• 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.

DF: Tales of Arise PS5 vs Series X/S Tech Breakdown.

onQ123

Member
Now you know full well the touryst launched a whole year later on PS5 with a completely rebuilt engine, made specifically for PS5, not a port of the PS4 title. Series X got an Xbox One X port on the older engine. Not the same circumstances for development, and despite the simplicity or unique nature of the touryst we know full well it isn't a more demanding game than a whole set titles that demonstrate a series x advantage. I would imagine if there was an advantage to be had, the PS5 would display it far more often in more demanding titles if it truly had 100% pixel pushing worth of an advantage over Series X. That's the difference between 8K and 6K. Do you actually believe that to be the case when weaker GPUs than what the Series X has can run the game's 8K at 60fps mode on PC? Clearly something doesn't add up about the touryst. And it's not like Series X has any performance dips, suggesting it has some headroom to go higher. Perhaps with more optimization it would also hit 8K. Not really a guess. Of course it would.

And as for ghostrunner, Doom Eternal runs its ray traced reflections at a higher resolution with respect to its native output, is just as fast, if not faster, with loads more action happening on screen at any given moment, but somehow manages to achieve way better, basically flawless, performance. So obviously we have an example of a game not too dissimilar in certain respects from ghostrunner, but doing a far better job on series x, suggesting the issue is with the dev's optimization on series x.

Isn't Doom 1800p with raytracing
not as technically minded as you or many in here. I would have thought the fatter gpu with more CUs would be stronger at raytracing? Something just doesn’t add up with ghost runner, which we know is an outsourced console port by a small indie studio with 30 employees and they still outsourced the console version.
It would be if it didn't run into other bottlenecks
 
Last edited:

Fredrik

Member
Some games will run better on PS5, some will run better on Series X, some will just run like shit on everything.
Exactly. And the real kicker - We can never know if the differences are related to different hardware or different code base. And we never know if the optimization is as good. It’s not like on PC where the same code runs on the same rig and all you swap out is the graphics card when you bench different graphics cards. All we know is that there can be differences, the real reason will always be unknown.
 
Last edited:
You mean simpler and a lot less demanding? It doesn't favor PS5. It was horribly optimized on Series X. You don't get that kind of performance variation with the hardware in the two boxes.

In this game, however, the differences align more with what you would expect of the hardware specifications. That said, to be fair, they should be using dynamic resolution, or they should at least run at a lower resolution on PS5 to guarantee it can maintain its performance better. Higher resolutions, especially native 4K is always going to favor Series X with proper optimization because it's got a beefier GPU by quite a bit that is going to perform better at higher resolutions. PS5 draws equal in framerate or ever so slightly better with similar visual settings by running at a more manageable resolution befitting its GPU strength.

This game on the whole is doing quite a bit more than Ghostrunner is. It's got a whole lot more going on and not as straightforward a design approach. Everything Ghostrunner did as a game, including ray tracing 1/3rd reflections resolution of the game, is already done at a superior level in Doom Eternal, and Doom Eternal is just as fast paced, if not more so, with a lot more action going on at any given time, and yet performance was flawless on both consoles.

Ghostrunner's engine isn't id software's highly well optimized engine, so that could have been the culprit there, but I'm more willing to blame optimization failures. It stands as one of the more obvious examples of bad optimization on Xbox Series X this entire gen because even if the PS5 is to have an advantage - although the hardware differences suggest it should realistically never have any - only insufficient optimization on Series X could possibly result in the performance picture in that game. I don't doubt better optimization can be done on the PS5 version of this game without lowering resolution, but depending on their time, that or a dynamic res solution might be their best option.

But I reject the idea that the game setup of Ghostrunner favors PS5. Is not Doom Eternal doing nearly everything that game is doing and way more?
His entire theory is based on the fact that PS5 has higher pixel fillrate/ROP throughput and that this pixel fillrate allows for better performance at native 4K60 with RT. It's complete and utter nonsense because pixel fillrate hasn't been a bottleneck in games since maybe ten years ago.

The one exception I can think of is The Touryst that uses a deferred renderer, which means the game is ROP-bound and the PS5 has an advantage in it.
 

DenchDeckard

Moderated wildly
Isn't Doom 1800p with raytracing

It would be if it didn't run into other bottlenecks

sorry if you’ve already mentioned them, but what are the bottlenecks on series x that could cause such an issue with this Specific linear title? I thought both machines look extremely balanced, the Xbox seems to be a design masterpiece really when you consider everything, size, power draw, noise levels. It’s world class Imo.

I imagine it’s the split memory? A full fat 4k image with raytracing might be exceeding the v ram? Similar to an 8k tourist image?
 
Last edited:
sorry if you’ve already mentioned them, but what are the bottlenecks on series x that could cause such an issue with this Specific linear title? I thought both machines look extremely balanced, the Xbox seems to be a design masterpiece really when you consider everything, size, power draw, noise levels. It’s world class Imo.

I imagine it’s the split memory? A full fat 4k image with raytracing might be exceeding the v ram? Similar to an 8k tourist image?
We have Control as proof that this can't be true. Runs better in the unlocked raytraced photo mode. And Control has much more complex raytracing than Ghostrunner, where raytracing is super light.
 

DenchDeckard

Moderated wildly
We have Control as proof that this can't be true. Runs better in the unlocked raytraced photo mode. And Control has much more complex raytracing than Ghostrunner, where raytracing is super light.

yeah I was just thinking maybe the native 4k image buffer on ghost runner might be causing it. I have no idea haha. I do think an outsourced game from some tiny European studio has a higher chance of being optimised than a full sized Japanese dev studio though. Just my opinion of course.
 

Md Ray

Member
You pointing to the PS5's higher pixel fillrate due to its clock speed, correct?
It also has higher rasterization throughput:

8QXpMZC.jpg
 
His entire theory is based on the fact that PS5 has higher pixel fillrate/ROP throughput and that this pixel fillrate allows for better performance at native 4K60 with RT. It's complete and utter nonsense because pixel fillrate hasn't been a bottleneck in games since maybe ten years ago.

The one exception I can think of is The Touryst that uses a deferred renderer, which means the game is ROP-bound and the PS5 has an advantage in it.
This is why its hilarious to see people talk about fill rates like wtf is this 2004 lmao
 
He clearly shows that there are drops in the framerate mode on PS5 whilst Series X is a locked 60fps.
He then says 60fps is a "rarity" on PS5 in graphics mode whilst normal on Series X.
It's pretty obvious Series X performs better.
Sure it performs better on Xbox (and we already knew this from a previous analysis). But technically both games can hover in the 50s and drop in the 30s in quality mode.
That's not bad considering xsx and ps5 has pretty much identical resolution only a 10fps difference between them both.
Yes which makes a 20% framerate advantage for Xbox. Similar to others games with 20% resolution advantage. But for more precise numbers we'd need VGTech analysis as he usually uses like for like scenes.
 
Last edited:

DenchDeckard

Moderated wildly
Can’t believe there’s so much salt in this thread with people arguing over performance of old style games.

Honestly all of you should wait for actual next gen games to drop. Most if not all of these current games were designed to run on old architectures that won’t exploit newer efficiency tech.

We aren't going to see next gen only games for ages though :( everything that's released so far could run on anything with a big of work. Definitely on pc. This will remain for a while I think

Sadge
 

Loxus

Member
If you go by the different fps readings in the most demanding battle shown on the graphics mode for both systems, Series X is maintaining roughly a 46% framerate advantage. So a bit closer to about 18fps.

I got 56.2fps for Series X to 38.3 for PS5 in graphics mode. Rough averages based on the different fps readings during this battle. I also got the Series X's lowest fps counts of 42 and 43 fps factored in. Would have been a higher Series X average if I included the exploration bits, but I focused strictly on the most demanding battle I saw on both consoles, which was the one below. Series X looks like it can get away with being 60fps at native 4K, but PS5 at native 4K is better off being 30fps. Now unless they can optimize more for both to improve performance, the solution is simple. Lower graphics mode res a bit on Series X, so it stays 60fps more often, and do the same for PS5, but to a more serious degree.


q56YbUS.png
FjOz3nM.png
Yk9fYWd.png
DXbxhtk.png
IKCIc70.png
2NDBS4t.png
umpzmtu.png
ldXmn7m.png
EFXo4IJ.png
rJKsICA.png
vGFBVVm.png
vmd80Td.png
Dude, stop with the misinformation.
@9m19s mark
Ezq2qTE.png

82m0GIM.png


I don't know why Digital Foundry failed to mention this, but it's kinda messed up.
 

Riky

$MSFT
Dude, stop with the misinformation.
@9m19s mark
Ezq2qTE.png

82m0GIM.png


I don't know why Digital Foundry failed to mention this, but it's kinda messed up.
Why you cut the framerate counter from the top of the image?

Show us the framerate counter in the 30's please.
 
Last edited:

Loxus

Member
But you never questioned DF’s results in the thread where PS5 was winning, and you had a lot of posts where you could have done so. You just assumed it was correct because it was what you wanted to hear.
There was no reason too question them.

But here you can clearly see DF messed up.
All I'm trying to do is figure out why they only mention PS5 drops in the 30s, but their own frame graph shows something different.

You see it also @9m19s (sorry couldn't get the full screen shot)
Ezq2qTE.png


Why don't you also question this?
This isn't even the same scenario as Ghost Runner.
This isn't even about favouritism, but DF credibility.

I wish the Mods would chime in on this.
 

Riky

$MSFT
With no side by side comparisons, I can't agree with you on this one.
I don't think it stopped at 38fps as the line points downwards.
Show us the screenshot where the framerate counter is in the 30's, then show us one where it hits as low as 32 like the PS5 screens quoted do.
 

Riky

$MSFT
Are you for real?
WTF, I'm done with this shit.
Clearly you guys have an agenda.
I love GAF and technical talks, so I'm not about to get myself banned.
You all can continue being blind console warriors.

th4qS0A.png

It's a simple request, just show it. You're calling the credibility of DF into question yet have deliberately cut off the frame counter from the screenshot. You do know when they turn it on it starts from zero and goes up don't you?
Just show us the frame counter in the 30's like you claim, easy.
 

Loxus

Member
It's a simple request, just show it. You're calling the credibility of DF into question yet have deliberately cut off the frame counter from the screenshot. You do know when they turn it on it starts from zero and goes up don't you?
Just show us the frame counter in the 30's like you claim, easy.


Clear as fucking day.
AMJlJkq.png


If you can't pick it up, just pause the video @9m19s.

It don't make any sense explaining anything to some of you all anymore, just full of console wars.
 

Riky

$MSFT


Clear as fucking day.
AMJlJkq.png


If you can't pick it up, just pause the video @9m19s.

It don't make any sense explaining anything to some of you all anymore, just full of console wars.


That doesn't show the frame counter, that's at the top right.
Show us the full screenshot with the frame counter in the top right in the thirties, like this
BieREEA.jpg
 
Last edited:
There is nothing deep about this if you can clearly see XBSX also drops in the 30s, but DF never mentioned it.
I think it's just due to rounding. Footage shows ps5 at around 31 and XSX at around 38. I don't think it's that big of a deal to call one 30 and the other 40. 7fps is around 20% difference so those lows make some sense for this game judging by the rest of what we seen.
 

DenchDeckard

Moderated wildly
With no side by side comparisons, I can't agree with you on this one.
I don't think it stopped at 38fps as the line points downwards.

Bro, seriously. There's tons of footage with the ps5 struggling with 60 fps? Are you saying digital foundry are being selective with all their footage. Coz if you're on some conspiracy theory shit like that I'm out.

Isn't this video done by the same person who did ghost runner? Why wouldn't they cherry pick footage then?
 

Loxus

Member
That doesn't show the frame counter, that's at the top right.
Show us the full screenshot with the frame counter in the top right in the thirties, like this
BieREEA.jpg
Come on Riky, your more intelligent than this.

So your telling me right.
When you look at this image,
You can't see where XBSX drops into the 30s?
0PsBNWA.png

Also, I was on a mobile device that shows the notification bar when taking a screen shot, so I just cropped it out. That's why the top is missing.
 

SkylineRKR

Member
I platted the PS5 version and performance mode is fairly perfect. Quality mode however doesn't feel good on PS5 but certain graphical effects do stand out more. Might've gotten XSS version instead as 1440p quality mode retains the better effects and the framerate holds up well on Xbox consoles.

It might have to do with VRR which no-show has become a prank on PS5 and is one of the most baffling things this gen so far.
 
Last edited:

Loxus

Member
Bro, seriously. There's tons of footage with the ps5 struggling with 60 fps? Are you saying digital foundry are being selective with all their footage. Coz if you're on some conspiracy theory shit like that I'm out.

Isn't this video done by the same person who did ghost runner? Why wouldn't they cherry pick footage then?
0PsBNWA.png


So your telling me your not going to question why DF is not mentioning XBSX is also dropping into the 30s?

This is not the same as Ghost Running, if you don't understand this. It doesn't make sense explaining anything anymore.
 

Riky

$MSFT
Come on Riky, your more intelligent than this.

So your telling me right.
When you look at this image,
You can't see where XBSX drops into the 30s?
0PsBNWA.png

Also, I was on a mobile device that shows the notification bar when taking a screen shot, so I just cropped it out. That's why the top is missing.
You're original post is that you don't "see" the performance difference that everyone else in the thread has accepted.
That shot may be when the frame counter is turned on, if you watch a lot of these videos it goes from zero up to the framerate, or it could be for a frame or two the Series X hits 39fps. However the PS5 hits 32fps as we see in the video and is shown by me.
So regardless the performance difference that you can't "see" is clearly there.
They also show you a run of footage in that boss battle and clearly the Series X holds a better framerate, as well as performance mode where both the Series X and S hold 60fps and he states he found no drops but the PS5 does have drops that are shown.
The facts are all there in the video, just accept it.
 
Come on Riky, your more intelligent than this.

So your telling me right.
When you look at this image,
You can't see where XBSX drops into the 30s?
0PsBNWA.png

Also, I was on a mobile device that shows the notification bar when taking a screen shot, so I just cropped it out. That's why the top is missing.
There is another part where it drops in the low 30s in a cutscene (I think), maybe something like 33-34fps (so about the PS5 min which is 32fps). The game runs on average better on XSX but it can also drop very low (30s) in both cutscenes and gameplay on XSX. To compare both we'd need averages using like for like scenes as DF comparison was a bit weak. Also it's clear DF seemingly avoided showing 30s drop (on the fps counter) happening on XSX while they were very keen to show the same drops happening on PS5.

HfzKCfu.png
 
Last edited:

Loxus

Member
You're original post is that you don't "see" the performance difference that everyone else in the thread has accepted.
That shot may be when the frame counter is turned on, if you watch a lot of these videos it goes from zero up to the framerate, or it could be for a frame or two the Series X hits 39fps. However the PS5 hits 32fps as we see in the video and is shown by me.
So regardless the performance difference that you can't "see" is clearly there.
They also show you a run of footage in that boss battle and clearly the Series X holds a better framerate, as well as performance mode where both the Series X and S hold 60fps and he states he found no drops but the PS5 does have drops that are shown.
The facts are all there in the video, just accept it.
Like I side, I'm done with this Riky.
If you don't understand what I'm talking about, let's just move on and end console wars together.
 

Loxus

Member
There are another part where it drops in the low 30s in a cutscene (I think), maybe something like 33-34fps (so about the PS5 min which is 32fps). The game runs on average better on XSX but it can also drop very low (30s) in both cutscenes and gameplay on XSX. To compare both we'd need averages using like for like scenes as DF comparison was a bit weak. Also it's clear DF seemingly avoided showing 30s drop (on the fps counter) happening on XSX while they were very keen to show the same drops happening on PS5.

HfzKCfu.png
Thank you very much for this.
I was just wondering why DF never mentioned the drops into the 30s on XBSX.
 

DeepEnigma

Gold Member
That doesn't show the frame counter, that's at the top right.
Show us the full screenshot with the frame counter in the top right in the thirties, like this
BieREEA.jpg
There are another part where it drops in the low 30s in a cutscene (I think), maybe something like 33-34fps (so about the PS5 min which is 32fps). The game runs on average better on XSX but it can also drop very low (30s) in both cutscenes and gameplay on XSX. To compare both we'd need averages using like for like scenes as DF comparison was a bit weak. Also it's clear DF seemingly avoided showing 30s drop (on the fps counter) happening on XSX while they were very keen to show the same drops happening on PS5.

HfzKCfu.png
Poorly Optimized: The Thread
 
Last edited:

Riky

$MSFT
I can't tell if you are being serious... It clearly shows it drops to the high 30s. I thought it was a Tom video for a second.

I've just said maybe it does at its lowest hit 38/39 FPS for a brief moment, however the video clearly shows the same boss battle running around 32fps on the PS5, so the performance difference he can't "see" is clearly still evident.
 
There are another part where it drops in the low 30s in a cutscene (I think), maybe something like 33-34fps (so about the PS5 min which is 32fps). The game runs on average better on XSX but it can also drop very low (30s) in both cutscenes and gameplay on XSX. To compare both we'd need averages using like for like scenes as DF comparison was a bit weak. Also it's clear DF seemingly avoided showing 30s drop (on the fps counter) happening on XSX while they were very keen to show the same drops happening on PS5.

HfzKCfu.png
So they show that part for ps5?
 

Md Ray

Member
His entire theory is based on the fact that PS5 has higher pixel fillrate/ROP throughput and that this pixel fillrate allows for better performance at native 4K60 with RT. It's complete and utter nonsense because pixel fillrate hasn't been a bottleneck in games since maybe ten years ago.

The one exception I can think of is The Touryst that uses a deferred renderer, which means the game is ROP-bound and the PS5 has an advantage in it.
This is why its hilarious to see people talk about fill rates like wtf is this 2004 lmao
Pixel fillrate isn't the only thing that gets bigger. Higher clock speed also means higher cache bandwidths.

Games can be limited by L1, L2 throughput before they can be CU/SM bound. Here are two games I profiled on my PC. DOOM Eternal is predominantly CU/SM bound then memory bandwidth and then L1 and L2, which means the more SM/CU and memory bandwidth you throw at it the better it will perform. Death Stranding, on the other hand, is almost always L1 throughput bound across multiple frames even at higher res like 4K and less so by the SMs/CUs or other parts of the GPU. In scenarios like Death Stranding where it thrashes cache bandwidths, PS5, due to its higher clock speed therefore higher cache bandwidths/FLOP, will benefit more and have similar or better perf than XSX.

HbTt8y5.png


i8uBQER.png
 
Last edited:
So they show that part for ps5?
That's the whole point. There is no like for like comparison which could have being easily doable using cutscenes (even if it's less representative, it's at least a fair comparison). Which is what they mostly did in many previous PS5 / XSX videos as it's easy to see the gap between both machines and prove any statement.

Here they state PS5 often hovers at 50s and can drop to 30s while XSX runs often at 60fps. But guess what, XSX can also hover at 50s and drop down to 30s, so what they should have done is showing like for like moments to prove their statements (which most of us believe). The comparison side of this video has being clearly rushed and poorly done.

Md Ray Md Ray Very interesting post just above mine!
 
Last edited:

Mr Moose

Member
I've just said maybe it does at its lowest hit 38/39 FPS for a brief moment, however the video clearly shows the same boss battle running around 32fps on the PS5, so the performance difference he can't "see" is clearly still evident.
I think they are arguing that DF said something else, I couldn't find the website page about it.
 

Mr Moose

Member
The easiest comparison is the frame rate mode, since they report a locked 60fps on Series X and S and didn't find any drops, on PS5 they do,

E9LNqgZ.jpg
It runs better on Series consoles (or at least that's what it seems without a like for like comparison). The game doesn't look all that good, not sure why it would be demanding on any of them. A framerate mode you would hope would be at least close to locked.
 
Top Bottom