Guys, come to talk about tires

Discussion in 'General Discussion' started by fufsgfen, Oct 26, 2018.

  1. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    5,877
    I find this interesting, I feel that there is something I might learn or maybe something we all can learn.

    This is my attempt with base 1.5DX Covet and Beater Covet:


    Next I test Hirochi as that was mentioned too.
     
    • Like Like x 1
  2. rottenfitzy

    rottenfitzy
    Expand Collapse

    Joined:
    Mar 7, 2015
    Messages:
    529
    Not that different, IMO.
     
  3. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    5,877
    I did fail at Hirochi, because of Covet over steered:


    Earlier video I think I did manage to drive corner better with beater, DX probably would go faster, but I'm not that great driver anymore, slow and blind :D
     
  4. B25Mitch

    B25Mitch
    Expand Collapse
    Vehicle Artist
    Staff Member

    Joined:
    Aug 27, 2013
    Messages:
    257
    Here's a video of the SBR4 beating the Pike's Peak production car record by 11 seconds. Totally factory config with street tires, too. Put slicks on and it'll do at least 30 seconds faster.

     
    • Like Like x 15
  5. ¿Carbohydration?

    ¿Carbohydration?
    Expand Collapse
    Banned

    Joined:
    Mar 19, 2017
    Messages:
    1,710
    Tyre is a District Capital in the South Governate of Lebanon.

    It was a Port and Mercantile city, created in 2750 BC and had an important role in the Phoenician Empire.
    It was last recorded to have a population of 117000 in 2003.
     
    • Informative Informative x 2
    • Like Like x 1
    • Agree Agree x 1
  6. corvette81

    corvette81
    Expand Collapse

    Joined:
    Oct 17, 2012
    Messages:
    270
    But as a dev pointed out before "non official map" also who knows if the surface type/grip is correct, map is too scale etc, etc.

    I still strongly believe the tire model needs ground up rework.
     
    • Agree Agree x 1
    • Informative Informative x 1
  7. Ai'Torror

    Ai'Torror
    Expand Collapse

    Joined:
    Aug 29, 2015
    Messages:
    1,653
    Are those new engine sounds that I spotted? :D

    What is wrong with them other than not so realistic behavior at really low speeds (due to the fact of ideal circles not being possible in virtual world)?
     
    • Like Like x 2
    • Agree Agree x 2
  8. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    5,877
    That was beautiful drive to watch, I can't remember those corners at all.

    @corvette81 That Pikes Peak is probably closest to real track you can have in any game, it is very accurately made LIDAR track, but of course I don't know about grip levels, someone else might know about those.

    Here is Covet Zxi at Hirochi with 1 attempt of fast lap:


    Of course understeer is present as it is FWD and I hammer the throttle, but I find it pretty much in line for what to except, nothing excessive.

    With gamepad I used to have issue of turning too much and that is what makes whole lot of understeer, it is quite strong effect, I wonder if that could play part of this understeer thing, there certainly are reasons why people are experiencing that and feeling it being a problem.
     
    • Agree Agree x 1
  9. B25Mitch

    B25Mitch
    Expand Collapse
    Vehicle Artist
    Staff Member

    Joined:
    Aug 27, 2013
    Messages:
    257
    The map was built off LIDAR data, so it's obviously to scale. And the surface type is the exact same default asphalt groundmodel as the rest of the maps use. You could have easily confirmed these two facts yourself.
     
    • Like Like x 4
  10. estama

    estama
    Expand Collapse
    Developer
    Staff Member

    Joined:
    Aug 7, 2012
    Messages:
    233
    When the argumentation devolves down to "beliefs", then it can become infantilistic like. My belief is bigger than yours and all that stuff.

    This is why i prefer scientific arguments. Scientific results (data) can speak for themselves, so they don't need anyone to believe anything.
     
    • Agree Agree x 24
    • Like Like x 2
  11. Capkirk

    Capkirk
    Expand Collapse

    Joined:
    Nov 19, 2017
    Messages:
    397
    I agree with this. BeamNG tire models aren't perfect, and there are probably better ones out there, but they're pretty good. If I do a handling test in real life, and compare it to beamng, I am willing to bet the results will be very similar. Of course, I can only test so much, I'm not destroying my tires to prove a point.
     
    • Agree Agree x 3
  12. corvette81

    corvette81
    Expand Collapse

    Joined:
    Oct 17, 2012
    Messages:
    270
    So the science is based on mis skewed tests is absolute but anyone with any knowledge of how complex a tire is and how it reacts with the world is wrong? kek

    I am not saying I am right, hence why I said get race drivers in and do a trial, there is a lot that could be improved upon, like the lack of mechanical grip on throttle causing loss of traction and the understeer that makes no sense. A tire model is not just a patch like this |_| its a very complex thing, tires need to lean on themself, use side wall, have a large surface patch to grip to the surface, etc. The basic box node layout on the wheels isn't good enough for a complex softbody sim.

    You don't have to listen to me, nothing/nobody is saying you do but think about it this way, thats hundreds of thousands of views on Jimmy Broadbent's channel, all mostly under the impression there is something worse with your physics so it might pay to investigate improvements.

    This is my last post anyway, I would seriously look into the claim though.

    Cheers,
    Corvette
     
    • Agree Agree x 2
    • Informative Informative x 1
  13. Cheekqo

    Cheekqo
    Expand Collapse
    Banned

    Joined:
    Feb 6, 2016
    Messages:
    1,533
    I think the reason why my driving is all sloppy is because I have to fight the G27 wheel because the G27 has this feature that prevents you from spinning out or cocking up, and it isn't working properly so it tries to go it's own direction.
    May I have advice on this? Then I'll drive the LXi around the Hirochi Raceway.
    EDIT: Here's the thread explaining the issue:
    https://www.beamng.com/threads/g27-wheel-tries-to-go-its-own-way.58723/#post-949318
     
    #13 Cheekqo, Oct 26, 2018
    Last edited: Oct 26, 2018
    • Informative Informative x 1
  14. stenyak

    stenyak
    Expand Collapse
    Staff Member

    Joined:
    Aug 6, 2012
    Messages:
    1,119
    Using 'race' drivers is not necessarily a good way to benchmark physics realism, at least when you're past certain realism threshold. Specially not if they are not used to simulators in the first place, or if they expect to compare fictional vehicles with other different real life vehicles.

    However the basics of physics can be checked easily using factual data, numbers. As opposed to opinions or fuzzy descriptions like "it would undesteer more over here in real life".

    Just record a video of an example that you think is wrong. Put some UI apps showing G-forces, the tire loads, tachometer, pedals and steering inputs. Provide track/car/mods information so we can replicate on our side too. Only then we can start talking about whether or not the behaviour is utterly incorrect, very off or just slightly off (like you would expect from any realtime proper simulation).
     
    • Agree Agree x 8
    • Informative Informative x 1
  15. ManfredE3

    ManfredE3
    Expand Collapse

    Joined:
    Jan 9, 2016
    Messages:
    1,007
    You guys are talking about all this science-y stuff and I’m just here like “when will we get square blocks of wood to replace our tires with”

    I don’t actually care for this as official content, just wanted to make this comment. Last time I made a comment like this (the Fiat 500 monster truck) I think Falkrum took me slightly too seriously
     
    • Like Like x 1
    • Agree Agree x 1
  16. MrAngry

    MrAngry
    Expand Collapse

    Joined:
    Sep 5, 2013
    Messages:
    634
    Then again, if the tire simulation is claimed to be really good, how come BeamNG isn't popular among the simracing community? Aside from the obvious lack of multiplayer, of course. I'm going to try the game out again with FFB and maybe record a video.
     
  17. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    5,877
    That is countersteering effect that most cars have to some extent.

    Someone with similar wheel might share settings with you, but it is important to disable most in Windows control panel and have wheel setup not too strong because if it is too strong, then you feel just one kind of force.

    Also too weak is bad, wheels are bit challenging to setup well.

    For example with real drift cars, you can let out wheel and it countersteers to opposite lock really fast, you kinda learn to catch it at right moment.

    I have FFB strength set to around 140-160 depends a bit how I drive, but I have T150, there is lot of thoughts on my FFB thread which might be helpful for you, wish that people would post their settings and wheels used there for others to find:
    https://www.beamng.com/threads/guide-to-ffb.56886


    I did Green hell with a pickup truck, slowly those corners are coming back to me. D15 Custom classic is bit of a dog as that does get understeer, heavy motor at front makes corner entry understeer happen, but light rear makes it bit nervous at corner exits, almost lost it at one corner:


    Something like 10:23, don't know how good or bad that is for a bone stock pickup truck, could drive much faster with skills, but I lack skills :p

    Testing however is something very interesting to do, after all driving is pretty much fun in this or other sims, so I'm all up for testing and attempting to figure out if there would be something that could help people to get better experience.

    Also there probably is lot that I don't know and testing always teaches more. At least it has already reminded me how much fun Green hell is to drive, had forgotten that during so long driving on that track :)

    Also one thing I have learned is that Hillclimb SBR4 is very unstable at high speed, using normal steering instead of quick steering it is bit more manageable, but would like to adjust aero balance of that, or something, at green hell those super fast small turns at dowhill, first half of the track, rear end grip just is not enough I think, but that is hillclimb config and not track racing config of course.

    I would be all up to auto-x day or two with Covet like vehicle, if we just could take BeamNG with motion sim there and each one of us would have hell of good time, while doing some science, sadly that is not realistically possible, but maybe someone has chance to do some proper testing of such nature :D
     
    #17 fufsgfen, Oct 26, 2018
    Last edited: Oct 26, 2018
    • Like Like x 1
    • Agree Agree x 1
  18. ManfredE3

    ManfredE3
    Expand Collapse

    Joined:
    Jan 9, 2016
    Messages:
    1,007
    Probably because it’s not really possible to do any racing other than hot laps. Just look at how competitive stuff like iRacing is. I can’t imagine anyone getting that competitive over their PB around Hirochi Raceway.

    Also there aren’t that many proper race cars in BeamNG, which is what I see a lot of people racing in the more realistic sims.
     
    • Agree Agree x 5
  19. estama

    estama
    Expand Collapse
    Developer
    Staff Member

    Joined:
    Aug 7, 2012
    Messages:
    233
    In general we do not hide/cover up our problems. We strongly believe that what you drive should be coming from the physics as best as we can do it. Hence we didn't even try to cover up the problems of our FFB using effects and so on, before we made our breakthroughs that fixed these problems.

    Also our simulation approach is to simulate the different phenomenons, starting from first principles of physics. Tires are an extremely complex phenomenon to approach using first principles, as the principles aren't even that well known yet and there are very complex interactions that happen between the different parts of the tire. Due to these, there is validity in the approach that other simulators have, to use huge spreadsheets that try to describe the behavior of the tire in different situations.

    Even so, we avoid that preferring to push with our "from first principle" approach due to various reasons. One of these is that we are aiming our simulation not only towards normal driving/racing situations but towards any situation, no matter how extreme and wacky it is. And by any situation i mean situations that are way way beyond what can be described in these spreadsheets. Towards this, our tires have many contact points, the hub is also being simulated and we also do full pressure simulation inside the tire. Our tires, expand at speed or at altitude, allow waves to develop around them, when they loose traction they can even start "pulsing" until the regain traction and they reach steady state. And all these at 2khz. To give you an indication of what 2khz is, we could create an audible waveform out of that.

    So if the complains around our tires, want from us to do what everybody else is doing and use spreadsheets that cover a narrow field of situations on top of a single (or a few) contact points. Then i suggest that there are lots of options following that approach already existing, so please visit them.

    What our simulation is about, it is to allow experiencing whatever you wish for. You want to race on top of road bumps? Going up and down sidewalks, rockcrawling? Racing while putting 10 other vehicles on top of your own while towing 5 more? All these while you change the gravity and the ground models? We want you to be able to do all of these and more. This is why we are doing everything in our power to have every little detail be as physically accurate as we can make it. And this is why we don't use spreadsheets to describe our tires [*]. Because spreadsheets cannot even describe a tiny bit of what the fantasy and ingenuity of people can create and do.

    [*] By the way, I don't want to diminish the approach of other games, i have immense respect for their efforts as i understand how hard tires are.
     
    #19 estama, Oct 26, 2018
    Last edited: Oct 26, 2018
    • Like Like x 25
    • Agree Agree x 1
  20. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    5,877
    I think there is one key parameter that could explain some of these understeering complains.

    Steering lock.

    In most racing simulators, steering lock is half or less of what it is in BeamNG, to limit steering lock to same as your favorite racing sim, you need to set limit steering by speed in filters page so that it does not allow all of steering lock.

    There is huge difference of say 22 degrees or 46 degrees.
     
    • Agree Agree x 4
    • Like Like x 1
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice