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

Simucube 1
41

Your view on “pure” FFB without filters/dampening seems to be more about thinking you will lose detail in the signal and deaden the wheel if you apply any of these. I agree with Brion on that part.

The twitchiness of the wheel you prefer includes all the overshoots/corrections the system has to make which I think you are confusing with “detail”.

Luckily we can set these things up to our own preferences! :slight_smile:

I agree with @Spinelli that the centering force is way too strong sometimes. Since I got my OSW I can often just catch dumb mistakes/slides in my trusty Skip Barber by letting the wheel do the work. It almost feels like cheating. I have never driven a real car on a track really pushing it though, nothing to compare with.

42

i am not doing it for what you call detail. but for the speed and freedom of the wheel.
this is the williams fw31 @ pretty high strength 30nm in iracing with small mige @ 10A.

and virtually no overshoot anywhere.

here a different game with virtually the same settings. which gos nuts

these games are just kind of broken and not made for direct drive at all even damping wont fix this.
you need too seriously change the game ini to even make it function properly a bit.

even after fixing its still bad https://www.youtube.com/watch?v=faI0jilkU0g
on some cars its less bad but f1 cars show better how a sim operates at its limits.
and iracing out preforms them by miles and that even with a 60hz ffb signal

43

I agree, a lot of these games are not ready for DD systems and their FFB systems seem to be designed for slow belt/gear wheels. Which makes sense, because that is the mass market.

But until these games do catch up I feel like it is needed to add some filtering/damping.
Again, it’s back to preferences. From your videos I can see you are a skilled driver, better than me :slight_smile:

But for me it would be unacceptable for my wheel to ever start wildly oscillating when I let go of the wheel at speed or even when just standing still in the pit lane.

44

Automobilista physics are developed using DD wheels, not necessary to play with .ini files as soon as you use 100% ingame and ajust final gain on SimuCUBE config tool.

With mmos there was a lot of oscilations and the force you had to do against the wheel when the car start to oversteer was unrealistic because was very high.

With SimuCUBE firmware is really good, and i would say perfect in latest versions.

45

that niels uses one doesn’t mean the game is developed for dd wheels they dont even have a dd wheel profile in the game as far as i know which says enough. niels runs terror amounts of damping on his bodnar wheel to fix his broken ffb game :wink:

46

i have been working with Niels and believe me , he doesnt do his work thinking in comercial wheels , what he feels is correct with his bodnar is what he gives ok for.

at least before he was kidnapped by the brazilians…

probably the problem is on my side, because i enjoy AMS ffb but have never feel the same with iR ffb.

tried tons of guides, videos, examples and configs for the mmos and SIMUcube config tool , but have never one smile on my face with iRacing ffb, even if some of the best moments on simracing for me have been in iRacing, but not because of the ffb.

:sweat_smile:

1 Like
47

In my view the following statements by @SKeijmel are missing the point:

“these games are just kind of broken and not made for direct drive at all”

“iracing out preforms them by miles and that even with a 60hz ffb signal.”

“niels runs … to fix his broken ffb game”

Sims were not developed to work with DD wheels. It was the other way round… DD wheels were developed to work with sims. If they do not, it could be equally argued that it is the DD system that is broken.

As I understand it, DD systems were developed by clever people who were primarily using the sim iRacing. It follows therefore that DD wheels are likely to perform better with iRacing than other sims. The fact that a sim does not perform well with DD wheels does not mean that the sim is broken.

48

Just look at my video in the formula v10 in AMS too see how broken it is. i think that says enough.
maybe you get less oscillation if your turn off realfeel but then you are really stuck with old rf1 canned effects.

if you know niels maybe ask how to fix it. but damping until it doesn’t do it anymore isn’t the solution :wink:
i invested quite some time to make ffb settings for this game you can see that in my ams video guide. but even with max smoothing in realfeel its unfixable

49

i have very good FFB on Raceroom thanks to you SKeijmel , so dont think that your opinions for me are not valid for me, i think that same you did some great settings on other games, you did not do it in AMS.

and let me say that you should not use a 26 cm wheel with those high levels of torque.

i have a modified G25 (27 cm) OMP trecento (30cm) Kubic (31) and
320 Alu s (32) , and with the same FFB settings, the feelings can go from terrific to absolutely disgusting.

Regards.

50

Don’t worry i am just very critical and i am just as critical on iracing :smiley:

51

I hear you, Gonzas. For every current sim other than AMS that I’ve tried (AC, PC2, iRacing, rF2, RR), ffb seems like an overlay of effects, with varying degrees of usefulness in communicating what the car is doing. Maybe it’s a combination of the intricate sounds and visual cues along with FFB, but AMS uniquely feels like it has FFB as an integrated component of its physics modeling.

On my large mige in open wheel cars, I’ve seen the busy jolting and oscillation in your video, SKeijmel, particularly on higher power settings. But candidly, a simple addition of 2.0, -2.5, 0.1 to the three notch settings in Simucube has eliminated the jolting nature of the detail without eliminating detail in all the cars I’ve tried in AMS. The front straight of Goiania in the StockCar is a perfect test of that. With no notch filter, the bumpiness wobbles the wheel disproportionately. With the above three entries, I experience just normal bumpiness.

All that said, too—Reiza (Niels) noted early in development that they set the Realfeel MaxSteeringForce entries for the cars a bit low (stronger) to improve feel on consumer wheels. By simple stroke of a few keys in cockpit for each car, I increased that entry by 25 to 35 percent. That spreads out the forces more appropriately for a high-powered wheel. If I use Motec to track the steering arm force after a few laps, that gets the Realfeel just almost precisely where the Motec peaks fall at or just around the entries for each car, telling me Reiza set these up with attention to the actual steering arm forces being generated by their physics.

Maybe this is purposely-engineered “broken” ffb in relation to DD wheels out of the box----clearly to accommodate the G27’s and T150’s that are common in the market—but the DD fixes for these market compromises are built into the Realfeel and Simucube firmware. The adjustments translate to, IMHO, an outstanding (and extremely immersive) experience.

There are hundreds of posts on the iRacing forums comparing and debating FFB settings for all levels of wheels, just as there are on the Reiza and RaceDepartment forums for AMS, AC, rF2, etc.

52

There is no actual centering force in the SimuCUBE when in game (there is now a center spring at desktop) so this is a completely game controlled situation, I am not sure what Telemetry manipulation is being done in the AF to implement a centering force feature… But basically the game determines how much force to apply to the wheel given how much deviation there is between the optimal wheel position (as determined by the game) and the actual wheel position (where you have the wheel positioned).

#2 - Is simple the game sends a higher force than you can counter or react to which causes the wheel to drive you. In this case you would probably be best turning down the power or you will have to add filtering to counter the quickness of the wheel… The car also has a huge effect on this as some cars have a massively snappy traction loss…

When I set up my wheel the one thing I concentrated on was to get the feel across center constant by testing and weaving the car back and forth, In doing so you will find that the center forces do let loose at some point (this is what makes it difficult to catch and SAVE an oversteer situation) to get it to go away you have to balance Damping which will drop force the closer you get to the intended position and inertia which will try to keep the wheel going as speed…

53

Yep, Brion is right, the excessive centering comes from the physics model of the game itself. Stationary car is a difficult thing to model. Assetto Corsa’s usage of the friction effect on slow / stationary car is a good solution.

54

Yesterday, I installed the AccuForce to test the newer FFB-settings - one of them being (G-Force Effect). The results were interesting because increasing the effect is meant to represent increased cornering steer-resistance without directly impacting SAT in a constant way so it doesn’t seem to increase the rotational forces during return-to-center. Increasing the effect beyond a certain point resulted in reduced SAT (perhaps a form of clipping?).

It seems there have been quite a few DD-wheel users recently expressing ideas about how best to address some of the forces being out of proportion. Some titles allow editing FFB files to make such adjustments while others do not. Assetto Corsa has very few parameters that can be changed under the hood in regard to the FFB but, the way the “Gyro” effect works seems to improve the results with DD-wheels in general. I’ve seen several requests for a dedicated “SAT” adjustment.

Perhaps, a Gyro-effect might be worth considering on the firmware side (?)

55

Response within…

I think you are missing the point here ALL servos have a torque control problem and will have Overshoot to some degree unless they are perfectly matched to the game (which none are at the moment because none have a brake system to control spin down it is all done with opposing force) so the issue has to be dealt with with filtering, if you don’t you are technically adding movement that is not present in the games telemetry signal. Which some find pleasing and they use that added movement to enhance there experience.

Two different things - When you increase power the overshoots get more dramatic and stronger so if they are not controlled they will wear you down. The bumps in the track cause forces to be applied to the wheel which if an overshoot is present WILL cause an oscillation which you either have to control with your arms, wrists, hands (which will eventually wear you down) OR you set up your filters to FIX the issue.

If set up correctly there is no handicap relative the telemetry the game produces. It only would correct the inherent issues of the servo’s moment of inertial and overshoot, making the experience more correct to the telemetry trace and allowing you to stay on the optimal inputs making everything more natural to drive and actually lowering the force used by the system (so even if you are at higher forces on the wheel you would be using less)

I do understand where you are coming from but without some damping/inertia (I try to stay away from friction except on the Small Mige as it is a REALLY fast servo and Damping alone does not control it) you are actually imparting more movement to the wheel than the telemetry calls for which in your case might give you that added extra feel you need to be able to catch a slide.

2 Likes
56

The SimuCUBE reconstruction filter will help but that characteristic of the fast opposite lock is a trait of a car going against caster… It happens a lot when you get too out of shape on an Oval… The way to counter it is counter intuitive in that you have to force the wheel straight and HOLD it there (don’t try to correct at that point) or do as you do and let go… Hence hart of the reason real race car drivers let go at times early in a spin is they have already received that “there is NO WAY to save this thing” message through the wheel. The best thing though for catching is like I mention above is to adjust the filtering so that you do not feel a center transition… A good car to do this with is the iRacing Skippy as it is a very direct car without power steering influence.

57

In some cases, it’s possible to reduce caster to affect the SAT level. I tried a custom FFB-preset in PCars where the caster-effect was near zero and SAT was almost non-existent. It can even have the affect of the steering falling off center like reversed-FFB. Of course, most titles don’t have such feature and many road cars in games may not offer caster adjustments in the set-ups.

58

That would be magic, that i would need overshoot to get some extra feel to catch a slide :wink:

The faster the wheel the more control you will have. a akm52 has even way less inertia and is even faster than a small mige if i am correct.
Maybe than, you can get away with damping on a akm. i guess that’s also a reason why bodnar uses them.

You are already holding the wheel so there already is damping. you can clearly see in my iracing williams fw31 video there is no overshoot or oscillation. and if the wheel moves on the straith thats because of a bump.


Even here with quite a bit of damping on a akm AMS oscillates @ sim expo @ heusinkveld Engineering simrig. Which should have perfect settings for DD
even with damping it overshoots, it just makes it a bit less worse :wink:
(and that car is way less erratic than the Formula v10)
(You can hear Beano talking in the back )
59

That would be interesting but if I understand the Gyro effect it would be calculated by wheel rotational speed (increasing the faster the car goes) and damping the signal or adding another effect to help have essentially speed sensitive steering… This would be something that would not be able to really be able to be done in firmware without a full telemetry translation in the firmware, The Accuforce probably does have some degree of this built in somewhere as that is where the SimCommander software works well in the it allows some after the game telemetry manipulation IF the game allows it, iRacing does, not sure about any other.

60

The reason why I said the above re the overshoot giving you a clue to oversteer situation is that any signal that is variable can be tuned to your brain (or anyone elses) to use as a predictive factor. There are some that run with the reconstruction filter OFF due to that the Variation in the 60Hz stair step allows them to “feel” what the car is doing more even though that digital variation is NOT actually part of the true telemetry signal from the game.

The overshoot issue is not MASSIVE but it is enough where when using no filtering you cannot generally take your hands off the wheel and have the cars not go into oscillation (yes some of these issues are in the games themselves)… As I have said before when holding onto the wheel you act as a dampener… So you are controlling the overshoot which also means your muscles are micro firing to control the ticks and twitches, which imparts physical wear on you as a driver, you may be fit enough and used to it enough that you do not feel that wear… The Question is can you let go of the wheel at speed and not have it go into oscillation… If you can then the overshoots are controlled if you cannot then you are doing the work of the dampening filter. As long as you are holding the wheel it is very hard to “feel” overshoot.

It is in the end what the driver likes best but the damping filter does not as you say slow things down and diminish the accuracy enough to make a major difference in being able to catch a slide or sustain a drift.

As you happened to mention Beano in the background he just released a video running The Ford GT around Brazil and mentioned his settings with his 65L which is a monster running at 30Nm Max power (which is turned down). The 65L has similar inertia levels as the Large Mige though it has a higher potential top power… I run my Large Mige at 3.25% Dampening and .5% inertia, 0 friction with recon 5 at full power and .6:1 Specific Output in iRacing… Beano was running his 65L at 1.5% F and I and 5% D with recon 5 which is quite high in my mind even for me but it controls his oscillation.

Everyone thinks he is crazy for running those sorts of levels and some think I am crazy as well (I fun 40% less specific output than him most of the time) but the reason we can is due to the servo NOT killing us through setting it up to be as Passive and telemetry direct as possible, Eliminating overshoot and controlling the reruns speed through inertial and friction.

I have compared telemetry with people running G25’s and 27’s and lower powered DD’s and in many cases due to the set-up I am pulling significantly less Nm in the corners and the steering curve is much less jagged (meaning the wheel is bouncing and fighting less)… I will agree that the bouncing and fighting can in some cases actually be your savior as it will bounce the wheel to the proper position faster than you can to correct a slide, but, that would be the wheel actually driving you if it does that and not you driving the wheel (Car).

Also NOTE: the Bodnar systems are highly internally dampened and adjusted to be VERY smooth, They were designed to be Professional sim systems and actually allow for very little control over the total feel of the system… If you think those are not Dampened heavily you would be wrong, Beano has actually expressed his displeasure at the fact that they seem to over control the signal to the point where it cannot be adjusted well, He even got Bodnar to release a Beta firmware that opened up control over settings more which I guess made things more to his liking but not exact. Given the amount of Dampening and everything he runs with the SimuCUBE you can imagine what Bodnar actually has in theirs if it is too much for Beano.