1. Trouble with the game?
    Try the troubleshooter!

    Dismiss Notice
  2. Issues with the game?
    Check the Known Issues list before reporting!

    Dismiss Notice

What do you think about driving physics

Discussion in 'General Discussion' started by ilkermurat, Feb 5, 2018.

  1. atv_123

    atv_123
    Expand Collapse

    Joined:
    Aug 5, 2012
    Messages:
    1,711
    Well, naturally, that is what the devs are aiming for all the time... to make the most realistic physics possible.

    I am just curious where you seem to feel that the game is currently lacking, because it may just be in a feature that hasn't even been added yet.
     
  2. vinkkelikasi

    vinkkelikasi
    Expand Collapse

    Joined:
    Sep 4, 2017
    Messages:
    10
    The physics in BeamNG are stunning but I just started playing Gran Turismo Sport and I must say GT Sport feels even better. Yes BeamNG is more diverse, I know, and there are a lot of things more realistic. I love the game. Anyway comparing the two I think BeamNG needs better force feedback and tire physics, (and maybe a bit more work on suspension geometries). Drifting in BeamNG is near impossible whereas in GT Sport it feels natural and realistic. And the overall handling just feels good.

    I’ve said it before but flicking the steering wheel, letting the car turn to the desired angle and the steering wheel to opposite lock and then grabbing the wheel to maintain the drift angle can’t be done in BeamNG. Not with a Logitech G25 or G29 at least. I always end up spinning out even though I have some experience drifting in real life. In GT Sport it is very possible and feels natural
     
    #22 vinkkelikasi, Feb 7, 2018
    Last edited: Feb 7, 2018
  3. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    6,774
    I'm thinking poor setup, wheel + FOV + incar camera + car setup, with those set to same way racing sims have them set, BeamNG is really good.

    Of course opinions are opinions, I did a test, very unscientific test, mostly to support my own opinion of course and so on.

    Drive into turn, then apply throttle so that rear end starts sliding, after that hit brakes, from real life experience I can tell that rFactor 2 has it quite nicely covered, so does BeamNG.Drive, but Assetto Corsa, no, it is nothing like real life, it is not even simulator class, you see this kind of stuff in Need for Speed (try to ignore opposite lock message in rF2 had recording start in F2 and that makes it activate help function in rF2, it just countersteers or something):


    But I must say that I do like Assetto Corsa, even it has certain misgivings, it gives wonderful immersive experience and I don't care that it is illusion and fake, driving just is so nice in that, but I would not call that a simulator, more like emulator or something. Even basic things in driving physics seem to have such flaws, so I'm happy with BeamNG real physics. Also rF2 seems to have improved a lot during a year or then I'm becoming senile, because it is not that bad.
    --- Post updated ---
    New video includes also LFS:
     
  4. ilkerrbr

    ilkerrbr
    Expand Collapse

    Joined:
    Dec 23, 2016
    Messages:
    205
    [QUOTE = "SebastianJDM, posta: 795358, üye: 223435"] I'm guessing you've never driven a car fast/ QUOTE]
    i know kung fu:p:p:p

    i love beamng but not good enough now:cool:
     
  5. Slammington

    Slammington
    Expand Collapse

    Joined:
    Aug 14, 2014
    Messages:
    1,460
    This test is totally useless. You're comparing a full size boaty sedan, with a dodge viper, and then a corvette, with massive variances in your line, and general weight transfer with the cars.
     
  6. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    6,774
    Do test yourself and you see that in AC car magically straightens itself when you apply brake after initiating power slide with throttle during cornering.

    FR with big engine was my only criteria, cars are GM V8 sport, C6, C7, XR_GTT

    Don't know which car would mach better to vette in BeamNG when need is to have FR with big engine, but pick any car in BeamNG and you will see they follow laws of nature, where as AC cars don't.
     
  7. SebastianJDM

    SebastianJDM
    Expand Collapse

    Joined:
    Apr 9, 2017
    Messages:
    856
    GT Sport's graphics, sounds, and other aspects of immersion are all far superior to BeamNG, but I think their physics are very close in my opinion. However, I think drifting in GT Sport is horrendously unrealistic compared to drifting in Beam, which feels (and looks) rather natural.
     
  8. Slammington

    Slammington
    Expand Collapse

    Joined:
    Aug 14, 2014
    Messages:
    1,460
    Is it possible that you're using traction control or other assists?
     
  9. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    6,774
    In BeamNG there might be ABS (has GM V8 sport that as a standard?), also I think LFS had ABS, but others had none to my knowledge. Around 30% initial brake application, kinda hard to know with the gamepad.

    Thing is that when car's tail starts to come around, car starts sliding, worst thing you can do is to touch brakes, however in AC it straightens up car and you can keep driving like nothing would of happened, just losing some speed, when in reality touching brakes in such situation spins you out.

    I found out this in AC with ABS and TC activated too, some cars do it less than others, but it is not very convincing, especially with addition of car not rolling to hill, coasting down stops all of sudden as car decides all of sudden it has to be in stopped state etc. Lot of telltale signs there is something awfully weird going on.

    I have tested it now enough that I can tell AC is very different from other sims in situation where rear end starts to come around and you apply bit of brakes, which should send you sideways really good.

    Update: It seems Semi slicks in AC with C7 do that much worse than street tires for example, so there are differences. Maybe that better tire model is not quite so good then? At least with semi slicks there seem to be some quirks.
     
    #29 fufsgfen, Feb 7, 2018
    Last edited: Feb 7, 2018
  10. atv_123

    atv_123
    Expand Collapse

    Joined:
    Aug 5, 2012
    Messages:
    1,711
    Hmm... I dunno... there is a good point that these cars a vastly different... honestly with Assetto Corsa it would be a closer match between the Bolide and the Countach probably... perhaps before we start chucking random stuff at the wall to see what sticks in our favor we actually try to come up with cars that are very close across all games we test. Same for tires and if possible, try to setup the suspension the same. The C7 is an incredibly well balanced car, and while I agree that hitting the brakes while in a slide is the worst thing you could do (unless you want to tighten your line while sliding, then by all means) I also think that the C7 might actually be able to hold its own in a case like this. We need older, less thought out cars... I will take a look at the car list tonight and see what I can come up with I guess unless someone here can come up with one first.
     
  11. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    6,774
    I don't like to include mods, reason being that for example there is Road Atlanta for BeamNG, AC and rF2, but they are bit different, AC version is converted from rF2 version, but it is different.

    Then there is GT3 cars, BMW Z4 GT3 is available as a mod to rF2, but it is default content in AC, they look bit same and that is about it.

    Where Corvette C6 and C7 are official content for rF2 and AC respectively and are somewhat close to each other actually when driven without doing stupid tests, both have semi slicks and street tires available if my memory serves me right.

    Zandvoort is track that is available for AC and rF2 as official content I believe, but they are different, at least visually a lot different, but maybe enough close.

    Now of course BeamNG does not have Zandvoort, so I guess Road Atlanta is best track choice for such testing.

    K-Series is closest to Vette, I guess, with supercharger and race tires it might be something of such, but it is still quite different, engine is quite small.

    Now each game has their faults, every one of them, even BeamNG of course and I do like different cars and different tracks in each game, but even I do not think AC as being quite as full blown sim as other two, I do like it a lot, those little fake things are so well though out that they do add a lot to immersion and it is pretty, I admit that.

    BeamNG lacks just bit of texture quality, that whitish blue shine on car and all cars should have rear light power like Hopper, desaturation of distance objects could do good too, also there is too much overexposure on whites, but it is not that far really.
    upload_2018-2-7_22-59-16.png upload_2018-2-7_22-59-47.png



    What is bad in BeamNG, I think is setups of cars, they are quite lot on oversteering side and tires don't bite quite hard enough when applying power, but lot of that is to do with a setup.

    Now about that Z4 in AC and rF2, same setup is bit hard to do, because suspensions are not equal way set up, so with same values you end up different values at the wheel, only way is to build setup for car from almost scratch. In rF2 Z4 is not very good, steering is very aggressive, I'm having issues even with 9.4 degrees steering lock with gamepad, you might get better experience with a wheel, but still it is very oversteering compared to AC Z4, I think that Z4 in rF2 could be improved quite a bit with remaking it, but I doubt we have time for that.

    rF2 has new tire model update recently, that is not applied to mods or many ingame vehicles yet I believe, so that might also have something to do with why I so much like from rF2 C6 and hate so much rF2 Z4.

    I do like that Ferrari in AC though, also Z4 GT3 is nice in AC.

    In BeamNG, K-series would need GT3 spec setup to be tested against Z4, Nomi wing etc. SBR4 is rear engine/mid engine, so is Bolide, but AC has 288GTO of course which would be match for Bolide, however it would need DLC or then mod that might not be again quite as good.


    I hope everyone did read when I wrote video being unscientifically tested and mentioned it just supports my opinion, don't take it too seriously.

    Lot has to do with brake balance and very sensitive brake control, but still those C7 semi slicks as well as some other car's tires which I forgot do cause AC to behave rather strange way, but it is enough good really, just not simulating quite the way BeamNG simulates.
     
  12. Goosah

    Goosah
    Expand Collapse
    Global Moderator
    BeamNG Team

    Joined:
    Aug 30, 2013
    Messages:
    790
    I think the problem with that test is you are clearly locking the front tires in Assetto Corsa, which should cause the car to straighten. If you try a BeamNG or LFS car with no abs and make sure the brake bias is forward enough, it will do the same. Maybe you could make the argument that in Assetto Corsa the car tested has too much front brake bias, but that is a setup issue not fundamental physics.
     
  13. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    6,774
    You are very much correct, that is pretty much what it is, there is also something related to tires, maybe cold semi slick later grip is not very strong as street tires compared to those seem to do less of that.

    What I find interesting though is that F40 does not exhibit this behavior, despite having quite forward bias brakes and tires are named same as C7 tires, who knows how different those are. Nissan GT-R is almost impossible to make straighten up with a brake, despite very forward brake bias, it has hypercar tires, so there is difference in tires.

    I can make it happen with LFS quite easily and also with BeamNG, but with BeamNG there is more roational inertia around Z-axis I believe when body starts to rotate it does not quite so easily want to change direction despite tires lock up.

    AC is not bad at all, but I think BeamNG is ahead in many ways, most trouble with hotlapping I get with rear end grip, but also there is this thing that BeamNG is more dynamic, bump is not just bump for tire, it sets a force that affects whole car, which have more interesting effect to grip, imo. Somehow AC seems precalculated, but it does really good job not to make it apparent when trying to get best possible time.

    Hillclimb SBR4 without wings is just something very hard to tame, but with some changes to setups and lot of practice it is manageable :D

    We all of course wish that it would be possible to get sounds and graphics of AC to BeamNG, I think even developers would want that, but sadly it is not quite possible.
    ------------


    ETK K-series setup attached which I did work a little, it is not even nearly done yet, but if atv_123 or anyone else wants some base for setup building there is one with race parts and supercharger. Not sure actually if supercharger comes with vanilla content or if it is a mod, this memory is so good sometimes :(

    Update: I think that best setup in BeamNG is in 200BX track version, well at least for me and my gamepad that is, with little brake adjustment it is thrilling car to dance around the race track, it is still quite loose, but considering the power it has, it is incredible how you can balance it with throttle and brake, keeping it at that small slide trough the corner.

    Quite bit is probably because of those racing tires in 533 Clockwise wheels, when I put those to ETK K-series it starts to behave a lot better.
     

    Attached Files:

    #33 fufsgfen, Feb 7, 2018
    Last edited: Feb 8, 2018
  14. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    6,774
    It took something like 75% F and 25% R brake balance (semi race pads with race brakes) to get BeamNG car to straighten itself, even then need to hit brake just right. If I would put street tires to front and race tires at the back I would get very similar to AC straightening effect.


    I could not get default setup K-series to straighten itself, maybe someone more skilled could, but I couldn't.

    Then again, if having warm tires in AC that straightening is not going to happen even nearly that easy, with cold semi slicks effect is really strong compared to street tires for example, I'm not sure how slippery cold semi slicks are then, but that is big contributing factor.
     
  15. stenyak

    stenyak
    Expand Collapse
    Programmer & Global Moderator
    BeamNG Team

    Joined:
    Aug 6, 2012
    Messages:
    2,021
    There are many different factors in play, but ultimately it boils down to what's the natural balance of the car and what you are doing to change it.

    For example, braking with cold tires means you cannot load the front axle as much/quick, which means there's going to be less grip bias in the front compared to hot runs. Suspension setup also affects the tilting/rolling behaviour. Tire compounds, tire shape&size, actual amount of grip due to varying contact patch characteristics due to camber/toe effects during suspension compression, how good the ABS or your foot is, aero balance, distribution of mass, amount of engine braking and what axle it ends up in, how smooth you are with the steering...

    If you were testing in completely level roads, same road conditions (grip) on smooth, unpatched tarmac, with the same vehicle in all sims (a well made real life vehicle reproduction, or a ported vehicle with ported physics), using a steering wheel for better accuracy and making sure your inputs are exactly the same in both sims, then you might have a chance of comparing different sims.

    Anything else would be too random, and only usable for very rough/general comparisons, for very broken or fake behaviours (such as "the car is unable to roll over" or "a magic force stops the car with twice the grip when i run wide in corner entry", etc). With AC, rFactor, iRacing, LFS, BeamNG.drive and similar realism-oriented sims, most of the differences can be easily explained by differences in car design, track features, tiny input differences, and even if those were all the same, setup differences :)
     
  16. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    6,774
    This is exactly true!

    For example if we change one parameter in original comparison video, that is if we put warm tires under the car, situations changes completely and that is really a point here, all claims about tire models etc. are really funny, but setup is very much what is different between sims.

    Car's are different so making same setup is really challenging, but it is possible to make car behave similar way or really weird way too.

    Controller setup is also one big aspect, in racing sims usually low steering lock is used and in BeamNG, when I want lower steering lock, I use steering limit by speed feature, with 900 degree wheel difference is perhaps smaller, but with gamepad that helps a lot to get similar feel like in racing sim as whole motion range of controller can be used when driving around the track.
     
  17. CommandoAir

    CommandoAir
    Expand Collapse

    Joined:
    Jan 24, 2014
    Messages:
    117
    I think in terms of racing physics, AC has BeamNG beat every time. When it comes to pretty much anything else, BeamNG is miles ahead.

    I'm not sure what you mean by AC having "fake" physics, but when racing (which is what AC is designed for), it certainly feels more realistic. I feel like the cars have more weight to them, and the grip levels are very well balanced. When racing around the Hirochi track on BeamNG, some cars feel far too understeery, as if you would have to slow to crawling speed to take the corner, yet other cars (like the race covet) feel FAR too grippy. That thing zipps around the course at light speed. Even if it is a race version, I shouldn't be able to go flat out around corners in a FWD car.

    Somebody also mentioned something about break temperature simulation, and how AC will give you the same thing every time if you drive the same way, which certainly isn't true. AC also has break temperature simulation, and it also has tyre wear, temperature and flat spot simulation, which I'm pretty sure BeamNG doesn't have any of.

    What it boils down to IMO, is intent. AC is a racing simulator. Simulating racing conditions is what it does best, and IMO, better than BeamNG. But BeamNG is pretty much the ultimate vehicle physics sandbox. It might not be as good at racing conditions as AC, but it's better in every single other aspect. It's a jack of all trades, master of none (except maybe destruction physics). In BeamNG, crashing and wrecking your car is part of the attraction, but in a game like AC, that would mean game over. You generally don't want one simple mistake to end your race. It's generally better to take a time penalty and perhaps some minor damage while losing positions, but stay in the race, than it is to go flying into the wall, writing off your car and ending your race all together.
     
  18. Brother_Dave

    Brother_Dave
    Expand Collapse

    Joined:
    Aug 16, 2012
    Messages:
    1,740
    Id say the understeer feel you talk about in BeamNG is about AC has a stronger sense of speed, which makes you pay better attention to it. Its so easy in BeamNG to go flat out and brake to late, steer to strong.

    What i guess he means with that AC has 'fake' physics is that AC is coded to replicate how racing physics feel, while BeamNG replicates physics itself, if that makes sense. In that way i guess you could say AC has 'fake' physics. I do think AC has great physics, for its intent. Its way easier to go full race mode in AC then in Beamng, and that is down to, as you say, its intent. IMO BeamNG does feel as good when racing but the game itself isnt set up to only do races, menus, camera views, gfx, gfx effects, sounds, tracks so that the fell of racing is duller. BeamNG has a much larger field of work, which needs to do compromises (such as menus, race leagues and what not) so that its doesnt focus too hard on one area.
     
  19. fufsgfen

    fufsgfen
    Expand Collapse

    Joined:
    Jan 10, 2017
    Messages:
    6,774
    Of course it would be better to have single crash to end the race, that would keep racing clean and it would be a lot more fun when you could not use your 'elbows' to make way. Bumping other cars off the track is very viable tactic if there is no realistic damage.

    Some day people will understand that tires are not everything, accurate physics simulation is required to get real racing physics, but that day is not today.

    Good luck overheating these 'godmode' brakes:
    upload_2018-2-12_13-25-21.png

    Sure, some cars do have brake temps, but you really can't overheat them, there is no tactical element like real world racing where you push hard few laps and then have to take it bit easier and especially with street cars where you just can't push your brakes at maximum lap after lap.

    Subjective is of course always subjective, I have played Kunos products since Netkar 0.99 Namie and they have their own way of creating this illusion of driving something on your computer screen, one thing he has always known is how to make driving enjoyable.


    Do a good setup in BeamNG and try then, I don't think that setups are quite as optimized for racing as are setups in AC. Now BeamNG does not have tire temperatures to my knowledge, if those are added, surely BeamNG will behave again bit differently.

    I'm not sure actually which are sticky racing slicks in BeamNG, I have been using Clocwise533, but I get only about 1.3G's or so while I would think closer to 2 with sticky slicks should be excepted?

    Anyway I have been trying to make a race setup similar to racing sims and it is still not there, but it's getting closer, rear wing is perhaps bit too much, so it understeer too much in high speed corners.

    Then controller setup, in BeamNG steering lock is too much for racing use, at least with gamepad, so I use this setting to limit steering lock to race sim levels, also I'm using filter here, but it is tad too slow, without filter tad too fast, could use 50% filter of what gamepad filter is:
    upload_2018-2-12_14-52-20.png

    Because my throttle, steering and brake are all in one thumb control, I'm also getting issues of steering when hitting brakes, to compensate that I need this, that is same with AC, I use quite high gamma there:
    upload_2018-2-12_14-55-14.png

    View I have set like this, it is relative camera, I just press 9 from keypad and then page down to set view like this, because racing sims have view like this:
    upload_2018-2-12_14-57-9.png (from some reason replay is very choppy, so I doubt that I upload that video)

    With setup attached and not steering too much with that controller setup (reason for understeering in BeamNG) I get pretty good racing experience, but default settings are not giving best possible experience in BeamNG from my opinion.

    Greatest difference is that steering amount, where in racing sims you get only something like 17 degrees of steering lock, in BeamNG you have close to double of that, which makes it quite easy to steer too much.

    Try limiting that steering amount and give BeamNG racing a go, it is surprisingly good, but also difficult, much difficult because physics are raw real thing and not just something to give you great experience in racing.

    West Coast USA is a better race track as there is grippier racing line and slippery off the line, also somehow I think Hirochi is tiny, I get only to use couple of gears during whole track.

    Road Atlanta is nice as it is also in AC and rF2, albeit last two turns seem to be somehow different, but maybe it is just me.

    Update: Sorry I attached wrong setup file, now there is correct one, it is still bit understeery and can be bit touchy if too agressive on throttle at corner exits, but it is WIP.
     

    Attached Files:

    #39 fufsgfen, Feb 12, 2018
    Last edited: Feb 12, 2018
  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