Little Sense of FFB Dynamic Range while Applying Steering. Link to Tyre Model / Physics?

Discussion in 'Automobilista - General Discussion' started by Spin, Dec 27, 2019.

  1. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    I've been wondering about this for quite a few years now but haven't really mentioned it until now.

    I've never gotten too much of a sense of FFB range while going into corners with AMS. As soon as I put small amounts of steering input into the corner, the FFB loads up so much compared to the little amount of steering lock I put in. Then, as I'm adding more steering lock, I don't feel the steering resistance increase even more. It's like the steering resistance / FFB peaks at way too low steering/tyre angles. BTW, I'm 100% sure I'm not clipping. I know I'm not because I'm very familiar with clipping, AMS, and sims in general but I also confirmed it with telemetry just to be safe.

    A quick example. I'll be going into the first corner of the Hockenheim "stadium section" (3rd sector) - a fast right-hander - and I'll feel the FFB increase it's resistance a lot (a very noticeable amount) the first small amount of steering input, then, as I'm heading into the corner and adding more steering lock, I don't really feel the FFB increasing more. I get this at all sorts of corners, not just fast corners. Of course on slower corners the peak-FFB is generally expected to be lower on winged cars but it's not the peak FFB that's necessarily the problem, it's how quickly or early into the steering-lock / tyre-angle that peak-FFB arrives. Basically, as I'm adding more steering lock, I'm barely feeling an increase in force once past a relatively small amount of degrees. It feels strange and unlike any real-life roadcar or racecar I've ever driven as well as some other non-ISI physics engined games.


    I hate to mention physics because I know it's a sensitive & often controversial topic but the FFB model in AMS is highly dependent on the physics so it's difficult to mention one without the other. I've often felt like most of the pure tyre's grip is used up at too low steering/tyre angles in the ISI physics engine / tyre model. Now, if steering force is partially related to tyre grip/load, then maybe what I described in the paragraph above (peak, or close to peak, FFB happening at seemingly too small amounts of steering/tyre angle) is evidence of this?

    Basically, the FFB feels like it loads up to (eg.) 80%-100% way too early into the steering/tyre angle application. It feels extremely, extremely non-linear. Is it just a coincidence that I've also felt the exact same way about tyre-grip itself (regardless of FFB)? Not to mention tyre grip and FFB are related. All just a coincidence? I often felt the same with RF2 as well and I remember lots of people before were even recommending lowering rFactor 2's FFB sensitivity/linearity (controller.ini file) down to 0.2-ish from fully linear (1.0) for exactly the reason I'm describing (FFB builds up too much in relation to steering/tyre angle applied). All just a coincidence?


    Current FFB setup (although I've felt this with all wheels in the past but I'll list current specs anyways)
    - vehicle = F301
    - steering ratio = 360/16. Although 360/16 is an unrealistically fast steering-ratio compared to real-life (real F3s are around 14.5/360), it is a slower ratio than AMS' default and slower than the extremely unrealistically fast steering-ratios most people (especially fast AMS/ISI drivers) seem to use (18/360 - 20/360 or the equivalent steering ratio).
    - Wheel = Simucube 2 Pro
    - realfeel.ini = all MaxForceAtSteeringRack values increased by 25%
    - FFB strength: in-game 100%, wheel 40% (10.0 Nm), or the opposite, in-game 50%, wheel: 80% (20.1 Nm). I test both combos. Sometimes I feel like lower in-game % but higher wheel % feels a little less aggressive but I'm not 100% sure so it could be placebo on my part.
    - FFB low force boost (in-game): 0%
    - wheel driver FFB settings (picture below)
    Capture.JPG
     

    Attached Files:

    Last edited: Dec 28, 2019
  2. Heitor Facuri Cicoti

    Heitor Facuri Cicoti Good Member AMS2 Club Member

    Joined:
    Mar 8, 2016
    Messages:
    858
    Likes Received:
    838
    I think this video can be useful (if you haven't seen it already).

     
  3. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    Thanks. If I try extrapolating what I learned in that video into what I'm sensing from the game/FFB, then it seems like it's got something to do with AMS' tyres using up too much grip (relatively speaking) at too low steering/tyre angles. When I drive into corners, I don't really feel the FFB build up. It almost instead feels like there's only 2 FFB strengths: low (when driving straight) and high (once wheel is turned more than, let's say, 30 degrees). Shouldn't there be more of a gradual FFB strength build up? Curiously enough (coincidence or not?), I also sometimes feel this even regardless of FFB but just purely with the tyre grip physics; I'll notice seemingly too high amount of grip drop-off at what seems like quite low steering/tyre angles (like the tyre begins to use up too much of it's grip at too low steering/tyre angles) - maybe not full grip drop off but it's like the grip starts dropping off (and possibly at too fast a rate thereafter) at too low steering/tyre angles The FFB I'm describing could be a result of those physics - or it could just be a coincidence that I feel something similar.

    You know when the throttle or brake sensitivity/linearity is raised to, let's say, 90% (50% being linear like in AMS)? That's what it feels like is happening but with the FFB output sensitivity/linearity for each corner (this is why many people started lowering RF2's FFB sensitivity from 1.0 [linear] to 0.2). It so quickly goes from little FFB (wheel straight line) to suddenly maxing out (or almost) for that given corner.

    I'm going to try and see if I can setup Motec for live telemetry so that I can have it running live while I drive. I'll then check the steering arm forces to see if they seem to match what I'm saying.
     
    Last edited: Dec 28, 2019
  4. Marc Collins

    Marc Collins Internal Tester AMS2 Club Member

    Joined:
    Mar 12, 2016
    Messages:
    6,091
    Likes Received:
    2,930
    AMS has the most subtle and progressive FFB of any sim for me. It is precisely because it has what you are missing that makes it so glorious to drive and that allows for user control of oversteer and spins. Since we are using the same software, the polar opposite experience is down to hardware and/or settings.
     
    • Like Like x 4
  5. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    I don't agree. I've felt this with the T500RS, TX, CSW v2.0, CSL Elite v1.0, OSW, SC1, and SC2.

    I wouldn't say it must be hardware and/or settings that's the issue, that's implying your opinion is automatically correct while mine is automatically incorrect. BTW, I'm not talking about saving slides - how vehicles behave & react during oversteer - as those are physics-based aspects (you know what my general thoughts are about that but that's a totally different story). I'm discussing how FFB loads-up in relation to turning the steering wheel & tyres, front-grip, etc.

    My intention is not to have an argument about how good/bad x game is. I'm trying to have an honest, technical discussion and see if anyone has tips or ideas how to test/analyze this.
     
    Last edited: Jan 6, 2020
  6. Marc Collins

    Marc Collins Internal Tester AMS2 Club Member

    Joined:
    Mar 12, 2016
    Messages:
    6,091
    Likes Received:
    2,930
    I stand by my comment. Compared to rf2, AC, ACC, R3E, iRacing, AMS has more progressive and natural-feeling FFB "load-up" with my hardware and settings. Since the FFB in AMS generally gets rave reviews, I am implying that something on your end is preventing you from feeling what most of the rest of us feel. How the FFB loads-up in relation to turning the steering wheel, etc., is literally the number one selling point of AMS and the one consistently identified area it competes at the top of the rung with any other sim available. AMS is not perfect and has issues related to its aging foundation, but the FFB is not one of them.
     
  7. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    I dunno...I just tried on another track (same car) and same thing, the FFB just loads up to what feels like 80%-100% resistance after such a small amount of lock (let's say 20 or 30 degrees of wheel rotation) and then it's just the same, constant (or almost same) resistance from that point all throughout me turning the wheel.

    I've added 25% to the MaxForceAtSteeringRack for all cars so it's definetely not clipping but it feels like it's clipping in-terms of steering/turning resistance FFB. On high FFB settings, this just makes the car's steering always sooooo heavy from 20 degrees of steering input to 90 degrees, from low speed corners to high speed corners. It even happens the opposite way too, the "return-to-centre" force is so disproportionately forceful - the game applies very heavy and unrealistic FFB power when the rear comes around, it's like there's an invisible weightlifter standing next to me who grabs my wheel when the rear comes out and is applying all his might to turn the wheel opposite lock. The car's steering shouldn't always feel so constantly powerful and stiff. It's like the FFB forces from 0-100% have been compressed and scaled to truly output in the 80%-100% ranges. Not bumps and stuff like that; I feel the lighter forces and fidelity with that, I'm just talking about the steering forces - either steering the car into the corner or the "auto-return-to-centre" force.

    Now that I seem to have AMS back up and running good, I'm going to reinstall the Data Acquisition plugin and have a good look at the telemetry to see if the steering arm #s are matching what I'm saying.

    I reported the same thing 3 years ago with the original OSW (or it may have been an SC1 by then).

    OSW Help Needed with FFB
     
  8. Marc Collins

    Marc Collins Internal Tester AMS2 Club Member

    Joined:
    Mar 12, 2016
    Messages:
    6,091
    Likes Received:
    2,930
    This is exactly what usually feels well-balanced to me in AMS. Though I have driven modded (usually quick hack conversions) cars that feel more like what you are describing. Even within the Reiza vehicles, there are some that are more sublime than others, but even the worst cars still feel more progressive and natural than most cars in other sims. Whatever S397 have done recently, I find most of the joy of rF2 FFB has been lost in favour of the "constantly powerful and stiff" school of design. Do you find that with the latest rF2 vehicles, too? It is always fascinating how elusive good FFB is with all the various programming, simulation and hardware combinations.
     
  9. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    I'll be honest with you, I've barely played RF2 lately so I can't really comment. However I may have a "band-aid" fix for you. Try setting the STS (steering torque sensitivity) from 1.0 (linear) to different values like 0.2, 0.4, 0.6, etc. That setting is just the force output curve. If at 1.0 (linear) it doesn't "feel" linear but rather way too boosted/compressed, then lowering the linearity (aka sensitivity) may help as it adjusts the FFB output curve (just like the pedals' sensitivity setting in the controls section only that's for an input rather than an output but same thing).

    RF2's STS setting is basically like the Low FFB Boost setting in AMS only what I'm suggesting would be like a negative FFB boost (ie. the output curve would be an increasing positive slope rather than a decreasing one).

    Back to AMS, I admit, this is all with the F301. I could be wrong but I think that car is one of the cars with the least modern AMS updates; maybe, hopefully, it's just a specific car thing. What I'll do is try 7 or 8 different cars - newer, more widely used cars - and see if I feel the same (hopefully not).
     
    Last edited: Jan 10, 2020
    • Useful Useful x 1
  10. Marc Collins

    Marc Collins Internal Tester AMS2 Club Member

    Joined:
    Mar 12, 2016
    Messages:
    6,091
    Likes Received:
    2,930
    The F3's are one of the favourite cars in AMS, so don't think there is an issue with them. Trying others will help, though, always.

    My rF2 crapped-out about a week ago and I haven't bothered to spend an hour or two to figure out why? It was working perfectly after the switch to the new beta UI...now suddenly it fails to even start with a .JSON error. Sigh...
     
  11. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    The F309 is one of the favorites, not the F301. They're quite different in AMS. I just always use the F301 cause it's slower, and harder/trickier; I like competing online in slower cars (eg. F301 while everyone else is in the F309). Helps makes things more interesting.
     
  12. Marc Collins

    Marc Collins Internal Tester AMS2 Club Member

    Joined:
    Mar 12, 2016
    Messages:
    6,091
    Likes Received:
    2,930
    Hmmmm, I usually use the 301 as well, but cannot say I found anything too different when using the 309. Just one more gear in exchange for slightly less power. What is noticeably different between them? (I know there are actual suspension differences, etc., in real world, but they are more alike than different compared to every other car in AMS.)
     
  13. Dean Ogurek

    Dean Ogurek "Love the Simulation You're Dreaming In." AMS2 Club Member

    Joined:
    Mar 9, 2016
    Messages:
    1,399
    Likes Received:
    644
    At one point, I had an issue with rF2 FFB where the constant-force resistance seemed very exaggerated. I didn't know what changed at the time but, the solution was to reduce the "Spring" effect (by 70%) in the SimuCube settings and I haven't had any issues since then.

    I share my experience in case it is somehow relevant to this discussion but, I'm sorry I don't have any technical data to explain it. I too, find the FFB in AMS to be at or very near the top of my "best-ever" sim-racing attributes list.
     
  14. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    Oh they're very different, the F301 & F309, Marc. Of course not different like going from a 60s F1 to a 90s F1, or from a Metalmoro prototype to an Aussie V8 but, yes, they are very different (to me at-least). I guess it's all relative (eg. big difference for me may be tiny difference for you).

    Thanks Dean. Going to try that now. I'm pretty sure the wheel's driver''s spring should have no affect in AMS if using FFB low but I'll try it anyways. I'll also zero-out the spring and damper #s in the controller file and see if that does anything.
     
    Last edited: Jan 10, 2020
  15. Marc Collins

    Marc Collins Internal Tester AMS2 Club Member

    Joined:
    Mar 12, 2016
    Messages:
    6,091
    Likes Received:
    2,930
    Well, you must be more sensitive than me :) If I try to equibrilate the down-force and Realfeel (the 309 has a default "lighter" setting for some unknown reason), the cars are as close to identical as you can get. Even with default settings, they feel like the same car just with a different set-up to me.

    Your higher force wheel may be transmitting more subtlety than my Fanatec.
     
  16. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    I did some more tests. I realized I already had wheel-driver DirectInput spring set to 0. I raised it to 150% and played around with the other DirectInput effects (damping, friction, etc.) but they didn't make a difference (as expected since AMS doesn't use these).

    I then changed the following lines in the controller file to all 0.00000:
    FFB steer friction coefficient
    FFB steer friction saturation
    FFB steer damper coefficient
    FFB steer damper saturation

    Doesn't seem to have made a difference either (I also didn't really expect this to do much since I'm pretty sure these are DirectInput effects).

    Next, I'm going to experiment with the controller file's FFB Steer Force Exponent setting. This is what's called FFB boost in-game. I'm going to see if I can adjust the numbers in the controller file to give it the opposite effect.

    Ya, there's definitely a difference in FFB too but I meant more in-terms of pure physics. The car drives quite differently for me. I used to drive with a CSW V2.0 and (and very short period a CSL Elite). It's irrespective of FFB. I remember when I stopped using the F309 for a while and went to the F301 a long while back. Coming out of the pits and into the 2nd corner at Suzuka was an eye-opener. Huge difference, the F301 seemed so soft, so wobbly & imprecise. There may have been updates to the F301 since then as this was a while ago but, man, I was shocked at how sloppy and slidey the F301 drove compared to the F309. I even spun, twice, in a matter of a minute or so (spins are fairly rare for me).
     
    Last edited: Jan 11, 2020
  17. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    This really sucks. Reiza edited the rFactor/ISI FFB Steer Force Exponent setting so that it only works from 0.0 (or 0.5) - 1.0 therefore we can only make the FFB non-linear in a boosted way. I tried 1.5, 1.9, and 2.0 but it doesn't lower the FFB sensitivity like it does in every other iteration of the RF/ISI engine. I don't see why they would purposely block that.

    Lowering the FFB linearity/sensitivity is almost certainly the best "quick-fix" to this FFB problem I'm experiencing. The only other way to lower the FFB sensitivity would be if the wheel's manufacturer had a FFB sensitivity/linearity option in their drivers which the SC2 doesn't (at the moment).

    All we need is for Reiza to unlock the FFB Steer Force Exponent setting so that it works above 1.0 like it always did before. Pleaseeeeeee Reiza. This could be of great benefit to those with high powered wheels (anything about 10-ish Nm)...
     
    Last edited: Jan 11, 2020
  18. Marc Collins

    Marc Collins Internal Tester AMS2 Club Member

    Joined:
    Mar 12, 2016
    Messages:
    6,091
    Likes Received:
    2,930
    Try playing with set-up long enough that they have equivalent down-force (even if it's just by feel). Then make sure the Realfeel "weight" is the same/similar (not the numerical value!). I think it will feel pretty much like the same car that it is in real life (an F3).

    The default set-ups do feel different, but either the 301 or 309 can be adjusted to feel like the other or to be similar (I believe).
     
    • Informative Informative x 1

Share This Page