Physics discussion thread

Unfortunately, this test isn't very good if you look closely.
Unlike in my test where i copied the actual real life wet line, in this test the "wet line" is actually just the dry line for the most part. Let's give it a closer look:
(to avoid confusion: in the following i will use "quatation marks" when talking about the "dry / wet line" in the video and no quation marks when talking about the actual dry / wet line)

-Braking into t1 was different
-Through t1, t2 and t3 the lines were identical
-Braking into t4 different
-Line in t4 identical
-Braking t5 identical
-Line in t5 different
-Braking and line in t7, t8 identical
-Line in t9 identical exept that the car ran wider on the "wet line" so the "wet line" was actually more of a dry line than the "dry line"
-The lines in t10 left me confused because they ran wide and avoided the rubbered line on the "dry line" but kept it tight and on the rubber on the "wet line". So it is the wrong way around. The "wet line" was the dry line and the "dry line" was the wet line
-In t12 the lines were pretty identical exept that, again, they ran wider on the "wet line" than on "dry line". Once again taking a more dry line approach on the "wet line" than on the "dry line"
-In t13 it was basically the same as t12. Lines very similar exept that on the "wet line" they actually hit the apex more. Once again, taking more of a dry line than in the "dry line"
-And finally in t14, the lines were different.

Overall, you can see that the test did a pretty bad job at taking the intended line with the "wet line" being more of a dry line than the "dry line" in t9, t10, t12 and t13.


So where does the difference in lap time come from?

Firstly, you can see on the dasboard that in the "dry line" the tires are not up to pressure and temparature yet. But in the "wet line" they are in the perfect window. JUST THIS ALONE makes the entire test completly worthless. The conditions are clearly not identical for both laps. Tire pressure makes a huge difference in ams2. Also the fuel was higher on the "dry line" than on the wet line. Presumably they just set one "dry line" lap with cold and underpressured tires, then switched to the "wet line" once the tires were in the window.

Secondly, it immediately becomes quite clear that the "wet line" just simply extracted more lap time by pushing harder and simply driving better.
Just taking a look at t9. Despite taking a similat line, the minimum speed on the "wet line" was 150 and on the "dry line" it was 142, simply pushing harder on the "wet line". Same thing in t12. Similar line but 93 min speed on the "wet line" and 87 min speed on the "dry line". Again the same thing in t13: same line but doing 161 on the "wet line" and 153 on the dry line.

So in conclusion the test:
-Didn't even take the wet line
-Didn't have identical conditions
-Simply drove bad on the "dry line"

Can i proof that the "dry line" was a really bad lap?
Absolutely.
I just jumped in and did a lap in the LMP2 G2 in wet conditions with heavy rubber preset using the dry line, completely ignoring the rubber and even with a very messy lap, i instantly beat both the "dry line" and the "wet line" lap time in the video.

With a bit of practice you could easily do a 1:45 in these conditions. Since the lap times in the test are not competitive, it doesn't say much anyways.

In my original test i put in lots of attention to detail to ensure it is a fair as possible (such as waiting in the pits with a stop watch to ensure the rain falling onto the track in both sessions is identical). You'll also find that, unlike in this test, the lap times in my test were actually competitive.

Unfortunately, this test wasn't representative and as you could see, you can go much quicker on the dry line.
Really hoping the rubbered line will become much more slippery in the future, so that the wet line will actually be faster. For the moment the dry line in the wet is still the way to go.

 

Completely agree.
I just have a doubt,since in the past the weather progression and water accumulation was a bit "unclear".
I don't have the time to test it today,but what if in this condition there is just "not enough water" on the track yet (i am talking physically, not visually)?
I want yo run the same test,just setting as you did real time and 1 weather slot.
And then trying to set 2 identical time slots and setting weather acceleration to x60....just to see if there is any differences

 

Not saying I disagree with a lot of this what you wrote, but couple of points.

1) Kerbs don't get slippery on wet on AMS2, this is known limitation currently and will surely impact what lines you want to take and where. I suspect the driver was using the wet line in place it felt it gave better grip and performance and dry line in places where it made sense with the available grip, kerbs and handling characteristics of the car otherwise.

2) It's a lot easier to push if there is more grip, so saying "didn't push as much" doesn't mean much, you cannot know where and how the driver felt the grip limit.

In any case, I suspect you are right that there is not enough difference for it to be impactful enough compared to how it can be IRL. Like I wrote already earlier, sometimes it's the AI calibration challenges that force them to be more conservative on things like this. Other thing where I similar style of issues (being conservative) is tire wear and it's impact.

Anyway, I appreciate the effort and detail from you on this subject. iRacing from what I have seen, is clearly a benchmark on this wet weather stuff currently, but at least on AMS2 there is a lot of dynamics available (and it's not like they didn't spend years on iRacing to make it).

 

While i really do appreciate efforts to improve the AI, what caught my eye how natural the car's behaviour looked.
For example at 2:45 the car has at first slight understeer at corner entry, but it eases of gradually as speed decelerates.

 

For me, RaceRoom is too much on the understeery side, and I feel this is the way that the latest AMS2 Beta has gone. I just don't get along with understeer, at all. But too much oversteer can be almost as bad. But it is all preferential, right?

 

I made the Porsche Cup and the GT4 from awful floating on the rear to very believable and natural to drive with 5 clicks on the diff setup page.

I really don't understand why the default setups have so much diff locking on power, which creates serious unnatural behavior out of corners. It really is solved with a few clicks. Less clutch plates and more preload.
I couldn't stand AMS2 behavior anymore and this really solved it for me. I couldn't stop doing laps.

Just my 2 cents.

 

It's almost impossible and very time consuming to explain the chaises behavior in AMS2. I will be happy to share my personal experience with the developers if I ever get a chance even if it takes hours of my time because I care about this sim. I believe AMS2 could be the Sim of the decade.

Grab a box of matches and put it on the table. Now move it left and right without steering the front. A race car or any four wheelers will need steering to achieve that movement and that's the weird part in AMS2. That disconnection of the left & right movement of the chassis and I give you an example.

Grab a Porsche GT3 or whatever, any LMDh (low/high DF cars don't matter for this experiment). Now head to Nürburgring. Your tires will be at proper temperature by the time you get to Brunnchen corner. Exit that corner and you'll be pushing 98% of the time full throttle through Pflarzgarten. I had the privilege to be in Sabine Schmitz's M5 in 2010 and saw her rip through that part of the track full throttle while talking to me and my stones fell on the floor and no, I don't remember a thing she told me so I felt that part of the ring with every cell of my body when a legend and a legendary car was in charge of that force. The car was not jumping left and right w/o your command. That M5 (close to factory specs with good tires for the time) was listening and reacting to your input and demand. Not just jumping left & right as if you're flying a helicopter.

You don't get that in AMS2 and I tell you why. Let's go back to our box of matches on your table. It goes left and right in a nasty way...you're not connected and that movement is unrealistic and unqualified for a race car. You're just rolling and surviving the madness and it doesn't feel good! You're basically flying with a few centimeters above the tarmac.

Again, not bragging here about my experience and I want to help the title to shine, but I've owned race cars in the past. BMW 2002 Turbo, Datsun 260Z fully race qualified in the 90's all the way to Miatas, BMW E46 M3 now M4 and I've raced with my friends rides including Porsche and Corvettes.

There's a major issue in AMS2 when it comes to tire connection and cabin behavior in my view. I've no issues with understeer/oversteer and the setup could fix some of that in most cases and I believe they killed it with updating the FFB. Currently the Default FFB shines, but not the phisycs unfortunately.

Sorry for the long post

 

I find this topic so amusing. Look, AMS2 is slidey? Yes, any other view is huffing copium, but..it's totally fine if YOU enjoy it and have managed to drive within that envelope. No I do not need to be a GT3 or F1 race driver to know it is slidey, being a normal driver at 45 years old and driving every day for 28 years is enough to know. Why? because AMS2 doesn't just make cars rotate in circles at high speed, it does it at low speed as well. I have had some very silly low speed slides in AMS2.. In RaceRoom cars do not do this, they actually do slide though, actually pushed it too far yesterday in RaceRoom and did a 360. But I could feel that it was a real 360, it actually felt like, "Yeah I stuffed up there" ... we are allowed to have our opinions, and we don't all have to prove them. But whatever the fundamental vectors are in the equations that AMS2 uses in order to calculate these slip angles is not the same underlying fundamental formula that other sims are using, there should be in theory a scaling factor or dial in the physics equation they can turn down to increase grip and also to have a more realistic physics slip and stopping time after losing grip. Because the other issue I feel is that after loss of grip in AMS2 the cars slide for DAYS ! ...

(Awaits attacks)

That being said, I find that super high horsepower and high downforce combination, like an F1 car, it is not possible for me to comment on, since these cars actually have good aerodynamic grip in AMS2. And I would assume that in real life these cars have terrible grip at low speed. And so whilst I think the tin top low downforce cars seem super slidey.., It is hard to say if the F1 cars are as a person without any real life F1 driving experience, because they are the easiest cars for me to drive in AMS2 and they shouldn't be. They should be the hardest. Hence, aero grip in AMS2 seems fine to me.

 

That being said, wouldn't it be great is we had RaceRoom and AMS2 combine!!! ..., RR guys could really learn about graphics and day night cycles, modding openness, CPU and GPU efficiency..!! from AMS2 guys, seriously RR is really last gen game now.. They keep saying they will update it, but all we see is more cars and tracks made to sell ... so I am getting pretty over the fact they keep charging new money from existing players without polishing the game up to standards of today.. If we had the handling and physics of RR in AMS2, I would never play RR again.

 

I think I've finally figured out why AMS2 feels "normal" to me. Here's a description of the first car I owned @Nuck Chorris was totally normal to have low speed spins in the old Holden VN Commodore

You'd have to baby them through roundabouts in the wet or they were guaranteed to spin/fishtail. Mine was an '89 model and it was great fun, and between it and a FWD car with 4 of the worst tyres known to man being the first two cars I drove a lot, I became quite a safe, controlled driver

 

Well I had a VN Commodore, and VR Commodore, and I changed pretty much every part on those cars myself in the years that I had them, so I actually do know a lot about these cars, and yep..they certainly handled like a wet sponge at times and you could slip on a roundabout (if it had just started raining on a summers day), and you better park them with the rear wheels down the hill or good luck reversing up a hill with no traction, ....but the slides in AMS2 are next level, we aren't driving 3.8L V6 ****box commodore executives here though...lol.. we are driving cars which should have a lot more grip. lol,

 

I don't honestly know what you mean by this. I even spent over an hour just now to test what you asked (used the McLaren GT2 gen2) and I don't feel the car jumping left and right on that section or it feeling like it's flying over a tarmac. The road is not flat, so there is some little bit of movement that isn't from my inputs directly, but nothing that to me would feel out of the ordinary. I'm sure a lot of that could be tuned with car suspension setup changes

Then I tried to watch replay how it looks (as feeling is so subjective), from different angles too and to confirm if the moment is somewhat in line with another sim, I did similar for rF2 and to my eye, there wasn't a huge amount of difference on how it looks on the replay. Not sure how much you can rely on the replay showing things correctly though, best would be to analysize telemetry data and compare it to real life data.

Third thing I tried, was to do the same on the GT1 2005 Lambo that feels really good to drive on AMS2 and nothing weird on that either. On replay the main difference I noticed to the McLaren GT3, was that the front of the car was bouncing quite bit more compared to the McLaren

People seem to have very different experiences and perceptions on how correct the sim feels to each other, especially with AMS2, but it's the same with other sims too. That also applies to those who do IRL racing.

Not saing that everything is correct, but it's annoying to read "physics are bad" style of comments that don't go into detail and try to prove it or even explain exactly what in them is wrong and more helpful to work with maybe a single car or car class as example case. (not saying this comment I'm quoting is such exactly, but there are plenty of those that are).

ps. not sure why you and some others insist on continuing to write these things on this thread, which is more likely missed by the devs instead the dedicated physics thread Physics discussion thread.

 

Somehow my post from FFB was moved here and I'm glad it did. Maybe there are some eyes on this page that could make a difference in the future. I've never digged into this forum to see what else is out there and I initially came here to experiment some custom FFB. Anyways, happy to be here and this forum has helped me to understand sim racing better.

First of all, I'd like to share my experience about Raceroom and I know many won't agree with me as I see some friends here are having fun with it.

Raceroom is way easy to drive and it makes you feel over confident in many scenarios. It's a good thing for a casual racer and helps people to stay on the track, finish races and have fun. Maybe that's why many like playing it after the latest update and what you get for spending ~$30. Also, the graphics look very cartoonish and outdated. Those sparks from the exhaust pipe took me back to Sega Genesis days and the overall feel of realisim in terms of look and immersion is nowhere close to AMS2. I don't want to get into details because that title didn't last long on my PC. I was running it maxed out on graphics all around btw.

There are currently three racing sims on my PC. AMS2 (which I support and own all the DLCs), Assetto Corsa Evo and Dirt Rally 2.

Evo is fun and I see happy days ahead for that title. But I've never had more login hours in the gaming world than I do in AMS2. That's why I personally set my expectations very high and I demand perfection from that title in my mind because it has that potential.

I'm 100% sure that I'm not having a hardware issue because I own three bases from my favorite T300 belt system to the underrated Thrustmaster T818 (and when they said unfiltered FFB, they really meant it) and finally Simagic Evo. They're not the best bases on the planet by any means, but the variety is there. Also, I've spent months to study input lag and how to eliminate it and I've achieved that goal beautifully. At least that's what my tested reliable softwares tell me.

You can't rely on watching a replay to decide if the phisycs are right or wrong. It's all about the feeling in the cockpit view. You keep asking for proof and if you think AMS2 is exactly where you want it to be; then stick to to that belief, enjoy it and have a blast because that's the only reason we are sim racing. It doesn't matter what everyone else is saying. If you find that joy in Raceroom or any other sims, who cares what others think about your experience? But at the same time, there are always people that have some experiences that you and I never had the opportunity to be a part of. That goes for anything in life and combining those experiences and feedbacks could solve problems and bring greater experience in life for everyone.

Do I think AMS2 feels great at the moment when it comes to phisycs? Absolutely not. Have I shared my real life experience many times before? Yes, I have. Is it the best looking title in VR, triple and single screen? Hell yea it is. Do we have the best cars in the market? Yes, we do. Does it run beautifully on most PCs? Yes, I don't think if there's any other sim that could beat AMS2 in that regard. Is the crash model perfect? Yes, it is. How about the Audio? Besides the redicilous curve rumble effect and engine cutoff sound that feels like you have a drummer on the backseat; it's perfect. Do I dig the day/night cycle and that gorgeous looking fog and sun rays in the game? I can't live without them.

All that combined will increase some of us expectation for Reiza and AMS2 to step up and solve a few issues we have.

We have a hidden gem in AMS2 that could be used as a handling model to improve all other cars/classes and that's our Formula cars. They distroy EA F1 titles in term of handling, tire management, DRS behavior and how effective the downforce setup could be. It's unbelievably good driving those cars. What went wrong with the other classes? I dunno

The goal is not to make AMS2 easy to drive. That'd be a disappointing move. Racing has never been an easy sport and if any of us want to be a part of easy silly stuff, we should jump on Gran Turismo and call it a day.

 

@Peykan I have to say I did a couple of laps in the spot you mentioned and maybe it's just I still don't quite understand what you mean but I didn't feel what you're talking about either

In layman's terms are you trying to say that the car skips laterally/sideways across the track when it shouldn't? Unfortunately I've never understood the 'helicopter' analogy. Comparing to IRL driving for me is difficult in the sense that I can't feel it with my body so I can't tell if it's "weird" due to having to feel it through the wheel instead. It would be interesting to get the opinions of someone with a good motion rig.

 

I'm glad you don't feel like being thrown around on the track at high speed S corners and the car not following your mind and input from your feet to your hands. I wish I could claim the same.

Unless you use the motion rig for Rally and Truck Sims; I don't recommend it to anyone who's in it for lap times and consistency specially for long sessions and endurance races. They can make you slower actually. I've been there, done that and sold mine long ago. In addition, a motion rig will not help you to understand and analyze the game engine phisycs

 

This has nothing to do with motion. You won't feel real G-forces even with motion. Motion is for immersion, and it's cool, especially when you use it in rally racing. And if you combine it with VR, you have perfect immersion. But it doesn't help you understand physics. The only way in a simulation is to understand how a car should feel through the steering axis (your hands) and your pedals (your feet). Everything else is immersion; admittedly, there are many great things. By the way, I always tune my files without any immersion hardware because it gives me additional feedback that has nothing to do with physics. Only afterward do I use motion, VR, or other immersion hardware to increase the fun factor.

 

Well boys, the unthinkable happened... I bought the new car pack.

As a long-time harsh AMS2 physics critic, I went into the weekend expecting to be underwhelmed, but that was not (fully) the case. These are my unrequested (and probably unwanted) opinions as someone who is generally not thrilled with AMS2 physics.

The GT1's are very fun and feel like a big step forward in how GT cars can feel in this game. I don't have nostalgia tied to these cars, but have played GTR2 and drive the RSS cars in AC pretty often, which i very much enjoy.

The Pro's:

1. The 2005 GT1's! The Viper, the Corvette, and the DBR9 drive very well. I really like how they feel powerful, heavy, dangerous, and have grip. If you hate the 'AMS2 slideyness', i would recommend giving these a shot (it still pops up occasionally, but not nearly as often as other cars). These cars alone are worth the $10, imo, and are the best GT cars in the game. If i was to show someone how much AMS2 has improved from the early days to now, i would show them these.
2. The BMW M3 GTR is also really good, it feels lighter and more nimble than the GT1's, but convincing and seems like it will provide really good racing. I would also say the physics/tires are acceptable here, but not as good as the GT1's overall. Still better than most cars in AMS2, though.
3. The LMP's are also good, but i am not a huge prototype guy and i think AMS2 nailed the prototype physics back with 1.6 / IMSA release, so no surprise there.

The Cons:

1. The new Milano and Porsche GT2 class - Not sure what happened here, but these feel like the opposite of the GT1's. They are slidey, spin if you look at them wrong, and i just can't connect with them. Maybe these cars do have significantly less grip in real life than the GT1's, but these sadly feel like your standard AMS2 GT car from the old days that just don't feel right. They just always feel like you're on a really dirty/dusty track, or have snow tires on the car,, etc etc. I'm not saying they are 'wrong', as i've never driven one in real life, but if you didn't like physics in AMS2 before, you won't like these cars.
2. The modern GT3's -- With the 'physics pass' and tread adjustments, i had high hopes that the GT3's would feel significantly better, but I am honestly not sure I can even tell a difference here. I'm starting to think these cars are just permanently cursed.

Anyway, promising signs of things to come (i hope!?). I don't know how many more physics adjustments we can expect before career mode / major updates stop, but hopefully we get a few more because they seem to be improving every step of the way.

I do still think we should be able to lean on the tires a bit more before they break traction and slide, but the GT1's are closer than we've ever been, imo.

 

I tend to think that the Madness engine favors cars without TC and ABS, but the new 2005 GT2 cars are an exception to this.
I tested them a while today and to me they are unpredictable snappy. Maybe thats how these cars drive and of course it‘s just my opinion.

The GT1 cars are a good drive, the LMP1 and LMP2 cars seem a little bit easy to me.

GT3 Gen2 also feels the same to me as before, bur they are ok to me in this state. And again: The GTE without ABS feel much better.

 

I'm curiously interested in your backround. Have you been a race driver or an engineer? Whereon do you ground your criticism?

The reason I am asking is the following:
I don't have any car or motorsport backround so I compare AMS 2 physics to my experiences with my own cars and to what I can see in the media and what professionals say in interviews/documentaries. This likely is the oppostite of how physics should be examined if we want to give valuable feedback to Reiza.
So if you have a professional backround and/or you can provide "hard facts"(data) that would be helpful.

It seems that you only describe your feelings. And if you ask me about my feelings the Porsche 996 GT3 RSR is one of the best car in the whole simracing universe because it handles exactly like it's been reported by race drivers of this era in real life. It's very hard to get the braking right and you have to really put into place to get quickly out of corners. You have to balance the Porsche rather with your feet than with ylur hands. I simply love it and I'm very happy that it handles like real life race car drivers say. But again: This is my opinion; these are my feelings. They do not help Reiza improving on the physics front.

 

Did this line from my original post give it away? "These are my unrequested (and probably unwanted) opinions"...

If Reiza want to close the physics discussion thread to only include race drivers and engineers, they will do that. Maybe they should also close the beta program, because feedback based on users feelings/opinions clearly don't matter?...

I am not going to argue physics or validity based on if i'm a 'race driver or engineer'. I've read comments like yours since AMS2 1.0, and the physics from that point to today are entirely different. Some people like them, some don't.

My post was mostly positive, and i'm not going waste time defending that. I essentially said good job reiza, i like the physics changes of some new cars when i previously haven't, and gave them $10.

If you like the physics all around, cool, enjoy it. If someone previously hasn't and wants to see if the new update has changed anything, maybe they will find value in my post.