Guide: Optimal FFB settings for rFactor 2 - The key to being in the "Zone" :D

if you are still interesting:
http://www.youtube.com/watch?v=cauNbx5M_Xg
you can see settings on the wheel, driver and in game setting (these you can find at the end)
first try with STM 0,00315. You can notice tendency to oscillate and deadzone feeling. I prefer the word feeling because it is not true deadzone, the wheels reacts instantly to whatever movement of the steering wheel, also within "deadzone" range
then with STM 0. I provoked somethig like deadzone feeling also with this setting, it has its own sound manifest but feeling is different - initial movement is stiffer and "border" is weaker. So I think I can confirm that this deadzone feeling is built in already but increasing STM amplifies it.

 

Yes, there is no deadzone on my wheel normally. With STM, it causes a deadzone that is proportional to the size of STM. Axe describes it well.

I have never had an oscillation issue with my CSW wheel, but I certainly know what it is and have experienced it with previous wheels. I don't need to adjust FFB to prevent clipping--everything is set at max or defaults and it feels great.

The only issue is a bit of vagueness near centre that STM cures. I don't have time now, but I will experiment going back to 0 and then comparing with a very small number like 0.01000. Perhaps with this wheel all that is needed is to make it non-zero? Or, we need some other combination of adjustments to keep the good part of STM which is that the lower powered forces feel more alive.

Also, since not everyone uses the same Windows driver settings or settings on the wheel itself (if you have Fanatec), it's difficult to compare and not everyone will have the same experience. And even when there are absolutely identical results, two people describe them and experience them differently enough that it can be confounding.

 

@DrR1pper
Nice thread, thanks for sharing your knowledge! Thanks too for clarify the dead center question.
IMHO this thread deserves to be fixed.

@Andrea
Nice numerical approach. Excellent point of view!
But dont take me wrong, please. I will take some words.

Fact is that FFB and Steering Rack Forces must have a linear relationship (no doubt) but this isnt sufficient if you consider that the ideal solution for this case is when Y = a + b.X is found at a=0 and b=1 or simply Y = X. See graph.

Any angular coefficient (b) different from 1 will introduce biased results at one or both ends. In other words, we are looking for parallels close to the origin. How much b must be close to 1 ? The more, the better and if I can make one note about your data, is that I really dont see so big different r2. Certainly a deviation of 20% (0.8 and 1.2) is too excessive. So I suppose that the best approach is to consider both r2 and b.

My opinion is that low end forces are way too more important than the high ones, since the extreme forces will happen in situations where you are over curbs or out of track.

WheelCheck give forces as percent (times 100) and DeltaXDeg could be easily transformed to percentages in order to plot a graph of force against displacement (XDeg) comparable to the ideal Y = X

Still a WIP but I suppose one can validate the setup numbers running the TechAde's rF2 Telemetry CSV Plugin v2 found here http://isiforums.net/f/showthread.php/18286-rF2-Telemetry-CSV-Plugin-v2
Plot the telemetry data of the Force at Steering Rack (in Newtons) by the resulting ffb (as % from total). Graph to see the actual dead zone you're running. Sorry to dont post my data, Im out of home.

 

Yes, to be clear the STM is not introducing an actual deadzone as my earlier post said/suggested, but a "slop" zone of strange lack of feedback. The connection to the wheels in the sim in terms of turning motion is unaffected and has no traditional deadzone.

 

Hold on...

Axe, you said "it is not true deadzone, the wheels reacts instantly to whatever movement of the steering wheel". I just want to be sure we are talking about the same thing here.

I don't mean an "deadzone" as in an initial region where you turn the wheel and nothing happens to the ingame wheel. I meant an "ffb deadzone" where the initial region of turning the wheel does not produce any ffb forces at your wheel despite the fact that it is turning the wheel in the game. These are two very different things.

 

Yes, a FFB dead zone, though it's not literally dead, just weird. We're having trouble describing because it is strange. Definitely artificial/digital feeling that ruins any immersion/realism. I am about to test some more so will report back later.

 

By the way, does the main controller.ini override settings in custom controller.ini's (in the Controller folder), or do we have to make changes in both places every time? To be certain I have been doing that, but it would be more convenient if testing only had to affect one file.

 

I honestly don't know but i've only touched the controller.ini file.

 

OK, I confirm, we are talking about FFB deadzone. Did you see the new video? Is there anything suspicious in the settings?

 

Sorry, i just watched it now. Aside from the slight oscillation i can see in your wheel which i presume is from the stm being too high(?), if your talking about the ffb deadzone which feels a little resistant when your stationary in the car, this is perfectly normal.

Road cars do not feel like this because they have power steering which stops you from feeling the resistant forces of the tyres to turning when stationary (due to geometry of the steering pivot to the tyre contact point causing scrubbing). When the car gets moving though (even just a little bit) this resistant force to turning reduces dramatically but you would not notice this change if your car has power steering.

This is also why good teachers insist their student to only turn the wheel when the car has started moving because otherwise turning the steering wheel when stationary can damage your tyres and power steering.

I am suprised though by how little stm you used (only 0.35%) that caused oscillation in your wheel. It doesn't make sense and despite your showing me the evidence for it, i can't help but wonder if it's something else causing this. I had the CSR-Elite and although i sold it before i could test the stm with it, i know it had a terrible ffb deadzone problem. Maybe it does but only needs a very small amount of stm.

What was your ingame ffb multiplier as well?

 

In fact, i'm certain there is an ffb deadzone on the CSR-Elite because someone on the forums pm'ed me for the tool to check their own wheels ffb output. They did the test but didn't have excel to make a graph of it so i offered to. Here's the graph from their csr-elite:



This is almost double the amount of the T500's.

Also, the CSW shares the exact same base unit which means the exact same motors. The only difference i read about was an altered belt for the CSW (perhaps a double v-belt) for more grip i think. But that belt should not make a bit of difference to the ffb deadzone issue.

 

if you load a controller in settings it becomes controller.ini in the player folder.
controller.ini is always your 'current' one,even if you saved your controller as 123.ini

Are the CSW owners loading the 'clubsport wheel' controller 1st?
Does it even matter which controller is loaded?
Are all the FFb settings the same,just the pedals,gears etc different?
__________________________________________________________
That's me DrR1pper is talking about above.
I got to 0.05000 for the STM and I may know what the CSW owners are talking about.
I have had chance to fine tune and properly test as yet.

 

I believe my G25 setup deserves some fiddling... LOL

Some positive forces (righ turn) are giving neg FFB and some left turning result into pos FFB

But everybody can note the FFB deadzone boundary (-.05 to .05)

 

Hey Mysta,

So you found 5% STM works for you? Get any oscillation still? Is the ffb deadzone removed when in motion?

 

How did you measure the actual FFB torque output at your wheel?

 

see post #123

and I dont know if its FFB numbers are torque, force, displacement or anything

Edited:
See post #146 for the reference units of the variables FFB, SteeringArmForce and Un/FilteredPosition

 

How did you get this data?



Where did the input variables come from?

 

somewhere round about 0.05000 stops oscillation,was going to fine tune it.
I'll do a back to back load of testing over the weekend as the STM may have introduced a G25/27 esq center deadzone?
can't quite remember as it was a while ago but will put in the hrs over the weekend.

 

I don't understand. What do you mean by "25/27 esq center deadzone"?

Also, your 5% STM should be roughly the same for CSW users as well.

edit: sorry, i got you on the "25/27 esq center deadzone" now.

newtonpg, perhaps that deadzone your seeing on the graph is a bit of a gap/play between the gears in the g25 (maybe). To test if it goes away, just ramp up the STM and go out and repeat your test again and if it doesn't go away then it's most likely that i think.

 

DrR1pper, Im using TechAde's "rF2 Telemetry CSV Plugin v2"

You can find it @
http://isiforums.net/f/showthread.php/18286-rF2-Telemetry-CSV-Plugin-v2

Im affraid you are right about G25 gear/deadzone. ATM all I can say is that the low forces (at steering rack) are present.
I'll make some runs over the weekend to confirm

Some additional info about the posted graph:
Data are raw values, not a single outlier was deleted
Dataset name: Panoz LRPwC
Dataset source: Unknown
Number of Points: 39925
Number of Columns: 2

Statistic SF FFB
-----------------------------------------------------
Minimum -39.166 -1
Maximum 42.677 1
Range 81.843 2
Average 1.2114 0.1036
Std. Deviation 5.18 0.43

Equation: a + b*x
a = 0.00678
b = 0.07998
Standard Error : 0.1143
Correlation Coefficient : 0.9640
Run time : 0.0060 seconds