AMS 2 | Kyalami Grand Prix Circuit Headlines New Content Additions In Latest Update

Paul Jeffrey

Premium
Reiza Studios today deployed a brand new update to their Early Access Automobilista 2 racing simulation - adding new content, improved netcode and a raft of fundamental improvements to the title.

Update time! Yes, Friday, the start of the weekend, for most of us two days to rest and relax and more importantly - Automobilista 2 early access update time!

In this new build from Reiza Studios, the developing AMS2 simulation enjoys quite the array of changes and improvements, not least of which is some gratefully received netcode advancements for those with an online fancying, and AI balancing tweaks to be enjoyed by those that like their racing thrills against computer controlled opponents.

In content news, the latest build adds the impressive modern layout of Kyalami, plus the P2 specification Sigma P1 (with the updated 2020 P1 variant expected to arrive in sim once the real world version races on track). In old school content news, today's build adds the Passat to the Copa Classic FL category - however this car is know to cause a CTD at present, an issue that Reiza are currently investigating and expect to hotfix in the very near future.

Quick note - This is the last weekend to buy Automobilista 2 with the current 40% discount over its v1.0 price - as we have crossed the halfway point in our Early Access roadmap, by Monday pricing will be adjusted accordingly to be 20% off its v1.0 full price, remaining that way until its official release.


V0.9.0.1 CHANGELOG:

CONTENT

  • Added Kyalami Grand Prix Circuit
  • Added Sigma P1 to P2 Class
  • Added Passat to Copa Classic FL Class
GENERAL
  • Update App Key used in Live Motion SDK
NETCODE
  • Reduced Fade Accel parameter to minimise chances of "ghost" car-to-car collisions in a Multiplayer Session
UI & HUD
  • Fixed back button on TimeTrial screen
  • Amended all brake bias labels to include F / R qualifier and all values to include front and rear.
  • Fixed telemetry HUD screen anchors (affects ultrawide positioning)
  • Fixed in-game menu anchors on various pages (audio, camera, controls & gameplay options screens, save/load setup, quick/full setup edit screens, VR)
  • Added Track Altitude information to loading screen
  • Added current time of day to Session Overview / Pause screens
  • Added Imola 1972 & Modern trackmaps
  • Added Stock Car Driver names
  • Updated missing tire names for Caterhams, F-Ultimate, Copa Classics
  • Added chat box to lobby page, multiplayer & pre-event screens
  • Fixed bug with incorrect engine torque statistic in vehicle selection & loading screen due to reversed metric / empiric system conversion
  • Adjusted naming for various Metalmoro MRX models
  • Fixed track country label on lobby page
  • Added exit confirmation to pre-event screen
  • Coodown lap now defaults to manual player control
PHYSICS
  • Standardized ranges, adjusted drag & lift increments per setting for cars with adjustable radiator & brake ducts
  • Adjusted brake heating for all prototypes & GT cars
  • Minor adjustments to Ultima Race, StockV8, F-Vintage tyre treads
  • Reduced roll inertia for StockV8, SuperV8, Ultima Race, Ultima
  • Slightly reduced grip off the ideal racing line
  • Reduced Metalmoro AJR diffuser efficiency
  • Adjusted StockV8 diffuser center of pressure slightly rearwards
  • Adjusted Roco, MRX, ARC Camaro engine torque curves
  • Moved default brake bias rearwards for Sprint Race, StockV8, Ultima Race
AI
  • Slightly increased AI awareness of human players
  • AI performance tuning for GT3, StockV8, P1, P2, P3, P4, Ultima Race, Sprint Race, F3, F-Vintage classes
TRACKS
  • Spielberg: fixed hole in garage floor; replaced incorrect pit building walls & pit doors; remapped garage and pit stall locations; removed excess grid assignments (32 cars supported); rebuilt triggers.
  • Brands Hatch. Fixed bug with cars appearing to float over the track; adjusted rolling start location; revised track limits; fixed garage door collision
VEHICLES
  • AJR: Adjusted Driver Animation; Added Imperio #175 & Mottin Racing 46 liveries
  • MRX: Corrected RPM LEDs for all variants
  • Ultima Race: Adjusted collision mesh
  • Puma P052: New liveries/updates/redesign for #01 #02 #04 #06 #07 #08 #10 #11 #12 #13 #19 #42 #53 #54 #69; Added community skins #17 #91


Original Source: Reiza Studios

AMS 2 is available now in Steam Early Access.

If you need help and support getting the most from AMS2, start a thread in our AMS2 sub forum here at RaceDepartment and let our great community help you out.

AMS 2 Kyalami.jpg
 
Last edited:
With "yaw velocity" do you mean "the rate of rotation around the vertical axis through the CG of the car"?

I'm unsure what you mean to imply with "a pervasive resistance in keeping traction on the rear"? Are you suggesting that the rear axle has poor traction and that the car wants to oversteer (haven't tried it myself yet)?

Yes, the yaw motion is the one around the vertical axis.

And no, the Opala 76 has very high traction and it's very difficult to 'persuade' it to power overseer.
 
However, I think he may also be onto something with the ability to overdrive the front tyres at very high slip angle -- if you want to try it yourself, take the 620R and drive it slightly too fast into the corners and then put on way more lock (like 2-3x more) than you'd normally do while stepping on the throttle and note how the fronts keep sticking as opposed to them giving up and serving you with a nice trip head-first into the barriers?

I've tried this and you can provoke terminal understeer if you want. You have to be pretty brutal with the steering input.

As an avid reader of Milliken and Milliken I'm sure you've read the chapter on understeer amd oversteer, and 8.5 Limit Behaviour. So if you are testing a car close to the limit you have to be aware of its general handling balance. A car which is set up to understeer, will understeer more if you apply more lock. A car set up to oversteer will oversteer if you apply more lock.
 
I've tried this and you can provoke terminal understeer if you want. You have to be pretty brutal with the steering input.

As an avid reader of Milliken and Milliken I'm sure you've read the chapter on understeer amd oversteer, and 8.5 Limit Behaviour. So if you are testing a car close to the limit you have to be aware of its general handling balance. A car which is set up to understeer, will understeer more if you apply more lock. A car set up to oversteer will oversteer if you apply more lock.

Yes but I would add that for this exact reason not even race cars (for sure not RWD cars) can be set-up IRL for steady state oversteer without (sooner than later) killing the driver.
 
Last edited:
I've tried this and you can provoke terminal understeer if you want. You have to be pretty brutal with the steering input.

As an avid reader of Milliken and Milliken I'm sure you've read the chapter on understeer amd oversteer, and 8.5 Limit Behaviour. So if you are testing a car close to the limit you have to be aware of its general handling balance. A car which is set up to understeer, will understeer more if you apply more lock. A car set up to oversteer will oversteer if you apply more lock.

Yes, and Avo is suggesting that it appears that on some cars, if they are already understeering, you can pile on more lock without understeering more, that is, the front axle retains the same amount of grip, which means the turning radius doesn't change when applying more steering and slip angle to the front wheels at steady velocity.

If what you are quoting from M&M were the case, keeping constant velocity while increasing lock would result in a greater turning radius (= "understeer more") due to the decreased abilty of the tyres to generate lateral force beyond the peak grip/load/slip point. This is also what Avo is suggesting should happen.

Now, I've tried the behaviour with both Opalas and the 620R and if I've loaded the car to the point of understeer and then apply more lock (2x-3x) and keep the velocity steady (by applying more throttle to counteract the scrubbing off of speed) my turning radius doesn't increase; it stays the same.

I'm testing on VIR Grand, particularly the slower, longer corners.

I'm still in the process of working out the best, most repeatable sequence with which to contrast and compare the "normal lock" and "excessive lock" scenarios.
 
Last edited:
To get a theoretical perspective on Avo's objections, this article might provide some insight. IIUC, Avo is arguing that the line in Figure 4 in that article in the frictional part of the tyre regimen is too flat in STM tyres and ought to slope more downwards as more slip angle is added.

Have you any idea how flat the curves are in AMS1? The files are encrypted I think but I do have an unencrypted tyre file from Stock Car Extreme

For the 2014 Stock Car tyre

grip drop off at twice peak slip angle = 0.0655%
grip drop off at three times peak slip angle = 0.1772%
grip drop off at four times peak slip angle = 0.3778%

That is pretty flat! Steep drop off in slip curves went out of fashion back in 2005.
 
In AMS 1 I frequently found the 60 second delay from the time you entered a session to the time the track went green to be annoying. The more I play AMS 2, however, the more I feel I’d really like to have that time gap back.
 
The other plausible explanation is that:

1./ It takes developer time to implement this. If it requires significant new software engineering efforts (design, implementation, testing, bugfixing), then that is a barrier in and of itself.

2./ AIUI, it is well known for people studying User Interaction that exposing too many options in an "advanced settings" page is no bueno because it confuses the less technically adept users and actually causes more issues than it solves, among which is a heightened user question and support burden. According to latest cognitive research I've seen (which may already be outdated!), the human mind can apparently hold up to 4 concepts at a time in its working memory, which means that e.g. exposing more than 4 knobs to the user on a single page is a recipe for confusion.

3./ Changing design and code at a late stage in a project's development is almost never a good idea -- particularly close to the release of a 1.0 version. In this development phase, you generally want any changes to be minimal and be reserved to bug-fixes and very small tweaks to existing subsystems and parameters.

1) hmm no not really, it's just a UI addition in most cases which fits alongside the multitude of other UI options. Other Sims have added in some hidden options in the past so it's not actually that hard to do.

2) That's why I would say have Pre warning with a big Reset To Default button, again not that much work and provides the user the option to escape the madness if they dig too deep.

3) See #1, the games development doesn't stop after EA, future UI updates may well come further down the line generally speaking.

I think you're somewhat missing the point that these options are there in the background and already working, they just need adding to the UI, there's no major coding work to be done that hasn't been done already.
 
Have you any idea how flat the curves are in AMS1? The files are encrypted I think but I do have an unencrypted tyre file from Stock Car Extreme

For the 2014 Stock Car tyre

grip drop off at twice peak slip angle = 0.0655%
grip drop off at three times peak slip angle = 0.1772%
grip drop off at four times peak slip angle = 0.3778%

That is pretty flat! Steep drop off in slip curves went out of fashion back in 2005.

Without telling Milliken and Milliken about it?! Heresy!

In any case, from your last two posts, you appear to state both that peak slip drop-off should be almost flat and that it shouldn't be (corollary of the paraphrased quote from M&M).

I take it you lean more towards the former (relatively little drop-off)?

Also note that I haven't stated that *I* think one is more correct than the other; I'm just analysing what Avo and now you are stating and offering my observations, which happen to be in line with relatively little-to-no drop-off "falling out" of the STM, which unlike Pacejka is a dynamic model (as you are no doubt aware) and thus doesn't rely on lookup tables to generate slip curves to the best of my knowledge.

@Avo77 : Are you at liberty to share some examples of ballpark drop-off values from your own personal experience?
 
1) hmm no not really, it's just a UI addition in most cases which fits alongside the multitude of other UI options. Other Sims have added in some hidden options in the past so it's not actually that hard to do.

2) That's why I would say have Pre warning with a big Reset To Default button, again not that much work and provides the user the option to escape the madness if they dig too deep.

3) See #1, the games development doesn't stop after EA, future UI updates may well come further down the line generally speaking.

I think you're somewhat missing the point that these options are there in the background and already working, they just need adding to the UI, there's no major coding work to be done that hasn't been done already.

Are you by any chance a software developer/engineer by trade?
 
1589302984831.png


Let's stay on sim racing for a moment. Thanks to the massive work of @mclarenf1papa, in the graph you may find in red a typical slip vs grip curve for AC, that is already considered one of the most forgiving sims out there (it's for longitudinal silp, but the same stands for lateral one).

In black a possible more accurate (opinion) approach: it's in testing and we will see what will happen but the first impressions seems to be good with no "it's very difficult so it must be realistic" feel.


Schermata 2020-05-12 alle 19.50.30.png


Speaking about real life data, in the black graph you may find a lateral slip vs grip curve (slip numbers are meaningless and grip is normalized to one) shared by one of the major tire manufacturer for one of the major world championships.
This don't mean automatically that the numbers are perfectly fine (they more than often require a lot of 'massaging'), but at least the trend is reasonable.

It's a modern race tire tested on purpose at load higher than the nominal one but still...

From my experience in general no, the slip curve should not be (too) flat after the peak.
But even if probably both AC and AMS are too permissive in this regard, the driving experience is plausible anyway, maybe because there are several other effects influencing the grip distribution, maybe because... who cares?

They are fun and good enough not to rise major suspects with the strong advantage that if someone don't like something he is also free to modify at will in order to compensate (all my favourite cars in AC have custom tires, for instance...).

AMS2 at the moment (in my opinion) is not yet in the 'good enough' status...
 
Last edited:
@Avo77 : Cheers for sharing the graphs (that's what the 'Like' was for).

It looks like the black line (the new AC tyre in testing) has a nice, wide and thus presumably "user-friendly" peak, even if it drops off harder at higher longitudinal slip percentages? So in this case, it would mean that it would generate the most acceleration at around 10-13.5% slip (both under braking and acceleration)?

It's a pity that tyre data is considered so proprietary that it's difficult to get an idea of what is up and what is down in this field of engineering...

EDIT: Re. the real-life black tyre lateral slip(x) / grip (y) graph: If the x-axis was expressed in radians, the peak would be in the range 0.26-0.32 radians which corresponds to 14.9-18.3 degrees? And if it was purposely loaded too high, the peak would be at a lower, more reasonable range for nominal loads, yes?
 
Last edited:
@Avo77 : Cheers for sharing the graphs (that's what the 'Like' was for).

It looks like the black line (the new AC tyre in testing) has a nice, wide and thus presumably "user-friendly" peak, even if it drops off harder at higher slip angles? Out of curiousity, what is the scale of the X-axis measured in for this example?

It's a pity that tyre data is considered so proprietary that it's difficult to get an idea of what is up and what down in this field of engineering...

I won't expect a noticeable difference from the wider peak (not for the tire in the graph, at least).

Let's wait and see, the guys are making wonders to the AC engine and even rain physics are in the work.

Regarding tires the available data are limited for several reasons and the main one is that making reliable tests is almost impossible, in particular with combined slip and high loads or speed.

This makes also the tire manufacturers very jealous and protective about data...
 
the real-life black tyre lateral slip(x) / grip (y) graph: If the x-axis was expressed in radians, the peak would be in the range 0.26-0.32 radians which corresponds to 14.9-18.3 degrees? And if it was purposely loaded too high, the peak would be at a lower, more reasonable range for nominal loads, yes?

No those are not radians and I stretched also a bit the scale on the x axis...

Btw the peak for this kind of tire arrives much, much sooner than 15 deg IRL.
And yes, at lower loads the peak would be anticipated and the curve flatter.
 
Without telling Milliken and Milliken about it?! Heresy!

In any case, from your last two posts, you appear to state both that peak slip drop-off should be almost flat and that it shouldn't be (corollary of the paraphrased quote from M&M).

I take it you lean more towards the former (relatively little drop-off)?

You take it correctly. To be fair to Milliken and Milliken, their curves from real tyres don't go beyond the peak. As Niels and Dave Kaemmer have previously said - data for this region of tyre behavior is almost non-existent. A slightly more recent book "The racing and high performance tyre" by Paul Haney does have some slip curves which show very little drop off beyond peak (p114 if anyone has the book) but other parts of the book shows the traditional drop off beyond peak.

I worked for Blimey! Studios during development of GTR2. Doug Arnao their physics guy managed to get hold of a slip curve for a GT tyre that went beyond peak and he said it was pretty flat. The result was that the drop off in grip at twice peak slip angle went from about 8% drop in the original GTR to less than 1% in GTR2. Of course there were accusations of "dumbing down" and "arcade" from some but real life drivers felt GTR2 was more realistic.

There is still this split in the simracing community between those who feel harder to drive is more realistic, epitomized by shall we say iRacing and those who feel you can drive racing cars at and beyond the limit without dying, with no doubt some in the middle.
 
Back
Top