Gran Turismo's Future: "4K Resolution is Enough", But 240fps is the Target

Wrong. 50-60% of PS4s do not have PS+. Single player games are some of the best selling games on PS4.

PD have already rightfully said that GT7 will be a mixture of past, present and future GT. Having a great SP campaign is what most people want.

Yeah its clear, abundantly clear the PS4 has succeeded partly due to their strong portfolio of single player exclusives.

I think PD learnt their lesson when they released GT Sport without leagues and then they promptly added it two months later at Xmas and then, partly due to some of that, ended up selling 10 mil copies.

Its clear they will have GT Leagues back day one. I disparage PD a lot but I think they wont do that again.
 
As in graphic area, the most important aspect is the talent programming, developper decisions and optimization.

You can create a very bad AI with 5x physics calculations than other.

That's why AI and physics are better in some old games than in some new games.

Agreed, but I’m still convince proper physics will help to make AI more realistic.
 
Last edited:
Otherwise we'd dump every trick in the book we possibly could into super-realistic rendering, at a low res, then do the UI on top at full resolution.

I've been hoping that some indie dev would try this approach, to mimic mid-90's in-car footage - the level of 'realistic' effects you could get at those low resolutions would be incredible.

Sort of like a Cuphead approach, maybe?

(also cheers for the expert contribution)
 
I don't think people need to worry that they'll target 120fps. 120hz screen ownership is way too low.

My general hope for next gen racing titles is that they'll optimize the graphics and experience for 4k@60hz and then add a setting for lowered graphics at 4k@120hz.

Since getting a 120hz TV, I've made the jump from Dirt Rally 2 in 60hz 4k on Xbox One X to 120hz 1440p on PC, and it really does make a big difference. The combination of lower input latency and high update rate makes hurtling down narrow roads at crazy speeds at the edge of control that much easier.

For future titles of Forza or GT6 that I'd intially try out and want to be wowed by the graphics, I'd opt for 60fps initially - but if it's a game I stick with and want to get competitive at, I'd eventually change to lower graphics and 120fps to get the extra responsiveness and temporal resolution.
 
This actually doesn't tell us anything and entirely relies on the assumption that people play online for trophies. I dont do SP at all, only MP and I could care less about trophies, which shows how that metric is meaningless.
GTS has an trophy for unlocking the good sports mode. Many people didn't even bother unlocking it.

If anyone has technical questions, just message here and ask.
Just wanted to thanks you for the post. Imo just "like" the post wasn't enough. Everyone should read it :)
 

Attachments

  • Screenshot_20200218-112044.png
    Screenshot_20200218-112044.png
    19.7 KB · Views: 47
Last edited:
Wtf...GT sport is NOT 4K now! Ps4 pro and 4K tv I can see very ugly jagged edges!!! What is real 4K e.g. Forza Horizon 4 on PC. Tack sharp everywhere...
 
No...no it doesn’t. Dow sampling is far from a “required” feature for graphical fidelity.
It's required if you want to completely eliminate aliasing, in that I've never seen it achieved without that. Rendering at the display resolution with 8x MSAA leaves significant aliasing. Even the PS4 Pro's 4k rendering of GTS downsampled to 1080p has significant aliasing because it's not a full 4k rendering, and doesn't have enough AA applied at 4k. You need significantly more power than a PS4 Pro has just to produce a perfect 60fps 1080p image, never mind a perfect 240fps 4k image.

Also why bring PC here?
The relevance of games running on PC is it gives us an idea of what is possible. It seems unlikely that PS5 is going to massively exceed the performance of a 2080Ti. From the specs published so far, it seems it will perform more like the 5700 XT I've got in my gaming PC, i.e. 9-10 TFLOPS. That can manage around 90fps 4k with 4x MSAA for FH4. It's a nice improvement over current gen consoles, but 240fps 4k with really nice AA is PS6 or even PS7 territory IMO.
 
Last edited:
The relevance of games running on PC is it gives us an idea of what is possible. It seems unlikely that PS5 is going to massively exceed the performance of a 2080T
Wait, not in my context actually. He was talking about GTSport and thus he meant current gen console, PS4 and Xbone, which has concrete specs at this point.

For PS5 and Xbox X, we not really sure what their specs are so speculating the next game relative to PC is fair.

I just wish next GT will support 120 FPS (120hz) or stable 75 FPS instead and give the remaining graphical power to elsewhere, like better antialiasing, smoke effects, better reflection, better track rubbles, etc. Not adopting 8K fad is actually a step forward already in my opinion and I'm glad Kaz realizes that.
 
My eyes are too horrible to see 4k very well, same with seeing 120 fps, I can however perceive the reaction time differences of 120. So a 120fps GT gives me a reason to upgrade from a 60hz display to a 120hz display.
 
Point is you guys are in the growing minority. Racing sims are overall abandoning the single player relative to the online play.

ACC 8pm last night per Steam.
1,652 total
Online players 384

You’ll find the same percentage difference across all the sims favoring career/championship/whatever excluding of course, iRacing.
 
What is offloaded to the CPU specifically? Afaik GPUs are great at parallel processes and most linear algebra can be done in such a way. So why not have everything on the GPU?
I'm not sure which part of the post you're referring to, so I might be answering the wrong part.

If it's about rendering, even nowdays, with GPU particles, tessellation, etc, the process of setting up for executing the render is still CPU. Less than previously (5-10 years ago) but it's still there.
Think of it like the CPU being the secretary and the GPU being the CEO/Associate/Partner/whatever.
Agreed, but I’m still convince proper physics will help to make AI more realistic.

No. There's still so much room left to fine-tune the AI as it exists, without 'creating' any new information. Think of it like this: the AI will hit the curb at the apex of T2 at Red Bull Ring, causing it to hope on the left 2 wheels a bit, then oversteer on power-down. It'll react in (say) 1/50th of a second to countersteer; a micro-correction.

What do humans do? We predict that oversteer is a likely outcome for such an action (curb-hitting), and will...queue up (???) some counter-steer in readiness of that.

I said 'create' like that cause it could be able to have that information from the previous lap. Whether it's learning things like that, I doubt. But the point is, it has more than enough information and variables to drastically improve upon what it is now. And not just improve, but start to show behaviours like that, which we associate with human driving and driver development.

I'm not saying our B-Spec Bob has a brain that big, but it would be very interesting to see Big Think Bob, who was trained on SS replays.

I've been hoping that some indie dev would try this approach, to mimic mid-90's in-car footage - the level of 'realistic' effects you could get at those low resolutions would be incredible.

Sort of like a Cuphead approach, maybe?

(also cheers for the expert contribution)

It's unlikely that an indie dev would try, simply because it's not just a problem of running the application, but building it. Paying the rendering performance cost is one thing, the R&D cost for integrating the latest and greatest teach into a realtime engine, then creating content for it, would still be the same as the R&D + content creation cost a large VFX studio would incur.

Just wanted to thanks you for the post. Imo just "like" the post wasn't enough. Everyone should read it :)

No problem. :)
 
I'd rather them focus on their crap tire model. The game looks good enough. Sure, more fps would be great. But I want the "real driving simulator" not the "real visual simulator".
Personally want both so i hope they don't sacrifice visuals for the sake of super hight framerate, since also this is part of the simulation. To me it's better 60 fps with hight graphics fidelty with good reflections,shadows,lod,trees,textures,dynamic weather and tod etc. than 120/240 fps game with GTS graphics (i know GTS has top visuals but surely the competition won't stand by)
 
I think GTS will be the baseline of what is supported graphics wise... ie. it wont get worst than what we have.

I mean it seems like GFX is ending to a point where not much else can get better.

One can only hope that things like HDR 1,080/120hz and 4k/60 with dynamic lighting and weather conditions is the standard.
 
Personally want both so i hope they don't sacrifice visuals for the sake of super hight framerate, since also this is part of the simulation. To me it's better 60 fps with hight graphics fidelty with good reflections,shadows,lod,trees,textures,dynamic weather and tod etc. than 120/240 fps game with GTS graphics (i know GTS has top visuals but surely the competition won't stand by)
Wouldn't a high framerate act as a buffer when there's a lot happening at one time e.g. a group of 24 A.I. cars bunched together at turn one with pouring rain and time progression accelerated. A huge drop from 120fps like to 90 is still less noticeable than any drop from 60, right?

Also, people bring up the physics calculations a lot but do they actually require so much power that things have to be sacrificed?
 
I mean it seems like GFX is ending to a point where not much else can get better.
I disagree, because real time ray tracing can improve games a lot and in GTS or other racing games environmental detail is still not very close to real-life.
 
I love ray tracing, I had loads of fun playing an open source ray-traced version of Quake 2, hex-editing the maps to get some cool effects etc.

But it will continue to be an "add on" to, rather than the basis of, full frame rendering for at least the next generation. Nevertheless, even just applied to certain specular reflection and direct shadowing effects, it is pretty much "transformative".

Combine that with all the fancy upscaling tricks now available, there's no real need to worry about performance.


On the display side of things, the next consoles are said to support adaptive refresh. So whilst you don't strictly need new hardware to take advantage of that, it does work better if it's been properly tested and tuned. I'd say that's a good sign that the industry is waking up to the issues with feedback-breaking latency figures and other framerate issues.
 
I love how people suddenly decided that they don't care about graphics at all, turns out.

The same people that moan 24/7 back then about the ugly trees at Nurburgring when the early build of GTSport was first unveiled.
I bet that these people who are saying they prefer high framerates and if PD will listen to them sacrifying visuals (let's face it, with a 500 euro hardware it cannot be otherwise) they will start bashing the game because it's to similar to GTS and things like that.
 
But I want the "real driving simulator" not the "real visual simulator".
Suddenly, I was reminded that they couldn't even implement turbo lag properly in the game, as well as the characteristics for a diesel car before.
 
I bet that these people who are saying they prefer high framerates and if PD will listen to them sacrifying visuals (let's face it, with a 500 euro hardware it cannot be otherwise) they will start bashing the game because it's to similar to GTS and things like that.
Not this people, no sir.
 
I think there are going to be a lot of disappointed people when the next Gran Turismo game comes out.
If they release just a GTS porting with some more cars and tracks and without real evolution since they want to reach 120 fps, than yes probably you are right
 
I would like 60 car multi class races at Sarthe and 150 car multi class races at N24. That would make FIA a lot more interesting.

This question goes out to anyone interested in multi-class racing: why?

In reality, it's to solve an problem that simply doesn't exist in GTS, or anything virtual.

For Le Mans, multiclass was considered a Really Dumb Idea for a long time (and arguably still is) because the speed difference between cars was and is so high. This wasn't just a safety concern (closing speeds between vehicles) but also for the quality or racing (GT cars being mobile chicanes to the top tiers, means competitive cars get separated too easy).
Also in GT Sport, you don't have to worry so much about the Gentleman Drivers needed to prop up teams either, though that point could be argued at CA-BA rating online....

Why do they do it? Because the top tier class was too small.
Is that a problem in GT Sport? No.


For the Nurburgring 24h it's kind of similar - except you're not limited by the number of entry slots (pit spaces, track limits). I think they had to cap some classes, from memory. You can't run multiple 24h races over a month to fit in everyone in every class who wants to enter.

Why do they do it? Because they can only run the one 24H of Nurburging race a year.
Is that a problem in GT Sport? No.


Multi-class is a solution to a problem that simple doesn't exist in this context.

You DON'T add it to improve the racing - a full grid of competitive cars is going to product better racing than only a handful of competitive cars.
You DO add it only to fill up small grids in real life, typically because the category organizers suck at their jobs.
 
Back