I agree, but I still criticize the lack of consistency when it comes to console-centric games...
Until PSSR started being discussed, comparisons between console and PC upscalers were almost nonexistent. If you look closely, you'll notice they talked more about ML upscalers with the PS5 Pro + Switch 2 than about DLSS in the last 5-6 years.
Until yesterday, on consoles, the distance from the TV hid even 720p, but not anymore. This is consistent with the usual things that are only amplified when they arrive on consoles, like the more present 60fps, SSD, VRR, etc.
Now the sensitivity threshold has been raised. If we need an upscaler that performs like PSSR2 on RE9, meaning it's not wasted, it means that the image quality issues on consoles were downplayed before. After all, a PC only has a little compilation stutter and the discussion begins and ends there. 9 times out of 10, the superior image quality isn't praised, and without a PS5 Pro and Switch 2, cross-checks don't exist.
When the story is better understood by considering what's missing, it's a bad sign before considering the reasons.
They're already starting with the off-topic distinctions and the overall approximate weight. So, overall, it's worse than upscalers on PC, right? I figured it out easily: when something on consoles looks similar to PC, it means it's worse, before I get dragged by the neck.
Sharpness on PC is easily changed; RIS2 on AMD is excellent. So the increased sharpness of PSSR2 doesn't exist, and it's possible that it can be achieved using an internal sharpening filter.
These "details" are always skipped, if it's the PS5 that suffers...
Most console gamers I know play two steps away from the big screen, usually placing a chair between the couch and the TV. Who said the average PC gamer is glued to the monitor? All assumptions aimed at minimizing...
But even if that were the case, could one easily argue that certain details on consoles are useless, since what do you want to see from five meters away?
Or does the remote screen hide the flaws but not the strengths?
We don't understand each other. I'll give you some more examples...
XSS has been heavily criticized, as if the threshold below the major consoles wasn't decent, now the Switch-2 still seems good. How many times have you seen XSS results benchmarked against price and/or theoretical power? Almost never, I tell you...
XSS was awful regardless. Don't tell me you really think the difference now could be the difference between TAAU and FSR-2, when it was awful in any way?
What is being considered in this test? Do you see the criticism consistent with the sense of proportion you're talking about?
Shouldn't a company that claims to be a pioneer of technical tests always consider technical differences? That's what I was talking about, not the difference in mass perception, which is often influenced by the industry press.
If it were as you say, this test would be about hairs crossed, not problems. You can't see the double standards because you've been trained to think like that.
I was talking specifically about Oliver, who with the PS5 Pro, once said that from a distance you can't see the glowing red doll (shimmering on Astrobot), another time that you can't see the lack of sharpness. Consistency is the essence, and the only consistency DF can afford is to amplify and/or minimize as needed, tending to have the mid-range consoles at the epicenter of the universe.
Why has there been more talk of ML upscaling with the PS5-Pro and SW2 than in the previous 7 years? Why are there no cross-tests with PC? Why was FSR1 only a flop on the road on PC?
Does it seem natural to you to continue carrying the Ryzen 3600? They test it more than anything else, just like in the PS3/X360 days, they forced PC resolution to 720p, and said the assets were the same even at 1080p (which isn't always true), 60fps wasn't essential, etc. Except for celebrating GOW-3 on PS4, where the assets seemed to be made to be better, and the mipmapping worked wonders—which was also a given on PC, but there wasn't even a mention of it there.
You mean to tell me that perception improves with age? Old age is supposed to improve your eyesight...?
And then the PC alone has a market that rivals the rest of the world, yet there are a lot of PC-only tests that failed, like Mafia, Stellar Blade (how many articles for each patch and for the PS5-Pro version?), Wuchang, etc.
You've veered into the unsolicited. I was talking about DF, and Oliver specifically; you've changed the subject by bringing up the masses' perception of time.
You've been happy with the narrative they've cooked up for you; if you don't follow the same path, it will seem strange, and you're led to justify the real inconsistencies without needing any encouragement.
Do you think that when the new consoles come out, they won't make the differences in upscaling and starting resolution count? You're just wrong to assume so...
By the way, look at the "significant difference" in Hellblade 2 on PS5-Pro, for just a few extra pixels (with accompanying music by Ennio Morricone).
In theory, your point makes sense...
In practice, so looking at the history of the value weighted by DF: It has the usual flaw, that is, the usual double standards depending on what and how you're looking.
Explain to me why Oliver had a big grin when talking about TLOU-2 on PS5-Pro with PSSR.
Remember:
"Hey, it looks native, hippy hippy yeah yeah! Incredible! Impressive!"
In the early days when only the Xbox Series S could handle 720p, there was a sarcastic smile followed by a comparison with the PS3.
There, it wasn't good enough; it was simply poor and confusing.
Now that standard consoles are running at 720p (but also 504p, as we said above), 720p isn't bad after all. From a reasonable distance from the TV (I don't know which one, and it's better not to say...) you can't see much, but ultimately it's good, and at times excellent.
The same concept can be "exported" to the 60fps, VRR, SSD loading, etc. Until they arrived on consoles, it wasn't very useful, and not having them wasn't a problem. After that, it was indispensable, and not having the plus—or any more—was a failure.
Forgive my frankness, but with such a distracted user base, they can easily make donkeys fly; maybe they'll even explain that they fly because of their big ears...
@themightyant When you'll have to start at 720p + FSR2 to reach 60fps: You won't be able to help but worry about this while playing.
The PC and console tests are increasingly separating, more or less like they did at the end of the previous generation's lifespan. Maybe it's because it's better to analyze the consoles with a closer look, perhaps saying that with the above settings it's not bad, before looking for nitpicks in the best upscaler out there, and by far.
Oliver called Dragon Age Veilguard pretty good, starting at 504p. What are we talking about?
Coincidentally, most upscaler tests are done on consoles, almost never directly comparing them to PC.
Anyway, a question arises spontaneously: If the lame DLSS is better than FSR, how does an RTX 3060 stack up against DLSS4? I'd say the PS5 can easily handle it, and could even compete with the PS5 Pro, or is it better not to say that?
Why is the case where the DLSS half performs worse missing from the paper? There's the case where DLSS performs better and the one where both perform poorly, but the counterpart is missing. So even the conclusion seems absurd, because if you show the best and the worst, why conclude with an overall result? Perhaps because the unwritten result is also considered?
I've long noticed the inconsistency of your variable values, the unit of measurement that changes depending on what you're trying to measure. Whenever there's a comparison, otherwise it's like all the missing PC tests, which alone takes up as much space as all the other consoles.
So, double standards, and when it's not convenient, it's better to leave the measurement out, one might say...
First of all, you should criticize the person who wrote this article, not because of the CPU limitation, because it's true, but because it's ridiculous to make such a sensational title for a game that drops to 20fps. Then, on PC, at the first problem, the game would be rejected...
After that, but it would be earlier... If it's CPU-limited, you can render it internally even at 240p, and it will still be CPU-limited. If you don't even have a basic understanding of what it means to be CPU-limited, you shouldn't point out other people's mistakes.
In his review of Mafia: The Old Country, Oliver advised against using performance mode. It's not clear that that mode is a significant downgrade, and a reasonable frame rate means nothing; it was unstable in driving and shooting, yet he said it played well at 30fps.
It's interesting that 2 out of 3 of you played it on PC, but that's the only analysis missing. It joins Wuchang and Stellar Blade, where each patch was analyzed for Sony consoles.
The honorable mentions dedicated solely to the Switch 2 are laughable; they're the icing on the cake... Said as if it weren't a clear stretch, a bottleneck that most will interpret differently.
DL The Beast wasn't clear that consoles were missing what was already in the incomplete PC version. The RT patch came out two weeks ago; you see a partial award without even a dedicated analysis.
Before we digress from current events (e.g., the Snow Drop engine and retro games...), it's best not to select tests based on what's missing from consoles...
Oh, the hair strands didn't seem interesting in Oliver's Pro test; he spent five seconds pretending not to know they weren't there.
You're not invisible, you'd better keep that in mind...
Comments 10
Re: It's Official: Resident Evil Requiem Uses Sony's Brand-New PSSR Upscaler
@MattGPT
I agree, but I still criticize the lack of consistency when it comes to console-centric games...
Until PSSR started being discussed, comparisons between console and PC upscalers were almost nonexistent. If you look closely, you'll notice they talked more about ML upscalers with the PS5 Pro + Switch 2 than about DLSS in the last 5-6 years.
Until yesterday, on consoles, the distance from the TV hid even 720p, but not anymore. This is consistent with the usual things that are only amplified when they arrive on consoles, like the more present 60fps, SSD, VRR, etc.
Now the sensitivity threshold has been raised. If we need an upscaler that performs like PSSR2 on RE9, meaning it's not wasted, it means that the image quality issues on consoles were downplayed before. After all, a PC only has a little compilation stutter and the discussion begins and ends there. 9 times out of 10, the superior image quality isn't praised, and without a PS5 Pro and Switch 2, cross-checks don't exist.
When the story is better understood by considering what's missing, it's a bad sign before considering the reasons.
Re: It's Official: Resident Evil Requiem Uses Sony's Brand-New PSSR Upscaler
They're already starting with the off-topic distinctions and the overall approximate weight. So, overall, it's worse than upscalers on PC, right? I figured it out easily: when something on consoles looks similar to PC, it means it's worse, before I get dragged by the neck.
Sharpness on PC is easily changed; RIS2 on AMD is excellent. So the increased sharpness of PSSR2 doesn't exist, and it's possible that it can be achieved using an internal sharpening filter.
These "details" are always skipped, if it's the PS5 that suffers...
Most console gamers I know play two steps away from the big screen, usually placing a chair between the couch and the TV. Who said the average PC gamer is glued to the monitor? All assumptions aimed at minimizing...
But even if that were the case, could one easily argue that certain details on consoles are useless, since what do you want to see from five meters away?
Or does the remote screen hide the flaws but not the strengths?
Re: DLSS 4.5 "Preset L" Tested: How Good Can a 4K Upscale from 720p Look?
@themightyant
We don't understand each other. I'll give you some more examples...
XSS has been heavily criticized, as if the threshold below the major consoles wasn't decent, now the Switch-2 still seems good. How many times have you seen XSS results benchmarked against price and/or theoretical power? Almost never, I tell you...
XSS was awful regardless. Don't tell me you really think the difference now could be the difference between TAAU and FSR-2, when it was awful in any way?
What is being considered in this test?
Do you see the criticism consistent with the sense of proportion you're talking about?
Shouldn't a company that claims to be a pioneer of technical tests always consider technical differences? That's what I was talking about, not the difference in mass perception, which is often influenced by the industry press.
If it were as you say, this test would be about hairs crossed, not problems. You can't see the double standards because you've been trained to think like that.
I was talking specifically about Oliver, who with the PS5 Pro, once said that from a distance you can't see the glowing red doll (shimmering on Astrobot), another time that you can't see the lack of sharpness.
Consistency is the essence, and the only consistency DF can afford is to amplify and/or minimize as needed, tending to have the mid-range consoles at the epicenter of the universe.
Why has there been more talk of ML upscaling with the PS5-Pro and SW2 than in the previous 7 years? Why are there no cross-tests with PC?
Why was FSR1 only a flop on the road on PC?
Does it seem natural to you to continue carrying the Ryzen 3600?
They test it more than anything else, just like in the PS3/X360 days, they forced PC resolution to 720p, and said the assets were the same even at 1080p (which isn't always true), 60fps wasn't essential, etc. Except for celebrating GOW-3 on PS4, where the assets seemed to be made to be better, and the mipmapping worked wonders—which was also a given on PC, but there wasn't even a mention of it there.
You mean to tell me that perception improves with age?
Old age is supposed to improve your eyesight...?
And then the PC alone has a market that rivals the rest of the world, yet there are a lot of PC-only tests that failed, like Mafia, Stellar Blade (how many articles for each patch and for the PS5-Pro version?), Wuchang, etc.
You've veered into the unsolicited. I was talking about DF, and Oliver specifically; you've changed the subject by bringing up the masses' perception of time.
You've been happy with the narrative they've cooked up for you; if you don't follow the same path, it will seem strange, and you're led to justify the real inconsistencies without needing any encouragement.
Do you think that when the new consoles come out, they won't make the differences in upscaling and starting resolution count? You're just wrong to assume so...
By the way, look at the "significant difference" in Hellblade 2 on PS5-Pro, for just a few extra pixels (with accompanying music by Ennio Morricone).
Re: DLSS 4.5 "Preset L" Tested: How Good Can a 4K Upscale from 720p Look?
@themightyant
In theory, your point makes sense...
In practice, so looking at the history of the value weighted by DF: It has the usual flaw, that is, the usual double standards depending on what and how you're looking.
Explain to me why Oliver had a big grin when talking about TLOU-2 on PS5-Pro with PSSR.
Remember:
"Hey, it looks native, hippy hippy yeah yeah! Incredible! Impressive!"
In the early days when only the Xbox Series S could handle 720p, there was a sarcastic smile followed by a comparison with the PS3.
There, it wasn't good enough; it was simply poor and confusing.
Now that standard consoles are running at 720p (but also 504p, as we said above), 720p isn't bad after all. From a reasonable distance from the TV (I don't know which one, and it's better not to say...) you can't see much, but ultimately it's good, and at times excellent.
The same concept can be "exported" to the 60fps, VRR, SSD loading, etc. Until they arrived on consoles, it wasn't very useful, and not having them wasn't a problem. After that, it was indispensable, and not having the plus—or any more—was a failure.
Forgive my frankness, but with such a distracted user base, they can easily make donkeys fly; maybe they'll even explain that they fly because of their big ears...
Re: DLSS 4.5 "Preset L" Tested: How Good Can a 4K Upscale from 720p Look?
@themightyant
When you'll have to start at 720p + FSR2 to reach 60fps: You won't be able to help but worry about this while playing.
The PC and console tests are increasingly separating, more or less like they did at the end of the previous generation's lifespan.
Maybe it's because it's better to analyze the consoles with a closer look, perhaps saying that with the above settings it's not bad, before looking for nitpicks in the best upscaler out there, and by far.
Oliver called Dragon Age Veilguard pretty good, starting at 504p. What are we talking about?
Re: Is Switch 2's "Tiny" DLSS better than FSR 3?
Coincidentally, most upscaler tests are done on consoles, almost never directly comparing them to PC.
Anyway, a question arises spontaneously:
If the lame DLSS is better than FSR, how does an RTX 3060 stack up against DLSS4? I'd say the PS5 can easily handle it, and could even compete with the PS5 Pro, or is it better not to say that?
Why is the case where the DLSS half performs worse missing from the paper?
There's the case where DLSS performs better and the one where both perform poorly, but the counterpart is missing. So even the conclusion seems absurd, because if you show the best and the worst, why conclude with an overall result? Perhaps because the unwritten result is also considered?
I've long noticed the inconsistency of your variable values, the unit of measurement that changes depending on what you're trying to measure. Whenever there's a comparison, otherwise it's like all the missing PC tests, which alone takes up as much space as all the other consoles.
So, double standards, and when it's not convenient, it's better to leave the measurement out, one might say...
Re: Review: Microsoft Flight Simulator 2024 Gets A Solid, Unsurprisingly Great PS5 Port
First of all, you should criticize the person who wrote this article, not because of the CPU limitation, because it's true, but because it's ridiculous to make such a sensational title for a game that drops to 20fps. Then, on PC, at the first problem, the game would be rejected...
After that, but it would be earlier... If it's CPU-limited, you can render it internally even at 240p, and it will still be CPU-limited.
If you don't even have a basic understanding of what it means to be CPU-limited, you shouldn't point out other people's mistakes.
Anyway, best wishes and a happy new year.
Re: Review: Microsoft Flight Simulator 2024 Gets A Solid, Unsurprisingly Great PS5 Port
@PGR
A better upscaler could have helped reduce shimmering, and perhaps enable RT features, starting at a much lower resolution.
I think you completely misunderstand that the upscaler won't "overcome" the processor's limitations; it will "only" improve the GPU's bottleneck.
Re: The Best and Worst PC Ports of 2025
But which ports?
Re: Digital Foundry's 2025 Graphics of the Year Awards
In his review of Mafia: The Old Country, Oliver advised against using performance mode.
It's not clear that that mode is a significant downgrade, and a reasonable frame rate means nothing; it was unstable in driving and shooting, yet he said it played well at 30fps.
It's interesting that 2 out of 3 of you played it on PC, but that's the only analysis missing. It joins Wuchang and Stellar Blade, where each patch was analyzed for Sony consoles.
The honorable mentions dedicated solely to the Switch 2 are laughable; they're the icing on the cake... Said as if it weren't a clear stretch, a bottleneck that most will interpret differently.
DL The Beast wasn't clear that consoles were missing what was already in the incomplete PC version. The RT patch came out two weeks ago; you see a partial award without even a dedicated analysis.
Before we digress from current events (e.g., the Snow Drop engine and retro games...), it's best not to select tests based on what's missing from consoles...
Oh, the hair strands didn't seem interesting in Oliver's Pro test; he spent five seconds pretending not to know they weren't there.
You're not invisible, you'd better keep that in mind...