whatchu mean by "iracing-like ffb", like throw basically every ffb detail and every suspension detail to trash and rewrite everything?
Hey there, thank you for the mod, it definitely helped me understand what was feeling wrong with my setup ! Looking at your modifications, it appears that the only change in FFB calculation is the introduction of a 2x multiplier for the g-forces (couldn't find anything different in sensors.lua): forceAtWheel = FFBsmooth:getWindow(max(min(forceAtWheel, limit), -limit), wheelFFBSmoothing, wheelFFBSmoothing2) - GforceVelCoef * (2 * sensors.gx) * M.GforceCoef To my understanding, M.GforceCoef is the parameter we're changing within the settings, so why not just double the slider max value and remove this change altogether ? Otherwise results are really solid, and I think the slider limit should be increased by default in a future update !
Hello! I tried this mod, drove for hours. The first thing I have question about is the numbers. I have a Moza R9 turned up to use all 9nm of torque. Normally I run BeamNG FFB on about 250 most cars most of the time to avoid excessive clipping. With the mod I found that if I turned the side acceleration strength up to 40 then I had to turn the FFB down to 100 or so to avoid absolutely excessive clipping in sharp corners. Even if I had a stronger wheel then I would just be getting an undesirably high wheel force rather than clipping. So how is it even possible to run at 570 ffb strength etc? Do different wheels give different torque for the same nominal ffb strength in the beam menu? The second thing I noticed was that 40 any sense of understeer through the wheel (there's already little enough of that in beamng) was washed out by the side accel strength, aside from the sound of scrubbing front tires of course. I actually ended up experimenting with the side accel back down and the strength more normal, seeing if I can tell any other effects from the mod. Maybe I'm just missing something but that's been my result. What does the mod do besides give increased side accel strength as an option? If a mod could be made to *vastly* drop the steering force after the point of understeer is reached, that would be good, because right now it's subtle at best in BeamNG, outside of the auditory cue. One thing about AC ffb that's great is how you can really feel the increase in turning potential from say subtle trail braking that shifts weight to the front of the car in terms of very noticable increased steering forces. Then when you brake too much you can feel the reduction in steering potential again through a very noticable drop in steering force. I believe that and very detectable vibration right BEFORE brake lock or other tire slip situations is most of what makes it so intuitive.
the most accurate clipping indicator is yourself, you should know how clipping feels like (btw using 250 strength on a 9nm wheel is absurdly overkill lol, i see most people using about 50 strength on those strong wheels)
Some people say owning a 9nm instead of a 15 is *underkill* it takes all kinds I guess. On 50 with the Moza on most default cars the self-correction would be a bit sluggish for recovery from getting sideways (rally/drift) and even less chance of feeling understeer overset. So on most vehicles in beam I use close to as high as I can go without clipping at maximum cornering. Assetto Corsa FFB by comparison speaks clearer so I can turn it down quieter.
that doesnt make sense lol, a t300 is all you would need to do anything and it "only" has 4nm, even a 2nm logitech wheel is pretty fine for most stuff
We're thinking alike alot, ive missed having a better feel of where the tires run out of grip and also what the rear end of the car is doing. What you are talking about is afaik the ffb force window that the g-forces isnt a part of from factory so to speak, theyre just as a multiplier on top of the window which i assume is why it cuts off all detail. Not sure why its done like that in the game but ive had ideas how to fix it, amongst other things. Try the files int the zip below, just replace the ones you have EDIT: Scratch that, that was worse IMO, updated file below!
Okay, looking at the forces ui app confirms that the more you raise side acceleration feedback the less the forces drop off with understeer and the more it becomes a flat line. Beyond 15% side accel there is no manually detectable change when you "over the hill" into understeer land. The first change I'd make if I knew how would be to make the drop off in forces once past peak front tire grip *much* steeper.
Good input! Ive got an idea to scale the side accel forces in a more dynamic way but itll take a few days before i can test it out. Ive already tried to change it but the result it gave was a much less predictable behaviour of the car (still more than vanilla IMO) and i dont want to loose the feel of the car wheight.
Alright, IMO the main thing that needs to be done with the FFB is to somehow make the forward/back weight transfer more detectable. That is the most important piece of information. In AC1 if you accelerate, suddenly lift off, or trail brake, the increase or decrease in weight to the front wheels and thus grip and turning potential is very, very detectable. In BeamNG even with your current FFB mod it is borderline undetectable to completely undetectable, depending on car. This probably means that the calculation for steering force needs to be something other than a simple physics calculation of how much torque it would take to turn the rack in the real world. Again, it is my philosophy that providing FFB that gives clear information about weight transfer and such to the player that they otherwise lack (unless they have a motion rig, ha ha) is more functionally realistic than trying to feel *exactly* like a specific car's steering column driving down the highway.
i was wondering why my wheel was counter-steering so slow, then i found out i accidentally uninstalled the mod lol
Yea, same here, if I cant feel what the vehicle is doing forces-wise i dont care if I can feel a grain of sand under the tire. Ive been trying to cheat a way for getting longitudinal forces but that wasnt with the latest version of my mod so it might work better now, got a few ideas cooking. Done the same thing, huge difference lol
Tested my old idea i had, with my new code change and it led to the expected result this time; ive managed to map longitudinal forces to the force feedback. Also expected this result; increased longitudinal force yanks the wheel one direction, decreased yanks it the other. Now its a matter of using it to scale (or add/remove value from) the correct force output to the wheel (atleast thats how i imagine it being implemented) since it should not scale the side gforces (atleast not too much?) since that would counter what i made this mod for to begin with.
You’re the guy who knows how to do the work, I’m just comparing to AC1 and spitballing, all credit to you bud. Another thing I notice in AC1 is that in slow driving low G force no heavy transfer the steering forces are actually quite light. This gives a lot of overhead for. Information transfer without clipping. You don’t need a lot of force feedback for pulling into a parking lot. There is something of a feel of the steering wheel getting tighter as you go faster and the same amount of steering or whatever does more. This is fine because you simply don’t need a lot of feedback and counter stress etc for moving very slowly and statically, like pulling into a parking place. BeamNG (quite realistically) models steering forces as being quite heavy at slow speed lin a no power steering rack and then they get lighter and lighter as the car gets faster, even with “reduce strength with parked” checked .. As I say this is quite realistic, but it very much more limits the range that can be used for information transfer because there’s a low speed/high G turn that will always be your clipping limit. Whereas as I said AC1 leaves the scale of forces weak enough at very slow speeds that it can be a lot stronger for high speed and hard cornering where you actually need a lot more feel and countersteer strength without clipping. I hope that made some sense.
