Incorrect stock suspension setup on vast majority of cars

[szejok666]

I've written about this issue before on tuning forum, but I think it is a quite big problem so I've decided to create a thread about this.

The recent update (1.09) changed stock camber and toe setup on just about every factory stock car in GT6. It's been discussed already many times, that now stock values are quite radical (for instance I don't have a knowledge of a car, that has 0.60 rear toe-in as standard, it is just too much) and the same for cars with similar type of drive (I mean FR, MR, FF). This values can be however set correctly as we all know, by fitting full customizable suspension.
There is I think much bigger issue, that I've noticed recently - spring rates.
In GT6, factory fitted springs in vast majority of cars have different values than in reality. Often they are much stiffer. There is loads of data available on the web so I will post only a few examples:

1990 Nissan Silvia K's Dia Selection (S13)
GT6 Spring Rates: F: 3,12 kgf/mm R: 2,97 kgf/mm
IRL Spring Rates: F: 2,00 kgf/mm R: 2,00 kgf/mm
Source: http://zilvia.net/f/showthread.php?t=26680

1991 Mazda Eunos Roadster J-Limited (NA) '91
GT6 Spring Rates: F: 3,39 kgf/mm R: 3,07 kgf/mm
1991 Mazda MX-5
IRL Spring Rates: F: 2.94 kgf/mm R: 1.73 kgf/mm or F: 2.76 kgf/mm R: 1.68 kgf/mm
Source: http://users.telenet.be/miata/english/suspension/Springs.htm
http://fatcatmotorsports.com/FRC_1_6NA/FCM_MSDS_1_6NA.htm
This have been spotted previously by bread82 in this post:
https://www.gtplanet.net/forum/threads/stock-spring-rates-incorrect.295487/

All of the NSX's available in GT6 also suffer from this issue, You can compare them with this data:
http://www.nsxprime.com/wiki/Suspension
What's worth to mention is how softer is 2002 NSX Type R in GT6, than it is in reality.
Corvettes are also problematic in this case. In GT6, C6 ZR1 is much stiffer than C6 Z06. In reality it's the other way around.
Almost every car I've compared has different spring rates in game. The only one I've found so far, that has exactly same values is 1997 Toyota MR2 GT-S.

The most strange however is the fact, that in previous GT games these values were correct, or very close to real in most cases. I've even fired up GT1 to check this, and yes, they the same as in reality. I have also played GT5 recently to compare how the same factory stock cars roll in corners, and apart from completely new physics model, the GT6 equivalents are noticeably stiffer.
The question is:
Is it because PD messed with them for casual players (Cars come with much grippier tires in standard, so if suspensions were soft as in reality, stock cars would move in corners too much), or there is some more physical stuff included in stock spring rates by PD, so they behave more true-to-life?
Sorry for a long post, but I just can't understand this logic, and have no idea how these not even close-to-real settings works in GT6.

 

[beezball]

Great post, it would be a very good question to ask. Such an enormous and universal oversight would seem to be done on purpose. Either they changed the values on purpose like you said, or the physics engine is fractionally off from the real world and these values are adjusted to mirror their real life counterparts.

 

[g0ld]

Curiously enough, the car you mentioned having correct spring rates is (if I recall correctly, from other users' posts here) one of the hardest to drive to the point of being uncontrollable.

 

[Jimmy_Pop]

Some cars like the cobra r mustang cant even be adjusted to real life springs.

 

[TexasTyme214]

Could it be possible that PD has the incorrect motion ratios to some cars, leading to different spring rates, yet causing, generally, the right effect in handling? Also, some cars probably have the wrong weight distribution, leading to incorrect spring rates to make up for the data they probably couldn't get.

 

[akyp]

Quick question: Is the indicated spring rate in GT6 before or after applying the motion ratio?

 

[szejok666]

Curiously enough, the car you mentioned having correct spring rates is (...) one of the hardest to drive to the point of being uncontrollable.

Yes, that is a fact, and it's quite interesting. Does it mean, that being a car with correct spring rates, it's too soft in the game? Hard to say...

Some cars like the cobra r mustang cant even be adjusted to real life springs.

Actually that's a problem with most cars. Real factory values are much lower, than available to set in GT6 with full customizable suspension.

Is the indicated spring rate in GT6 before or after applying the motion ratio?

I really don't know, maybe someone with knowledge and understanding of GT6 physics could answer this.
After physics and suspension changes made for GT6, it's even more confusing.

Another thing that I've spotted is the change in lap times, that I can achieve, due to 1.09 update. I'm slower at Tsukuba by 0.5 to 1 second in a lap in most cars.

 

[nickg07]

Mr2 gts is fine in 1.09 I can take it to 550pp and it's controllable. I did the seasonals in it. But I did add a bit of ballast and went 48:52 weight distribution. I owned mr2 turbo for 3 years, is okay.

 

[schmocke]

First off: yes, you're right.
But did you already read the info screen which shows up before main menu of GT6!?
There you will see PD claiming, that in game cars can't be compared to real counterparts.
Why is that so? Because it's a game, no more - no less.
End of story.

Personally, I like the post 1.09 physics more than 1.08 and before ones.
Since I like PC gaming as well and playing Assetto Corsa and Race 07 right now I can absolutely assure you, that GT6's physics are a joke.
However, I like the game and will play it further, online with some buddys in streetcar lobbies.

 

[Cosbuster]

First off: yes, you're right.
But did you already read the info screen which shows up before main menu of GT6!?
There you will see PD claiming, that in game cars can't be compared to real counterparts.
Why is that so? Because it's a game, no more - no less.
End of story.

That's sort of like trying to hide behind your finger. No, a disclaimer doesn't make everything right. A stock Focus coming with a jetpack wouldn't be alright because there was a claim cars can differ from their real counterparts.

 

[bread82]

It's weird that the numbers for spring rate seem so precise, they're not just rounded to the nearest kgf.m or something. This leads me to think that there is a deliberate calculation going on which has some unknown fudge factors. As people have pointed out, usually motion ratios make the wheel rate(WR) less than the spring rate (SR), so it's not that WR are being quoted by PD. I have a couple of theories but as yet no way to prove/disprove them.

a) The default ARB are 1/1 (or 3/3 for FC suspension). In my thread (linked in OP) I wondered if the real ARB stiffness had been combined with the SR to simplify the suspension model. The reason I don't like this is that this would presumably make no differentiation between roll and pitch stiffness, for default settings.

b) PD use a pretend generic suspension geometry which has individual fudge factors to make it fit the RL geometry and dimensions for each car. The SR we see is the one that the PD model uses before applying the fudge factors to make it equivalent to RL.

This would have to be pretty strange to give the often higher than RL SR we see. I can't think of how this would work other than modelling the suspension as (PD default length very short wishbone, very stiff spring) = (RL longer wishbone with softer spring). Doesn't seem to make much sense to me. Why not let the user input the RL spring rate and then apply the fudge factors retrospectively behind the scenes? I wonder if there is a link between the difference in RL and PD SR, and car size or suspension type?

Why the change from GT5 which I believe had largely correct SR? Must be something to do with the new physics but it seems like a step backwards.

Edit: Random link to a data logger ride height and spring stiffness post I made some time ago. I've not had time to work on this since but it might help shed some light on roll bar vs spring stiffness.
https://www.gtplanet.net/forum/threads/data-logger-what-uses-have-people-found.306705/

 

[MOTORTRENDmitch]

PD was wrong about something!

in all seriousness, when I make a car to be completely realistic I find the suspension setup and try and replicate in on GT6, because it's pretty obvious some of the settings they gave cars were way off.

 

[Rotorist]

Yes, that is a fact, and it's quite interesting. Does it mean, that being a car with correct spring rates, it's too soft in the game? Hard to say...


Actually that's a problem with most cars. Real factory values are much lower, than available to set in GT6 with full customizable suspension.

I really don't know, maybe someone with knowledge and understanding of GT6 physics could answer this.
After physics and suspension changes made for GT6, it's even more confusing.

Another thing that I've spotted is the change in lap times, that I can achieve, due to 1.09 update. I'm slower at Tsukuba by 0.5 to 1 second in a lap in most cars.

Slower in tsukuba was a must. Before 1.09 we were able to smash RL laptimes. Now it is closer and cars are more prdictable and with much better feedback. But this topic includes a question for PD that is nice to have an answer.

 

[szejok666]

Before 1.09 we were able to smash RL laptimes.

When fitting more realistic tires, we weren't actually.
Most cars achieve most realistic lap times on comfort medium. I'm not the best driver in the world, but on this tires I could achieve nurb record times with no problem ('90 NSX - 8.16, '89 Skyline GT-R - 8.22, '02 NSX Type R - 7.56, C6 Z06 - 7.42). On Tsukuba however, it was impossible due to the fact, that the most records were done in winter, when temperature of track and air was about 1 Celsius (I think I've watched all Best Motoring shows, that I could find ). In these conditions times were about 1-2 seconds faster than in normal ones, let's say in spring time of the year. Conclusion is, that I couldn't achieve record times on Tsukuba wearing CM, but on all other real tracks represented in game, I could.

 

[Robbks]

Another thing worth pointing out,
Spring rates differ across different worldwide markets for the exact same car.
i.e JDM, AUDM, USDM, EUDM

For example,
Australian roads are generally rough, and bumpy, as such our cars have softer spring rates than the equivalent JDM or EUDM variant.
Aswell as us having lower highway speeds than other areas (110km/h or 65mph)
For higher speeds, they generally select a stiffer spring.

I'd probably expect the USDM cars to be set up softer than the JDM counterparts

 

[meek]

When fitting more realistic tires, we weren't actually.
Most cars achieve most realistic lap times on comfort medium. I'm not the best driver in the world, but on this tires I could achieve nurb record times with no problem ('90 NSX - 8.16, '89 Skyline GT-R - 8.22, '02 NSX Type R - 7.56, C6 Z06 - 7.42). On Tsukuba however, it was impossible due to the fact, that the most records were done in winter, when temperature of track and air was about 1 Celsius (I think I've watched all Best Motoring shows, that I could find ). In these conditions times were about 1-2 seconds faster than in normal ones, let's say in spring time of the year. Conclusion is, that I couldn't achieve record times on Tsukuba wearing CM, but on all other real tracks represented in game, I could.

another thing upside down in GT6 if so... (you run faster laps the hotter it is).

Do you know why cars are faster when it's colder, by chance?

 

[Robbks]

engines "generally" make more power when it's colder,
but the tracks might produce more grip when hotter....

 

[Le_Manuz]

when it's cold, the air is more dense, so you have more combustive air,
the calculator put a bit more of benzine, the result is that you have more power
the opposite is valid
and it can also be extended to the altitude
at see level, there is more air, at the top of the mountain there is less

so at sea level in a cold weather, you have the max power available
on a mountain in a hot weather, you have the less power available

that said, back on topic
i don't know were they took the specs from but ... after 15 years, it's still not really accurate

PoDi, i personally think it's easier to work on a new Physics Engine when you have the correct numbers ...
and easier to work on a standardized unit system than using all of them

look at the loss of Mars Observer, Mars Climate Orbiter and i think there is a few more
even the best make mistakes

 

[Wide Boy]

Does PD use a fudge formula that adjusts the spring rates according to the type of suspension fitted to the car? I have no idea if it models different suspension types but doubt it does more than live axle & independent.
Also with active suspensions as fitted to some modern cars have they modelled the sport/track setting.
There are also for sure some hidden "settings" we don't see though whether these impact on suspension I have no idea.

On cold air as said density is the key, more oxygen to burn in the same volume, mist increases power as its fine enough but rain doesn't. Or not noticeably on anything I have driven anyway.

 

[meek]

engines "generally" make more power when it's colder,
but the tracks might produce more grip when hotter....

that makes sense, temperature in GT games is probably the track's (I always wondered, it seems low for some), and is not off, then.

i don't know were they took the specs from but ... after 15 years, it's still not really accurate

PoDi, i personally think it's easier to work on a new Physics Engine when you have the correct numbers ...
and easier to work on a standardized unit system than using all of them

very true

 

[Robbks]

and easier to work on a standardized unit system than using all of them

look at the loss of Mars Observer, Mars Climate Orbiter and i think there is a few more
even the best make mistakes

One day the U.S. might catch up with the rest of the civilised world and use SI units....

 

[meek]

One day the U.S. might catch up with the rest of the civilised world and use SI units....

one day, maybe... hopefully get rid of the am/pm system too!

 

[szejok666]

I would like to ask someone to post here stock spring rates values for couple of cars from GT5. Unfortunately my PS3 laser is broken, and I can't check it by myself. Maybe there is some pattern, because these suspension numbers in GT6 really seem very precise, and I would like to compare them with these in GT5. Cars I'm interested in are:
1999 Dodge Viper GTS, 1997 Toyota Supra RZ, 1990 Honda NSX, 1991 Nissan Silvia S13 K's, 2002 Honda NSX Type R, 2010 Chevrolet Camaro SS, 1997 Toyota MR2 GT-S.
I will be very thankful for help.

 

[szejok666]

First of all, sorry for double post.

But since the last one I wrote, I've done some research and testing, and my conclusions aren't very optimistic.
The spring rate visible in GT6 suspension setup is an universal figure from what i have experienced. In simple words, if spring rates are soft, car will be soft, nevermind the model, or type of suspension it has. And it's actually a great thing, because the suspension model alone seems to be working quite good. We don't have to wonder what type of suspension we have in car, we only have to know values, set them on car in question and everything seems to be working very well.
But actually, we can't set real, true to life figures in most of the cars. Stock ones are; like I've said in previous posts; often much too stiff, and even with fully customizable suspension, we can't set the values as low as they are IRL.
If You think, that only the numbers shown in GT6 suspension setup are wrong, but the whole system and suspension model works good on these wrong numbers, please compare it for yourself. Here's a great example:
Buy 1990 Honda NSX (japanese model to be exact) and fit comfort medium tires. Then go to Tsukuba track, disable all aids except ABS and run a few laps to achieve a time of 1.09-1.11 per lap. Feel free to drift on the last corner. After that follow this link:

Please watch it, then play a replay of your free ride, and compare how the exact same car behaves in corners. The fact is, that virtual recreation is much, much stiffer. This of course has it's translation in figures:
Real NSX has Front Spring Rate of 3.04 kgf/mm and Rear of 3.94 kgf/mm.
GT6 counterpart has F: 5,64 kgf/mm R: 7,97 kgf/mm
Conclusion is, that the numbers PD set as stock are horribly wrong, and this problem affects almost every stock car in this game.
Maybe I'm overreacting to this, but it is for me the biggest and the most critical flaw this game actually has. At least over 3/4 cars in game are handling, like they are on tuned, very hard spring straight from the factory. I know PD has done many things not quite perfectly before, but this one is VAST and affects most of the cars available in game. I'm thinking about posting a new thread about this problem on the main board, but I'm worried that hardly anyone will notice.

 

[FSportIS]

Wow, I support your research. I like to run stock set up myself as it is simple and the cars are more real life. Interesting finding. Please keep up your good work. I think you should contact PD directly on this. Their programming team may input wrong values or something.

 

[szejok666]

Thank You very much I really would like to contact Polyphony Digital about this issue, but I really don't know how. Also, I'm not feeling that important to tell this to Them directly Maybe there is someone on this forum who feels, that he/she can do this, and knows English language better than me

Another interesting fact is, that in GT5 as far as I remember most values were real, but not all of them. Premium Acura NSX from 1991 was much stiffer than standard 1990 Honda NSX. Similar thing was with Viper GTS ('99 standard and '02 premium). And strangely, the premium ones behaved unrealistically compared to the videos I've seen in the net, not the standards. Also I've just remembered that premium 1970 Challenger R/T was also very stiff. Other premiums that I remember for now had real life values for spring rates.
It looks like for GT6 most of cars were stiffened, instead of being softened to behave much more realistically in the corners.