rF2 FFB system and philosophy

Discussion in 'General Discussion' started by DrivingFast, Nov 4, 2020.

Tags:
  1. UGM 133A

    UGM 133A Registered

    Joined:
    Mar 24, 2019
    Messages:
    106
    Likes Received:
    56
    Why is this even an argument? This is a basic stuff you learn when you go to almost any tactical driving or race driving schools.

    The reason why the steering lightens after a certain slip angle in either case has to do with both tire grip being lost and load acting on the suspension. It's the same reason why the steering will lighten slightly if you lock up the tires under braking. Slip angles can be in both the under or over steer direction as it refers to the individual tires not the vehicle as a whole, so it doesn't matter to the laws of physics which way you're screwing up, it'll give you the same result.

    https://www.steeringwheelguy.com/racing-steering-wheel-theory.html

    From the link above:
    [​IMG]
     
  2. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,926
    Likes Received:
    3,876
    Yeah the tire forces and steering forces going lighter when understeering or oversteering is quite fundamental stuff. It is a bit strange to see people claiming otherwise. It is a bit different than you explain, particularly for oversteer.
    • In case of oversteer
    Rear end looses more cornering capability than front, thus looses more friction, or traction, or grip. Front end has more grip. The lightening of steering comes simply from the fact that steering (or front wheels angle to the car, if to be more direct) has to straighten out before it winds up to opposite direction because of rear end overturning. This will obviously depend on grip balance per front vs rear during the process of overturning. The harder the oversteer the faster steering should pass through "lightening" phase, and the more neutral overturning aka neutralsteer the longer lasting lightening phase of steering should be.

    Things like caster, geometry, tires, steering lock to lock, steering wheel lock to lock has very much influence on steering specifics.

    And it is simply this curve for that, nothing really about it:
    [​IMG]

    • In case of understeer
    There are several reasons for steering getting lighter. First is that as tire contact patches sliding area is increasing tire reduces its force just because sliding friction is lower. Second thing is that because same thing pneumatic trail gets reduced, which will be the more noticeable the lower the mechanical trail is (which caster has to do a lot with). And extra thing is that self aligning forces peaks before cornering forces peaks, so if a car has a steering that you can feel getting lighter during understeer, you should do extra little bit more steer to get most of the tire.

    Also, however, the big hook up of FFB we should be talking about is not the fact what fundamental effects work on a car steering, but how they work...
     
  3. UGM 133A

    UGM 133A Registered

    Joined:
    Mar 24, 2019
    Messages:
    106
    Likes Received:
    56
    Yes, and all of this is mainly due to the fact that nearly all modern cars having positive caster in their steering geometry, which gives them self aligning torque, and causes an effect known as the pneumatic trail. That's what you're feeling when the tire goes through those phases you're talking about, which all goes back to the slip angles of the front tires.
     
  4. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,926
    Likes Received:
    3,876
    I think all cars ever had positive caster. At least at the moment I don't know a car with negative caster, maybe just my limited knowledge. Perhaps there could be a car with negative or neutral caster angle and still have proper mechanical trail for proper stability, I don't know. Pneumatic trail is independent from geometry of the car, it is all coming from mechanics of elastic deformable contact patch, majorly caused by resultant force of cornering forces plus reactions caused by twisting tire. Caster is just additional leverage to pneumatic trail that is completely mechanical, thus it gives "mechanical" trail which is hopefully positive. And yes, of course everything that goes on with moving car is related to slip angles and slip ratios, otherwise the car would be stationary or in the air.

    By the way, there are also gyroscopic forces that has some effect for resisting direction change, I don't know how strong and how significant those forces are.
     
  5. UGM 133A

    UGM 133A Registered

    Joined:
    Mar 24, 2019
    Messages:
    106
    Likes Received:
    56
    Some old cars used negative caster. I don't know if it was designed into the cars to make the steering lighter, or if they simply didn't understand the mechanics involved, but it was done. Yes, you're right that the tires have pneumatic trail independently. I think with a negative caster the pneumatic trail would get swamped by the caster trail, which is why it's so unstable in a straight line with negative caster.

    I think also the gyroscopic forces would be swamped by the caster. Maybe if the vehicle was extremely light it might be noticeable, but I think in real world scenarios for most cars it's negligible.

    The point remains that, what you're feeling is a combination of the self alignment torque caused by the caster and pneumatic trail.
     
  6. pferreirag60

    pferreirag60 Registered

    Joined:
    Jan 22, 2015
    Messages:
    11
    Likes Received:
    17
    I have an 4x4 car, and i can tell you, every time i lost the front, the wheel felt much lighter (wet conditions)
     
  7. UGM 133A

    UGM 133A Registered

    Joined:
    Mar 24, 2019
    Messages:
    106
    Likes Received:
    56
    Most hydraulic power steering used in road cars are non-linear, it could be that you don't feel a change in resistance due to that and not anything to do with what the tires are actually doing. I think it's only been relatively recently that some sports cars have started to have linear steering with the use of electrically assisted steering.

    Apparently when power steering was originally introduced, people all over the world said it made it more difficult to feel what the car was doing. More recently I've heard essentially the opposite claim from people who've only ever driven cars with hydraulic power assist after they try a car with a more linear electric power assist. I've also heard many people say that cars without power steering have absolutely no steering feel, which is laughable. In all of these cases they're people who aren't experienced with driving a car at it's limits and who therefore probably don't fully understand the nuances of steering feedback to begin with, they're just used to feeling a particular thing and anything else feels "wrong" to them.
     
  8. Alex72

    Alex72 Registered

    Joined:
    Dec 27, 2016
    Messages:
    1,317
    Likes Received:
    992
    They just try to find something to attack in order to undermine it. Its sinister. Like dirty SJW weirdos who try to attack that which will undermine the whole. Physics and FFB like everyone knows is superb in rF2 compared so some dummies will try to attack that by coming up with lies. Its dirty and i dont know why anyone would do it, but there you go. Some people are just dummies who suck.
     
  9. David O'Reilly

    David O'Reilly Registered

    Joined:
    Jun 12, 2012
    Messages:
    1,050
    Likes Received:
    756
    I was so impressed with @Lazza and his evolution draft that i was inspired to go into more specific detail.
    I'm not drinking coffee at the moment so instead of primary research I used a Wikipedia entry on the evolution of mammals and used "search and replace" feature and some basic editing.
    Here is what we got......

    The evolution of Force Feedback has passed through many stages since the first appearance of their bungee cord ancestors in the Racing Gaming sub-period of the late Arcade period. By the mid-Sega-Rally period, there were many wheel species that looked like Force Feedback. The lineage leading to today's Force Feedback split up in the Jurassic; Wheels from this period include Madcatz, more closely related to game controllers and keyboards than to mid period Force Feedback, as well as some early Thrustmaster products, more closely related to controllers.[1] Later on, the Bungee wheel and Non Force Feedback lineages separated; the Non Force Feedbacks are the animals more closely related to the marsupials, while the Bungee wheels are those more closely related to the placentals. Since Juramaia, the earliest known Bungee wheel, lived 50 years ago in the Arcade period, this divergence must have occurred in the same period.


    After the Playsation 1 extinction event wiped out the coin in slot arcades (sit-in games being the the only surviving examples) and several controller groups, controller and wheel Force Feedback diversified into many new forms and ecological niches throughout the GT Legends and Need For Speed period, by the end of which all modern orders had appeared.


    Force Feedback are the only living Wheels. The direct drive lineage became distinct from the gear and belt drive lineage in the late force feedback period, between 6 and 4 years ago.[3] The game controllers are today's reptiles and birds along with all the extinct animals more closely related to them than to Force Feedback.[3] This does not include the rumble effect wheels, a group more closely related to the Force Feedback.


    Throughout the PC2 period, the Wheels included the dominant peripherals and several important controllers. In the subsequent Triassic period, however, a previously obscure group of Wheels, the self-build Direct Drive, became the dominant Wheel. The external control box appeared during this period; their superior sense of pneumatic trail, backed up by a large brain, facilitated entry into oversteer niches with less exposure to total spins. The nocturnal sim racing lifestyle may have contributed greatly to the development of force feedback traits such as steering rack forces and high frequency feedback. Later in the Rfactor2, after load cell brakes replaced simple springs as the dominant high end domestic rigs, Force Feedback spread into other ecological niches. For example, some became high end arcades, some were E-Sports, and some even were adopted by professional racing teams.


    Most of the evidence consists of Youtube videos. For many years, fossils of non Force Feedback and their immediate ancestors were very rare and fragmentary; but, since the mid-1990s, there have been many important new suppliers, especially in China. The relatively new techniques of un-canned physics force feedback have also shed light on some aspects of FFB evolution by estimating the timing of important divergence points for modern species. When used carefully, these techniques often, but not always, agree with the Ebay used selling price hierarchy method.[citation needed]


    Although steering rack forces are a signature feature of modern Force Feedback, little is known about the evolution of this as these software records are not often preserved in the fossil record.
     
    Filip and Korva7 like this.
  10. avenger82

    avenger82 Registered

    Joined:
    Dec 29, 2016
    Messages:
    811
    Likes Received:
    342
    I prefer sliders rather than json files as it's in rF2. I think in rF2 some FFB effects like road feel (bumps) are exaggerated (which is good for most consumer wheels). I've read that someone said rF2 was much more bumpy than IRL with default settings. Also there are a some issues with particular cars like with ridiculous BMW M2 CS braking and infamous Mercedes AMG GT3 FFB.

    But overall I always got the best feel in rF2, but I guess ACC FFB is more pure. I bet no one here knows which one is more realistic. For sure both are not really that accurate due to some DirectInput limitations and cars' physics itself not being that accurate.
     
    Last edited: Nov 13, 2020
  11. Comante

    Comante Registered

    Joined:
    Nov 20, 2013
    Messages:
    1,694
    Likes Received:
    1,218
    Actually, you should not mess with Json anymore. At the present stage, unless you particular wheel has some specific needs , messing with Json is not recommended.
    Rf2 does not exaggerate anything, is just your multiplier that is exaggerated probably. Today you just need to tune minimum torque and FFB multiplier. Stop. You can more probably disrupt FFB.
     
  12. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,926
    Likes Received:
    3,876
    One word: physical versus empirical.
     
    Rui Santos and mesfigas like this.
  13. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,346
    Likes Received:
    6,572
    @davidporeilly that's lazy journalism mate :p

    I bolded the troublesome thing you said. That's where most of these arguments start.
     
  14. avenger82

    avenger82 Registered

    Joined:
    Dec 29, 2016
    Messages:
    811
    Likes Received:
    342
    this magic word doesn’t say much about being close to reality. IRacing had physical approach from the beginning which supposed to be revolution. I think pCars2 and 3 also has physical tire model.
    Empirical can be more accurate than physical and often is. It’s more about input data which most often isn’t there.
    Multimilion dollar simulators like in F1 used empirical approach long after physical TM were used, because it’s easier to work with real data. Perhaps some still do.
    Besides FFB is doesn’t only come from tires, but also suspension has a huge role.
     
  15. avenger82

    avenger82 Registered

    Joined:
    Dec 29, 2016
    Messages:
    811
    Likes Received:
    342
    Agreed
     
  16. David O'Reilly

    David O'Reilly Registered

    Joined:
    Jun 12, 2012
    Messages:
    1,050
    Likes Received:
    756
    I can give some perspective on Assetto Corsa vs RF2 but not ACC.
    I run an MX5 Supercup car in the UK in real life which I can use for a real life comparison.
    I bought Assetto Corsa on sale for curiosity. For me the feel of chassis movement just wasn't there. Also when encountering understeer the wheel response didn't provide FFB cues. I would have to just look where I was on the road. I would have to stare at the delta to see if the setup was faster, I couldn't feel the grip change. There was FFB yes but the subtle changes that telegraph what's happening were missing in my view.
    So for me it's a hands down win to RF2.

    To compare it to real life; in my car there isn't any more ffb than in RF2 and probably less. EG As semi slick tyres come in its felt more through the seat and chassis than the wheel. Kerbs are gentler irl than in RF2!
    Lock ups are telegraphed as much through tyre smoke (front) or sliding (rear) as via the wheel. A rear lock up in the wet is noticed by the revs dying as the rears stop turning.
    In my view these other stimulI irl add to the awareness but the actual wheel tells me less than in-game.
     
    pkelly, Korva7, avenger82 and 2 others like this.
  17. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,926
    Likes Received:
    3,876
    @avenger82 It does definitely say much, but you are right it is not everything. I can not imagine empirical simulation having more potential in simulating fine details of vehicle handling. Details that gives most of the driving immersion, emotion being exact or almost exact to real thing. Empirical stuff has this advantage of having better relation with data and being less complex, but simulation is simplified too then. Without stepping into more complex territory and biting some of the chaos it is impossible to get better. Also probably there should be ways how to reverse engineer empirical data to physical parameters. It must depend on how well made the empirical simulation is, and how badly made is physical simulation. Obviously.

    @davidporeilly That is very interesting. I suppose you compare to existing rF2 MX5 mod ? I wonder how much of the things that mismatch with your RL experience could be improved by modifying only parameters of the car physics parameters and its tires parameters, perhaps even configuring a whole different tire. P.S. interesting about revs dropping with rears locking up, it makes perfect sense, but I can not recall having such thing in simulation, perhaps I have never locked up rears that much.
     
  18. David O'Reilly

    David O'Reilly Registered

    Joined:
    Jun 12, 2012
    Messages:
    1,050
    Likes Received:
    756
    Hi, No the Spec miata mod doesn't convey the handling of an MX5 Supercup car very well. Its a Mark 2 and the Supercup car is a mark 3. The real car has more grip and better brakes, the RF2 Spec Miata mod has more top speed. So lap times are similar but feel is different. It's possible that modding the various parameters might bring it closer. A lot of other cars give a closer feel to the MX5 Supercup car than that one.
    I am making a general observation of how a real car feels versus in game FFB.
     
  19. Pawel44

    Pawel44 Registered

    Joined:
    Oct 1, 2020
    Messages:
    63
    Likes Received:
    43
    Such fanboys like you make me sad. There are serious flaws in rF2 physics and they're present since years. Even one of the biggest rF2 supporters (fanboys is more appropriate in this case ;)) is talking about them:

    "Cranked ABS & very low TC (since these systems aren't simulated very realistically). Rear wing as low as possible (so car is sliding around every corner), balanced by nearly detaching the rear sway bar. Tyre pressures at minimum, and camber as low as you can have it. Minimum ride height, since bottoming out isn't really a problem."

     
    avenger82 likes this.
  20. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,926
    Likes Received:
    3,876
    "issues will be addressed" hahaha

    Some of the stuff what happens might not be issue with rF2 core physics, but parametrization (I love this word) of cars/tires.

    And rF2 with those issues still kicks ass. I wish developers adressed that stuff and sent it in full attack. But it is how it is...

    By the way, I would never fully trust what youtubers say about complicated matters such as physics. They have great success rate at misinterpreting stuff, lacking critical thinking, lacking simple fundamental knowledge. They are not nailing those subjects as much as they would like to. And it goes in both ways - when they say sweet and sour things. But they has great impact, with one video greater influence than thousand posts I would write everywhere, and I am just random guy too, perhaps bit less random. Kool to have some occasional content by guys that are known of experience with physics such as N.Heusinkveld, it is just unfortunate that he has no interest in non-empirical way that rF2 has went. I know Aris does some talk and drive stuff to share some light. I would be super interested in developers being more active in forming opinions and explaining fundamentals. Actual engineers that at least some of them must be into simracing, it is impossible that none are interested in simracing. Pro racers... and gntelman racers like David above. But they aren't doing that much. I guess it is not important. Or it is too complicated and too easy to be incorrect. And the more they know, the less they know. This is nice quote by Einstein: “As our circle of knowledge expands, so does the circumference of darkness surrounding it.”

    Everybody keeps on mentioning Dunning-Kruger effect, but I haven't seen the curve shared :D
    [​IMG]
     
    Last edited: Nov 14, 2020
    mesfigas and Rui Santos like this.

Share This Page