Give your opinions please, witch one nails car physics?

[W_Akira]

I have my favorites, but I'm curious to know how you people feel about those simulators in terms of car physics.
Just driving the car in a dry course to even things.
Feel free to post why you choose your favorite.

 

[W_Akira]

For me, despite the akward slidding when you loose control, is iRacing that pass the feeling I'm driving a real car. Weight transfer, braking, accelerating feel very natural.

 

[Eric Bergeron]

for me is rFactor2, Imho (i am not a race driver) it is the Sim who has the best car control if I compare at this video. starting at 15min.

 

[April Dillon]

rFactor 2 for me also, though mainly because of the transparency of it, and what I've genuinely seen working in it, as opposed to just 'felt'.

Racing sim physics fascinates me, so I tend to look closely at them when they let me. I can take a look at a lot of physics data in something like rF2 to see what's going on. It's great when that kind of thing's possible, and I genuinely enjoy a sim more if I can look closer at the technical details like this.

There's some very elegant solutions in how wear and temperatures work in rF2. It maintains temperature and wear at numerous places on each single tire, and details like flat spots just occur organically within the system.

I also quite like how the tires are springy in all directions, not just vertically. I'm pretty sure pCars has this also (and LFS, if I recall correctly), but I'm not sure about others. In anything rF1 based, the springiness is only vertical. The vertical spring is definitely the most important, but I'm pretty sure other directions add something. Especially any kind of flex around the 'roll' axis, which is good for properly simulating camber and why we want it on race cars.

The turbochargers in rF2 are especially crazy IMO. Assetto Corsa approximates turbos quite well with just a few parameters. But for a turbo in rF2 you need an entire compressor map... Now that's thorough.

 

[Gijs van Elderen]

Nailing the physics is not the same as giving a believable driving experience.

AC, Pcars, RRRE give a believable driving experience. You can even adjust some setting on the FFB to give you the believable driving experience you are looking for. With much less detailed physics as rF2 and Iracing.

Physics = input, output data: things you'll notice in motec data. tire temps, spring loads, alignment dynamically while driving VS your static alignment in your car setup, weight distribution, aerodynamics, etc,...

So if i would answer your question correctly:
Iracing should nail the physics with the most accurate car/track data.
But it doesn't give me the believable driving experience at all.

rF2 is mainly focussed on the physics and let the driving experience just be a result of the that.
The FFB is just a result of the physics calculation.

 

[W_Akira]

What I meant by feel like driving a real car is to use real drive technics to go fast, being able to know what the inputs on wheel, accelerator, brake and clutch will make to the car (weight transfer, understeer, oversteer and others). Also the car must behave realistic.
Maybe I should put more time on rFactor2. I've mainly played iRacing for online, Assetto for the blast that is drive in Nordschleife on road cars, rFactor2 for the best formula experience I've seen, R3E for the fun of WTCC 2014 and PCars for career. I've done some great online league championships on GSC too.

 

[Ghoults]

I think with good data all sims can do really good job. But with bad data you can get bad and strangeness from every sim. Imho rf2 has the best physics engine whereas ac has basically all of its cars done well from data point of view. Game stock car I don't own (waiting for discounts on steam) so I can't comment much on that. My impression is that rf1 engine can do better job than the ac engine in certain ways simply because ac lacks so many basic features. But overall I'd drive a car in ac than rf1 if the car is done really well.

When it comes to being the best sim I don't really see rrrrrre or pcars even trying to be sims. They are driving games and whereas someone may think this distinction is just wordplay I think it is important to have words and terms that have a meaning instead of everything being labelled as sim just because some pr team thinks mentioning that moves more products.

Physics is also a term that needs to be defined. It is 2015 and not having rain for example is not really acceptable anymore. I don't think driving in the dry and how it feels should be the measuring stick for racing sims anymore in 2015. The basic stuff should already be there (correct handling of high slip angles, aero ground effects, tire wear and heat, physics based ffb and from game pov good usable multiplayer, good ai and game options available in the game instead of random collection of hidden config files) and we should discuss about features and finer details and not using the same metrics we used when gpl was still the king.

I think overall driving sims are doing great. There is so much choise. But I don't think anybody needs to choose just one. Just for good physics there are couple of good alternatives that all have their strengths and weaknesses and what one likes is more likely to be based on how those strengths and weaknesses match your wishes and not-care-about-that-feelings.

So for me it is ac and rf2. When rf2 gets it right it is the best. Often times though I feel that ac nails it and rf2 doesn't.

Anyways, here are my bullet points for each:

rf2
-variety of racing surfaces and environments
-best ffb of all sims
-complex support of different car physics
-physics that can be the best in the business

ac
-all cars use good data making them all use full potential of the (limited) physics engine
-best feel of any driving sim (I have trouble feeling the cars in rf1 and rf2 at times, it is all about something being different in the visual aspects of how the sim conveys the car movement via the screen)

 

[Lou Sytsma]

rFactor2 & Stock Car Extreme - not just the cars but the road too. Feels very organic.

Many rave about Assetto Corsa but I find it has little road feedback - except for bumps and the car handling while informative feels digital.

 

[Emery]

R3E's sim strength is the default brake pedal mapping. There is so much better control with my ancient G25 pedals in R3E that I'm still trying to get rF2's brake pedal to feel as good. Weird that the same controls can feel so different across programs!

Edit: Hah, went back and fiddled with sensitivity (linearity), turned brake down to 75% in rF2 and that helped. Did that long ago with throttle.

 

[P*Funk]

This is a topic that involves much more granularity than a single choice poll can represent. It also involves some fairly potent design philosophy and the nature of experiencing things.

Certainly anecdotal belief in one over another is the best way to describe preference, but its not the best way to judge an overall simulation. Its been mentioned, what good are physics if the interface that allows us to judge the car's handling is poor or flawed?

Also, the notion that a more complex model is inevitably more accurate isn't necessarily true and the presumption that it is is in my opinion a huge part of why people denigrate Stock Car Extreme, even without having tried it.

Perceptions are subjective of course, but what if I can judge the validity of my own perceptions, the usefulness of them and their inherent biases, by trying to be more objective about racing simulation?

I love this one quote by a statistician that is very relevant to this discussion I think.

Since all models are wrong the scientist cannot obtain a "correct" one by excessive elaboration. On the contrary following William of Occam he should seek an economical description of natural phenomena. Just as the ability to devise simple but evocative models is the signature of the great scientist so overelaboration and overparameterization is often the mark of mediocrity.

https://en.wikipedia.org/wiki/All_models_are_wrong

Therefore I suggest that Occam's Razor implies that there is a great degree of relevance to using a simplified model over a complex one for achieving more accurate results. So, the question then is how is say rF2's Turbo model better than Assetto Corsa? Why is the objectively less complex tire model for Reiza's game not just as valid as the more complex ones in say AC or rF2?

Food for thought I say.

 

[David Ignjatovic]

With G27, I vote SCE as the best, it also has good gfx and good sounds.
rf2 requires a high torque wheel, this might also explain iceracings problem, ie, too much clipping on the low end wheels to enable the feel of the rear end.

 

[Emery]

At their very hearts, GSCE, R3E, rF2, and, as I recall, pCARS are all the same. They're all derived from ISI's pMotor. Where they differ is what data is fed to the core, how the output is displayed/interacted with, and differing approaches to the tire model. The tire model and the output presentation are the two things that influence how we feel when we use them.

So gearboxes are gearboxes, engines are engines, drag is drag, friction is friction, and masses are, er, massive. Change the masses and the inertias change, etc. As long as those numbers are good, then the end result will be in the ballpark up until you get to the tires because tires are still just shy of black magic.

The presentation to us in the driver's seat, is dependent on update rates and strength of force feedback before clipping occurs. Slow update rates, low available FFB are problems and there are different approaches to dealing with them.

GSCE stuck with proven technology and Niels is a whiz with it. R3E also stuck with proven technology and have several experts; slightly different presentation than R3E that's suitable for their goals. Update rates for both are low and they're prepared to handle low FFB. rF2 and pCARS added new tire technology and neither are completely at home with low update rates and low FFB, though they can be adapted.

 

[cristianno]

I believe that everyone is very good, if I had space on my HD settled all of these, plus the rFactor 1 Racer Free Car Simulator and Grand Prix Legends ... but AC is the best ... for old low end hardware Racer is the best

 

[P*Funk]

I will give AC the award for best optimized sim with modern graphics. I have an 8 year old PC with a 5 year old graphics card in it and I can run AC at 120 FPS while it looks pretty fine.

 

[Arne Dopudja]

Pcars, LFS. Because of the tyre systems with various forces applied to them, unlike rfactor/race07 which I never really fell in love with despite spending hundreds of hours in each.

 

[Mark L]

At their very hearts, GSCE, R3E, rF2, and, as I recall, pCARS are all the same. They're all derived from ISI's pMotor. Where they differ is what data is fed to the core, how the output is displayed/interacted with, and differing approaches to the tire model. The tire model and the output presentation are the two things that influence how we feel when we use them.

So gearboxes are gearboxes, engines are engines, drag is drag, friction is friction, and masses are, er, massive. Change the masses and the inertias change, etc. As long as those numbers are good, then the end result will be in the ballpark up until you get to the tires because tires are still just shy of black magic.

The presentation to us in the driver's seat, is dependent on update rates and strength of force feedback before clipping occurs. Slow update rates, low available FFB are problems and there are different approaches to dealing with them.

GSCE stuck with proven technology and Niels is a whiz with it. R3E also stuck with proven technology and have several experts; slightly different presentation than R3E that's suitable for their goals. Update rates for both are low and they're prepared to handle low FFB. rF2 and pCARS added new tire technology and neither are completely at home with low update rates and low FFB, though they can be adapted.

one clarification - doesn't Pcars the title utilise a upgraded version of PhysX Madness engine by Nvidia as used for Need for Speed?

https://en.wikipedia.org/wiki/Project_CARS

 

[Emery]

GTR --> Project Ferrari --> NFS --> pCARS
Simbin/Tentacle............Blimey/EA...SMS
ISI supplied the pre-rFactor stuff for GTR. Ian Bell in 2012 said the pCAR roots are in Project Ferrari rather than NFS. Andy Tudor is quoted as saying the roots are in NFS. Ian Bell is most proud of GTR.

So... I tend to go with the theory that there's been a lot of modification, but I think the core physics go back to ISI's work that was licensed by Simbin and Ian Bell has never let go of that control. He has to distance himself from the NFS series and he doesn't dare admit older heritage.

 

[Tobbe Bergman]

Assetto Corsa top my list and RBR comes next.
I've tried rF2 but to me it felt like a half finished Sim.

Somehow the poster forgot RBR but mention pCARS.
Now that's funny

 

[Dan Allen]

AC has pretty cool physics, as does pCARS.

 

[Arne Dopudja]

I come back to this thread to offer a very simple, artistic even, opinion on physics. Now we cannot really compare physics to the real deal, especially without any movement simulator so for me, saying that some physics are a bit better than others is empty talk.

I like to gauge physics from my body to to the road in the game. To explain, I put an imaginary distance between eyes-hands-wheel-car body-tires-surface. So if you can imagine this as a car; closer they are overall the better it is and equal spacing between them is the perfect way to go. Like a gokart is very nimble and direct opposed to a heavy slow long stretch limo.

For instance AC: For my feel, the relation between my wheel and car body is very distant and I don't mean FFB wise. Otherwise all other connections are close enough. The cars feel fine but I don't feel involved in the car.
PCars: A bit stretched out overall but equal distance between all the components. I feel involved almost perfectly, though it is all a bit hazy.
Race07: Very direct connection to the car, but then between the car and tires and the surface it's a bit distant. I feel very "in the car" but not so much "on the track", opposite of AC.