chromatic9
Member
benzy said its time trial, in case he doesn't replay himself, his pics.
-
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.
Next-gen Racing Graphics Face-off | (Next-gen means current-gen)
- Thread starter Gestault
- Start date
benzy said its time trial, in case he doesn't replay himself, his pics.
1.21Gigawatts
Banned
Eh, we were all both wrong and right.photomode models are still used in real-time in time trials, in which case they are still valid for comparison purposes.
I see the model swap in the bottom gif, but in the top one there is no model switch.
Photo-mode enhanced:
So, I tried to replicate this shot in GT6 just to compare the models. They're pretty close, but it does look rounder in GT6 thanks to tessellation. (It's the photo travel model though)
Texture resolution looks about the same lol. That's one of the things that's bugging me with these next gen racers, a lot of the textures on the cars are still not as high-res as I expected. Hopefully GT7 is gonna change that, since PD is already playing with vector textures in GT6. If they can implement them in all the cars then the issue will be pretty much gone.
GT6 car models are amazing.
They are. Which model is more accurate though? The GT6 one has way more coils in the spring.
SolidSnakex
Member
What's probaly most impressive about the cars is the accuracy of the different materials
Summer Haze
Banned
I know you posted this GIF to compare car models...but goddam, the vegetation in this game gives me wood (lol). I love how lush and varied the vegetation is, especially on the Canada stages. And the way the sunlight hits it just looks so good.
Eh, we were all both wrong and right.
Yea I've been saying that time trial uses photo mode models while other modes use a lower quality model since the very first post I made that had the comparison of different LOD models.
Heh, the thing with the photo travel models is that they aren't ever used in anything else, I'm not even sure if the models in the rendered pictures are even the same ones used when photo travel mode is running in real-time. With DriveClub running at 30fps I wonder if even GT7 will use them during actual racing.
chromatic9
Member
Looks like real is in-between after looking at xbow pics, they don't show any having different springs either
icecold1983
Member
Is this your own picture? If so you clearly aren't using a very good AA method. And with PC games you can't really go off of a picture, obviously it varies from rig to rig and the game can look stunning if the right settings and ToD are used.
Here I have a similar angle with better AA and a more overcast setting
The game also looks a lot better in the rain, similar to Driveclub.
Now the game can look better than my pictures as I only have a GTX 760 but I really feel the game is stunning at times and while Driveclub is clearly superior overall, the IQ of Pcars and weather effects can give sometimes give Driveclub a run for its money. I really think you guys are too harsh on it.
the poster now replied that it was taken on a 780 at max settings. looks like he has MSAA enabled to me. your shots dont show any less aliasing, its just a lower contrast TOD. anyways the shot gives a good representation of what IQ you can actually achieve on a reasonable system. very few people will be playing this game at an IQ even approaching all the incredibly misleading shots you see people post, the people posting them are often included. 32 fps on a 780 with that level of visuals is just....wow.
The environment and assets/lighting still look just as bad in your pic as well, its not a TOD issue.
Heh, the thing with the photo travel models is that they aren't ever used in anything else, I'm not even sure if the models in the rendered pictures are even the same ones used when photo travel mode is running in real-time. With DriveClub running at 30fps I wonder if even GT7 will use them during actual racing.
Isn't that why they're using tessellation though? To push a higher polycount without it being too demanding? It probably doesn't work that well on ps3 because it's an ancient hardware and it's already a miracle they got tessellation running at all. But on ps4, maybe it will be possible to use these models during gameplay. We'll really have to wait and see.
Neue Regel
Banned
Weird because I get 60fps max settings with MSAA, 30-40 when 2x downsampling, of course in the rain it's lower. And I never see jaggies like that on the spoiler, even without downsampling. You are very harsh on the game, have you played it at all?
icecold1983
Member
on a 760? yeah im not buying it
http://www.pcgameshardware.de/Project-CARS-PC-238576/Specials/Technik-Test-Benchmarks-1136058/
an overclocked 780ti doesnt even get 60 fps with no hardware AA at 1080p. heres another
http://gamegpu.ru/racing-simulators-/-gonki/project-cars-test-gpu-2014.html
even worse performance in their testing sequence, which they provided a video for.
your images have similar jaggies to his, just have to look at different areas due to the lower contrast. see the powerlines for example.
Neue Regel
Banned
on a 760? yeah im not buying it
http://www.pcgameshardware.de/Project-CARS-PC-238576/Specials/Technik-Test-Benchmarks-1136058/
an overclocked 780ti doesnt even get 60 fps with no hardware AA at 1080p. heres another
even worse performance in their testing sequence, which they provided a video for.
your images have similar jaggies to his, just have to look at different areas due to the lower contrast. see the powerlines for example.
Racing on with clear weather at max settings with MSAA I get 60fps, there can be dips if there are a lot of tree shadows but for the most part it's stable.
Those benchmarks either don't explain what weather settings they were on or I can't find them. Stormy weather can reduce framerate, as I said in my original post. Again you are going off of screenshots and benchmarks, I'd actually try the game before criticizing it, you may be surprised how good it looks.
icecold1983
Member
benchmarks done by 2 completely different benchmarking portals actually hold a lot more weight than an individuals claims on a forum. especially when your absurd claims are the EXTREME odd man out while their numbers mostly align. both benchmarks were done in rain. the first link was a 16 car race on belgian forest, the 2nd link is a single car time trial and contains a video of the exact sequence they benchmarked on:
https://www.youtube.com/watch?v=Kr6jZGgZmkI
if you want to try and make a legitimate claim that you get 60 fps w/ max settings and MSAA, upload a video showing a full lap against 11 other cars(same as driveclub) that includes you showing us your settings screen. make sure the framerate is visible. shadowplay is nearly free so the results will be perfectly accurate
https://www.youtube.com/watch?v=Kr6jZGgZmkI
if you want to try and make a legitimate claim that you get 60 fps w/ max settings and MSAA, upload a video showing a full lap against 11 other cars(same as driveclub) that includes you showing us your settings screen. make sure the framerate is visible. shadowplay is nearly free so the results will be perfectly accurate
TheAdmiester
Member
So it turns out I forgot an important thing with pCARS due to its unfinished nature.
Apparently you're meant to delete your user folder with every update right now. I did that, then put all my settings to maximum with the AA set to MSAA like the above poster, and I'm getting a capped 60FPS on the same track and car combo.
Apparently you're meant to delete your user folder with every update right now. I did that, then put all my settings to maximum with the AA set to MSAA like the above poster, and I'm getting a capped 60FPS on the same track and car combo.
Neue Regel
Banned
Well there you go, I don't get 60fps in the rain at max settings, I just don't get how that guy was getting such low FPS on a clear day with a 780, something must be wrong, or it's an older less optimized build. I just find it weird you are beating on the game so much, maxed maxed out at 60fps with clean IQ is a sight to behold, I really doubt you'd disagree if you actually played it.
Ah! Glad to see you got it fixed.
icecold1983
Member
provide screens using the 2 areas i specified were tested in the 2 benchmark links i posted,
Well that's your problem right there. You're using old videos trying to prove your point. Games been optimized a bunch since then.
https://www.youtube.com/watch?v=0Z_-u46143U
There's a more recent video of a 780ti running it at ultra at 1440p with 2x downsampling with a full field of cars at 60+fps. A 760 should have no problem hitting 60 at 1080p with no downsampling. Game runs great and looks amazing.
icecold1983
Member
the track is an empty field with a polygonal wall with tree textures painted on it.
TheAdmiester
Member
My lowest framerate captured on a rainy Belgian Forest with a 16 car grid, behind all of the other cars with all of their spray visible was 28FPS. I tried to capture it but it was only brief and I caught 29, but this is the scenario that caused it.
Great, next time I'm at the Nurburgring GP I'll ask them to plant a few more forests there to make it a bit more lively, eh?
On the real track at car height you'll only see the outer layers of the forests that surround the track. Why bother modelling them entirely when you aren't going to see flyby cameras like Driveclub?
Great, next time I'm at the Nurburgring GP I'll ask them to plant a few more forests there to make it a bit more lively, eh?
On the real track at car height you'll only see the outer layers of the forests that surround the track. Why bother modelling them entirely when you aren't going to see flyby cameras like Driveclub?
icecold1983
Member
at 1080p on your 780 w/ msaa? what was performance generally like on average? mind uploading a shadowplay lap?
i can see how badly it looks during gameplay, what are you talking about? when they have so little to render due to it being a real racetrack, theres no excuse for what little is rendered to not be tremendous quality. especially when they have 3 to 5x the rendering budget over a ps4.
https://www.youtube.com/watch?v=4JWoua1Da_U
1440p, no hardware aa on a 780 stock(no idea if yours is oc'ed.) this is from an older build but i wouldnt expect anything to be too different on the current build.
TheAdmiester
Member
I didn't record the lap because I had Fraps running at the same time for the framerate counter that can be screenshotted.
I took another screenshot of a single car run running at 3840x2160 downsampled to 1080p.
As you can see the game has near perfect IQ because of this, and it's still running comfortably over 30FPS.
Scaramanger
Member
Just tried this myself and there is definitely some swapping going on when in race mode.
Race mode: -
Photo mode (pending post processing): -
Interesting stuff and to be honest entirely expected, I'm not sure why this kind of thing really bothers anyone?
Off topic but doing this brought out my biggest problem with Driveclub, after playing around with this for a couple of mins I drove off and still won the race! The AI had waited for me and within a lap and a half I had caught them! Grrrrrrrrrrrr
Edit: - Actually these pictures don't show the swap very well but on my TV screen you can see a difference in the rendering of the engine bay quite clearly!
Race mode: -
Photo mode (pending post processing): -
Interesting stuff and to be honest entirely expected, I'm not sure why this kind of thing really bothers anyone?
Off topic but doing this brought out my biggest problem with Driveclub, after playing around with this for a couple of mins I drove off and still won the race! The AI had waited for me and within a lap and a half I had caught them! Grrrrrrrrrrrr
Edit: - Actually these pictures don't show the swap very well but on my TV screen you can see a difference in the rendering of the engine bay quite clearly!
DirgeExtinction
Member
I really do wonder if Rushy will explain why he said that there no model swap at all.
Scaramanger
Member
I wonder if its a terminology thing? What we are really talking about here is different LOD versions of the same model, it is probably technically correct that they are the same model.
DirgeExtinction
Member
That's quite possible.
Pretty much. It's like when he assured everyone there's absolutely no rubberbanding whatsoever... despite this...
... and the AI suddenly turning into Schumachers the second you're in front. He later clarified that yea, they'll wait for you, and if they're behind they'll use perfect drafting, which when combined, for the average player, will have exactly the same effect as pretty much any other rubberbanding mechanic.
If you're not super careful in framing the question to avoid there being any technical loopholes, he'll just omit the relevant information and give you the answer that sounds best, because you can technically make the argument that he's not lying.
Scaramanger
Member
To be fair to Rushy you can understand why. The modern world is very prone to band wagon jumping and things like this can be blown out of proportion very quickly and dramatically effect sales. I really wish all devs could be 100% upfront all the time but I fear it may be corporate suicide!
DC is a fun game, a graphical showcase, but for me its not a racing game because the racing is faked. PCars may look weaker graphically but for me its a hell of a lot more fun to race because you are racing rather than trading places in a procession moving at a speed controlled by your own pace.
Speaking of PCars I need to get on there later and try and take some shots that capture visuals that represent how it feels to play!
DirgeExtinction
Member
I think he's talking about the "rubberbanding" in the game, ie: AI slows down to let you catch up, go full speed when you're in the pack.
Btw, didn't Rushy say they were gonna tone down the drafting speed for AI?
Scaramanger
Member
Because of the AI problems we just discussed. I can completely understand that for some people its fine (or even preferred) but for me when the AI are slowing down to wait for me that is not racing.
Anyway its really an OT discussion so we should either leave it or move it to another thread.
eso76
Member
Eh, we were all both wrong and right.
So forza 4/5 should be too ?
I honestly never noticed the model swap, must be because i race in time trial most of the time
Only if you can provide example pics.
Isn't the xbone supposed to get a image capture setting in an update? It really needs one.Yeah, the density and detail of the foliage and a large variety of tree models (including variants within the same species) is great. Really adds a lot more realism to the environments.
Psycho_Mantis
Banned
What's probaly most impressive about the cars is the accuracy of the different materials
This is from GT5:
PD's models are soo good.
flipswitch
Member
But were they kept at the level of detail ingame (GT5) or just for photo travel mode?
SolidSnakex
Member
They're just photo travel. And they're most likely not even the highest end models that PD has as Kaz has stated that the models that they made were beyond the PS3's capabilities and more in line with the PS4.
I made a video of all 4 racing games I like, it's in 4K so they are all 1080P (the 4K option may take a while to show).
http://youtu.be/QFNIMGZdFDk
Forza 5 actually impressed me a bit in this, it isn't bad looking compared to the others.
http://youtu.be/QFNIMGZdFDk
Forza 5 actually impressed me a bit in this, it isn't bad looking compared to the others.
Psycho_Mantis
Banned
I was just talking about the models PD have which will no doubt be used in GT7 on PS4. Iirc photomode models are more detailed than gameplay and also PD have even more detailed models.
Of course!
In game:
PjotrStroganov
Member
I don't always follow discussions to the end and rarely participate in them but it irks me that people who were plainly wrong just keep quiet, play down their part in the argument or just carry on arguing about other aspects.
Respects to Jett for just admitting it. I have nothing useful to contribute so ignore if you will.
Respects to Jett for just admitting it. I have nothing useful to contribute so ignore if you will.
Yea, I understand the dangers these sorts of questions pose, and I don't blame his answer in regardless to model-swapping, as if we're being entirely accurate, that's not actually what's happening (even though he almost certainly knew the question referred to the LOD changes, and could have answered that properly). I do however think that the rubber-banding stuff was less justifiable personally, as rubber-banding doesn't have to mean that the AI cars break speed limits, it's simply any game mechanic implemented to artificially cause a race to be close, even if the player themselves varies significantly in performance. It's not even limited to racing games really, and what Driveclub implements is most definitely a form of rubber-banding. With that said though, it's still a term that many people use to describe a specific implementation (the Mario Kart or Manx TT Superbike kind), so I can't fault him too much for this either really.
There's been other situations too though, such as when Driveclub was apparently close to releasing with the PS+ unlock still tied to your PS+ subscription. It turned out Rushy had actually gone back to edit a very old post that stated the PS+ unlock wouldn't require a continued sub to remain functional. And then there was that thread with downsampled b-roll footage, to which he tweeted:
There wasn't even a need to say anything at all in this case, let alone something so misleading.
Basically, I just think that when discussing stuff like this Rushy isn't really a source worth quoting as an authority, as he's actually likely to give you a less accurate answer than most of us here could give ourselves. There's pretty much two camps any of these queries fall into... either we can determine for ourselves whether something is true or not (as we've done with both the rubber-banding and the car LOD changes), or we can't determine it for ourselves, at which point the issue is obviously minor enough to not even really warrant questioning.
SneakyStephan
Banned
I was just talking about the models PD have which will no doubt be used in GT7 on PS4. Iirc photomode models are more detailed than gameplay and also PD have even more detailed models.
Of course!
In game:
So tired of this gif, I put 200 hours in gt5 and the game looked like that(just with worse image quality that is hidden by postage stamp sized downsampled gif) for maybe 0.1 percent of the time.
A lot of the time gt5 was an ugly ugly game, a lot of the time it was a mixed bag.
Some of the tracks were hideous too, especially the 'standard' tracks
The quality of the (premium, not talking about standard cars, just quality within the premium cars) was also wildly inconsistent.
Some were great (bmw 3 series for example) , some were awful
Reclaimer.
Banned
Hey guys, I'm new here. I wanted to comment on the subject of 60 vs. 30 fps. I've been lurking around here and a lot of people say that a racing game at 30fps isn't good because it affects gameplay.
In my opinion, and this goes for all games really, is that good gameplay isn't a definition of enjoyment, but rather a parameter of enjoyment. Visuals should be concidered. There should be a balance between viuals and framerate that cater toward enjoyment of a certain game. For example, I enjoyed PGR4 tremendously, while it was 30fps. Juding by what a lot of people say, that game would have been more fun at 60fps. I disagree, because at 60fps they would have had to dial down the visuals, and visuals in that particular game are a big part of how I experience the game. Depending on the nature of the game, I'd take 30fps over 60fps, and this also holds for Driveclub. Driveclub is a game in which there is no performance penalty by playing it at 30 fps compared to 60, but the immersion due to better visuals is much greater, which makes me perfectly fine with the framerate.
There are more factors to the enjoyment of a game than just the framerate. If the PS4 was powerful enough to push current Driveclub graphics at 60fps, I think I would have preferred 30 fps but even better graphics.
In my opinion, and this goes for all games really, is that good gameplay isn't a definition of enjoyment, but rather a parameter of enjoyment. Visuals should be concidered. There should be a balance between viuals and framerate that cater toward enjoyment of a certain game. For example, I enjoyed PGR4 tremendously, while it was 30fps. Juding by what a lot of people say, that game would have been more fun at 60fps. I disagree, because at 60fps they would have had to dial down the visuals, and visuals in that particular game are a big part of how I experience the game. Depending on the nature of the game, I'd take 30fps over 60fps, and this also holds for Driveclub. Driveclub is a game in which there is no performance penalty by playing it at 30 fps compared to 60, but the immersion due to better visuals is much greater, which makes me perfectly fine with the framerate.
There are more factors to the enjoyment of a game than just the framerate. If the PS4 was powerful enough to push current Driveclub graphics at 60fps, I think I would have preferred 30 fps but even better graphics.
Neue Regel
Banned
Of course!
Isn't that an AI car? They are lower poly than the players.
Terbinator
Member
What? You haven't played it at 60FPS so can't possibly make that claim in addition to the 30/60 trade-off having a *direct* affect ob gameplay.
Sure, but your blanket statement doesn't do anything for this argument (and is beyond the scope of the thread).
Reclaimer.
Banned
What I mean by that statement is that when you play a fighting game for example, a high refresh rate is a must in order to do good combo inputs, blocks, counters, reactions etc.
In a racing game like DC, I don't really see how it could benefit from a higher framerate.
Scaramanger
Member
I don't think there is a right or wrong with this but for me I would rather have a 30fps fighter than a racer
I don't play a lot of fighters and when I do its basically button mashing however I play a lot of racers and shaving 100ths of seconds from lap times requires as quicker response as possible. So for the same reason as you'd like 60fps fighters I would like 60fps racers, I guess it just depends what type of games are your 'thing'.
Edit: I should add though more arcade racers (like PGR, FH2) I don't find it an issue because they are more bombastic than anything - DC I'm a little torn because underneath it all there is quite a serious racer (bar the AI!).
Psycho_Mantis
Banned
Why don;t you read the context of that reply instead of going off on some anecdotal tangent.