Automobilista 2 V1.5 Physics Development Update

Maybe Reiza should secretly do one car with the full authentic tires and see what the response is. Be an interesting experiment.

 

For all we know, they did already and didn't tell us and they are already monitored the results

 

Maybe that was the car i looped.lol

 

I've been on-and-off with the update since it's released and tried a bunch of content - I haven't had time to get a taste of everything, but it's been sort of a mixed bag really.

To start - the improvement across the board is noticeable. Cars I didn't enjoy too much before (Cadillac DPi, all Formula V10 cars) are much better, cars I already enjoyed (CART) are somehow even more enjoyable and the Formula Junior is a real gem that I've been loving.

I noticed there's some general discourse around the Group C cars and how they've been tamed out, but I don't necessarily agree - a full boost Sauber C9 is still wheel-spinning into 3rd gear at Spa for me if I try to put too much power down too quickly. I think the theme of the tyres being too forgiving across the whole platform is probably being muddled into the physics of these cars - the general car behaviour correlates with what I know of the cars when you start doing some setup work to get some of the safeness out of them from the baseline setups.

I really still do not like the GT3 and GTE cars. That's fine to me, because they're not what I play the game for, but it is disappointing to have heard the hype for them and yet I can drift around the track without a care. They're heavy cars with a thin sidewall, with particularly punishing tyres, with an insane amount of downforce in real life that I think too many people in sim racing circles tend to underestimate. When you watch these cars in real life they're either glued to the ground, or, if the limit of grip has been surpassed, they spin. There's some play on the edge of grip, but past the peak slip angle it's a cliff in real life because of aerodynamic sensitivity and the tyre itself.

I'm glad that there's some recognition that could go towards this - GT3 is the most popular class in sim racing right now for good reason, so it'd probably be a good idea to get ontop of that sooner than later - but regardless, a modern GT3 car today is not an easy thing to drive quickly... when it's setup to go fast. There's a reason why the Am cars are significantly slower even with professional drivers at the helm sometimes.

Other than that, I think my main gut problem is inconsistency in FFB feel across some cars. Some cars that have power steering (Formula V10 Gen 2) feel great with lots of detail, and then others I just feel nothing but wheel weight despite the caster in the car setup being at a reasonable level. The GT1 cars are enjoyable to drive and race, but the outright feel in the wheel is somewhat disappointing. Not particularly interested in custom FFB files - just using Default+ with very minimal LFB and damping on an original CSL elite. Strength is appropriately dialled to not clip per car.

Overall I'm pretty happy. It's a great step forward and I'm glad there's a lot of positivity around the title now. Hope the positive momentum keeps being positive, though.

 

Exactly, that's what I been saying...they might as well change the game title to Automobilista 2: A Physics/FFB Simulator lol.

 

Our roads are pretty rubbish but I feel all those things through my bum. Maybe I just don't drive fast enough.

 

maybe a test car ?

 

To be fair, this is a thread relating to the physics updates. Physics talk should be expected, and honestly the conversations have been mostly very good and informative. I mean of course it hasn't been perfect, I feel some of the slidiness is being a bit exaggerated. Like the GT3 cars are not "impossible to spin" (if someone even did say that) nor would I call them "drifty" but perhaps too forgiving. While I do feel the forgiveness in GT3 cars I really don't for the Group C ones and maybe a tiny bit for the GT1s. But that also could be because I'm not that good of a driver.

But despite these disagreements I can tell that most of these critisms are in good faith. Explaining the exact problems they have while also complimenting things about the physics they like. No whinging about people disagreeing, just a lot of respect.

To those who think "you can't criticize AMS 2" should look at these past few pages and you'll see a masterclass of how to criticize in good faith. This is awesome stuff and I hope to see more of it, world would be a lot better of a place if this was the standard of online discussion instead of the exception.

 

Very good to hear your feedback on this Renato, more standard setups are always welcome, but considering that good pilots are not necessarily good mechanics, just like me rsrs, the introduction of an "engineer" in a similar way that there is in PC2 would be very welcome, making adjustments to your car according to predefined questions and answers, perhaps introducing an AI. on this that would be simply magnificent.

 

Tested a bunch of cars today and they were all fantastic

nice to see the f-v10 gen 1 finally getting fixed

mclaren F1 road car is still maybe a bit too nervous

but there was one car that is really odd now, the F-Reiza it is not pointy it is kinda sliding in high speed even on third gear, ironically behaving kinda like pre 1.5 f-v10 gen 1

 

I've found the secret code to the locked door...

Roll back to the previous build and update via "NONE" instead of RC...

I need to do more testing but I can see why the religion of "AMS2 1.4-1.48 drives great" has been split into 2 groups...



There are many many factors at play here... Including expectations and perceptions... My perception is not going to be the same as someone elses... Nor is everything you are going to read going to be 100% correct... This is just my perception of what has been a really surprising week of being stuck on the outside looking in and being unable to do anything but annoy people and ask how does car X feel now?... Anyway onto the wall of text...

Some want to feel like they are as good as their heroes on TV, whilst others want to leave a sim racing event with more respect for their heroes on TV...

Some really love the handling of GT3 cars and all the extra grip they've gotten from driver aids over the years... Others like me find that GT3s are the most boring things to drive in the motorsports world... And anything with TC or ABS on immediately gets that turned off before anything else in the set up so we can have fun...

Some just want to rock up and race on default settings, others go through every possible setting change to find a sweet spot for them for each car at each track...

Some were able to tune their FFB to feel the flatspots and find a sweet spot, therefore found EVERY car in AMS2 driveable when driving to the limits of that car prior to 1.5... Others couldn't feel flatspots or couldn't get their FFB into a sweet spot and found many cars undriveable or floaty...

Plus there is an issue with the track rubber state polluting peoples views... I've found with test day that there's no difference in track state between green, light, medium and heavy rubber with my best times usually coming on green... And when you are expecting a much harder experience on a green track than what you are receiving it's going to warp the view a little... It seems to be fine when on track in progression it's just the starting value is the same almost like Test day is the same as TT...

We also have to keep in mind that when working with tyres the tiniest change makes a massive difference to the outcome of grip... I've worked with AC, rF1, AMS1 and rF2 tyres and although I haven't touched anything in the Madness engine or to do with the SETA tyre model, from past experience I know how easy it is to go from something wonderful feeling to there's something that feels better to me but just not right for someone who hasn't been testing it religiously for months... Most of my work was for leagues where different talent levels were at play leading to some very different feedback on the same tyres... And that was done on platforms without massive updates on many fronts to the physics engine itself...

rF2s tyres were a big eye opener for me when it came to editing them, as even though there were similar parameters to change the overall effect was different... Because there was more knock on effects through the simulation of the tyre flexing is my main reasoning behind this shift in the way I had to attack my calculations when I wanted to fix something for a league or mod... So when talking about the numbers involved here it's the difference between 0.00000001 and 0.00000002...

I said those last 2 paragraphs to explain how big of a job this would of been for Reiza... There's so much content... 1 class alone could take me a year to properly tune from scratch when starting work on a new engines tyre model... Especially when there's changing track conditions to take into account as well different drivers putting different levels of force into a tyre... And in my opinion Reiza are adjusting to a massive shift in what the numbers were after fixing this bug, which could be as massive to them as adjusting to a new engine was for me...

The biggest thing that everyone has to keep in mind just how broken the physics were in pCARS 2... It was a complete mess and along with the name Ian Bell pCARS 2 is a trigger word for anyone who takes the physics representations of the cars seriously... Mention either anywhere in a positive light where there's a lot of sim racers from different platforms and you'll get those who are triggered by how broken it was... There's a complete excess of grip within the code for pCARS 2 to compensate for all of the broken elements within the engine, and that's before you even look at what is in the code from the cars perspective... Working with aero to find the grip I wanted out of the pre 1.48 tyres had me believing that there was an odd multiplier happening to the expected grip changes at anything from -0.15 to -0.6 in the lift parameters for the wings (negative numbers for lift = downforce)... This could of been excess lateral grip in the old tyres, or simply a correct representation that is different to other engines, so I would need to run a lot more tests with different physics params in controlled conditions to confirm, but that's another thing to consider... What other crazy things did Ian Bell decide to do to make pCARS 2 sell more copies...

And the engine also leads into another point to make about expectations... Some thought this game changing release was going to be a 2.0 style release for AMS2... The reality with a bug that effected so much within the physics is that it's really a 1.0 release for the Madness engine... That's not a slight at Reiza, it's a large compliment for the work that went into the back end here to get it to where it is... We're in the middle of their plan anyway...



Now that I've got that wall of text out of the way... If you read this far congratulations... You now get to read my drivers eye feedback on the first real running I've done with AMS2 1.5... I've now gone past the 10 hour mark on in game time... I tried a bunch of cars mostly in Test Day, but I did a few online races as well as one offline practice session with the AI to check on the track progression and how that is...

There's a lot to like and a lot of corrections to the physics engine that are immediately noticeable so the amount of work that went into this update has improved the simulation of many many aspects of the title... There's so much being conveyed correctly in the Madness engine that Reiza should get a lot of the glory that they deserve for finally making the physics do what they should on this engine... The tyres just aren't tuned yet...

As soon as there's heat in the tyres I don't need to think with my right foot anymore... All of the corrections are in my hands... If I drove the car to what I expected were the limits and did something really outside of those limits I'd get punished... But as soon as I started exploring the limits I found I could drive each car easily out of the corners without thinking with my feet... There was only a need to do something with the feet when I got it really wrong with my hands... Maybe a week spent doing largely DR2 helped me when it came to controlling a slide, but there were far too many moments that real life drivers would of been off track with that I could catch and continue on with...

Most noticeable with the effect are the Brabham BT52 at the Gesamtstrecke and the Caterham 620R at Brands GP... Both had a lot of soul when the tyres were cold... I read some people say these cars are hard so these were the first 2 I tried... They are a little bit of a handful when on cold tyres but as soon as I could put heat into them I was able to push further and further and further... Very little consequences from such a big turbo on the BT52... Whilst the 620R was still fun I couldn't really find the limit of adhesion on traction but it was a lot closer to what it was before the update than anything else I drove...

I did a race online at Jerez Moto in the F3... And even though I saw a lot of people going past the limit I just felt there was very little perception of the limit for me... Even though so much more was being conveyed through the FFB and that was so much easier to tune... I just had no idea where the limit of grip really was... The only real evidence I could find of a limit of grip was mid corner at Sito Ponds, Angel Nieto and Peluqui and that was when I messed up my entry point and carried too much speed into the corner...

All of the torque related responses were in my arms out of Lorenzo and Dry Sack... I kept trying to find the limit with incorrect adjustments and caught them far too often then I feel they would of been caught by my heroes on TV who would know what to do in these situations far better than I...

I have been trying to think of the best way to explain what I'm feeling on traction...

The best way would be to say that the slip behaviour of tyres pre 1.48 in understeer scenarios where you could just steer further into them is effecting the traction zones now... There's grip, grip, less grip, slightly less grip, still some grip then you are out of trouble... There's no point where I feel the grip has gone low enough for there to be a disconnect and the car is uncontrollable because of wheelspin... On cold tyres there is a limit in a few cars, but when there's heat in the tyres it's there in some shape or form in every car I have tried... I can keep pushing further than I would expect to be able to push the boundaries of the expected behaviour of the car... If I do something really stupid I pay for it, searching for laptime is a very different animal now...

Which brings me to the next issue...

This effects the way I feel about the default set ups now and what I do on the set up page... Whoever made them is wasted in software development if they are the root cause of the way these cars drive now... They should be a race engineer somewhere helping drivers win championships...

All I ever really felt I had to do was turn the steering lock to 14 on most cars... Some cars there was a need to dial out understeer so I lowered the front ARB and/or raised the rear and others had their brake bias too far forwards for me... I spent a few races in the F3s and there was no need to think about consequences of my set up changes outside of the mid corner phase and corner entry... There was no real perceivable penalty for making the car more pointy on exit... And once the brakes were in the right zone for me the only real trouble I had on corner entry was the timing of when to brake... Which makes my set up work almost entirely about laptime and not about the effects of weight transfer causing me to spin...

I'm not Shane Van Gisbergen... I shouldn't get out of one car and be able to drive something from a different era and different level of horsepower and feel I'm able to put down laps within the time it takes to get heat in the tyres... There should be an adjustment period that was there before that isn't now... This could be what is causing so many people to say that each car can be driven the same... There's less of a re-learning of how to drive another car...

I've read a lot of reports of some kind tyre wear related consequences for spinning the rears on exit... I have yet to find this... There's more heat than there is in the front tyres but there's never a rear limited situation on wear, I was always front limited and it didn't matter what car... And there was never too much heat in the tyres leading me to have less grip...



A bug that has such wide ranging effects that also came with so many advances in other areas is a huge workload for any team... iRacing would struggle to get through this in the time Reiza have and get it right without a wide variety of testers providing massive amounts of feedback in different scenarios... And there were massive content projects on top of this workload...

From the timeframes given of when the bug was discovered I've traced an excess of grip back to the release of Group C Nissan R89C and the Brazilian Legends DLC which gives me reasons to believe there's been compensations made after the bug was discovered (I could be wrong again this is just my perception of the situation without raw data from the tyre files)... And it most certainly isn't the first time there were compensations in grip made for an error elsewhere in the physics code of the Madness engine...

With the Group C tyres the Nissan had/has different tyres to the other 3 cars and I've been tuning different sets of physics to the Group C tyres... The pre-1.5 version of the Nissan tyres had too much longitudinal grip and a lot of what is described above could be said about them when you took the 918bhp engine out of the equation... The Mazda was described as boring over many changes because of the very issues I describe above with 1.5... I had a very similar problem tuning the Indycar Speedway physics on the 1.48 build as soon as there was heat the twitchy behaviour I was going for was gone... With those tyres I was hoping that the 1.5 release would give them more definition of the edge of grip... I am yet to test any of the F-USA outside of the gen 1 cars at Cleveland... Which again was a handful with cold Hard tyres, but as soon as there was heat all I felt I had to do to the car was steering lock and brake bias and it's boundaries could be pushed further than I thought should be possible...

I also think the last changes for the final pieces of 1.5 tyres that were early releases on 1.48 gave extra grip to the driver... The P1 Metalmoro Judd is a fine example of this... I loved the 1.48 version of this car and the 1.5 version isn't that much different... But there's more grip out there and I'm not really punished for a flatspot like I was... Like the Caterham 620R I found it to be one of the more fun cars that I have tried...

I also found with the Group A Mercedes that flatspots weren't really a problem outside of my wheel rattling... It was the only car I really experienced a flatspot with and at Hockenheim I felt I could finish the stint as long as I could stop locking it up and take tread off the tyres...



TL-DR version

This was a massive undertaking that requires massive amounts of testing to get everything right, tyres are not a simple part of the physics process and with a track that evolves there's even more variables to think of... Anyone expecting this to be a 2.0 style release of the game is going to be disappointed... This was always going to be a long haul process and a fix to a bug like this is far more about the engine side of the equation than the game side...

We should be using this chance before the TTs are reset to give as much feedback as possible for Reiza to make the right changes to these tyres... Mathematically they are most likely only off by the tiniest of margins and making that small change could throw something else off within the tyre code or the default set up... Which then requires more testing...

My driving thoughts are condensed in this fashion...

If you drive it like you stole it you'll pay, but the middle ground is far too forgiving...

A lot was fixed, but the feeling of respect I had for the drivers that get paid to do this in real life after every event I did on 1.4-1.48 is gone...

 

very interesting but I just spent my 3 weeks of summer vacation reading your article.

 

You were warned

 

I already lost you on page 37…
Just kidding, nice article and a good read. Very informative…

 

(Obligatory remark on size of comment)
With that out of the way, I've gotta say I wholeheartedly agree with what's being said here- there's always been something funky with the amount of grip you're able to deal with and honestly at this point I think it's the only real notable flaw I can point out in the physics: everything else has been refined very well and AMS2 is a lot of fun to rag cars around in! However, I just feel it needs a bit more tweaking in how the car loses grip, particularly the aforementioned "just steer more to avoid understeer and get nothing more than some tire wear/overheating" thing that's been around forever
I reckon Reiza can probably get it figured out eventually given it's been mentioned that it's an already noted problem, I guess we'll just have to see on if they can properly dial it in

 

I find myself in a very similar spot, there's no doubt that you should be able to control a race car on slicks when it's on the limit, however the easiness you're able to do so in AMS2 can feel offputting, took the Stock car brasil 2019 yesterday to a hot Indianapolis road course, and I was able to drift the car around with full throttle just doing corrections on the steering.

The track condition is another thing that I've noticed, where there doesn't seem to be that much difference between a rubbered track vs a green track on the start of the session (Don't know if a green track at the beggining of practise is equally as fast as a fully rubbered track just selecting the preset).

Lets hope like you said that with further tweaking on the tyre front we can get to a point where this behaviour can be corrected.

 

chatgpt has summarised this:


- Various factors affect the driving experience, including expectations and perceptions.
- Different drivers have different preferences for sim racing events and car setups.
- Track rubber state can impact perceptions, but it's not always consistent.
- Tuning tires is a complex process, and small changes can have significant effects on grip.
- The physics engine in pCARS 2 was broken, and its excessive grip influenced AMS2 physics.
- The update to AMS2 1.5 improved many aspects of the simulation, but tire tuning remains a challenge.
- The slip behavior of tires has changed, and some cars feel too forgiving, leading to less differentiation in driving styles.
- A bug with the grip in Group C cars may have influenced physics adjustments in the game.
- The AMS2 1.5 update is a significant step forward, but more feedback is needed to fine-tune the tires.

 

All this talk of improving physics is all well and good but when are you going to fulfill the ultimate destiny of AMS2 but giving us the 38 mile Isle of Man TT Mountain Course? If I could race the Porche 911 gt3-r around this track I could die happy.

 

RIP Andrew Curtis. Cause of virtual death: 38 mile Isle of Man TT Mountain Course.

Destiny awaits!

 

RIP - Racing In Perpetuity.