Thank you so much 
1 Like
Let the test restarts!!! 
EDIT
I found SteeringFFbSmooth realy useful and seeing your test result VS real car, I found that im at 25% of iR current force with a value of 0.25 on that line, and coincidentally matches your findings 
Interesting, is that for the corner forces or the curb peaks?
It is more noticeable on the responsiveness of the forces.
Start from 0.1 then 0.15, 0.2, 0.25…
From 0 to 0.2 i can feel lag/latency
Thanks a lot!! You are always very kind and nice. Is a pleasure
1 Like
There is ONE MAJOR factor missing here and it is something that you don’t have in the data you have provided and that is the set-up…
If you are using the baseline setup or one based on the Baseline set-up for the DW12 then it is generally using an 8tooth steering pinion…
From conversations with Real world Indy Car drivers they will use ONLY the 6tooth pinion on the real car as otherwise you get extraordinary forces like what you are seeing… When using the 6 tooth pinion with the DW12 you should see Steering output numbers MUCH closer to the real numbers of the car output… As well setup in general caster angle, Camber, Toe, etc can also cause difference in output through the steering column.
I have seen data from Barber Motorsports Park that with the same setup on the DW12 the Steering torque graphs aligned almost identically. ecxept for a few outliers… Unfortunately I am not at liberty to show this data as it is not mine to show AND Yes it could be a little out of date given the newer iRacing Tire Model Data as it was early in the DW12, so before the IR18…
The DW12 though with the 8 tooth pinion will Spike VERY VERY high in torque due to its quicker steering ratio. I think Iracing settled on thins in regard to their baseline set-ups due to that higher torque output translates well to less powerful wheels by giving more active indication of the amount of torque.
I racing smoothing setting in the app.ini is much like the Reconstruction filter in a way but done on the telemetry feed… It can be useful to create a smoother output from the game but it comes at a cost as it must do additional processing before being output as a % to the device.
The interesting thing is that the close you get to your the wheel being natural in feel the spikiness will come closer to that of the real cars line as the changes you make in the wheel will allow you to control the wheel movement (eliminating the spikes) and therefore be closer to the real output…
@CLAYREGAZZONI - With regard to force from iRacing you would increase the MAX force in iRacing but with the DW12 IF you are getting 120NM spikes you would need a much higher number to not be beat up still as the telemetry coming from the game doesn’t really change (other than with Set-up of the car)… SO TD would be the place to reduce the force for this particular car.
I would be curious for you to try the settings I have for the SC2 Pro along with the set-up with a 6 tooth pinion in the DW12 set-up to see how close they fall… Granted the other suspension set-up issues would be different but the 6 tooth is the MAJOR factor difference in all the DW12 Road Settings.
For me the car is fully drivable with the MAX Force at 42Nm (.595:1) but Because the DW12 is a beast I usually will end up with at the wheel force of about 60 Max force or .417:1 because at .595 I struggle to make it through a race.
With the 6 tooth pinion, it had a massive impact on the high spikes dropped from 120Nm down to around 60Nm but for the corner forces not so much. They are still around twice as high as the real car. The graph is with the HDF iRacing setup the same I used in the first picture, the only change I made was the steering pinion.
It would be interesting if had more real data for other cars as well.
I still think it shows that we can’t have a true 1:1 ratio at least not for the Dallara DW12 and that there is no need for a much stronger wheel base even if we could set it to a true 1:1 ratio.
There are definitely some things that are off in the iRacing output as of course the iRacing models cannot take every factor into account with regard to car construction… I am not sure if you ran that last test with the settings I provided or the settings that you have for the SC2 as there should be a little bit of a difference there as the settings that I have above are more weighted than most… The other thing to consider is that the actual car telemetry has been set up to run LongBeach very well most likely, while the Default iRacing Set-ups are generic… Long Beach is a VERY bumpy track and from the past talking with people who have set-up indycars for the race (not the DW12, prior indy cars) generally speaking in real life they run very Soft suspension to absorb curb strikes and undulations in the road… So I am guessing with a fully optimized tune to absorb running over the curbs on the suspension many of the larger changes in force would actually be a lot smaller…
The running Average force is what you need to be looking at to determine how on target things are for both telemetry outputs… as the oscillation factor can come from as mentioned above the suspension settings and how the car is being driven (ie over the curbs, major road bumps, etc…)
In looking at the graph you posted the averaging would probably place the peak average output from the car at around 27Nm which is about where I would expect it with variations based on all the above.
If you look at your original graph you will see that the the steering angle on the wheel is moving back and forth a lot even when the car is going straight on the iRacing Telemetry and that you are actually tending to turn the steering wheel more in the corners… The steering wheel moving all the time actually means that you are probably a little under dampened on the filters which is allowing the forces that come from the telemetry to be harder than they would if the wheel were not moving as with FFB the further the wheel gets off of its target position the more it tried to add a correcting force… This is unfortunately the nature of FFB that is not present in a real car. This situation is self triggering in most cases and is part of the reason why FFB wheels end up feeling heavier than the real car… A Real car doesn’t fight you going a bit off of actual 0 torque, a FFB will. This is a big reason why the real car steering angle is generally much more consistent and smooth when looking at the graph. The key here is to dampen the FFB wheel enough so that the wheel has a hard time moving around on the straight but is not sluggish to react when you want to change direction… You can set up the car with additional caster which will help have a smaller self centering window BUT more caster in a car also makes the steering much harder to turn.
The Second thing is that in the first graph you are turning the wheel much more that the driver of the real car (5 - 10 degrees is a big difference in a race car). This would generally say that the car is not rotating well and causing understeer… In this case the cornering forces will generally also be greater because you are essentially trying to force the tires into doing something that they do not want to do so with that added angle brings more resistance until you actually start to understeer and then resistance actually drops since the tires are resisting the road less and begin to slide… SIDE NOTE: The interesting thing is that many tires actually get the most Grip right at the point they start to understeer, so, optimally, both for the speed of the car and the forces you get through the wheel if you can keep the tire right at that slip angle you will be the fastest… This is where the loose is fast term comes from as having the rear tires in this slight state of lost grip (due to the optimal slip angle) it puts a car right on the edge of spinning but at the same time you have the most grip.
In the Barber telemetry that I saw there were definitely some quick spikes that went well above the actual car telemetry output, which were explained to me as calculation errors that do occur from time to time (which is inevitable really as there are a lot of in the moment calculations going on).
The basic thing is you can have a true 1:1 but your FFB wheel does have to be set-up perfectly which is almost impossible… You can get close though if you come from the standpoint that the Wheel and its filters are there to control the way the wheel reacts and NOT to change the feel coming through the wheel… Many set-up the wheels very active (which is fine if that is what they want) but when you do so it amplifies the FFB loop which causes the massive spikiness and the tank slapping sort of return of the week… this makes the forces feel to you as the recipient that they are much higher than they should be… this is ALSO where the issue is that you mentioned above with the speed of acceleration of the wheel to a certain Nm which is VERY true DD wheels in general are faster at getting to a force than a real car would be, again it is reflected in the back and forth nature of the telemetry line of the steering angle.
Personally from my experience and the issues around FFB with regard to the recreation I find that .6:1 or close to that is about real in terms of the amount of fatigue you would feel in a race. It is not an easy level of force but it is not overwhelming either and as I said I can’t even deal with the Indy cars at that level for the full length of a race as they are so RAW. I have found that .4-.5:1 is generally where most people blindly fall into what they feel is right (mostly in the mid to low of that range).
In all reality with the 6 tooth pinion in the setup and a good setup a SC2 Ultimate would pretty much give you a 1:1 output with the DW12… No other commercial wheel could do that currently. but the Large Mige SC1 and the New Asetek may come close as well. Otherwise you would be going custom like Beanos, which, BTW is a VERY Dampened setup even mores than my settings that I posted above. I would probably almost guess that his set-up would fall even closer to the Real Car output as far as reaction than mine.
First I’m not used to that car, so yes my inputs isn’t the best, but even with 5-10 degrees more input it shouldn’t increase the corner forces that much, because I’m much closer to under-steer which would have give a drop off in the corner forces. Most tyres have a drop off in the resistance/forces before it reach the maximum grip and start to under-steer which will then have a much steeper drop off in the resistance/forces.
The car setup (caster, toe, camber etc) plays a role but not so it goes up twice, for the corner forces at least not if you run a somehow realistic setup.
That is true at least the near impossible part
I know from my own experience with different wheel bases how hard it can be, but it also depends heavily on the wheel and the options you have in the driver if you get a good FFB or not.
The main problem for me is you need to choose between either the right behavior and very little feedback (small forces). Or a more active wheel where you have the forces all over the place, in both cases you don’t get the small forces. It’s either tuned out to much (damped) or it is to active and the spikes takes over.
I hope this will give people a better understanding of what they are dealing with when it comes to tune in their wheel bases in iRacing.
That’s why I posted it.
I do appreciate your thoughts and I’m learning a lot from it, the technical par of FFB which i know you know a lot about.
I have experienced what iRacing can do when you get it right, (which is nearly impossible). In my experience we are not there yet.
It’s way better than people give it credit for
What I would like to see is a tool in the driver where we can actually manipulate the strength of the different forces to be able to either lower or raise them. That way we can use the damping, friction and inertia purely to get the right behavior in the wheel, regardless of what specific output we are using. As it is now we are forced to use damping, friction and inertia to tune out peaks. Because of that we are losing all the small forces.
I think this shows the need for such a tool
That is sort of the thing though… is that it will increase until you actually get right at understeer but that transition happens fast if you are not exactly online. With a car like the DW12 with high downforce and quite sticky tires with no power steering when you start to get where the tires protest the forces will increase significantly. actually happens in all cars but power steering can mask this significant increase. This is why it is generally easier to drive the car online as opposed to off as offline you are asking the tires to do more and they will protest until they just give up (understeer or oversteer). Caster can easily add a significantly noticeable amount of weight to the steering… But at the same time when you add caster you do change the way the car reacts in the corner… i.e. minimal caster will be very light steering but will also tend to cause the tires to lose grip quicker.
The one thing to remember when looking at the graphs you posted is that the average is not double… You are seeing the spikes in the FFB as double. Those spikes come from your wheel position changes, induced wheel position changes (road changes), and as well the FFB corrective changes (which is a nature of FFB). As you lower the force of the wheel in your hands you would probably see a reduction in the spikiness of the output as you would be “fighting the wheel less”
This is the unfortunate balancing act with some devices and it is one of those areas where I do Believe that the SC2 has an advantage as they have enough filters and options to be able to for the most part balance things… (which is why I am fairly vocal when they talk of removing or combining settings) But it also comes down to personal preference. The issue in many cases is that in a sim we are unable to feel and chassis vibration which is where MOST of your information about the car comes from (through your butt and other senses, Seat of Pants feel), very little of this information is transferred through the steering and almost none of it is transferred directly through the steering Rack. The vibration felt in a real car is generally through the chassis into the steering column and then into our hand… However, as sim racers we want to feel these feelings through the wheel … but with a game like iRacing they just don’t produce this information to be sent through the steering so we liven the wheel up allowing it to be somewhat unstable and in some cases VERY unstable to try to recreate that information that we are not getting… This is where a large majority of the issues come in.
We are effectively trying to recreate information that isn’t present by boosting the minor references to that information that come through the steering. which intern boosts everything to the point of being a bit brutal, then we have to turn down the wheel and in doing so we start to loose more important information because its amplitude becomes too low to feel. Then we complain we can’t feel these things and it is the Game,Wheel, etc… fault when in reality it is our fault as we are asking the output to be something that it is not.
The best thing for FFB currently that I have seen is actually Logitechs TrueForce… in essence they have separated the Telemetry and the chassis vibration and merge them to the wheel. This minimizes the effect seen on the FFB as they can be controlled and balanced more like a real car… They use the Audio Rumble (like SimVibe) to create the chassis vibration effect separate from the FFB signal and then inject that into the Wheel output to recreate the chassis vibration and give a more true sense of the wheel feel. It is Very smart and it has been done effectively by SimRig manufacturers for a while where they will place a vibration puck on the side of the steering uprights to send a vibration to the wheel.
What I would like to see is a tool in the driver where we can actually manipulate the strength of the different forces to be able to either lower or raise them. That way we can use the damping, friction and inertia purely to get the right behavior in the wheel, regardless of what specific output we are using. As it is now we are forced to use damping, friction and inertia to tune out peaks. Because of that we are losing all the small forces.
I think this shows the need for such a tool.
Brion has given good comments earlier. Something small from my side: IRL, you will not feel those sharp high-amplitude spikes, they will be dampened by the steering system anyway. Thus using Damping, Friction and Inertia as a 1st step to simulate almost what you’d feel IRL, then lastly Slew to control the ramp-rate of the servo output to create softer or harder race-tyre feel, you’re good to go.
Those small forces people are after in the steering column is mostly noise and not really felt at all IRL. If you wish to get road-effects and those in iRacing, it is best to use tactile transducers via LFE output. That gives the best of both worlds.
Anyway, as Brion has alluded to earlier, I typically (used to) run quite high forces on the DW12 and set quite a few practice WR’s, but, my wheel was set up to replicate real life feel, thus the sharp spikes were filtered, at which point the high torque ffb becomes significantly more manageable.
You are still able to feel when the front tyres are at the limit in the steering wheel IRL.
iRacing used to have them in the steering wheel and they were not noise. They were to distinct to be mistaken for dirt or noise.
If iRacing has taken that away from the steering wheel it’s the single biggest mistake iRacing have ever done.
Some of those forces aren’t in the LFE so you can’t get them with transducers either.
Even on my big AKM65 servo with more inertia than any of the current SC2 servos, and with my higher than normal D, F and I settings, running Recon Filter at 3, I can very clearly feel when the front tyres let go at the edge of traction, or when you run on higher profile tyres, and you go in to hot, the sidewall starts buckling…those are rather distinct and clear to me.
Perhaps it might be because one gets to ‘hear’ what the FFB tells you after 8 years of testing things at the limits, not sure, but I have heard before guys cannot tell when the fronts hits the limits…I definitely can feel that….most likely because those cues are not masked by overly noisy ffb settings due to way to low filter settings…
I find iRacing ffb cues quite accurate. Just IMHO of course…
Thanks, that is the kind of feedback I’m after in the steering wheel. Including the ability to feel derbies level of sand on track and all that kind of feedback. I had all that with my other wheels but I can’t get it with my SC nor the 1 or 2.
They are as you say very distinct when you get it.
To be honest when I had it, iRacing may actually have had the most detailed FFB of all sims. I didn’t had the need for transducers or motion rigs and more important I wasn’t dependent on the sound for tyre scrub, I could feel it in the steering wheel.
Today I don’t have any of those details and I can’t get regardless of what I do. Because there is too much dirt that needs to be tuned out or introduced when you apply some settings.
1 Like
One of the other things to not remember but a lot of people don’t know is that over the past 4-5 years iRacing ahas been moving to a “real” tire model where they can specify all aspects of a tire that is placed on a certain car (including sidewall build, tire compound softness, wear traits, etc…), This means that they are trying to duplicate real life in the way the tires actually react to surfaces… These are not real tires and will never be due to that in talking with iRacing staff and noting a bit about the industry the tire companies WILL NOT give up any data directly so they have to work it out themselves… In the Past and with most games they do use what is called a theoretical tire model which doesn’t take into effect as many different parameters. it is vey probably that this model Due to being more theoretical it is probably that this model gives more feedback in certain respects.
If you have ever had to change the tires on your real car you would realize that changing brands, compounds, tread patterns, sidewall ratio, etc… all can have a dramatic effect on the way the car feels both in driving and in response… Some tires will communicate grip in a much more direct way where others will mask things. This is true of race and road tires.
The other thing is a lot of what you are describing as wanting to feel (sand, debris, grass, even in some cases curbing) with regard to vibrations and things truly do not come through the steering system, they are translated through the chassis (which is where LFE Pucks work wonders). Things like the tires sliding on the surface due to sand will change the weight of the wheel but the scrub vibration is translated through the wheel to the hub to the suspension to the chassis to the steering column and then to your hands. It is not actually a force that is coming through the steering tie rods in the car and through the rack.
That is something I really like in WRC9: I can feel really good how the surface changes, Fe when it starts/stops to rain. Or how tyre performance is dropping on Tarmac, how different it is to drive on snow: with the exact same settings the car feels much lighter, steering inputs are needing more time to get recognized (of course a car on snow drifting sideways behaves very different, and this is well done in the game).
When I read about the Iracing tyre model I thought by myself “how appropriate can it be without having data from manufacturers”, and then you mentioned exactly this point.
What you think about the fact that we all use different steering wheels? It’s where we get almost all FFB from, and we never really focus on this fact.
Last question: Iracing has a 60 Hertz FFB frequency (if this is still correct), where other games have 333 Hertz. ATS changed to 120 Hertz in 21, and also started a new FFB physics engine which is not only based on artificial effects, but real life data. When will Iracing make this step?
What I had and are missing these days.
Road texture.
I could feel the difference of how smooth or rough the asphalt was, difference between asphalt, concrete, gravel, grass etc. they all felt different. These days everything feels the same, super smooth even dirt feels the same as asphalt. It’s just the bumpiness of the track that’s different.
Tyre scrub.
I could feel when the tyres were under- over- or at the limit. I could feel the impact of tyre pressure and if it was right even the difference before it came up to the right pressure. When the sidewall started to bend. Tyre ware. How much steering input I had based on scrub. All of those small differences.
Acceleration/deceleration.
I could feel the rotation speed of the tyres spinning up/down. I could feel the amount of braking, if one tyre had less grip. If I was close to or locked up a tyre and which one or both. I could even feel and tell when the car come to a full stop just before it happened.
All those small details.
Dirt on track.
Rubber, sand, gravel and the amount of it, more sand felt different the less. I could feel the tyres collecting the dirt and when it started to fall of the tyres. I could feel the gap between the asphalt layers and how it moved from side to side over the tyres.
The gyroscopic effects.
I could feel if I had an unbalance in the car, if there were a damage or miss align in toe or camber and things like that.
I have probably missed some effects or put some of them under the wrong category but you get the hint.
Today I don’t have any of it and to be honest it feels like crap and like an arcade game in comparison.
I wish more people have experienced what I had to understand the level of details we had in iRacing, and how good the FFB really was. It was in line with rFactor. Personally I hold iRacings FFB back then higher than rFactors.
It was so immersive compared to what I have today.
I can honestly say that none of the settings (all settings in True drive and iRacing app.ini) or profiles from forums and paddock (I have tried more than 200) so far with SC1 and SC2 have been able to give me any road texture including dirt on track or tyre scrub including acc. deceleration. When it comes to the gyroscopic effects only a few settings have been able to give me that unless I’m about to lose a wheel.
If any of you can truly claim you can feel road texture and tyre scrub in the steering wheel, please share your settings.
Why is iRacing spending a lot of money on trying to get the perfect tyre model and dynamic tracks if we are not supposed to feel it in the steering wheel anymore? LFE, butt kickers and motion rigs are just enhancement, and should not be mandatory to be able to feel those effects.
If iRacing has taking all that feedback away from the steering wheel (which I don’t think), it is the single biggest mistake they have ever done. If iRacing hasn’t taking it away from the steering wheel (which I hope they haven’t) then there is a lot of work to do, to get it with a SC2.
You do get the overall force feeling differences and the details of the road are very subtle in iRacing when you don’t have full vibration through LFE’s. So you do get all of the appropriate information. it just doesn’t come through quite as bluntly…
The settings i have for the SC2Pro are very good at communicating when the car is getting loose or understeering and I can easily anticipate and correct slides. Rarely do I ever spin the car without knowing that I have done something to make it spin and what it is that I did… There are some cars however that are not as good as others at giving you certain keys through the steering… i.e. the Mercedes GT3 2020 can get very dull when getting close to understeer or over while the F488 will give you quite a bit of information due to changes just in the weight of the wheel… so it varies greatly by car… Which is completely normal in the real world as well, some cars just communicate things better…
The thing with iRacing is they are creating real tires in a digital form, It’s complex, and it works… But the tires are “Different” than real life. What they ar trying to get to is something to the effect where they can essentially build an in sim tire that will react if THAT spec tire would be built that way in real life… But as I said Tire Companies are extremely secretive about their data. Even Race teams have no clue as to all the details about the tires they use… They put them on the car, Hope they feel good, and if they don’t they Complain… and Drivers do it worse than anyone when they don’t feel like they think they should… But they have to live with it unless the company is willing to change it and then again hope they get what they want…
We as Sim Racers are pretty much like the drivers, We want what we want… but then that might not be what we get… But then we try to solve it in other ways. Some of which MAY work… Some of which we try to pull information out of specs of information that isn’t really there or supposed to be there (i.e. errors or artifacts in th FFB signal) by using other settings we have at our disposal.
On Steering wheels… While there is merit to the leverage on the wheel creating more or less force. My personal opinion on different size wheels is more about steering ratio, I generally don’t find that a smaller wheel causes me more physical fatigue than a larger or visa versa, so the amount of force exerted by the driver is probably about the same but maybe over different time elements based on ratio. The smaller wheel you use the quicker you will get your arms to the angle that you might want to be at. This does cause a mismatch in the speed at which the steering commands get to the car as opposed to the cars actual steering wheel. i.e running a formula car with a Nascar wheel you will always feel a little behind the steering or that you have to make really drastic inputs because you have slowed the steering ratio by a small amount. This makes it harder to whip the car around hairpins and makes the car harder to catch because if you flick the steering with a small wheel you have turned more than the same flick with a larger wheel. Going the other way has sort of the same effects except you may find yourself causing the car to understeer prematurely because the wheel is turned to fast or that the back end gets upset very early because you have turned the wheel more and the front end grips on you quicker… These are things that can of course be learned to compensate for by using slower movements with a smaller wheel or flailing your arms around with a larger wheel… Sometimes I have found going to a smaller wheel on a car that I just can’t seem to get to turn makes the car turn better, this would be me just not getting my brain to wrap around just how much turn is needed in the wheel with the larger wheel. If the wheel and car are matched though you would get the most accurate mapping and ratio.
As for the Refresh rate for iRacing… That will probably not be updated anytime soon. The reasons are mainly that it would cause a complete rewire of the physics engine. If they do ever move to one of the standard physics platforms it would probably be easier but at the same time it is going to be a large job… Part of the issue is their support of slower computers… Physics unfortunately cannot be multithreaded due to the complex timings required for the interaction of all tracked parts (road, all suspension and chassis dynamic, relationship to other cars, and as well the output to the feedback systems) this means that increases in the speeds could have detrimental effects on the speed at which information actually is output to the devices on a slower computer as well as affect network races depending on how much interaction occurs with the other cars (worse NetCode Issues).
Personally while I believe it would smooth out some of the artifacts which can help with oscillatory behavior making for a nicer straight line and certain circumstance driving attitude (much like the higher resolution encoders smoothed out the servos with the OSW) I’m not sure it would truly revolutionize the FFB. Basically meaning I don’t think you will all of the sudden FIND a curb feel or an understeer feel or all of the sudden it will be easier to hold a drift. As currently the information is all there and IF your motions are on target with the point to point information provided by the telemetry at he current rate. system the overall feel of the information would be virtually the same. However if you are off the points since the points would be closer together the reactive counterforce would probably be less making the wheel feel overall lighter and then people would be craving higher forces as they wouldn’t be getting as beat by the wheel (which goes back to the active set-up of the wheel to less active) A more active wheel would be calmed… a less active wheel it wouldn’t change much…