Gran Turismo 6 general Physics Discussion(as well as video)

Torque steer is an effect of something being uneven/unequal. Whether it be the driveshaft length, tyre's or tyre pressures, suspension etc. But probably the biggest cause is an uneven road surface. Which is why I find it miffing as to why all examples are being swiped aside because on a perfectly flat surface the cars are not torque steering.

In the real world you don't get perfectly flat road surfaces, least not on a public road. In GT6 your cars don't have issues with uneven suspension geometry or uneven tyre pressures and the physics model does not simulate tyre deformation (most sims don't). So if you're on a perfectly even surface with a sim that does not simulate tyre deformation, that is why you aren't getting torque steer at Route X, while you do on other circuits.

As I've said, my own car only torque steers on noticeably uneven surfaces. Atleast only to the point where I can notice it.
 
Btw, (and please excuse my caveman-english) torque-steer happens mostly on FWD or AWD cars because of tire and road grip. In powerful RWD cars the wheels almost never have enough grip to make for enough friction for the wheel to have any difference. Just look at any top gear video with standing starts....

RWD cars have torque steer when they spin the wheels, caused by the rear trying to rotate around to the front.
 
FML...

That is not torque steer......



Torque steer should not only happen because of the camber or bumps on the track. It should always happen. If it doesn't always happen, that is because it's not simulated...

If it's not simulated, the "real driving simulator" is missing a basic vehicle dynamic and completely changes the way cars are driven in an unrealistic manner.

without torque steer simulation the real driving simulator is not real enough!
 
RWD cars have torque steer when they spin the wheels, caused by the rear trying to rotate around to the front.

I know what you say, but torque-steer doesn´t apply only to lateral movement of the car because of wheelspin. Torque-steer happens because of the movement of the crankshaft in only one direction. That goes to the differential and then to the wheels and it´s almost corrected there. Almost....

When you have all the motor spinning in one direction, that´s what cause torque-steer. NOT wheelspin.

I have IRL a 220 hp turbo FWD and you can feel it cornering, when you are pushing on the accelerator. And I don´t mean oversteer or understeer. It´s hard to describe, but you can feel it on the wheel. And only happen in one direction, because of the crankshaft that I mentioned before :) (please note that my english is quite limited... I know that in a FWD car one part is out of the equation, but I can´t find the word... )
 
Last edited:
FWD cars in real life have the additional problem of unequal length drive axles which causes torque steering during hard acceleration.
A good explanation here: http://www.aa1car.com/library/torque_steer.htm

This is interesting:

CONDITIONS THAT CAN INCREASE TORQUE STEER

Any or all of the following conditions can aggravate an existing torque steer problem as well as cause it to appear in a FWD vehicle that previously didn't experience torque steer:

Unequal front tire inflation pressure side to side
Unequal camber (the car will pull towards the side with the greatest amount of positive camber).
Toe misalignment (both static toe or toe-out when steering).
Unequal caster side-to-side
Incorrect ride height (nose too high).
Misalignment between the transaxle, driveshafts and chassis.
Front tires with unequal wear.
Loose or worn wheel bearings
Looseness or worn parts in the steering linkage or rack.
Worn or loose control arm bushings
Control arm bushings that are too soft and allow too much movement under load
Excessive front wheel offset (deep dish aftermarket wheels)
Rear axle misalignment
A dragging brake caliper on one side
Increased engine torque (hopping up the engine in a FWD car that already has a torque steer problem will make torque steer worse).


This is not simulated in GT.
 
That tire squeal is annoying!

Dodzzz, have you been on a circuit before with your car? Even cornering slowly, the only thing you´ll hear is tire squeal ! And going 130 mph cornering on a circuit, with Confort tires (like normal tires on a car) would be like the video...

Watch any TopGear or 5th Gear video, and you´ll see that while annoying, it´s quite real :)
 
Dodzzz, have you been on a circuit before with your car? Even cornering slowly, the only thing you´ll hear is tire squeal ! And going 130 mph cornering on a circuit, with Confort tires (like normal tires on a car) would be like the video...

Watch any TopGear or 5th Gear video, and you´ll see that while annoying, it´s quite real :)
Ah in that case, glad that it's not yet another bug in GT6 :D
 
I like the tire squeal too. It reminds me of Enthusia Professional Racing's tire squeal, which I really liked because it sounded realistic and made the game feel real together with its physics.
 
Wow, It actually works. So far I've done all three comfort tyres with the C7 Stingray, and gaved it a tiny nudge and it slowly torque steered. I'm going to try sports tyres next.

With Sports tyres, the hards remain giving my C7 torque steer after nudging the control stick. The meduims somewhat remain so as well, but with softs it turns alittle than goes straight.
@Wolfe - while the car is spinning it's tires, any steering inputs will initiate what you'd expect; the car will start rotating in that direction, even if it was a simple insta-tap followed by no further steering input.
Thanks guys. That sounds more like Enthusia, then. RWD cars in EPR would launch straight if you began on a perfectly flat and even surface without disturbing the steering, such as Airport Square, yet it only took a bit of camber or a bump or a nudge on the steering to send a powerful RWD spinning of its own volition. It doesn't mean the physics model doesn't incorporate torque steer at all, it means there are no imbalances in the parameters it's been given to "tip" the torque flow to one side or the other. It's like balancing a coin on its edge.

Maybe GT6 still doesn't go far enough in getting a RWD car to dance around under power, and I wouldn't be surprised if it fails to account for uneven-length halfshafts where it should (to be honest I can't recall off the top of my head if EPR did), but I think the issue is more a matter of the game skipping over the various inequalities inherent in a real life car, which @shirakawaa pointed out in the list he shared. 👍

Based on this I wouldn't call it "no torque steer" so much as "somewhat simplistic mechanical modelling". It's certainly sounds like an upgrade from GT's former behavior as seen in the opening section of my old YouTube video about GT4.
 
Thanks guys. That sounds more like Enthusia, then. RWD cars in EPR would launch straight if you began on a perfectly flat and even surface without disturbing the steering, such as Airport Square, yet it only took a bit of camber or a bump or a nudge on the steering to send a powerful RWD spinning of its own volition. It doesn't mean the physics model doesn't incorporate torque steer at all, it means there are no imbalances in the parameters it's been given to "tip" the torque flow to one side or the other. It's like balancing a coin on its edge.

Maybe GT6 still doesn't go far enough in getting a RWD car to dance around under power, and I wouldn't be surprised if it fails to account for uneven-length halfshafts where it should (to be honest I can't recall off the top of my head if EPR did), but I think the issue is more a matter of the game skipping over the various inequalities inherent in a real life car, which @shirakawaa pointed out in the list he shared. 👍

Based on this I wouldn't call it "no torque steer" so much as "somewhat simplistic mechanical modelling". It's certainly sounds like an upgrade from GT's former behavior as seen in the opening section of my old YouTube video about GT4.
You're welcome @Wolfe. 👍 It was an honor to help ya, and also enjoyable as well. :)
 
I am not a professional simracer, but I have noticed that to go inside a turn lifting the throttle quicker or slower really changes how the gar goes inside the turn... Also combining that with braking force and braking slowly/quickly is fantastic and with some cars having a slight oversteer entering the turns is fantastic and also helps to achieve better lap times.
+1 for GT6, great handling for me
 
Could someone more knowledgeable try the 3-2 Goodwood Festival and tell me if there is torque-steer in GT6? If I plant in first my car most certainly moves to the side. I'm seeing different traction levels on left and right wheels in the Aston Martin One-77.
 
Wow, I was finally able to log into gtp.

Guys, regarding "torque steer", use the Mustang or C7 (SH tyres or lower) from the arcade mode and test it on the main straights (start/finish line) of Monza, Silverstone, Indy road course and Fuji. Then test it on Rome circuit, Madrid, SSR5 and SSR7. Finally, test it on Trial mountain, Grand Valley, GT Arena and SSRX. See if you guys can notice a pattern.

My personal theory is on how the tracks are created/modeled. Real world/location vs fictional tracks (laser scanned tracks vs perfect flat straights). Just a theory based off my observations and tests.

@Bigbazz
I think you explained it very well and it fits perfectly with the results from my tests. If the Mustang has torque steer on a real world track like Monza (which is laser scanned) but no torque steer (main straight) on fictional tracks like GT Arena or SSRX (which can never be laser scanned because they don't exist) then it could only lead to one conclusion which is what you wrote, my friend.
 
Last edited:
I was a bit surprised by the Acura NSX. It has a very interesting setup for a road car. You get oversteer when going on throttle quite easy. I would have expected better traction and maybe even slight understeer on corner exits.

Makes it really fun to drive and challenging to drive fast on the limit, I like it. Not sure if realistic but could be. They weren't so obsessed with making every car understeer a little for safety in the early 90s I guess :P
 
I've been thinking over things for a bit and I have a theory that might explain some of the physics quirks we've been seeing.

In GT5 a lot of people complained that the cars weren't dynamic enough. They didn't exhibit enough pitch/roll and weight transfer wasn't really something you could experience to the degree of say, iRacing.

My thinking is there was some fundamental flaw in their dynamics algorithms that didn't allow this behavior to be recreated accurately using the data provided for each car. To counteract this they artificially raised the CG of the car a substantial amount to compensate for the weaknesses of their algorithm. Raising the CG would allow them to keep the old code while having the cars appear more lively (Which they are as we can see from the increased body roll). The raised CG would be consistent with what we've been seeing in game, cars lifting the rears during hard braking and the issue of flipping when hitting curbs or other features that can upset the cars balance.

Thoughts?
 
That tire squeal is annoying!

I'm discovering that PD has actually done a lot of work, impressive work actually, on relaying information about what is happening to the car through the tire noise. I have a very good sound system and play it fairly loud so I can hear everything that is going on with great sound separation between the channels. The other night I was tuning and hotlapping a Ford Focus around Brands Hatch and Bathurst and realized I could tell what was happening to the front end just from the differing pitch and sounds coming from the tires. There is a lot more here than meets the eye and anyone that plays the game without a decent sound system will be missing out somewhat. I would assume that some good headphones might accomplish the same thing although I don't use them, but if you're concerned about things being too loud at home that's the route I'd go.

Gotta give credit where credit is due in this case, well done PD!!:bowdown:

NOTE: I don't want to get into the middle of the MR car debate except to say in the 3 or 4 I've driven, I'm absolutely convinced there is some kind of coding issue with the cars I drove. With the G27 you can feel that the problem isn't actually with the back end, it's with the front, it has way way too much grip giving the cars a severe grip imbalance front to rear. I think it'll be patched somehow and fairly quickly. The cars aren't undrivable, tuning can and does help, like sticking harder tires on the front...but it doesn't make them "right" if that makes sense.
 
It seems to me that GTs continued problems and oddities stem from it not having a "full" physics engine. It should have an environment with surfaces to move across where the car is a loose object that simply sits on the ground freely and reacts based on gravity, tyres, weight transfer, suspension and power transfer to the ground.

Instead it appears to have an engine where the car is anchored to the ground (as opposed to sitting on it) and exhibits strange behaviour because the engine really only wants the car to move within a limited range of predetermined values. The car only wants to do what it's invisible tether will allow it to do. So you get stoppies, cars floating mid corner and an inability to simulate a range of behaviours outside of normal driving including things as simple as being able to do proper doughnuts.

In reference to the torque steer argument earlier, it should be happening whether the surface is bumpy as hell or flat as a billiards table. If it's not happening it's because the physics engine doesn't allow for it.

I realise that a lot of people have strong loyalty to the GT series and it's still a good game but you need to step back and look at things more objectively.
 
It seems to me that GTs continued problems and oddities stem from it not having a "full" physics engine. It should have an environment with surfaces to move across where the car is a loose object that simply sits on the ground freely and reacts based on gravity, tyres, weight transfer, suspension and power transfer to the ground.

Instead it appears to have an engine where the car is anchored to the ground (as opposed to sitting on it) and exhibits strange behaviour because the engine really only wants the car to move within a limited range of predetermined values. The car only wants to do what it's invisible tether will allow it to do. So you get stoppies, cars floating mid corner and an inability to simulate a range of behaviours outside of normal driving including things as simple as being able to do proper doughnuts.

In reference to the torque steer argument earlier, it should be happening whether the surface is bumpy as hell or flat as a billiards table. If it's not happening it's because the physics engine doesn't allow for it.

I realise that a lot of people have strong loyalty to the GT series and it's still a good game but you need to step back and look at things more objectively.

Torque steer is an effect of something being uneven though, if nothing is uneven or unequal in the car model/data then there would never be torque steer on a flat surface. The uneven road surface is creating that uneven suspension for example and the result causes torque steer, just as it is one of the main factors in the real world. If the physics engine was not capable of producing it then it would not happen on those surfaces either.

As I've tried to say before, it's the car data itself where it is missing and not the overall physics engine. Real cars have tolerences and imperfections that result in torque steer, GT6 car models are missing these imperfections.
 
As I've tried to say before, it's the car data itself where it is missing and not the overall physics engine. Real cars have tolerences and imperfections that result in torque steer, GT6 car models are missing these imperfections.

Neither of those things are what produce the torque steer. Imbalances are what produce it. Even in a perfectly flawless and symmetrical car on a perfectly flat surface it will happen because the vibration of the engine and rotation/oscillation of its components will cause an imbalance. It is a physical effect, not the result of an imperfection in car modelling. If it's being added to cars individually or if cars of a certain drivetrain are having to be patched (as another poster alluded to) then that would speak to a massive deficiency in the physics/simulation engine.
 
Neither of those things are what produce the torque steer. Imbalances are what produce it. Even in a perfectly flawless and symmetrical car on a perfectly flat surface it will happen because the vibration of the engine and rotation/oscillation of its components will cause an imbalance. It is a physical effect, not the result of an imperfection in car modelling. If it's being added to cars individually or if cars of a certain drivetrain are having to be patched (as another poster alluded to) then that would speak to a massive deficiency in the physics/simulation engine.

You just misunderstood and then reworded my post, essencially. That engine rotation/oscillation is part of what I mean by imperfections. A perfectly symmetrical car with no imperfections is not going to vibrate in a way that causes torque steer, that isn't how the engine data works in GT6.
 
You just misunderstood and then reworded my post, essencially. That engine rotation/oscillation is part of what I mean by imperfections. A perfectly symmetrical car with no imperfections is not going to vibrate in a way that causes torque steer, that isn't how the engine data works in GT6.

You're right I did. What you're saying is that the things that cause the imbalance are not there to begin with so even if the physics engine IS able to take account of such things it couldn't possibly do so because those things are not there (unless PD adds them).
 
It seems to me that GTs continued problems and oddities stem from it not having a "full" physics engine. It should have an environment with surfaces to move across where the car is a loose object that simply sits on the ground freely and reacts based on gravity, tyres, weight transfer, suspension and power transfer to the ground.

Instead it appears to have an engine where the car is anchored to the ground (as opposed to sitting on it) and exhibits strange behaviour because the engine really only wants the car to move within a limited range of predetermined values. The car only wants to do what it's invisible tether will allow it to do.
So you get stoppies, cars floating mid corner and an inability to simulate a range of behaviours outside of normal driving including things as simple as being able to do proper doughnuts.

In reference to the torque steer argument earlier, it should be happening whether the surface is bumpy as hell or flat as a billiards table. If it's not happening it's because the physics engine doesn't allow for it.

I realise that a lot of people have strong loyalty to the GT series and it's still a good game but you need to step back and look at things more objectively.

Look at how the car floats at 1:37 and the shifty physics when the car i about to lose control.

Mark 1:37

 
Idea for PD for GT7 or GT6 spec 2 physics. Improved weight distribution! 💡 They can have a place for the engine, and how much it weighs and chassis, body... etc etc. So that cars also have a vertical weight to them. Who is for vertical weight distributions? :)
 
Guys, just some little impressions that I want to share with all of you (and please excuse my caveman english)

There´s something quite odd with MR cars.

I´m driving with a DFGT, and I took the Huayra for a drive on Ascari... And dear god, the car just won´t stay on track.

CS tires, no ABS, no TCS, no ESP, No nothing... and it just won´t turn not even at 55 mph without trying to kill me... Of course I don´t have a Huayra IRL to compare (:dunce:), but we´ve all seen videos about the handling, and everybody says it´s pure marvel. Here with the most gentle gas, or if you dare to lift-off while turning, and the back end just goes...

Then I tried with ABS 1 + TCS 2 and it´s just the same.

The next turn was with SH tires, and it was a little better, but there´s no doubt a weight transfer problem here. There´s a thread about the R8 AWD doing the same... and if there´s a car that want´s to be driven hard without complaining is the R8 AWD....

Everytime we see a FR car doing stoppies, at least the driving wheels have weight. But whenever you brake on a MR car the weight goes way up to the front, and the back starts lifting.... and on an MR car that´s not what you´d want. There´s no more traction on the rear wheels and that´s why they snap. And then when the weight shifts back the wheels suddenly gain traction, and then it´s worse because the car it´s sideways already.

Then I go and race on arcade, on professional, (slider to 10) in Deep Forest. With the same car. And believe me, I was holding the DFGT for dear life. On that track (as Ascari) the elevation changes make it quite hard to drive. I just manage to get a 3rd place on a 5 lap race. And I have to use TCS 1 - ABS 1 - SH - NO ESP or Skid recovery and REAL off track traction.

They should tune the weight transfer a little down. On FR cars was different because that allows you to trow the car into the apex and let the back slide a bit and then with a gentle lift off, the car straights, but on MR it´s just too much. WAY too much.

(BTW, I LOVE GT6.... I´m not complaining for the sake of doing it :gtpflag:)
 
Back