The online racing simulator
Default Setups
2
(48 posts, started )
#26 - tpa
Quote from Tege :Yes like Bob said the RAs default set is as close to real as I could get it. Only thing that isn't from real value is the dampers because I don't have the values. On the real thing it's just adjustment "ring". But I did like if the the "value" was like 7/14 I put to slider close half way in LFS. That setup that the default in LFS imitates is the setup used on the track and it's easy to drive and you can use it on the normal roads too.

So you think that I should make the dampers softer?

I'll make changes to the default set after the physics patch is ready and I have more kms on the real car.

hey, cool to know someone's one it

Do I think you should make the dampers softer? I think you should adjust them so that you feel like it's closest to the real thing
OK here's my analysis of the default RA setup:

3.02Hz front, 3.07Hz rear, with equal track widths that means: oversteer (before anti-roll is taken into account, at least)

Not to mention that sports car generally aren't much stiffer than 1.8-1.9Hz. Even my race setups don't go higher than 2.6Hz.

The anti-roll bar is actually contributes to roll stiffness more than the suspension does - this also seems too high since this can be attributed to snappy over-the-limit handling. Admittedly it does take away the oversteer from the springs, but too much, making the roll stiffness point towards aggressive understeer (due to the lack of a rear anti-roll bar).

Damping:
For the given spring rates, the rear rebound damping seems a touch low, but is OK. Bump damping looks too high though. At the front, bump damping is a touch better, but still a bit high, and rebound damping is over critically damped - also NOT GOOD for handling.

Transient damping analysis shows corner exit oversteer, even before throttle is applied.

Diff:
60% locked on the power side. Jeez! That's more than a GT3 Porsche. That's a TBR of 4:1 btw. Read: oversteer (when on the power).

Finally tyre pressure being equal on a car with that weight distribution seems a touch off, but nothing too serious.

Everything else seems fine.

But on the whole - that thing is set up to spin. Assuming you converted everything properly then the guy who set it up is nuts. Frankly I don't fancy driving that until it's fitted with something closer to the road setup I made for it.


... how's that?
Oh, Bob, don't waste your time criticizing the setup ESPECIALLY if you haven't driven the result... It's well documented that even with your 'road going' setups or Niels' equally well-inpired 'real' setups, the big problem with LFS remains the tire physics in the sim!

Nothing wrong with that from a devolepment point of view though. It makes a lot of sense IMO that Scawen decided to tackle the finer points of the tire model only after suspension and dif modelling -- those can be checked against published data much more easily, but the tire model will need more tinkering. So this way round is the most time-efficient way to go about it.

But for now, it's like Stefano Casillo said:

Quote :
If one would have to come up with a sort of "ladder" to indicate what is the most important element in a sim and how big is the contribuition of that element on the overall feeling then it would look something like:

1) Tyre math model - 80%

This is THE most important element in a car simulation, remember it is the only contact with the road and the main source of every force acting on the car. The implementation could go from a sad "no tyre modeling at all", still typical in arcade games, where the car behaviour is mapped against a series of predefined curves.. like to say, if the car is going at speed X then it car rotate around its center at speed Y. This is usually refered to by the racing community as "central pivot problem". Going up in complexity there could be a 2 point (or bycicle) model, where the car is simulated and simplified into a 2 wheels veichle, where each virtual wheel is responsable for an "axle". This was common in early racing sims. Then you have the full 4 contact points with each wheel simulated separatelly and generating separate forces. From there, you can have a good or bad model, but once that's right, you've nailed 80% of your sim's general feel. In other words, let's say you come up with an amazing 5 arm multilink suspension in real time but your tyre math model sucks, then your sim will suck as well, the suspension improvement won't save a bad tyre model. This is the same for wings and every other part of the sim.

The highlighted part reminds me of when the clutch-pack dif was added to the sim and the devs said it would help with RWD handling. Maybe a bit, but it didn't save the incomplete tire model. Same with downforce.
#29 - Woz
Quote from durbster :I still disagree.

Just like a lot of sporty road cars, race cars are tuned for agility and stability but without the compromises necessary for public roads. Basically, if you don't have confidence in your car when you turn in, you have no chance of performing well in a race.

And it is your right to disagree.

On the stability/agility issue. What is considered stable by one driver might be considered a nightmare by another, as also pointed out by Bob.

The norm is to set up the car to be as agile as the driver can cope with. Also depends on the track. A high speed track you might take out some of the twitch to get better top speed etc.

For example when I first tried a FWD car with oversteer I found it a nightmare and I was slower than with my balanced set. Once I had learnt to keep the car in balance I soon learn it would turn faster etc and now I find an oversteer FWD car stable.
Quote from Silly :Oh, Bob, don't waste your time criticizing the setup ESPECIALLY if you haven't driven the result... It's well documented that even with your 'road going' setups or Niels' equally well-inpired 'real' setups, the big problem with LFS remains the tire physics in the sim!

I don't consider basic analysis such as that to be time wasting, only took 2 or 3 mins anyway. I admit I haven't driven the car IRL, but that doesn't mean the principles aren't the same. For instance, my Dad had the tracking sorted on his car. He handed me the detailed specs for before and after, and I told him how it should have affected the car. From his drive home, he totally agreed and had certainly noticed most of the affects. This is all stuff I've picked up from LFS.

Yeah I've read that quote before and I certainly agree with it. LFS is the main issue. But even then that setup just doesn't look right. Also, another consideration is that, even with perfect tyre modelling, real life tyres vary a lot too, so it's impossible to get it exactly right for any car.

Tege has pm'd me though which explains a few things.

But, both my own road going setups (which are based on typical real life values) and Niels' set, improve driveability massively (albeit through different methods), so this shows that the default setup doesn't appear to suit the car. Neither of us used excessive understeer or such to get the car controllable, just fine tuned things here and there.
Quote from Silly : the big problem with LFS remains the tire physics in the sim!

I can't wait for an update that starts to address that.
I'm really glad I posted this. Two reason's:

1. I've now got four sets of sets that I hadn't found previoulsy. 3 are from a known 'driving feel I like'. Can't wait to try them out.

2. Now I know I'm not nuts when it comes to the physics. I had not previously read certain statements. The lack of which lead me to believe the Devs were on a path to leaving the sim 'as is'.

I will drive the sets soon, thank you Bob, for your efforts. Given the current state of the physics, the setups need to cater to them. Not reality. Even in the end, the same will be true. We're driving LFS, not a RAC or LX6.

I have one other question for you but I'm not precisely sure what it is yet. I'm getting closer and when I do, I'll post two short replays demonstrating what it is I like, and don't.

Quote from Woz :Yes a race car has far more grip and much higher limits than a road car but step over that edge and it will punish you. The thing with racing is that you try to sit on that edge 100% of the time so a race car in any situation will be harder to drive on the edge than a road going car. The margin of error is just far lower and the higher speeds mean the time to react to an error is smaller.

Correct. But your entire argument is flawed. A race car is tuned to handle the physical demands of it's given class. When it comes to road racing cars that are based on mechnical grip, they are so freaking easy to drive over the limit it's not funny. Everything is smaller and quicker, as you say. The difference being that they are so direct, youd have to be numb from the hairline down not to know what the car is doing. The limit is more norrow. So subtlety is key. If you're smooth and accurate, you can write your name with those things if you have balance. I am, and I do. I don't expect you to believe me on the internet, who the heck am I? But I know what your missing, whether you take my word or not.

I've had terminal understeer at 100mph. I never saw it coming. (You do reduce your numbness eventually, starting with getting in and driving.) I went in too fast by a few of mph, steered too much, too quickly, didn't lift enough, and paid the price - skipping through gravel and having to clean out my air box and replace a plastic tube. The BIG question is... in the same situation, how do you think a sports car would've faired? A family sedan? There are many other mistakes I've driven through with glee that I guarantee you, in the same situation a road car would have been death.

After saying all that, it's also important to realise that a four wheel, rear drive vehicle is a four wheel, rear drive vehicle. They all handle the same. I personally prefer the racing breed. All that waiting around to see what a road car is doing makes me nervous. The thing that I'm on about is the extremities. My hands and feet. They are trained to do certain things in certain situations. Things that always work, within reasonable limits. In certain situations in LFS, the extremities are doing... well... seriously? Next to nothing. Up, down? Just doesn't do what I expect. Bob's LX6 set gave me a lot of realisitc response to realistic inputs that I've never experienced in LFS before. It was an eye opener. Both in that the physics are darn close, and the sets exploit what isn't. Sorry for being Mr. Munsty, but before that, I really didn't believe LFS was going to be able to simulate rear drive. Now all my rear drive sets are getting better. (I'm starting to learn how to set up a car, at least in LFS.) But the defaults left me - as someone who doesn't know much about setup - unconvinced.

As far as what we can do in a sim... well, I have no problem knowing exactly how much grip is left on each tire in GP4 and you don't see that chassis moving a whole lot.

Quote from tpa :Why not just make a set for the RA with the real thing's values and then take the real RA and the virtual one, to an identical test area. Then compare the results ... shouldn't that maybe give an idea where to start with physics updates?

Maybe not. Because reality updates much more frequently than computer and you can't simulate everything anyway. But that is certainly the way to do it. It's called A/B comparison and it works. Whether the setups match or not is of no importance. How much the thing dives with a certain amount of steer angle and braking... is. It would certainly lead to ideas of where to change the physics, but maybe not directly from the settings.
Quote :Both in that the physics are darn close, and the sets exploit what isn't

I just tested the easy race and road going setups- The RA ad FZ50 are still ridiculously "over-steery" .. and once the rear tires break loose, they don't regain traction until way after they would in a real car- any car. In any road car you can go around a right angle corner in the city at 25 to 30 MPH.. try that in one of these cars and the rear tires break loose and don't regain traction- look at the "F" forces view- the indicators go red and stay there way too long.. only on the rear tires though. They also go instantly from green to red with nothing in between. It's like a light switch, instantly switching to no traction and staying there way too long. Go to the test area and test it out.

That makes me think it isn't just the tire's physics that are the problem, but maybe something with the inertia of the car rotation too.. who knows.

Why is it that the front tires always stick like glue and the rears are always breaking loose? It is like that in all cars in LFS, even the FWD cars..

With rFactor and GTL coming soon, it would be nice to see some improvements before too long. I guess we will all just sit and wait.
Quote from BWX232 :I just tested the easy race and road going setups- The RA ad FZ50 are still ridiculously "over-steery"

OK I played with the FZ50 and worked out how to get rid of most of the oversteer when off the throttle. The RA seems fine to me though in this respect. When the rear does go though, it doesn't hang about.

Anyway I updated the zip, have another play with the FZ50.
Quote from Bob Smith :OK I played with the FZ50 and worked out how to get rid of most of the oversteer when off the throttle. The RA seems fine to me though in this respect. When the rear does go though, it doesn't hang about.

Anyway I updated the zip, have another play with the FZ50.

Ok cool.. thx.. -yeah I figured it could be tuned out.. Even though your previous sets were way more forgiving than default, once it stepped out, it was gone.

One thing to note about all of this.. I raced about 100 races today (it seemed like) with the FOX, and when the aero model takes hold at higher speed- it feels so much better.. At low speed it is still very touchy. it just seems like the mechanical grip is too low or something.. and doesn't go from "grip" - to "lack of grip" - back to "grip" again like it should.

Best example I can think of.. Get in a RWD car and take a sharp slow speed turn, and give it too much gas to break the rear wheels loose just a little, the instantly let off the throttle. The tail end will instantly stick again, and grab the small slide you initiated. I am talking under 30 MPH here.. In LFS if you do the same, the force indicators go red and you will have a weird long slow speed slide that ends when you are sideways or almost backwards. Then at about 7MPH or so it grabs again, the indicators go green. But at such low speeds, you shouldn't have a long slow slide like that on tarmac, especially with good soft sport tires.

I was experimenting a lot in the test area with the force view on the other day.. Some strange things happen. Skids at 8MPH and whatnot..
I guess I will have to make some replays and pinpoint exactly what I mean.

The good thing is, everyone has the same disadvantage, so online racing is still very fun.
I tried the road-biased setups and was really impressed, but i felt the diff's were a little loose on a couple of cars.

In the RA, i'm using a 20/40 diff, and it has less sloppy power oversteer. With the roady setup, the car would spin the inside wheel in fourth gear turns and the car would throttle steer inwards in a big lazy arc. With it turned up to 20% for the power side, i was getting a lot more obediance from the tail.

Same can be said about the FZ50, which i was running at 20/50. I use a higher coast side for the mid/rear engined cars to help with the lift oversteer. It keeps the car more controlled when braking or entering a corner (or both).

One of the other factors i find with the lfs handling traits is down to driving style aswell. That corner entry oversteer is associated with weight shifts, for sure. You see people coming off the brakes from 100% just as they turn in, and whammo, backwardsville. If you are more gentle with the application of the controls (as you would be in reality), the whole thing feels much more dependable. The softer sprung setups definitely help in this regard!

At a recent trackday, a corolla driver spun into a lake from a 150kmh corner. He went through an understeer/lift oversteer phase before shifting down mid-corner (because he'd lost speed and the corolla has a 2rpm powerband), and the tail jumped around. He said it had never happened before, but this was his first day with a stiff set of cusco suspension in it. Nuff said about that.

Beyond these things though, there does seem to be an inherent problem with the transition from grip to slip, which is only a problem when you break traction (so everything is peachy until then).
Road cars rarely have particularly tight diffs, this is why those setups are like that. The easy race setups have tighter diffs, since they're meant to be raced with.

Anyway it's not like I'm stopping you from altering them if you so wish.
Quote from Tege :Yes like Bob said the RAs default set is as close to real as I could get it. ... That setup that the default in LFS imitates is the setup used on the track and it's easy to drive and you can use it on the normal roads too.

It's easy to drive IRL? That is very good to know. So, given Bob's analysis that attributes very aggressive handling characteristics to the RA (wasn't that sort of a design goal?), is it safe to say that even an aggressively tuned car is much easier to drive than it is in LFS S2P?
Quote :I'll make changes to the default set after the physics patch is ready and I have more kms on the real car.

Sounds excellent. Is Scawen using real world data to tune the physics model then?
Ive been wondering about one aspect of lift off oversteer and this sim, on a 'normal' roadcar, when you 'lift off' the throttle it is actually 'damped' and comes of much slower, i think this is to stop such lift off occurences and aid gear changing, infact if i change gear very close to my ref limiter on my own car during the gear change it will sometimes hit the rev limiter as the throttle is left open enough to 'rev up')

Im wondering if a 'Damp' factor on throttle shuting would be a good idea to add to the setup of the cars.

It probably wouldnt aid me as i match the revs and trail brake anyway but could be an aid to new comers and right foot brakers, what do ya think?

Troy Mclure
#40 - Tege
Quote from Silly :It's easy to drive IRL? That is very good to know. So, given Bob's analysis that attributes very aggressive handling characteristics to the RA (wasn't that sort of a design goal?), is it safe to say that even an aggressively tuned car is much easier to drive than it is in LFS S2P?

Atleast I didn't have any problems handling it. It was my first time at the track on a real car and infact on a rwd car like that. The real car doesn't have so much power as in LFS so that makes it a bit easier, but if all goes well the '05 version will have even more than in LFS now. Couple of times the rear came out coming out of the corners but it was easy to control the small slides. Handling was kinda understeery on slow corners and neutral on fast ones.

Yeah design goal was to make a real sports car that you can tune for track use. The older version of the car had handling problems like oversteering and it was kinda "edgy".
There are some very interesting points in this discussion.

1) I agree that the slow speed slides that never stop are... odd, to say the least. I think this is a problem in the sim engine, but who knows what right now. I do think that it has always been this way. I have never driven a race car and I know you can't feel the g's, so I could be wrong about this. I think this issue does happen with all the cars too, not just the ones in question here; just not as bad.

2) Bob's easy race sets are simply wonderful. I did find the FZ50 to still be too oversteery in certain situations. Maybe the update he just did fixed it. I found the set that I did better (for me), but I would really like to try the new set. The RAC is outstanding with Bob's easy race set. It is really fun to drive. It is still difficult to drive fast, but that is as it should be.

Getting rid of the throttle off dramatic oversteer tendency is critical for driveability of these mid/rear engine cars or the rear weight balanced cars. With the way the back end can be so hard to recover once a slide starts, it is important to keep this under control. I was able to drive the LX4 and LX6 relatively well in S1, but it took a good set that behaved more like a front wheel drive car than a rear wheel drive car. You still had to be ultra careful not to spin wheels while turning though.
For the hell of it I downloaded netKar to see what it was like.

The Mini is very realistic for a front drive car. It handles exactly like all the sporty front drive cars I have ever driven in RL.. I had a relatively powerful one with stiff suspension too- the kind that rattle your teeth over road bumps..

Anyway, not comparing game to game here.. just that car in that game and the way the back and front end are very balanced and predictable. Especially the way the rear end will start to slide, and then if you counter steer and let off a little, it tucks back in and regains grip. Just like my VW Gold Wolfsburg, Honda Prelude, and Dodge Charger Shelby 2.2L Turbo (the fast stiffly sprung car) did.

Just an observation I thought worth posting in case anyone wanted to try it-- hey, it's free, why not.
Quote from BWX232 :For the hell of it I downloaded netKar to see what it was like.

Quite off topic but try the Formula cars. Dare I say they're possibly even better than LFS :eek:
Here are setups for GTT and FZ50. Just try it. It feels so natural. This is how the cars should act. No more unrealistic sliding.

The key is to set normal front tires and soft rear tires.
Attached files
Test.rar - 595 B - 239 views
I just want to chime in to say that I tried Bob Smith's Easy Race setups last night and they are a huge, HUGE improvement. I've already sent them off to my friends who play LFS and they had similar comments. I haven't had a chance to try Bob's "road" sets yet, but I expect they'll be even easier to handle and more realistic. To be quite honest, if these types of setups were the game default I don't think we'd hear hardly any complaints about tail-happy cars and loss of rear traction. Using these setups the rear end seems more firmly planted and it's easier to tell when you're on (or over) the edge. Once you DO go over the edge, the car is easier to reel back in and handles more predictably.

My hat's off to you, Bob. I truly hope that the devs consider using your setups as the default in future releases.
Quote from Cue-Ball :
My hat's off to you, Bob. I truly hope that the devs consider using your setups as the default in future releases.

Ditto, put my vote in for that one too.
#47 - avih
thx for the sets bob, great improvement. Niels' sets are great too. at last, the "natural feel" (at least imho) is here.

grats
Quote from Hallen :I personally really enjoy tweaking sets, but I like to have something close to start with.

That's exactly what I was going to say. I have no skill (nor time) to build a fast setup from bottom (i.e. the default setups, which isn't actually even near the bottom ), but i don't like to take the world record setup and just drive neither.

Back to the topic, i'd like to see 'better' (= more balanced, easier to drive) default setups too. I can say i'm such a beginner on LFS, and I was really scared testing the RaceAbout for the first time, before I realised it was just the setup making the car quite hard to drive. Smoother default setups would not scare the beginners away.
2

Default Setups
(48 posts, started )
FGED GREDG RDFGDR GSFDG