Physics discussion thread

Initial impression of the latest patch;

Definitely an improvement, many cars doesn't feel as "cheesy" while maintaining forgiving nature and effortless corner entry & rotation which are trademarks of AMS2.

Though i can still get away hazardous inputs with the steering wheel without unbalancing and upsetting the car.
And honestly why would i even complain about that anymore.
As a busy familyman it's nice that atleast the sim car is compliant and forgiving when i fire up the rig.

 

Really curious to hear about those GT3 changes! I’m out of town for a couple days and can’t try it !

 

Hello, I am a die-hard-LMU-fan. Let me put it this way: I never touched AMS2 because I didn't like the handling/physics. Yesterday, after the update, I fired it up quickly to test it out and haven't stopped since (except to sleep).

The GT3 Gen2 (with a few changes to the setup) now reminds me a little of the very first physics in LMU (right after release). Detailed, relatively forgiving, still more slip angle than it should be, but overall incredibly fun to drive.
I'm completely rediscovering AMS2 now.

 

I would be interesting to see members cite their negative impressions about any and all handling issues with examples and if possible, data. (apply that rationale to any sim.) I try to enjoy every sim I have in my library, despite the issues that lead to a "lack of immersion or sim-believability. All of this stuff is fake. Well-made auto mechanics with beyond most people's understanding of the mathematics used to make these one's and zero's appear on our monitor and work with the hardware used to simulate (not replicate, please always remember that) auto racing. We need to trick our brains into "believing" the visuals, auditory cues, wheel-response, pedal response, and anything else not mentioned; that we are simulating the feeling of auto racing. Most Sim Devs may not have all the data needed to create this illusion, so they have to discover some kind of secret sauce to assist the lack of data not provided by real racing teams and tire manufacturers.
I would like to think every sim racer understands this. However, mashing the keyboard with parroted-statements without defined-examples is not helpful for any sim developer.
To those who feel that is their prerogative to do so, ask yourself if you are just frustrated that you want the sim to improve, or just want to be heard in this social media generation. But, if you want the sim to improve, I'd hope one could start using examples of what you are experiencing and better yet, prove that data (like Steel had mentioned above) can be used to understand your subjective, opinion. We all have them.
My suggestion has been put into action in terms of a member creating a program to help analyze what AMS2 is doing. Perhaps you should look into the really interesting work Matus Celko has created; Second Monitor. Second Monitor - Timing/Telemetry App Others have created the ability to have been able to get a certain amount of AMS2 data into Motec. (vid cited by UniSimRacing)

Being realistic, things probably won't change. "Complain, instead of explain" is the most simplistic. I have my own complaints about this sim, as does everyone who considers AMS2 to be their go-to sim.
Everyone wants AMS2 to be tailored to their personal, liking. That will never happen. It's the same for all sims. At the end of the day, I hope the immersion and enjoyment experienced in AMS2 (as well, all sims) will out-weigh the "complaints." If you agree, then please cite specifics and examples with your one-word complaints. An observation. Please correct (Devs) if I'm wrong. It seems default dates correlate to the year the DLC tracks were released, as well to specific track events. i.e. Road Atlanta. I'm not at my race rig, but I recall October, 12th, 2024. Live Track gives the ambient and track temp, of course.
If one is to test physics and tire handling, would it be best to provide Devs the optimal time of the season per-track, along with affirming the track rubbering setting details and car set-up information (be it default or edited) information to go along with explanations of corner speed and braking mark distance, etc...?
I do bring this questioning up, because if one is honest, there is potential for everyone here (self, included) to pillage/maximize the limits of the physics engine. Do things that you can get away with in a sim car, and not attempt in a real race car, including the braking later than the best professional race car driver in the world; including entry speed into specific corners, with less than favorable wear of tires. I try to repeat this when I sim race;
"Simulate" not "Replicate."
50 cents, spent...

 

Guys, the slip angle thing is just my personal opinion, backed up by a few statements from the web, compared to other sims, but still purely subjective. If you read my post, you should have noticed that I spoke very positively about AMS and said that I would enjoy it a lot.
I didn't expect that it would still trigger everyone.

 

This is very reassuring to read. There is a very strong body of opinion in many discussions that AMS2 should be more like X or Y sim, making assumptions that they are somehow more realistic, which always makes me nervous. Surely the only comparison is with reality and that is what to aim for. Some Youtubers especially are very bad for this, to the point where I find myself screaming at the screen 'Have you ever driven an actual car?'. Glad to hear Reiza are still focused on reality rather than the opinions of bedroom racers.

 

I dont' know do you have access to be better data, but at least I had hard time finding information what kind of slip angles something like GT3 tires allow (from different manufacturers) when I tried while back.

In any case, IMO it's quite obvious that the tires modelled in GT3 gen2 (haven't tried the gen1 in a while) feels quite generous and the tires don't feel very peaky at all (seems the apparent peakiness earlier was because of other reasons, which the last few updates have changed). The "window" where it feels it feels the tires feel/seem being slip angle beneficial way for rotation seems quite easy to achieve and maintain. It might not be the maximum optimal angle, but at least there is not very sharp peak if you go past such optimal angle.

Which might be correct, like many have stated, IRL tires from different manufacturers can be quite different what they allow and behave as well.

Whether is correct or not for these cars/tires, there are plenty of people who find it more rewarding when it's tighter are requires more skill and also exiting for those who like "living on the edge" with danger of losing it (without overdriving the car too much). A bit like if you have some action game with a lot of frames to pull of a dodge/move succesfully feels less rewarding vs. one that has tighter window (as long it's not too unfair and too frustrating).

It's a simulation though, so it should aim for whatever is being simulated. With tires though, developers can occasionally have a choice which brand / model they want to go for..

EDIT: this isn't a critique, just observation and thoughts. I do think the GT3's have improved and now drive more like many expect them to.

 

I think it doesn't even need to be telemetry, but I would love to just see recorded ingame footage that shows the slip angle and their laptime since if you're not at the limit of car it might react different than when it carries the maximum laternal forces it can use in a corner.

And again I can only point out that especially modern GT3s are not the 100% planted racecars some people think they are - most likely because other sims and bigger content creators claim they are or are just getting used to how certain sims handle it. Here's a clip from the recent Battle on the bricks. I marked the 2:42 timestamp, people should just watch closely how the body of #81 Dragonspeed Ferrari moves, in the next scene watch the Lexus closely. These cars are supposed to have a certain slip that the drivers constantly work with.

Here's another one where it's even more obvious in the Pfaff Lambo

The key element to slip angle is not if it's allowed but if slows you down. Slip should always be slower than getting through a corner clean and that's exactly what I love about AMS2s physics. It allows a certain amount of slip to work with, but you are actually faster going through corners clean. And that's the point why I don't enjoy other GT3 sims as much. Having these insanely planted GT3s that can sudden snap oversteer does not feel realistic. It feels to me like other sims adjusted their physics to have this insane snap oversteer since otherwise slip angle would just be faster than driving clean. If they would crash out on the slightest overdriving nobody could race a 6, 10 or even 24 hour race in a GT3 car which is done all over the planet

So TLDR, I would just love to see more footage where AMS2 allows GT3s to be driven very fast while maintaining slip angle that is often brought up. Especially with 1.6.7 and now 1.6.8 I can't relate to any of this from my experience and just want to see how anybody still can use slip to an abnormal amount here

 

Whatever Reiza did in the latest update is fine for me. I think it feels good in many ways and they should probably just leave it there so we can get used to it.

 

I think i can now word more precisely my issue with the physics, slip angle and the "slideyness".

It is most noticeable when you're turning the car using steering inputs alone, without touching the pedals at all.
Turns out, with aggressive enough steering inputs you can initiate little slip angle not only to the front axle, but also to the rear axle.
It feels like my steering inputs doesn't turn only the front axle, but the whole chassis.

Or am i completely wrong?
Can you actually get the rear loosen up with just steering inputs in a real GT car?

And sorry for everyone who wants a video.
If you want to replicate it, i used the AMG GT3 on the Nordschleife under standard practice conditions, with the standard setup but less fuel.

 

I think I fully agree with you. For me it feels like this: when I start loosing rear end, it feels like I am loosing it in slow motion. And at that moment I also countersteer with my steering wheel at steady relaxed speed. When you look at IRL onboards they do countersteering a bit more sharply. James Baldwin in his videos mentions tires being peaky in GTWC.

To be specific, I am not talking about if and when rear is loosing grip. But when it does, in my personal opinion, it could have sharper drop

 

I never realised they had tyre temperature readouts on the dash, I just thought that was a simracing cheat assist thing because we couldn't feel the car the same as in real life.

 

Yes, I notice forgiving handling like being able to go through Schumacher 's' corner on nurburgring with the gt3s flat out with no need to lift off accelerator . I'm assuming this can't be the case in reality.

 

I notice a significant reduction in torque. Can I assume that the polar moment of inertia in the vertical axis has been reduced? Or to put it another way: The car's mass is now closer to the rotation axis. Does this also mean that the car's mass distribution has changed? Honestly, it would have been easier if Reiza hadn't released the physics update and FFB update at the same time. This makes it impossible to distinguish between the two.

 

About the “bite” feeling... I’m curious how PMR will represent that, as it’s one of the main claims about their tyre model.

In AMS2, I can easily feel the car’s rotation caused by a slipping rear, but the rotation from a biting front end is much less evident.

An obvious observation: on a wide track (like Cleveland or Galeão), if you’re at a stable high speed and sharply turn the wheel in one direction, the car tends to go straight, with strange oscillations that eventually bring it back to the straight line — but there’s no sign of front-tyre-induced rotation. Other sims handle this much better, to the point that I’d say the real car probably behaves closer to those. In AMS2, though, I can confidently say a real car wouldn’t act like that - after an initial "bite" (that would make the car turn sharply), there should always be some “residual” directional friction making the car keep turning, even if it’s in a strong understeer manner.

Yep, the “bite” feeling. These are excellent drivers, they make it look easy and planted, but when pushing hard, the car’s “unstable” behavior becomes obvious.

Yes, the slow-motion, filtered/auto-correct AMS2 behavior. My main concern is whether it’s like this by design or if it’s still a pain point... I mean, Reiza might have reasons for leaving it that way to please the current, more relaxed and less critical user base.

PS – my car uses UHP tyres (so, more like mid-to-high-end road tyres, still far from slicks), and the bite is quite obvious in the way the car’s moment of inertia changes, leading to sharp ESP interventions (without these, the car would go wild in a wide number of ways...)

 

Which car and did you change the setup from default? When I was testing these (all GT3 gen2), on that track primarily, I had to lift just a bit to avoid going wide and invalidating the lap.

Baldwin's best lap from IRL below, he lifts at the start fo the Schumacher esses, probably to avoid going wide on the right hander (which where I went wide if I didn't lift or lift enough).

You probably can change the setup to do it flat out, but not sure is the setup then good for the whole lap (there ain't that many faster corners like that on the GP layout).

Not saying it's absolutely correct on AMS2, but I don't think that example is that far off either from reality.

 

I can‘t take Schumacher S without lifting, too, even after the update.
Maybe you can do, but how big is the effect on the tires? Will they heat up and will be ruined for the rest of the lap? Especially when you enter Nordschleife?