Is ISI still working on the tire model ?

because irl they trying to catch it before it happens. in sim you do that after it. why? because irl they feel forces announcing incomming car behaviour. in sim you can onlybuse what you see and hear. even to feel a feedback on ff is too late - it is consequence of what car is doing right now.

 

No, they're not. They're using the same base tech but they have modified the code to some extent, it's not the exact same physics engine anymore, just like Shift, GTR2 and GTR Evolution, which are also based on the same tech, but aren't the exact same engine anymore. If they were, you could take the cars from those games, import them to rFactor, not change any numbers and everything would work just as well as it does in the original games. But that doesn't happen. They feel different, because the physics engines aren't identical anymore.

 

100% right. Anyway, growing sound of pre-skidding (not sure about my English words here) can also prevent you before it happen. As you said, when you feel it with ffb is too late

 

graphs : no. i hope this will be possible in future. but you can make 2 or 3 tests and see what happen.

when i saw the TGM i was lost, and i'm still but i think we forget that we understand pretty fine the rf1 model because we work on it during months or years.

it looks like pacejka so "we" learn this model. today there is a new model, so we'll learn new things. i'm not saying this is easy, and i'm not sure this a good way. i'm just saying we have to be carefull about "oohhh it was better ".

moreover, on rf1 tires physics, as others parameters, we see in first the slipcurves, loadsens, peaks etc. ok but there are others important points where it's impossible or very hard to match with reality and who give some huges differences in game.

 

I bet it is the exact same engine. Keep in mind how are their tracks modeled, very accurate with realistic 3d bumps, unlike 95% of rFactor tracks, which are using the wave bumps from the .tdf file.

 

thats not what i am talking about

What I was talking about was

1) the car is already under steering off the track in a linear and progressive way ( its already in a mild slide)

2) the driver wants to gain a small amount of front end grip to ajust the amount they are under steering off whist still being in a moderate slide

as I said you can do this in NKP not RF1 and 2 with default grippy cars cars.

In real life you don't feel it before you just have a 1-1 communication / constant feed of what the car is doing , for the specifc slide and adjustment I am talking about above even given hardware and software latency you are at close enough communication with the pc to feel the give in real time and respond to it , weather visually ore through ffb.

The latency issue I would have though will most likely come into play with FFB devices when the car gets into a tank slapper however and unless you are using a leo bodner wheel or there as some give put into the tire model to account for this then that latency will always throw people off compared to what they would expect from a reel car.

 

Yeah I'm not saying it's worse. Just that it's more difficult to work with it than with the Pacejka model, where you can change some numbers and see the results in a moment. That tire model has some flaws, indeed, so no one is perfect.

Anyway the slip curves are possible to check with telemetry, the next time I drive with rF2 I will try to check how they look.

 

EXACTLY, ive been trying to get this through to people over and over.

I said it 100 times and ill say it again, in sims we probably feel about 75% LESS than we do in real life, AND the little 25% we do feel in sims, well, we feel them WAYYYYYY later than we would in real life.

This is also why 90% of feeling a car and its understeer and oversteer CANNOT (or barley) be seen by viewing videos, bcause most all the feeling of understeer and oversteer towards the driver happens BEFORE A VIEWER CAN VISUALLY SEE IT. This is why I laugh when people use onboards to judge driving, onboards only show when a driver goes over the grip too much to the point where we can visually see it, well ill say it again, 90% OF THE FEELING OF UNDER OR OVERSTEER HAPPENS BEFORE YOU CAN VISUALLY SEE IT. I dont know how many times I have to say this, real life is so much easier to correct and compensate for under or oversteer because you can feel it happening or juuuuust abooouuutttttt to happen before you can visually see it.

Also, in real life you feel everything through your body BEFORE you feel it through your steering wheel and BEFORE you feel/see it visually, ive even heard a driver say the same thing, you feel everything before you even feel it through the wheel and visually.

You are never going to get the same feedback and sense of grip in sims as good and as early (and therefore more easy and more time to react) as in real life. Its impossible, unless there is some compensation made into the physics/vehicle dynamics/tyre model in order to give you this same sensation. However, theoretically speaking I would rather have a sims physics EXACTLY PERFECT like real life whoch would be harder for gamers(due to everything I mentioned above plus a few other things), rather than some small compensation/fudging to the physics in order to maybe "feel" more accurate to us in our hands. This would then be more un-accurate/un-realistic if you analysed the physics and what the car was REALLY doing in game, rather than judging by your hands and brain.

Afterall, arent we all looking for the most perfect physics (of a vehicle in performance driving) that we can get? Dont we all want the best, most realistic, most complex, most well...REAL physics model that we can get? Well the most perfect physics engine would be harder to drive than in real life, again, due to everything I said above (plus a few other things).

 

This is why FFB in consumer simulators dont simply emulate what a real stearing colum does in a car , a good consumer simulator does an abstract virsoin of ffb thats a mixture of what a car wheel would feel like and do , as well as convaying some aspects of the mass of the car.

so in fact evan though there is still a latency you can in a simulator know near enough 1-1 what the car is doing and the level of grip before a tire gives out. ( the other separate point and something you can see from videos and is the main issue with rf2 is that even after the tire has accentually given out and the car has started to slide significantly or a small amount there is still a good deal of control this you can see in videos )

The real world is not magic you don't feel things before they happen yes there is more information and yes you have the advantage of Gs though your bum as well as independent forces and information from the steering column but you are still reacting to things as they come to you.

Other than the basic hardware latency and the rate the software runs at you should be able to know how much grip you have at any given point in time in a driving simulator accentually 1-1. ( even if that is communicated in a different way from real life)

I'm not saying latency is'nt an issue i'm just saying hardware latency is not the core issue at play when it comes to balancing most cars especially slower cars with large wheel bases in simulators.

outside of that I'm fairly confident you could drive a real world car remotely through a video screen with 150ms of latency with no FFB and it would be infinetly more stable and predictable to drive than the grippy cars in rf2. ( wish i had the cash to test this idea out lol )

As another example you can also see that the better RC helicopter simulators have more depth , more progressive movements in how the vehicular changes direction and reacts to wind or sharp movements , even if those movements are very sharp and happening in a very short time frame.

 

In that case using converted tracks or mod tracks that are available for both games should make them feel identical. It doesn't. They're using the basic engine ISI made, but have added their own twists to it. They might be very small changes, but very small changes can have big results.

 

You always take everything I say and change it, I never said you feel something before it happens, i dont even know why you always feel the need to change things around just to start another argument.

Anyways, you feel way more through your body, sitting in the car, being one with the car, etc etc etc you feel way more through that than you do in the actual steering wheel itself or with your eyes, I cant believe you would even argue with that one, out of all our different opinions this one truly blows my mind WOW are you for real right now? yes you take things as they come to you, but you get much more info sent to your "body feeling the car" than you do through the actual steering wheel itself, you feel much more information, much earlier and much more defined and subtle, you can feel the chassis and 4 corners just slightly squirlying around just as they are on the edge of grip, not even enough where you have to make any corrections but you can just feel that your getting there, that is just a quick example though. You feel way more and it gets transferred to you quicker/earlier, than if you magically disconnected your entire body from the car and just went by the feeling through the wheel. Im pretty sure anyone would agree with me on this matter, regardless of which sim, tyre model or whatever they prefer.

Input lag from all our electronics and hardware is a completely different thing, I havent even brought that up yet thats why I said "plus a few other things" in my last post, but sure add that on top of everything I already mentioned and it gets even harder for us sim racers.

 

Where have I done this ?

Me saying its not magic and that it all happens 1-1 is stating that logicaly if you can feel tire flex and other aspects of G forces as it happens in a real car you should be able to communicate that through ffb ( for example you can put a force on the FFB moter that is not enough to move the wheel but the driver can feel it in there hands and so know the gravity state of the car without moving the wheel)

I didn't say you don't feel more through your body in a real car my point was that you can give the player the same idea of grip through abstraction and this is what nearly all consumer FFB does so evan though it is not the same as a real car it should allow the same near 1-1 communication with the car and the grip it has at any point in time .

yes in a real car the steering column communicates front tire grip to some extent aspects of speed due to gyroscopic effects of the wheels but primerally what it does is attempt to straighten up the car , in a real car the driver is not generally using the wheel to gage the overall momentum and feel of the car. ( as you have said and everyone has said and noone has disagreed with in a real car you use g forces and vibrations and how they feel on your body to gage the overall car state)

But the point I was making is that you can abstract the mass of the car and still present it through the FFB of the wheel whilst still maintaining the basic self correcting nature of a real car wheel this is what prity much all consumer sims do at the moment its just a case of which ones do it better or worse and how the developer decided to abstract it.

aside from all that - The issue of car stability is largely separate to FFB , this is why you can play nkp , GSK, RBR with no ffb and they still drive more convincingly than the grippy cars in RF2.

I bet you when ISI finish with the tire model and release more final builds the cars will handle closer to that of the cars in NKP , GSC , ( hopefully better) I also bet that the most critically acclaimed and regarded mods will be ones that drive simular to NKP, GSC and RBR.

 

i agree. it's very very hard to manage ! personnaly i'm not sure to be able to make some nice tires from scratch with this new system. but from what i see during this beta, i think ISI will provide a better base than we had on rf1.

moreover, after projects with some racing teams, i would say there are many many parameters more importants before some details on tires. i'm not saying tires are a detail, not at all, just that the important point is to have a good base of each category.

anyway, the pacejka model is a semi empirical system, even if you have the complete model of the tires, you can't be sure this is accurate.many teams say this is a nightmare to have the datas, and after this, they often see some problems when they try to use them in a driving simulator or "just" to estimate some setups. this is the same thing about the aeromaps of cars in many classes.

from what i saw, everywhere, you can't be sure the data from manufacturers are rights, because ... nobody use them how we try to do. maybe things will change in future, as more and more teams want some simulators.

i have pacejka models of 3 differents openwheels classes. from this, and i don't say this is the truth, but this is what i can see, a open wheel tire is an open wheel tire ^^ sure there are some differences, but you find the same logic in all models.

to finish, rf1 looks like pacejka but it's not pacejka. we easily can believe that a switch is good enough, but it's not

 

Yes, but as simulations, they should take into account the limitations of being a sim racer, and we also have other sims that are quite responsive to car behaviour, for example, RACEON US Muscle Cars at Road America is impressive thru the high speed carousel, very good range of car behaviour and very controllable under throttle, also, GSC 2012's Camaro is a brilliant machine by nearly every criteria and would make any noob rage quit within minutes, but by the same token, an experienced sim racer can get that car under control and become familiar with it's braking points, weight transfer, throttle response etc, a truly magnificent sim car despite an older engine, talk about on steroids!!

Maximum on paper physics suck if that means the driving experience is well beyond the real deal, over and above the existing imitations of sight and sound.

 

I would just liek to point out that one of the most important skills in any debate is to be able to advance your argument energetically but show respect to the other person. You may totally disagree with their view , if so politely and respectfully dismantle their argument.

We really need to attack the concept or argument and not the person. They are two totally different things.
The moment one starts to insult or denigrate the person, wounds are opened that may not heal. You also bring your own character and by association your view into disrepute. Now thats not what you want is it?

 

Pacejka-based tire models have a problem with low speeds, maybe what you describe it's related to that.

http://www.edy.es/dev/2011/12/facts-and-myths-on-the-pacejka-curves/

The lateral force is based on the slip angle. However the slip angle does not account for any variation with the speed. Intuitively, the forces generated at high speed must be greater than the forces at low speed with the same slip angles. But V = (0.2, 0.1) produces the same slip angle (and thus the same lateral force) as V = (20, 10).

 

exactly

 

So if I understand correctly, you are sort of saying you want the physics to be slightly fudged/slightly easier/slightly more time to react, slightly wider/more forgiving slip angle etc etc to compensate for us being behind the screen, our crappy hardware (compared to real life), input lag from our wheels, monitors, electronics etc etc. (By the way when I say all this fudging stuff I mean just a tiny tiny bit to compensate, not fudging in a "dumbing down" way, im pretty sure I am right in assuming you, and none of us here for that matter would want that so please dont be offended when i use the word "fudge" lol). That philosophy is definetely a great route you could go down, and would definetely make sense, it could even possibly make the feel to us in our hands and brain seem more authentic/natural, but I would rather go down the "car is behaving EXACTLY/as close to real life as possible" route, and then just know in my head "of course its going to be a little harder to react, catch slides, anticipate and react to grip changes etc. etc. because we arent in a real car, so we have less feel, we get the feel later, hardware lag etc etc etc". I would rather go down that route, than the SLIGHTLY compensating for all these problems and therefore adjusting (or as others call it "fudging") the physics SLIGHTLY in order to compensate route.

That way, we could strive to make the sim as 100% accurate as possible so that what the car is doing in the game is 10000% as accurate as you can make it relative to exactly how the car the tyres etc etc behave in real life. Then as hardware gets better, faster internal electronics, faster communication from pc to hardware, less input lag, less lag with monitors, quicker more responsive interal motors, higher fidelity rates so the physics and ffb etc etc are all updating more times a second and etc etc etc. as all that stuff gets better and better than we know that the sim is as good as it could be, rather than then the sim may be slightly over-compensating because you got better hardware than the other guy. Not to mention the fact that everyone has different hardware so will have different.

I would rather the sim be as close to perfect as possible and then rely on (or blame) our hardware for the translating the sim to our hands/brain part.

I completely respect what your saying though, because you can look at it from your point of view and say that the VERY SLIGHT "fudging" to compensate, would actually make our driving experience more realistic because the slight physics fudging is compensating for all those issues i mentioned above, so that now when your in real life where obviosuly the physics arent fudged, well the feeling isnt fudged either so the ratio between the 2 would stay the same, therefore the experience and overall feel from sim to reality should theoretically be VERY similiar. So I completely agree with you in that method being very good aswell, but I just like the other method better and decide to choose that in this particular discussion . I would prefer the racing sim developers to "make the game as 100% close to real life as possible", and then rely on (or blame) the hardware for the rest.

By the way, when I say fudging I strictly mean it only to compensate for all the issues mentioned above, that would be the only acceptable way a true racing sim could/should fudge. Any other "fudging" or "dumbing down" would be unacceptable for a hardcore sim (I feel they did this in Ferrari Virtual Academy, you can tell its a "hardcore" sim as its pretty much netkar pro, but it just feels less "hardcore" than nk pro, feels like they slightly made the physics of that car easier and less delicate/complex/dynamic, probably because Ferrari wanted many people to play/buy it, so its still a hardcore sim underneath but they sort of "dumbed down" the physics/driving characteristics of that particular model, even though the physics engine driving the game itself is definetely a "true sim")

 

To some degree yes, but I guess it's my view that the more feel/precision translated thru my controller, the more realistic a simulation it is, ie, regardless of requiring some practice, the better the software/hardware synchronization is, the more realistic the sim is as it gives me the opportunity to duplicate most of what happens in reality.

Regarding your views, I'm not sure how they'd translate in practical terms right here and now other than as ultra difficult over and beyond adjusting to the downsizing and delaying of reality.

Considering what you've said, it reinforces that sims will always get better as both SW/Hardware do, but we should always be mindful of what HW people use, and I think that a G27 probably still dominates the overall sim market.
I guess to some degree, this is a case where like it or not, the lowest common denominator needs to be one of the benchmarks, even though it's obvious that rf2 is built around the higher end wheels.

 

And the more better that hardware-software synch gets, the less compensating will be needed, but there will of course always likely need to be some, even if its super duper tiny. Your view makes total sense, im not saying your wrong, just 2 different opinions thats all, and I honestly think both could work in sims, not saying your way is wrong.

Again, I hope you read the entire post, I explained how I dont mean to offend you or your views when i use the word "compensating" or "fudging", as I dont mean it in a "dumbing down way", but just that tiny tiny bare minimal amount specifically needed to compensate for our "being behind a computer/monitor/controller" way.