No idea and don't care. All I know is that disabling DLSS fixed this issue back then. Again even a game update can add motion vectors to deal with DLSS issues. We jump through these hoops all the time when somebody mentions some issue DLSS causes and it's always the same deflection from the usual crowd.
-
Hey Guest. Check out your NeoGAF Wrapped 2025 results here!
PS5 Pro is getting PSSR 2.0 between January and March 2026
- Thread starter IbizaPocholo
- Start date
- Rumor Hardware
Zathalus
Gold Member
Man, they make half a dozen videos for a ton of Sony games on the regular. Even have exclusive DF directs just for Sony state of plays. How is that proof of anything? Once again, Mark Cerny, fan, listens to the podcasts, exclusive interview. You're just mad because they point out the flaws with some PSSR titles.I don't follow you they have to say ps5 is a shitty hardware and despise the ps5 exclusive? I just reported outside John they were always the MS bitches until the thing start to going wrong, they make like 6 videos analises for Forza releases either and if I'm not wrong also for the other MS ip, are you for real?Nothing wrong to prefer a brand but I thought I stated the obvious, they never were exactly impartial and they used a very different tone for the 2 brands.
Gaiff
SBI’s Resident Gaslighter
Lol at the irony. You and your posse are always crying about DF the moment they bring up Sony. Thankfully, the likes of James Sawyer Ford, Mibu Okami, and a few others of their ilk bit the dust, so we don't have to deal with them anymore.
Last edited:
Wut
of course they have to cover the ps5, how they can ignore it, the fuck you are talking about I just said their narrative always favoured the xbox brand for years also saying things are changed lately for obvious reason, the hell it has to do the pssr with that
Last edited:
Zathalus
Gold Member
What game update? The game hasn't had any updates since they added FSR 3.1. The hoops you go through to justify your fanboyism sometimes is funny. Instead of admitting the far more likely scenario (a bug), it's somehow a super secret DLSS version or secret patch that fixed the issue.
I already did come up with something substantial. Video evidence of the exact test showing the issue in action and my own experience of it that I know was fixed by disabling DLSS. The fact that you can't reproduce it means fuck all to me. Especially when I can't actually check anything you're doing in your methodology when producing these "conclusions". So if it was a "PC bug" you can easily reproduce it too correct? Can I claim it never even existed now because you can't?
So if they didn't update it what happened to the "PC bug", it fixed itself? Again just because you don't know the exact parameters doesn't mean shit.
Last edited:
Gaiff
SBI’s Resident Gaslighter
One fucking example isn't substantial. Half a dozen with different rigs, set ups, and like four DLSS versions is.
No one gives a damn what it means to you. The world doesn't revolve around you. We're trying to reproduce the issue and none of us can under the same parameters where the only changing variable is DLSS. What matters is objective reality.
No, because not everyone gets the same bugs lol. A bug can be slippery and variable enough not to be reproducible even under very similar or identical scenarios. DLSS is not. It's consistent. Could it be an interaction problem between a certain light setting and DLSS? Could it be that some weird stuff happened that cause something not to load properly? Yes to all. However, we have attempted to isolate DLSS and have found it not responsible for this visual issue. You therefore can't claim it was the cause, unless you can reproduce it.
Zathalus
Gold Member
A bug is obvious. Because it's appears to be so hard to replicate in the first place. That is kind of how a huge amount of bugs work. Maybe switching AF to 4x after changing the resolution scaling causes motion vectors to mess up. You know, like a bug.
All I know is you can use the in-game DLSS, 3.7, 4, 4.5, and none of them recreates the issue. Simple.
Last edited:
omegasc
Member
could have been a driver issue
Possibly, but all I know is that disabling DLSS fixed this issue back then.
Yes and disabling DLSS fixed it. Not switching AF to 4x. If you think it's that then try to reproduce it since they obviously haven't updated the game or anything so you should be able to.
Zathalus
Gold Member
What narrative? The very sensible one where they thought the XSX would outperform the PS5 before launch? Which most people thought? Yet they were literally the very first outlet to report on the PS5 absolutely dunking on the XSX in AC Valhalla and DMCV? And had a massive Bluepoint interview lined up for Demon Souls?Wut
And PSSR was brought up as an example. Because you would likely twist them pointing out PSSR flaws as some sort of anti Sony agenda. When they are the very ones who point out all the games it does well in as well.
Zathalus
Gold Member
I mean, you literally don't know if disabling DLSS fixed the issue. It's just what shows up in his capture. For all we know the next time he booted the game the issue was fixed. Because, you know, bug. And I never claimed 4x AF fixed or caused the issue, it was just and example of how easy it is to break behaviour in games.
DeepEnigma
Gold Member
Wut
HeisenbergFX4
repeated so many times and provided receipts to mods about them bringing DF "close to the vest." Along with many other online publications. Yet the selective memory regulars on this forum pretend it never happened.It was quite obvious, really. But let's all pretend, it was "merely organic."
Zathalus
Gold Member
If they thought they had manipulated DF than did did a tremendously shit job of it, considering the very two console comparisons they did after each console launched. The AC Valhalla one caused a fanboy meltdown on Twitter.
It was quite obvious, really. But let's all pretend, it was "merely organic."
Did you not read that the way I fixed this issue was to disable dlss? Even got a suggestion from Boji I believe that "changing preset" would fix it back then.
Bojji
Gold Member
I don't remember telling that but who knows...
If I had game installed at the time I would easily debunk it. Remember that Alex used third party tool (internal resolution change for DLSS) to do comparison with PSSR, he introduced more complexity so something might just stopped working correctly.
Its the stupid idea that you think my own experience of disabling DLSS and not getting the issue is not "controlled". Or that when DF did their comparison they didn't do "controlled conditions" for a fair comparison. It's you who has failed to show your hypothesis of "a bug on PC" under any conditions or with any evidence. Throwing out my evidence with nothing but conjecture.
Last edited:
Gaiff
SBI’s Resident Gaslighter
It's exactly a bug because we can't reproduce it, but have seen it.
Your conclusion that DLSS was the cause is conjecture, because if it was the case, we'd be able to reproduce. My conclusion that it's a bug is much stronger, because we can't.
So it can be a bug that you can't reproduce but it can't be "an issue with DLSS" that you can't reproduce. I don't think you know how this works. As I said before there are many parameters but DLSS was the cause of it because I have my own evidence of that and one even captured on video in a controlled test.
Last edited:
Believe whatever you want but there were a lot of evidences about Leadbetter to be an xbox chill especially in the golden era just please doesn't insult the other intelligence talking about a persecution complex.
Last edited:
Gaiff
SBI’s Resident Gaslighter
Oh, for fuck's sake. How many times have you been told? Bugs are by nature slippery and can often be very hard to reproduce. That's why bug fixes can take so long, because you might fail to see them under identical conditions. DLSS isn't a bug. It's consistent. If there is ghosting caused by DLSS, it'll still be there after shutting down your PC or reloading the game. It won't appear one time and be gone the next. That goes for other issues.
Your conclusion is incorrect because if it was correct, we'd have no issue reproducing your problem as you assert that it was DLSS, not a bug.
Could it be that you changed your resolution and it went away? Maybe you flicked between multiple upscalers and it triggered something that fixed it? Whatever the case, we know it can't be DLSS, because none of us can replicate it. Therefore, bug is the only logical conclusion, because it covers every other possible scenario.
Just like a bug you haven't tested all parameters, drivers, hardware, framerate, exact DLSS version, exact game version, exact preset, exact resolution. Like I said I don't care of these parameters all I know is that disabling DLSS fixed this back then. This was reproducable too consistently.
Bojji
Gold Member
So who confirmed that and where are the receipts?
Gaiff
SBI’s Resident Gaslighter
Yeah, which is my point. A bug has a million different possibilities, DLSS does not.
You assume disabling DLSS fixed it, but it could have easily been something else and almost certainly was, since, you know, none of us can replicate it.
No it couldn't have because it certainly did. Why is your assumption correct when you certainly known you haven't replicated the parameters?
The receipts are my own evidence and the video I've shown you. Just because you can't replicate those conditions doesn't mean shit. Like I said it's like the "works on my pc" junior devs throw out at QAs. That's all I've seen here.
Zathalus
Gold Member
Evidence like the PS5 box tearing? Alright then.
Gaiff
SBI’s Resident Gaslighter
Except I did replicate the parameters. Why don't you do it again and show it instead of whining about us not being able to reproduce your issue?
How when none of the parameters are given? You're talking nonsense.
Gaiff
SBI’s Resident Gaslighter
The game, the game's version, DLSS version, settings, resolution, and upscaling level are all known.
Last edited:
MatrixReloaded101
Member
Hopefully this update is worth the wait.
Last edited:
Horseshit. how do you know which versions these were exactly? let alone the exact settings, hardware and drivers? As I said nonsense.
Senua
Member
Gaiff
SBI’s Resident Gaslighter
1. The game hasn't been updated since June 2024 with the FSR 3.1 update. This was over a 6 months before that video
2. The DLSS version hasn't been updated either
3. The upscaling level is literally given in the video as well as the resolution
4. The settings are also known. It says "approximate PS5 Pro settings" lol
If it's the hardware or drivers, then DLSS isn't the problem, is it?
Last edited:
Ah yes because I'm sure you had the same framerate, same "ps5 pro approximates" and everything set exactly the same. According to
Zathalus
the "default DLSS" is 3.7 but 3.7 launched a year after the game, the default was 3.1. The only explanation is time travel obviously! Or somebody is making shit up. Not sure who.
Gaiff
SBI’s Resident Gaslighter
What the fuck does frame rate have to do with this lol?
Yes.
Yes, the DLSS version hasn't been updated since the last June 2024 update (even before that I think) and the DF video came out well after that. It does use DLSS 3.7.0.0 as of now and as of the DF video.According to
I mean, you say you were able to reproduce it and you weren't even trying to match their settings.
Last edited:
It was quite obvious until well into the generation they were still convinced on the horsepower of the series x meaning it should be outperforming ps5 and they couldn't quite understand it still, then about 3 years in the point was pushed less and don't think it was until 2024 they actually spoke the unspoken out loud that they had been wrong with their original analysis and firm beliefs that we would see the Xbox series x versions consistently outperforming ps5. Once that was said it's been fine since, but I watched every episode of the podcast and all their videos 2020-now and wasn't aware of the behind the scenes discord agendas and stuff until after it had all failed.. think Alex was the main one behind that vs Richard, but I only paid attention to that on a cursory level so never bothered about the nitty gritty but the evidence is out there for all to find if you really can be bothered. They always reviewed all versions of a game where they could and even went out day one and paid rrp if necessary to make that happen. Any bias there was didn't actually compromise the content they made that I've seen and at the end of the day we all have opinions of the hardware we own, which is why although I own switch 2, series x and ps5 pro I picked up resident evil village, 2 & 3 remake on series x and only chose re4 remake on ps5 once digital foundry showed that it was handling it better, but re9 is fully paid up deluxe edition on ps5 as I want the best experience and I believe capcom will deliver that on pro with their outlined pro mode
It's always developer dependant and going back to the subject of the thread, new version means Jack all unless Sony covers the costs of the devs going back and patching it to bring out the magic - which personally given how Marc way oversold what they delivered is the least they can do to fix the tarnished image of what the pro is all about. Sony should also be paying capcom for a pro exclusive re9 vr mode (because I don't think re9 would be able to hit the fidelity level necessary on base ps5 without re projection and that just sucks)
It's always developer dependant and going back to the subject of the thread, new version means Jack all unless Sony covers the costs of the devs going back and patching it to bring out the magic - which personally given how Marc way oversold what they delivered is the least they can do to fix the tarnished image of what the pro is all about. Sony should also be paying capcom for a pro exclusive re9 vr mode (because I don't think re9 would be able to hit the fidelity level necessary on base ps5 without re projection and that just sucks)
Thats how I know your tests are uncontrolled bollocks because you don't even know how any of this works
Framerate can hide temporal issues in an upscaler. If I'm running a higher framerate my accumulation is different.
Try learning from this video
But you just told me
"2. The DLSS version hasn't been updated either"
So you were talking shit. DF do not specify which version they used in that part they compared 3.1 and 3.7 but did not explicitly specify which it was afterwards.
Don't want to enable VPN to look at whatever it is you're posting anymore. Yes I had the issue. Was not trying to reproduce what's your point?
I mean, you say you were able to reproduce it and you weren't even trying to match their settings.
Gaiff
SBI’s Resident Gaslighter
It can hide temporal issues when there are temporal issues, not when you make shit up and draw incorrect conclusions.
No chance of that happening, since my game was running at 60fps too.
Yes, I said the game hasn't been update since the June 2024 FSR 3.1 update and that DLSS hasn't been updated either, you know, it means since that exact same date. Not that DLSS was never updated at all.
Unless they deliberately downgraded the DLSS version to 3.1 to make it look unfavorable, why wouldn't they just use the default one that ships with the game?
My point is that you're full of shit and an absolute embarrassment of a poster. You weren't even trying to reproduce it, yet you found it. I'm trying to reproduce it as closely as I can, and you go "bu bu but, X and Y", when I shouldn't even need to match anything closely to see it for myself. By all means though, keep crying about DF and moving the goalposts. You can't fucking read, can't analyze, and can't debate worth a lick. You've been schooled and badly at that.
Over and out.
Last edited:
Too difficult to admit you incorrectly thought framerate doesn't affect DLSS ghosting issues huh?It can hide temporal issues when there are temporal issues, not when you make shit up and draw incorrect conclusions.
No chance of that happening, since my game was running at 60fps too.
Horseshit you claimed to know the exact setup because DLSS hasn't been updated. They did intentionally downgrade to 3.1 and test it as I already said (though I think they actually did the opposite and intentionally upgraded though admittedly I don't know on what game version this was tested on anyway)
It shipped with 3.1 though, not sure which version they used and when it was updated. You claimed DLSS version hasn't changed. Df showed both 3.1 and 3.7 and didn't specify. But they clearly intentionally either downgraded or upgraded already.
No you're not trying to reproduce it by testing everything. You're claiming you have and that the issue I've seen for myself is not attributed to something that was clearly related to DLSS. You're so embarrassing that you can't even admit you were wrong about the framerate affecting ghosting issues in an upscaler so even if you were to reproduce it I have no faith in you coming here and admitting it. Over and out indeed. You're tiring.
Last edited:
Gaiff
SBI’s Resident Gaslighter
lmao, just fuck off. I said DLSS hasn't been updated either in reference to the June 2024 update. I then proceeded to show you a screenshot of it running DLSS 3.7.0.0. I never made the claim that DLSS was never updated since release, you made that shit up because you're getting annihilated.
Last update was June 2024 that introduced FSR 3.1. DF used that version which has 3.7.0.0, so unless you think they deliberately downgraded the DLSS version, they were using 3.7, so fuck off.
Jesus Christ, I wish they took you out with the rest of the fanboys like Mibu and James Sawyer.
Last edited:
They did either downgrade and compare to DLSS 3.1 or upgraded and compared to 3.7, either way it wasn't said what exactly they used. Either way they intentionally did one or the other and didn't specify which ones they were testing in that specific clip. Depends when they did the analysis exactly.
Last edited:
SegaSnatcher
Member
We really need proper upscaling for Cyberpunk 2077 on PS5 Pro. I doubt we will get this update, but man the current FSR is holding that port back.
PaintTinJr
Member
Before the subpoints it seem you are still saying that PlayStation's upscaler just needed longer to cook, implying this is purely a software issue and you some how think a 22.6TF SM + 181TOPs of tensor cores (sparsity enabled 360TOPS) of a 4070 is parity with a 16.7TF GPU that shares that resource with its TOPs, and in a 60fps game uses 18 TOPS for PSSR and leaves just 15.7TFs for the game.
The first DF comparison was with DLAA for Ratchet running on a 5090 unlocked frame-rate. and see above for the comparison between a 4070 and Pro for resource comparison for equal software, even without taking account for DLSS3 and 4 using far more TOPs per frame than the Pro's zero latency 18-30TOPs for PSSR per second.
I agree they will replace it with a superior scaler, but it is mostly for games that are bad and are bad native and still look AA with any scaler on 5090. So I draw the line where you blame PSSR for being under cooked for what is a classic garbage in garbage out scenario, but don't take issue with the garbage in and use it for comparisons.
Bojji
Gold Member
Arguably even their own upscaling (TAAU based) looked better before they patched in FSR2. Game also was reconstructing to full 2160p, that was reduced ~PL launch.
winjer
Gold Member
The leaked version of FSR4, used a 100.000 parameters on a 39-layer U-Net architecture.
This is a very small model, especially considering what it can do.
SegaSnatcher
Member
If PSSR 2.0 is anywhere close to FSR4 then we could see a dramatic improvement, but I just wouldn't bet on CDPR touching that game again anytime soon.
Bojji
Gold Member
Even PSSR1 would look better than what they use right now. I think they won't do any update but they really should, they already contacted some third porty firm for latest PC patches.
Similar threads
- Locked