• 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 Face-Off

burgerdog

Member
Yikes...

grassgateaduvr.jpg


Grassgate 2.0?

Forget the grass, the lack of AF is disgusting.

edit
Talking about the ps4, although it isn't good on xb1 either.
 

Aceofspades

Banned
Most likely because they couldn't. blending two frames together is ROP dependent and the XB1 may struggle with it.

I don't know if anyone has noticed but there is something wrong with specular reflections on the XB1 version. Like it is missing tone mapping or something.

mXkHMh.png

JGMfma.png

X1 aliasing is seriously bad o_O
 
Me thinks Dark10x & blimblim should have conversation about hosting uncompressed video face-offs on gamersyde, instead youtube's horrible compression.
 

Tenebrous

Member
Forget the grass, the lack of AF is disgusting.

edit
Talking about the ps4, although it isn't good on xb1 either.

It's pretty terrible... I know it's just one part of a much more complicated image, but I can't imagine that certain part of the picture is any worse on the same track on GT5/GT6.
 

hesido

Member
Most likely because they couldn't. blending two frames together is ROP dependent and the XB1 may struggle with it.

But the Xbox One has high quality motion blur, which has several samples. You no longer would need as many samples for it, since PS4 version seems to be set to 2-3% motion velocity compared to Xbox One and PC anyway.. Maybe an eSRAM limitation? They'd be much better off getting rid of previous buffers as soon as possible, no?
 

Marmelade

Member
I installed windows 10 TP yesterday and the game runs much better than on seven (r9 290 tri-x)
Really sounds like a driver issue on AMD's part.
 

Stacey

Banned
You know what? If the Wii U comes @ 720p/LOCKED 30fps, I'd def be more interested than the PS4 version, especially if something is done at least a little interesting with the gamepad screen.

A rear view mirror without the entire track disappearing in front of your eyes?
 
Pretty substantial win for the PS4.

I have a 6850 and i5, I was debating if I do buy the game which version to get but I'd probably go with the PS4 version in this case since AMD sucks balls and PS4s gpu is better than mine.
 

thelastword

Banned
Forget the grass, the lack of AF is disgusting.

edit
Talking about the ps4, although it isn't good on xb1 either.
I don't think this is a lack of AF but moreso extra blur induced by temporal AA, it appears that both console versions are using the same level of AF. Of course when you're racing, the blur is even less apparent.

mXkHMh.png


What is really apparent on the starting grid and in motion though is this woeful aliasing that the XBONE has, that's on-top of the sub-native resolution already at play.
 

Three

Member
But the Xbox One has high quality motion blur, which has several samples. You no longer would need as many samples for it, since PS4 version seems to be set to 2-3% motion velocity compared to Xbox One and PC anyway.. Maybe an eSRAM limitation? They'd be much better off getting rid of previous buffers as soon as possible, no?

Well SMS kept the same number of motion samples on the PS4 even with the temporal AA so I'm not sure why or if that's true if it is then they really should have done it. I don't know what it would look like but if they could have reduced it without a hit in IQ I trust they would have since they could have improved other areas like foliage on consoles, or the bad shadow maps with the freed resources. If they were not ROP bound, and I suspect they were, one aspect they really needed to improve was that. I don't know why they didn't remove the blur and and use Poisson Sampling for the shadows but maybe the sense of speed it gives is more important than seeing the texels in the shadow which really needed improving. I assume they made the right call for each platform in any case.

WUk0n4.png
 

Cidd

Member
Well SMS kept the same number of motion samples on the PS4 even with the temporal AA so I'm not sure why or if that's true if it is then they really should have done it. I don't know what it would look like but if they could have reduced it without a hit in IQ I trust they would have since they could have improved other areas like foliage on consoles, or the bad shadow maps with the freed resources. If they were not ROP, and I suspect they were, one aspect they really needed to improve was that. I'm don't know why they didn't remove the blur and and use Poisson Sampling for the shadows but maybe the sense of speed it gives is more important than seeing the texels in the shadow which really needed improving. I assume they made the right call for each platform in any case.

WUk0n4.png

Wow that's Awful wtf.
 
Most likely because they couldn't. blending two frames together is ROP dependent and the XB1 may struggle with it.

I don't know if anyone has noticed but there is something wrong with specular reflections on the XB1 version. Like it is missing tone mapping or something.

mXkHMh.png

JGMfma.png
Jesus, the Xbox One version looks pretty rough. It's probably not that apparent in motion though, I'd imagine.
 
Well SMS kept the same number of motion samples on the PS4 even with the temporal AA so I'm not sure why or if that's true if it is then they really should have done it. I don't know what it would look like but if they could have reduced it without a hit in IQ I trust they would have since they could have improved other areas like foliage on consoles, or the bad shadow maps with the freed resources. If they were not ROP bound, and I suspect they were, one aspect they really needed to improve was that. I don't know why they didn't remove the blur and and use Poisson Sampling for the shadows but maybe the sense of speed it gives is more important than seeing the texels in the shadow which really needed improving. I assume they made the right call for each platform in any case.

WUk0n4.png
Wow, the Figth is going to be between Xb1 and Wii U. That looks really bad.
 

Fredrik

Member
Yikes...

grassgateaduvr.jpg


Grassgate 2.0?
o_O
There is no nice way to say this, but these consoles are clearly too weak to be around for 3-4 more years, this faceoff is almost embarrassing to look at, neither of the console versions are even close to be okay imo - framedrops, tearing, lost details, ghosting, etc - What a mess :(
 
I wonder how many people will complain that this game uses the hardware inefficiently or something, rather than the fact that this game is doing some pretty crazy effects and simulation work.

The way reflections are done for example... is just heavy. Sorry console :/
 
As it stands, there is a trend of AMD cards falling short of their Nvidia's equivalents; even the top-tier R9 290X reportedly struggling against a lowly GTX 760 at 1080p and high settings
Damn this sucks for AMD GPU users. Will PS4 and Xbox One get performance hits since both consoles also use AMD GPU.
 

Stacey

Banned
I wonder how many people will complain that this game uses the hardware inefficiently or something, rather than the fact that this game is doing some pretty crazy effects and simulation work.

The way reflections are done for example... is just heavy. Sorry console :/

Ummm, the reflections on consoles are disgusting, cars aren't reflected in the wet road and the environment reflections are of such a low quality they're literally made from blocks that flicker as you approach them.

The car reflections simply aren't worth mentioning.......
 

Three

Member
I wonder how many people will complain that this game uses the hardware inefficiently or something, rather than the fact that this game is doing some pretty crazy effects and simulation work.

The way reflections are done for example... is just heavy. Sorry console :/

I don't think anybody has complained about the game using hardware inefficiently, at least not yet, but I'm curious to know what method they use for reflections. What do they use? I remember seeing a gif where the reflection wasn't actually correct even.

meMQRb.gif


is it a PC specific method?
 
Slightly Mad Studios' racer leaves a high watermark for Polyphony Digital and Turn 10 to match in the coming years


I still think DriveClub is the better looking game with its dynamic lighting and all
 

Shin-Ra

Junior Member
I think it uses the hardware inefficiently... compared to what the first party devs will achieve, but I wouldn't complain because I don't expect better from this developer.
 

JAYSIMPLE

Banned
Xbox looks so jaggi. Ps4 is ruff compared to pc too. Does the pc version dominate the consoles on respectable hardware too?
 
Ummm, the reflections on consoles are disgusting, cars aren't reflected in the wet road and the environment reflections are of such a low quality they're literally made from blocks that flicker as you approach them.

The car reflections simply aren't worth mentioning.......
They are primarily tied to the CPU rather than the GPU. I think that is why it makes sense that they are of such low quality on consoles.
I don't think anybody has complained about the game using hardware inefficiently, at least not yet, but I'm curious to know what method they use for reflections. What do they use? I remember seeing a gif where the reflection wasn't actually correct even.

meMQRb.gif


is it a PC specific method?
I am partially trying to pre-empt that discussion (which infects DF threds like a cancer IMO).

This game is using an old school method of doing reflections (not SSR). It leads to better reflections for things not on screen, but it has other limitations.

It is one of the better ways to do it for a racer but I personally would preffer a combo of SSR and this way. So you have more accurate higher res planar reflections, but still high quality off screen reflections.
 

nib95

Banned
o_O
There is no nice way to say this, but these consoles are clearly too weak to be around for 3-4 more years, this faceoff is almost embarrassing to look at, neither of the console versions are even close to be okay imo - framedrops, tearing, lost details, ghosting, etc - What a mess :(

Lol, maybe you should look at the rest of the comparison images instead of this cherry picked worst case example.

I am shocked at how little difference there is between the console and PC versions. In most of the comparison images, there barely is a difference.
 

hesido

Member
Well SMS kept the same number of motion samples on the PS4 even with the temporal AA so I'm not sure why or if that's true if it is then they really should have done it. I don't know what it would look like but if they could have reduced it without a hit in IQ I trust they would have since they could have improved other areas like foliage on consoles, or the bad shadow maps with the freed resources. If they were not ROP bound, and I suspect they were, one aspect they really needed to improve was that. I don't know why they didn't remove the blur and and use Poisson Sampling for the shadows but maybe the sense of speed it gives is more important than seeing the texels in the shadow which really needed improving. I assume they made the right call for each platform in any case.

WUk0n4.png

Well, they seem to have reduced the motion blur velocity so much, that its effect is literally imperceptible. If they are using the same number of samples as the other platforms, it's a waste.
JWuwAye.png

You can even see the road texture pattern on the PS4. The blurring on the car is caused exclusively by the temporal AA. The motion blur does not touch where the cars are (even when there are objects passing in front of them)

The pre-temporal AA shot shown by the team to prove there's motion blur was a specific case when the camera was moving at >150km/h and 30cm above ground, and only the road seemed to get that blur. I'd really like to see the blur on other platforms on similar angles, because it's reduced so much that the road-side objects receives almost no motion blur.

Also, properly set motion blur conveys speed much better than ghosting can.
 
Top Bottom