Support NeoGAF

[Three]

The DLSS3 video does say "4K120fps and beyond" compared to AMD where it states that frame generation has problems. Maybe that?

Yeah that's the one.
"Exclusive First Look - 4K 120FPS and Beyond"
"Promising Image Quality, But There Are Problems - DF First Look"

No guesses about which one they got an "exclusive first look" on.

Also, I think people should form their own opinions instead of listening to DF's conclusion. Watch their analysis, but the final answer belongs to you. From what I've seen in this video, the PSSR issues are relatively minor and not too distracting, so I take it as a huge win even though Tom's opinion is more mixed.

It's like, thanks for highlighting some issue for it to get fixed but the tone is always very different even when nvidia has similar issues in their implementation. With nvidia the problems take a back seat and just get mentioned in the vid with an overall hyperbolic vibe of how great it is. They're not emphasised the same. With others it's always "there are issues", "there are problems" front and centre.

 

[Topher]

Look, the pro has image quality issue, but it's more preferable here than checkerboarding on the ps5.

Even Tom highlights the grass but which one look better here?

The Pro is clearly a significant improvement. This is what Pro *should* be compared to.

 

[Three]

I have played a ton of games with DLSS, and almost a dozen now with PSSR. (Jedi Survivor, Avatar, GT7, Stellar Blade, Spider-Man 2, Last of Us Part 2, Ratchet & Clank, Veilguard, Ragnarok, Dragon's Dogma 2) DLSS never had the flickering and stability issues to the extent that PSSR has with RT. GT7 in RT+PSSR mode is quite bad, but almost every single game with PSSR and RT present flickering and temporal instability far in excess of what DLSS ever did. The flickering and instability gets diminished with less noise and/or higher resolutions, but that requires 1440p or higher, and with anything lower becoming a mess. DLSS has none of the instability and flickering is far less severe, even with lower resolutions.

DLSS did still suffer from more ghosting before 3.7, but that was not nearly as distracting as the visual issues we are seeing with PSSR. It's not just my, or DF's opinion on this matter, just look at the general reaction to both. DLSS 1.0 was widely considering the laughingstock of upscaling, but DLSS 2 (and especially since 2.5) onwards has been extremely well-received, whereas PSSR has not been, because the visual distractions are extremely obvious and doesn't require 4x zoom to notice.

On what hardware and what settings? DLSS performance or balanced on lower end cards? You can look up the same distracting issues on DLSS. The same shimmering. 3.5 onwards had Ray Reconstruction for better temporal stability but had trails and ghosting artifacts instead. Again not a major issue to concentrate on and mention in a headline though.

 

[Bojji]

On what hardware and what settings? DLSS performance or balanced on lower end cards? You can look up the same distracting issues on DLSS. The same shimmering. 3.5 onwards had Ray Reconstruction for better temporal stability but had trails and ghosting artifacts instead. Again not a major issue to concentrate on and mention in a headline though.

You can't tell the difference between denoiser problem and DLSS SR problem.

He changed denoiser and problem was solved (but other problem appeared that was later fixed by developers)

PSSR doesn't change game denoiser, it just breaks with RT lighting and some forms of SSAO - seen in many different games/engines. DLSS SR never had problems like that.

You are trying hard.

 

[King Dazzar]

I've only played 30 minutes of DD2 since the Pro update. But in that time the Pro looked far better than the base PS5 did. Much sharper and running much better. I'm not saying there aren't issues, but where as Janky Jedi's and SW Outlaws issues are very pronounced, to the point where by I've shelved them. DD2 looked far better than base PS5 with nothing to stop me enjoying my limited time with it.

Or to put in another way - I'd rather have the Pro patched version as it currently is. Than go back to how it looked on the base PS5.

 

[Three]

You can't tell the difference between denoiser problem and DLSS SR problem.

He changed denoiser and problem was solved (but other problem appeared that was later fixed by developers)

What exactly are you highlighting here with that image. It says "this kind of shimmering is a temporal artifact introduced by DLSS". The OP/user even says it happens when he is using DLSS. Stop your nonsense deflection of the issue.

 

[James Sawyer Ford]

And the people who say baked lighting can look just as good are lying. Especially large sandbox or open world games. My case in point:



Does anyone think the lighting of this in motion looks anywhere close to the games with RTGI.

The game overall looks far better, yes

 

[Bojji]

What exactly are you highlighting here with that image. It says "this kind of shimmering is a temporal artifact introduced by DLSS" he even says it happens when he is using DLSS. Stop your nonsense deflection of the issue.

He changed denoiser to RR and problem was gone. I never said that the answer was 100% correct.

You clearly have no idea how to tell the difference between denoiser problems and upscaling problems, have you played many games with RT lighting?

 

[Zathalus]

On what hardware and what settings? DLSS performance or balanced on lower end cards? You can look up the same distracting issues on DLSS. The same shimmering. 3.5 onwards had Ray Reconstruction for better temporal stability but had trails and ghosting artifacts instead. Again not a major issue to concentrate on and mention in a headline though.

That is DLSS quality at 1080p (720p internal). It’s basically one quarter the output resolution and a quarter to a third of the rendering resolution compared to the titles I mentioned. On top of that the issue being highlighted is less severe than what you would see in Silent Hill 2, GT7, or Jedi Survivor. You don’t see temporal instability in the whole image either.

At no point have I ever switched from FSR/TSR to DLSS and thought the image looked worse. That has happened for a number of games on the Pro now, which is quite disappointing.

 

[Kangx]

Even with the PSSR issues with this game, the PS5 Pro version runs smoother and looks better than the PS5 version.
It's interesting to compare the clarity from CBR 4K (4.14M pixels) and PSSR 1440p to 4K (3.68M pixels).
If DF was going to highlight and point out the shortcomings of the PSSR, it should have also refocused on its strengths.
When playing games while sitting on the couch, the image quality on the PS5 OG is blurry and the frame rate is more stuttering.

Yup. See my post also. On the grass issue which Tom seems to fixated on, PSSR seems to blur the grass, Checkerboarding seems to make it fuzzier.

Overall, When playing normally, the Pro looks much sharper and better especially in motion which is a trait of PSSR.

Not agree with Tom on this game about putting back ps5 setting for some people. The trade off here is clearly worth it compare to checkerboarding.

Why would I want to get back to an worst image quality and sharpness especially in motion and also run at worst frame rate?

 

[Feel Like I'm On 42]

One would think it’s way worse than fsr2 shite we used to get. But it’s still way better

Overall yeah but when specifically talking about shimmer? I'd argue it's just as bad ...all the evidence that's needed is Avatar, AW2 (quality mode is bad), Sw Outlaws and SH2 ....this is really disappointing

 

[Three]

He changed denoiser to RR and problem was gone. I never said that the answer was 100% correct.

You clearly have no idea how to tell the difference between denoiser problems and upscaling problems, have you played many games with RT lighting?

You clearly have no idea what you're talking about. RR is a DLSS tuned "denoiser" created because DLSS introduced these artifacts.

"Basically, the game does its ray sampling per-pixel of your internal render resolution, pre-DLSS.

This shimmering (or, alternatively, ghosting when using ray reconstruction) occurs when there aren't enough actual samples being calculated, so DLSS/RR can't make accurate predictions about the image."

Hence DLSS introduces this shimmering as correctly stated and why he sees it when he enables DLSS.

It wasn't even available until DLSS 3.5 but the shimmering doesn't happen without DLSS 1-3. It's DLSS not playing nice with the game pre-RR. I don’t understand why you're so desperate to distance this issue from DLSS. The issue Zathalus is seeing in GT7 in tunnels is the exact same issue being pinned on PSSR yet you don't have an issue with this distinction or lack thereof.

That is DLSS quality at 1080p (720p internal). It’s basically one quarter the output resolution and a quarter to a third of the rendering resolution compared to the titles I mentioned.

Exactly and he's playing at that res because of the hardware hence why I said DLSS has similar issues but it's mostly mitigated by higher end hardware. Same is true for RT settings/hardware on AMD cards. The RT settings are lower to begin with. performance is mostly better with PSSR/DLSS with comparable image quality or better though. DD2 is way better on PS5 Pro than the PS5 base yet there is this emphasis on some minor artifact.

 

[Bojji]

You clearly have no idea what you're talking about. RR is a DLSS tuned "denoiser" created because DLSS introduced these artifacts.

"Basically, the game does its ray sampling per-pixel of your internal render resolution, pre-DLSS.

This shimmering (or, alternatively, ghosting when using ray reconstruction) occurs when there aren't enough actual samples being calculated, so DLSS/RR can't make accurate predictions about the image."

Hence DLSS introduces this shimmering as correctly stated and why he sees it when he enables DLSS.

It wasn't even available until DLSS 3.5 but the shimmering doesn't happen without DLSS 1-3. It's DLSS not playing nice with the game pre-RR. I don’t understand why you're so desperate to distance this issue from DLSS. The issue Zathalus is seeing in GT7 in tunnels is the exact same issue being pinned on PSSR yet you don't have an issue with this distinction or lack thereof.


Exactly and he's playing at that res because of the hardware hence why I said DLSS has similar issues but it's mostly mitigated by higher end hardware. Same is true for RT settings/hardware on AMD cards. The RT settings are lower to begin with. performance is mostly better with PSSR/DLSS with comparable image quality or better though. DD2 is way better on PS5 Pro than the PS5 base yet there is this emphasis on some minor artifact.

If lower resolution causes this then why you say it's DLSS problem? It would look the same with in game TAA upscaling or FSR - how it's DLSS related?

While switching from TSR/FSR to PSSR causes issues in some pro games.

 

[Zathalus]

Exactly and he's playing at that res because of the hardware hence why I said DLSS has similar issues but it's mostly mitigated by higher end hardware. Same is true for RT settings/hardware on AMD cards. The RT settings are lower to begin with. performance is mostly better with PSSR/DLSS with comparable image quality or better though. DD2 is way better on PS5 Pro than the PS5 base yet there is this emphasis on some minor artifact.

Of course lower resolutions have more issues. That’s to be expected when you have less pixel data to work with. But comparing like for like in terms of resolutions and it’s clear that PSSR fares noticeable worse compared to DLSS and can have significant issues with RT. Dragons Dogma 2 I’d say is certainly better vs checkerboard but it presents issues that DLSS simply doesn’t have, even at lower resolutions. Other games fare worse, especially some of the examples I mentioned.

 

[winjer]

The issue with RT and AO might be an improper setting of the resolution of each effect.
Similar to LoDs and Mipamaps, these effects depend on the base resolution of the game.
So when it's being used with a temporal upscaler, the dev has to create an offset to these effects to render at the proper resolution.

The issue with vegetation might be an issue with the game not generating motion vectors for the grass or vegetation.
Having a reactive mask, or a pass to output velocity for these objects, could solve some issues.
It also doesn't help that the performance mode renders at 720p. This resolution is very low for a temporal upscaler to deal with.

 

[Three]

If lower resolution causes this then why you say it's DLSS problem? It would look the same with in game TAA upscaling or FSR - how it's DLSS related?

While switching from TSR/FSR to PSSR causes issues in some pro games.

I've mention this to you several times in the other thread. It's DLSS/PSSR hallucinating from the lower res image that cause this particular shimmer. it doesn't have enough information and introduces that temporal shimmer . It would not be the same shimmering without the upscale. It would just be a lower resolution image without PSSR/DLSS. And you still didn't answer the question in the other thread if this shimmer is not DLSS related to you how is it "PSSR related" in GT7 with the exact same thing Zathalus is experiencing?

It wouldn't be, right? So why do you have a problem with my later post in particular and not Zathalus and PSSR?

 

[Three]

Of course lower resolutions have more issues. That’s to be expected when you have less pixel data to work with. But comparing like for like in terms of resolutions and it’s clear that PSSR fares noticeable worse compared to DLSS and can have significant issues with RT. Dragons Dogma 2 I’d say is certainly better vs checkerboard but it presents issues that DLSS simply doesn’t have, even at lower resolutions. Other games fare worse, especially some of the examples I mentioned.

has anybody compared like for like here to even draw any conclusion like this?
RT is better on nvidia cards in general and often in these comparisons it's against high end cards at max settings or RR enabled with ghosting trailing. Maybe set low settings and similar upscale disable RR and see what happens.

 

[Bojji]

I've mention this to you several times in the other thread. It's DLSS/PSSR hallucinating from the lower res image that cause this particular shimmer. it doesn't have enough information and introduces that temporal shimmer . It would not be the same shimmering without the upscale. It would just be a lower resolution image without PSSR/DLSS. And you still didn't answer the question in the other thread if this shimmer is not DLSS related to you how is it "PSSR related" in GT7 with the exact same thing Zathalus is experiencing?

It wouldn't be, right? So why do you have a problem with my later post in particular and not Zathalus and PSSR?

It wouldn't be, FSR/Xess also reconstructs the image.

You can see what is caused by PSSR by seeing games when developers switched from FSR/TSR to PSSR - issues appeared. What more proof do you need?

CP example you show is caused by denoiser first and foremost, this game has full path traced lighting so denoiser is VERY important.

 

[Three]

It wouldn't be, FSR/Xess also reconstructs the image.

You can see what is caused by PSSR by seeing games when developers switched from FSR/TSR to PSSR - issues appeared. What more proof do you need?

And yet somebody enabling DLSS and having this issue only when enabling it isn't enough proof? Go figure.

CP example you show is caused by denoiser first and foremost, this game has full path traced lighting so denoiser is VERY important.

So GT7 is the denoiser too when RT/PSSR is enabled, yes? you're not making sense as usual.

 

[Kangx]

Hey, what do you know. DLSS also seems to exibit the same issue as PSSR where it blur small grass here in this scene. Go to 20:48.

Interesting, FSR is the best here.

Edit:

 

[Fake]

PS6 will fix that I hope.

 

[Kangx]

PS6 will fix that I hope.

It should. Higher setting plus higher resolution given significant gpu and especially cpu upgrades that help with RT.

 

[rofif]

Overall yeah but when specifically talking about shimmer? I'd argue it's just as bad ...all the evidence that's needed is Avatar, AW2 (quality mode is bad), Sw Outlaws and SH2 ....this is really disappointing

Aw2 and sh2 i agree. Both suck on pro. Aw2 is improved partially but sh2 not really

 

[Bojji]

And yet somebody enabling DLSS and having this issue only when enabling it isn't enough proof? Go figure.

So GT7 is the denoiser too when RT/PSSR is enabled, yes? you're not making sense as usual.

You use reddit post as proof for something? Author clearly don't know much about this stuff.

GT7 shows the same issues as other games with PSSR (but it's only in some specific situations?) - clearly something doesn't work as should between denoiser and pssr. While you don't see the same issues with other reconstruction techniques.

 

[Three]

You use reddit post as proof for something? Author clearly don't know much about this stuff.

Now you're suggesting it simply doesn't exist because it's reddit posts. Great, it's easy enough to test is it not? Is that why you disappeared for a while?

GT7 shows the same issues as other games with PSSR (but it's only in some specific situations?) - clearly something doesn't work as should between denoiser and pssr. While you don't see the same issues with other reconstruction techniques.

Yes its an issue with a noisy image and an upscaler hallucinating trying to upscale it. just as it showed in cyberpunk with DLSS it is PSSR upscaling a noisy image (low light/low samples and hallucinating) . You can test this yourself. Once again, they're both the exact same issue. Stop pretending they're not and making handwavy remarks about "specific situations".

 

[Feel Like I'm On 42]

It should. Higher setting plus higher resolution given significant gpu and especially cpu upgrades that help with RT.

Sony better come up with a solution to this for the Pro or else there will be a lot of angry customers ...the fact that they didn't learn from Nvidia and the way DLSS is kept OPTIONAL is such a big fuck up, now we have 5 or 6 games that look worse than before and we're left waiting with our thumbs up our asses hoping the devs will fix or add back in the option of the original upscaling solution ...

Unbelievable that Sony couldn't foresee this and before someone says "its not in Sonys hands" that's just plain BS, they easily could've prevented this situation by mandating devs keep the original modes just as has been done on PC all the way back in 2018 when dlss came on the scene

 

[Bojji]

Now you're suggesting it simply doesn't exist because it's reddit posts. Great, it's easy enough to test is it not? Is that why you disappeared for a while?

Yes its an issue with a noisy image and an upscaler hallucinating trying to upscale it. just as it showed in cyberpunk with DLSS it is PSSR upscaling a noisy image (low light/low samples and hallucinating) . You can test this yourself. Once again, they're both the exact same issue. Stop pretending they're not and making handwavy remarks about "specific situations".

Why don't you test it yourself and show proof to your words?

TAA/TSR/TAAU/FSR - no issue
PSSR - issue

So what is the problem here Mr. Detective? Of course it can't be PSSR, hahaha.

 

[Radical_3d]

Hey, what do you know. DLSS also seems to exibit the same issue as PSSR where it blur small grass here in this scene. Go to 20:48.

Interesting, FSR is the best here.

Edit:

“Big FPS Boosts But DLSS Issues Need Fixing”​

 

[Lagspike_exe]

They are massively overemphasizing PSSR's deficiencies. Yes, it has flicker issue, but overall the IQ is much better than FSR/CB in motion. Every game i tried looks much better in motion vs. PS5 base.

 

[Three]

Why don't you test it yourself and show proof to your words?

TAA/TSR/TAAU/FSR - no issue
PSSR - issue

So what is the problem here Mr. Detective? Of course it can't be PSSR, hahaha.

JFC are you stupid? I've been saying it's the upscalers (PSSR/DLSS) hallucinating from lower res/sample noise over and over again.

The test would be on you since you seem to love nvidia. I assume you have an nvidia card to be this defensive so go disprove the people reporting this issue with low res/samples settings and RR disabled. You can enable TAA and compare.

 

[Bojji]

JFC are you stupid? I've been saying it's the upscalers (PSSR/DLSS) hallucinating from lower res/sample noise over and over again.

The test would be on you since you seem to love nvidia. I assume you have an nvidia card to be this defensive so go disprove the people reporting this issue with low res/samples settings and RR disabled. You can enable TAA and compare.

I have Nvidia card but I don't like Nvidia at all, they have the best tech so what choice do I have?

Do I need to install cyberpunk and do tests for you? You bringed dlss into this thread and showed the same "example" of dlss issues at the last time, last time you also tried to sold frame generation issues as Dlss SR issues.

DF already tested few games with dlss vs pssr and they don't show the same problems with Dlss. Do you still think that Dlss has the same issues as pssr?

 

[Three]

Do I need to install cyberpunk and do tests for you? You bringed dlss into this thread and showed the same "example" of dlss issues at the last time, last time you also tried to sold frame generation issues as Dlss SR issues.

You expect me to do it though? You're the one trying to call other people's posts or experiences false or lies. Not me.

DF already tested few games with dlss vs pssr and they don't show the same problems with Dlss. Do you still think that Dlss has the same issues as pssr?

Maybe you missed the recent posts above? But let me guess, that's the grasses fault not DLSS!

 

[omegasc]

Overall a nice upgrade! I've been thinking of getting this, but still unsure due to my backlog.

Bojji T Three I will install and test Cyberpunk for you guys. Just send some PSN credit my way so I can buy it. Case closed

 

[Physiognomonics]

They are massively overemphasizing PSSR's deficiencies. Yes, it has flicker issue, but overall the IQ is much better than FSR/CB in motion. Every game i tried looks much better in motion vs. PS5 base.

Because this game broke their "CPU limited games won't be improved on Pro" narrative. So they had to destroy this game in others ways, notice that they don't talk about this game as been CPU limited since the first Pro footage. But it was their second favorite CPU limited game with BG3 before.

They are fake news owned by the biggest corporation in the world. This is what they are.

 

[Zathalus]

has anybody compared like for like here to even draw any conclusion like this?
RT is better on nvidia cards in general and often in these comparisons it's against high end cards at max settings or RR enabled with ghosting trailing. Maybe set low settings and similar upscale disable RR and see what happens.

There are already plenty of data points to compare with. Silent Hill 2, Jedi Survivor, Alan Wake 2, Avatar, Dragons Dogma 2, Star Wars Outlaws, and Veilguard. It’s pretty clear PSSR has some sort of issue with denoising or reconstruction when certain RT elements (anything that appears to be using some form of GI/AO be it lumen or something else) are in a game. These contrast with games that don’t use RT and rely on older rendering techniques, there PSSR looks great with only minor issues to speak of.

 

[Zathalus]

Because this game broke their "CPU limited games won't be improved on Pro" narrative. So they had to destroy this game in others ways, notice that they don't talk about this game as been CPU limited since the first Pro footage. But it was their second favorite CPU limited game with BG3 before.

They are fake news owned by the biggest corporation in the world. This is what they are.

They never claimed CPU performance wouldn’t be improved, they claimed if a game is CPU bound to something like 30fps then the Pro won’t be able to make it 60fps. Obviously.

Baldurs Gate 3 still drops to the high 30s in act 3, and Dragons Dogma 2 drops to the high 40s in certain areas of the city even on the performance mode (so 720p internal and while NPCs still pop into existence 3 meters in front of you).

So yes, the Pro is bound by the same CPU limitations as the base PS5, unless you think a 10%-15% boost to CPU performance is transformative.

 

[Three]

There are already plenty of data points to compare with. Silent Hill 2, Jedi Survivor, Alan Wake 2, Avatar, Dragons Dogma 2, Star Wars Outlaws, and Veilguard. It’s pretty clear PSSR has some sort of issue with denoising or reconstruction when certain RT elements (anything that appears to be using some form of GI/AO be it lumen or something else) are in a game. These contrast with games that don’t use RT and rely on older rendering techniques, there PSSR looks great with only minor issues to speak of.

There are games that the scaler can be tested on obviously but nobody has done a match for match settings analysis to test just the upscaler capability. Nobody has really done a controlled test.

For example in Jedi Survivor you have an identical pulsating foliage issue in DLSS



However, this is with high settings RT on one and absolutely no mention of the framerate it's running at either. Higher fps is usually better for a more temporaly stable image but again things like high RT and fps requires hardware power and not a better upscaler.

In SH2 it's pretty bad too with shimmer with RR disabled and with RR you got that trailing instead


So with identical settings and performance is the upscaler performing that differently is the question.

 

[Zathalus]

So with identical settings and performance is the upscaler performing that differently is the question.

Silent Hill 2 using Lumen and DLSS has some denoising issues, but not nearly to the extent of PSSR. PSSR is worse than standard TSR here. Jedi Survivor has as well, RT effects in quality mode are all there and it has the same issues still. DLSS has the foliage issue as mentioned, but not nearly to the same extent. Alan Wake 2 is another good example, stability in both quality and performance mode is bad and the reflection noise in the quality mode is quite frankly terrible. This is with identical settings vs the game on DLSS as DF did. Or just look at the recently released Avatar:



It’s worse than FSR.

Does DLSS also have some trouble with denoising issues at lower input resolution? Yes, but nowhere near to the extent PSSR does. But that is only one issue, PSSR temporal stability (the noise or film grain like effect on games) is just bad compared to DLSS or even FSR for that matter. It needs improvements.

Even more telling is that GG decided having their own upscaler instead of PSSR leads to better results. That’s a Sony 1st party studio deciding that PSSR is not worth using.

 

[Three]

Does DLSS also have some trouble with denoising issues at lower input resolution? Yes, but nowhere near to the extent PSSR does. But that is only one issue, PSSR temporal stability (the noise or film grain like effect on games) is just bad compared to DLSS or even FSR for that matter. It needs improvements.

Where are these tests with identical settings and fps though to show this? Temporal stability is anything in a moving image. when you mention film grain or noise are you perhaps referring to lower quality shadows and specular noise? Again lower sampling.

Even more telling is that GG decided having their own upscaler instead of PSSR leads to better results. That’s a Sony 1st party studio deciding that PSSR is not worth using.

GG worked hard on their own temporal stability since early PS5 launch when DF and co made a similar emphasis on artifacts of its 60fps mode before too. They didn't want to throw that away and have people complain again. Foliage isn't DLSS or PSSRs strong point so it's completely reasonable for GG to do what's best for IQ for a foliage heavy game. Says nothing about relative upscale performance between the two though.

 

[Zathalus]

Where are these tests with identical settings and fps though to show this? Temporal stability is anything in a moving image. when you mention film grain or noise are you perhaps referring to lower quality shadows and specular noise? Again lower sampling.

Alan Wake 2 is you want perfectly identical settings between DLSS and PSSR. DLSS (not 1.0) has been out for over 4 years at this point and is in 600+ games yet games do not have the same temporal or noise issues that PSSR has shown in a number of games when the output resolution is at 4K like these PSSR games do. Noise grain is the temporal stability of the entire image, it looks like the games have this slightly noisy or jittery look to them. It’s a pretty good telltale sign of PSSR, even Ratchet & Clank has it.

Silent Hill 2 PSSR is worse than TSR at the same settings. Avatar PSSR is worse than FSR at the same settings. Alan Wake 2 is worse than DLSS at the same settings. There is video evidence for all of that. DF and GamingTech has videos for them.

 

[Bojji]

Overall a nice upgrade! I've been thinking of getting this, but still unsure due to my backlog.

Bojji T Three I will install and test Cyberpunk for you guys. Just send some PSN credit my way so I can buy it. Case closed

Hahaha, sadly PS5 version is not needed here.

Because this game broke their "CPU limited games won't be improved on Pro" narrative. So they had to destroy this game in others ways, notice that they don't talk about this game as been CPU limited since the first Pro footage. But it was their second favorite CPU limited game with BG3 before.

They are fake news owned by the biggest corporation in the world. This is what they are.

It's still CPU limited, this limit is just lifted since launch. Look how 720p mode performs vs. standard PS5:

Compared to this PC version has much worse CPU performance, lack of optimization in PC build or Capcom DRM/Denuvo. I don't know.

 

[Ironman_Bad]

Still waiting on Callisto feature this game looks fine on ps5pro definite upgrade over ps5 we have just reached that it's trendy and cool to bash pssr point with some people...let's just focus on fixing the bad patches like Sh2 and AW2

 

[Three]

Alan Wake 2 is you want perfectly identical settings between DLSS and PSSR. DLSS (not 1.0) has been out for over 4 years at this point and is in 600+ games yet games do not have the same temporal or noise issues that PSSR has shown in a number of games when the output resolution is at 4K like these PSSR games do. Noise grain is the temporal stability of the entire image, it looks like the games have this slightly noisy or jittery look to them. It’s a pretty good telltale sign of PSSR, even Ratchet & Clank has it.

There have been comparisons of AW 2 at 864p base resolution and identical settings? Haven't really seen them. Specular noise is low sample RT. DLSS does force a more stable image in R&C at the cost of not actually showing things that are meant to be moving and a softer/blurry image.

Silent Hill 2 PSSR is worse than TSR at the same settings. Avatar PSSR is worse than FSR at the same settings. Alan Wake 2 is worse than DLSS at the same settings. There is video evidence for all of that. DF and GamingTech has videos for them.

Obviously but it's poorer performance on TSR. DF doesn't really have like for like controlled tests of these upscalers.

 

[winjer]

In just 2 weeks, we went from PSSR is as good or better than DLSS, to its worse than TSR and FSR2.
Holy crap.

 

[Bojji]

In just 2 weeks, we went from PSSR is as good or better than DLSS, to its worse than TSR and FSR2.
Holy crap.

FSR is worse for the most part but it doesn't have issues like PSSR with RT and SSAO.

TSR was always very good.

 

[Zathalus]

In just 2 weeks, we went from PSSR is as good or better than DLSS, to its worse than TSR and FSR2.
Holy crap.

Small caveat, I’d say it is worse than FSR or TSR when certain RT effects (GI of some sort?) are present, as there is something going on there that is causing problems with the final resolve. Not all aspects of the image are worse, just that noise/flickering becomes very distracting.

Any game that relies on light maps or non RT techniques are really very good. PSSR bests FSR there any day. The Crew just got a PSSR update and it looks quite a bit better.

 

[bbeach123]

This game had terrible upscaling (both dlss and fsr) , one of the few game I had to turn off all of the upscaling because how blurry they cause.

 

[YeulEmeralda]

In just 2 weeks, we went from PSSR is as good or better than DLSS, to its worse than TSR and FSR2.
Holy crap.

Sony was never going to beat Nvidia on this- they are the king of AI.

 

[DenchDeckard]

Sony was never going to beat Nvidia on this- they are the king of AI.

There were definitely some that thought they might.

It doesn't even seem like its performing too badly to be fair. Lots of similarities with DLSS it seems going by the posts in here.

 

[Mister Wolf]

I wonder if Rebirth would look just as good or better with TAAU instead.