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

Discussion in 'General Discussion' started by DrR1pper, Mar 26, 2014.

  1. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,286
    Likes Received:
    22
    Nice Marc.

    Remember this though, these results of the initial ffb deadzone are deadzones from a static steering wheel start position which means these are the deadzones for static friction. Static friction in higher than rolling friction, so the actual initial ffb deadzone for rolling friction will be less than for the static friction which is what your wheelcheck graph results are showing. On my t500 it was 8% but when/if you use stm in rf2 you don't go higher than 4-5% because when you add just a little movement to the wheel you take it out of the static friction and into the rolling friction and now the amount of force needed to keep turning the wheel by itself is reduced (because rolling friction is less than static friction).

    Now as you can see from your results, using 110% instead of 100% reduces the initial ffb deadzone 75% and 120% reduces it 100% down to no initial deadzone. But whilst you've found a value to fix the initial ffb deadzone in the profiler alone, you've caused another problem in the process…..ffb non-linearity. Just look at (for 110% overall effects strength curve) from 0% - 22% force sent to the wheel correlates to 0% to 60% force output from the wheel output. The remaining 22%-100% force sent to the wheel resulting in the remaining 60-100% force output at the wheel. It's quite non linear indeed.

    Now even 100% is not ideally linear as well but at least it regains some more linearity than 110%. Now if you can find a value between these two (maybe 106-107%?) that still gives you a very minimal deadzone from the profiler, then you can add less stm from rf2 and combined you can achieve the best of both worlds. This is what i do with my t500 (using 70% overall effect strength where the default is 60%) and add only 3% stm instead of the 4-5% i would have to otherwise.

    Alternatively, if you don't wish to use stm, perhaps you could try using the 110-112% in the profiler and add some negative ffb non linearity in the controller.ini file. The curves for all the overall effects strength on your g27 are characteristically positive, i.e. like this:

    [​IMG]

    So you want to add a layer of opposite non linearity by choosing a less than "1.0" value in the controller.ini file that applies a layer of negative non linearity like this:

    [​IMG]

    and hopefully you can come to some good values that'll equalise each other out to produce a relative more linear ffb response curve.

    I still believe in using the stm though but if you have non linearity issues with your wheel's ffb response curve regardless of the overall effects strength % you use, then perhaps it's worth a look at attacking the problem from the reverse non linearity solution.
     
    Last edited by a moderator: Jul 12, 2014
  2. Marc Coyles

    Marc Coyles Registered

    Joined:
    Mar 21, 2013
    Messages:
    78
    Likes Received:
    7
    Ta for the indepth response - more tweaking to do (in progress)!
     
  3. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,286
    Likes Received:
    22
    No problem. One last thing of note as well though, despite what i say makes sense (at least to me :p) on paper and in theory, it all boils down to what you find feels best for you. In order to know this, it means you need to find the most optimal settings from both approach and then do some rigorous length comparisons flipping from one solution to another until you can make up your mind on which you think feels better and makes you drive better.

    And as Wernher von Braun once said: ""One test result is worth one thousand expert opinions."

    ;)
     
  4. Marc Coyles

    Marc Coyles Registered

    Joined:
    Mar 21, 2013
    Messages:
    78
    Likes Received:
    7
    Yep - I'll be at this for days getting it nailed, I can tell... but it's something to do whilst I work out how to mount my hydraulic brake mod! Can't fire up rF2 to experiment in any major way at the mo tho sadly, so it'll go down on my list of stuff-to-do.

    Finding wheelcheck fascinating just for visually exemplifying the effects of Profiler settings... for example, below is 120% overall strength with Profile Dampener Strength varied from 0-100% - irrelevant, am aware, but good to know what the settings are doing nonetheless!

    More dampening = more linearity, but sacrifices a huge amount of overall strength ultimately.

    View attachment 13557
     
  5. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,286
    Likes Received:
    22
    Interesting observation but yeah, not very good if it does that to your ffb force strength. :S

    The best solution for us all is if we could have an intermediary ffb response curve that we could use manually adjust (in detail) for each of our non linear ffb response curve wheels. It would be the perfect solution because we could then make any and all wheels perfectly linear and remove the initial ffb deadzone from in there (just like the stm function).

    Think of it like how you can manipulate the response curve linearity of the wheel input and pedal inputs. So like this:

    [​IMG]

    And if you where is outputting the negative curve then you apply a positive curve onto it to then get the linear curve/line ffb response (and no initial deadzone) that we all desire. Ofc this graph is just for illustration purposes and you would need to tailor the intermediary curve according to how the wheel outputs forces based on (something like) the wheel check.exe program.

    This is in fact completely doable and within rf2 if the devs thought it would be worth the time implementing. They could even program the equivalency of the wheelcheck.exe program to test the linearity of the wheel and auto generate the intermediary curve to correct the response curve to become absolutely linear. This would be brilliant!
     
    Last edited by a moderator: Jul 12, 2014
  6. Marc Coyles

    Marc Coyles Registered

    Joined:
    Mar 21, 2013
    Messages:
    78
    Likes Received:
    7
    Optimal FFB settings in rFactor 2 guide - The key to being in the "Zone" :D

    Was thinking that as I was playing with it - would indeed be fantastic if they could add (or license) it.

    But yeah, I get all that entirely. Effectively we want knee & threshold adjustment (if we were to akin it to audio compression / expansion).

    Next question re: killing clipping (once STM is sorted - attacking things backwards to your pt1/2 recommendation in the OP)... is it genuinely better to reduce the incar multi, rather than reducing effect max strengths in controller.ini instead, or are we looking at the multi as it's easier to deal with a single control rather than pulling down all the effect maximums? Ie: if we reduce steering max strength (and brake / clutch / gearbox if those are also mapped to wheel axis) does rF2 adjust the sweep of it's range accordingly, or does it just artificially clip it down and we lose some forces from the top end of it's range by doing so? Yes, I can see incar multi still necessary to fine tune individual vehicles... but would attacking the ini maximums not solve it on a more global scale?

    (Again, hypothetical, could test for myself if I could get the rig out n' setup for a few days, but just not doable at the mo... Doesn't stop my brain working on it tho. Feel free to tell me it's already covered elsewhere on the forum if it is!)
     
    Last edited by a moderator: Jul 12, 2014
  7. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,286
    Likes Received:
    22
    Regarding the controller.ini file variables you mentioned, i do not know. Regarding the in game ffb multiplier (providing it's not affected by the controller.ini file variables you mentioned) it is crucial that you set this to the right value (yes, per car unfortunately and each car is different you will need to test, sorry i didn't compile a list….rain check) to avoid ffb clipping (or to minimise it, you can have a little clipping when running over curbs for example but if you were to take the corner without doing so you want the car to producing around 95% of the peak force of you wheel, the last 5% for kerbs/bumps for example but you'll clip when you do so as it will would want a strength that is more like 120-130% over a kerb or bump….but not big deal, rather you sacrifice the detail on the kerb where you should be anyway or at least very minimally and let the rest be used for the actual track).

    I like your testing btw. :p
     
  8. Squeeekmo

    Squeeekmo Registered

    Joined:
    Dec 12, 2012
    Messages:
    113
    Likes Received:
    0
    Thanks for the write up. You've done ISI and the community a great service.
    Now if someone was able to develop a "Dial FFB" plugin so we can change the 'steering torque minimum' values in game, that would be amazing.
     
  9. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,286
    Likes Received:
    22
    ^ yeah that's a good idea, like trim control on a plane.
     
    Last edited by a moderator: Jul 21, 2014
  10. antgen

    antgen Registered

    Joined:
    Mar 15, 2012
    Messages:
    54
    Likes Received:
    0
  11. Paul_Ceglia

    Paul_Ceglia Registered

    Joined:
    Jan 14, 2012
    Messages:
    676
    Likes Received:
    13
  12. c.pucher

    c.pucher Registered

    Joined:
    Nov 4, 2011
    Messages:
    167
    Likes Received:
    0
    Hey :)

    running a G25 here.
    So far I have been running 100% in Logitech Profiler and a quite high STM value 0.08 - 0.1 to get rid of deadzone.
    Now I read that I could use a higher value 106-107% in Logitech Profiler and a lower STM to achieve a similar result.

    Does any of these options has a significant up-/downside?
    Thanks!
     
  13. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,286
    Likes Received:
    22
    well, yes and no. It depends on what wheel you're using as well. But generally….

    We all want to have the initial ffb deadzone reduced/removed. This can be done by either adding stm in rf2 or by increasing the overall effects strength in the wheel profiler (or some combination of both.)

    When adding it via the stm exclusively, you can reduce the deadzone without touching the response curve of your wheel. Having said that, stm alone is not the best method in my opinion anymore. The reason being that stm is only able to remove about half of the true initial ffb deadzone (due to the difference in the rolling and static friction)….going higher causes unwanted rattling/vibrations. Therefore it is advised to use a higher overall effects strength value in the wheel profiler (e.g. 106-107% in the logitech profiler) to significantly reduce the initial deadzone via this way (whilst not causing extreme amounts of non-linearity to the ffb response curve with higher values…e.g. 110+%). Then complimenting the remaining deadzone solution with a smaller amount of stm in rf2 to reduce/remove the initial deadzone further/completely.

    This is what i do with my t500. I must admit though that adding a little non-linearity is not such a bad thing really, especially on our common consumer wheels (even on the t500). A modest amount of added non-linearity has the benefit of also making the low end for forces feel stronger, which is a solution to a common problem for people complaining about the lack of low end force ffb feel. But you don't want to over do it and end up with a relatively flat high end (which effectively acts like ffb clipping then).
     
    Last edited by a moderator: Jul 22, 2014
  14. Eddy

    Eddy Member

    Joined:
    Jun 27, 2013
    Messages:
    1,634
    Likes Received:
    386
    With my g27 i used 106 in logitech profiler and stm 0.04000. Upside of that imho was wheel feeling a bit stronger. No downside discovered. Ingame ffb settings were ~0.8 depending on car of use to avoid ffb clipping.
     
  15. c.pucher

    c.pucher Registered

    Joined:
    Nov 4, 2011
    Messages:
    167
    Likes Received:
    0
    Thanks for your detailed answer DrR1pper, great big hug, and thanks to Eddy as well! :)
    Will try higher wheel profiler value then.
     
  16. Eddy

    Eddy Member

    Joined:
    Jun 27, 2013
    Messages:
    1,634
    Likes Received:
    386
  17. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,286
    Likes Received:
    22
    You're very welcome pucher and perfect Eddy, thanks for sharing your settings and confirming observation. :)
     
  18. coops

    coops Banned

    Joined:
    Feb 1, 2011
    Messages:
    1,680
    Likes Received:
    9
    steering torque minimum = why does is it look like 0.005 there is not 5 numbers like you say should i put an extra number in there ? it is working for me i dropped it to 0.003 it is working well on my csw with 0.98 ffb multi megans and my wheel ffb on 100%. tyvm for this info it has helped a lot. i know its different for each vehicle will just keep a log.
     
    Last edited by a moderator: Jul 23, 2014
  19. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,286
    Likes Received:
    22
    ;)

    Fyi, you may find stm is not compatible with your CSW. Just an observation of some CSW owners who had rattling issues the moment they used any amount of stm.

    If you do you stm, i keep it the same for all cars. Because if you set the car specific ffb multiplier to ffb clip at around the same point for each car, the stm required is almost identical for each car then.
     
  20. c.pucher

    c.pucher Registered

    Joined:
    Nov 4, 2011
    Messages:
    167
    Likes Received:
    0
    Think with the new JSON file format one can spare the trailing 0. So 0.003 should be ok :)
     

Share This Page