freshseth83
(Banned)
- 8,246
- freshseth83
The forza picture looks like a normal right turn not the pitch you see in the real life pictures. There seems to be no car lean unless its just a bad angle.
Tire Deformation.
- Thread starter townnet
- 88 comments
- 11,676 views
- 216
- 8,723
- 674
- M.I.A
How is it possible to see lean and weight transfer looking at a single corner of the car?
The picture was created solely to show tyre deformation, not car dynamics.
freshseth83
(Banned)
- 8,246
- freshseth83
The car would be at a dive on the left front. You can tell by the 'lean' in the car (positioned at a downward left angle). This photo in discussion doesn't seem to have that. Unless the angle of the picture is causing this.
- 22,551
- Arizona
- HamiltonMP427
That doesn't mean they don't have the tire model right, and to be fair we don't have the full car picture to see body roll. All it tells us is the engine for tires is right but the dynamics of the car's body/chassis isn't in simulation. Either one could be the situation. However, that type of flex isn't exaggerated and it has been shown so now people are saying well perhaps T10 don't have something else right. Let's keep grasping I guess.
This guys has it.
- 552
And once again, in a simulator tyres are by far the most important aspect to get right.
I understand that I really do, but youre looking at photos of old tyre technology, I'm pretty sure the law would deem those tyres unsafe for road use these days
Here we see very minimal tyre deformation, and as you can see thats some hard cornering, if this picture was from Forza those tyres would be curled under the car.
thats what I'm saying, not that it shouldnt be in the game if it can be, but that it should look good and not overly exaggerated so you can see it all the time
![hankook_ventus_s1_evo2_launch_18-1102.jpg[img]](/forum/proxy.php?image=http%3A%2F%2Fthemotorreport.com.au%2Fcontent%2Fimage%2Fh%2Fa%2Fhankook_ventus_s1_evo2_launch_18-1102.jpg%5Bimg%5D&hash=dc66e78232a0bcb9c6458a86becad632)
Last edited:
- 1,856
- Montreal
- oOJ-KiLLA240o
I never said they didn't right or that they exaggerated the deformation as a whole, I was just pointing out that in that particular picture (granted we can't see the whole car) it seems that the tire would not flex like that in a real life situation.
This is what I'm talking talk about.
freshseth83
(Banned)
- 8,246
- freshseth83
freshseth83
(Banned)
- 8,246
- freshseth83
Amazing...
Low profile tyres have significantly less sidewall to deform. Comparing it to high profile tyres is beyond pointless.
p.s the tyre is still significantly deformed. That's why you can hardly if at all see the sidewall on the far side of the front left wheel.
Without knowing the modifications or speed the car is traveling at you cannot know how much body roll it should or shouldn't have.
Do you have something to gain by defending a Forza picture? Theres no lean in the car. The body is 'flat'. If a car is turning hard enough to cause the tires to deform in that manner, there has to be a large amount of load to that area. The transfer of weight doesn't seem to even be represented, like the car is sitting still with the wheel turned to the right. If there was load on that tire to cause it to change shape, the car should show this with the front left of the car (such the bumper area) being lower than the area behind the tire (under the door).
- 216
The car could be accelerating so the front was being lifted..
- 2,321
- Wales
- TSR_GTWelsh
I think in response to the OP we need to all agree that there is indeed...
You simply cannot say its "just an error", with the evidence shown its clearly acting like its deforming and rebounding back, it just isn't visually deforming the tire.
It isn't important anyway as in games like LFS its like "oh cool" for 5 mins then I never notice it again or it looks over exaggerated and stupid.
Tire deformation is and will be modelled, just not shown, and yes it's very important
/thread
- AT LEAST a spring force in the physics engine acting as tire deformation
- OR
- Full tire deformation modelled but not shown
You simply cannot say its "just an error", with the evidence shown its clearly acting like its deforming and rebounding back, it just isn't visually deforming the tire.
It isn't important anyway as in games like LFS its like "oh cool" for 5 mins then I never notice it again or it looks over exaggerated and stupid.
Tire deformation is and will be modelled, just not shown, and yes it's very important
/thread
- 916
I think it says something about Gran Turismo that this is even being debated here. Tyre deformation is a massive factor in car handling! You can't "decide" whether or not to model it in a "simulation". It's a must. Although if they do it like they do tyre smoke I'd expect gentle cornering to make the car fall out a metre sideways on elastic band black tyre sidewalls (and then big puffs of smoke for no reason).
- 22,551
- Arizona
- HamiltonMP427
Which would mean that the model exaggerates and has the propensity to not be done correctly. I'm not saying you said it is exaggerated word for word, I'm just saying it to describe my point. Don't take my wording to such a literal level, I was just saying that if it isn't right, then it must be exaggerated.
- 216
- 169
- 8,723
If that was the case, the driving would be awful with the delays on that order. Usually, the physics engine runs much faster than the graphics, at least twice as fast. What would be causing the delay in that instance?
I can clearly see a two-step response when the tyre lands. That might be a bit of play in the model, or it might be some other form of compound spring, but it seems more likely that it's the tyre given it's relative squishiness.
PD have had access to a full chassis dynamometer, where they can characterise the dynamic spring rates, damping etc. of practically every part of the car. It's not much of a leap to assume that they're modeling the major separate components separately. Indeed, most PC sims have had a sort of "chassis" spring, from what I remember, for some time anyway.
What can't be ruled out at this stage is the action of the anti-roll bar, which is what causes the rear wheel to be in the air before the front has even taken off itself. This would come under "compound spring" above. We would need to see all four corners to get an idea of that. It's unlikely that the suspension geometry is anything but generic at this stage (and going by PD's past history, and the difficulty in characterising every car) so that might explain the odd camber behaviour you're seeing, although I don't see it myself.
EDIT: The steering is opened up slightly after the wheel lands. Given the caster angle modeling that was added to GT5, this would cause a change in camber. Usually, opening the steering would cause the camber to decrease (the angle becomes larger in the negative direction) on the "inside" wheel. There's also a visual effect just from the wheel presenting a different angle to the camera. Then there's all the road surface interactions, and the likely interactions with the other front wheel.
Last edited:
- 433
- MR2 Land
- rollingzoku92
Well this was a hell of a read. Half of you should be hired by some sim companies to do the calculations for the tire model.
- 169
A delay of 1 singular frame or less wouldn't even be detectable by the player in real time.
- 8,723
You may have to provide proof of that, although it depends on the length of a frame - 60 Hz is 16.6... milliseconds, but almost no (driving) physics engine runs that slowly.
As a point of comparison, when GPL's force feedback rate was increased from 36 Hz to 60 Hz, and the single-frame latency built-in for smoothing removed, it made all the difference. There was a similar, if smaller, difference, according to the guy who made it possible, when going to 120 Hz. Most of the physics interactions were occurring faster than that even (I think the base clock was 360 Hz, can't find that info now), and this a game that was made in 1998.
If the reporting of the physics interactions through the wheel are noticeable at different rates, surely the actual simulation of those interactions will also be noticeable at different rates, too. That latency, though, was the big killer, and GT5 seems to suffer from excessive FFB latency itself (which I may only be able to recognise thanks to the aforementioned mod).
Obviously the visual representation is a downsampled version of what's happening in the physics engine, owing to the graphics framerate, but that particular (30 fps) clip we're discussing is in slow motion (2x or more, minimum 60 fps equivalent, probably closer to 90 - 120 fps) and there is no sign of stepping or other aliasing in the motion (quite the contrary, there are even subtle oscillations in the wheels' movements visible, like you see on TV), implying that the physics rate is at least of the order of twice (or whatever the slow-motion scaling factor actually is) the normal framerate - see Nyquist limit / theorem.
I don't think what we're seeing is any kind of physics latency at all.
- 674
- M.I.A
Not surprising. The leaf wasn't built to have sporty handling, thus the mushy soft damping and spring rate combined with long suspension travel and ARB's made of chocolate means you'll be trying fruitlessly forever.
- 433
- MR2 Land
- rollingzoku92
I lol'd so hard when I read this.I'm Motarded
- 552
Sorry to dig up old dirt, but I would still like to push that Forzas tyre deformation model is overly done... and while on this photo of the Lotus you can see the tyres deform, its so slight it wouldnt really be worth modelling, because as I said the only time your'll see it is if youre literally right up against the tyre. Wouldnt it be better to put the data and the effect of tyre deformation instead of wasting valuable processing power on something you would hardly see...
http://www.topgear.com/uk/lotus/exige-s/road-test/roadster-driven
http://www.topgear.com/uk/lotus/exige-s/road-test/roadster-driven
- 8,723
The trouble is, that Lotus is in a relatively low-grip regime (it is sliding, after all). With some tyres, the deformation is pretty obvious.
I agree though, I'd rather the deformation were in the physics but not the graphics, than the other way around. Ideally we'd get both, but the processing budget might not stretch. The problem is: how do you know tyre deformation is in without being able to see it? Does being able to see it mean you can be tricked into thinking you can feel it, even though it's not actually there?
With some cars in GT5 (the 300SL is a good example, from memory) it does feel as though the tyres (comforts only) do sort of "bend" a bit before really biting. That might just be a force-feedback effect, though (which would be weird).
I agree though, I'd rather the deformation were in the physics but not the graphics, than the other way around. Ideally we'd get both, but the processing budget might not stretch. The problem is: how do you know tyre deformation is in without being able to see it? Does being able to see it mean you can be tricked into thinking you can feel it, even though it's not actually there?
With some cars in GT5 (the 300SL is a good example, from memory) it does feel as though the tyres (comforts only) do sort of "bend" a bit before really biting. That might just be a force-feedback effect, though (which would be weird).
- 169
How about some real time footage of a tyre going through racing conditions?
Sounds good doesn't it?
Tyres do deform and a lot when pushed to the limit in a racing situation. This is a simple fact. Tyres are the only part of the car that makes contact with the road so they are by far the most important part of a simulator. Get them wrong and everything will end up being wrong.
Now, GT is meant to be a simulator is it not? Why would it be ok for it to ignore one of the main facets of tyre simulation? Oh, I forget. Since GT is the "Real Driving Simulator" anything it doesn't simulate just isn't important and isn't worth simulating. Right?
- 1,233
- Patra
- Rotorist
I totally agree with you here. There are many cars when being pressed that show this delay in the reaction of the chassis, suspension or the tires. So, tha point is, if we like this delay that when we press a racing car is vanished.
