• 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: Hands-On with Project Cars

op_ivy

Fallen Xbot (cannot continue gaining levels in this class)
Game runs at solid 60 fps with downgraded graphics.
GAF reaction: stupid downgrade, looks like shit, not buying this.

Game drops to 30 fps with 44 cars on the track in rainy weather.
GAF reaction: unacceptable, runs like shit, not buying this.

who is this gaf guy? i dont agree with him
 
So long for those "dozens of cars with close enough graphics at 60fps" shitposts on Driveclub threads.

Curious to see how high the final build will go on PC (without brute-force stuff like DSR and SSAA)
 

Yiazmat

Member
Hopefully they'll have everything sorted out by launch.

Digital Foundry: Hands-On with Project Cars
UPDATE: Build not intended for technical analysis. Our apologies!
GuLLVda.gif
 

chadskin

Member
Wow... You'd think you'd verify the version of the game WITH the devs before posting the article? Especially when the devs stated their intended targets and the build didn't reach its goals.

That's not how it works. Pre-release, the publisher sends you a preview copy with a note attached of the things you can/can't do with it. There was clearly some miscommunication going on as to what EG/DF were allowed to do with the copy they received and that's ultimately more on Bamco than on DF. Unless, of course, Bamco outright said no to any DF analysis and they went ahead anyway which I find doubtful.
 

p3tran

Banned
Seriously, at least give the option

I am thinking maybe all the effect toggles menues that we learned of yesterday in the consoles,
maybe they are not only and strictly for style preference, but may affect a little performance too.

plus, maybe final version is indeed much improved.

we really dont know yet.

but at this point, I agree with both of you, if they cant have it run at 60 locked, either a ..deforestation of environments according per console to reach the goal, or an option to keep framerate locked at 30 are both preferable for me to a fluctuating game
 

VGA222

Banned
It's not DFs fault that they analysed an old build of the game. They were told that the build that they got the footage from was of submission quality.

Also it's very strange to see people upset that the game drops quite a bit below 60 when you turn the more demanding settings up. I wonder if people would have thought that this is well optimized if SMS capped the current gen version AI counts to 35 and disabled the advanced weather effects.
 

_machine

Member
I am thinking maybe all the effect toggles menues that we learned of yesterday in the consoles,
maybe they are not only and strictly for style preference, but may affect a little performance too.
It's not the reason:

They have been there for years, they were always planned to be in the console versions (they weren't in the performance category and had been designed ages ago) even when they had no idea what the performance levels would be. Again, they are post-process effects that make absolutely no difference on PC and there is no reason to expect any noticeable difference in framerate either.
 

dark10x

Digital Foundry pixel pusher
Well, I hope the reaction by SMS suggests that this build really WAS out of date and not representative of the near final game.
 

_machine

Member
Well, I hope the reaction by SMS suggests that this build really WAS out of date and not representative of the near final game.
It was, I remember seeing the old UI in the videos and that was before some really major improvements to the game. I've had over 20% performance increase and the perceived smoothness has been much better since February (and looking at the date of my posts on WMD it's before that console build with the old UI).

EDIT: looking at the pictures not sure about the UI, but Andy seemed sure on the forums since apparently no build after the camera render update had been delivered to anyone.
 

prag16

Banned
I guess that Wii U version will sacrifice the amount of cars racing to preserve frame rate.

720p30 target with stripped down effects and less cars. And it'll still drop into low 20s in some situations. At least that's what I'd expect.
 

Javin98

Banned
Well, I hope the reaction by SMS suggests that this build really WAS out of date and not representative of the near final game.
Exactly what I was thinking earlier. Hopefully it runs much better than this and their reaction wasn't for damage control.
 
There's new information regarding this article that I'm not sure I can post, once I get an all clear I will edit this post. (Link for members)

Edit: The render team over at the WMD forums just posted a few comments on the article:

The preview was based on an older build of the game and several of its observations are unfortunately either wrong or due to bugs or items which have subsequently been addressed by the dev team during the finalling process. Notable:

- The level of Anisotropic filtering was increased from 4x to 8x on both consoles, significantly improving the general image quality and sharpness of the road.
- Both consoles are set to use motion-blur at the equivalent of the PC medium setting – the differences seen on PS4 were due to bugs which were transient with on-going optimisation and not because PS4 uses object based motion-blur.
- The shadow differences highlighted were largely down to the various slope-scale DX11 issues we had at this time, like many WMD users had reported on the PC builds.
- The build reviewed was also lacking the render bridge camera work, which as you all know significantly improved frame pacing and the general feeling of smoothness especially when cornering.

In the article it is mentioned that Project Cars on Xbox One is using post-process anti-aliasing which is equivalent to the PC's higher FXAA settings. This is incorrect, the Xbox One version uses MSAA, or more precisely AMD's EQAA (Enhanced Quality Anti Aliasing) with 8 fragments and 4 samples, which equivalent to MSAA 4x on PC.
 

dark10x

Digital Foundry pixel pusher
THIS SOOO MUCH THIS
I rather them wait for the day it comes out also so it doesn't sway people waayyy before the game launches and potentially giving the game an actual chance.
Wait wait wait, you're saying that developers should be excused for day 1 patches now?

I'm starting to get sick of this, to be honest.

As someone who prefers buying physical, I find this trend of broken games on launch to be unacceptable. It's happening far too often to the point.

I don't think anyone should ever defend that practice. I feel for the dev teams put in those situations, of course, but publishers (and anyone else responsible for this behavior) need to be called out.

When I covered The Evil Within last year that's why I went back and did a piece on the unpatched version to highlight just how broken the retail version was. The patched version is still awful but the original version they shipped? Basically unplayable - bad N64 game level of performance, really.
 

artsi

Member
Well, I hope the reaction by SMS suggests that this build really WAS out of date and not representative of the near final game.

Hopefully so, for their sake. If the final release has the same results then the angry pitchforks will turn from DF to SMS.
 

Three

Member
Well, I hope the reaction by SMS suggests that this build really WAS out of date and not representative of the near final game.

Any idea why the video was removed? Was it by request of the publisher?
Just seems kind of strange to keep an article that refers to the performance in the now removed video and keep the article intact.

Hopefully so, for their sake. If the final release has the same results then the angry pitchforks will turn from DF to SMS.

I don't get why anyone would point any blame towards DF to begin with.
 

_machine

Member
The render team over at the WMD forums just posted a few comments on the article:
The preview was based on an older build of the game and several of its observations are unfortunately either wrong or due to bugs or items which have subsequently been addressed by the dev team during the finalling process. Notable:

- The level of Anisotropic filtering was increased from 4x to 8x on both consoles, significantly improving the general image quality and sharpness of the road.
- Both consoles are set to use motion-blur at the equivalent of the PC medium setting – the differences seen on PS4 were due to bugs which were transient with on-going optimisation and not because PS4 uses object based motion-blur.
- The shadow differences highlighted were largely down to the various slope-scale DX11 issues we had at this time, like many WMD users had reported on the PC builds.
- The build reviewed was also lacking the render bridge camera work, which as you all know significantly improved frame pacing and the general feeling of smoothness especially when cornering.

In the article it is mentioned that Project Cars on Xbox One is using post-process anti-aliasing which is equivalent to the PC's higher FXAA settings. This is incorrect, the Xbox One version uses MSAA, or more precisely AMD's EQAA (Enhanced Quality Anti Aliasing) with 8 fragments and 4 samples, which equivalent to MSAA 4x on PC.

EDIT: wrong person
 

mrklaw

MrArseFace
Namco for making stupid claims to DF and allowing them access to the build in the first place.

DF for being so slow in updating the article.

By 'so slow' you mean updating within about 6 hours? That seems pretty good to me to be made aware of it, contact the publisher and developer, pull the videos and update the article.
 

ironcreed

Banned
Yikes. And that's the 'submission quality'.

Yeah, this is going to be a nice game to look at (other than the tearing), but I am doubting it will touch playing Forza or GT for me. I'll check it out on the cheap and I bet I will not have to wait long.
 

_machine

Member
Yeah, this is going to be a nice game to look at (other than the tearing), but I am doubting it will touch playing Forza or GT for me. I'll check it out on the cheap and I bet I will not have to wait long.
It isn't submission quality, see posts above.
 

dark10x

Digital Foundry pixel pusher
In the article it is mentioned that Project Cars on Xbox One is using post-process anti-aliasing which is equivalent to the PC's higher FXAA settings. This is incorrect, the Xbox One version uses MSAA, or more precisely AMD's EQAA (Enhanced Quality Anti Aliasing) with 8 fragments and 4 samples, which equivalent to MSAA 4x on PC.
If that's true then, yikes, seems like a bad choice.

I wonder how demanding it is? One would think, if performance was an issue, you'd stick with FXAA or SMAA. That is, unless they can deliver EQAA with proper performance in the end.

I actually wish we'd see more games take a dynamic approach to AA like Resident Evil 5 did back on 360. It could adjust between 4x MSAA to no AA depending on the load. Worked brilliantly.
 

artsi

Member
- The build reviewed was also lacking the render bridge camera work, which as you all know significantly improved frame pacing and the general feeling of smoothness especially when cornering.

That feeling of higher FPS.

So can that kind of huge differences in FPS be caused by incorrect frame pacing?

I don't get why anyone would point any blame towards DF to begin with.

Well I agree with that if they got incorrect info from the publisher.
 

Three

Member
Yeah, this is going to be a nice game to look at (other than the tearing), but I am doubting it will touch playing Forza or GT for me. I'll check it out on the cheap and I bet I will not have to wait long.

Can I ask why? It's not like GT6 had a consistent framerate.
 

Javin98

Banned
The render team over at the WMD forums just posted a few comments on the article:


That was from Andy over on WMD.
Nice! So AF was bumped from 4× to 8×. That's pretty good if you ask me. And the XB1 version is using an AA algorithm about comparable to 4× MSAA on PC? That's amazing for a 60FPS game. But strangely, the motion blur is a match for the PC's medium setting? Why? What are the other settings on? Also, it doesn't mention the AA on the PS4 version.
 

Noobcraft

Member
Ugh Xbox One version is definitely out for me. Drops from 60 to 29 fps? Granted there are over 40 cars on the track but it's still really disappointing to hear. I'll wait for performance patches before picking this up.
 
Wait wait wait, you're saying that developers should be excused for day 1 patches now?

I'm starting to get sick of this, to be honest.

I have no problems with a day 1 patch (unless its more than 20 GB, that is).

I do have a problem however with day 180 patches (connotation on patch, not content update).

Btw, as far as I know no day 1 patch was announced for PCars.
 

dark10x

Digital Foundry pixel pusher
I suspect it is submission quality but there is a day one patch.
Honestly, if that is the case, they deserve to be called out on it.

A day 1 patch is not a free pass, as far as I'm concerned. It only really flies for games that are entirely online. It's just bad practice for consumers.

I have no problems with a day 1 patch (unless its more than 20 GB, that is).
I have a HUGE problem with it myself. It's not acceptable especially as a collector and someone that enjoys returning to games from years gone by. When those patch servers go offline what then? We'll be left with a broken game?
 

Javin98

Banned
If that's true then, yikes, seems like a bad choice.

I wonder how demanding it is? One would think, if performance was an issue, you'd stick with FXAA or SMAA. That is, unless they can deliver EQAA with proper performance in the end.

I actually wish we'd see more games take a dynamic approach to AA like Resident Evil 5 did back on 360. It could adjust between 4x MSAA to no AA depending on the load. Worked brilliantly.
Didn't TLOU on PS3 also use some form of dynamic MLAA? I'm pretty sure I heard some people saying that. Anyway, I agree. 4× MSAA is way too demanding.
 

Kama_1082

Banned
Framerate *is* king in a racing sim. This is not about options, but about having the gameplay hold up so people can drive properly.

You keep making exaggerated alternatives. Nobody is asking them to do 1080p or 90fps or only have one car. But not letting the framerate drop into the 30's or even low 40's would significantly improve the actual act of driving, which is what this game is all about in the end.

In my experience, large grids in these sorts of games run into CPU limitations that cause these framerate drops. Cutting down the grid to like 24-30 could potentially increase performance dramatically without having to do anything extreme like you keep suggesting.
Driveclub is 30 fps and its fucking awesome.
 

ironcreed

Banned
Can I ask why? It's not like GT6 had a consistent framerate.

I am talking the overall experience as well. I have yet to find a racer that touches Forza and GT for me. While this looks gorgeous, I am just not confident that it will provide quite the same highs. But I will check it out at some point and see for myself.
 

Jimrpg

Member
That's not how it works. Pre-release, the publisher sends you a preview copy with a note attached of the things you can/can't do with it. There was clearly some miscommunication going on as to what EG/DF were allowed to do with the copy they received and that's ultimately more on Bamco than on DF. Unless, of course, Bamco outright said no to any DF analysis and they went ahead anyway which I find doubtful.

Oh bamco is at fault for giving an old version and telling DF they could analyse it. But DF should have double checked that the version was acceptable with the devs. It's clearly not as the dev posts showed.

The article I read stated the build was a near final build that was comparable to the final version which was obviously not the case.

The analysis is basically useless because nobody is going to play that. I doubt that DF want to post incorrect information but it honestly looks like they didn't do their due diligence on their end. In my line of work, if something looks off, we need to check why.
 

dark10x

Digital Foundry pixel pusher
Can I ask why? It's not like GT6 had a consistent framerate.
Well, to be fair, if you played at 1280x720 it really did have a very consistent frame-rate. Not 100% perfect but pretty darn close, from what I recall.

It was only in the 1440x1080 mode that it really fell flat and struggled heavily. That mode should have been optional rather than default.

The analysis is basically useless because nobody is going to play that.
You really think so? You don't find pre-release information interesting? I'd rather collect as much data as possible over a span of time. Looking at pre-release stuff gives an idea of how a game is shaping up and also gives us something to reflect on in the future post release. If more games were covered pre-release in this form perhaps some issues could have been avoided? Frame-pacing problems, lack of AF, etc. These are all things that devs ARE patching into their games that were broken at release. A pre-release analysis could have highlighted that and resulted in discussion around the net triggering the attention of the developers.

I understand WHY developers wouldn't want pre-release stuff out there but consumers? I find it odd that people are so actively protecting the interests of publishers and developers when we're being screwed more and more these days with broken launches.
 

fresquito

Member
Can someone post the SMS info in the OT. I know nobody reads, but still.

People are discussing basing their opinions on something that is factually wrong.
 
Top Bottom