• 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: Project CARS Performance Analysis (PS4/XB1)

SkylineRKR

Member
And people wanted Evo to target 60fps with Driveclub.

I'm happy SMS did it. I can't play 30fps racers, I love DC but 30fps always fucks with my cornering anticipation. If is like GT5, with some dips during crowded scenes and weather effects.. i'm fine.

The ghosting sounds worrying though, but perhaps it can be ironed out.
 

Dargor

Member
It quite obviously ghosts though even at a flat 60. I am not sure why they thought it was worth the weird trade off in IQ.

Probably because ghosting is one of those issues that are hardly noticed by most people. And even when people notice them, the complaints are rather inexistent.
 

omonimo

Banned
Actually yes, I hadn't read any comments or the text of the DF article before watching the 1080P videos DF put up, and straight away (the first corner in fact) it was night and day comparing the image quality of the XB1 vs the PS4, with the XB1 looking significantly sharper despite a lower resolution. That's when I looked into why and discovered the reasons.
Incredibly sharper and incredibly jaggies. Aliasing in some edge it's really horrible but I imagine it's better for someone.
 

nikos

Member
Why does everyone ignore the fact that performance issues only occur in extreme cases, where in all probably you can easily avoid?

Do we know this for sure? We don't really have any other PS4 gameplay footage to go by, other than replays.

I can't see the ghosting being situational. Unless I'm not understanding correctly, it's always there. Screen tearing could also still be an issue, which is what is bothering me the most.

I'm still reserving judgement until I see additional actual gameplay footage.
 

omonimo

Banned
Do we know this for sure? We don't really have any other PS4 gameplay footage to go by, other than replays.

I can't see the ghosting being situational. Unless I'm not understanding correctly, it's always there. Screen tearing could also still be an issue, which is what is bothering me the most.

I'm still reserving judgement until I see additional actual gameplay footage.
That's what DF article said. It drops when the cars on screen are around 40 and with weather effects.
 

bombshell

Member
Been away for some hours. I thought my OP was pretty clear about what was stress-tests and what was the common performance you'll be experiencing in the game (mainly the first video).

In this video the performance is a nearly locked 60 fps on PS4 and slightly worse performance on XB1 albeit at a lower resolution.

So what's the verdict on this game..I want to pre order it but still conflicted as to what system to get it for

The verdict is that PS4 version has higher resolution and higher framerate.
 

Noobcraft

Member
So what's the verdict on this game..I want to pre order it but still conflicted as to what system to get it for

PS4:
+1080P
+More stable frame rate
-less AF
-Ghosting
+ Higher res shadows

Xbox One
-900P
-Less stable frame rate (especially when there are lots of alpha effects like in the rain)
+Better AF
+No Ghosting
-lower res shadows

Frame rate is pretty variable in both under stress. Go with whichever console has the friends you want to play with.
 

hesido

Member
- You stated that PS4 was using object based motion-blur when it does not - the motion-blur is the same between PS4/XB1.

OK.. I downloaded an Xbox 1 video from gamersyde.

The Xbox 1 has a nice motion blur. You should be proud because you said you implemented it.

The PS4 does not have, in any single frame, a proper motion blur. To you, it may seem strange, but I don't have to pause the game to tell this. But it would make my life so much easier if someone pointed to me a gamersyde video with the replay camera on both Xbox 1 and PS4. There's actually a very nice shot I shared but it's compressed youtube shot. Since replay camera follow the car, surroundings pass about at real high speeds and this makes it an easy case for demonstration.

PS4 on left, Xbox 1 on right.
jCDQor3.png
tgrSo4W.png


These are both when the car is turning. Does the PS4 apply any linear blur to the cones (barriers?)? No. They just look like sharply rendered plastic barriers, only doubled. Do the trees moving sideways have any blur (aside from ghosting) to them, or their edges are pretty much sharp? They are sharp. Do we see double motion blurred trees, if they were to have the same motion blur as Xbox One? No. We see double sharp trees.

Does the Xbox 1 have blur applied to the cones? Yes. Do the trees have sharp edges? No, because they have motion blur applied to them. The Xbox one shows a subtle yet nicely executed blur.

It's hard to make judgments from still shots of what is going on behind the scenes, but I haven't started out with still shots. Also, when you say implementation of blur is the same, I expect to see some evidence of it. So far, there is none. Xbox 1's is a still shot, PS4's is a still shot. They are on equal footing. PS4 does not display any linear blur.

How about someone drive the same car around a track with a PS4 and an Xbox One, upload that video in it replay mode? Because my spider senses are warning me that someone will complain they are not exactly the same shots. But I still challenge anyone to show me a linear motion blur on the PS4, be it in still shot, or a video.
 

system11

Member
Incredibly sharper and incredibly jaggies. Aliasing in some edge it's really horrible but I imagine it's better for someone.

Yeah, I don't really care about low AA, prefer it to be honest in a few cases. They look ugly on a static screen but when games move at 60fps I don't notice them at all. A few posts up someone mentioned a blur slider, if there's a 'zero' setting I'll pick this game up after all when the patch hits.

I could understand and ignore a bit on (for example) a car zooming past me, I don't like the effect but it would make some kind of sense. I noticed even the surrounding treeline moving into the screen was blurry straight away on the 480P footage. Upped it to 1080 hoping it was just due to YT and nope, still there.

If this doesn't bother people I'm happy for them. It's not a choice I'd ever make as a developer though because I've been gaming long enough to know that image sharpness in long gaming sessions is a really crucial factor, and indeed with fast moving games.

Yes I'm that guy in meetings at work who ends up fixing the projector focus during the coffee break because nobody else gets it right.
 
PS4:
+1080P
+More stable frame rate
-less AF
-Ghosting
+ Higher res shadows

Xbox One
-900P
-Less stable frame rate (especially when there are lots of alpha effects like in the rain)
+Better AF
+No Ghosting
-lower res shadows

Frame rate is pretty variable in both under stress. Go with whichever console has the friends you want to play with.

Been away for some hours. I thought my OP was pretty clear about what was stress-tests and what was the common performance you'll be experiencing in the game (mainly the first video).

In this video the performance is a nearly locked 60 fps on PS4 and slightly worse performance on XB1 albeit at a lower resolution.



The verdict is that PS4 version has higher resolution and higher framerate.


Well I dont know anyone that has the game or will be getting the game but from what I can see the PS4 version will be better..as far as the ghosting goes, is that issue resolved in a day one patch?
 

Noobcraft

Member
Well I dont know anyone that has the game or will be getting the game but from what I can see the PS4 version will be better..as far as the ghosting goes, is that issue resolved in a day one patch?
That's the question. I don't think they've mentioned fixing it, but if the demand is there it wouldn't surprise me. I don't know if it would make launch though.
 

Conduit

Banned
Well I dont know anyone that has the game or will be getting the game but from what I can see the PS4 version will be better..as far as the ghosting goes, is that issue resolved in a day one patch?

Not yet. But the dev said PS4 version will have toggle option for that.

SMS :

It's not a bug. It's a side effect of the way we do AA on PS4, and is arguably a good thing (hear me out!). It's there, and very noticeable from static images of non-direct feed footage especially, but also undeniably in the game itself (mainly when paused). I spotted it and reported it as soon as it got checked in long ago (so we didn't "miss it" as many would have you believe).

However, I also "spotted" that the PS4 version has really lovely image quality, a very smooth and slightly soft (not low res!), "non-gamey" look, which is very natural, and (imo!) rather fantastic, due to this very clever AA approach. It also greatly minimises the distant "shimmer" you often see in console racing games, which is a side effect of low quality AA on thin vertical objects such as fences.

So yes, the ghosting effect is there, if you look hard enough. The advantages far outweigh this small disadvantage however. Notwithstanding that, if customer feedback demands it, we have a plan to add a UI slider in a future update, which will enable you to tune out the ghosting, at the expense of more aliasing and distant shimmer.
 

bombshell

Member
Oh so its done on purpose? Why? And why doesnt the XBOX ONE version have it..

Its mentioned about that the PS4 version has "Less AI" what does that mean ?

Because it costs more than the AA being done in the XB1 version and that version, even at its lower resolution, already has 10-15 fps deficiencies in some cases if you watch the videos.

AF, not AI.

AF = Anisotropic Filtering, the detail of textures in the distance when viewed at an angle.
 
Between this and "The Witcher looks bad" hyperbole, I'm starting to really dislike where the zeitgeist is headed in my beloved hobby. :(

I'm not sure why people holding developers to performance standards is something to be lamented. Are half-functioning games now something to be desired?

I swear, the game could be hitting 10 frames per second for 90% of races on a clear day, and people would still be telling everybody that they should be happy that they have the game at all. Why people defend a game losing 50% of its performance at any point in time is beyond me. Couple this with the complete lack of content on the console versions unless you download a day one patch, and it's no wonder why we see so many publishers regularly raking gamers over the coals with horrendous business practices -- you can always find a contingent of people willing to accept anything that they're given if you just get them hyped enough for it.

If you can't deliver a particular feature or promise -- say, 40 cars on a single track on consoles -- then don't deliver it at all; this sort of half-assed, "well, something is better than nothing" approach is the reason why a series like Gran Turismo has been in the toilet for the last couple of installments. I'm sorry, but something is not always better than nothing; sometimes, nothing is actually preferable to something which is broken or does not work as intended.
 

Metfanant

Member
Oh so its done on purpose? Why? And why doesnt the XBOX ONE version have it..
Its mentioned about that the PS4 version has "Less AI" what does that mean ?
According to the SMS dude, different people coded the two versions (he worked on Xbone and PC and said he quite likes the PS4 approach)
 
Not yet. But the dev said PS4 version will have toggle option for that.

SMS :

It's not a bug. It's a side effect of the way we do AA on PS4, and is arguably a good thing (hear me out!). It's there, and very noticeable from static images of non-direct feed footage especially, but also undeniably in the game itself (mainly when paused). I spotted it and reported it as soon as it got checked in long ago (so we didn't "miss it" as many would have you believe).

However, I also "spotted" that the PS4 version has really lovely image quality, a very smooth and slightly soft (not low res!), "non-gamey" look, which is very natural, and (imo!) rather fantastic, due to this very clever AA approach. It also greatly minimises the distant "shimmer" you often see in console racing games, which is a side effect of low quality AA on thin vertical objects such as fences.

So yes, the ghosting effect is there, if you look hard enough. The advantages far outweigh this small disadvantage however. Notwithstanding that, if customer feedback demands it, we have a plan to add a UI slider in a future update, which will enable you to tune out the ghosting, at the expense of more aliasing and distant shimmer.

Because it costs more than the AA being done in the XB1 version and that version, even at its lower resolution, already has 10-15 fps deficiencies in some cases if you watch the videos.

AF, not AI.

AF = Anisotropic Filtering, the detail of textures in the distance when viewed at an angle.


Thanks!! Sorry for all the questions guess Im just really picky..heck I wouldnt know what those things mean if I saw them up close, just want the best version for the console..so I guess decision made then, PS4 version..that way I can pre order it and get the DLC included :) gamespot is still having their review in progress so maybe in a few hours?
 
This. They should have done whatever was necessary. Turn down the shadow quality, tone down rain effects, turn down number of cars on-track, reduce the poly-count on the cars...anything that needs to be done to get a stable framerate and to get rid of that awful screen-tearing, which for me is even worse than framerate drops.

I don't understand. If your view point is that they should have done anything to get a stable frame rate and you would be happy with anything they did to achieve it. . . .

Then why don't you just use the many options they've given you to reduce the number of cars and reduce or remove the rain?
I mean, isnt that what you are asking them to do?
 

bombshell

Member
Why doesn't the PC have that version of AA then?

I don't know, ask SMS :) The PC version probably already use even more expensive AA methods.

By my understanding, the PS4 in motion has less aliasing than the XB1 version because of this extra temporal AA pass. But for some reason DF doesn't comment on that. Hopefully they do in the face-off.

In screenshots it looks like shit though.
 

MaLDo

Member
OK.. I downloaded an Xbox 1 video from gamersyde.

The Xbox 1 has a nice motion blur. You should be proud because you said you implemented it.

The PS4 does not have, in any single frame, a proper motion blur. To you, it may seem strange, but I don't have to pause the game to tell this. But it would make my life so much easier if someone pointed to me a gamersyde video with the replay camera on both Xbox 1 and PS4. There's actually a very nice shot I shared but it's compressed youtube shot. Since replay camera follow the car, surroundings pass about at real high speeds and this makes it an easy case for demonstration.

PS4 on left, Xbox 1 on right.
jCDQor3.png
tgrSo4W.png


These are both when the car is turning. Does the PS4 apply any linear blur to the cones (barriers?)? No. They just look like sharply rendered plastic barriers, only doubled. Do the trees moving sideways have any blur (aside from ghosting) to them, or their edges are pretty much sharp? They are sharp. Do we see double motion blurred trees, if they were to have the same motion blur as Xbox One? No. We see double sharp trees.

Does the Xbox 1 have blur applied to the cones? Yes. Do the trees have sharp edges? No, because they have motion blur applied to them. The Xbox one shows a subtle yet nicely executed blur.

It's hard to make judgments from still shots of what is going on behind the scenes, but I haven't started out with still shots. Also, when you say implementation of blur is the same, I expect to see some evidence of it. So far, there is none. Xbox 1's is a still shot, PS4's is a still shot. They are on equal footing. PS4 does not display any linear blur.

How about someone drive the same car around a track with a PS4 and an Xbox One, upload that video in it replay mode? Because my spider senses are warning me that someone will complain they are not exactly the same shots. But I still challenge anyone to show me a linear motion blur on the PS4, be it in still shot, or a video.


Man, I have to admire your patience here. I know you're right too, but I gave up after the myopic comments. Damage control is on a new whole level nowadays.
 

OmegaDL50

Member
And people wanted Evo to target 60fps with Driveclub.

I'm happy SMS did it. I can't play 30fps racers, I love DC but 30fps always fucks with my cornering anticipation. If is like GT5, with some dips during crowded scenes and weather effects.. i'm fine.

The ghosting sounds worrying though, but perhaps it can be ironed out.

I love Driveclub but I doubt the engine Evo crafted could handle 30 cars on track with their weather system.

The game gets crazy enough as it is when you are doing those Thunder Storm Hypercar Races with a full pack of 12.

At 30 Cars it would be too much, TOO much.
 
I don't understand. If your view point is that they should have done anything to get a stable frame rate and you would be happy with anything they did to achieve it. . . .

Then why don't you just use the many options they've given you to reduce the number of cars and reduce or remove the rain?
I mean, isnt that what you are asking them to do?

No, I'm asking them to ship a product that doesn't dip and tear as excessively as it does. It's not my responsibility to make sure their game runs well...

And the first video in the article has no rain effects and only 20 cars on-track...yet it still dips, and screen-tears like a motherfucker. Not to mention tweaking certain post-process effects only gives you a ~2 FPS gain. A shame, but I know I simply won't be able to enjoy it like that. I really hope they bring out some performance tweaks via patches in the future, because I really want to play this.

And I'm no framerate/performance snob. I'm really not. I'm a console gamer after all...but for racers I need stable performance. All of those dips will be horrible for precise inputs, especially in a sim, and those tears will be distracting as fuck to someone who is as sensitive to them as I am.
 
Hilarious the outcry over the consoles performance that was stressed test at extreme levels.

Come on people get a grip, Game looks astounding and will play great on our measly £300 consoles.
 

Cuyejo

Member
OK.. I downloaded an Xbox 1 video from gamersyde.

The Xbox 1 has a nice motion blur. You should be proud because you said you implemented it.

The PS4 does not have, in any single frame, a proper motion blur. To you, it may seem strange, but I don't have to pause the game to tell this. But it would make my life so much easier if someone pointed to me a gamersyde video with the replay camera on both Xbox 1 and PS4. There's actually a very nice shot I shared but it's compressed youtube shot. Since replay camera follow the car, surroundings pass about at real high speeds and this makes it an easy case for demonstration.

PS4 on left, Xbox 1 on right.
jCDQor3.png
tgrSo4W.png


These are both when the car is turning. Does the PS4 apply any linear blur to the cones (barriers?)? No. They just look like sharply rendered plastic barriers, only doubled. Do the trees moving sideways have any blur (aside from ghosting) to them, or their edges are pretty much sharp? They are sharp. Do we see double motion blurred trees, if they were to have the same motion blur as Xbox One? No. We see double sharp trees.

Does the Xbox 1 have blur applied to the cones? Yes. Do the trees have sharp edges? No, because they have motion blur applied to them. The Xbox one shows a subtle yet nicely executed blur.

It's hard to make judgments from still shots of what is going on behind the scenes, but I haven't started out with still shots. Also, when you say implementation of blur is the same, I expect to see some evidence of it. So far, there is none. Xbox 1's is a still shot, PS4's is a still shot. They are on equal footing. PS4 does not display any linear blur.

How about someone drive the same car around a track with a PS4 and an Xbox One, upload that video in it replay mode? Because my spider senses are warning me that someone will complain they are not exactly the same shots. But I still challenge anyone to show me a linear motion blur on the PS4, be it in still shot, or a video.

I know you are completely right, just showing some support.

It looks AWFUL on PS4.
 

Hendrick's

If only my penis was as big as my GamerScore!
I swear, the game could be hitting 10 frames per second for 90% of races of races on a clear day, and people would still be telling everybody that they should be happy that they have the game at all.
I swear the game is hitting 60 frames per second for 90% of the races and people are calling it a disaster.
 
Personally I think it would be prudent if SMS gave players a toggle to lock 30fps. That means those who are bothered by heavily fluctuating frame rates at least have the option to solve their problems on larger grids and in the rain.

It won't help those who adopt a 'constant 60 fps or die' extremist attitude but having a lock would satisfy those who just dont like the heavy fluctuations and tearing. It sounds like smaller grids suffer a lot less anyway.

I saw this on PS Live (or whatever it's called) last night. There were quite a few streams actually.
I thought it looked pretty good, the audio seemed impressive too. Probably good enough to appeal broadly to people who don't get quite as exciteable about fluctuating framerates as some of the sim/PC crowd on here. I don't feel there has been that much of a buzz about this game though, so who knows?
 

wachie

Member
Looks like the devs pushed both consoles and got their best out of it, within the extended release timeline.

The Xbox One version is borderline sad however.
 

Hedrush

Member
Hilarious the outcry over the consoles performance that was stressed test at extreme levels.

Come on people get a grip, Game looks astounding and will play great on our measly £300 consoles.

This.

Some people are going to make themselves ill with the amount of time they're spending looking at screenshots and slowed down youtube footage just to try and convince people there's a problem with the game when that is quite clearly not the case. Also the game is being praised highly by the gaming media as shown in their reviews.
 

SMSRenderTeam

Neo Member
OK.. I downloaded an Xbox 1 video from gamersyde.

The Xbox 1 has a nice motion blur. You should be proud because you said you implemented it.

The PS4 does not have, in any single frame, a proper motion blur. To you, it may seem strange, but I don't have to pause the game to tell this. But it would make my life so much easier if someone pointed to me a gamersyde video with the replay camera on both Xbox 1 and PS4.

PS4 on left, Xbox 1 on right.
jCDQor3.png
tgrSo4W.png


These are both when the car is turning. Does the PS4 apply any linear blur to the cones (barriers?)? No. They just look like sharply rendered plastic barriers, only doubled. Do the trees moving sideways have any blur (aside from ghosting) to them, or their edges are pretty much sharp? They are sharp. Do we see double motion blurred trees, if they were to have the same motion blur as Xbox One? No. We see double sharp trees.

Does the Xbox 1 have blur applied to the cones? Yes. Do the trees have sharp edges? No, because they have motion blur applied to them. The Xbox one shows a subtle yet nicely executed blur.

It's hard to make judgments from still shots of what is going on behind the scenes, but I haven't started out with still shots. Also, when you say implementation of blur is the same, I expect to see some evidence of it. So far, there is none. Xbox 1's is a still shot, PS4's is a still shot. They are on equal footing. PS4 does not display any linear blur.

How about someone drive the same car around a track with a PS4 and an Xbox One, upload that video in it replay mode? Because my spider senses are warning me that someone will complain they are not exactly the same shots. But I still challenge anyone to show me a linear motion blur on the PS4, be it in still shot, or a video.



Ok,

Here is a capture from PS4 I just made without the temporal AA step applied. (Ignore the yellow text from the development build)





hHrgtic.jpg




You can clearly see our standard PC/XB1 motion-blur shader visible - it's exactly the same runtime code, shader code with same number of motion samples and the same full HDR texture resolution!

The road has that "lovely" motion-blur speed look that you get with this method using multiple samples.

Now you might note that the armco's and some of the scenery are less blurred than some of the images posted around here. This is because the motion *velocity* shader parameter is reduced on PS4 - the temporal AA post process (through persistence of vision) also adds to the impression of motion blur - so if we kept the motion velocity scale the same as XB1/PC there would then be to much MB.

If we implemented an option for allowing the temporal AA effect to have sort some of slider as has been mentioned then the traditional Motion Blur velocity scale would be change inversely to the scale of the temporal AA effect. The motion velocity scale could be also be on a seperate slider on all platforms - since it's the exact same code.

I hope this explanation gives the community here some further clarity on why isolated image analysis isn't always the full picture........ (badum!)
 
I swear the game is hitting 60 frames per second for 90% of the races and people are calling it a disaster.

DF and thus this threads are for that, people make of this what they want. In my case I can neglect the issues as they are under stress, the game works fine 90% of the time. I also can lower the # of cars, no issue with that.

+1 to SMS, Congrats on a great PS4 release! I have faith they will ironed out the other issues or add more options. This game has great promise to continue delivering content and become huge.
 

Skyzard

Banned
There isn't going to be an fps lock on the consoles - they just answered the question on twitch. Apparently it would cause tearing and would be worse. And it runs at 60 the majority of the time he added.

They also don't discuss these tests from random sites, apparently.
 
No, I'm asking them to ship a product that doesn't dip and tear as excessively as it does. It's not my responsibility to make sure their game runs well...

Nooooo. I wasn't implying you were being a snob at all. Lol

In your other post you said they should have cut the number of cars (or polys for cars - same effect) and cut back on rain and stuff.

I can't be arsed with PC gaming at all (Elite Dangerous is the only thing to even remotely tempt me otherwise in the past 10 years) so I'm all for not having to tweak games settings yourself to alter performance. I agree that devs should, by and large, have the optimal settings in console games with as few toggles and sliders as possible.

I was pointing out that this dev has decided to include options that can improve performance similar to the suggestions you made.
 

danowat

Banned
Ok,

Here is a capture from PS4 I just made before the temporal AA step is applied. (Ignore the yellow text from the development build)





hHrgtic.jpg




You can clearly see our standard PC/XB1 motion-blur shader visible - it's exactly the same runtime code, shader code with same number of motion samples and the same full HDR texture resolution!

The road has that "lovely" motion-blur speed look that you get with this method using multiple samples.

Now you might note that the armco's and some of the scenery are less blurred than some of the images posted around here. This is because the motion *velocity* shader parameter is reduced on PS4 - the temporal AA post process (through persistence of vision) also adds to the impression of motion blur - so if we kept the motion velocity scale the same as XB1/PC there would then be to much MB.

If we implemented an option for allowing the temporal AA effect to have sort some of slider as has been mentioned then the traditional Motion Blur velocity scale would be change inversely to the scale of the temporal AA effect. The motion velocity scale could be also be on a seperate slider on all platforms - since it's the exact same code.

I hope this explanation gives the community here some further clarity on why isolated image analysis isn't always the full picture........ (badum!)

So are the render team looking into ways of making the double imaging of the temporal AA less apparent?,or is it a case of, that's what it is, and that's what it's going to be?

In motion, it looks fine, but in static shots, it looks abhorrent, it's a shame to see something with such a nice IQ be represented by static screenshots that look so nasty.
 
Top Bottom