• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Digital Foundry: Cyberpunk 2077 Phantom Liberty - DF Tech Review - PS5/Xbox Series Tests + 2.0 Upgrade Breakdown

Vergil1992

Member
That video is a year old.
Yes, but I put it to point out that it doesn't seem like the situation is much different. In that video from a year ago the performance difference was very similar, the XSX resolution was higher and it also used VRS. Also, it was also to point out that minimum and maximum resolutions don't tell the whole story: in VG Tech's pixel counts (from that video), even though the minimum and maximum resolutions were identical, in most situations they saw higher resolutions in XSX.
 

Tripolygon

Banned
Hello, it's my first message. My English is basic-medium, I hope I can be understood correctly even if I sound strange.

I think the performance difference between PS5-XSX is being greatly exaggerated, it's a difference of more or less 5 fps in areas with multiple NPCs or driving. Essentially, it's a very similar difference to the one before the patch:


XSX always had a higher DRS and the difference seems to remain very similar. The minimum and maximum resolutions do not tell us all the details. In that VG Tech video it is clear, the minimum and maximum resolutions were the same on PS5/XSX, but however, the average resolutions higher on Xbox:

Kabuki Entrance - PS5 2176x1224, Series X 2304x1296 Near Police Station - PS5: 2435x1370, Series x1544 Streetkid Start - PS5: 2062x1160, Series X: 2062x1160

It is not that simple. While XSX has a slight edge in terms of overall output resolution PS5 has the edge in internal resolution on surface details. On XSX using VRS it is rendering at half the resolution. That is why you get situations like below when PS5 looks higher resolution even if on average the overall output resolution should be higher on XSX. When you are playing on XSX you wouldn't notice it because you are not comparing it to something that is better side by side, but it doesn't mean the difference is not there when you compare. It is a tradeoff between overall resolution vs resolution of surface details and performance.

Cyberpunk-2077-Phantom-Liberty-DF-Tech-Review-PS5-Xbox-Series-Tests-2-0-Upgrade-Breakdown-4-15-scree.png
 

shamoomoo

Banned
Hello, it's my first message. My English is basic-medium, I hope I can be understood correctly even if I sound strange.

I think the performance difference between PS5-XSX is being greatly exaggerated, it's a difference of more or less 5 fps in areas with multiple NPCs or driving. Essentially, it's a very similar difference to the one before the patch:


XSX always had a higher DRS and the difference seems to remain very similar. The minimum and maximum resolutions do not tell us all the details. In that VG Tech video it is clear, the minimum and maximum resolutions were the same on PS5/XSX, but however, the average resolutions higher on Xbox:

Kabuki Entrance - PS5 2176x1224, Series X 2304x1296 Near Police Station - PS5: 2435x1370, Series x1544 Streetkid Start - PS5: 2062x1160, Series X: 2062x1160



I think the problem we have here is similar to The Witcher 3. I also had problems maintaining 60fps in areas like Novigrad, especially on XSX, but after a few patches (some even made performance worse on all platforms) the performance was 1 :1 with PS5, but the resolution difference did not change.



My theory is that on the CPU side it is more difficult/problematic to optimize on Xbox Series, probably because of the API. On PC there was a lot of CPU overhead when The Witcher 3 was released. It's also possible that the resolution difference in those scenes is too high for similar performance and the problem seems to be only in Night City, not in its DLC, that the performance of PS5/XSX is very equal (I would like to know the minimum, maximum and average resolutions). Honestly, I don't think the original game code has changed much on an optimization level. It seems that on PS5 the framerate has been improved compared to the old version, but on XSX there does not seem to be an improvement at all. Maybe the PS5 version takes priority and the XSX patches will fix it? We will see it in time.



In the use of VRS, Oliver himself indicates it in the video. That is a screenshot taken in a specific situation and "looking" for the artifacts, in the normal game it is impossible to see. I'm playing the XSX version and I'm pretty sure I haven't seen objects with such a low resolution. I don't think this has much relevance; VRS doesn't improve performance much, and what Oliver has captured probably won't be visible to anyone. VG Tech also pointed out that Cyberpunk used VRS in the previous version of the game, the difference is that it didn't capture any anomalies!




Also as an added detail, loads are significantly faster on PS5/PC than XSX/XSS. Something curious because the previous version, which was essentially identical on a technological level, loaded just as fast (around 9 seconds on all platforms). We could say that the difference is due to the faster PS5/PC SSD, but that would be the case if the loads were faster than before, but on Xbox platforms they have gone from 9 seconds to 17-18 seconds. I think something is wrong here.




At this time, it can be summarized that the XSX has a higher resolution and the PS5 has a more stable framerate. At least according to DF and ElAnalistadeBits. The exact difference in resolutions is very difficult to know. In principle, when you have a higher dynamic resolution it is because you have more power and better framerate (if it works well), if in XSX it works worse despite having higher resolutions, I only see three options:

DSR is not working properly and is higher than it should be on Xbox.

DRS works, but the drops are not on the GPU side.

Everything works as it should, and CD Projekt will address this difference with some update (such as The Witcher 3).



Honestly, I think everything is practically the same as in the previous patch. Higher DRS on Xbox, more stable framerate on PS5 and on XSX it already used VRS from before. The only "new" is that the Xbox Series framerate seems to be the same or worse than before the update, while the PS5 code has received improvements. Even loads are now slower on Xbox. As I said before, it could be argued that they have now made better use of the PS5 SSD, but the thing is that the loads remain exactly the same on PC/PS5, only those on the Xbox platforms have become slower.

I think these are issues that CD Projekt will address.


I hope I don't offend anyone or have my comment be considered "console wars." I have both systems and I don't consider myself a fan of either.

Interesting. But Cyberpunk uses VRS,if VRS can be used to improve performance and visuals then how would that factor into your hypothesis? Nevermind.
 
Last edited:

Vergil1992

Member
It is not that simple. While XSX has a slight edge in terms of overall output resolution PS5 has the edge in internal resolution on surface details. On XSX using VRS it is rendering at half the resolution. That is why you get situations like below when PS5 looks higher resolution even if on average the overall output resolution should be higher on XSX. When you are playing on XSX you wouldn't notice it because you are not comparing it to something that is better side by side, but it doesn't mean the difference is not there when you compare. It is a tradeoff between overall resolution vs resolution of surface details and performance.

Cyberpunk-2077-Phantom-Liberty-DF-Tech-Review-PS5-Xbox-Series-Tests-2-0-Upgrade-Breakdown-4-15-scree.png
The problem with what you say is that VRS was actually here before. These VRS artifacts and anomalies were already there before, because VG Tech confirms the presence of VRS in their video, and if they confirm it it is because they had already observed it. The only difference is that no one had searched for and captured this, and of course no user had noticed it.

The difference is that Oliver from DF has had a better eye than us, he is dedicated to this, and has looked for the characteristic artifacts of using VRS (decreasing the resolution in non-visible areas). He himself says that in the normal game it is not noticeable, and the capture was taken on a specific object, in a very specific situation (low contrast and camera rotations). The proof that VRS is not noticeable is that it was already used on Xbox long before and no one had complained about the image quality. We see it when someone with a "better eye" than us takes a screenshot. And that is when we believe that it reflects reality.


I personally wouldn't put much stock in VRS artifacts. Unless you look for them, you won't see them. I also don't think almost anyone can tell the PS5/XSX resolution differences, honestly.


@

S shamoomoo

I don't know if I understood you correctly. Yes, VRS allows for improved performance, although unfortunately, it seems that it was more of a promise than something realistic. In Dead Space it was completely eliminated and the improvement was negligible.

In this case, we don't know, maybe the difference in average resolution is reduced between PS5 and XSX (or does not exist), or maybe they eliminate VRS and it is not noticeable at all.
 
Last edited:

Vergil1992

Member
私は PS5 と Xbox Series X の両方で Doom Eternal も所有しています。
何を言っているのか理解できません。
実機の画像を添付しましたので、見比べてみてください。
mOzFWde.jpg
q12cq2P.jpg
vU3FM0E.jpg
It doesn't look like that in XSX...

Captura-de-pantalla-2023-09-27-183430.png




The capture is compressed by Twitter. But it looks like it matches the PS5 version. I'm not trying to say that this user is lying, just that maybe the textures hadn't loaded.
 
Last edited:

Tripolygon

Banned
The problem with what you say is that VRS was actually here before. These VRS artifacts and anomalies were already there before, because VG Tech confirms the presence of VRS in their video, and if they confirm it it is because they had already observed it. The only difference is that no one had searched for and captured this, and of course no user had noticed it.
Now apply that to the final resolution, you cannot tell the difference between 2176x1224 and 2304x1296, yet you conclude that XSX has a higher resolution to PS5 because you were told so. Will you notice that PS5 has an average lower resolution compared to XSX during gameplay? No, unless someone pixel counts and tells you the difference. It doesn't mean the difference is not there. It's the same way XSX using VRS renders at half the resolution of the final output on internal surfaces giving PS5 an edge when compared side by side but when you are just playing you wouldn't notice if you aren't looking for it. That is how our vision works, our vision is only sharp in a small 5% window which is why foveated rendering works so well.

I never used to notice these things, but I do now, I never use to notice frame drops but now I do. There are many people on this forum who do.
 

Vergil1992

Member
Now apply that to the final resolution, you cannot tell the difference between 2176x1224 and 2304x1296, yet you conclude that XSX has a higher resolution to PS5 because you were told so. Will you notice that PS5 has an average lower resolution compared to XSX during gameplay? No, unless someone pixel counts and tells you the difference. It doesn't mean the difference is not there. It's the same way XSX using VRS renders at half the resolution of the final output on internal surfaces giving PS5 an edge when compared side by side but when you are just playing you wouldn't notice if you aren't looking for it. That is how our vision works, our vision is only sharp in a small 5% window which is why foveated rendering works so well.

I never used to notice these things, but I do now, I never use to notice frame drops but now I do. There are many people on this forum who do.
I agree with you. In fact, I said it. We cannot notice the differences in resolutions between PS5 and XSX, nor the use of VRS. They are still technological curiosities. Nothing else. Even if we put them side by side, probably 99.9% of us couldn't tell which is which.

But it is also not honest to say about a version "it has a better framerate" while omitting that it probably runs at lower resolutions. I'm not saying it for you.


@

Tripolygon Tripolygon

The screenshot that I published is identical to the PS5 version, put one next to the other and you will see it. It is true that it is much more compressed, but the textures and shading are exactly the same as the screenshot he posted from PS5. Although you are right that the fingers do look the same. I have the physical PS5 version (I played the XSX version on Game Pass), if I have time I will do some testing.
 
Last edited:
I agree with you. In fact, I said it. We cannot notice the differences in resolutions between PS5 and XSX, nor the use of VRS. They are still technological curiosities. Nothing else. Even if we put them side by side, probably 99.9% of us couldn't tell which is which.

But it is also not honest to say about a version "it has a better framerate" while omitting that it probably runs at lower resolutions. I'm not saying it for you.


@

Tripolygon Tripolygon

The screenshot that I published is identical to the PS5 version, put one next to the other and you will see it. It is true that it is much more compressed, but the textures and shading are exactly the same as the screenshot he posted from PS5. Although you are right that the fingers do look the same. I have the physical PS5 version (I played the XSX version on Game Pass), if I have time I will do some testing.
1) How does VRS enhance framerate?
2) Is it honest to say the resolution is definitively higher when said VRS is in use?
3) see what i did there?
 
Last edited:

Vergil1992

Member
Your image shows the same problem.
I'm watching PS5 videos (I'd have to play it again to take screenshots haha) and the same thing happens, the fingers also have lower resolution textures:

Captura-de-pantalla-2023-09-27-185053.png


Video:



@

Die Namek Ability Die Namek Ability

Good point. Maybe that's not entirely fair, maybe VRS is what allows XSX to have the advantage in resolution.

We can't know. It's also fair to point out that VRS seems to work "better" on XSX, or at least easier to implement. But we do not know what the real resolution gap is, nor how much the use of VRS improves performance.
 
Last edited:

Mr Moose

Member
I'm watching PS5 videos (I'd have to play it again to take screenshots haha) and the same thing happens, the fingers also have lower resolution textures:

Captura-de-pantalla-2023-09-27-185053.png


Video:



@

Die Namek Ability Die Namek Ability

Good point. Maybe that's not entirely fair, maybe VRS is what allows XSX to have the advantage in resolution.

We can't know. It's also fair to point out that VRS seems to work "better" on XSX, or at least easier to implement. But we do not know what the real resolution gap is, nor how much the use of VRS improves performance.

PS4 version. It wasn't released until 2 weeks after that video.
 

Tripolygon

Banned

Tripolygon Tripolygon

The screenshot that I published is identical to the PS5 version, put one next to the other and you will see it. It is true that it is much more compressed, but the textures and shading are exactly the same as the screenshot he posted from PS5. Although you are right that the fingers do look the same. I have the physical PS5 version (I played the XSX version on Game Pass), if I have time I will do some testing.
Since you asked i guess here. Look at the fingers, the texture resolution and detail is higher on PS5. What you posted is a YouTube capture with a different color saturation and less brightness.
2wUEDhK.jpg
 
Last edited:

adamsapple

Or is it just one of Phil's balls in my throat?
The detail on the finger is not VRS, that's a difference in sheer quality of the texture. Likely a relic of the games cross-gen nature.

VRS doesn't do that, it runs the same texture but lowers internal geometry resolution.

Y'all are using the Doom image wrong.
 
Last edited:

Vergil1992

Member
Since you asked i guess here. Look at the fingers.
2wUEDhK.jpg
Okay, you're right, I especially looked at the face and it looks like the finger textures match the PS4/One version. I have watched more videos and as another user says, what confused me is seeing the PS4 version working on PS5.

In any case, it seems more like an oversight. It looks like an unloaded texture or graphical settings inherited from Xbox One. I don't think it has anything to do with VRS or Doom Eternal's performance on XSX (which I think was superior to PS5 overall). At the end of the day, textures are just a matter of memory.
 
It could be I/0 related as NXGamer IGN videos shows the load times and the PS5 loads in 8 seconds and and the Series' consoles are loading in the double digits.

Yes I noticed it. PS5 is loading almost 100% faster than XSX and about on par with high end PC. It's odd but it has nothing to do with framerate drops during gameplay with normal max 33ms frame-times.
 

jroc74

Phone reception is more important to me than human rights
Hey...if VRS issues arent visible when playing, we no longer need to have 5-10 pages about 100-200 difference in overall resolutions between the consoles then.

Analysis threads about to become alot smaller.
Now apply that to the final resolution, you cannot tell the difference between 2176x1224 and 2304x1296, yet you conclude that XSX has a higher resolution to PS5 because you were told so. Will you notice that PS5 has an average lower resolution compared to XSX during gameplay? No, unless someone pixel counts and tells you the difference. It doesn't mean the difference is not there. It's the same way XSX using VRS renders at half the resolution of the final output on internal surfaces giving PS5 an edge when compared side by side but when you are just playing you wouldn't notice if you aren't looking for it. That is how our vision works, our vision is only sharp in a small 5% window which is why foveated rendering works so well.

I never used to notice these things, but I do now, I never use to notice frame drops but now I do. There are many people on this forum who do.


Exactly.
 

sendit

Member
It could be I/0 related as NXGamer IGN videos shows the load times and the PS5 loads in 8 seconds and and the Series' consoles are loading in the double digits.

Can we verify that CDProject is using the latest Xbox Tools?
 

Vergil1992

Member
Yes I noticed it. PS5 is loading almost 100% faster than XSX and about on par with high end PC. It's odd but it has nothing to do with framerate drops during gameplay with normal max 33ms frame-times.
The strange thing here is that on XSX it previously loaded almost as fast as on PS5/PC.

1.png



In XSX the loads have apparently increased. And I say "apparently", because my results since I hit "load game" are not the same as those of IGN or elanalistabits. It just took me about 10 seconds, approximately. If someone can test and contrast their results it would be interesting.
In any case, I still think that there have not been many changes in the game at the technical level (VRS was on Xbox BEFORE the patch, the "drama" now is because DF has captured a frame). It seems that the PS5 version has been polished a little more. It is not the first game that updates take longer to stabilize the Xbox version.

But essentially, it is still the same. More solid performance on PS5, higher resolution in XSX. But in no case is it a drama. The game has framerate drops everywhere equally. And even if I don't say DF, I think it has an important stuttering due to asset transmission, at least in XSX. And the resolution difference is extremely small.
 
Last edited:
Top Bottom