So....who has PCars?

Pro`s and Con`s about pCars?

Con`s
- AI controlled Pitlane
- FFB very difficult to set up
- Physics seems Console optimised
- Modding unfriendly
- more?


Pro`s
- changeable Weather
- environmental ambiance
- career
- driver stats and info
- ready to drive, no addons
-
-
-

 

I'm willing to bet, besides the new tyre model, real turbos, and chassis flex, that RF2 is quite different than RF1.

The RF2 exe (the core program that only ISI can see) could be doing all sorts of new things and have all sorts of new systems in place to do all sorts of new/different things with the numbers from an HDV file. Unless you or anyone can see/analyze the coding/programming of the RF2 EXE then I'd hold off on saying the whole "it's almost just rFactor 1 with a better tyre model and chassis flex" thing.

 

Yeah that could well be that the exe does different calculations with those HDV numbers. All I'm saying is that you can basically convert a HDV file directly from rF1 mod to rF2, there are only a couple of new lines, diffuser/aero/wings have all the same parameters, which is what Tim implied as well (with taking most Corvette parameters directly from Niels). I have done this sort of basic conversion and it works pretty well, apart from the tire model.

 

The laws of physics haven't changed as far as I'm aware.

 

What agenda might that be?

---------

Let me correct you there, by the way:

simracing wise, I have mainly worked (both as modder and as racing teams consultant) with ISIMotor2, and specifically with rF1, GSC and Race07. So, where that "gmotor1 is the greatest simulator in the whole wide..." oddity came from I have no idea, maybe you'll grace us with an explanation for that as well.

Many seem to equate "higher complexity" with higher accuracy and higher precision, which is, to anyone informed on the matter, a fallacy. I have discussed that very same issue several times in the past at RD and NoGrip, even participated in articles published on the matter, not going to bother to revisit it again. People choose to believe "more/bigger is better", or that "more complexity" yields better results or provides higher physical fidelity (actually, it usually is the opposite), and I fully disagree.

Yes...but no need to "tweak to fit the data" (once you have a reliable engine).

That's one of the virtues of ISIMotor2's tire model (and other seemingly "simple" sim packages I know of). When you have the data (yes, I have dealt with incomplete, noisy or simply out-of-range data, but it's not always like that), the so-called "simple" models usually and easily allow for a 1:1 correlation, whereas the more complex models (especially the "physical/theoretical" ones) often fail to deliver the same results. Cruden's CTO, myself and other developers/engineers, discussed this issue a couple of months ago, our conclusions are pretty much similar - Pacejka/empiric/semi-empirical work quite well and can get us a long way, even if it takes a LOT OF KNOWLEDGE to analyse/tweak/verify (the data packages).

Er...I "can't stand rF2"?! Why?

----

Ah, modders as "old dogs", how polite...

In the past 4 years or so, I have worked alongside with race car engineers, data analysts, tire engineers and drivers, in different platforms and with different tools (sim modules, data acqs, etc).

I come from a Theoretical Physics background and prior to the simracing "virus" entering my system, I worked as a computer programmer in Telecommunications, Airspace & Defence companies.

My old dusty bags are full of "new tricks" that would last you a lifetime.

As for the "old timers"... much of the racing simulation software you have in your computer is being or has been produced/programmed by "old timers". You really do not know what you're talking about.

-------------------------

As my mate (F2Chump) stated, I have done all I can to promote simulations and dispel myths and dubious claims. Hence my praise for Stefano's NetKar Pro, my admiration for David Kaemmer's body of work, and all that Gjon and his team has done with the ISI engine (and for simracingdom at large).

I am a firm believer that you don't need (ever) to increase complexity in order to achieve good results. In my experience, the more you "complicate" things, the more likely you are to induce unforeseen side-effects to the systems modelling you're working on. [Dr. Feynman spoke of this often, too bad his lessons seem forgotten...]

Given how tire companies work (rubber and tire research/construction being highly guarded secrets), more complexity (in simracing models) only makes matters worse due to the obvious requirement.

The ISI engine, particularly its version 2, may not be flawless, may be older than the current next-gens making simracing headlines, but it still is one of the best tools for modelling a car's behaviour (not pretty, but certainly one of the best).

And while many simracers, eager for the eye-candy, decry ISIMotor2's age -when age has no bearing on the qualities that make it a reliable, all-round solid physics engine-, I see not one or two, but many professional drivers making use of it even today. Which is ironic and puts simracingdom's views in the proper perspective.

 

So what's your opinion on PCars? And I don't mean the business model...

 

The thing I would love to see most instead of this macho physics jargon bs talk.

Is to someone to explain in plain English the real differences between rf1 and rf2, regarding the handling the tyres etc. When I think one car tells me the difference another just disappoints and feels no difference to a car in gsc. Take away real road, I'm talking down to pure handling and physics. Because stuff that I hear was in rf1 (low speed grip loss for ex) is still in rf2, so I'm baffled.

If someone physics guru can come out and tell me this is plain English. Then I respect that more than that crap above!

 

Pro`s
- *environmental ambiance, just a few.... "varying levels of rain /audio, thunder and lightning, sunlight filtering through trees .......... "
- *career
- *driver stats and info
- *ready to drive, no addons

 

Not supposed to be a argument or insulting people either.

Get 3 physics gurus and the best 3 online racers .........they will all have differing opinions.

 

"I am a firm believer that you don't need (ever) to increase complexity in order to achieve good results. In my experience, the more you "complicate" things, the more likely you are to induce unforeseen side-effects to the systems modelling you're working on."

Yeah, but with that logic, how is progress made? A simple aero model that just relates speed to downforce would be easy to deal with, and easy to get a decent correlation to real life. But making a complex model that takes yaw into account and wind speeds would be possible to make more accurate, but could take more time to refine and get accurate.

 

So after pages and pages of posts and theories and speculation. No one can truly say what the difference they feel or if they even call it placebo.

I think some cars in gsc feel better than rf2. Reizas cars just seem to have this consistent quality which rf2 cars lack. Lets be honest rf2 cars areall over the place!?

Apart from spinellu can anyone else see a mountain difference between rf 2 and gcs cars anymore?

 

Oh no! One more of this self-proclaimed physics-guru´s!
Sorry, who is not able to see the progression between gmotor2 and 2.5 or even negate the advantages that it offers can´t be taken seriously (by me)!
LOL !

 

Faster you go the better it gets.

Nothing is like rF2 for me, not even close.

 

Mmm, coz how much of this would be opinion and speculation on our part? I use a G25 wheel, what level of forces can I feel? Piss all.

And as much as you are looking for a definitive and scientific answer, all that is being put up is opinion. All that would be countered with is opinion.

Not to mention it seems like you are in the middle of another swing of what's "real" to you. Drive whatever gives you a feeling of "reality". Who cares about trying to prove yours is right. Page after page of endless "debate"...

 

Not interested in discussing its business model.

Haven't tested pCARS on my own machine for well over a year now (am/was a junior member), so as always I'll refrain from commenting on its "handling", "realism", etc.

From all that was promised and said, given who's behind the physics engine, I -and probably many simracers as well- expected the ultimate racing sim from pCARS. By the end of the year we will all have the definitive opinion on that. For sure.

That is the problem right there. Progress does not necessarily translate to higher complexity.

Anyway, your comparing apples to oranges. The "simple aero model" you talk about above is not an aero model, it's just a simple equation. Pacejka-based models, other empirical/semi-empirical models are far more interesting and encompassing than a model that relates a certain quantity to another.

 

You know perfectly well what agenda I mean - what I said, apart from writing "gmotor1" when I obviously meant "gmotor2" (yes, I know it's ISImotor but most people on forums recognise it that way). Over the years I've seen you enthusiastically jumping in to agree with criticisms of Kaemmer's physical modelling strategy, and have been decidedly lukewarm about the idea of rF2's new model, whilst claiming these race professionals you "know" verify the accuracy of isimotor2 all the time.

That said, it is ironic that Kaemmer thinks temperature is such a critical element and has been modelling from the fundamental physical properties of rubber, but the tyre temperatures are less reliable in iRacing than any other sim last I checked.

What sort of programming were you doing in "Telecommunications, Airspace & Defence". I only ask because I'm suspicious. LOL

Regarding professional race teams and simulations. I suspect the use of isimotor2 isn't as impressive as it sounds. I hear they're not particularly concerned about accurately modelling tyres when sliding i.e. when drivers have screwed up. Up to that point, standard models are accepted by Kaemmer to do a reasonable job. It's because games need to simulate the complete experience that this matters more. In effect, games are more demanding than pro simulators. And Kaemmer and presumably the guy at LFS have concluded physical modelling is the way to do this. The trouble is it's obviously a LOT more complicated than they thought. Apparently, the LFS guy was last spotted on the event horizon of a black hole, examining individual photons with a magnifying glass. (I might have made that last bit up).

 

This "all over the place" argument is getting old. People said the same 2 years ago when ISI released new Howston, Civic, updated Megane etc. Because back then those cars felt much better than anything else. Now ISI releases updated CPM cars that all feel better. So the old content (Civic etc) suddenly got from awesome to useless? No, as a year ago they were still considered the best sim cars. The new content is just better, nothing wrong here. If ISI didn't release CPM you would still consider those old cars good, they didn't suddenly became unusable.

 

Funny you mention Chronus. I just had a PM conversation with him a few days ago on the RD site and, from what he said, I have to say I fully agree with you.

I explained some odd behaviors that at times occur in the ISI engine (ISI engine is the best in my opinion but every game/sim has odd moments in certain siutations/moments - nothing is perfect). Despite how "in-your face" and obvious it can be if you watch real-life cars and observe these particular moment/situation behaviors to in-game, Chronus pretty-much just totally dismissed everything I said just because I couldn't provide "hard numbers" (as if I'm some physicist or engineer or something).

In summary, Chronus pretty-much said what you said (RF1 engine, best ever, etc.). Chronus pretty-much said that the limitations of the RF1 engine are only the data that's put into. So everything can behave great all the time if the data entered is accurate and well correlated and such. I then said how could that be when every single car from at-least F1 2002 to RF1 (and RF2 as well, but things are improving) have the EXACT same behaviors no matter how good or bad the mod is? That must mean that every single human ever making a car in the last 15 years must all have used the same data which is obviously not the case. And also, how come not a single car out of the 100s, 1000s I've tried has ever, ever been free of these particular issues? It's clearly a result of the physics engine when the same thing happens to every single car for 15 years from all sorts of different people and data. But nope, not according to Chronus. He then proceeded to end the convo instead of trying to have an open-mind and look into and consider what I was saying. Heck, within 5 minutes of Saturday's real-life F1 qualifying session I already spotted the difference.

Right......Because other than tyres, RF1 simulates real-life absolutely 100% accurately. Infact, other than tyres, AC, PC, NKP, LFS, RBR, RF1, RF2, etc. are all pretty-much running identical vehicle dynamics and vehicle kinematics modeling systems....right...

How can people be so ignorant?

 

-----------

David has a thing about rFactor.

Only thing I see he says about rF2 around various forums is that ...... " online numbers are bad "

lool


He inferred rF2 can't match the others because they don't have the money ( resources, staff, etc. )

Fair enough......

SMS claim to have sold a 1,000,000 copies.
Never helped the physics did it.

 

that's a misquote DD, go up to see his actual post. Hes got a point though, the laws of physics hasn't changed.