RealFeel and Steering Arm Force (n) in Motec

Discussion in 'Automobilista - General Discussion' started by Alan Nash, Apr 2, 2018.

  1. Alan Nash

    Alan Nash New Member

    Joined:
    Jun 19, 2017
    Messages:
    21
    Likes Received:
    14
    Hello all.

    Have been fiddling with Motec and its Steering Arm Force readings in various cars in AMS to determine the ideal MaxForceAtSteeringRack settings in the RealFeelPlugin.ini file for my large Mige Simucube system. Using data logging, I've run all the Formula 1 level cars at Patrick's Nurburgring, because it has some high speed, high downforce corners which induce soft clipping at default RealFeel values. I also did the MCR2000 car. Here are some findings about the peak Steering Arm Force for several cars on the same track:

    MCR2000 . . . 1895
    Formula Classic . . . . 4225
    Formula V12 . . . . . .2262
    Formula Reiza . . . 10656
    Formula Extreme . . . 9850
    Formula Utlimate . . . 7847
    Formula V10 . . . . 3374

    And for good measure

    CART Extreme . . . 4291


    My intention has been to set up each RealFeel MaxForceAtSteeringRack entry just above these maximums from Motec. Finding the wide variations in settings and driving the cars reveals this: the MCR2000 on my OSW is almost violently jolting in ordinary driving, and it seems this is likely due to the much smaller range of force (1,900 units) within a wide range of strength on a DD wheel. Then, the Formula Reiza, set at 11,000 in Realfeel, has periods of strength but much more smoothness and subtlety because it has 11,000 units within that same range of strength on a DD wheel.

    My question is this: are the physics of each car and the resulting Steering Arm Force figures uniquely relevant only to that car, or are they related among different cars? If one were to set a universal MaxForceAtSteeringRack of, say, 11,000 in each of the above cars on a linear DD wheel, and crank the Simucube power up, would the variations in FFB feel among the cars be realistic? This would make the MCR2000 quite light, the Reiza potent, and the rest in between.

    Maybe it is realistic for the MCR2000 to be unpleasant to drive, but with tiny tires, lower speed, and modest downforce, it does seem like it would be proportionally lighter at the steering wheel than an F-Ultimate.

    Of course, power steering or lack thereof would be big factor in ultimate feel . . .

    Sorry for long post---but wanted to share these findings and invite any insight others have.
     
  2. Alan Nash

    Alan Nash New Member

    Joined:
    Jun 19, 2017
    Messages:
    21
    Likes Received:
    14
    In this conversation with myself, I'll post the answer to my curiosity regarding whether a fixed MaxForceAtSteeringRack across multiple cars using a potent DD wheel would be effective: ummm . . . no. Using the MCR2000 as an example, and another car which has a small Steering Arm Force output number, the Formula Retro: putting a high number in the MaxForceAtSteeringRack in RealFeelPlugin.ini leaves the FFB pretty lifeless.

    What I am finding is that taking the Motec Steering Arm Force data, collected at 100hz, then shaving off the most sharp, abrupt peaks using a filter of 5 or so samples (CTRL SHIFT F), leaves a pretty good top number for use in the RealFeel ini file.

    Two other things that have been consistent among all the cars on my DD wheel: 1) turning the Smoothing in the RealFeel ini for a car down to zero actually lessens the harshness of bumps and tendency to oscillate at high speed. It also results in more road detail, as would be expected. I didn't expect turning off smoothing would actually smooth out the wheel. 2) in controller.ini, turning Skip Updates under FFB settings to zero actually has an effect of smoothing out the "Pure 360" FFB. I am convinced it's not placebo.

    When I'm done experimenting with Motec and RealFeel settings, I will post the numbers in case a future fan of this fantastic sim want to try them out.
     
    • Like Like x 1
  3. Domagoj Lovric

    Domagoj Lovric Moderator Staff Member

    Joined:
    Feb 15, 2016
    Messages:
    1,092
    Likes Received:
    1,225
    Monitoring this, can't comment everything else now, but let me just tell you that turning that to zero is EXACTY the same as it is at "1" - let me explain further - game obviously loads everything, minimum range for that value is 1. So anything less then 1 (zero, or even negative numbers), game will just assume 1, and use that in dynamics.
     
    • Like Like x 1
  4. Alan Nash

    Alan Nash New Member

    Joined:
    Jun 19, 2017
    Messages:
    21
    Likes Received:
    14
    Ok, Thank you Domagoj. That settles it, then . . . it was placebo.

    By the way . . . these tweaks I'm attempting are like polishing a diamond. This sim is simply outstanding as is.
     
    • Like Like x 1
  5. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    So F1 2009-2013 has 3.5x as powerful steering forces as the F1 1995? And go-karts are usually less than 1000 so the F1 2009-2013 has over 10x the steering forces as karts?

    hmmmm.......
     
  6. Biucci

    Biucci New Member AMS2 Club Member

    Joined:
    Mar 8, 2016
    Messages:
    11
    Likes Received:
    4
    It's actually the opposite
     
  7. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    Ya. Those game numbers don't seem to make sense unless I'm missing something.
     
  8. Alan Nash

    Alan Nash New Member

    Joined:
    Jun 19, 2017
    Messages:
    21
    Likes Received:
    14
    I'm not schooled in the physics of the rfactor engine, but I think the steering arm forces are produced through the physics of each car---which is unique. It doesn't seem that there is any relative relationship among the cars' ffb output and the steering arm forces. Each car must be set on its own to feel right. Accordingly, the F-Ultimate peak numbers being high and the MCR being very low doesn't mean the F-Ultimate is putting out more FFB. In fact, setting the RealFeel numbers in line with the Motec data results in the Utlimate being rather subdued in comparison to the cars that have low numbers---as the dynamic range is much greater.
     
    • Like Like x 1
  9. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    I thought they weren't relative to each other either but when I asked here, someone from Reiza corrected me and said they are...

    Question:
    "Is a force of 2500 in car A the same as 2500 in car B? Or is there more to it than that since the unit used is not Nm nor is it something that can be converted to Nm therefore maybe something different is being calculated which means 2500 in one car may equal a different output-FFB-strength from 2500 in another car?"
    Automobilista - How the FFB works and what is new about it

    Answer:
    "2500 in car A is same as 2500 in car B."
    Automobilista - How the FFB works and what is new about it

    Weird, right?

    P.S. I was trying to figure out the same thing as you, if setting the same MFASR for all cars means all cars have their forces properly relative to each other just like setting the global Nm in iRacing or rFactor 2.
     
    Last edited: May 20, 2018
  10. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    Going to reply to my own message here. I think I may understand now what Reiza meant...2500 in car A = 2500 in car B = 2500 in car C and so on. OK, makes sense, the same number is the same force-output (assuming MFASR are equal of course)...but, if Car A's physics should be outputting the same force as Car B then why, for example, is car A only outputting a max of 600 while car B outputs 8000?

    Why would cars that are supposed to be outputting the same/similar FFB forces have such vastly different numbers. If car A is supposed to have similar FFB forces from car B, why in the physics engine is car A only reaching 600 while car B reaches 8000? It's like saying a guy with a mass of 10 pounds actually weighs the same as a guy with a mass of 90 pounds because the way 10 pounds is measured for person A means he actually has the same mass as the way 90 pounds is measured for person b...
     
    Last edited: May 20, 2018
  11. Heitor Facuri Cicoti

    Heitor Facuri Cicoti Good Member AMS2 Club Member

    Joined:
    Mar 8, 2016
    Messages:
    858
    Likes Received:
    838
    Because it's max force at STEERING RACK (I'm not shouting, just highlighting), not at steering column. There's still one more force multiplier, which is the pinion radius, until the force reaches you (as torque), and it should be different for each car.

    I don't know if it can explain all the magnitude of the differences in your numbers, but it's something to take a look.
     
    • Informative Informative x 2
  12. Heitor Facuri Cicoti

    Heitor Facuri Cicoti Good Member AMS2 Club Member

    Joined:
    Mar 8, 2016
    Messages:
    858
    Likes Received:
    838
    And the same thought can be applied to the other direction. Forces on the tire aren't directly related to MFASR, because there's another multiplier: the length of the toe link (not the bar, the connection on the upright [or spindle])
     
    • Informative Informative x 1
  13. Domagoj Lovric

    Domagoj Lovric Moderator Staff Member

    Joined:
    Feb 15, 2016
    Messages:
    1,092
    Likes Received:
    1,225
    To get some usable results out of telemetry data, what you should do is ignore temporary high peaks like that readings. Watch for array of steady force numbers in corners, that should give better idea what to set force output too.

    Or in case of DD wheels, i would go this route: set in game FFB gain to 100%, and if you want "full dynamic range" base all cars around one, in this case f-ultimate, use wheel's own settings (not familiar with any, but guessing they can amplify forces quite high). Would not recommend going for very high dynamic range as that could result in some nasty jolts. Logically assuming.

    How realfeel per-car MFASR works is quite simple formula:

    Code:
    int A = controller.ini MFASR
    int B = actual force from car at given moment
    int C = realfeelplugin.ini MFASR for given car
    int D = force returned back by realfeel to game
    int E = force scaling by ffb slider
    int F = final force
    
    D = ( A / C ) * B;
    
    F =  ( D / A ) * E;
    
    So you can see that the only thing realfeel does is rescales force produced by physics such that when B == C --> D == A, this in the end results in 100% calculated force. Slider then can scale that down, or leave it at what it is.
    Hope you guys now have clearer picture.

    Getting accurate comparison data is (dare i say) impossible without dedicated track, on which you should be able to run (or let AI) drive exact route through a desirable perfectly flat corner (no microbumps at all, same camber everywhere (or lack of)).
     
    • Informative Informative x 2
  14. Alan Nash

    Alan Nash New Member

    Joined:
    Jun 19, 2017
    Messages:
    21
    Likes Received:
    14
    I did some more experimentation with Domagoj's thoughtful points about "peaks" in mind. Using Motec and the Formula V10---I used Motec's filter to identify the maximum steering force at Magny Cours. I played with the filter at various numbers of samples, as well as fractions of seconds, and settled on .05 second to filter out the jolts and peaks. I got a max steering force of -3600 with that. I then set the maxsteeringforce in RealFeel.ini to -3500. This lets the game soft clip more jolts and peaks, yet leaves headroom in high speed, high downforce corners to relay bumps and curbs, etc. I think this is the nicest, most informative feel so far. Every bump is still conveyed through the FFB, but the intensity (peak value) of the spikes are clipped.

    Having tried setting RealFeel values at or above the raw, unfiltered peaks, I found the raw settings to obscure the nice feel of high sustained force situations (like high speed corners), because bumps had to turn into extreme, abrupt, rather busy jolts.

    I hope Reiza keeps the RealFeel system in the next title. It feels great out of the box, but to be able to tune based on telemetry is a real benefit compared to any other sim I've tried.
     
    • Like Like x 1
  15. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    [delete]
     
    Last edited: Nov 22, 2019
  16. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    If that's the case then why does lowering in-game FFB have the same/similoar effect (possibly 1:1 inverse) as raising RealFeel's MaxForceAtSteeringRack? If a vehicle is software clipping at 100% in-game FFB (let's say MFSR = 3000 but car is hitting 3500 on long turns) then lowering in-game FFB - and not touching MFSR - removes the software clipping and allows the full range of forces.

    With regards to range of forces, lowering in-game FFB seems to do the same thing as raising MFSR...

    This won't work because as was mentioned (I think by you or another Reiza representative), the steering N that Reiza uses to represent steering force would not be the final number in a real world situation because that number does not represent force on the steering column and is therefore missing multipliers.

    If setting MaxForceAtSteeringRack to the same global number for all vehicles (in an attempt to get 1:1-relative forces between all vehicles), you'll be able to drive the karts with your pinky finger as it's MFSR is only around 400 or so while some vehicles are 6000, 8000, 11000, etc. So the MFSR and steering rack N forces AMS uses for steering force doesn't seem to be an accurate # to use when trying to compare relative steering forces from vehicle to vehicle from the driver's point of view.

    Eg. A kart is not 7.5 times lighter (400 N VS 3000 N) to steer than an F3 car. If I set the same MFSR for the kart as the F3 (let's say 3000 N), the kart's FFB will be almost non-existant even if I set my DD wheel to ridiculously strong forces. Therefore these steering numbers reported by AMS' telemetry/physics and realfeel cannot be used to compare steering strength and FFB from one car to another car...
     
    Last edited: Nov 22, 2019
  17. Marc Collins

    Marc Collins Internal Tester AMS2 Club Member

    Joined:
    Mar 12, 2016
    Messages:
    6,091
    Likes Received:
    2,930
    I made this point many moons ago in some other thread. The proof, which you can find without even testing anything directly in AMS, can be found by comparing the default rates for the F3 301 vs the F3 309 cars. These cars are almost identical relative to differences across car classes. Yet, their values that produce more or less the same steering feel are quite different.

    But, does it matter? Not really. Use whatever combination of settings minimizes clipping and maximizes "feel" given your particular hardware. This is not an exact science due to the wide variations in hardware (design and strength) and even in how vehicles are constructed/simulated in the sim.
     
    • Like Like x 1
  18. Spin

    Spin Active Member

    Joined:
    Apr 15, 2016
    Messages:
    297
    Likes Received:
    33
    I agree Marc, of course it's nice that we can customize this stuff from our wheels and also from AMS itself (overall power, realfeel MFSR, realfeel "smoothing" which by the way I feel is a little underrated and doesn't get the credit it deserves, etc.). But having that relative difference in force from car to car just adds more variation and character without us having to try and makeup our own different relative forces from car to car.

    Not a huge deal, it just would have been cool.

    P.S. You're also right that we already established we can't compare AMS' steering #s from car to car in a relative fashion. I should have looked up the page and read older posts. My bad :oops:


    What I am more curious about though, is how overall FFB affects RealFeel. I know for sure that lowering in-game FFB has the same (or same-ish) affect as raising RealFeel's MFSR therefore I'm very curious when I saw Niels H. pull up telemetry to set his car's MFSR at just around the clipping point (only avoiding some very extreme peaks). The part that confused me was that his in-game FFB was already lowered (50%) therefore he'll have way more range than whatever he set MFSR. Eg. If you're clipping at 100% in-game FFB, then all you have to do is lower the in-game FFB down to somewhere usually around 70-85 % (depending on car, setup, track, etc.) which gives the same effect as raising MFSR (adds more FFB dynamic range). So why was Niels adjusting RealFeel's MFSR to a particular value when his in-game FFB was already reduced to 50%?
     
    Last edited: Nov 23, 2019
  19. Domagoj Lovric

    Domagoj Lovric Moderator Staff Member

    Joined:
    Feb 15, 2016
    Messages:
    1,092
    Likes Received:
    1,225
    @Spin Point above about lowering slider restores clipped range is indeed oversight on my end. Lowering slider will restore what had been clipped if it's on 100%.
    (And before you question it, yes calculation is correct as at it has been compared back to back with realfeel output and calculating this directly in game, without realfeel. FFB out level was equal).




    The number output by game / telemetry is force on steering rack. Note: force on rack, not torque on steering column.
    F309 and 301 produce different force level as these have different suspension geometries.

    When i suggested in other post if one was to go by telemetry to choose something that looks like general "average" value, i never suggested to choose one global value for all cars.
     
  20. Supa

    Supa Active Member AMS2 Club Member

    Joined:
    May 5, 2016
    Messages:
    271
    Likes Received:
    43
    I always run my DD at %100 strength in wheel software and just reduce in-game slider to whatever suits. For minis I might have slider down to 30 or 40 max. Realfeel is default or I might reduce a bit. This gives me best feel. Same with RF2. Wheel at 100 and in-game ffb multiplier turned down to suit.
     
    • Agree Agree x 1
    • Winner Winner x 1

Share This Page