Automobilista 2 V1.5.5.0, Le Mans & Endurance Pack Pt1 RELEASED - Now Updated to V1.5.5.6

AMS2 is pretty good already and I think tyre model is not the predominant fault right now.
To me, there's something to be improved in inertia department, deep in the physics.

While testing my custom file, I often seek unusual situations (crash the car into walls or other cars, going over the grass and gravel, etc.) and, while many times things are ok, many times they're not (they're just wrong, not just imperfect).
It's harder to unsettle the cars compared to other sims (things like kerbs, small car jumps) and the crashes often miss something, like the resulting trajectory vectors of the two crashing cars is an impossible situation. Many times a perpendicular crash results in one car (at least) moving on as nothing happened (when I expected a huge rotational momentum should result from the impact).
I can feel much more tyre grip variations through FFB than physical impacts in car trajectory or stability, so tyre model really captures those grip details but they don't propagate to car chassis very well (too much inertia!).

Also, AI cars behaviour also make things look 'wrong', again inertia...
The way Ai catches spins or rear end loss of grip (driver errors or crashs/bumps) should be revised (too little inertia! And too fast driver's reation time, leading to pendulum motion).
Because the crashs/bumps are too forgiving, I don't "respect" AI cars the way I do in other sims and that also contributes to more (unrealist) aggressive moves from me.

Given the progresion we've seen over the time, I really expect Reiza to solve all those issues. It's just a matter of time .

 

The car was already unsettled before going into the corner and sliding imo and he kept hitting the throttle. He already had the wrong line in the first kink I think and was also going to fast. And even after that he hit the bumps wrong and gone he was.

 

it’s no drama, just trying to work out what you mean , as there is no spin . just a very unsettled car that has its weight shifting the wrong way to make a left turn .

I’ve encountered many times that coming off a curb unsettled the car to a point that i had to bail on trying to turn and just had to go straight to try to allow the car to settle or i would definitely spin the opposite why that i wanted to go initially.

I often admire the replay of my crashes as they behave much as i would expect and often look very similar to what i’ve seen on tv from a similar camera angle.

 

Yes, telemetry would be a better deciding factor, but who has access to real world telemetry for most of these car / track combos.. and even then, setup differences. So, as a result, I think these "discussions" will go on forever without resolution and as previously stated, it can be part of the fun (until people get bent out of shape because such discussions are even taking place)

BTW, I checked out some videos regarding the Porsche RSR '74 and it does indeed seem pretty rock solid under heavy braking and initial liftoff.

 

I know you are saying this as a bit of a joke but this is how I have been testing the latest tire changes (and prior) with the GT3's and you can tell its miles better with this apparent "minor" tweak.

 

I won't lie, and I witnessed it, a Cadillac did something similar (almost hit the wall) at Le Mans and I almost hit it xD And I found it spectacular.

 

There are, obviously... it happens. A far more interesting statistic would be laps successfully completed vs the amount of spins not caused by contact, wet weather, obvious driving mistakes or completely fumbling something like curbs and throwing the car completely off balance, etc. etc. - which obviously is pretty impossible to obtain but it'd be interesting.

In general you do not see GT3s spinning or flying off the track left and right into walls during regular running under dry conditions. You will see mistakes, excursions into run-off areas and consequently completely messed up laps without major disaster far more often.

This has nothing to do with how well or not well AMS 2 simulates anything btw, I'm just curious where this obsession with "I should fear losing the car at any moment" comes from, since it is the one thing that is very much not desired in real life, teams/drivers will go out of their way to avoid by any means, since it is very expensive and can be quite hazardous to health also. However it appears much craved by some that race cars need to have murderous intentions at all times even with a relatively safe setup and not driving completely on the edge.

Also no need to stop, it's a forum: people will discuss things, human beings being human beings... nobody will agree on anything anytime soon. It's a known "issue" that no patch or physics adjustments can fix

 

No joke/not joking. I wouldn't use TC or ABS to test any race car (of course not the generic aids ever nor even the authentic ones).

The TC and ABS on GT3 cars are there primarily for wet weather or unusual conditions (something spilled or dirt on the track). Everyone makes snide comments, but I think even the pay-for-a-seat drivers are competent enough to drive a relatively tame race car without needing to rely on TC or ABS on a dry track. We should be the same. Once you know how the car behaves raw, then turning on the assistance should allow us to see exactly where and how it is utilized. Then, when it saves our butts, or does not, depending on where we set it, we will know why.

The TC and ABS are dials... meant to be turned up and down as required. Of course AMS 2 has only one default set-up, so some middle/compromise setting is chosen. First thing I would do in the car after confirming it is a nice, dry day is to turn down both of these at least two notches and maybe more. First thing I would do if I saw a wet track was do the opposite--turn them up a couple of notches.

 

Maybe everything should be reversed. Have you ever thought that all the mistakes come from you? Maybe the user is the source of the error, not the game or physics. With that in mind, have fun continuing your search

 

Do you have any sources to back this up? I think both Daniel Morad and Nicki Thiim* explained how GT3 drivers rely on TC and ABS in real life. Also, in the ACC setup menu, the TC and ABS charts show how one should dial in their aids according to the track conditions, wet and other low grip conditions being only at the extreme end of the spectrum.

Don't want to nitpick, I am really interested in your statement if it is supported by facts.

Cheers!

*Edit: I'm still searching for at least one video to support my claim. I'll update my post when I find it.

 

This!!
I'll add: those discussing are the most interested people in AMS2 and it's possible improvements, not the other way around (those out of this forum are the true AMS2 lovers? I don't think so...).
Reiza (and AMS2 users) should not see criticism as a bad thing, and if something is often pointed out... perhaps that's something deserving been checked. That's the nature of evolution.

 

Yeah, your right about that, i have seen/heard the same from several drivers, that they always use a bit of TC and ABS during a race and adjudt according to the track/weather situation.

 

For GT3's I've always heard that TC/ABS is on, and drivers adjust it depending on the conditions, but it depends on the car, for example here's a clip from Moritz about the LMP3 in iracing:

So my logic is that if an LMP3 can have his TC off, why wouldn't a GT3 benefit aswell? My guess? it depends on the car, not just the class. Maybe with a more pointy and aggresive 911 GT3 you need TC but with a more tamed and balanced M4 you don't? at the end of the day, thats one of the least important things IMO.. ACC pros now run without TC on qualy after the 1.9 update since it's actually slower, and that sim does only GT3's

 

On virtual racing.org (Rig Vorstellung - Peter Winkler) Peter, who has IRL track experience and profound sim racing know how, posts regularly interesting insights about both worlds. Even serious IRL racers use e.g. AC to prep and train for real races.

The boundaries between Consumer Sims and Pro Sims are quite blurred. (the difference lies in the manual adjustment of vehicle-specific details).

Ultimately, real driving is "different" from simulated driving, and vice versa, and yet simulations are so good that they are used for training. And of course, if you know the behavior of the real vehicle, the sim can be adjusted astonishingly well to match it. So I think it's possible to speak of great realism according to the motto "Same, same but different". Either way, the many different changing external factors in both worlds always make a direct comparison difficult.

I can only say that the first time I drove a KTM Xbow with 550 hp (in race trim) on the Panoniaring, for which I had previously practiced in AC (in VR), I was amazed at how well the experience from the sim could be transferred to the real track. (Except for the sudden brutal snappiness of the real car, which is surprisingly easy to drive 5-6% below the best lap time limit. And this kind of snappiness can hardly be avoided even with traction control, which this car does not have). And above all, that it was possible to adjust the sim after the trackday so that it could "simulate" the IRL experience even better.

One thing was particularly interesting for me with regard to "same, same but different": In the real car I got on pace (for the fastest lap times possible for me) just after a few laps, while in the Sim I needed many more laps to reach such lap times (given that I matched the sim's tire grip characteristic to the real car). G-forces and the real Seat of Pants feedback help a lot... I'm sure that it can also be the other way around individually, but I understand the often-heard argument that real race cars are easier to drive than in sims. However, if you try to drive at the (personal or absolute) limit, I think - especially then - the difference between sim and reality becomes smaller and smaller - in both cases it becomes damn difficult. In this tunnel, our brain doesn't care any longer about the distinction between sim and reality .

I'm just amazed at what is possible these days. From a technical/coding perspective it's a miracle to me how far we got.

 

Nice find. He also mentions the iRacing peculiarity that tyres are very susceptible to flash overheating, which reduces their grip for several corners. rFactor 2 has something similar, you can easily overheat your tyres to 300+ C in a single corner simply by sliding a bit (idk how that is possible given the melting point of rubber, it should lead to a blowout at that point) and then you start skating around. Since we can't see the surface temps in AMS2 it's hard to compare directly, but it's possible that this behaviour is what creates the feeling that the two aforementioned games are harder to drive than AMS2.

 

I think there are several temperatures, not visible on the HUD, available in the telemetry data... depends how interested in looking into it you are but you could do some experiments with an appropriate app for getting the telemetry.

 

All I can see there is the "Tire Temperature" (equal to what appears to be a core temperature) and "Tire Carcass Temperature", which is close to 400 C(?) on GT3 cars for example.

 

Imho, it doesn't have to be mechnaical failure, all it takes is hit bump hard mid-rotation. The car was rotating as he hit the bumb, so, as per 1st newton law, it continued rotating while the rear was unloaded. After it landed, the car has already rotated too much, so he was out of options and basically slammed the brakes.

Also, while racing tyres have more grip than some sims would like us believe, they do have their limits. Especially laterally loaded low profile tyres with quite a stiff sidewall.

I've had similar crashes in AMS 2 LMDh, at that spot. When the rear goes, it goes in AMS 2 LMDh, especially with low-downforce setup.

 

In the SharedMemory setup presented in ".../Automobilista 2/Support/SharedMemory/AMS2_SharedMemoryExampleApp/SharedMemory.h" there are:

Code:

float mTyreTemp[TYRE_MAX];
...
float mTyreTreadTemp[TYRE_MAX];
float mTyreLayerTemp[TYRE_MAX];
float mTyreCarcassTemp[TYRE_MAX];
float mTyreRimTemp[TYRE_MAX];
float mTyreInternalAirTemp[TYRE_MAX];

I hope one of these is relevant to what you're looking for; unfortunately there are no comments i can see in the code explaining what each of these are precisely, although the names seem clear enough. TYRE_MAX is a constant that looks to be set to 4 given it's the 5th element of a C enum... so I guess no Tyrrell 6-wheeler for now

 

Speaking only for myself, being actually 'on the limit' is relatively rare, and I notice when it happens lol. In the Indycar 2023 I had a couple lovely moments in mid/low speed corners where I could feel I was on the proverbial 'knife-edge' wiggling to keep the grip, and it was cool and scary at the same time

Mario Andretti I think who said if you're not feeling out of control you're not going fast enough?..