"Light cars"-VP review of rF2

Can't really understand why some people always say rF2 cars feel a bit too "light"... VP is not the first one I heard saying that over the years. Wonder if there are any truth to some sort of issue with some inertia and weight transfer code? Just really curious. If one comes from AC and ACC it's normal to feel that way though. There many cars are tank like from what I remember.

 

I remember a few over the years who were very quick to question weight and inertia, though apparently without actually doing any tests or basic modding to check (It's rFactor... It's not like modding values is difficult!).

Weight, inertia, are also very basic concepts well covered by high school level physics, so it's amusing that people think a game would have it wrong.


On the other hand, there are many other factors contributing to handling and the experience/perception of it (ie I think the exact same physics could "feel" different between games), so if that's the best way to describe it maybe that will do. Very much doubt it's actually those things though.

 

I don't think that the basic physics are wrong, but as I have said in the past I do agree that at times the cars can feel a bit light. This is what I think...

1. On lower end wheels (such as G920), the ffb can feel very weak compared to some other sims. Other sims really load up a lot more when applying steering lock, whereas rF2 only really feels heavy under braking. (rF2 ffb is still absolutely the best and most accurate imo, but maybe a DD does it a more justice and makes it feel a bit heavier). The slight centre deadzone on lower end wheels seems a bit more pronounced in rF2 aswell, which doesn't help the situation.

2. The cars in rF2 are very very quick to change direction on initial turn in (relative to other sims... not saying that this is unrealistic). Look at a recent video by our friend Henri Sinik at the Nord... The way the front end of the BMW Class1 is changing direction almost seems to defy the laws of physics.

I think both of these points come down to the tyre model being quite sensitive to load (and the raw ffb) - a lot of load on the tyres and there is a huge amount of grip, less load on the tyres and the grip is reduced massively. So at times the car changes direction like crazy, at other times the front is not loaded up and the ffb feels light.

On top of this, if the caster is even slightly on the low side (such as the Radical SR3) the self aligning torque really drops of dramatically and even ever so slightly reverses.

 

The reason is quite obvious and simple: in RF2 you don't want clipping because you don't have any means to "autotune" your FFB output, in other titles where you can configure FFB output to your liking (that can even be completely unrelated to reality), this is less of an issue: you want the wheel heavier? You got it, you want to feel curbs more or less? you can have it. RF2 is unable to employ only the upper half of the torque of a low tier wheel without massive clipping. Compromises have to be done: RF2 compromises is that is hardware dependent, other titles compromises are that the FFB effect have a "layer" between the physics and the player, if you mess with that layer you can mess with the whole experience. Every solution has pro and cons.

 

I couldn't understand many things that people are talking about physics of cars in simulations for a long time too. But one day after listening to Norbert Singer, I understood it. The reason is simple - they don't know what they are talking about precisely. Of course they try to communicate what they feel, what they perceive and what they think is going on, but due to complexity of the subjects it is common that what people speak about what car does, and what they actually mean can be different. There is where genius of N.Singer comes in, he realized that and didn't necessarily take literally what drivers spoke, instead interpreted it so it made sense at the end. And thats about getting a feedback from professional, top level drivers !

In simulation we get one step higher in terms of complexity, because IRL physics just happens, so no one has to bother thinking about if physics are incorrect. In simulation we have lots of confusion because it could be anything. But the most common mistake is lack of realization that each car has its own parameters, and all these creators of content mostly generalize everything down to physics code of game. A lot more attention really should be given to each cars individually. Unfortunately it is just not what content creators do most of the time. Most of them doesn't have sufficient expertise anyway...

As mentioned before the weight of FFB has a lot to do with perception of cars weight/amount of friction at the tires. Secondly, I don't get why 1000kg car is assumed as heavy. Then the tires, depending on construction, compound and tread they should produce very different response to steering. Then the speed that car is traveling at, at higher speed relaxation length will be less perceivable, so cars will respond quicker, and lesser steering inputs will also result in more deviation from straight line at the same fraction of a time. Then there are springs, dampers, geometry, roll bars... It all together with tires will cause natural frequency of a car to be higher or lower, higher frequency will cause more responsive quicker transients. Soft for roll car will seem heavier due to slower transients, good thing it will give better traction, bad thing is that with sharper better hooking tires it is more likely to cause possibility of driver induced oscillations and loss of control, as body roll continues after tires gripping up already. Then there also is steering ratio, depending on it car turns quicker or slower in relation to degrees of steering wheel turning in cockpit.

LAST BUT NOT LEAST. HOWSTON DISSENTER HAS BROKEN STEERING LIKE ALL CARS WITH OLD HDV STEERING ENTRIES ARE SINCE SOME RF2 UPDATE LAST YEAR. ACTUAL STEERING DOESN"T MATCH THE STEERING IN GAME AND IT MAKES CARS FEEL WORSE. IF YOU DON"T BELIEVE IT GO AND CHECK IT OUT YOURSELF. AND SPREAD THIS MESSAGE BECAUSE I HAVE LIKE 70 DIFFERENT CARS PHYSICS WITH OLD HDV STEERING ENTRIES LOL !!!!111

 

I would be interested if people would assume that a light car, or heavy car physics ?

In my mind what the car does in this example you shared is correct. But the better question would be is it correct how it does it ? In my mind this is too nerfed and too smoothed out, it doesn't look like Aris needs to put any effort in driving like that, it just isn't edgy although it oversteer and understeer greatly in these couple of turns as he demonstrates how car behaves when overdriving, which is most important aspect of racing cars simulation.

I don't think it is unrealistic for this car to be understeering at 150km/h with smashed throttle. It depends on load distribution on tires - aerodynamics, road incline, actual weight distribution of a car, tires mechanical grip front vs rear as it comes from tires themselves and body roll stiffness distribution.

In my opinion it does what it has to, but not edgy at all. Feels like there is minimal danger of a tankslap when oversteering, so steering can be slow, minimalistic, no guessing range in it, it is too sure, too slow, too much passive from a driver. But sadly its what people likes, even in virtuality avoiding responsibilities. Feels like there is minimal danger of snap oversteer happening when fronts eventually hook up at highspeed, in that case chances of understeer immediately switching to oversteer should be large. But these things can't be happening if physics adjusted for gentle fading from sliding to static friction, would it be to aero lacking dynamics, or friction lacking dynamics. Or perhaps Aris just is this much of a god at driving that he doesn't even need to put any effort in it anymore, but I think it is just because "three-months simracers" likes it softcore these days.

 

Thanks for the comments guys.

Taking a little bit of everything from all of you above I have my own theories I've actually propposed before on a simillar post - I just found it interesting to revisit this since it's a video from a youtuber with a good reaching and understands one or two things about vehicle dynamics and has experienced many simulation and racing games.

The first thing to observe is that when he is making the comments on the light car feeling and lack of lateral and longitudional inertia and weight transfer is on a older car, with tires and dynamics wich we all know are of cars that drifted quite a bit to be driven fast and taken advantaged of, since tires at the time weren't as peaky as they are now with modern slicks. This leads me to the conclusion that when I'm driving simillar cars is that the cars rotate so much (yaw inertia) and the tires seem to laterally hold no grip at all, it makes it all feel like only the yaw weight transfer is at play, since tires are the contact with the road that make the cars lean or not in terms of roll inertia.

Other than that the way suspension and things like that are simulated, when compared to AC and ACC for example, is that it's much more smooth when it gets compressed and decompressed. When looking at onboard gameplay comparisons and real life, there seems to be a bit more of an "humf" in AC/ACC suspensions (GT3 cars in this case, but it's a bit general in the sim, even if it's of course content related, some suspension compoents and stuff will have to share common code).

@mantasisg

What you mean by softcore simulation can also be fully applied to rF2 in many many cars, probably the majority of them.
Yesterday I was thinking about it while driving my road car at the highway at around 140kmh. I was wondering that it's quite a good feel of speed and I could feel some edginess in the air outside influencing my car. I was thinking that if I lost the car in a moment like that while I could possibly still save it, it would be much more abrupt and not as smooth as saving casually some cars in rF2 that you don't even have sometimes to countersteer at the propper angle, due to how much leeway there is in the initial moments of slide wich take a bit more time then I think to accelarate to a nastier slide...

While I think the steering thing is a problem, I don't think it will justify these perceived issues with the cars though.

When it comes to FFB... Well, VP has not really made it about how it feels on the wheel, he is actually looking directly at how the game is handling in despite of the FFB and even gives an example of crashing into a barrier.

The good thing is that many many more cars in the game lately (mods and official content) have a much better feeling of "mass" to the cars, wheter is tire related or not.

 

WOW that real onboard is fantastic, you can see that driving fast is taking place there. And ACC as expected, lazy, boring, no snapping, no hooking up, tires transfer from sliding to static griping just too ideally, the transients are slow and so the smooth driving. People could argue that it is also a driving style thing, could be some truth in there, but I would refuse it as complete explanation of such massive difference to reality.

And I disagree with theory that this is consistent in AC engine. It might seem so because all cars are made to be like that, there is consistency in content. I had some mods in AC with which I tried to break the "consistent" logic originating from main devs, and AC people just doesn't like experimentation. DRM mod team initially faced plenty of negativity even disrespect because they also did build physics in a bit of more their own way (which I liked a lot), but AC people are fixated on their original paradigm and BASTA.

ACC physics are actually quite advanced over AC, and was so from very beginning. The very first ACC release had rather impressive and very realistic handling in my opinion, well perhaps a bit on "too hard" side, but not crazy too much. The Huracan GT3 at Nurburgring GT layout was VERY VERY alive, it was snapping into oversteer from understeer. The corrections of oversteer had to be very well executed, quickly and precisely. There was great sense of neutral steer when fronts and rears were at about equal slippage, and you had to do this quick sawing action in slight guessing range of steering motion to find the correct front wheel direction quickly. The feeling of traction was very bold, you knew that you have a plenty of grip, but also knew that if you overdo it a little it will drop massively, and then hook up strongly after, so you better correct car well. The moment rears hook up when exiting the corner already in straight line felt amazing, you really felt that moment when wheelspin ends and acceleration ramps up at that moment - an amazing feel. The car in the rain worked spectacularly realistic. Tweaks of a setup meant a lot, everything was just working together. Having tires and/or aero not responding realistically, transients being way off definitely reduces importance and accuracy of setup changes, since usually more than half of setup is exactly meant to alter the transients, change the speeds and amounts of tires loading and alignment in every scenario as best as possible to achieve most possible net friction and/or stability per lap, or per duration of a race. Proper tires and aero is almost everything.

I have spent over 60hours in it. Interestingly enough first day or two of EA it felt horrible. They did a little hotfix quickly, and it started being awesome to me. IDK how much it was due to hotfix, and how much due to me getting used to it.

Unfortunately it was breaking out of AC paradigm way too much, as expected AC fans rejected new paradigm. But I was surprised by how much Kunos reurned to their original AC way in very next build, I was expecting some stepping back closer to AC handling character, but they went back massively, and I think they continued on going back, up to a point where they probably have handling even more spoiled and softened, but IDK, I stopped carrying about ACC, way more interesting to me is the dynamic of public perception of reality and how developers are adjusting to it in simracing as a whole. IMO it is a problem. And you know what was a common theme all the time - the FFB, people couldn't comprehend that physics are different, they kept on chanting: "something is about FFB"...

@Nieubermesch You are right there are tons of rF2 cars that are a bit too soft (not in terms of suspension), but forgiveness to loose unfocused driving.

YEAH driving at 140km/h with average road car already feels like speed. It wouldn't be same thing with a modern race car and on wide racetrack. However, loosing control of road car 140km/h would feel safer, more forgiving than race car. Because race car would be loosing so much more friction, and in so much smaller fraction of a second, due to massively higher initial friction and stiffer, higher frequency chassis. With race car you'd probably have to react five times faster and precisely assuming loss of control would be equally large, and you'd have to react fast coming into slide, and then returning from it as well. These things are absolutely the most related to dynamics of tire friction and aero. Unfortunately they are also probably hardest to collect data, reliable information - so many devs can do whatever they want. They can make a tire that performs like modern top tier racing tire, while it will be as demanding from driver as street car tire. And then if no one calls out them on suspicious handling, and even worse LOVES it, then devs do it that way.

Speaking of classic cars they were actually light, also their racing tires were probably closer to modern street tires, than to modern racing tires, IDK about slicks, but 60s crossplies with tread probably even lesser performing than above average modern street tire. Again it is highly difficult to interpret what VP actually has in his mind when he spoke about that, I think there are gaps in his logic. Although it was interesting comment about collisions, but I refused to even think about it - it is absolutely different subject, the walls in rF2 can have different collision damping, the angle of collision is important, perhaps slight deformations of car are taken into account of damping too, although no deformations takes place. I personally don't crash too much, so I am lacking experience on this subject

 

Yeah, I know with a race car losing it would result in a much snappier slide and my reactions needed to be faster for sure. I'm not quite sure how hard it would be to save it in real life at that velocity, but I've lost the car in the rain in a roundabout going probably only 60kmh and while I saved it, it was after some snap oversteer two times after initial corrections. Rain is a different thing entirely, but still, can only imagine how scary it could be to lose it on the highway when I can feel all that air compressing the car and making it a bit wobbly and my car is like 1500kgs or something. Some other road cars existing in rF2 at that velocity are a joke to correct honestly, even after some good degrees of yaw angle on that slide. GT cars by S397 are simply a joke.

Them having a much more sliding feel to them and not thight at all even compared to road tires of today, it's why I think VP driving something like the Dissenter felt so much lacking in the roll inertia department, because they seem to really be holding almost no load to make the car roll inertia act, as the tires can't bite very hard at all lateraly, and then it becomes a bit of a "yawy" car to drive, a bit like a plane but with the tires. Some of it can also be down to differentials? If the car has open differential - any car - then you will feel so much yaw the lateral side of the tires also don't tight up that much at all and it leaves you feeling like something is missing in the weight transfer department if you expected more roll on the car.

 

One very interesting comment by VP on his video:

So, I'm back. I assume you and your friends won't agree on what I'll say that because you start with the assumption I'm wrong and you're right -and that's quite disappointing already because basically I'm wasting my time to tell you the truth which you won't accept. But I'm meticulous guy and I'll show you that in the following test:

I made a test:

7 laps on 60°C and 7 laps with 0°C with a Megane V6 on a very technical track to use more the tyres (track is Mores short). on both the realroad was set on medium. No drift: clean driving style, not forcing it. Sunny weather on both attempts. Stock car setup on both conditions. On both humidity was set on 30%
The results?

on the 3°C: Engine temperature was max 73°C , the average temperature of the tyres was 59°C and my average laptime was 0:50:800

On the 60°C: engine temperature was max 118 °C, the average temperature of the tyres was 66 °C and my average laptime was 0:49:900

According to that:

- YES, unlike you said, ambience temperature has an influence on the tyres temperature.
- YES, unlike you said, with hotter tyres you get better performances.
- YES, unlike you said, temperature has an influence on the performances.

And you know what? because I want to give you the benefit of the doubt just because I don't want to sound rude or ignorant, I REPEATED the same test one mroe time! Similiar results (obviously I can't repeat exactly the same laptimes), but same conclusions. but... just to let you know, I made the same test with a Tatus for my video to give you my thoughts about the game. I don't have time tonight to test also (one more time after the review) wind and humidity. Same for the tyres pressure. But when I'll have time, I will. But what makes me angry, it isn't your toughts about the technical parts of the game, because I agree with the fact you could not agree on what I said on it. What make me angry is you and your friends accuse me of superficiality despite I PROVE my toughts with actual tests, spending energies and efforts to give an accurate review to the people who are watching me. I make my conclusion with *my several tests*, which I try my best to have the most objective and accurate results as possible. As you can see, I put passion not just on what I'm doing but also at answering to the comments. Just to show you my honesty on what I'm saying, I could assume the tyres puncture isn't featured in game. But as I told you in the video, I never experenced it (because it's pretty rare and difficult to get a tyres puncture in all games). I just assumed it could have been present because the elements to simulate it are present (you can deflate a lot the tyres). that's it.

 

Yeah, I'm not sure how he tested it for it to work somehow or something? The guy is trying to cover up his mistakes in the video

 

Warning! Warning! Errant discussion/argument ahead!

I believe 99.6% of @Nieubermesch 's post was a quote. That could probably be made a little clearer.

 

Oh ok ok, dynamic temps just don't affect tires. Well, he still thought they were affecting it and probably it was just the driving as he is talking about an average. It should be higher...

 

Oh yeah yeah, I didn't even consider that he was taking it as mine, as I left a VP comments before.

 

One things is better physics and features, another thing is how they decide to make certain behaviors the way they want them to be, specially the tires, where I think Mantas is totally right. In ACC saving cars is weird as hell and things slow down. You still have to fight it to save the car and lose time and can't do the same things as you do in rF2 as you lose much more time, but in the other side of the coin they made grip dynamics of static/sliding friction pretty weird in specific areas of the slide.

 

As above I think you were replying to someone not here, but you've got me curious.

I know the files define an optimum temp, but there's no parameters defining the range or extent of performance gain/hit through temperature and I've never taken the time to try and measure any hardcoded effects. Does engine temperature really affect performance, outside of damaging and eventually killing it at high temperatures?

 

S397 cars aside, I am more interested in Dissenter at this moment. Hopefully S397 updates its GT3 fleet to improved tires build, but who knows... thats a lot of cars, and they already put lots of time in them...

It has been a while since I drove Dissenter more than for twenty seconds just to test whats going on with steering. I remember liking this car, but I also remember being floaty, and being compliant. But perhaps it is best not to rely on memory too much and drive it again lol... I also mostly drove it before I started working on my own cars, learning stuff , and becoming better at understanding and noticing things.... This being said, as I read your words "lacking at roll department" I instantly remembered all solid rear axle cars I have worked on (a lot of cars). And then I thought.... wait a minute, isn't Dissenter a solid axle car ? I should check it, I have probably been missing it all the time. I just briefly checked 70s Ford Maverick, and for sure it is:


So Dissenter should be as well. And if everything aligns my observations on all rF2 solid axle cars should apply to Dissenter as well. They do indeed lack severely in roll department, even with crazy weak suspension. Anyway... I should just test it. If it is then Dissenter has two of my "favorite" car bugs at the moment - mismatching steering, and restricting compliance solid rear axle.




In this random pic in google roll is observable, although isn't much. Looks slightly less than in real pic, but not much, I think real car in above pic is with rather stiff setup. Need to try it tomorrow. Anyway, I think solid axle cars really should roll less than cars with independent suspension and same suspension setup, but solid axle cars still should have proper amount of compliance. And of course finally it does come down to tires, and if they just slip off all the time, and don't hookup quickly, then of course it will feel floaty, and some will describe it as car feeling like it lacks mass, I guess. It doesn't actually mean that it is unrealistic, maybe it is, but maybe not. But there is mass, and it generates forces on a car, saying that car feels like it is an RC car is just a bit weird and not accurate, but perhaps the observation that something is lacking isn't unreasonable.

 

Interesting. I also have doubts it's anything other then tires characteristics of that era and also could be something with some suspension gemotry types, just as you have seem to find- I remember you had a thread about that but I didn't really read much of it. I would attribute some of those feelings of lightness to mostly some floatiness of the tires of that era and in rF2 they could be even more then they were... Can't know for sure.

Some cars onboards I see from time to time to compare to how suspension is apparently behaving is why I feel some cars suspensions feel a bit on the "stable" side, lacking a bit of impact or something. Can't quite explain it. Some other cars seem perfectly fine though, so that also is something of setup or even just the way the suspension components have been setted up.

I think it was a Hwoston too that I tried and the car felt very floaty and rotaty. At the same time for the type of floaty and slidy car it still felt the steering responses were too fast, and it kinda mirrors whay VP said. Also could be related to tires. Things can get so weird, even with a physical tire with real world proprieties they can be made into a strange mix of high grip and peakiness but still feel like a road car slip curve...