RFactor 2: Best overall physics in simracing... FR3.5: ATM, one of the worst sim-cars

First, I believe the following to be a list of issues that plague either the ISI physics engine as a whole (going back to even before RF1, but also, slightly continually improved ever since, all the way up to and including rFactor 2), or just specifically the FR3.5 itself:



1. More of the vehicle's momentum/inertia should continue forward in the original direction of travel while the rear-end swings out, rather than so much of the vehicle's direction of travel changing so drastically as if you had a ton of grip and literally turned the car sharply left or right.

It's like the front-end just turns-in sharper and sharper with too much of the vehicle's original direction of travel changing to, and therefore following, the direction that the front-end is pointing towards. You can even easily see it when watching from a 3rd person, or trackside/TV cam; the car hardly looks like it's drifting, but rather just following the nose. Because of this, many times at the end of a slide you'll find yourself halfway, or completely, on the other side of the track because too much of the vehicle's original direction of travel was lost so you therefore find yourself just turning-in sharply and driving almost perpendicular relative to the track's direction and therefore heading towards the wall/side of the track nose-first.

The other sims do this too, but I feel ISI's engine may do this the most. The sim that did this, by far, the best and closest to real-life was the alpha/beta of an AMAZING driving game called Driver's Republic which sadly never came to fruition (the driving dynamics potential of that engine was mind-blowing). LFS is probably 2nd best (in this very specific area).

Is it strictly an issue with the core tyre physics, or something deeper in the overall physics engine? I don't know, but it's been here in every ISI-engine-based sim since F1 2002 and maybe even before (maybe SCGT? - Never played it).



2. At very slow speeds, when you crank the steering wheel very hard left-and-right (I like to call this, Fernando Alonso-style warm-up driving, lol) the vehicle will start to loose too much rear-lateral grip and start sliding even if you are applying neutral-throttle, and only a little amount of throttle at that. Again, this seems to happen with way too many cars and may be more of a core-thing rather than specific car-modelling.



3. On many cars - even at just 25%, or so throttle - once the rears start slipping just the most minute amount, the slippage, almost instantly, becomes WAY over-exaggerated. What do I mean by "over-exaggerated"? Well, for eg. the split-second you get just a bit of rear-slip - even at, let's say, 25% throttle - the car's engine/tyres want to increase their RPMs/speed WAY too quickly as if the car just gained 6000 lb/ft of torque, or as if you suddenly smashed the throttle to 100%. The RPMs / tyre-rotational-speeds just spike up insanely quickly - WAY too much and WAY too suddenly.

This last issue isn't experienced so much with less powerful cars like, for eg. The Skippy - I guess, due to it's engine (motor) modeling behaving as if it has a heavier flywheel and just generally more intertia in the entire drivetrain - however, the seeming flaw exposes itself more and more the more "hardcore" the car is. A perfect example this is the FR3.5 which is ABSOLUTELY ATROCIOUS in this specific area.



4. NOTE: I feel that the following issue is NOT a problem with the overall physics engine, but rather, just a car-specific problem with the RF2 FR3.5 (and maybe some other cars, not sure):
Adding a bit of initial throttle to plant the rear-end down in order to shift weight to the rear has literally almost no affect, or maybe it just seems like it doesn't have a large enough effect due to other physics issues (mentioned above)? I have no idea, all I know is that this hardly works in this car. Again, I have tried different setups, including ones with softer (not ridiculously soft, but soft-ER) rear slow-bump dampers, rear springs, and a softer rear anti-roll bar). Nothing reflects this effect to be working. You don't see it in the car behavior, you don't sense/feel it, nothing, it just doesn't happen.



5: NOTE: I'm not sure if this is a separate issue, or just down to one, or a combination, of the points listed above:
When tyres start loosing grip, it's sometimes sort of like they become elevated and are floating almost above the ground and barely coming into contact with the road, rather than what (I believe) a real tyre does which is literally just scrubbing/sliding against the road-surface and sort of "wanting" to grip. This may not have anything to do with issue #1 in my post (above), but may very-well be the reason (and therefore fixing it could be the cure) for issues #2 and #3. ALSO, this can be sensed a lot from the rear of the FR3.5, and it can be sensed a lot from all 4 tyres with the Howston G4/G6 models, especially the lower grip models. I believe that fixing this issue will also cure the Howston of the car's behavior where it is wandering all over the track. Remember, low grip doesn't mean a car has to wander all over the road as if there's a 100MPH crosswind combined with the car only weighing 100lbs.




HOWEVER ...
I still feel that, overall, RF2 definitely has the best physics; it's the most alive and dynamic, and reminds me of real-life hard-racecar-track-driving more than any other physics engine, especially in terms of all the small throttle/brake/steering inputs that can make such large differences which can make or break a driver. All the complexities, all the slips and slides (small and big), the way you can influence the car balance with your technique for better or for worse, and so on, and so on. Not to mention, RF2's FFB is #1 in the industry. However, ISI really need to get to fix those particular issues I noted above as I believe those are the main areas that ISI-physics dislikers have issues with, especially point #1.

[HR][/HR]
The following is a video, of the RF2 FR3.5, which I feel especially showcases point #s 3 and 4 (above) extremely well. Point #1 is noticed on just about any car, so, for now, just try to focus on point #s 3 and 4.

Note: I was using a slightly altered car-setup for the first part of the video. The setup included one extra tick of rear-wing, as well as a softer rear (softer by a tick or two rear springs and dampers, and I think anti-roll bar as well). Part way through the video I revert back to default setup with identical results. The car was using the "Medium" downforce package.

- If possible, watch at 60fps

P.S. Sorry about the sometimes-choppy video. Not sure what happened. Rest assured, the gameplay was 100% perfectly normal and this is definitely not the only time I've driven this car. Results are easily repeatable, as well as on other circuits.

[HR][/HR]
UPDATE - NEW VIDEO (Feb 2, 2015)

Again, I will quote what kamikaze has already stated:

ISI 2014 FR3.5 v1.1 - medium downforce package

- PLEASE, pause the video for each description and try to actually fully read and understand each description. REMEMBER, the fact that the car lost some grip isn't always the problem, it's the behavior of the physics once even the slightest amount of grip is lost that is the problem

- preferably watch @ 60fps

 

Is it a modded F3.5 2014? Or is it the FR3.5 2012 ?

The red text on the tires...


Some slip and drift actions with ISI F3.5 2014.

 

Almost looks from car behaviour that your running the smaller wing config Spinelli, obviously with your knowledge of RF2 Id assume your not., my set drives much more like Gijs video, either very well or controllable drift if required?.

 

From the throttle and brake trace, you are hitting it pretty suddenly through the corner with the throttle, or are going from a coast to then giving it a fair old whack, then the rear steps out, and the throttle trace stays steady. Once they are spinning, they aren't just going to grip again if you keep it spinning up lots.

 

As Bwana said, you have three different aero packages, and the high-downforce package is probably used for most circuits in real life, whereas low would only be used at Monza, i don't think anyone would use the low package even at Spa. Medium is probably rare too.

[MENTION=6003]Spinelli[/MENTION]; from your video, it honestly looks like you're just not correcting for the slides, you're also using quite a bit of trail braking, which if you're not even going to balance the car properly for, will obviously put you in a bad situation. Even if this car has the low downforce package, it shouldn't be hugely difficult to drive, so i have no idea what you're having problems with.

 

Gijs van Elderen...

- the snap-spin in your first video, especially the initial part of it, showcases exactly what I am talking about in point #1.
- the quarter spin at 2:57 in your second video also showcases point #1
- Video #3...same thing as the other two

Those vids haven't really show me much other than further showcasing my point #1.

The rotation does not happen at any time during braking. It's a case of atrocious vehicle behavior once you get some slip. I am correcting, of course I am.

Once you get some slip the revs shouldn't blast off to redline as if you have 6000 lb ft of torque. Watch some real racing man, I even see F1 cars, even in the rain, hold the revs as long as the driver does some very slight modulation with the throttle, it doesn't instantly blast-off to redline at the speed of light just because you went from a grippy 25% throttle to a slip-inducing 30% throttle.

And just because you learn to drive a car the way it wants to be driven in a sim, doesn't mean the car works properly. I could obviously just be super gentle, under-drive, make some massive setup changes, whatever, but the point here is to showcase some major flaws, and not that I personally can't drive a car.

 

The corner starts the moment you hit the brakes, that's when you start to change the balance of the car and at this point you either balance the car properly with your inputs and go through the corner, or you do what you did in your video and spin.

Maybe you should tell the devs to go watch some real racing, since you're telling them that they've got this wrong, it's not my responsibility to make sure this is 100% accurate. I can drive the cars just fine anyways.

 

You think I don't know what you're telling me? Honestly your the biggest fanboy of em all and praise every car as the best thing ever and as "I just can't see how it could get any better/realistic than this". You're blinded, don't spit racing technique 101 at me as if I'm some clueless newcomer.


- The car is balanced and has absolutely no sign of slip during braking, I have done this even while way off the brakes.

- There is also the major issue I described in point #4.

- You can see the car is not slipping at all at, for eg., 25% throttle, then as soon as you just slightly go over the limit, let's say 30% throttle, it just goes haywire as if I applied full throttle, rather than just a few more % from when the grip was just fine.

- You can even see that the throttle increase from the amount of throttle I had before slip, to the amount of throttle after slip, was only very little increase.

- Some of the spins are with even less than a quarter % throttle and practically almost fully neutral throttle where the car just hits a "dead zone" and wants to go crazy and acts as if I have the throttle floored.


Try giving the driving and car-behavior a deep and thorough analysis rather than just continually spitting out your biased fanboyism, and arrogant racing technique 101 as if everything is the driver's fault rather than the "perfect" car/game physics.

 

Well this is going well.

 

Yes. This is going well, isn't it. Calm down mate, I know you are serious about your simming, as are the rest of us. You do sometimes bring up good points to ponder and question...

In your video, for me, it clearly is the drivers fault. A stab up to 1/3 throttle is gently loading the rear up? In the middle of a corner? Tyres loaded up near their limits? No wonder the tyres have enough and slide. But to keep the same throttle down, of course it will start spinning up. They had started to spin, they spin more, they have less affect and grip, they spin more, you still have the throttle on and it keeps spinning, engine may be at a different part of its power band noelw, making more power, keeps spinning.
On one of those corners, you even add a little bit more throttle instead of giving it a fighting chance to grip.
Yep, you're doing that to show us the problems...but in doing it, those are the actual problems.

What would be interesting is a g meter of some kind, see what's actually going on in that respect. And even more interesting would be some kind of tyre load reporting stuff and engine reporting details.

In terms of drifting too, I noticed looking at a replay that whenever I'd spin off, I'd go slow motion to study it, and I'd see all my reactions were always slightly delayed in regards to the car movements. My guess is that any little bit of lag that is always going to be present makes it harder to control slides. Hence why I always slide, then snap back, then snap again the otherwise. Each correction was coming too late, and I was again late with each rolling off of the opposite lock, and rolling on the other direction.

Again...I hate to be labeled a fanboy, but I think ISI know what they're doing...
How I'd love a post from the physics engineer...at least to give an inside glance at what is going on.

 

You're talking about changing very slight amounts of things, like exactly how much momentum a car should have at a given rate of rotation. You know this by simply driving the car in the sim? How? What reference do you have for this particular car that makes you think it's wrong? Also, how wrong is it exactly?

If you can't answer these questions with facts, i won't bother replying.

 

Sorry guys, but the reason why I was so ticked off is because I knew - the minute I started thinking about posting this (before even beginning to type it) - that I would get EXACTLY that same response (learn to drive, it's your fault, car is fine, setup, etc. etc. etc.) from EXACTLY that same person (Paul Loatman).

You're kidding right? I'm not the only person in the past who has complained about these issues. I remember back a few years ago when quite a few people were complaining about issue #1, in-fact, I've had a few people agree about this on the forums in just the last month or so, and people have been mentioning it since the day I started sims with F1 2002 back before rFactor 1 even had a preview screenshot. It happens in every car, you can feel it, see it, can notice it even from trackside/TV cams and 3rd person view. I see it in F1 cars from ISI, 3rd party mods, your Howston drifting videos, classic 3rd party F1 mods, GSCE, the very best rF1 mods, GT Legends, the NSX (it was the very, very first thing I noticed in Tim's NSX @ Tiger Drome preview video), the karts (which I have a great time with), everything, absolutely everything and anything in an ISI-engined sim.

There are so many oddities in the behavior in my video, I can hardly even get through the first minute of watching it, it's a joke.

Issue #s 3 and 4 are so obvious.

 

The fact that you think about me this much is quite disturbing.

I wasn't the only person who said that it's probably your driving, either.

I generally don't block people on forums, but i make a few exceptions on occasion, and honestly Spinelli; your comments just sound like they stem from frustration or something, and you make everything so personal, if you have real reasons to be making these claims you would have already given them, rather than trying to pass off your opinion as fact. I just don't want to deal with that sort of behavior anymore. Sorry.

 

Fully understandable. I take this stuff seriously, and you are one of the people with whom I've had most discussions on here with (especially regarding FFB). I'm not thinking of you "this much" or hardly at all, the idea just suddenly came to me as I was spending time thinking about how to write-out my original post, that's all.

 

I know from experience it's easy to convince yourself of things (I have, on a couple of occasions over the years in rF1&2, tried setups before and after a tweak and felt the difference...then later discovered I'd stuffed up and they were identical). When you're looking for something, or even trying to test something, and your own inputs are involved, it's very difficult to draw definite conclusions. It's not a matter of wanting to influence things and being deceitful.

Before rF2 came out people were trying to judge it from preview videos and I didn't think it made much sense then, or now. People feel what they feel, see what they see, can learn to drive a virtual car whether it's realistic or not, and sims are good enough these days that the supposed flaws are very difficult to properly analyse.

Not worth it IMO.

 

until Isi releases another update and: oh my god, its perfect. "Most" perfect than before, now we have to update real world physics to match rf2
edit: if was for some people we could be in build 46. Never see perfection get so many updates
lol

i see that a lot. always puzzle me and i don´t think it helps game/sim development (i´m not saying you´re automatically right either because i have no clue about technical stuff- my pov is always subjective since my real car is fwd and cant reach 160 km/h)

back on topic

 

@ 0:47
Point #4
- He gives a decent stab of throttle to get some rear-end down, the car does not have an instant freak-out attack, not to mention it's in the wet

@ 1:25
Point #4
- First he gives the car a decent amount of initial stab to plant-rear in order to accelerate out, no car freaking out like it has 9 million lb-ft of torque @ small amounts of throttle (even if that throttle amount was a bit of a stab)

Point #3
- Later, as he is more straightened out, he goes to what seems like full-throttle (or thereabouts) too quickly and therefore get's wheelspin, however, the wheelspin doesn't shoot to redline at the speed of light, EVEN while he has wheelspin under power WHILE RUNNING OVER BAD CURBS, let alone just the regular track

Point #1
- While he gets wheelspin and lateral slip (sideways movement) from the rear, the car DOES NOT start darting around on it's front-end's axis and start shooting-off into the side of the track. Instead, most of the car's initial direction of travel/momentum continues forwards rather than changing according to the direction the nose is pointing. Again, as in the previous note, even while he's on the brutal Monza-Turn-1 curbs, the car doesn't just turn-arc into the side of the track according to the front-end, but instead, most momentum of the car's initial direction of travel CONTINUES TO MOVE FORWARD IN IT"S INITIAL DIRECTION OF TRAVEL, RATHER THAN HARSHLY TURN/ARCING LEFT OR RIGHT. Not to mention, he keeps the throttle pressed a substantial amount (maybe not 100% throttle [I obviously can't confirm] for the entire duration of slip, but you can tell it's definitely at fairly high amounts of throttle for just about the entire duration of wheelspin WHILE the rear-end still slithers left-right)

@ 2:18
Point #4
Very greasy conditions - Initial stab of throttle does not induce any rear slip, let alone a rear freak-out attack.

Point #1
About 1 second after the initial throttle stable, the car does get into a bit of a snap, however, once again, the rear-end swings and (important) THE CAR'S DIRECTION OF TRAVEL MOSTLY (BY A LARGE AMOUNT) CONTINUES ON IN THE SAME DIRECTION THAT THE CAR WAS MOVING TOWARDS BEFORE THE SLIDE OCCURED. That's why, after the nice quick correction, the car was literally hardly offline from the trajectory it was travelling towards before the slide occurred. However, in the game, much, much, much more of the car's direction of travel would have erratically changed and started turning/arcing left as if you were purposely turning the car left; therefore the car would have headed towards the direction the front-end was pointing to as the rear-slide occurred, and the in-game driver would have then been on the much more left-hand part of the track after correcting the car, and maybe even facing more to the left, rather than in real-life where the original direction of travel from before the slide, compared to after the slide, hardly changed and kept the driver on the same general path of momentum.

@ 3:10
Point #4
Again, pretty/quick initial throttle stab, not to mention in greasy conditions.

@ 3:30
Point #4
Again, more wet conditions, another stab, it's not the hardest stab of the bunch but one nonetheless

@ 3:33
Point 3
This time the initial stab, combined with the very high amount of lock he still had on the steering wheel, combined with the greasy conditions, did make the rear-step out, HOWEVER, when the rear stepped out the RPMs did not blast-off to a million RPM, AND (important) you can hear that the driver still held throttle down. Holding a bit of throttle can actual help settle the rear as you add some weight back there, obviously at this point the throttle is sensitive and if the driver wanted to he could press the throttle much more and make the RPMs shoot-up, but if you hold the RPMs and modulate and play with them a little you can play with the throttle and the weight WITHOUT HAVING THE RPMS BLAST-OFF TO REDLINE. I see this all the time in racing, in every series, even in F1, even in the rain, in every car in every condition. Anyone who watches racing will see this all the time. It's completely normal in real-life even for very sensitive and powerful cars.

Point #1
Again, even after the slide, the vehicle is still continuing on the original general motion that the car was travelling towards before the slide took place, rather than darting around or turn/arcing on it's front-end and therefore having it's direction of motion dictated by the direction that the front-end points to.

@ 0:15
Point 3
- He has very, very little initial throttle, and holds it at that very, very little initial amount, then, as he starts running further out wide, he very, very quickly goes from very light to very hard throttle. The car steps out but the revs don't blast-off

Point 1
- The car continues in the original direction of travel rather than suddenly and sharply turning/arcing to the right "on it's front-end".

@ 0:49
Point 3
- Hard on the power, so hard that not only does the rear come out, but it comes out quickly. However, the revs do not skyrocket as if the car has 100,000 lb-ft of torque, and ALSO (important), the driver hardly reduces the throttle at the sudden moment of grip loss; again, the revs and rear-tyres DO NOT skyrocket as if he has 100,000 lb-ft of torque and as if he is flooring the throttle to 100% even while still maintaining a darn-heck of a lot of throttle in the middle of the rear-slide.

Point 1
- The car continues outward in the original trajectory of travel rather than suddenly and sharply turning/arcing to the right "on it's front-end" during the car's slide-rotation. Drive, slide, correct, continue on in the same direction you were travelling before the slide took place, done.

@ 1:05 - 1:13 (the exits of both right handers) - this is more subtle than all the above - and even more subtle in the first right-hander - with much smaller angles and slips, but still showcases some points very nicely but in a slightly different way than the others
Point #3
- You can see quite obvious corrections from high lateral-load rear-slip under hard power. Again, he hardly reduces the power (if at all) yet the revs don't shoot up and start bouncing off the limiter into a huge rev-bouncing rear-tyre burnout spin.

Point #1
Once again, no front-end sharp turn/arc to the right. Instead, the car is heading to the left, he then does quick rear-slide corrections, then - with hardly any car-travel direction change - he just continues on and on as if nothing happened (in terms of the car's original direction of travel). No sharp front-end turning/arcing to the right, no weird front-end darting around dictating the direction of motion of the car, no weird, un-natural snap wobbles. Just an ever natural continuous flow (for the most part) of the original direction of vehicle-motion, with some corrections in-between.





NOTE: I have - as I'm sure most, if not all, of you have - seen this type of behavior (shown in the videos above) in all sorts of cars, not just FR3.5s - F1, GP2, F3, etc. Wet, dry, super wet, greasy, etc. I was going to also post a ton of GP2 videos - since they have very similar performance to the FR3.5 - but then figured I've shown enough, plus, I'm sure some of you would have just then said "well that's a different car, you can't compare the two" as an excuse/defence before even looking/thinking things over . Lots of these issues are not only specific to the FR3.5.

 

As many have stated before we have no solid facts to base our opinions, so my opinion will be totally subjective.

I disagree with the general affirmation that all ISI based sims have the cars spinning around front tires (my RF1 mods with an adequate setup dont show this behaviour neither).

I can drive the F3.5 without any particular problem, did it on the same track as the OP (with the high down force aero that seems more suitable in my view to this track) before replying to this thread, I don't do this often because this car doesn't report ride heights to Motec (a big no for me, so when I want to drive an modern OW I choose another car).

Is the tire model perfect at the moment?

I don't think so, even ISI doesn't think so, and even when the sim will be perfected, it's complexity may mean that we never will drive a car with adequate tire behaviour due to it's complexity, the lack of data available for the tires of this car, or the modder skills.

It would be enough that the thermal inertia of the tire to be slightly off to cause the tire to loose grip to fast after a strong solicitation, and make spins more difficult to catch than necessary.

I agree with the OP when he says that in real life, tires give you the impression they are fighting hard to maintain their grip, in RF2 I have the impression of driving on a smooth glass track (only exception Sebring because it is bumpier) the tires happily glide smoothly laterally without any fight (Silverstone is a perfect sample). May be they fight, but you don't know (no variation in noise, nothing in the steering, no vibration nor motion in cockpit).

Cheers and please maintain a cool discussion.

 

Good points Spin, I agree with you about this.

There is loads of topics about FFB but hardly non about pedals and as a mater of fact pedals are as important device as you wheel and when you are driving real racing car, pedal control is as important as wheel control, drive train , throttle response, clutch, normally aspirated-or turbo etc.car are not simulated at all.

ISI has many good cars some is hard to drive and some easy but lack of proper drive train it is hard to simulate real racing car and if you take C6R GT2 in real life and and use pedals as you do in rF2 you are dead in first corner , there is no RS bottom in real life.

 

For me the FR3.5 is the most slippery car ever in a sim and I can't do a single lap....but then again I'm a super noob so this must be the reason. Pity, as I can never drive this car....