GT5 physics problem

  • Thread starter Thread starter red piramyd k
  • 43 comments
  • 3,948 views
Bottoz
I don't think you people want Realistic Damage... all you want is a crumpled up car. Realistic Damage would drive you away. One little tap or bump and your race is over.

Best answer .
 
I don't think you people want Realistic Damage... all you want is a crumpled up car. Realistic Damage would drive you away. One little tap or bump and your race is over.

Really clever answer for this question...
 
I don't want both, GT5 is perfect as it is.
I don't want to watch replays of crumpled up cars :yuck:
 
We're not talking about realistic damage. We're talking about realistic physics and we're wondering what the 🤬 is happening when cars in GT5 do septuple head spins.

And sure, it's not vital to the gameplay. It also doesn't happen very often or affect us much. We know! But it is worth noting and discussing!



LFS crashes are HELLA buggy, but you can tell the car's weight, inertia and the plasticity of the body is still being simulated. The result are crashes that are a bit glitchy and sometimes completely screwy, but at least you know your car is moving around in a real simulation. This video perfectly shows a physics engine that is buggy, but free of hacks. To me it seems like GT5 has a bunch of weird dampening hacks to stop crazy bugs like this, but the result is equally weird, but dulled crash / roll physics..

In GT5, the physics doesn't seem to apply when a car is airborne. I don't know about programming, so is it like missing code, to save processing power?
 
Last edited by a moderator:
as far as driving simulation goes its all about rubber on the road. Everything about the cars behavior relates to how the tires are contacing the road surface. Once the tires are no longer in contact with the road surface, its no longer a driving simulation, its a flight simulation. The fact is PD's flight simulator isn't nearly as developed as their driving simulator. Thats not to say the driving sim is perfect by any means but the focus is obviously on what happens under the 99.9% situations where the car is not airborne. People worry a lot about whats going on "under the hood" - concerns about this or that element of the physics being "fudged" and what-have-you. In the end it really doesn't matter how they get from point A to point Z - its point Z that we care about. If the end result is a convincing approximation of what it might feel like to drive a Lambo around Nurby I don't care if the underlying physics engine is full of all sorts of crazy fudging. If that means things can go screwy when the car ceases to be a car and becomes an airplane I'm not terribly concerned really. Also, When you add net lag into the mix the potential for weirdness just skyrockets. The prediction code is always going to have a hard time dealing with what happens under the most extreme situations...
 
LFS crashes are HELLA buggy, but you can tell the car's weight, inertia and the plasticity of the body is still being simulated. The result are crashes that are a bit glitchy and sometimes completely screwy, but at least you know your car is moving around in a real simulation. This video perfectly shows a physics engine that is buggy, but free of hacks. To me it seems like GT5 has a bunch of weird dampening hacks to stop crazy bugs like this, but the result is equally weird, but dulled crash / roll physics..

In GT5, the physics doesn't seem to apply when a car is airborne. I don't know about programming, so is it like missing code, to save processing power?

GT5 does know the weight of a car.

I used to know a video where a GTR crashes and comes on top of a Gallardo and the Gallardo trys to go on with full power.

It was Daytona, and it happend before the 3rd corner and the happy couple made it to the 5th.

I'm curious for a Caterham on top, or a GTR on a X1, to see what's going on.

By the way, the crashes CAN be bad if you hit a solid tyre wall.

Again, i could not find a video, but try getting pitted by a opponent in Nascar.
 
Just thought I'd point out that the crazy barrier collisions have been massively improved in the latest version of LFS. Numerical discontinuities are fun!
It is sad though, when I think back at the fully-formed sense of physical interaction (although still often glitchy in "high energy" situations, or online) in games of GT's vintage like GPL or Carmageddon 2, that we're still not getting that with GT. Mind you, it has come a long way.

And whilst I totally agree that the end result is important, Zevious Z, it doesn't take very much for the suspension of disbelief to be broken by GT's frankly odd collision / solid body physics, and it points to a fundamental shortcoming in the overall physics system.
Maybe it's part of the reason body pitch / roll is so slow and under-represented, or that the low speed physics are so screwy, or there's a lack of any real drivetrain interaction with the road, or that jumps are self-levelling. Were all of these to be "fixed", it would make for a better simulation experience, so hopefully all of these will be addressed in turn; unless there's a fundamental reason they can't be with the current system.
Let's not forget that there are plenty of situations in real life that at least one wheel will come off the ground; e.g. most FWD cars cock a rear wheel in the corners, and some older FR cars would pull the inside front wheel up in corners, too. GT exhibits neither of these traits in the appropriate cars, either.

It's come a long way, but it still has some way to go yet. :)
 
In GT5 cars don't crumble because there's no damage so the momentum from the crash can't get absorved which then becomes motion as we see on the video.
In a crash cars crumble because they absorve energy from the crash, if they couldn't absorve energy (no damage) - crashes in real life wouldn't differ too much from this video in the op.
Also, play iRacing and you'll see the same problem - no game has realistic crash physics.

Is this why they act like rubber whenever they crash?

GT5's damage is more suited to deformation and dents, rather than pieces breaking from impact:dunce:

Which is why this seems valid.
 
I think gt5 has alot to learn from iracing on this topic. iracing is not perfect but it has defently the most real crash physics of the games I have tried.
 
I think gt5 has alot to learn from iracing on this topic. iracing is not perfect but it has defently the most real crash physics of the games I have tried.

PC has a lot more ram than console. PC is a lot better multitasking than consoles as well.
From what I've read good crash physics demands some serious amount of hardware which is why even PC sims uses short cuts.
P.S At first Nascar 2011 was suppose to have some good crash physics but didn't seem to make it in the final release probably due to the serious hit to the framerate.
 
Last edited:
i dont know if this affects the driving physics , but this is just annoying, crashes are totally "fantastic"

http://www.youtube.com/watch?v=89Oq_aMq0sY&feature=related


just play nascar and you will see lol

PD have 2 serious things to improve before getting new DLC:

-IMPROVE THE ENGINE SOUNDS
-IMPROVE THIS PHYSICS (DRIVING ISTS FINE)

im just gonna say pre 2.0 / before 2.0 was much better in terms of physics.
 
*puts on David Attenborough voice*
Here we have the common "fix this PD" thread, inhabiting the Gran Turismo Planet forum. It's common predator is the Super Moderator. Most of these threads don't grow to be over 30 posts, and are usually killed of in the earliest stages of life.

lollollollol:tup:👍👍
 
Back