S2U Community Physics Experiment

[ermo]

@offom:

Sounds interesting. I've mentioned it before, but in my experience, stiffening the S2U tyres to between +50 to +75% of the original stiffness is a pretty good compromise between stiffness and controllability.

Is that something you've considered testing out for yourself with the NTM mod?

EDIT: Street_1 has a SelfAligningStiffness=500 at the rear, while the front is 6000. Typo? The CG adjustment may or may not be a good idea. Only testing will tell, I suppose.

[ermo]

Originally Posted by dequaded

I found some bug in Apollo suspension ride height - RideHeightRange=(0.080, -0.001, 100). That means if value will be 0 then suspension height will be -20. Car body will be lying on the road.
There is some official data from Apollo site "ground clearance adjustable between 40 and 120 mm". It may be useful.

Right you are. This goes for all RideHeights < 100mm actually -- good catch!

[offom]

Originally Posted by ermo

@offom:

Sounds interesting. I've mentioned it before, but in my experience, stiffening the S2U tyres to between +50 to +75% of the original stiffness is a pretty good compromise between stiffness and controllability.

Is that something you've considered testing out for yourself with the NTM mod?

EDIT: Street_1 has a SelfAligningStiffness=500 at the, while the front is 6000. Typo? The CG adjustment may or may not be a good idea. Only testing will tell, I suppose.

Since Shift 2 reflects rear grip's % 40 to the steering wheel, I reduced rear's self aligning stiffness. It got nothing to do with the steering feel as I am turning both 4 wheels, it just straightens out the car a bit. I sacrificed it for more alive force feedback.

Yes, I implemented the changes that I told, CG multiplier is changed to 1.

[ermo]

Originally Posted by offom

Since Shift 2 reflects rear grip's % 40 to the steering wheel, I reduced rear's self aligning stiffness.

How/where did you determine that? And based on what is this a good idea? (Just curious)

[offom]

Originally Posted by ermo

How/where did you determine that? And based on what is this a good idea? (Just curious)

If you change controllerdefault xml's writeable and overwrite them in the game menu, exe writes its coded values into the xml. Gripweight is defined as 0.6, this is the percentage of front grip through the steering wheel. I don't just want to feel rear's self aligning force.

[offom]

Hi all again, I corrected out front right suspension problem on Lotus Elise and Exige.

and nearly all the cars have 3.0m wheelbase...

Here is "some" corrected/meaningful suspension geometries.

[dev_null]

Any chance of having offom's suspension work included in the UCP?

It would be nice to get solid improvements that would help everyone in the UCP.

That way offom's NTM can focus on tires, and the suspension fixes can stay within the UCP.

Just an idea, as it seems like there are several different improvement threads.

[ermo]

You and offom have sort of arrived out of the blue, so we (the existing community of S2U modders) need to get to know you and him and your work before deciding whether we trust you to generally do the right thing most of the time. Call it inertia, call it skepticism or just call it human nature.

IMHO, the guiding principle for physics changes is that they have to make sense and that people should explain what was changed and why in such a way that other people can understand it.

If offom's changes make sense and if other people get a chance to look at his work in this thread and don't think it looks totally strange, then we can look at including it in the UCP after a certain grace period where other people get to comment on it.

How does that sound?

[offom]

It sounds right. I want my tire mod be seperated from UCP.
But
Suspension geometry changes can be in account.
I'm an automotive engineer and trying to be sure what I'm doing

[ermo]

Originally Posted by offom

It sounds right. I want my tire mod be seperated from UCP.
But
Suspension geometry changes can be in account.
I'm an automotive engineer and trying to be sure what I'm doing

Well then, sounds like it would be prudent to take your comments and suggestions under serious consideration at least.

In your educated opinion, to which degree are the stock suspension files and chassis settings sound? Does everything look half-baked as it is, or is there a method to the madness? I'm far from an automotive physics expert myself -- I just try to integrate bug fixes and tweaks into S2U as best I can.

The goal with the CPE is to experiment and see what kind of behaviour we can get out of S2U when carefully calibrating the physics input data.

Whether any proposed suspension changes should go into the UCP is probably a bit premature at this point. Why not just rework a few cars at a time and take it from there?

[offom]

There are real suspension geometry design problems. We know that Shift 2's materials are shared among too many games.
There are some problems as:
Undertray parameters
Pushrod body - pushrod suspension connection points
... etc

I've been playing SU2 for years and finally came up with modding idea.

Suspension designs are so wrong, if I can edit them (testing on Lotus) to be realish, we are on the book of honour. Trust me. This suspension corrections will change all the behaviour of the game in a realistic way.

I already edited tires for real data and for game implementation, I like cornering stiffness values, some are a bit high while some are a bit low but generally I will not touch them to make game "sharp" .

Let me just say that for example:

Cars have reversed kingpin inclination response (reason of floaty drive seen on most of the games)...

[dequaded]

2 ermo
I think you need reedit all ride height and height range. I tweak Porshce GT2 today and figured out that the RideHeightRange=(0.115, -0.001, 100) is not actually 11.5cm but 12.5cm and 12.5cm is too low to be of maximum for GT2. I read at forums that actual normal ride height for GT2 is 110mm front and 130mm rear. And methinks 100mm tweak is too much for stock cars. For GT2 i use that settings
Front
RideHeightRange=(0.130, -0.001, 70) //70-14.0 30-10.0 0-7.0
RideHeightSetting=30 //Ingame - 70(0.130(140mm))-30(30mm)=40(110mm)
Rear
RideHeightRange=(0.135, -0.001, 70) //70-14.5 30-10.5 0-7.5
RideHeightSetting=15 //Ingame it will be 70(0.135)-15(0.015)=55(0.120)

[offom]

This is how SU2 makes calculations...Bah...

[MEFF1091]

waiting for your suspension mod really. i always tried to change the suspension behaviour by myself but no result if the game will change to realism with your mod, so let it be done then

[ermo]

Originally Posted by offom

This is how SU2 makes calculations...Bah...

What is your working hypothesis? What is causing this? If you post the changes you made (in text or native file format) which leads to the behaviour in the screenshot, JDoug might be able to shed some light on it as he has a lot of experience with how MADNESS Engine tech treats the numbers...

[offom]

Originally Posted by ermo

What is your working hypothesis? What is causing this? If you post the changes you made (in text or native file format) which leads to the behaviour in the screenshot, JDoug might be able to shed some light on it as he has a lot of experience with how MADNESS Engine tech treats the numbers...

Should I send a private message to JDoug ? How we can implement kingpin inclination on the front axle, this is my "just one" question.

Front wheels does not turn according to the spindle axis, but somehow on some cars it acts slightly reversed.

[offom]

@JDougNY

Thanks for information indeed, but cars do not act like there is kingpin inclination, just try that, shift upper arms inner for 300 mm (!), and relocate spindle/wheel joint for new geometry. Due to the new (and exaggratedly wrong) geometry you would expect to lift car's nose up when turning steering wheel lock-to-lock. But it does not happen. Also, changing joint axis, shifting wheel position, adding caster offset etc does not effect wheel's stand-still turning behaviour (they must). They just change driving dynamics.

After all those geometry editing, wheel still turns on the vertical axis like (0.72,0.0,0.0) because of joint&hinge definition. Wheel must have some relative motion to the other axises too. (I assume there is zero caster and zero camber)

[offom]

Ok, I noticed that there are some different things to do too, thanks again JDoug

[ermo]

@offom:

A guy called Andrew Weber (nickname AJ) is currently heading up the pCARS steering rack and FFB development. He mentioned in passing after inspecting Eero's rack code that Eero's rack (the rack in S2U) doesn't take KingPin Inclination/SAI into account.

Not sure if that rings a bell with you, but I assumed that AJ's remark was made in relation to the forces felt via the FFB that are picked up from the rack.

@JDoug:

pCARS is upping the physics tick rate to 600 from the stock 180 in S2U in order to combat some integration step issues that were partially responsible for causing cars to veer left/right under braking and when going over bumps using AJ's new, more 'pure' FFB/Rack code. I'm considering upping the frequency to 360 in S2U as I noticed you have decoded the corresponding value in physicstweaker.mrdf. Do you have any opinion on this?

[Kazumi]

Well, yeah... moving the upper arm to the inside like that will move the inner wheel back while turning. Should work.

[offom]

Originally Posted by ermo

@offom:

A guy called Andrew Weber (nickname AJ) is currently heading up the pCARS steering rack and FFB development. He mentioned in passing after inspecting Eero's rack code that Eero's rack (the rack in S2U) doesn't take KingPin Iinclination/SAI into account.

Not sure if that rings a bell with you, but I assumed that AJ's remark was made in relation to the forces felt via the FFB that are picked up from the rack.

@JDoug:

pCARS is upping the physics tick rate to 600 from the stock 180 in S2U in order to combat some integration step issues that were partially responsible for causing cars to veer left/right under braking and when going over bumps using AJ's new, more 'pure' FFB/Rack code. I'm considering upping the frequency to 360 in S2U as I noticed you have decoded the corresponding value in physicstweaker.mrdf. Do you have any opinion on this?

Originally Posted by JDougNY

@Ermo

In Shift 1, with 15 AI cars on the track, the upping of the tick rate to 360 made my car more responsive. It wasn't as noticeable in a solo race event. DJO had mentioned once that the tyres reacted differently and not to his liking at tick rate 360.

If you want to test it out, try a solo race then a race with 15 opponents at tick rate 180. Then switch to 360 (restart the game) and try the races again to see if there's a benefit...or if the side effects are creating new problems.

I already increased tick rate to 360 in my NTM mod 1.8 (racing games have it around 360 - 400), it essentialy smooths fast driving ride, makes the feel responsive.

With realish tire datas tick rate increase works well I think.

Anyways what is "pCARS" ??

[MEFF1091]

pcars is our new project. sms and we (community) working together to make the ultimate sim come join us

[offom]

Originally Posted by MEFF1091

pcars is our new project. sms and we (community) working together to make the ultimate sim come join us

Great to hear that we both are in Turkey. Count me in !

[ermo]

@offom:

As I understand it, the S2U physics engine has a tickrate of 180, but is substepped x2, so the effective physics update frequency is 360Hz. Upping the tickrate to 360 is effectively 720Hz. I'm not sure if AJ's recent change caps the tickrate at 600Hz or if it is really 1200Hz. He did mention doing x2 substepping when tyres are spinning at <40 Revolutions Per Second.

pCARS is shorthand for Project C.A.R.S -- Community Assisted Race Sim.

If you do join, you should get in touch with a guy called speed1 -- he and you would hit it off, I think.

But I'm sort of liking the idea of you, JDoug and Kazumi exchanging ideas and experience on your S2U suspension tinkering, so I'm not sure if I look forward to either of you forgetting S2U for pCARS ...

[offom]

@ ermo

I think phsyics tick rate means calculation frequency, like time step in explicit analyses. So using both high or low steps may result in inaccuracy or divergence problems. Optimal value seems as 360. But it is not neccessary to calculate 360 loop per second at the low speeds. So, substepping at lower speeds is meaningful.

my new mod will be on just suspension geometries for SU2

[Chenuka]

hey guys!!! I've made an Enzo mod...and I need the assistance of a genius to help me with custom physics....believe me guys I've endlessly tried to dabble in making it on my own replacing them with duplicates of Maserati MC-12 physics (which apparently is a similar ride under the hood) so believe me I wouldn't be asking if I had no other option. can someone collaborate with me on this?? For a share of the credit of course. I will send you the files..

[offom]

Dear friends, I came to conclusion that I will not be a part of pCARS project. Just for now I will work on SU2 subject. Thanks.

[Kazumi]

Originally Posted by ermo

@offom:

As I understand it, the S2U physics engine has a tickrate of 180, but is substepped x2, so the effective physics update frequency is 360Hz. Upping the tickrate to 360 is effectively 720Hz. I'm not sure if AJ's recent change caps the tickrate at 600Hz or if it is really 1200Hz. He did mention doing x2 substepping when tyres are spinning at <40 Revolutions Per Second.

(...)

But I'm sort of liking the idea of you, JDoug and Kazumi exchanging ideas and experience on your S2U suspension tinkering, so I'm not sure if I look forward to either of you forgetting S2U for pCARS ...

- I just noticed I had S1 at 200 for a while now. rFactor uses 400 so that's the highest I find for chassis / general physics, but just ~ 11% higher then 360. Well, maybe except Ring of Rods as they are doing crazy things with body deformation (but leave out some other things that are important in racing)

I also remember that's one of the first things in modding S1 someone tried, but unfortunately it didn't cure all issues!

- I won't 'leave' either at least not for that project

[offom]

Hi again, I "feel" that inertia values under chassis files are a bit high, what do you think ?

[ermo]

From the inertia values that JDoug has calculated using his own method, I think he agrees that the values in S2U are a bit high...

[ermo]

Thanks Joe!

That spreadsheet should come in handy for many a curious soul (including yours truly).

Just thinking out loud here, but isn't it fair to suggest that if someone were to measure the inertia of a real car, say, 10 times, they'd end up with roughly the same inertia measurement values, suggesting that there really is one true mathematically correct and sound method for this?

And perhaps the issue is not so much the formulas (assuming that there's a limit to how many variations there can be on this) as it is a question of digging up the relevant measurements and dimensions with sufficient accuracy and plugging them into the 'correct' (set of) formula(s)?

I know you well enough to understand that you have your reasons for not sharing your formulas. But would you be willing to comment on someone else's methods in terms of whether they align with yours?

[tulib202]

What about using kangaloosh car factory to determine the inertia values??
That's what i do for my mod. The utility calculates inertia using wheelbase, front track, rear track, weight values and engine position of the car.

[ermo]

Joe,

In SHIFT, the Cobalt was really, really awful to drive. In S2U it's not quite as awful, but still not as good as, say, the Civic Si.

Have you given it (and any other FWD cars) a thorough overhaul in your own game, including suspension mass, suspension links, car mass and inertia values? Or have you concentrated on the more powerful RWD cars so far?

I ask because I have a hard time understanding how it can drive so differently from e.g. the Civic Si and thought that maybe you would know why...

[ermo]

Joe,

When I reflect upon your posts about your investigations, it's hard not to wonder what sort of deadlines and quality control goals EA and SMS agreed to in their contract. To be honest, I think EA mismanaged the SHIFT brand, when in reality they had a golden opportunity to create a brand able to compete with Forza and GT5.

The stated goal of the CPE is to see where picking the low-hanging fruit will get us, though I am aware that the iteration of the MADNESS Engine tech used in S2U has known flaws (acknowledged and subsequently fixed in the ongoing pCARS development I might add) and that, as such, there are limits to what we as modders can achieve given S2U's underpinnings.

SMS have categorized S2U as an 'Action Racer' in the pCARS Game Design Document, even though they are very adamant that the MADNESS Engine tech is capable of competing with iRacing/NkP/LFS et al in terms of how advanced the simulation code is. And on the basis of the evidence from pCARS (in the form of software deliverables and developer commentary), I'm inclined to take that at face value with the proviso that SMS are currently implementing various updates in the physics and FFB system in preparation for AJ's new dynamic tyre model landing probably a few months from now.

So while S2U still has some life left in it, I don't think we should deceive ourselves into thinking that we can woo the hardcore sim audience or one-up pCARS/NkP/LFS/iRacing. But on the evidence so far, I do think that S2U can be made to feel better than it did when patch v1.02 was delivered.

P.S. Yes, I'm interested in the cobalt files and I'll get in touch via e-mail shortly. I'm in the process of building a new PC and I'm helping my GF set up a webshop for her store, so everything is a bit up in the air atm.

P.P.S I'm wondering if it would make sense to begin publishing screenshots of the 'as shipped' vs. 'overhauled' 3-D suspension models to give people an idea of just how crazy the suspension data appear to be and why we're bitching so much about it? Or would that be poor form?

[tulib202]

Originally Posted by ermo

Joe,

When I reflect upon your posts about your investigations, it's hard not to wonder what sort of deadlines and quality control goals EA and SMS agreed to in their contract. To be honest, I think EA mismanaged the SHIFT brand, when in reality they had a golden opportunity to create a brand able to compete with Forza and GT5.

The stated goal of the CPE is to see where picking the low-hanging fruit will get us, though I am aware that the iteration of the MADNESS Engine tech used in S2U has known flaws (acknowledged and subsequently fixed in the ongoing pCARS development I might add) and that, as such, there are limits to what we as modders can achieve given S2U's underpinnings.

SMS have categorized S2U as an 'Action Racer' in the pCARS Game Design Document, even though they are very adamant that the MADNESS Engine tech is capable of competing with iRacing/NkP/LFS et al in terms of how advanced the simulation code is. And on the basis of the evidence from pCARS (in the form of software deliverables and developer commentary), I'm inclined to take that at face value with the proviso that SMS are currently implementing various updates in the physics and FFB system in preparation for AJ's new dynamic tyre model landing probably a few months from now.

So while S2U still has some life left in it, I don't think we should deceive ourselves into thinking that we can woo the hardcore sim audience or one-up pCARS/NkP/LFS/iRacing. But on the evidence so far, I do think that S2U can be made to feel better than it did when patch v1.02 was delivered.

P.S. Yes, I'm interested in the cobalt files and I'll get in touch via e-mail shortly. I'm in the process of building a new PC and I'm helping my GF set up a webshop for her store, so everything is a bit up in the air atm.

P.P.S I'm wondering if it would make sense to begin publishing screenshots of the 'as shipped' vs. 'overhauled' 3-D suspension models to give people an idea of just how crazy the suspension data appear to be and why we're bitching so much about it? Or would that be poor form?

That's not very optimistic, still going to try though.
10 manufacturers left till I'm done with the cars. If not with physics i can make it a heck of a racing experience with the events and the improved AI
Anyway what's up with the lack of any specs for teh shelby terlingua mustang. The only thing i can find is it's hp

[tulib202]

Oh my thank you how did you find that?

[dequaded]

Originally Posted by ermo

P.P.S I'm wondering if it would make sense to begin publishing screenshots of the 'as shipped' vs. 'overhauled' 3-D suspension models to give people an idea of just how crazy the suspension data appear to be and why we're bitching so much about it? Or would that be poor form?

Will be very interesting to see and know what are going on.

[offom]

Do I have a right to upload edited suspensions ? Are we creating another mod named CPE or is this thread just for discussing and sharing opinions ?

@JDougNY

Can you take a look at suspensions in my NTM Tire mod ?

[ermo]

@offom:

I'm maintaining the CPE mod, which is currently not available for download because my UCP work took precedence and the earlier CPE mod is not compatible with the new UCP version -- so for now the CPE only lives on my PC.

As JDoug posts, this thread is meant to be a focal point for discussing the calibration of the physics data in S2U and to test and experiment with any and all changes.

So if you have things that you are working on and tinkering with and you want to collect feedback from people interested in the physics, by all means do publish them as an easily testable add-on to the UCP either as a .zip here or in the form of your own NTM mod. The CPE is meant as a proving ground for ideas and tests which could later be either merged back into the UCP or be used in a separate overhaul style mod based on the UCP v1.1 foundation.

I'm still hoping that the CPE can be the physics calibration mod for S2U, but only time will tell how that will pan out.

[Kazumi]

Originally Posted by offom

Hi again, I "feel" that inertia values under chassis files are a bit high, what do you think ?

They are not that much on the higher end of realistic values. But of course I can't know what you feel or what you want to achieve and there's also aerodynamics to include as it will change the "feel" of rotational inertia (or more precisely the rotation itself)

Also if you get more responsiveness from lower inertia, you also get less drivability on slight sideways controlable (10 - maybe 18% ) slip and weight transfer. That's why mid-engine cars with low inertia feel so good while driving but they can bite you !

If I'm correctly you increased the CoG globally in your mod so with the current set-up that also means a loss of responsiveness / agility...? I'm not saying that's wrong but it kinda changes everything with regards to suspension, roll center and so on.

Originally Posted by ermo

P.P.S I'm wondering if it would make sense to begin publishing screenshots of the 'as shipped' vs. 'overhauled' 3-D suspension models to give people an idea of just how crazy the suspension data appear to be and why we're bitching so much about it? Or would that be poor form?

Mhh I don't think the wrong layouts itself described by JDougNY create that much "real" issues here. It just results in an inaccurate car compared to the real counterpart and, unfortunately, some ingame suspension changes make less sense (toe, camber) especially when you not expect that layout.

[Kazumi]

No, dude

My point is in current UCP and the vanilla game to some extent it works out in a gamey sense. There are no real drivability issues when you use aarm or strut suspension on FWDs as you could build the same car, it just handles more weird then expected. Maybe not with upgrades incorporated... the last FWD I tried was the Scirocco and the car didn't had much issues with the default suspension.

I'm not saying it couldn't be better . I actually said the ingame (or other things which can't be accessed ingame) set-up behaves weirdly when you expect an accurate car but it is on a layout not really suited for that car.

But it's kinda funny you say this after you said that you're not calling S2U a sim

If it's not what's the point in fitting suspensions.

You can't build a lot of those cars accurately in say, Live for Speed in terms of suspension layout, driveline, maybe other things. There are always limitations. From my point of view the cars build in a sim should work out in real life. That does includes fitting suspension layouts... but of course that also means you - or SMS in that case - can build cars that won't ever leave factory in that form. I don't know why one would want that, except it kinda drives "better" for them

If you want an answer just look at the poll on what's #2

[Kazumi]

Then simply don't listen to me. Listen to the people you trust and like

[ermo]

@JDoug:

I think Kazumi is actually agreeing with you, but I'm guessing that there's maybe some sort of language barrier at work here; I must admit that I also had to read and re-read Kazumi's post to understand what he was pointing out.

As I understand it, he's saying that the suspensions in stock S2U might not be accurate but they sort of work from a computer game perspective in the sense that you can drive the cars even if they don't always behave like you would expect.

But, when you begin to tune the suspension set up of car X in the game on the assumption that car X in the game is on the same suspension as car X in real life (which is often not the case as you have observed), you might run into issues because that assumption doesn't hold; in other words, you could end up confused because the in-game car X on the inaccurate suspension will not respond to setup changes (such as camber adjustment) in the same way that the real car X would.

He's also pointing out that in a proper sim, he would expect that you could actually build a real car with the same suspension geometry as the in-game car and have the real car actually work and behave like the in-game car with that suspension. However, turning that around, some sims -- such as LFS -- cannot accurately represent certain real life suspension layouts, so the LFS developers need to work around that limitation when they set up the in-game version of car X and that sometimes means that they need to put car X on a different suspension and fudge its setup to make it behave roughly like the real car X.

In short, Kazumi -- like you -- is also curious if overhauling the suspension geometry means that the in-game cars will behave and react to tuning the same way you would expect from the real cars.

Did I get that mostly right Kazumi?

[bt48]

So, errr... The thread has stalled for the past week on a rather tense note.

I'm actually fairly casual but I've some thoughts:

1) One way to see if the new RX8 ratios are accurate is to see if it will do around 103/104 Kph in 2nd gear at 9000 rpm. (SMS messed up the wheel visual model so much it makes me wonder if the tyre diameter is accurate.)

2) The Z06 has a very tall 6th gear that is not useful for sporty driving or reaching top speeds. (Along with DB9, Camaro, Viper SRT and Maser GTS. I think IS-F's 8th gear as well, possibly 7th too.)

3) The actual production Koenigsegg Agera has a 7-speed dual-clutch tranny.
The game version has only 6-speeds and shifts as slow as a manual. It could be due to the car being modeled on a prototype with an older gearbox.
(Personally, I've changed the shift speed to 0.1 for now and may try adding the 7th gear.)

4) The Mitsu Evo X has the appearance and number of gears (6) of the Twin-Clutch-SST. However, like the Agera, it was originally given a manual shift speed of 0.3.

5) The Lambo LP640 has the appearance of the gated manual version but has the faster shift speed of an automated manual. (Opposite problem of the Evo X.)

6) Is there any convincing way to simulate a torque converter automatic? And does semiautomatic=1 have any effect? It seems to be given out randomly for certain t-converter and automated single clutch and double clutchers.

7) Can't wait for CPEs eventual 1st release.

[dev_null]

Originally Posted by bt48

...
4) The Mitsu Evo X has the appearance and number of gears (6) of the Twin-Clutch-SST. However, like the Agera, it was originally given a manual shift speed of 0.3.

I agree, as I have a Ralliart that has pretty much the same 6-speed twin-clutch tranny as the Evo X. With the tranny in sport mode, real-world shifting is quite a bit quicker than in the game. As well, being a twin-clutch, the transmission does not shift to neutral between gear changes, as found in the game.

Is there a way to correct this by editing an unpacked file?

[KKWL]

Note sure if I should post this here or in a new thread but I was playing around in-game, on pen & paper and excel to map out the FC3S's gear ratios + Final drive.

The CDF will list all the default ratio values but when I was doing some calculations I noticed that not all of the values are 100% accurate. Not to mention the RPM's are wrong I think.. The engine for the FC3S says the Rev Limit is 8500 RPM (150hp 1.3L Wankel) and max torque is @ 5500 RPM. Cept the game only lets me rev to a limit of 7500RPM when I was doing in-game tuning lol

Anyways this is what I've been writing down in my excel file:
Possible final drive ratios for the FC3S is:

FD Ratio Value (0-8) NOTE: I pretty much guessed on FD ratios from 5-8 but it looks correct to me.
0: 4.71
1: 4.63
2: 4.56
3: 4.5
4: 4.44
5: 4.33
6: 4.22
7: 4.11
8: 4

Default gear ratios from 1st to 5th gear:
1st: 3 (cdf says its 3.5 might've confused myself lol) 2nd: 2.125 3rd: 1.421 4th: 1 5th: 0.71 (when I calculated some 5th gear ratios I was getting a value of 0.67)

I will attach my excel file (2k7 version)

EDIT: Forgot to mention the FC3S I'm playing with has some upgrades applied (tier 1-2 stuff).

[ermo]

@bt48:

Yeah, there are a few personality clashes going on in here -- but don't worry about it too much; JDoug sometimes morphs into a grumpy (but lovable) old elephant bull when Kazumi inadvertedly pushes the wrong buttons (which basically happens whenever Kazumi posts something that JDoug interprets as Kazumi second-guessing JDoug's methods) ...

As I see it, the biggest issue with modding engines/gearboxes/suspensions/etc. is that the vehicle definition files (.vdfm) are in a binary only format, which means that each modder tends to use his own file-names for his custom additions, which means that modders can't easily transplant other peoples' changes.

We probably need a framework with documentation for these custom files where the .vdfms are pre-edited to point to the custom files and the custom files are already present for easy editing by interested parties who are not into the nitty gritty domain specific modding details.

I am, however, a bit concerned about the apparent tyre diameter issues, as that means that providing accurate gear ratios and engines is not enough when it comes to getting proper terminal speeds in each gear.

[tulib202]

Originally Posted by ermo

@bt48:

Yeah, there are a few personality clashes going on in here -- but don't worry about it too much; JDoug sometimes morphs into a grumpy (but lovable) old elephant bull when Kazumi inadvertedly pushes the wrong buttons (which basically happens whenever Kazumi posts something that JDoug interprets as Kazumi second-guessing JDoug's methods) ...

As I see it, the biggest issue with modding engines/gearboxes/suspensions/etc. is that the vehicle definition files (.vdfm) are in a binary only format, which means that each modder tends to use his own file-names for his custom additions, which means that modders can't easily transplant other peoples' changes.

We probably need a framework with documentation for these custom files where the .vdfms are pre-edited to point to the custom files and the custom files are already present for easy editing by interested parties who are not into the nitty gritty domain specific modding details.

I am, however, a bit concerned about the apparent tyre diameter issues, as that means that providing accurate gear ratios and engines is not enough when it comes to getting proper terminal speeds in each gear.

Well how can we determine that the tire diameters are wrong?
Checking the vehicle information in the vdfm??
It's not a suprise though as most of the game data seems to be wrong.

[bt48]

Well, that and also needing to take into account how the physics engine interprets loaded tyre section heights at higher speeds (if it does, life just became epic-hard).

If we also take into account:

-The inaccurate suspension geometries and types

-Possibly inaccurate aerodynamic drag and downforce data

-The inaccurate torque curves and rpm limits

-The 'soft' tyres

-Possibly inaccurate wheel diameters if they affect driving characteristics

-etc etc

There is simply a megaton of work.

(Who knows, EA might resurrect the Shift brand by the time everything is fixed. )

[tulib202]

well :
- JDougny has the suspenisons nailed and i think he also has the downforce data corrected.
- I'm taking care of the engines and rev limits (almost done)
- there is ptmu when it comes to stiffer tires + i'm sure JDougny has those tires sorted out for himself
- the diameter correction work is probably untoched and the etc stuff