A test of GT5's physics model

TouringBubble

TouringBubble

601
ProjectTuning
The other night I decided to test out GT5's physics model by taking my car, an Evo IX MR, and setting it up as close as possible to my specific specs and compare it to my real car's handling. These are the settings and results.

GT5 HP set to approximately 430hp using various methods. My car is at around 400 whp, with most power coming from increased boost on the stock turbo and quality fuel. I had to use the stage 1 turbo upgrade to match this power, though mine is still stock. The GT5 car has more HP due to differences in flywheel and wheel HP measurements (drivetrain loss).

No weight reduction. Stock transmission. Sports flywheel and clutch. I have a lightened stock clutch and upgraded sprung 6-puck clutch.

Sport Soft tires. I autocross my car on Kumho XS tires, which are about as sticky as you can get without going to an actual race tire. However, GT5's Sport tires seem to be more in line with "street" tires used in actual race series, such as the Continental Tire series, which are a lot softer. So, a Sport Medium may have been a better choice in retrospect.

Fully Adjustable Suspension. My car has upgraded springs and sways with camber adjustment added. I could not get my actual spring rates using the height adjustable option, so I had to go a level higher.

Car lowered 10mm F and 6mm R. -2.7ºC F, -1.3ºC R, 0 toe F and R. Front springs at 5.2k (290 in/lb) and rear springs at 5.9k (330in/lb). Shocks left in the center position. Sways adjusted at 5F and 4R (I believe ... can't remember). The sways and shocks are an estimation, but the rest of these values are my actual suspension settings.
Click to expand...

I tested the car as Tsukuba using a G25 wheel and pedals.

Amazingly, the handling was very similar to my real car. It was very neutral in corners and responded to throttle modulation and brakes just as my car does. It seemed to have a little more grip overall than my car, but that could have been the tire choice mentioned above or even that I was doing this test in practice mode and not online.

Overall, I'd say that GT5 is 100% the best simulation of an Evo that I've tested. Many people say that PC sims like iRacing, GTR2 and rFactor have much better physics, and they may be right. But, with PC sims like rFactor and GTR, you have to rely on 3rd party contributions for new and fresh cars, and they are not all equal. Many have no where near realistic handling. Live for Speed only comes with fictional cars, and iRacing only has race-prepped cars to run.

I would feel comfortable saying that, having tried several sims on the PC, that GT5 is definitely the best simulator of production car driving and modifying on the market. With the addition of mechanical failures, it would be a serious competitor for the hardcore sim titles.

That said, I have not tried Forza with a proper wheel setup. I'm comparing to LFS, GTR2 and rFactor mainly. iRacing and the Race series, as well as GTR2, have great version of race cars ... but simply can't represent a street car like GT5.
 
cool to be able to test it this way.
I you take your car to the track sometimes you could turn the test around and use GT5 to work on a setup for RL.
 
To a degree you could, bit it would be very limited.

The issue is that some values are fairly arbitrary, such as the damper settings and sways. Dampers do have number settings, but different manufacturers use different ranges of numbers, so it's all a wash. Also, camber settings and such are set in real life based on wear patterns and temperature differences across the tread blocks, which you can't measure in GT5.
 
Always nice to see these GT5 vs. real world tests. Although there are far too many variables to make any of these a totally accurate like-for-like comparison, the basics are there, and look to be generating some good results.

There is a great video posted elsewhere on GTP of a BMW M3 driving the 'ring and the likeness is amazing. Same can be said for a similar video of the Ferrari F1 car around Monza. Not only are the tracks very close to real-life, the cars behave the same too.

LFS actually does have some real world cars, though not many. I agree that GT5 strikes a great balance of road cars, race cars and specialist cars.

Allow me to put your mind at rest about Forza. I did enjoy Forza 2 with the MS wheel, but got the chance to try Forza 3 recently at a friend's house and it is no match for GT5. The cars do not handle well and do not feel realistic to me. Likewise, I have experience in LFS, GTR2, rFactor and netKar Pro: presently, Forza is a long way off these. This is just my own critical opinion. If I found Forza 3 better than GT5, I would say so.
 
VBR
If you wanna get the tyres right, try this:

https://www.gtplanet.net/forum/showthread.php?t=160821&page=5#post4638146


👍
Click to expand...

That's actually really good data. It's not quite in line with what I found though.

Logs show that I pull around 1.1g - 1.25g during an autocross. That would put me right at a sports soft tire according to that chart, which felt a little too grippy in GT5. Maybe if I tested it online the results would line up? He did state that modded cars kind of skewed the results, so maybe that's it.
 
Rhino_WJB
Allow me to put your mind at rest about Forza. I did enjoy Forza 2 with the MS wheel, but got the chance to try Forza 3 recently at a friend's house and it is no match for GT5. The cars do not handle well and do not feel realistic to me. Likewise, I have experience in LFS, GTR2, rFactor and netKar Pro: presently, Forza is a long way off these. This is just my own critical opinion. If I found Forza 3 better than GT5, I would say so.
Click to expand...

I played Forza with the Microsoft wheel as well ... but to be fair, the wheel sucks. But, from what I gathered at that point, it seemed more in line with Shift than a sim racer. Kinda that middle ground where you think it's sim-like until you realize you can beat understeer by just adding more steering ... =)
 
TouringBubble
That's actually really good data. It's not quite in line with what I found though.

Logs show that I pull around 1.1g - 1.25g during an autocross. That would put me right at a sports soft tire according to that chart, which felt a little too grippy in GT5. Maybe if I tested it online the results would line up? He did state that modded cars kind of skewed the results, so maybe that's it.
Click to expand...

Yeah, maybe all the tuning & settings gives the car a little more grip in GT5 compared to reality, a well set up LSD certainly would certainly do that. Try turning the car in a circle on The Top Gear Test Track & see what the maximum lateral force is, if it's higher than your autocross logs then try the next tyre compound down in GT5 to compensate for the unrealistic extra grip.


👍


Good thread by the way.
 
That would be a good comparison ... thanks for the idea. I wanted to do skidpad tests, but wanted a real skidpad. Why don't we have one!?
 
TouringBubble
That would be a good comparison ... thanks for the idea. I wanted to do skidpad tests, but wanted a real skidpad. Why don't we have one!?
Click to expand...

Dunno, there's a hundred & one things we don't have in GT5 that we easily could have if PD pulled their finger out, let's all just keep voting in Jordans exellent new Feedback section & hope they listen. Personally I want bump data in all tracks so that some are not just dead flat!

http://feedback.gtplanet.net/forums...-tracks-should-have-bumps-in-the-road-not-be-

As for skidpad tests, driving in circles before of after the tyre wall on The Top Gear Test Track is adequate & is what calan_svc used for his testing.


:)


EDIT: Also, having aero parts will increase Lateral G Meter readings in high speed corners in real life & in GT5 the effect may be a lot different than in reality. Just another thing to consider when doing reality vs GT5 tests.
 
Last edited:
TouringBubble
The other night I decided to test out GT5's physics model by taking my car, an Evo IX MR, and setting it up as close as possible to my specific specs and compare it to my real car's handling. These are the settings and results.



I tested the car as Tsukuba using a G25 wheel and pedals.

Amazingly, the handling was very similar to my real car. It was very neutral in corners and responded to throttle modulation and brakes just as my car does. It seemed to have a little more grip overall than my car, but that could have been the tire choice mentioned above or even that I was doing this test in practice mode and not online.

Overall, I'd say that GT5 is 100% the best simulation of an Evo that I've tested. Many people say that PC sims like iRacing, GTR2 and rFactor have much better physics, and they may be right. But, with PC sims like rFactor and GTR, you have to rely on 3rd party contributions for new and fresh cars, and they are not all equal. Many have no where near realistic handling. Live for Speed only comes with fictional cars, and iRacing only has race-prepped cars to run.

I would feel comfortable saying that, having tried several sims on the PC, that GT5 is definitely the best simulator of production car driving and modifying on the market. With the addition of mechanical failures, it would be a serious competitor for the hardcore sim titles.

That said, I have not tried Forza with a proper wheel setup. I'm comparing to LFS, GTR2 and rFactor mainly. iRacing and the Race series, as well as GTR2, have great version of race cars ... but simply can't represent a street car like GT5.
Click to expand...

Nice! thanks for your report.

My biggest gripe is what appears to be the ease of cars going offtrack into a gravel pit or grass and not losing any performance. very sad, but I think this is a compromise by PD to keep everyone in the race.
:indiff:
 
I've noticed a definite drop in speed in the gravel ... especially online. It seems to be worse with lower cars. But, it's no where near like hitting the gravel in real life ... you know, where your wheels sink in 5" and your bumper is ripped off while you slam your head in to the steering wheel ...

But, to be honest, a simulator can't actually use realistic gravel since there is no tow truck to pull cars out. If gravel was realistic, then half the field would be stuck in there with no way out. It could be a little stickier and still work.

Grass is done pretty well though.
 
Gravel is definitely a compromise, but I'm perfectly fine with it. I'd rather not spend five minutes trying to crawl through it back onto the track.

This is a great report. My Supra MkIII is stock, and while I've never autocrossed with such a heavy car, I have taken a few runs through rural country roads at night, and the GT5 version does handle remarkably similarly with hard Sport tires.

I'm not sure Turn 10 will mess with the Forza formula much more. While I don't find it to be as realistic, not having the flavor of a PC sim like I do even with Prologue, it is a blast to throw cars around turns way fast. If they toned that down and made drifting harder, I think the kids would throw a fit. So we could end up with a decent competition between the two series, with Forza being the more fun game and GT the more realistic.
 
I agree. Competition is great for the industry. Let each have their own niche but fight for attention. It's all better for us.

And I'd advise against Auto-Xing your Supra. I had a friend take his out once with ~800 miles on a rebuilt motor and a rod baring gave up due to oil starvation in corners. The MKIII is bad about that. Not sure if you can get a baffled pan for those. His was the 4AGTE or whatever I think ... the MKIII turbo engine.
 

Latest Posts

Back