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. Sherwin92

    Sherwin92 Registered

    Joined:
    Oct 31, 2013
    Messages:
    110
    Likes Received:
    17
    with the new JSON format, were looking for these lines right?

    "Steering torque minimum":0,
    "Steering torque minimum#":"Minimum torque to apply in either direction to overcome steering wheel's 'FFB deadzone' caused by friction",

    and to get the "old" "0.03000" for example, we now have to make the ""Steering torque minimum":0," line "0.04", correct? I haven't properly played rF2 yet since the last build, so I may be very, very wrong :eek:
     
  2. Rik

    Rik Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,174
    Likes Received:
    9
    lol, sorry
    before, with controller.ini was Steering torque zero-speed mult="0.30000" now is 0.3. the decimal places have the same sense?
    so for "steering torque minimum" value represented to 5 decimal places for it to work. So that's "0.03000" for 3% deadzone. it is still correct?
     
    Last edited by a moderator: Jul 9, 2014
  3. jimcarrel

    jimcarrel Registered

    Joined:
    Jan 15, 2012
    Messages:
    1,720
    Likes Received:
    24
    That is not my experience. with JSON you no longer have to carry out the multiple decimal places. Look through the the file and you will see it is not needed. Mine is set to 0.03 and it does work. You can experiment with absurd numbers with that format and you can see that it works.

    This is the actual line I have settled on.

    "Steering torque minimum":0.025,
     
  4. jjcook

    jjcook Registered

    Joined:
    Apr 12, 2012
    Messages:
    345
    Likes Received:
    69
    i just put it my old ini and it used that ;)....if you saved it :cool:
     
  5. mel1c

    mel1c Registered

    Joined:
    Jan 20, 2012
    Messages:
    112
    Likes Received:
    17
    Cookie, just save it again in-game and you will have the json version
     
  6. jjcook

    jjcook Registered

    Joined:
    Apr 12, 2012
    Messages:
    345
    Likes Received:
    69
    it did that automatically - the game backed up my file and made a new json-i think if you 'update' it does this as well- i did a clean install so had to put various things back the way i had them -
     
  7. buddhatree

    buddhatree Registered

    Joined:
    Dec 22, 2011
    Messages:
    1,700
    Likes Received:
    249
    First of all, great thread!

    2nd, is the pedal overlay working with the latest build (770)? I ask because my FFB bar never gets red. I have FFB set to 107% in Logitech profiller and even on full 1.0 FFB in-game, the bar never gets red?

    Another thing of note, by default my "Steering torque minimum" is set to 0 and I have zero oscillation. If I change it even to 0.01 my wheel starts oscillating. Is that odd?

    I'm having a hard time getting the balance of FFB + Filtering. I really need to get a belt-drive wheel again.
     
  8. jimcarrel

    jimcarrel Registered

    Joined:
    Jan 15, 2012
    Messages:
    1,720
    Likes Received:
    24
    The oscillation might be because you have 107% IN Logitech profiler. (yea, I used to run it at 106%, but I think it was learned tradition from Grand Prix legends day)

    (You may be correct that overlay might not ever turn red in this new build)

    After running the wheelcheck program mentioned in this thread, I am running Logitech profiler at 98%. with steering minimum at 0.250, I like it very much.

    I run the pedal overlay plugin but haven't been watching it a whole lot. I haven't noticed it turning red.

    **edit. The Mak-Corp cars are the few that I have to turn down ffb multiplier down to .20, everything else run at 1.0 with a few at .80
     
  9. buddhatree

    buddhatree Registered

    Joined:
    Dec 22, 2011
    Messages:
    1,700
    Likes Received:
    249
    Thanks. I think we have the same wheel so I'll try your settings. I must have missed the wheel check program, could you link to it? Thanks.
     
  10. jimcarrel

    jimcarrel Registered

    Joined:
    Jan 15, 2012
    Messages:
    1,720
    Likes Received:
    24
    It's in post #40 in this thread, download provided by DrRipper. (sorry, I keep forgetting how to post link on this forum)
     
  11. buddhatree

    buddhatree Registered

    Joined:
    Dec 22, 2011
    Messages:
    1,700
    Likes Received:
    249
    Thanks! I ran the test, but I have no idea what I'm looking for lol :) I've read post #40 over and over, but I don't get it? What is the sweet spot I'm looking for?
     
  12. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,294
    Likes Received:
    36
    I don't have time right now but if no one else explains it to you by the time I'm back, I'll post instructions. ;)

    And thanks Buddhatree. :)
     
  13. jimcarrel

    jimcarrel Registered

    Joined:
    Jan 15, 2012
    Messages:
    1,720
    Likes Received:
    24
    what I went by (and it was totally my made up theory) I ran wheelcheck.ini multple times (while each time reducing the ffb output in logi profiler) and when I found a setting in the logi profiler that gave me the least number of 0's before giving a number in the ouput graph. I ended up with 2 beginning zeros with a profiler setting of 98%. As a result of settling on 98% in profiler, I adjusted the "Steering torque minimum":0.04, for starters, then tested car on track, with it setting in garage. If wheel oscillated I set "Steering torque minimum":0.04, down to "Steering torque minimum":0.035,. I kept adjusting until oscillation was gone or just barely noticable..

    Just to keep things straight, my theory was STM of .01 for each 0 in the wheelcheck output graph.

    Now as to whether that is correct theory I don't know, but It sure made noticing wheel lock and loss of grip absolutely recognizable. Thats what I wanted out of the exercise.

    *Note.. when you get the Logitech Profiler FFB setting too low, you start getting bad inconsistent wheelcheck readouts at the high end of the readout.
     
  14. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,294
    Likes Received:
    36
    That's a really good idea Jim! A way of reducing/removing the initial ffb deadzone without using (or without using as much) rf2 steering torque minimum. This would also translate to other sims as well. :D Nice one!

    I might add though that you should use the exported excel file for the Logitech overall ffb strength value you've come to use to check how severe (if any) increasing it beyond 100% has caused non linearity in the ffb response (equivalent to increasing the steering torquelinearity in rf2). And if it's too high, you could drop the profiler down a little and use a little "steering torque minimum" in rf2 as well (I compensate) and that would be the perfect balance by using both.

    OMFG Jim, amazing!!! I just realised that's EXACTY what I've done with my t500 as well. Default for the t500 is 60% (which is the equivalent default of 100% for the overall ffb effects strength in the Logitech profiler) but I use 70% instead (like you using a little more than 100%, maybe 106%?) and as a result I use less steering torque minimum as well (only 3% instead of 4-5%) and I've subjectively observed the ffb felt much better. I didn't have a solid reason and understanding as to why this was so but you've just helped me realise now. Thank you! :D
     
    Last edited by a moderator: Jul 10, 2014
  15. buddhatree

    buddhatree Registered

    Joined:
    Dec 22, 2011
    Messages:
    1,700
    Likes Received:
    249
    Thanks for all the replies guys!

    So after hours of testing I've come to these settings: 106% in profiler, 0 STM. At 106%, any STM above zero causes oscillation.

    I tried Jim's setting of 98% in profiler, but found I could go all the way to .045 STM before oscillation. Driving like that felt very filtered?

    I think there's a point of diminishing return for STM. High values do not feel good at all.

    So 106% in profiler means I don't need any STM manipulation at all :)
     
  16. Spinelli

    Spinelli Banned

    Joined:
    Jan 28, 2012
    Messages:
    5,290
    Likes Received:
    32
    I've tried just about every FFB setting combination in this thread. It seems all we are doing is changing the strength-ratios of the different FFB effects relative to each-other rather than being able to change what from the physics engine actually gets through to the FFB, how to manipulate all those different FFB signals depending on what your car is doing/the vehicle physics, etc. I've therefore come to the conclusion that until we can tune the FFB with much more freedom and detail (like in rFactor 1 and prior ISI based sims) that I will be, unfortunately, "forced" to play the inferior physics rF1 based sims such as Game Stock Car Extreme (probably the best rF1 based sim) and so on.

    A month or so back I was driving the Howston G4 and G6 around Mid-Ohio probably worse than a rookie. Literally sometimes understeering onto the grass at the exact same corner for 3 or 4 laps in a row, missing apexs left, right, and centre, locking up brakes as if I was trying to play with a gamepad. Yet I can drive anything and everything in GT Legends at top 5 GTL Rank laptimes, I hardly lock brakes, hardly go off track, I basically just enjoy the drive and can feel all the individual 4-tyre-grip feelings through my wheel, and during all times as well (cold tyres, warm tyres, under braking, turn-in, mid corner, exit, feeling down-shift blips that were too small or too large, everything that's required to push the car to it's limits and keep it there).
     
    Last edited by a moderator: Jul 10, 2014
  17. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,294
    Likes Received:
    36
    I can't say that i have any of the problems that you describe. Maybe frOm having different wheels, maybe not.

    This thread is all about optimising the efficiency of our ffb wheels so we can a) put to use all of the ffb's toque output range without initial deadzone or upper deadzone (both forms of ffb clipping and b) to setup the ffb to be as linear as possible whilst allowing you to still induce a little non linearity should you wish to make, for example, the low end forces generally a little stronger).

    I think what you're after is the ability to manually manipulate the different contributing forces that make up the net force on the steering wheel. If that's so, I wouldn't hold your breath then because the type of physics model that allows for that are predominontly more of a look up table which is a simplified model compared to the solutions implemented in rf2 that is much more procedural meaning you don't have variables that you could manipulate in rf1 engined based sims because they probably don't exist in rf2. The fact you could manipulate them to your liking via coefficients and constants in some ini file was purely coincidencial due to the nature of the sims physics back then vs how they are now (both due to the significant increase in affordable computing power and the time, cost, effort and advancements taken to create more procedural dependent physics models...think rf2's procedural tyre model which I suspect is one of the most significant components responsible for the far more realistic handling characteristics and feeling through ffb that I feel vs all others I've tried, though I've not tried game stock cars so I can speak about it vs your own experiences.)
     
    Last edited by a moderator: Jul 10, 2014
  18. Spinelli

    Spinelli Banned

    Joined:
    Jan 28, 2012
    Messages:
    5,290
    Likes Received:
    32
    Go in a straight line and press the brake. Do you feel different "things" in the ffb telling you how close you are to lock up, lock up itself, how close you are to un-locking slightly before un-locking? Turn into a corner, other than the wheel loading up, do you feel the wheel doing "things" that tell you "hey you'r fronts are at about 2% slip", "now they are at about 10%", "now your heavily understeering", "now your understeering slight less at maybe 20% slip", "now you're at around 15% slip", "now your at 12%", now you're at only 6% slip and you actually have some grip left in reserve and you can turn a little harder to get another percent or two of slip". No, the FFB in rF2 is bland. It's raw as hell and requires lots of strength, it's amazingly alive in other ways, but driving the car and holding it in fine limits is a guessing game from a FFB feel point of view, it's all about visual feel and simply getting used to the car and what it will do based on your inputs, rather than literally having the FFB transmit the full physics of all 4 tyres to you.

    I was racing in the F3 classic at monza and Monaco. From initial corner entry where I had lots of grip left, to trying to turn to sheply and therefore gettin bad understeer, to slowing down and waiting for the understeer to go away, during all those different phases of grip/slip stages the FFB wasn't changing or doing anything different depending on if ai had lots of grip to spare, massively understeering, about to regain full grip, and everything in between, nothing, just the regular loading up/resistance of the wheel and that's it, no actual grip/slip physics info being transmitted to me.

    What a joke. I've had enough experience in real race cars to know that just driving by the steering column feel is a joke and not AT ALL what real racing is like. Anyone who thinks this is the proper way clearly has no idea what real racecar driving involves. ISI really needs to give us the option to customize how vehicle physics are transmitted through our FFB so that some of us who choose to do so can drive with the same type of car feelings and information that we get in real-life rather than driving through just bland steering rack forces because that's "apparently" more realistic.
     
  19. DrR1pper

    DrR1pper Registered

    Joined:
    Apr 29, 2012
    Messages:
    3,294
    Likes Received:
    36
    Ah, well this is where we fundamentally disagree for whatever subjective reasons on calling rf2's ffb bland but i think i know why you say so now.

    You highlighted what you want from a racing sim perfectly in your very last sentence i boldly underlined. You want the physics to tell you what is happening on ALL 4 tyres and i can kind of understand why you want this and associate this with "feeling" more realistic in your opinion and experience. Why? Because in a real car you may only feel the direct forces of the front tyres through the steering wheel but you can also feel the rear tyres through g-forces/seat-of-your-pants sensations which are a direct form of feedback (just like the steering wheel). So naturally a sim that has all 4 tyres directly affecting/contributing to the steering wheel forces rather than just the 2 front tyres will give you the closest sensation you seek based on your real world experience (where g-forces/seat-of-your-pants feel play a very significant if not 100% role in communicating the rear end to you.) rFactor 2 lacks this because it's aim is to be realistic to real-life meaning we only feel the net forces on the front tyres through the steering wheel and not any forces from the rear wheels which is not realistic to what you'd feel in a real car either. For that reason i can understand what your saying now (if i've understood you correctly?).

    If true, it would go some way to explaining perhaps why some cars feel better than others as a result of requiring less rear end sensation dependency for car control.

    Having said all that, i still prefer rfactor 2 to any other sim i have tried (but like i said earlier, i'll need to try game stock car) but for whatever problems having less/absent rear feel in rfactor2 has, i still think rfactor2's ffb and physics more than makes up for it through the front tyre feel alone since the sim feels so alive and realistic to me vs countless other racing sims i've tried. But i do understand what you are saying and in fact i think i may have kind of noticed it when i drove an F1 kart in Shenzhen China the other month where the g-forces are the most significant component of the sensory feedback about the kart vs the steering wheel (but that may also be partly due to the fact that i've not karted that many times as well as my fitness and strength levels at that time).

    Well, if ISI are listening or planning to do so, it'll probably take some time as what must to be done to incorporate such as solution is blend the already perfect front tyre physics with some rear end (somehow) to get the result your after. It's an interesting idea Spinelli.
     
    Last edited by a moderator: Jul 11, 2014
  20. Comante

    Comante Registered

    Joined:
    Nov 20, 2013
    Messages:
    1,694
    Likes Received:
    1,218
    Spinelli, you can't pretend to receive such kind of information from the wheel alone. If you (like me) would like to increase the level of feedback in your simulation, you should look at implementing other devices to your driving postation. There are interesting solution on the market, and today, you could even develop your's without too much trouble, as RF2 output plenty of data that can be promptly converted to signal to all the kind of actuators you would like. For example, I would like to develop a seat than in part resemble this : http://simxperience.com/en-us/products/motionsimulatorkits/simxperiencegs-4g-seat.aspx , but on the bottom, should have four actuators to transmit tire load and/or grip amount or slip angle, whatever you think it's best suited for your needs.
    I think your problem is of expectations, and not about the quality of the simulation, that is top notch. I'm sorry to tell, but if you expect from a 40 dollar game to have a real racing simulation, you are making jokes of yourself, if , on the other hand you invest real money to dress RF2 with high quality hardware, then I bet you will be much much much pleased.
     

Share This Page