GT5 physics model - how complex is it?

[nasanu]

That's from GTR2 - which uses the GMotor2 physics engine, and that's only the rear left corner. - just to give you an idea of how complex a physics engine can be

I could write a program with 1 million variables that just prints 'hi sekai', it would not be complex. All you have above are variables. There may well be quite simple equations running on them.

If you play a lot of varied games with an open mind you realize that these days there is not much difference between them. I recently took out a 3 month sub to iRacing 2.0. Going straight into it from GT5 was easy, they are both very similar physics wise. I've also been playing a little of GTR evo and that is basically the same too. They have all reached a level of realism that when someone comes along and states that one blows another way in terms of physics you can be certain you are dealing with bias rather than considered thought.

 

[vramak]

OMG. Just looking at section 8 has me ready for a drink.

Computing a cars velocity
"How fast it is going in x direction, ignoring its motion in the y and z directions."
The velocity of the race car formula.
race-car-vx car161
race-car-vy car161
race-car-vz car161
define vx(1(-x2x1)(-t2t1))

This stuff is insane.

You're not serious, are you ? That just says, velocity=distance/time

 

[robbo6]

Quite true! It would be easier to use and understand that way.
However this isn't really a "physics" issue, but more like UI/strange choices.

do we know that for sure tho, maybe doing it the way PD do it is the only way for it to be remotely realistic?

 

[vandaliser]

All i know is that driving my friend's honda civic matches exactly that which is in the game. And that is good enough for me.

 

[GoSpeedRacer]

I'm curious as to what effect the updated physics of 2.0 will have on current tunings. I recall one of the patches changed the transmission gearings and all gearing setups had to be redone.

This Spec 2.0 could be tough on Tuners that have a bunch of custom tunes, as well as users who have a bunch of their own. I guess we'll have to wait and see and then, take it one car at a time.

 

[GTP_Gearhead]

You're not serious, are you ? That just says, velocity=distance/time

Did you open up Scaffs link? Much more there. I didn't want to clog up the thread with wasteful text when you can click the link. Your opinion of my ability to understand simple velocity math is appreciated. Thanks!

 

[LP670-4 SV]

would be cool if they implemented the RAGE physics engine, so rather than constant parameters....the processor would calculate late things like speed/downforce/braking distance/tire grip on the fly (ex. GTA IV) it would also make the crashes EPIC lol

 

[gtnoobie]

If my friend who used to drive a alfa 146 got gold (by more than 0.8s) on the national b test with 147 on 2nd try but hardly bronze the golf gti on b license test, something must have clicked. Similarly, when I started off, I drove better in my own car, Mazda axela, than other comparable cars. I doubt any simulator can fully capture everything in real life else actual f1 race would be on simulator. Despite many millions invested with many manu different tuning and shared engines, some teams like red bull this season consistently perform better. So there must be subtle differences even the best engineers, equipment and simulator can't identify and match.

 

[Exorcet]

All i know is that driving my friend's honda civic matches exactly that which is in the game. And that is good enough for me.

Subjective tests are not as valuable as objective ones.

If my friend who used to drive a alfa 146 got gold (by more than 0.8s) on the national b test with 147 on 2nd try but hardly bronze the golf gti on b license test, something must have clicked.

Luck?

 

[Sselecta]

I'm curious as to what effect the updated physics of 2.0 will have on current tunings. I recall one of the patches changed the transmission gearings and all gearing setups had to be redone.

This Spec 2.0 could be tough on Tuners that have a bunch of custom tunes, as well as users who have a bunch of their own. I guess we'll have to wait and see and then, take it one car at a time.

Hopefully it's only some minor changes such as slipstreaming and nothing in general that affect the setups as a whole.

 

[Blood*Specter]

Each or the original stock, and racing cars in GT1 was individually mapped and miked for performance and sound. I am pretty sure, weight and distribution of weight was also taken into account. Torque figures gears as well. To do this type of work for a single car has to be exhausting and perhaps expensive. Also, think about the features added over the years.

I think Kaz would never settle for less than authentic. The real question is; do PD and Sony support him, or did they insist on cutting his budget so that he has to build cars off a standard base. When you think about that, you have to add in that many car manufacturers would not let Kaz misrepresent their cars. I wish I knew for sure. It would make tuning cars very interesting indeed. I am going to look for more information on this subject. Because it’s really the foundation of the game.

 

[Blood*Specter]

This was taken for "RIDE'S Magazine" It's a start but not real in depth questioning.

The game boasts more than 1,000 real cars. How long does it take to construct one car from start to finish? For a premium model in which even the interior is completely re-created, it takes six months for one car. For a standard model, it takes around four weeks.

What are some of the processes involved in constructing car models? We measure surfaces using a 3-D scanner; we put the paint color under a spectrophotometer, a device used to measure light intensity; we measure the power of the engine and do things like place cars on a dynamometer to record the sound.

The GT games are known for their meticulous realism and accuracy. Have you ever been agonizing about some minute detail—like, say, the sound a windshield wiper makes on a 2003 Mustang Cobra—and thought to yourself, Okay, now we’re just being silly?
I think if you look through GT5, you might find in many places that we’ve gotten “silly.” I think that over-the-top attitude is what gives Gran Turismo its character. We don’t want something that is just well made; we want to make it something that makes you feel a bit of crazy madness.

 

[godlameroso]

Tires can never make up enough grip to negate loss of downforce. Tires gain grip with normal force.

And I did not say there was no grip from aero, I said there was no drag. It doesn't matter if you're maxxed or minned on downforce, your top speed varies by ~2% at most.

True top speed doesn't change much, what does change is how fast you get there. It's not just the 190mph top speed but what happens from 120 onwards where trimmed wing becomes apparent.

 

[rexc]

Dan Greenawald was talking about how they did the physics for the cars in Forza. He said they basically got a sample number of cars with a good variety ie FR, MR FF etc and mapped them extensively, doing real life testing to test the modelling etc. Then for the rest they just plugged in the figures. ie that is an MR car with that type of suspension therefore we can make the assumption it will handle like this etc.

Give that whatever weight you want, but I would assume PD also take a similar approach.

The other issue is that its been said that with this generation of consoles it is perfectly possible to develop an extremely accurate handling model, but then this is always dumbed down to some degree to make an accessible game that can be played with a controller etc.

 

[Exorcet]

True top speed doesn't change much, what does change is how fast you get there. It's not just the 190mph top speed but what happens from 120 onwards where trimmed wing becomes apparent.

That doesn't change either. GT does not model drag very well.

 

[Bushby_23]

That doesn't change either. GT does not model drag very well.

Have you driven an x1 with both full and minimal downforce? Do that, then get back to me on aero drag not really affecting speeds.

 

[Exorcet]

Have you driven an x1 with both full and minimal downforce? Do that, then get back to me on aero drag not really affecting speeds.

I don't care about the X1. The X1 speed difference should be seen in LMP cars at the very least.

The X1 is the exception in that is makes so much downforce that the miniscule drag difference actually becomes visible. The X1 is just further proof that the physics are all wrong in this area.

 

[Bushby_23]

I don't care about the X1. The X1 speed difference should be seen in LMP cars at the very least.

The X1 is the exception in that is makes so much downforce that the miniscule drag difference actually becomes visible. The X1 is just further proof that the physics are all wrong in this area.

The fan speed isn't dependant on road speed, so that doesn't matter. Do it to one of the f1 cars, same effect applies.

 

[Exorcet]

The fan speed isn't dependant on road speed, so that doesn't matter. Do it to one of the f1 cars, same effect applies.

What are you talking about? The fan doesn't have anything to do with this. T he X1 has wings like every other race car.

I don't remember downforce making much difference for F1 cars, but that too is besides the point.

The point is aerodynamics are not modeled properly, because in GT the fastest way a circuit is put downforce to the max because you can double/triple/quadruple your lift coefficient and barely change your drag. It does not make sense. The optimum downforce is somewhere between min and max 99% of the time with high end race cars, but that is not reflected in game at all.

 

[Sele1981]

I dont know how good the aerodynamic stuff is, I dont know how good GT simulates TCS,ABS and other stuff...

But I must say the driving physics are still outstanding good! For me, the best on consoles.

Forza 4 is really trying it but it´s hard to describe, the FM physics feel "rusty" or not alive. I dont feel the dynamic of a car. If the car gets in understeer, the wheel locks in a strange way and the car only slides away. Probably the hidden steering assist ruins the physics, I dont know.

GT5´s physics feel totaly alive and dynamic for me. If you get understeer in GT, you can correct it in milliseconds.

It´s hard to describe, especially because english is not my main language...

 

[snowxdadx3]

All i know is that driving my friend's honda civic matches exactly that which is in the game. And that is good enough for me.

^^^^ what he said

 

[Exorcet]

But I must say the driving physics are still outstanding good! For me, the best on consoles.

Well I'd disagree there. FM4 is much better simulator. But that's another thread, in fact it's in the Q and A section of the GT5 forum.

Anyway, decided to throw in some numbers

F2007 speed vs downforce.

DF max
223 mph
DF min
212 mph

Essentially no difference. That speed difference is less than the mount the Dodge Viper suffers going from SRT-10 to ACR, and that car has a much less serious aero package.

Peugeot 908

DF max
226 mph
DF min
235 mph (I really hit 233, but I think the power started falling off)

I decided to do some further testing on the 908

2.0 g lift = -1064 lb, .1064 coeff (max DF at GVS tunnel)
1.8 g lift = -701 lb, .086 coeff (min DF at GVS tunnel)

Assuming 10% transmission losses
Max DF drag = 1020 lb (.02 coeff)
Min DF drag = 981 lb (.0178 coeff)

L/D rear wing est
Max DF = 5.32
Min DF = 4.83

The wing is getting more efficient with increased downforce. Highly unlikely. Induced drag rises with the square of lift.

 

[CSLACR]

Well I'd disagree there. FM4 is much better simulator. But that's another thread, in fact it's in the Q and A section of the GT5 forum.

Anyway, decided to throw in some numbers

F2007 speed vs downforce.

DF max
223 mph
DF min
212 mph

Essentially no difference. That speed difference is less than the mount the Dodge Viper suffers going from SRT-10 to ACR, and that car has a much less serious aero package.

Peugeot 908

DF max
226 mph
DF min
235 mph (I really hit 233, but I think the power started falling off)

I decided to do some further testing on the 908

2.0 g lift = -1064 lb, .1064 coeff (max DF at GVS tunnel)
1.8 g lift = -701 lb, .086 coeff (min DF at GVS tunnel)

Assuming 10% transmission losses
Max DF drag = 1020 lb (.02 coeff)
Min DF drag = 981 lb (.0178 coeff)

L/D rear wing est
Max DF = 5.32
Min DF = 4.83

The wing is getting more efficient with increased downforce. Highly unlikely. Induced drag rises with the square of lift.

Coulda just said aerodynamic physics in GT5 suck.
You forgot something in your test though, and try this out, it won't disappoint.
Aero adds grip off the line. Found this out "drag tuning", I'd thought lower aero would be the way to go in a 1.5 mile drag, but nope, the max aero makes the car take off like a rocket off the line compared to having minimum aero.

 

[Sele1981]

Well I'd disagree there. FM4 is much better simulator. But that's another thread, in fact it's in the Q and A section of the GT5 forum.

I was only talking about the feeling (physics) when you drive. Forza physics have way too much grip, it has hidden steering assist and it feels dusty.

Maybe other games simulate other things better. But I cant get this great feeling in any other game.

 

[TBR 427]

The wing is getting more efficient with increased downforce. Highly unlikely. Induced drag rises with the square of lift.

For the F2007, it's extremely likely. If the wing is more efficient at higher-downforce tracks, then it's useful because apart from Monza and Canada, and maybe Spa (where they run low-downforce-wings), having a good set up for tracks like Melbourne, Barcelona, Silverstone, Suzuka, etc takes preference.

For the 908 it's probably a similar thing, though not as much. I'm not too up to date with the various Prototype races (the cars, on the other hand...), but tracks could have a high role in this; there may be a bunch of races in a year, but usually only one 24 hours of Le Mans, where they run with a BESPOKE aero package, with the gourneys, dive planes (oh wait) and other rules.

 

[tinram]

I was only talking about the feeling (physics) when you drive. Forza physics have way too much grip, it has hidden steering assist and it feels dusty.

Maybe other games simulate other things better. But I cant get this great feeling in any other game.

Immersion, GT series always had this, specially when using a wheel.
FM always felt like a game (with a fanatec wheel), no matter how long/far i drove.

 

[Exorcet]

For the F2007, it's extremely likely.

It's not. There is a fundamental aerodynamic relationship preventing that. For a given wing, the induced drag increases with the square of the lift. In simpler terms, as downforce goes up, drag goes up much faster.

Also in F1, they tend to actually change the wing from track to track. They still need to pick the best operating point on the L/D curve for the wing in use though. In GT5 all you ever need to do is max [negative] L.

 

[cuco33]

I dont know how good the aerodynamic stuff is, I dont know how good GT simulates TCS,ABS and other stuff...

But I must say the driving physics are still outstanding good! For me, the best on consoles.

Forza 4 is really trying it but it´s hard to describe, the FM physics feel "rusty" or not alive. I dont feel the dynamic of a car. If the car gets in understeer, the wheel locks in a strange way and the car only slides away. Probably the hidden steering assist ruins the physics, I dont know.

GT5´s physics feel totaly alive and dynamic for me. If you get understeer in GT, you can correct it in milliseconds.

It´s hard to describe, especially because english is not my main language...

I would think that you would need to play FM4 to make any serious claims about FM4's physics or whatnot. I mean folks pointed out issues with your posts in the FM VS GT thread questioning if you even played it and you haven't posted much there ever since 💡

Personally, these games are a lot more complex and advanced today. Console racers are just as advanced as PC racers. Next gen is going to be even more epic!

 

[robbo6]

Console racers are just as advanced as PC racers. Next gen is going to be even more epic!

erm nope! Console racers look good but are still behind PC sims in many key areas. PC sims are MUCH better at the full race weekend, Also they simulate much more too, GTR2 is a game from around 2006 and factors in brake temperature, and you could blow the engine too. I havn't tried the stuff like iRacing either, and only had a go at the demo of NKpro, WAY more stuff going on. No matter how good F4 and GT5 are there are not there yet.

 

[cuco33]

erm nope! Console racers look good but are still behind PC sims in many key areas. PC sims are MUCH better at the full race weekend, Also they simulate much more too, GTR2 is a game from around 2006 and factors in brake temperature, and you could blow the engine too. I havn't tried the stuff like iRacing either, and only had a go at the demo of NKpro, WAY more stuff going on. No matter how good F4 and GT5 are there are not there yet.

I agree that pc sims do it better but that gap that used to be there between console and pc is so small nowadays that one can prefer the better visuals on console to the better physics on pc. I personally think they are very close, clearly.not 100% there yet but too close for the pc elitists from laughing