Lee bodnar's why current ffb doesn't work 2011 article

Often, in my opinion, discussions about the lack of “real” feedback through a FFB system is pointless. Any FFB systems function is unrealistic in it’s nature. It tries to translate an experience of motion and g-forces into a g-force free and motionless environment. Anyone with any experience on let’s say the Nordschleife (I live 10 minutes from this track) knows that a FFB system doesn’t even come close to real life driving. When I push my BRZ on the Nordschleife I hardly experience any force in my steering wheel. If you would translate that experience as close as possible in my OSW wheel, driving a sim would become boring as hell (for the feel part that is). People who think you can feel your back wheels loose grip while your front wheels keep grip in your steering wheel are lacking the real life experience, because you won’t (unless it’s really extreme).
What you want in a good FFB system is an alternative translation of forces you normally wouldn’t feel in a steering wheel.
You’ll never mimic the real deal with a FFB system, not now and not in the future. Don’t believe me, stop imagine driving a car on a track, go experience it.


So true imo; RL forces through our sim-wheels would be quite lifeless and boring. That’s the whole point of FFB for me; to in part, provide details and information that are not present in the virtual space. Granted, with DD-wheels, some force-effects need to be reduced and managed through manual fine tuning in most cases but, that’s not so much the fault of the hardware / DD-system design as much as it is about making FFB work for the majority of wheel users with mainstream hardware.

DD-wheel users should expect to have to make some effort to tune game-FFB to suit such powerful systems and fortunately, this and other sites offer good information on how to do that.


Yes, That is why many don’t like my settings for My SimuCUBE, I douse High Forces but not 1:! (.593:1 to be close to exact in iRacing telemetry) but I have the settings filtered in such a way that the wheel actually doesn’t give a whole bunch of ancillary feedback (minimal road noise, ect…) as this is how I remember it when driving real race prep cars back years ago. The car would wear me down over time with the strength it sent through the wheel but there was no real jolting for the most part and things were smooth… The main feelings came through your seat and the G’s which we do not get… This is where things differ with a Sim is that many want all those feelings through the wheel and in most cases many of those feelings are “fake” even in iRacings case much of the road feel is not from the telemetry but from the harshness of the FFB signal…

Right now FFB is as Mika mentioned all torque based BUT the fact that the sim has to know what position the wheel is in in order to send a “reactive force” does make the whole loop different than in real life where the ONLY thing really bringing the wheel back to center is Caster, Friction of the tires, and speed of the car. In a Sim the wheel tries to reposition itself to center sometimes with force because the game is saying the wheel should be centered BUT the wheel is reporting slightly off. This is actually where there are probably limitations with the Microsoft Direct Input resolution where we are pretty much maxing it out with the SinCOS encoders… If we had all the resolution available for positioning here we may actually be closer to reality as there will be less chance of misreads between game an wheel position.

I actually think that in a game like iRacing the Telemetry is pretty accurate but how it is being sent to the wheel and how the wheel is translating it does color it and in some cases (depending upon your settings) boosts it beyond what could be considered real…

But in any manner we are limited by the Technology and the Programming of the Feedback systems to some degree but we are getting a feel that is more real as it is something that will actually work you like you would actually be worked when in a car…

Road cars with power steering on the track help a lot. Because on sticky enough tires and without a/c and with helmet and fireproof suit there are enough troubles already to also fight a wheel.

But once you talk about regular steering on the open wheels or powerful enough karts - you have to fight a wheel a lot. Small Mige feels to be weaker somewhat, but I’m glad it’s easier.

On a track, you know that shit can end up bad driving 700hp. Really bad. And it’s part of fun.
Other than that - karts are more racing - you turn faster, you see better, you’re way more into the drive.

Sim should be fun and challenge. Fighting a wheel gives physical part of that. But I want a fair and realistic fight. And I don’t want lag, noise and weird stupid stuff happening. But I don’t care how close to reality it is - certainly, I know that in reality, those cars are not behaving like that. Who cares? I have a track car as well.

It’s not about using “1:1” forces or any certain ratio. The overall FFB behavior is different. 1:1 forces is often way to over-powered, accelerative, sharp, etc. compared to real-life this is largely due to the fact that real-life FFB is reactive while a video games’s FFB, for the time being, is active. Not to mention real-life has a ton more inertia and mass (tyres, wheels, suspension, etc) than our wheels. Moving over to a fundamentally new FFB system using torque-based FFB could potentially drastically improve the FFB behavior allowing us to use high FFB without getting the fake FFB behavior we have now. High FFB with today’s FFB wheels just want to snap your wrists and the wheels also want to do all the work for you rather than you do the work like in real life. Telling a wheel to move to a certain position is not the way to go and this problem gets more and more exaggerated the higher power the wheel is.

The simulators are not telling a position for the wheel to go. They are giving torque commands.

OK I guess it is really hard to explain what is going on but the feedback is sent via torque signals from the Sim based on how far off of the optimal position the sim determines the wheel to be at, So it isn’t putting the wheel at a position but it is nudging it toward that position through varying degrees of force… The problem comes when you have a light armature servo (and yes most servos are WAY lighter than a full rack steering system) it is easy for the reactive force to over power and overshoot where the wheel should optimally be and with that overshoot comes oscillation. The problem is that as you increase power with nothing to slow the servo you get a wheel that drives you BEYOND what is realistic because it will start to overcome the forces you apply to the wheel and intern the wheel feels more like it is driving you than you are driving the wheel…

I see SOOO many people running 0 damping on their servos which just shows off the issue… Damping is there to help stop that overshoot and counter the force as the wheel gets closer to its intended destination. Friction can slow the whole system which is like adding weight to the entire system. the only thing is friction is VERY heavy handed so it doesn’t take too much to add inertia artificially. Inertia is the filter if you want to add flywheel effect to the wheel so it is not as abrupt to stop on a reverse command, It also helps in getting the wheel moving faster so it is good in conjunction with minimizing the initial effects of Friction.

You can get a pretty natural feel out of the wheel but you aren’t going to get it out of a unfiltered servo UNLESS you actually mimic a rack and steering column and everything in between to up the inertia of the entire system and even then there will be some differences.

Yet you still don’t want to use damping friction or inertia it messes with the ffb too much and makes catching oversteer very hard. Its just a fake effect that masks the ffb and makes the wheel slow to react.

And as long as you hold the wheel there are no oscillation problems.
the problem is not that there is no mass from the steering rack its just how the games / ffb works, that is the problem.

even a Logitech g25 oscillates which is real weak. no matter what motor or the mass it has. You will still have overshoot but still you don’t want any damping types in any way. ( and also a motor with a lot of inertia will still overshoot)

Its also a bad excuse from people: i use damping because it feels more realistic. You will just have way less feel and control over your car in the sim. of course everybody can drive with the settings he wants.
But if you have problems controlling the car over the limit, its because of the damping people use.

I guess you can say i hate damping :wink: and i hope i convinced some people to try and drive without dampning again :smiley:

I agree, however, using no damping, friction, smoothing, etc. gives insanely overly harsh and fast FFB characteristics. If sims or a driver need the wheel to react and spin so fast in order to save oversteer then there is something wrong with the sim’s physics or the driver. We shouldn’t need a wheel applying opposite lock with so much acceleration and torque that it wants to break our wrists or almost automatically save the slide on it’s own for us. It’s also very unrealistic. In real life, the wheel of a car spinning will be sort of lagging behind the spin itself and it therefore requires the driver to actual apply lock himself to save the car, not just let the wheel “automatically” spin like with the OSW. There are some exceptions like some purpose built drift cars but it’s rare. Also, sometimes when the rear comes out, the OSW wants to keep applying a force against your wrists - this is a perfect example of “active” FFB trying to push through your grip - whereas in real life, the wheel doesn’t continuously try to spin against your grip because you can hold it at an angle if you want.

An example is in Automobilista. Sometimes I get the car too out of shape and the FFB wheel wants to instantly, very powerfully and quickly spin to opposite lock so much, all the way too the wheel’s stops. I have to let go of the wheel sometimes because the FFB wants to keep applying a force against my grip and wants to “break my wrists.” - all from just getting the rear out. This is not only very unrealistic but gives a negative impact on the experience.

I’m close to setting up my simucube for the first time (transitioning from an Ollie Ionicube kit) and hopefully the simucube combined with the reconstruction filter improve things.

You assume something wrong there. The steering wheel does not steer me, I steer the steering wheel.
I don’t run high ffb that steers me.

automobilista is just very broken with ffb and has crazy osscilation. you need max smoothing in realfeel to make it bearable. and default settings are real bad this might help https://www.youtube.com/watch?v=Ze5rTNpUxFQ

Sorry, I didn’t mean to say I was assuming for you. I meant to say it in general mostly with higher forces.

Thanks, I’ll look into that video later.

1 Like

its al good :smiley: everybody can set the wheel up to their liking!

I prefer to have a bit of dampening applied to reduce the rotation speed (SAT) a bit with most titles because the OSW is so fast. Wouldn’t reducing the rotational-speed in Granity be another way to do that without using dampening or friction?

Yes thats an intresting thing :smiley: what i also dont get max rotation speed is 1000 rpm for small mige is it possible to raise this ? or even lower this.
A big mige with more inertia can do 1500 rpm!?

And another thing :smiley: how close to the 1000 rpm do we actually get anyway :smiley:

That’s an interesting perspective. I have found AMS to be the most enjoyable FFB and require the least tuning and filtering in Simucube of all the sims I’ve tried. I have found notch filter setting of 2 and -2.5 to cure harsh effects and oscillation in pretty much all the cars I’ve tried in AMS.

To be fair, I run my large mige at around 40 % in Simucube and 80% in game.

1 Like

You can’t get 1000rpm from small Mige on SImucube. Simucube is 48V and motor is 140V/1000rpm, so you’re going to get 343rpm MAXIMUM or 421rpm from large Mige.

@Mika: I have reconstructed Leo’s original document to pdf, let me know if you still need a copy.


I have to say to a degree your view on Damping is a bit off. Damping is used to control Overshoot of the Servo System (which is an inherent thing). The fact is you want to use just enough damping so that the servo stops where intended after a force is applied. An example would be if you want the wheel to stop at 1 rotation and it stops at 1 and 1/8th rotation then you need Damping to get it to stop at 1 rotation. If it does not stop where intended then it is overshoot and IT is actually messing up feedback because it is causing the servo to be out of position relative the position it is supposed to be in… A real life Steering System does not have this problem with overshoot as it is passive so in a real car the wheel stops and starts exactly where it is supposed to be.

If your damping is correct you should be able to flick the wheel and let go of it and it will naturally come back to a stable position with maybe an oscillation or two (which IRL would be tire deformation). It will not go into an oscillation loop… If that happens you do not have good damping for your servo as what is happening is it is overshooting and causing the system to THINK you are turning the wheel past center so it tries to apply a simulation force in the opposite direction causing it again to overshoot in the other direction slowly amplifying (oscillation)…

Yes, if you can hold the wheel without oscillation being started it is due to you being the damping factor in the equation… Problem with this is if you run any decent force all of those little micro corrections that you are making to counter overshoot through your muscles will fatigue you over time, this doesn’t happen IRL. This is part of why simulation is felt to be too harsh and to strong compared to real life is that in real life the wheel dos not impart its own forces through our arms where with a sim it does.

1 Like

What about when the issue is not necessarily the wheel not stopping at the correct position, but the wheel just putting too much constant force?


  1. I’ve noticed OSW’s return-to-center (SAT?) force as wayyy too strong in any sim unless putting so much damping or friction on that the wheel becomes quite “dead” and feels like you’re trying to steer through water or mud. It’s like there’s a massive baked in centering force that I’ve never experienced even in non-power steering open-wheel cars in real life and never noticed in any video, ever, (besides some extreme cases like suddenly going off road, hitting a sudden massive curb which you’re supposed to avoid, etc.)

  2. I’ve noticed OSW’s trying to snap away from centre (initial part of oversteer) or snap back to centre (end part of oversteer) with that same constant force and, perhaps more importantly, it’s as if the motor ignores the driver’s input and just tries powering through the driver’s inputs. It’s as if the motor says “I’m going to rotate to this position with a lot of force and I don’t care who or what tries to stop or slow me down, I’m going to keep going.” It’s like the wheel tries to power through the resistance, or the holding of the wheel, that I apply myself. Saving oversteer, catching slides, etc. is not wrist-stressing in real-life. It’s as if the OSW is not picking the correct moments to apply such high forces but wants to apply them much too often.

If we had an option to apply negative spring in the OSW settings, or, like can be done with the Accuforce through their physics based FFB tuning, be able to lower the centering force (SAT?), that would be fantastic in helping to get around these issues even if it’s just a temporary band-aid.

P.S. I heard someone once mention that the overly exaggerated centering forces (when using high-powered wheels) are a result of the way sims and current FFB technology work. I don’t know how much truth there is to this - just thought I’d throw it out there.

1 Like

Yet you still don’t want any damping imo :slight_smile: you are slowing down more than just some overshoot. overshoot never has been a problem to me. the damping is a bigger negative overall for just fixing some overshoot. overshoot would need a real fix not an artificial one.

You talk about using any decent force strength, but you say you worry about getting tired of holding the wheel strait which hardly costs any effort at all even with no dampening, sounds a bit like a contradiction.

I know what people are trying to do with dampening but, Damping is just adding an extra handicap to the ffb muting all the rest too. and the overshoot is still there. or you would need to use such high amounts that makes the ffb totally useless imho.

And most probably everyone here uses such high ffb srength that it steers way heavier than the real race cars anyway.

(this is just for people so they know my way of thinking behind it)