In AC there are different settings to take gForce into consideration. With a Formula car like from RSS or Sim Dream I have up to 6G. I use real head motion (RHM) and sometimes race with VR. The speed of head movement will be adjusted based on data like gForces. NeckFX is a similar approach
AFAIK FFB is part of the game engine. If you use the same car on the same track in dirt 2.0 and WRC 9/10 the Simucube 2 will provide a total different sensation. So it’s more transmitting whatever is the basic approach of Devs (and the result of it) rather than dictating. If a games FFB is poor you can do whatever you want, you just can’t improve the signal/information with the DD. If a game has a good or even very good FFB it’s best to leave it like that. The purpose of TD settings is not to change FFB but to react to the quality/weight/flex of the steering wheel that is attached.
Sort of misses the point… The point is you can’t feel them, Taking into effect is their way of trying to give you gForce information without it “actually” being there. It may work for some it though it will not work for all as it is not feeling it through the seat of your pants and through your body. So in essence these are tricks and gimmicks to try to give you added information…
My point is that gForce information should not be attempted to be duplicated through the wheel where it naturally would not be. When doing so or attempting to do so the adding of information NOT pertinent to the road has a tendency to cover up important information needed to understand the actual grip between the road and the tire.
Lol, perhaps you guys should start another thread to sort this out. . .my issue was solved with one slider!
While true from a game perspective… if the game has goof FFB output then their signal is what should be recreated EXACTLY as it is output…
The problem is that “EXACTLY”… Each wheel will not recreate that output exactly… but if you know the developers intent (i.e. to be as real as possible) then based on your external knowledge you can De-color the wheel output through the use of the filters…
The wheels while they can deliver linear output to a torque event point, Meaning that when it says 5Nm it will get to 5Nm (if wanted) HOW it gets to that point can vary dramatically… Light wheels can spin up fast and cause it to reach 5NM and then hold at that until the timing of the point is reached where a heavier wheel might Ramp up slower giving you lower torque and then ramping to higher… BUT the big problem with FFB is built into the nature of FFB… FFB as it is now ONLY accelerates it doesn’t know how to decelerate. In order to decelerate it just sends a set point lower or opposite that which is in the direction it was before…
Due to this the wheel has to go Oh wait you need me to go in the opposite direction so the controller sends an electrical signal (acceleration based on a new endpoint) to spin the wheel in the opposite direction. When this happens you now have the “wheel coloration”… i.e. how long does it take the wheel to stop rotating in the direction of the first signal, before trying to go in the other direction and then catch up with itself to get to where it should be within a certain time before a new point is set.
Filtering helps solve this by placing additional friction or damping to help slow the wheel faster so that the change in direction can happen without Overshooting the Game intended directional change AND the resultant over acceleration by the FFB system to catch up with itself due to that overshoot of the change in direction. This reduces the time it takes to change direction allowing the wheel a better ability to get to where it needs to be without being behind trying to catch up. What you are trying to minimize is the Overshoot and resulting catch-up of the wheel which happens with 99% of direct drive wheels naturally.
This issue of coloration by the wheel is more prevalent when you consider different servo types and accelerations of those servos… if a servo accelerates very fast it will have a tendency to overshoot where it wants to be easier and therefore feel more active as it keeps trying to correct itself… (vibrations, sharp, responses, etc…) this is where the SC2 wheels are nice is that the servos have a relatively decent weight with the natural slower torque ramp of the in runner style servo as opposed to say the out runner of a Fanatec which has massive initial torque (due to a significant increase in poles) creating a very high acceleration rate or a stepper which is even higher than that.
Its sort of like two runners they start at the same time they need to get to a cone at a certain place. the first runner accelerates as fast as they can to get to that point the second though he can be as fast has a small parachute attached to him not allowing him to accelerate as fast.
about 2 yards from the cone the runners are told NO you need to return to the first cone… The first runner at speed shoots past the cone slowing down as fast as he can but has now gone significantly past the cone while the second runner with parachute attached starts slowing and only passes over the cone by a few feet. So now both runners have changed direction but the first runner has to make up the ground lost because he passed the cone by a large margin so he has to accelerate again harder and faster to catch up to the second runner who barely missed the cone because he accelerated slower but also had help stopping due to the parachute.
This would be a never-ending cycle for the faster runner they will ALWAYS be playing catch up, expending more energy and missing the actual location where they were supposed to be… This is how FFB works as it only sends torque commands without braking commands.
Granted you can put too large of a parachute on Runner 2 and then they would actually fall short of the cone before getting there which in terms of the wheel would mean you are over filtered and the results would feel laggy and slow.
So in simple terms the main filters are.
Friction is a Parachute,
Damping is a speed dependent Brake,
Inertia is a Tailwind, and
Slew is an Acceleration Limiter.
Torque bandwidth reduction - ignores cones if placed too close together.
Reconstruction - Adds cones between cones.
Unfortunately in this day and age Computers are just NOT fast enough to be analog enough to allow for realtime zero filtering as the setpoints for torque are not actually sent close enough together to eliminate overshoot…
This is why iRacings 60hz FFB does feel rougher than 360Hz. it is not due to inaccuracies… it is due to that the time between each set point allows the system to cause a more significant overshoot of the intended location so it keeps trying to correct itself.
If you get the filtering “right” then every time a new set-point is sent by the game the wheel will already be there (or very close) and the result is a smooth and extremely detailed response. Hitting this point though is DEFINITELY not an easy task by any means.
3 Likes
True but forums are sort of organic in nature lol and information is information and it is in essence all related.
Maybe a misunderstanding: the Gforces don’t show up at the wheel. They just have an impact on head movement
True but what I see are people trying to recreate gForce feedback through the use of add ins to the wheels… i.e SoP in irFFB and trying to boost certain signal types through the wheels… Granite has honestly refrained from doing these sorts of things but some games do try to do it… Head tracking and things like that are yes there to alter your view angle but honestly in a static rig in most cases they tend to feel off as all it makes you do is tilt your head… in a Motion rig though where you get small sensation of movement in that direction it can trick your brain into a greater sensation than provided by the external movement…
https://www.assettocorsa.net/forum/index.php?threads/real-head-motion-1-0-2.4755/
Biological facts
The feel of equilibrium, acceleration, gravity, and so bumps are not feel by the eye but with the inner hear. Wich is made of liquid and crystals in circular tunnels.
I think we are getting sidetracked a little more than the sidetrack this already has taken as the discussion moved from setting changes to catch a slide better - filtering and why it is needed (or Not) - to gForces which really don’t mean anything to the wheel conversation except that some believe they can recreate gForce effects through the wheel which can’t be done, as of course the wheel even in the real world doesn’t transmit or add gForce effects… Though gForces in the calculated effect on the steering can add weight to the wheel it is not something that affects your equilibrium.
What are your settings from the Direct Effect fine tuning
For Iracing that game doesn’t use direct input effects and you turn off all iRacing effects and then determine the strength at the wheel through the specific output which is the BASE OUTPUT / MAXFORCE slider setting… WHEELFORCE in iRacing is a safety limiter so as long as the MAX FORCE is not LOWER than your BASE OUTPUT you really do not need it.
Otherwise in TD I leave all of the effects at 100% and if the game has Direct Input effects I tune their intensity at the game…
That was a fantastic spalnation there friend 
Software is just as important as hardware. I bet game devs are scratching their heads seeing all the new wheel bases coming to market… Hopefully your comment about better device support eventuates.
Curious did you manage to find a way to get rid of those sharp torque surges in RF2? Its killing me slowly I cant figure it out. Low slew rate seems to help, have you had any success with the notch filter?
I have not driven in RF2 for a long time but the Skew filter is VERY good at calming the wheel without sacrificing the urgency or the details coming through as the wheel will still react quickly to a change but just won’t be as fast to get to the torque level commanded… I use 3.25 currently on the SC2 Pro but I also have some friction/Damping/ and inertia… My settings are further up this thread (May 4th) if you wish to try them out… I generally use this with iRacing hence the 100%Overall Strength… I would suspect you would want to use about 60% or less with RF2 if it uses a typical gain model for strength setting Though I did read somewhere that they have moved to an Attenuation Model similar to iRacing (though I have not Verified) if so you will want to probably run higher overall strength and then a higher clipping number.
Thanks bsohn,
Yes I’m running around 100% torque in TD with 10% damping/ 15% friction/ 10% inertia/ unlimited torque bandwidth/ 3 recon filter/ slew rate 1.33 on an SC2 pro with a HRS Momo 300mm GT3 wheel. The wheel is quite heavy!
FFB in game for cars is usually low -60%.
Did you try this in the .json?
“Steering torque sensitivity”:1.66, (change it to 1.66, not sure yet if it works as will be trying tomoz.)
“Steering torque sensitivity#”:“Sensitivity curve applied to representable torques: 0.0=low 1.0=linear 2.0=high”,
If you get to it before me let me know if it does anything. Apparently it will even out the peaks.
Which game is that .json adjustment for?? it appears to be an adjustment of the torque curve output (based o the description) which would boost lower signals or surprises them… i.e high being that it curves above the actual linear boosting the low to mid and compressing the higher strengths… This would smooth out peaks as the compression happens at the heavier side of the curve… Going below one would again based on description cause the mid and low signals to be reduced creating a greater difference in peaks at the high force end.
The only way to reduce force at the wheel is really turning down the output to the wheel or the wheel itself… the filters and things only fine tune things they do not generally make dramatic differences… HOWEVER at least with iRacing Car set-up can make a HUGE difference… ie my team was running the 488GTE at LeMans and the initial setups we got had 12 deg of caster… the steering Telemetry was putting out forces in excess of 50Nm which is REALLY HIGH… I adjusted the caster down to 9.7 deg or so and the Telemetry Force Output went down to about 27Nm or a little more than HALF strength reduction which means what was running the wheel at say 20Nm would be coming through the wheel at about 11Nm after the caster change… Nothing other than a set-up adjustment…
2 Likes
I was referring to RF2. Given if RF2 has moved to an Attenuation Model for ffb I have my SC2 pro @100% torque and adjust ffb in game per car. I’m not sure whether car setup i.e. caster etc drastically changes ffb static resistances or kerb telemetry. I still have more settings to try from tomoz so if anyone is having the same issue as me I’ll try posting again tomoz evening or so.
The massive spike isnt on all cars just certain ones that I enjoy racing so maybe it is setup related.
Anyways thanks for your reply all the knowledge helps.
Very interesting to me! Foa, was it straight away clear to you that adjusting camber is THE solution? Do you think that is something only so drastic in Iracing? I experienced some good results with camberextravaganca in AC but never had such a change in torque (it was never necessary to adjust 20% though). And was the difference for you so clear to feel at the steering wheel aswell?
It is Caster (Camber is different in its effects), but yes I have known for a long time that Caster is a HUGE contributor to feel at the wheel in a real race car and it translates into the sim as in real life… I just don’t always check the telemetry unless there is what seems to be an issue… Caster can also affect other things with regard to handling as well (i.e. how the car initially transfers weight from the inside to the outside tires and how it takes a set in the corners. the general thought though is more caster gives better mid corner grip, and it does but it effects entry and exit properties as well.
Yes the feel at the wheel is considerably different both due to the torque change and the mechanical set-up differences.
Now Camber is a different story, Camber will general result in less change in at the wheel feel but extreme levels of camber with soft settings on the tires (low air pressures) can cause more relative weighting of the steering wheel in the corner due to the way camber works is to keep the contact patch as low as possible on the straights to facilitate speed and then as the car puts pressure on the tire the contact patch widens out to allow for better cornering… This additional contact patch in the corners will generally raise at the wheel strength but not nearly as dramatically as Caster.
There is much more to it than this with regard to Camber though as it is also used to keep a tire in optimal temp ranges for grip ie if you are using all of the contact patch the temps will be lower than using half since you are spreading loading.
As for if it should be an effect in other games… It should be if the Car models are designed to translate these properties correctly…
1 Like
I mixed the terms up! But as both are active settings it is totally informative. Thanks for sharing these insights, very kind!! Simracing is getting so realistic, the level of experience will become even more important. I wonder if one day some professional eSports drivers will have their engineers? Anyway, I will read it carefully again, definitely worth it imo! Wish you, and your team mates all the best 
Thanks for taking the time!