A quasi-objective way to judge physics of sim cars

Honestly this isn't a good way of checking accuracy. The number of variables you can't account for is enormous, you've selected only one aspect of one small facet of performance with no particularly strong rationale and you've got no meaningful margin of error. The only real way we have is numbers and that takes an intense amount of research, working with ideal conditions and approaching compromises objectively based on the end goals of the product; basically what a sim developer does. People rattle on and on about 'looks right' and 'feels right' but these notions are catastrophically misleading.

I've literally seen people give huge praises so called 'improvements' in car dynamics that weren't even implemented and praise dynamics en-masse that in all numerical senses were laughably off-base because they 'looked right' or 'felt right'. It's a strange and frankly unhelpful aspect of the sim community that people take 'this feels or looks like my (at best) partially informed guess seems like it should' as accuracy. If you've ever seen experienced, skilled racing drivers trying new cars and cars from different eras the first thing they'll talk about is how different it is and the ways it surprised them.
 
Last edited by a moderator:
@Minibull
Interesting thought. You may have a point on this. this quick change on slip angle (in stage 2 and 3) as Chris Harris did on real car might not be well being modeled in sim or such "discontinunity" transition may not be well thought/learned indexed into the lookup table(??) who knows...?

@Guineapiggy
Your notion is well taken. However, the assumption here as stated in the beginning of post is that there is no field data (real data) for sliding so objective study may not be possible. Such assumption is not baseless though. If you watch the interview video from Co-Founder and Lead Developer Stefano Casillo at Kunos Simulazioni, Assetto Corsa. He said there is no theory to be published on how and where to slide. The sliding state is "black art", there is no data he can compare to.
In the Chris Harris video (how to slide on the BMW), he said too that there is no theory to be written about how to do that (sliding).
 
^cool post Paul, good reading eh :)
I'll have to delve deeper into my setup, but then again, why bother, until I get something decent XD

Make no mistake, I can drive the cars fine and catch my slides and twitches. Just that there have been more than a few times where I don't catch it and think "eh?", put the replay on and slow motion, where I see the "delay".

Just something that might play a part sometimes, especially when trying to balance the sliding car
 
The problem, IMO, is that sim cars (with the exception of LFS) don't recover from a drifts/big slides in a natural way. It's got nothing to do with sensory information or hardware/software delays IMO, because even when real drivers react and correct far too late, the car still moves differently even compared to when we anticipate in the sim.

In real life, when the drift ends and the car starts to snap straight, the rear swings like a pendulum about the front. Even when the countersteer is held on too long (see some of the Youtube "drift fails"), it's only once the 'pendulum' has swung some distance back that the front of the car starts steering off in the counter-steered direction. Whereas, in the sims, the moment the rear starts to return, the front of the car starts turning in the same direction of rotation. The result is that the car rotates about a central axis and looks like both ends are sliding rather than the front initially planted and tail wagging. No wonder it's hard trying to keep up with it if you wait for the FFB message. But as I say, even if you anticipate it (as you have to) the recovery phase of the drift looks all wrong; you still see the front trying to dart off immediately the rear has hooked up.

I've just fired up LFS for the first time in a year to confirm my opinion and it's definitely a different looking dynamics, although it's not nearly as violent as real life, and too easy to drift.

Word of caution: if you've a Logitech wheel, they can go off centre if you jerk them around (bits come loose), so don't bugger it up trying to convince yourself I'm wrong about this...unless you really have to. ;)
 
Paul Loatman said:
Sounds like tire model issues, i think there's too much lateral grip, or the range of lateral grip loss isn't large enough which might cause the issue you're talking about. That said, i'm pretty sure that with the Panoz i can continue a drift after a corner on a straight for a short period and have the rear pull back in instead of the front sliding out, provided that the tires are a bit worn out, because it's possible to do both in a real car and i have done both, both by accident. It all depends on exactly how you're driving and if the car is up for it. I prefer having the entire car neutral at corner exit, i point the car at the exit, open the throttle, and try to line the car up with the direction i want to be traveling in at the end of the corner. What results is called a 4-wheel drift, but you aren't forced to do that in rF2.

I prefer to do that myself since it's a good training exercise because it requires you to balance the car perfectly through the entire corner in order for it to be pointing in the direction you want by the end of the corner, and it's almost an exaggerated version of the way i would normally drive some types of cars while racing. It takes more precision to be able to do this and make it look seamless, than it does to keep the tail of the car hanging out or to regain traction too soon. So even if what you're feeling or seeing is a result of the tire models, it's probably not as big of a difference as you think. The fact that LFS is simply easier to drive probably is the reason why you feel it's somehow different in it's dynamics, when in reality, it's probably just easier to drive it in a certain way compared to rF2, not that rF2 can't be driven in a similar way. I think LFS is easier because things are a bit exaggerated, meaning that you will see the car do relatively extreme things and you'll feel it's easy to do.

EDIT: Another possibility for what you're describing could actually be the STS. If you have a high STS it has a tendency to pull toward the sliding direction much more, meaning that you'll regain traction and the car will stop sliding before your wheel actually starts to straighten out. Which could result in the front sliding out instead of the rear continuing. You can try lowering the STS and try the same thing, if you rely heavily on the steering feedback, you'll probably notice that you can continue sliding much longer.
Click to expand...

Nope, you're missing the point. If the sim driver completely fails to straighten the steering during the recovery stage (maintains full counter-steer), then in LFS you see essentially the same behaviour as you see in real life when the driver is too late. And that's completely different to what an ISImotor sim will do. It's got nothing to do with hard/easy, skilled or unskilled. I have never, ever seen the correct behaviour in another sim. Over 10 years of sim racing however, I have seen lots of videos where people think it's happening correctly when it isn't (IMO).

And I don't need a lesson thanks - I know what a 4-wheel drift is. LOL. ;)
 
@Paul,
Thanks for posting, but I somehow could not open the file.
Quick question: were you trying to slide just like as Chris Harris did on his teaching video (step 1-3)? If not, it will not help though. I would not be able to tell if the sliding is realistic or not without referencing his. This is whole purpose of this post.
 
Last edited by a moderator:
Who said that this forum is only for new to rF2? Using your brain before typing comments like that, otherwise make you look so retar$$. Surely, no short of idiots here who is willing to echo such ,,,,, repeatly.
"toxic"? you must be joking. I bet the subject of this post maybe worth more than your total sum of threads in your lifetime here.
 
Last edited by a moderator:
Joe said:
Who said that this forum is only for new to rF2? Using your brain before typing comments like that, otherwise make you look so retar$$. Surely, no short of idiots here who is willing to echo such ,,,,, repeatly.
"toxic"? you must be joking. I bet the subject of this post maybe worth more than your total sum of threads in your lifetime here.
Click to expand...

You know the 'toxic' comment wasn't directed at your post, right?
 
This thread is a completely objective way to judge physics, not quasi-objective. It all comes down to your interpretation, and the assumption that you have the skill to drift.
 
You could make a very arcade game respond "correctly" to the described inputs without having any proper physics behind it.

Or did you mean subjective?
 
@Paul, I watched your video. Some of nice sliding. By watching your video, I am still could not conclude however. I 100% understand what you said.

@PLAYLIFE and @Lazza, I guess a "survey" probably is more "scientific" way to get a meaningful result on this. It could be more or less like this:

In his video as he taught three stages: https://www.youtube.com/watch?v=-LpPc8qutCA

Can you carry out these stages on at least three rF2 cars (rear wheel drive)

Stage 1 (Y or N)
Stage 2 (Y or N)
Stage 3 (Y or N)

Does this sound more "objective" to you?
If majority results showed all "Y", then the physics is correct.

(PS: keep in mind "full throttle" he mentioned all the time and showed in video --- he did many cars like just like this).
 
GTR Evo in 2011

Proper setup, skills.
 
You could make a very arcade game respond "correctly" to the described inputs without having any proper physics behind it.
Click to expand...
Ignoring all other issues this right here is the killer for this test. You could, with a little bit of effort put the kind of drifting seen here in to Big Rigs: Over the Road Racing but that doesn't mean you won't accelerate infinitely in reverse or drive up vertical hills with no resistance.
 
Last edited by a moderator:
Old Hat said:
The problem, IMO, is that sim cars (with the exception of LFS) don't recover from a drifts/big slides in a natural way. It's got nothing to do with sensory information or hardware/software delays IMO, because even when real drivers react and correct far too late, the car still moves differently even compared to when we anticipate in the sim.

In real life, when the drift ends and the car starts to snap straight, the rear swings like a pendulum about the front. Even when the countersteer is held on too long (see some of the Youtube "drift fails"), it's only once the 'pendulum' has swung some distance back that the front of the car starts steering off in the counter-steered direction. Whereas, in the sims, the moment the rear starts to return, the front of the car starts turning in the same direction of rotation. The result is that the car rotates about a central axis and looks like both ends are sliding rather than the front initially planted and tail wagging. No wonder it's hard trying to keep up with it if you wait for the FFB message. But as I say, even if you anticipate it (as you have to) the recovery phase of the drift looks all wrong; you still see the front trying to dart off immediately the rear has hooked up.

I've just fired up LFS for the first time in a year to confirm my opinion and it's definitely a different looking dynamics, although it's not nearly as violent as real life, and too easy to drift.
Click to expand...
+11

Old Hat said:
Nope, you're missing the point. If the sim driver completely fails to straighten the steering during the recovery stage (maintains full counter-steer), then in LFS you see essentially the same behaviour as you see in real life when the driver is too late. And that's completely different to what an ISImotor sim will do. It's got nothing to do with hard/easy, skilled or unskilled. I have never, ever seen the correct behaviour in another sim. Over 10 years of sim racing however, I have seen lots of videos where people think it's happening correctly when it isn't (IMO).

And I don't need a lesson thanks - I know what a 4-wheel drift is. LOL. ;)
Click to expand...
+111


Cars in real-life don't all of a sudden act as if the laws of physics broke/changed just because a driver might have been too late with their inputs.

The issues are purely a vehicle behavior issue - nothing to do with FFB, input lag, driver skills, or, for that matter, whether you saved the car, spun, half-spun, over-corrected, etc.

The following explains it fantastically in my opinion:

"The physics is misleading. Drive one or two hours of both and you'll likely agree that LFS has better physics. However, drive two weeks of both, and the tables even out. Drive two months of both, and they both have irritating breaks from reality. They clearly are both based on multibody time integrated dynamic simulations. rFactor is NOT table based at its core, although it seems to have more layers of something going on above this core...for better or worse. LFS has a cleaner rawer feel, like GPL, and seems truer to its own model, and breaks down only when the core model itself breaks down, which only happens in subtle ways. rFactor, like with FF, seems to have more hackery atop of its model, but often for the better. Any simulation model is just that, a model, a simplification, and they all break down somehow somewhere. LFS gently just seems to follow its math where ever it leads, always natural within its own reality. rFactor on the other hand seems to explicitly make an effort to feel right, as in more calibrated to reality in feel, within its sweet spot...which is normal driving within low mistake slip angles. Within this window I have to say rfactor actually feels better. However, when rfactor breaks down it is jarring and unnatural, unlike LFS. Worst case is pre-spin slip angles and angular momentum, where rFactor gets downright whacky.

To summarize physics, if you want a full performance envelope that feels self consistent and natural within its own approximation at all times, LFS wins. If you want more RL realistic feel within nominal conditions, and can deal with odd behavior and strange transitions in extreme conditions (like pre-spin),..."
Click to expand...
source -->www.lfs.net/forum/post/253018#post253018


That is exactly what I sense/see/feel as-well - that ISI physics tend to fall apart and get "digital" and unnatural once you start pushing the limits. Until then, they are, for the most part, great - possibly the best in the industry - but then again racing is all about pushing the car, experiencing those limits, fighting those limits, controlling and playing with those limits, and it's at those crucial areas where some of the best physics in all of simracing (ISI physics) fall apart and become some of the worst, unnatural physics in the industry. This is what makes driving the ISI engine so great at times but so frustrating and annoying once you push the car and drive outside of the physics engine's sweet spot.



P.S. That GTR Evo vid looks horrid and highlights some of the issues discussed here (more-so the second half of the slide)
 
Last edited by a moderator:
Spinelli said:
That is exactly what I sense/see/feel as-well - that ISI physics tend to fall apart and get "digital" and unnatural once you start pushing the limits.
Click to expand...

Won't hear a argument from me mate, they all get a bit lucid , you think " how did I ever get through there" .....whatever, you know what I mean ;)

But having said that I think that is a basic flaw in them all, you have ever evolving engines at the bare end of limits.

Thing is what rF2 does at that point is just better then the others....imho.

So as you say yes take the good with the not so good, it is always going to be a compromise.
 
Doing burnouts and rollbacks tells me all I need to know about physics.

It is all tied up in there for me.

Simply put I believe the best physics always do the best most controlled and realistic burnouts , best flicks, rollbacks, etc etc .


Try doing burnouts.
pCars .... what a joke........

AC .........absolutely mind numb-lying canned, you don't need to worry about throttle control or tyre heat, it will just keep spinning the same.

You get nice car control and feeling but then after a dozen donuts when you suddenly lift off 500rpm it just magically grips ? Like the car is not even sitting on 1/2 melted tyres.


If physics can't replicate even a little of a realife burnout with interia and chassis and body roll and all the other stuff you get going on in a real burnout then it's lame physics for me

I have done 1000's realife in powerful cars.

That is where other physics break down for me.

No, I am not saying rF2 does it realistically 100% :rolleyes: ..... just much better then anything else
and that is down to the physics and tyres.
 
Last edited by a moderator:
Old Hat said:
The problem, IMO, is that sim cars (with the exception of LFS) don't recover from a drifts/big slides in a natural way. It's got nothing to do with sensory information or hardware/software delays IMO, because even when real drivers react and correct far too late, the car still moves differently even compared to when we anticipate in the sim.

In real life, when the drift ends and the car starts to snap straight, the rear swings like a pendulum about the front. Even when the countersteer is held on too long (see some of the Youtube "drift fails"), it's only once the 'pendulum' has swung some distance back that the front of the car starts steering off in the counter-steered direction. Whereas, in the sims, the moment the rear starts to return, the front of the car starts turning in the same direction of rotation. The result is that the car rotates about a central axis and looks like both ends are sliding rather than the front initially planted and tail wagging. No wonder it's hard trying to keep up with it if you wait for the FFB message. But as I say, even if you anticipate it (as you have to) the recovery phase of the drift looks all wrong; you still see the front trying to dart off immediately the rear has hooked up.

I've just fired up LFS for the first time in a year to confirm my opinion and it's definitely a different looking dynamics, although it's not nearly as violent as real life, and too easy to drift.

Word of caution: if you've a Logitech wheel, they can go off centre if you jerk them around (bits come loose), so don't bugger it up trying to convince yourself I'm wrong about this...unless you really have to. ;)
Click to expand...
Imo one thing that could cause that difference is what you do with throttle when you recover from a slide. In car review videos they apply throttle when car is recovering from a slide so the rear doesn't suddenly grip and throw car to rotate to other direction.
Also how and which part of the car the replay camera follows can make things look different compared to real life and other sims.
 
You must log in or register to reply here.
Back
Top