OSW Help Needed with FFB

Testing with F301 and F309 (Formula 3 cars) @ Brasilia short. I raised both cars' MaxForceAtSteeringRack to -6000. The F309 barely goes past 3000 (other than some extremely quick moments - probably curbs) but I raised it any way as it can potentially give more dynamic range.

I still can't get rid of the massive stiffness and massive return spring of the OSW. The wheel isn't free to move with tons of macro and micro corrections because you're almost always either doing 1 of two 2 things: holding the wheel hard in one place (after you're done your turn-in) since the OSW wants to spring back to centre with massive force, or you're preventing the wheel from returning to centre too quickly on exit. This forces me to drive in a very static, smooth way - turn the wheel for the corner, then unwind on exit, and repeat; there's hardly any micro and macro corrections going on because the wheel always has a constant force resisting any movements rather than only outputting those high forces at appropriate times. Instead of controlling and "battling" the car, you're more-so controlling and "battling" the steering wheel.

Also, it feels like there isn't much range of forces. It's like the 40-100% strength range of the game's FFB is all being output by the wheel in the 90-100% strength range. It's like there's only 3 or 4 forces, low, medium, and high with 80% of the driving and 80% of all FFB effects occurring in the high range.


P.S. I hit the inside of a curb (second last corner at Brasilia short) and spun and the insane return-spring of the OSW made me have to let go of the rim just from spinning or else it felt like it wanting to bend my wrists in half. This is not realistic. The wheel should not be outputting such massive forces all the time. It was just a harmless spun.


How can I make the FFB strong yet realistic in Automobilista?

 

Not sure what you are on about return spring force? I thought the wheel self centred only when you are moving just like in a real car. Are you saying your wheel wants to self centre when stationary?

 

It does want to return to centre but not in the way a real car does, it does it with an insanely unrealistic amount of torque and force.

I also tried the Toyota FT-86 in Assetto Corsa. I was just playing around with the car - bashing it around, being aggressive - and, on any oversteer situation, the OSW wants to return to centre so instantly and quickly. BUT, it's important to consider that it's not only the speed and instantaneous acceleration of the wheel that is wrong, it's the force/torque behind that speed/acceleration that is another huge problem. I've been in a couple cars setup for drifting (one was a quite modified late 90s/early 2000s BMW M3) and, yes, the wheel can spin fast during drifts BUT there is not much force/torque behind it so you can stop the rim spinning easily regardless of how fast it is, but in the OSW, there is always this insane amount of torque behind all movements and spinning - not just speed, but "force"/torque behind that speed if you know what I mean.

To further the problems. It feels like there's only 3 or 4 levels of forces - low, medium, high in Automobilista. I always felt like there was a lack of range in AMS/SCE even when using the Fanatec CSW V2. It's like 80% of the games' forces are compressed into the upper 20% range of the OSW's power. It's like a massive un-linearity is present and any game-forces over 30% become output at like 80% or 90% of the OSW's forces.

 

I just watched 3 real-life F3 onboard videos...it's wheel looks like it's outputting even more force than mine, BUT, only at certain times. I can also see that the real-life wheel DOES NOT have an insane, gimmicky, "baked in" centering force with loads of torque behind it.

Check out the following spin:

Notice the wheel is not trying to spin all the way to full lock with the force of 10 bears just because the car is spinning/oversteering. It's in fact quite serene and hardly trying to spin at all. Of course if the driver let go, the wheel would turn more but, with an OSW, it would instantly want to go full lock at insane speeds and with 10000 Nm if that where to happen in game. BUT, on the other hand, notice before the spin - during normal driving - that the wheel is definitely outputting high loads.

Here's another video with more spins:

The spin at around 1:12. Notice how the driver is doing the work; he puts a few individual corrections into the wheel whereas, with the OSW, the wheel would countersteer hard on it's own upon the very moment of initial oversteer rather than allowing the driver to "work the wheel" and play with those multiple inputs he makes in the video. It's sort of like the OSW dictates too much with it's inputs with the driver himself not being able to dictate enough through his own inputs. There's too much of just holding the OSW while it does the work with so much force/strength rather than you, the driver, actually controlling the inputs of the wheel.

Lowering strength would just make everything weak which is not what I bought the OSW for and not what I'm interested in. I want a proper "range" of forces rather than 90% of everything happening with what seems like a constant, "baked in", un-dynamic strength equal to 80% of the wheel's max force.

 

It seems these problems are not necessarily problems, or only problems, with Automobilista or any game, but with current FFB technology in general.


I found someone else complaining of the same thing:
Project BruteForce2 - IONI-based DIY DD FFB Wheel (Extensive worklog) - Seite 30

The post before the one I linked to referred to it as a "snappy SAT when oversteering".


According to Brion Sohn from the iracing forums:

source: Log in | iRacing.com™ Motorsport Simulations

So it sounds like what Brion is saying is similar to what I've been told by a few other FFB gurus. That the problem is FFB technology/software in general.

I then asked him what he thought about moving over to torque-based FFB technology instead of position-based - Leo Bodnar wrote an article a few years back explaining that FFB will never be correct until FFB evolves to torque-based instead of the current position-based. And mentioned how real-life cars can have wayyyy more rotational mass than our FFB wheels The following is part of what Brion had to say:

source: Log in | iRacing.com™ Motorsport Simulations

 

I'm using an OSW SimuCube system and I don't get the results as you describe here at all unless I'm over-driving the steering system with force-settings that are just too high and cause me to feel as though I'm the one being driven by the car.

Assetto Corsa used to have SAT that was over the top when using a DD-wheel but, they have toned that down nicely but, I don't get the super strong SAT and feel that there is very good dynamic range in the AMS FFB.

Perhaps you can share some of your settings, both in-game and MMos as a starting point. Generally, it's pretty easy to tame some of the return-to center behavior with some dampening or friction.

Most of us that have done some reading of FFB limitations are aware of the issues and we have to find ways to adapt or find some happy medium where we can stop focusing on setting the FFB and just enjoy the Sim-experience.

I also use an AccuForce wheel and have spent many hours tweaking FFB settings - not only in game-settings but, also in SimCommander FFB which has many more variables and finding good results with the OSW and AMS was very easy.

In regard to steering characteristics, perhaps you can experiment with car's castor settings to see if you can tune the SAT to something you like better.

 

Hi Dean,

Any chance you would post your OSW settings too? I'd love to give them a try.

I'm running a small mige osw and having inconsistent results with AMS. Having huge unpleasant, undriveable oscillations at high speed on the straights at Goiania in stockv8 and Imola in the fvintage. Not sure if this is from the track bumps? But on the other hand I have a really good feel at Cadwell and Oulton. I've tried a whole swag of settings: pure ffb versus canned effects in AMS, realfeel smoothing increase and damping settings, then damping in MMOS, and "overall effect" though to be honest I have no idea what this does.

 

Hi Tom i have small midge.. argon drive on 12 amps in granity and do not have the oscillations on those tracks my settings in mmos are overall filter 5 to 7% steering force minimum0% max 100% and no other effects ie damp ect..in game 0% low boost and 180%mhz or 360%mhz pure i prefer 180 and NO canned effects..as that can cause some oscillations...in game ffb about 40% to70% depending on the car...hope this helps also my tbw in granity is 470

 

Hi Tom, being that I'm using the SimuCube, it's recommended that the Effect Filters / Desktop Effect settings in MMos are turned off so the only one I use is the Overall Filter but, with AMS - I have mine at zero currently. I do have a small amount of filtering set in Granity and that seems to take care of most issues and provide a good starting baseline. By contrast, AC needs 4-5% filtering to reduce chatter and spikes in the FFB.

In AMS, I run 360hz Pure FFB with the master strength at about 65% and haven't noticed any significant oscillation but, I don't always try to dial it out 100% either (I try not to let go of the wheel though ). If the Oscillations are minor, I may live with them so I don't have to trade for details in FFB effects.

 

Hey Spin Check out Some of Neils Driving with the Caterham and especially the Ultima GTR and he appears to be spinning the wheel effortlessly when drifting just as you mentioned. I'm sure he only uses some damping to tame the wheel.Do you use any damping at all. I have the same wheel and damping changes everything in terms of making the car easier to drift and drive generally.Might want to ask Neils what his setting are.

 

The only way I can have these massive issues go away is with either lowering the force a ton or by introducing so much friction and damping that the steering feels like a constant, fake, resistance because that's exactly what it is. Using settings from Granity or MMos is just that: fake, not from the game, "baked in" settings.

As Dean mention, there are issues with current FFB technology/hardware/software etc. and FFB as barely evolved, if at all, from the 1990s.

Dean, I feel like the wheel is driving me too. That's why I sweat more and do more work myself using a Thristmaster T500RS than I do using an OSW at 15 (or so) Nm - because the DD OSW wheel drives itself while my hands just hold on for the ride whereas with the Thrustmaster I was actually doing the work just like in real life.

These systems, as John from the iRacing forums pointed out, are too powerful for the current FFB technology of games and that power really showcases these issues unlike something like a mainstream wheel which doesn't showcase the problems as much due to being much weaker.

Real cars have like 20x more rotational mass than our steering wheels through the suspension, tyres, etc.
____________________________________________________________________________________________________________________________________________________________________________________________________________

The following settings are for a Large Mige setup (10,000 PPR encoder, IONI Pro HC, 600w 48v PSU) unless otherwise noted!

Granity
Goals
CEI: Simucube - I am running in Simucube mode which is apparently better but I believe you need to do a few things before with wiring/cables and settings so research before setting to Simucube mode.
MUL: 100
DIV: 100

Machine
- MCC: 13.667
- MMC: 18.222
- MR: click the "Measure resistance & inductance" button below to automatically fill this setting
- ML: click the "Measure resistance & inductance" button below to automatically fill this setting
- FBR: 10000 (because I have a 10,000 PPR encoder)

Tuning
- TBW: 1000-3300 Hz - I keep experimenting. Currently using 1500 Hz & overall filter (see below) 5 as my starting "default."

- NFF: 7.0 Hz - Large Mige
- NFA: -3.3 dB - Large Mige
- NFQ: 1.0 - Large Mige
- NFF: 2.5 Hz - Small Mige
- NFA: -2.3 dB - Small Mige
- NFQ: 0.5 - Small Mige

Currently using (definitely not perfect but these wheels/sims are full of compromises):
- TED (Damping): 0.60%
- TEF (Friction): 0.70%
- TEI (Inertia): 2.50%

Others have suggested:
- TED (Damping): 1.75%
- TEF (Friction): 0.75%
- TEI (Inertia): 0.75%
and
- TED (Damping): 1.00%
- TEF (Friction): 8.00%
- TEI (Inertia): 4.00%

MMos
- Overall Filter: 2-10 - I keep experimenting. I usually raise/lower this inline with TBW (eg. TBW: 3300 Hz & Overall filter: 7-10, or, TBW: 1000 Hz & overall filter: 2-5). Currently using 1500 Hz & 5 as my starting "default."
- Min Force: 0.00%
- Max Force: 80% - The higher the %, the more non-linear/compressed (AKA "boosted"] the FFB feels to me
- All other effects/filters: off

RealFeel.INI file
Niels said that each car's MaxForceAtSteeringRack has been reduced by around 20-30% from it's "ideal point" because most people don't have high powered wheels. I'm therefore raising each cars' MaxForceAtSteeringRack by 33% (eg. F309 default = 3000, my setting = 4000). I then raise/lower my wheel's amperage (MMC) in Granity to a value I like for that particular car (and one that seems fairly realistic).

In-Game FFB @ 80%


Additional Info (not really important)

Why I set higher MMC and compensate with lower MMos and In-Game FFB
Setting MMos and In-Game FFB lower than 100% seem to offer a more progressive build-up especially in Automobilista (and anything based on the ISI engine, eg. RF2). It may have something to do with the physics engine and/or tyre model but FFB forces often seem to increase too quickly/much relative to how little actually tyre-lock has been input. Also, this method slightly helps with the nature of high-powered wheels with their unrealistically aggressive, fast, accelerative, sharp power.

Yes but damping adds a fake, "baked in" constant damping effect making it feel like I'm driving through mud and it also kills the liveliness of the wheel. There's all sorts of compromises to be made since these wheel's are much too powerful compared to the primitive FFB technology (and arguably physics) videogames use.

 

Heres a lint to Ultima GTR

 

Dont use any friction just damping and maybe some inertia....different for each car

 

[delete]

 

Regarding my changed settings from above, I set MMos to less than 100% and also set AM to less than 100%. I've mentioned this a year or two back but something in AM/SCE doesn't feel right when FFB is set to 100% in-game. It feels like there's a lack of range in AM/SCE. It feels like 80% of the games' forces are compressed into the upper 20% range of FFB output. It's like a massive non-linearity is present and any game-forces over 30% become output at like 80% or 90% of the OSW's forces. It's like the FFB ramps up way too much relative to how much the front tyres are turned.

I, and many others, noticed the same in RF2 that's why in rFactor 2, many people massively lower the STS (Steering Torque Sensitivity) which is basically the FFB output linearity as at the "apparent" fully linear setting of 1.0, FFB forces ramp up way too quickly. This could be an issue with regards to physics rather than actual FFB itself though as I find it curious that this only happens in ISI engine based sims (Automobilist/StockCarExtreme, rFactor 2, Project Cars). Then again, I noticed that - especially at low speeds - front tyre grip runs out at quite low tyre angles in all ISI sims so maybe that's what would lead the FFB output to ramp up so much even though front tyres are not turned much; this is just speculation for why this happens so don't quote me on that but there's definitely an odd relationship between grip and front tyre angle which MAY lead to the seeming massive non-linearity of FFB with regards to FFB build-up as you add steering/tyre angle.

Oscillations can be mostly ironed out...again, it's a fake, baked in way if doing it (I guess since FFB and physics are not perfect especially when using high-end equipment)...

For a Large Mige
- NFF: 7.0 Hz
- NFA: -3.3 dB
- NFQ: 1.0

For a Small Mige
- NFF: 2.5 Hz
- NFA: -2.3 dB
- NFQ: 0.5


By the way, I'm currently using a theoretical max torque of 18.48 Nm. MMC @21.845 A & MCC @ 20.753 A (MCC = 5% less than MMC) in Granity w/ MMos and Automobilista both set to 87%. So 21.845 / 1.414 * 1.58 (use 2.2 for Small Mige) * 0.87 * 0.87 = 18.48 Nm My theory behind this is that lowering MMos' and Automobilista's FFB feels like it gives the FFB a slightly more progressive and linear output for the same torque compared to setting MMos and AM to 100% and then lowering the Amperage in Granity to equal the same torque. Note: this is for current testing which is with the F309 @ Brasilia short with MaxForceAtSteeringRack @ -6000 (very important). I may lower MaxForceAtSteeringRack to -5000 or -4500 and then lower MMos and in-game, or lower Granity amperage, in order to equal the same torque.

 

Thanks for detailed replies and posting your settings guys! Seems I've got so much to learn. Soon as I build up some wife credits I'll give them a test and report back how it feels for me.

 

We should probably keep in mind that we are always comparing our Sim-experiences to realism and it's certainly a rather high standard to try to meet.

Of all the different wheels I've used over the years, the Direct-drive wheels have proven to be far superior FFB experiences - no question at all; within the context of what's possible with current FFB technology, this is pretty much living the dream.

 

Of course Dean . Still makes for some frustrating moments though like oversteer having way too much springiness and force to it which really puts a lot of strain on your wrists just from applying opposite lock to save oversteer (especially quicker oversteer and snaps) but it's still overall awesome

I tried lowering MCC to way less than MMC and it felt slightly different. I believe it made curbs and such less aggressive, possibly allow slightly more wheel "freedom" during turn-in and mid-corner, and also, maybe, just maybe, feel understeer (FFB drop-off) more. During many moments it felt the same but during other moments it did feel a little different.

The setting I tried was MCC @ 7.753 A. My other settings were unchanged: MMC @ 21.845 A, MMos: FFB @ 87%, Filter @ 8, AM FFB @ 87%, Granity: TBW @ 1500 Hz, Damping @ 0.60%, Friction @ 0.70%, Inertia @ 2.50%. F309 @ Brasilia Short with close to stock setup and steering lock (steering lock @ 360/17 instead of 360/18).

 

Hi. I'm having problem with my osw in ams. For some reason the Mmos damping effect doesn't work like in other sims. Usually something like 35% damper is good in ac and rf2, but in ams i have to put both damper and damper saturation up to 100% and i'm still not getting enough damping. I get some damping, so it works a little bit, but it is much weaker than in other sims.
Is this normal? Is there way to get the mmos damper work like in other sims?

I'm using 360 hz pure.

These controller lines don't seem to have any effect.

FFB steer friction coefficient="1.00000"
FFB steer friction saturation="1.00000"
FFB steer damper coefficient="1.00000"
FFB steer damper saturation="1.00000"