. get it?
"Yet you dont mind using ray reconstruction despite excessive sharpening mask"
like do you have any idea how dumb this sounds? the game is made around ray reconstruction. what am I to do? it has nothing to do with uncharted 4's situation
if uncharted 4 was a noisy mess without TAA but had a sharpened look with TAA, I'd also NOT MIND using the "sharpened TAA" there also. did I say anything otherwise? no. I just told you that I don't like using sharpening. get it? if it is the only option, I will be fine with it.
being fine with it doesn't mean that I will go and enable sharpening
get it? I hope so
black myth wukong had high amount of sharpening at launch without any setting to disable it. I was fine with its image quality at 1080p DLSS, I wasn't bothered. but a couple months later they added in game toggle to turn it off. guess what, I turned it off. get it? I was still fine with 1080p DLSS image quality. because sharpening had nothing to do with what I liked about it
did I say anywhere that I wouldn't be happy with Uncharted 4's image quality with sharpening? not liking it, or not using it doesn't mean "I wouldn't be happy with image quality"
so, no
we're done here
Last edited:
Corporal.Hicks
Member
The sharpening mask was applied to every single DLSS4 screenshot / game in your comparisons. You think you are not using sharpening masks, but almost every single TAA / DLSS game uses them. You are just not aware of it.
You said that Uncharted 4 TAA looks incredibly blurry, and you suggested that DLSS4 would look much better, so here's my comparison using the latest DLSS4. People can see if that's really the case.
Uncharted 4 TAA (TAA uses a small amount of sharpening mask and the game does not allow to disable it)
DLAA (transformer K preset)
DLAA with sharpening mask
Thanks for including PNG screenshot.
Here's my DLSS quality (transformer J preset) screenshot for comparison
PS5Pro
BTW. Star Wars Jedi Survivor is unplayable on PC with RT. The game stutters like crazy. RT mode is playable on the PS5Pro without stutters?
Last edited:
yamaci17
Gold Member
I don't like TV, reshade and NVIDIA overlay sharpening. it is not something I can control (what I feel).
whenever I actively apply sharpening through reshade (lumasharpen, fidelityfx cas, and more), nvidia app (sharpen, sharpen+, details), nvidia control panel (nvidia image scaling), the image looks wrong to me. when I tried sharpening on a couple different TVs, I always hated how it looks even in games, movies and regular content. I didn't try it on your particular TV model and I likely never will, so I will never know. I don't really care. I just don't like external sharpener filters. that much is clear
the reason for this is because it is external. it means I see what image looks like before applying that external sharpening. so when I apply it, I immediately feel like something is wrong. if a game ships with sharpening enabled by default, I either don't notice it or just accept it and move on (black myth wukong).
believe it or not, I had to stop playing last of us part 2 remastered on PS5 because it has a forced sharpening filter. I've played that game at blurry 1080p for at least 150 hours. I'm so used to its natural, non sharpened look, I just couldn't get used to how it looks on its PS5 remaster. now I look forward to its PC release where hopefully I can disable its sharpening. by the way, last of us part 2's base ps4 pro version on ps5 does not have the sharpening filter on ps5. so I played that and moved on
so being "unaware" is okay. it doesn't mean I'm fine with it. it doesn't mean I can enable an external sharpener and say "hey this looks better and great". over the years I've tried countless sharpening filters to fix "TAA" at 1080p. nothing worked for blurring in motion. I eventually gave up. up until DLSS 4
anyways, your uncharted 4 comparisons are made with character just standing still which gives TAA time to reconstruct to high quality non blurry image. TAA in RDR 2 looks reasonably sharp and clear while standing still, but becomes blurry in motion,
this is exactly what happened with uncharted 4 and tlou 2 on base PS4 when I played them. DLSS 4 however is sharper and clearer in MOTION. I can prove it to you if you want (or do it yourself, if you think you can pull it off yourself. and of course do it at 1080p)
i had to be elaborate because it seems like there's a miscommunication and I feel like you really don't understand what I am trying to say which is not your fault entirely but sharpening is a complex topic for me as I've explained above. for me to be fine with a sharpened look, I should not see it non sharpened, if that makes sense
(half life 2 rtx is a specific situation where non sharpened alternative is a noisy mess that looks horrible)
Last edited:
SKYF@ll
Member
Thank you for the screenshot from the PC version.
I played for about 5 hours on PS5 Pro (Performance Mode) and it seemed to run at a stable 60fps.
Apart from a bit of shimmering on the bright outdoor vegetation in Koboh, image quality was also satisfactory.
Similar threads
- Locked