Automobilista - How the FFB works and what is new about it

As we use realfeel, the steering arm forces are the source of our FFB. Steering arms are the rods in the suspension that physically connect the steering rack to the uprights / spindles, whatever you wish to call them.

At the end of the day the difference between this steering arm force and a steering torque in Nm is just a 'factor', a multiplication.

Also you can get a steering arm force of 8000 for a Formula Ford and 2000 for a Formula Extreme. This is a bit unlikely but look at the attached sketch, physically where you mount the steering arm changes its force. However it also changes how far it moves left /right, so at the end, if you get say 720 degrees of steering wheel rotation to turn the tires a total angle of 60 degrees, the steering torque would be the same.

We've also changed our use of pneumatic trail a bit, so a car that has a lot of pneumatic trail will generate larger steering arm forces on the same suspension as a tire with less pneumatic trail. This can all vary quite a bit and at the end of the day, the steering arm force in itself can be almost anything.

FFB output isn't anything more or less than a percentage. Send 100% FFB to a Logitech G27 and you may at most get 2Nm of torque. Do this to a big Bodnar and you may get 20Nm. And until a year or two ago, the Bodnar had to be set to ~66% force scale otherwise it would start clipping. And your Logitech software may be set to 50% or 150% FFB strength affecting how it reacts to this 100% FFB 'demand' from the sim.

So summing this all up we:
- We rarely have tire force or self alligning torque data
- And if we do, there is certainly a +25 -25% tolerance on this data
- We don't always have suspension geometry data
- And if we do, it isn't always completely accurate either
- We don't have any reliable way to send a torque to steering wheels
- FFB wheels are quite variable and not really made to work with a torque input, they just see a % and their output may not always be the torque you think they should be putting out
- Peoples tastes are different..

So even with good intentions we can be, say, 50% off with the torque we would feel at our wheel and what would be 'realistic'. And indeed we don't have any sort of power steering simulation at the moment either.

Probably 90%, or more likely 95+% of people are using Logitech / Thrustmaster / Fanatec wheels where the torque capacity just isn't there to realistically provide race car levels of feedback (on most cars anyway)..

So at the end of all with current technology it would LOOK nice if we somehow provided torque information, but there would be a considerable margin of error on this and for 95% of users it wouldn't really matter as they want each car to max out their FFB wheel just to feel 'something reasonable'.

Now in an ideal future this is all different, but for now the (relatively) few users of high end wheels may need to subjectively change the force feedback strenght from car to car.

If your starting point is a RealFeelPlugin.ini with little or no clipping, then you can simply adjust the in game FFB % strength from car to car. That is how I do it anyway!

 

Hi Neil, would you please be able to share the settings you use on your bodnar?? I have the v2 54 and struggling to set it up with AMS

 

Thank you for the explanation Niels; it makes sense to me.

Happy New Year!

 

And somehow AMS has consistently natural and plausible feeling cars that all feel like they live in the same physics universe as each other. Keep doing what you are doing. I'll happily ignore the numbers in favour of something that "just works." You have figured out how to get the most out of the simulation engine...please continue to do so

 

That's not my experience, The FFB of the Cars who where in game when AMS was published feels better then the FFB of the "New Cars"

 

Nope, not ranting but discussing and trying to understand partly out of curiosity/interest but, more importantly, in order to have the best AMS experience possible.

The reason why I'm trying to figure this out is because I have an industrial-style wheel (OSW) - not something like a Logitech - and I'm therefore trying to find out FFB strength DIFFERENCES for each car RELATIVE to each other. Eg. A Formula Renault 3.5 should have much higher forces, overall, than a road-car or even an F1 car. The only way to do this is to look at MFASR forces in telemetry but from what you and Niels said, it seems like this won't give us a good indication due to every car in the game having it's MFASR measured at a different spot RELATIVE to other cars (if I'm understanding correctly).

I don't really care to get an accurate number for FFB strength - it doesnt even have to be in Nm; all I'm trying to do is find a general FFB strength for each car RELATIVE to eachother.

Even when a rental kart isn't lifting it's rear wheel, it's not like 5 or 10 times lighter to steer and control than other open-wheel cars yet AMS's MFASR #s show just that: telemetry shows 200s meanwhile other open-wheels show 800, 1500, 2500, etc.

 

A Formula Renault 3.5 is not necessarily much higher forces than a road car, except perhaps at rest when the road car's power steering takes over, or if the road car is a 60's American boat that you could steer with one finger because of the over-boosted power steering.

Most non-power steering race cars and karts have steering force requirements within the same general range. You want lots of road feel, but it can't require a body builder to turn the wheel--because all racers are certainly not body builders. I would say a rental go kart and non-power steering open wheel beginner car (like a Skippy) have about the same amount of steering effort required in real life. Those cars are also very similar to a Radical SR3. They're all about the same which is to say your arms will be tired after a while if you are on a track that requires a lot of turning relative to rest time.

The current gen Indycar is suggested to be the heaviest steering of any modern race car. But there are women driving them and others who are not all Tony Kanaan muscle men, so it is still relative.

I think that for anyone who is used to non-power steering cars (which is just about no one of the current generation), race cars are one or two steps stiffer than a road car--mostly because of the larger and grippier tires. But they are not so much heavier or so variable from one to the next. Until we have power steering simulated, I don't think any of this is going to be too precise or accurate.

 

Marc, a FR3.5 (just an example) outputs some very high forces, so do Indycars; F1s are less. Apart from watching onboards, the high forces of many racecars are quite documented.

 

I wonder if anyone could get some real data on this--would be very helpful for modders and Reiza, S397, etc.

Would be great to see the range of forces for street cars, karts, and a variety of well known race cars.

 

So is there a simple way to know whether I should be reduce low force boost or ffb strength?

for logitech reiza recommends 80% lfb but i'm just curious if i should get more detail with less than 80% if i can reduce lfb without dead zone happening?

reason for asking is that i did some tests with vintage because i get much more detail (especially front grip loss/regain) from retro and ultima than vintage and don't know why so i though maybe some clipping.

also several other drivers with g27 are running very low ffb strength or at least much lower than 100%.

so should i keep 80% lfb and see how low i can make ffb strength without being too light or dead zone occuring?

or should i also try to reduce lfb? or both lfb and strength? you get the idea.

edit - and why is vintage ffb not as detailed as ultima and especially retro!?!?

 

I dropped Low Force Boost from 80% to 70% and felt more in control of cars, managing slides and not crashing all the time. After a bit more hot-lapping and trying different values, I have settled for keeping LFB at the recommended 80% but lowering the FFB to 95%. AMS feels great at the moment.

@Professionalsimracing.com, your suggested numbers (70% FFB and 80% LFB) allowed for very quick corrections but my trusty DFGT felt a bit light and numb.

 

Yeah I have to think that lowering lfb to 70% felt better because with 100% plus 80% is VERY strong ffb and you have to be clipping huge based on how that feels to me.

So right now I'm at 80% lfb and 80% ffb strength but really not much more detail if any than at 90% so yeah maybe not much cost for more strength at least with regard to direct impact on driving.

I've never had any issues regarding driving however as long as I can feel the rear of the car and center is not sloppy.

 

As far as I'm aware there is no relation to your wheel strength in your software to clipping. I thought its all about the realfeel maxforseatsteeringrack setting. This is what sends the signal to your wheel. If the signal tells your wheel to clip then your wheel will clip at whatever strenght setting you have in your wheel software.

 

I agree, Supa, I thought it was like that too - that the in-game FFB is just a sort of final multiplier of the force that the game has been told to send out based on MaxForceAtSteeringRack but, having said that, things always seemed to clip more/earlier when the in-game FFB was set to 100% compared to when setting it to 70 - 80-ish. This was with any wheel I used - Thrustmaster as-well as Fanatec. I lowered the wheel's control panel strength drastically to make sure it wasn't physical/hardware clipping but it still seemed like the clipping happened (pointing to game/software clipping) and less so when in-game was set to 70 - 80 % compared to 100%.

What I think is happening is the following:
Reiza, understandably so, has set the cars' MaxForceAtSteeringRack to a number which clips the last 20 - 30% otherwise the majority of players would be complaining of overly light FFB. The in-game FFB must have some connection to RealFeel and scale the FFB down relative to the MaxForceAtSteeringRack and therefore the last 20-30 % that would normally clip doesn't clip if the in-game FFB is lowered. So, let's say your car is hitting 3200 during a corner but MaxForceAtSteeringRack is set to 3000. Well, if your in-game FFB is set to 80%, the 3200 during that corner will actually be as if you were only hitting 2560 (80% of 3200). And 2560 is obviously not 100% of 3000 so you won't be clipping. It'll be as if the car is simply hitting 2560 while it's MaxForceAtSteeringRack is set to 3000. 2560 is around 85% of 3000 so the game is only sending out a 85% signal (ie. not software clipping).

I could be wrong about all this but, from what I remember, lowering in-game seems to affect/improve FFB range (ie. less software clipping) when I would have thought that only MaxForceAtSteeringRack could do that.


EDIT:
What I explained above can't be true otherwise lowering the in-game FFB would still allow a 100% signal (ie. your wheel's full power) to be sent from game to wheel. Eg. in-game FFB @ 80%, MaxForceAtSteeringRack @ 3000, car hits 3750 during a corner; 80% of 3750 = 3000 which means 100% signal. I don't know then...

 

As I suggested in another thread, if the busy developers can't spare the time to make a clipping app, could they throw us a fake graphic and call it the Placebo Tool? Release date: April 1st.

 

There is an interesting thread in the AC forums regarding the effects of clipping and the relationship between various FFB factors (Hardware vs soft-clipping). Hopefully, it may reveal some general formula that results in the best detail with relatively high force settings.

There are three general approaches being analyzed and displayed in graph-form; high game-FFB output / low wheel-force settings vs. low game-FFB output / high wheel-force settings vs. both mid-range settings.

Of course, different steering systems will produce different results but, it might be useful as a general guideline regarding which approach tends to yield best results.

 

Hey Spin. I used RF2 with the FFB meter to test this and come to this conclusion. You can clearly see if a car is clipping or not. Changing my Bodnar software has no effect on the FFB meter readings. If the RF2 FFB meter shows a car clipping then lowering the Bodnar to a low number like %20 doesnt stop the clipping signal being sent to your wheel.

 

After more time with Vintage at 80% strength I am beginning to notice improved FFB. If I get bored I may fiddle some more with reducing LFB and FFB strength some more. Really feeling good though so not very motivated which is a good thing!

 

Yeah, any clipping on the software (Game-FFB) side is carried over to the hardware side but, a game FFB-meter can't tell if there's clipping on the hardware side.

 

Guys, what about the "ConsoleEnabled=True" line on RealFeelPlugin.ini? Isn't that a way to know if the software side is clipping?