Gran Turismo Sport: General Discussion

  • Thread starter Formidable
  • 47,123 comments
  • 4,539,759 views
If this is a stress test then it is great already than GT on PS3.
It wasn't a stress test, and in many cases actually reduced the potential load (by not running in-car but bumper cam) and it (GT) still struggled to hold 60ps on the PS3.

frame-rate-jpg.327271


http://www.eurogamer.net/articles/digitalfoundry-gran-turismo-5-tech-analysis?page=3
Via updates I think they improved the performance also in those section of tracks.
It was a sticking plaster that helped but certainly didn't fully fix it.



They will easily achieve the 60fps target. Some months ago 1 cpu core was available as well to devs. Improvement to SDK and programmer spending more time in games I do not think this will be a issue at all for Polyphony
And you base this certainty on what? The last decade hasn't given me the confidence that will happen.


For PS VR 60fps in minimum which is something like 16.3ms and they said devs should target 15ms to make sure the experience is good in VR. GTS is probably the only game right now which will support it therefore I should think it will comply with the specification. I think one of the challenge is when lots of cars on screen.
It is, particularity when you are using 'head-cam' and interiors, hence the reason why I am not as certain as you that this will be easily achieved.


I wonder why they always have rolling start in GT.
I suspect its more of a physics (low speed tyre model) issue than anything else, even then it seems PD don't actually know how rolling starts actually work.


The only instance where I noticed fps dropping is during RB challenge on starting grid or many cars coming together in weather. They can play around some settings I am sure. I know Forza has the mirror and reflection of cars at 30fps. Not sure I would like them to do the same. But gfx and performance wise they should be fine. That is the probably the least worry TBH
I disagree on both counts, this is a significant worry in my opinion and I'm surprised that you only noticed frame rate drops in those circumstances (but then it's not as noticeable to everyone), however the hard evidence of them is certainly more than available.
 
I suspect its more of a physics (low speed tyre model) issue than anything else, even then it seems PD don't actually know how rolling starts actually work.
I am pretty sure the rolling starts are because of the AI, so that the player dont take the lead after the first corner. The low speed physics arent very wrong! (also we have standing starts in online mode)
 
I am pretty sure the rolling starts are because of the AI, so that the player dont take the lead after the first corner. The low speed physics arent very wrong! (also we have standing starts in online mode)
Umm the low speed physics from a standing start are very, very wrong in GT (unless no car in reality has any sort of front or rear torque steer/step out and tyres hook up once grip in established in a manner never seen in the real world).
 
Umm the low speed physics from a standing start are very, very wrong in GT (unless no car in reality has any sort of front or rear torque steer/step out and tyres hook up once grip in established in a manner never seen in the real world).
Can you further explain this? Apart from the missing torgue steer (which isnt such a big problem as the track surface has enough impact on standing starts) i dont know what you mean.
 
Can you further explain this? Apart from the missing torgue steer (which isnt such a big problem as the track surface has enough impact on standing starts) i dont know what you mean.
If you want standing starts to be even remotely realistic then the missing torque steer is a big issue and I don't agree that track surface compensates for it at all, you should be dealing with both not one.

As far as tyre hook up goes my main issue is the way in which the car just sits with the driven wheels spinning up (almost as if the brakes were still on) until the code decides you have enough traction and then launches you down the track with equal grip arriving at all four tyres in one go. It provides me with no sensation of the tyres being a flexible complex object interacting with the track surface, but of a solid block that is reacting to a preset variable for grip to arrive.

Now GT6 did improve on this slightly from GT5, however I've just fired up GT6 and tried standing starts at the Red Bull Ring Short (first track I came across) and an RT Challenger (rear live axle so it should be all over the place and a handful to launch cleanly). Tested using a pad to remove any FFB interference and all diver aids off.

End result is that you get a car that you can launch with full throttle and the the track surface pulls it to the left a fraction (its hard to notice its that small) and then corrects itself (I wasn't even touching the thumbstick) and tyres that go from no traction to traction in a totally digital manner ( the tyres show no sign of grip being a progressive increase on a tyre by tyre basis). Its as is rolling resistance is simply a value (for all four corners) that once its passed you get all the available grip in one go for all four corners equally.

What should be a challenge is simply a case of wide open throttle and let off the brake, I don't even need to touch the steering for a clean straight launch.

That to me is very, very wrong.
 
Last edited:
If you want standing starts to be even remotely realistic then the missing torque steer is a big issue and I don't agree that track surface compensates for it at all, you should be dealing with both not one.

As far as tyre hook up goes my main issue is the way in which the car just sits with the driven wheels spinning up (almost as if the brakes were still on) until the code decides you have enough traction and then launches you down the track with equal grip arriving at all four tyres in one go. It provides me with no sensation of the tyres being a flexible complex object interacting with the track surface, but of a solid block that is reacting to a preset variable for grip to arrive.

Now GT6 did improve on this slightly from GT5, however I've just fired up GT6 and tried standing starts at the Red Bull Ring Short (first track I came across) and an RT Challenger (rear live axle so it should be all over the place and a handful to launch cleanly). Tested using a pad to remove any FFB interference and all diver aids off.

End result is that you get a car that you can launch with full throttle and the the track surface pulls it to the left a fraction (its hard to notice its that small) and then corrects itself (I wasn't even touching the thumbstick) and tyres that go from no traction to traction in a totally digital manner ( the tyres show no sign of grip being a progressive increase on a tyre by tyre basis). Its as is rolling resistance is simply a value (for all four corners) that once its passed you get all the available grip in one go for all four corners equally.

What should be a challenge is simply a case of wide open throttle and let off the brake, I don't even need to touch the steering for a clean straight launch.

That to me is very, very wrong.
I see the problem with the missing torgue steer but dont see the "no traction to traction in a digital manner" case. When you hard launch a car and it revs to the limiter while spinning tyres you will see the rpm going down slowly for lets say 2k rpm until it gets full grip - and thats no "digital on/off" situation but a progress of the tyre getting grip.

For sure this depends on the car and the tyres. A 800hp will stay in the rev limiter until you hit the limit of the gear and softer tyres will have less progress in gaining grip.

(At least thats the way I see this)
 
But officially it doesn't exist yet. They've already announced GTS for PS4, why on earth would they not be allowed to show the game off on existing PS4 hardware? Because it might spoil how it looks a little better on the newer model?
Basically yes. They want to impress folks with flawless Neo+PSVR+GT Sport footage at E3. GT Sport will of course run on current PS4 and the beta will hit our machines soon.


;)
 
Guess this means the Super Premium cars become the new Premium and the Premium cars look Super Premium and the Premium-standard cars look Premium and the Standard cars look Premium-standard.
 
There shouldn't be a difference in definition like when Component over A/V debuted. Or will it actually be like HDMI vs A/V?

Edit: difference in PS4K over PS4 should be minute. Not like the difference of Component over A/V.
 
There shouldn't be a difference in definition like when Component over A/V debuted. Or will it actually be like HDMI vs A/V?

Edit: difference in PS4K over PS4 should be minute. Not like the difference of Component over A/V.
Yes. The difference would be for example in shades and lighting. PS4 version should be 60 fps too with less eye candy.
 
I see the problem with the missing torgue steer but dont see the "no traction to traction in a digital manner" case. When you hard launch a car and it revs to the limiter while spinning tyres you will see the rpm going down slowly for lets say 2k rpm until it gets full grip - and thats no "digital on/off" situation but a progress of the tyre getting grip.

For sure this depends on the car and the tyres. A 800hp will stay in the rev limiter until you hit the limit of the gear and softer tyres will have less progress in gaining grip.

(At least thats the way I see this)
A canned effect via a drop in the engine revs, but its not felt via the tyres nor does it need to be managed in any way (as it doesn't vary across the four corners in any way).

Give the RT Challenger a go as I honestly don't see how anyone could describe it as not being quite a serious issue with the low speed physics from a standing start. Every other sim (and quite a number of simcades and even pure arcade titles) manage to make a reasonable job of this.

You may be happy to accept launches that can be done in any car without the need to touch the steering, but I certainly am not, particularity in a title that wants to be classed as a serious sim.
 
A canned effect via a drop in the engine revs, but its not felt via the tyres nor does it need to be managed in any way (as it doesn't vary across the four corners in any way).

Give the RT Challenger a go as I honestly don't see how anyone could describe it as not being quite a serious issue with the low speed physics from a standing start. Every other sim (and quite a number of simcades and even pure arcade titles) manage to make a reasonable job of this.

You may be happy to accept launches that can be done in any car without the need to touch the steering, but I certainly am not, particularity in a title that wants to be classed as a serious sim.
Whats makes you sure thats its just a canned effect thats not felt via the tires?

Atleast the low speed physics are more believable than in gtr evo for example (one of the every other sims..) where you can hard launch a bmw gt3 race car in 5th gear from a standing start and it spins its tires till over 100 kmh...
 
Whats makes you sure thats its just a canned effect thats not felt via the tires?
Because when I plug my rig in and do it I can't feel it via the tyres anything other than canned FFB effect!

Even if I could the real issue is that it has no effect at all on the car as a body in motion.


Atleast the low speed physics are more believable than in gtr evo for example (one of the every other sims..) where you can hard launch a bmw gt3 race car in 5th gear from a standing start and it spins its tires till over 100 kmh...
I've not played GTR Evo in a long time, however if that is the case then I would rate them equally as bad, quite honestly the argument that something else gets it 'wrong' so that gives GT a pass isn't one I personally would accept. Personally I would rather that PD actually got it right, because right now it isn't (and that's not conjecture - its objectively wrong when you use reality as a benchmark).
 
Because when I plug my rig in and do it I can't feel it via the tyres anything other than canned FFB effect!

Even if I could the real issue is that it has no effect at all on the car as a body in motion.



I've not played GTR Evo in a long time, however if that is the case then I would rate them equally as bad, quite honestly the argument that something else gets it 'wrong' so that gives GT a pass isn't one I personally would accept. Personally I would rather that PD actually got it right, because right now it isn't (and that's not conjecture - its objectively wrong when you use reality as a benchmark).
The effect of the progress of gaining grip is that "the body in motion" accelerates faster! The only problem you seem to have is the lack of torgue steer.

You where the one who stated that every other sim manages low speed physics better. Just wanted to show an example of that statement being not true. And GTR is said to have very good physics (and way superior to GT6 of course) by many sim players.
 
Whats makes you sure thats its just a canned effect thats not felt via the tires?

Atleast the low speed physics are more believable than in gtr evo for example (one of the every other sims..) where you can hard launch a bmw gt3 race car in 5th gear from a standing start and it spins its tires till over 100 kmh...
Assuming that's the case then there are at least 2 games equally bad in this regard. Are you aware of any others?
 
The effect of the progress of gaining grip is that "the body in motion" accelerates faster! The only problem you seem to have is the lack of torgue steer.
Do you seriously believe that its that simple?

What about load transfer across three axis?
What about the tractive limits that then provides to each of the four points of contact?
What about the yaw rate that now puts into place on each of the four points of contact?
What about the affect on the tyre carcass (sidewalls and face), its internal temp and pressure (and how those two interact)?

That would just be for starters, but if you are (and its seems that you are) happy to accept that more grip = accelerate in a straight line as a complex enough model for a sim then good for you, don't however be surprised when others are not happy enough with that and look for more realistic models to be in place.



You where the one who stated that every other sim manages low speed physics better. Just wanted to show an example of that statement being not true. And GTR is said to have very good physics (and way superior to GT6 of course) by many sim players.
I will be honest in that I can't recall Race 07 (which I spent more time on) being anything like as bad as you claim, but if you actually took my post as being that literal then I will re-phrase it for you.

"The vast majority of other sims (and quite a number of simcades and even pure arcade titles) manage to make a reasonable job of this. Those that don't do it any better than GT (for whatever reason) does then give GT a pass".
 
Assuming that's the case then there are at least 2 games equally bad in this regard. Are you aware of any others?
What 2 games do you mean?

In dont see that problem in GT6, the only major thing lacking is torgue steer - which isnt such a big deal to me as the uneven track surfaces are effecting the cars behavior enough - i mean, people had to use the perfectly flat test track to test and prove that torgue steer isnt simulated.

I have played a few othet sims - but except project cars (with weird rollover physics [GT6s are weird too i know]) and dirt rally i didnt play them much, mostly just a few test runs. I wanted to see how much superior those games really are (and i didnt saw much of that..).

Do you seriously believe that its that simple?

What about load transfer across three axis?
What about the tractive limits that then provides to each of the four points of contact?
What about the yaw rate that now puts into place on each of the four points of contact?
What about the affect on the tyre carcass (sidewalls and face), its internal temp and pressure (and how those two interact)?

That would just be for starters, but if you are (and its seems that you are) happy to accept that more grip = accelerate in a straight line as a complex enough model for a sim then good for you, don't however be surprised when others are not happy enough with that and look for more realistic models to be in place.




I will be honest in that I can't recall Race 07 (which I spent more time on) being anything like as bad as you claim, but if you actually took my post as being that literal then I will re-phrase it for you.

"The vast majority of other sims (and quite a number of simcades and even pure arcade titles) manage to make a reasonable job of this. Those that don't do it any better than GT (for whatever reason) does then give GT a pass".

I dont know what GT6 simulates of all that fancy stuff you stated but the basics are -> more grip = more acceleration

See, your statement sounds much better now ;)
 
In dont see that problem in GT6, the only major thing lacking is torgue steer - which isnt such a big deal to me as the uneven track surfaces are effecting the cars behavior enough - i mean, people had to use the perfectly flat test track to test and prove that torgue steer isnt simulated.
I posted an example today that didn't require a perfectly flat test track to demonstrate it!

Are you honestly just going to ignore that or do you believe that a 400+bhp car from the '70s with a live axle can be launched full throttle in a straight line without touching the steering?


I have played a few othet sims - but except project cars (with weird rollover physics [GT6s are weird too i know]) and dirt rally i didnt play them much, mostly just a few test runs. I wanted to see how much superior those games really are (and i didnt saw much of that..).
Every FM that I have played, Enthusia and Richard Burns Rally on the PS2, PCars, AC, LFS, Race Pro, both of the Shift titles, Grid Autosport (and all of the Grid titles before that) DriveClub, Dirt, F1 2015, WRC 2015, all of the TOCA and RaceDriver series.

All of these (a mix of Sim, Simcade and Arcade) and many more had cars in them which required you to actually think about how you got them off the line when its a car that would cause issues due to its power and/or suspension set-up.

This was created by a member here a few years back and highlight what GT is still missing in places when compared to a PS2 title.

 
Yes. The difference would be for example in shades and lighting. PS4 version should be 60 fps too with less eye candy.

I don't care how they differ, it's the fact that they will differ that pisses me off.
 
In dont see that problem in GT6, the only major thing lacking is torgue steer - which isnt such a big deal to me as the uneven track surfaces are effecting the cars behavior enough - i mean, people had to use the perfectly flat test track to test and prove that torgue steer isnt simulated.

This just suggests that you don't actually know what torque steer is. It is not simply "cars veer off to the side a little on hard launches".
 
What 2 games do you mean?

In dont see that problem in GT6, the only major thing lacking is torgue steer - which isnt such a big deal to me as the uneven track surfaces are effecting the cars behavior enough - i mean, people had to use the perfectly flat test track to test and prove that torgue steer isnt simulated.
You mentioned GTR Evo. That and GT.

I raced a lot online in GT5 and I tuned a lot as well. I had a couple of hundred well tuned cars set up for online racing at all times. I liked to win so I practiced standing starts to see what the best way to start was because I wanted every advantage I could find. In all cases that I recall, throttle modulation was never necessary, you simply hammered the throttle, selected the right gear and off you went in a perfectly straight line in all cars on flat tracks. In some cars it was second gear, in others it might be third gear. This is also not realistic behaviour in cars on both street or race tires of any compound. I detected no differences in GT6 either.
 
I posted an example today that didn't require a perfectly flat test track to demonstrate it!

Are you honestly just going to ignore that or do you believe that a 400+bhp car from the '70s with a live axle can be launched full throttle in a straight line without touching the steering?



Every FM that I have played, Enthusia and Richard Burns Rally on the PS2, PCars, AC, LFS, Race Pro, both of the Shift titles, Grid Autosport (and all of the Grid titles before that) DriveClub, Dirt, F1 2015, WRC 2015, all of the TOCA and RaceDriver series.

All of these (a mix of Sim, Simcade and Arcade) and many more had cars in them which required you to actually think about how you got them off the line when its a car that would cause issues due to its power and/or suspension set-up.

This was created by a member here a few years back and highlight what GT is still missing in places when compared to a PS2 title.


Watched the first half of the video. I guess you want to show what enthusia could do compared to GT4?! Because GT6 simulates the things that are shown (donuts, starting from a banked surface and so on)!

You mentioned GTR Evo. That and GT.

I raced a lot online in GT5 and I tuned a lot as well. I had a couple of hundred well tuned cars set up for online racing at all times. I liked to win so I practiced standing starts to see what the best way to start was because I wanted every advantage I could find. In all cases that I recall, throttle modulation was never necessary, you simply hammered the throttle, selected the right gear and off you went in a perfectly straight line in all cars on flat tracks. In some cars it was second gear, in others it might be third gear. This is also not realistic behaviour in cars on both street or race tires of any compound. I detected no differences in GT6 either.

Why shouldnt it be possible for a very high powered car to launch in 2nd or 3rd gear? Depends on the gearing ratio and is definitly realistic. Its a different story with a GT3 car on racing tires (or whatever the default tires are in gtr evo) in 5th gear.
 
Watched the first half of the video. I guess you want to show what enthusia could do compared to GT4?! Because GT6 simulates the things that are shown (donuts, starting from a banked surface and so on)!
And its taken how long to come even close to that and the low speed is still not what it should be.

However I have been back on GT6 this afternoon and have some good news (in part), as I don't like leaving testing to a single example. So I ran a few more cars as well (some of which I have personal experience with).

  • Pagani Huayra - We have torque steer, honest a goodness torque steer in the kind of quantity and severity it should be, the tyres still hook up weirdly once you get grip but its present.
  • Renault Clio V6 (first generation) - Once again we have torque steer to the degree we should do, it still doesn't have any lift off oversteer however and it shoudl have a ton of it (this I know from personal experience).
  • Merc 190E Evo - A small amount of torque steer, but you have to bounce it off the limiter in first before it happens, it should be more and it should happen sooner (personal experience with the heavier, less powerful 190e 2.3 16v Cosworth).
As such it would seem that in an update at sometime PD have (and a big fat dollop of Kudos to them from me) made some serious headway with this. It would seem that MR cars (and this is based on a sample of two) are now simulating torque steer, while FR cars are starting to do so, but based on the two tested so far its too little and it happening too late after launch.

Opps - forgot one, also tested the Pug 908 LMP1. Very small amount of torque steer on cold tyres, that ones not even close to how it should be.

Why shouldnt it be possible for a very high powered car to launch in 2nd or 3rd gear? Depends on the gearing ratio and is definitly realistic.
It is possible and it does happen, it used to be the norm for some drivers in the BTCC Super Touring days. However that would be second gear, not third (torque is multiplied by gear ratios - third if set to anything realistic is going to bog down).

Its a different story with a GT3 car on racing tires (or whatever the default tires are in gtr evo) in 5th gear.
If the tyres were cold or worn then its not beyond possible at all, get your tyre pressures wrong on a sim and either fail to gain heat or lose it and it most certainly can happen.
 
Why shouldnt it be possible for a very high powered car to launch in 2nd or 3rd gear? Depends on the gearing ratio and is definitly realistic. Its a different story with a GT3 car on racing tires (or whatever the default tires are in gtr evo) in 5th gear.
Forget GTR Evo, it's completely irrelevant to GT. The point isn't launching in 2nd or 3rd gear, the point is that you simply floor the throttle, select the right gear and off you go, no throttle modulation necessary. That's not realistic in any case, especially with street tires.
 
Forget GTR Evo, it's completely irrelevant to GT. The point isn't launching in 2nd or 3rd gear, the point is that you simply floor the throttle, select the right gear and off you go, no throttle modulation necessary. That's not realistic in any case, especially with street tires.
Read my post above, PD have done something in an update, we actually have some torque steer now.
 
Just had a random thought. Now that we know about Neo I guess Sony will release the Reins on GT Sport and then it hit me...

What if Uncharted 4's seeming random and without reason additional delay was so PD could bundle the Beta in with the game!?

You can pinch me now
 
About the ps4neo and Poly : more power, more delay... Just a thought, for the announcement of GTsport, there were no screenshots, a trailer with very little footage from the game. And for months, not even a screenshot?
And for the psvr price announcement, no mention of Gtsport. You have time to talk about indies but not for one of your main IP :P

I feel that the next GT trailer is linked with the PS4 neo.
 
Back