Automobilista 2: New Preview Video Released

[Paul Jeffrey]

Not satisfied with the last Automobilista 2 video footage? Well wrap your eyes around this one...

Reiza Studios recently revealed a new comparison video of the developing Automobilista 2 simulation, showing the game in direct comparison to its older brother AMS 1, and featuring the work-in-progress Snetterton Circuit in the UK.

Despite obviously being a significant upgrade over the original title, many fans were a little underwhelmed by how development of Automobilista 2 looked to be progressing from a visual point of view, something that Reiza Studios have acknowledged when reading the comments section of the various articles featuring the July roadmap post.

Having another shot at wowing the crowds, Reiza have recently dropped a much improved video of the Ultima GTR lapping around the Snetterton track in various time of day conditions - and it's fair to say the visual improvements are considerable...

Capture has a higher bitrate which is more accurate to the game´s actual graphical quality" said Reiza Studios Renato Simioni.

Other differences in this new video include improvements to track lighting, updated road and foliage shaders, revised textures; smoothed cockpit model edges & updated materials; adjustments to audio code, rebalancing of engine sounds and various sound effects (also replacing a few placeholders used in the original video); updated physics and increased game graphical settings.

Should be stated perhaps that this still isn´t the most exciting car / track combo or settings to promote AMS2 which wasn´t really the point of the original video - obviously we miscalculated how closely people would be looking!

This is a better representation of AMS2 in its current stage of development but it´s still a WIP - things will change as we progress to release and continue to afterwards. Such is the nature of the thing...
​

So, what do you think?

Automobilista 2 will release for PC December 2019.

For the latest Automobilista 2 news and discussions, head over to the AMS 2 sub forum here at RaceDepartment and get yourself involved in the conversation today!

Like what we do at RaceDepartment? Follow us on Social Media!

 ​

 

[Salvatore Sirignano]

@Spinelli Have you tried the brilliant F1 71 mod? I sort of get what you mean, but in the above it is possible to get the rear sliding under braking, hit the apex sideways and power away still on opposite lock towards the outside of the track.

However, as mentioned earlier, I agree about the F V12 1995 in AMS. I think something changed since Niels posted this video of GSCE, skip to 25:00 and watch his last full lap, he drifts beautifully out of every corner.

I turned 200 laps of Montreal 2015, got a top ten leaderboard time (1:19.2) changed everything in the setup for controllable power oversteer, but not once did it handle like that.

@Niels_at_home Any (setup) tips? I'm desperate to be able to replicate the car behaviour in your video!

 

[Richard Wilks]

I believe what happened is that details in the physics engine were changed from GSC to AMS, and maybe this car was never properly updated by Niels himself. This is pure speculation in my part. Anyways, that video alone puts to rest the assumptions from Spinelli that its a flaw of the engine itself.

 

[Spinelli]

@SpinelliYou can't seem to make up your own mind, and i will repeat again, what you talk about is a "school" of doing a virtual car. TIme and time again i hear people wanting "crisp" steerings and tons of steering "feedback", and then accuse PCars2 of being "too slidy.

I can't make up my own mind? Again, no idea what you're talking about. And about steering feel, I'm not even talking about crisp steering / FFB nor "tons of steering 'feedback.'" I think you're massively misinterpreting some of what I'm saying or you're just making up your own discussion in your mind. The only time I mentioned mush was talking about grip/slide characteristics. I don't think I ever mentioned anything about actual FFB & crisp steering feel nor amounts of steering feedback in this thread.

@SpinelliYou keep harping your arguments over and over, but you have no proof whatsoever, no detailed analysis, no excel sheet with multiple slip angle or inertia changes, no different cars to compare, no nothing. I can make a car in AMS that behave just like those F3 you have in your video. But then you might say "oh the steering is mushy" like you said before. F1 cars in codies behave like that but "oh but its simcade garbage".

You seem to have a personal agenda against me. The fact that you're implying that I won't admit that F1 cars in codies is capable of this (or at least more capable than some other games/sims) and that, if I did admit it, I would then make an excuse like "oh but its simcade garbage" is proof of this because I never did/said either of those. You're completely fabricating things as a) i did admit and agreed with you that codemasters does this better than most sims (I've even mentioned this other times outside this thread), and b) I never then went on to say anything like "oh but its simcade garbage." Again, you're massively misinterpreting a lot of what I say or you're just making up your own discussion and replying to that.

Yes, I've given tons of detailed analysis in many long posts. This stuff can be sensed, even just visually if not playing, to a certain degree with basically every vehicle. It's literally easy to spot even just watching other people play online or replays, let alone actually playing yourself. Just use your eyes. We're not talking about a front wing or front suspension that should be dropping 8mm under braking at a certain load but it instead drops 6mm, or not enough downforce being lost when a car's in a certain angle of yaw. What I'm talking about does not require telemetry data, it literally requires just looking at vehicle behaviour, vehicle actions and reactions.

And if you can make a car an F3 car in AMS that can do that, are you saying every open-wheeler from every single developer and modder in the history of any game based on any version in the history of the ISI engine doesn't know how to properly make an open-wheel car in the ISI engine and if you make one you'd be the first person in history? Obviously I don't believe that. I would love to download this car of yours though. And, please, if you do decide to send it to me, make sure it's a proper F3 car with proper race "style" tyres. Not some car that's had it's physics purposely made or edited just to be able to try and achieve what I'm saying while then only "looking" like an F3 car.

@SpinelliYou are not wrong abotu the F V12 in AMS...

Thank you.

There is already another thread with this same F3 video. Even Niels Heusinkveld replied to that one. No need to repeat the same replies as are there. Yes, AMS can replicate that behaviour is the simple answer.

Just like you said in the other thread...all you said was "yup, I can do that in AMS." Please show me.

 

[Richard Wilks]

Your "detail analysis" is just your perceptions of what cars should and shouldn't do based on random corners at random tracks. And i dont have anything against you, you seem to be the one that has something against the ISI engine, because you say the same things in every thread.

As for cars that behave differently, i already showed. Like other users said, i have cars here for download. Not that F3 car, but others. You are free to try them.

I also find it interesting how you ignored Niels's Video of an early version of the Ferrari, in which he is doing all that you say it isn't possible on the ISI motor.

 

[Spinelli]

@Spinelli Have you tried the brilliant F1 71 mod? I sort of get what you mean, but in the above it is possible to get the rear sliding under braking, hit the apex sideways and power away still on opposite lock towards the outside of the track.

However, as mentioned earlier, I agree about the F V12 1995 in AMS. I think something changed since Niels posted this video of GSCE, skip to 25:00 and watch his last full lap, he drifts beautifully out of every corner.

I turned 200 laps of Montreal 2015, got a top ten leaderboard time (1:19.2) changed everything in the setup for controllable power oversteer, but not once did it handle like that.

@Niels_at_home Any (setup) tips? I'm desperate to be able to replicate the car behaviour in your video!

Ciao Salvatore.
Those slides look a little better but they're often still extremely on/off and crazy peaky. Also, you can see that upon any tiny modulation of throttle in the middle of an oversteer angle, the car's front end will "hyper-react" and start darting all over the place. You can see in so many slides where the INSTANT the revs start loosing grip, Niels does an instant full throttle lift and an instant steering correction snap and then he snaps the wheel back to centre. There are some other slides where he's able to slightly hold the angle and wheelspin but there's no or almost no modulating of the wheespin. It's all sorts of tiny little inputs (throttle and steering) which result in hugely over-reactive reactions from the vehicle. He's not so much controlling and manipulating the vehicle's oversteer and wheelspin but more just hyper-reacting to it. The issues are visibly still there, Niels (and other drivers) are just very used to the game and these limits and know how to keep the car within those limits. Another issue, is often slides require an unnaturally small amount of steering correction. There are slides there which have quite decent angles yet there's like 20 or 30 degrees of opposite lock and it's like you have to make sure to add less steering lock than you should and then just sort of wait for the slide to dye on it's own. You watch real onboards and you see all sorts of big steering angles required during oversteer. Also, often in that video when the slide is dying off, the car often snaps back to centre rather than being able to modulate the throttle and wheelspin while the slide returns to centre in a controllable fashion. Those 2 last points often result in snap-overcorrection which happens by far the most in ISI based sims (and iRacing).

I've had 3 real race drivers at my house who all agreed with me. One was a fairly high-profile karter on the world stage with Star Mazda wins (when it was high profile and full of future Indy Stars) and tests with Formula Atlantic teams (1 step below Indycar). They've all said, more or less, the same things as me. At first they're actually blown away by how detailed and good the physics feel as they start pushing but not too hard. Then as soon as they start really pushing the car and really playing with the limits and slip-angles, they all then have this confused look on their face like "what just happened to the incredible sim I was playing before the slide occured" look. Then, I put Netkar Pro or Live for Speed on to get them to try and they all start pushing the cars and getting oversteer slip angles and controlling them great. They all say they just drive naturally without changing according to the game whereas with the other games (ISI engine or iRacing), they need to adapt and control the car according to what works / doesn't work in the game.

Your "detail analysis" is just your perceptions of what cars should and shouldn't do based on random corners at random tracks. And i dont have anything against you, you seem to be the one that has something against the ISI engine, because you say the same things in every thread.

It's not based on random corners or random tracks. It's general overall behaviour of specific types of behaviors/actions/reactions. You make it sound like I just drove 5 or 6 cars over the past couple weeks and decided something is wrong, lol.

I also find it interesting how you ignored Niels's Video of an early version of the Ferrari, in which he is doing all that you say it isn't possible on the ISI motor.

I did reply. And it's not doing anything that I said is extremely difficult or impossible to. The fact you think that video shows stuff proving me wrong is proof itself that you're not even comprehending what I'm saying are the issues.

 

[MarcG]

Some things NEVER change!

 

[David Ignjatovic]

Years ago when I did a stint with Gtr evo I was able drift the F1 mod downhill with dfgt pedals.

I was playing 40hrs or more per week and appeared to overcome the limitations of the pedals, ie, I developed incredible sensitivity with my pedal inputs.

Trouble is, to get that level of control back I'd have to put in 80-100hrs over the next 2 weeks.

 

[motoliser]

From this footage we can clearly see AMS 2 is going to be another simcade! Back to AMS 1...

 

[Niels_at_home]

@Spinelli
However, as mentioned earlier, I agree about the F V12 1995 in AMS. I think something changed since Niels posted this video of GSCE, skip to 25:00 and watch his last full lap, he drifts beautifully out of every corner.

I turned 200 laps of Montreal 2015, got a top ten leaderboard time (1:19.2) changed everything in the setup for controllable power oversteer, but not once did it handle like that.

@Niels_at_home Any (setup) tips? I'm desperate to be able to replicate the car behaviour in your video!

Hey look at that, indeed the lap a bit after 25 minutes has some very calm and collected (pun intended) slides!

This has been years, it probably was pre-release physics, and while that lap went well, it wasn't easy. Then again when Michael Schumacher slides such a car on a very low grip Argentine track, doesn't mean it is easy, the guy knew how to drive..

Spinelli must have written megabytes worth of text over the years.
Whether that is commendable or worrying, well... ;-)

I'll just repeat what I've kinda said before.
To my knowledge there is nothing inherently wrong with the physics engine. Over the years most of the times we thought there was something wrong, it was just our lack of understanding. More recently I've figured out more of the source code. The tire model when it comes to combined cornering / braking / acceleration forces behaves as you would expect a good tire model.
An experienced and knowledgeable physics programmer literally said 'Don't worry about that tire model' (regarding the combined forces).

There are plenty of things NOT modeled in the engine, and who knows if there are some slight bugs in there somewhere. But all the evidence we have, and that also includes lots of real drivers, point in the direction that AMS can be pretty much amongst the top of available realistic sims when it comes to driver training.

Of course I would like to revisit all the cars in AMS, but the reality is that they were made over the years and tweaked with our tongue at the right angle. If we were to redo them all, I'm sure some cars would benefit as the numbers are just as vital as the engine the game runs on, but its also subjective so in the eyes of some people we'd make them worse..

Edit: this is about AMS 1 of course, as that was where the handling discussion was about, even though we are in the AMS 2 forum!

 

[Salvatore Sirignano]

Hey look at that, indeed the lap a bit after 25 minutes has some very calm and collected (pun intended) slides!

This has been years, it probably was pre-release physics, and while that lap went well, it wasn't easy. Then again when Michael Schumacher slides such a car on a very low grip Argentine track, doesn't mean it is easy, the guy knew how to drive..

Spinelli must have written megabytes worth of text over the years.
Whether that is commendable or worrying, well... ;-)

I'll just repeat what I've kinda said before.
To my knowledge there is nothing inherently wrong with the physics engine. Over the years most of the times we thought there was something wrong, it was just our lack of understanding. More recently I've figured out more of the source code. The tire model when it comes to combined cornering / braking / acceleration forces behaves as you would expect a good tire model.
An experienced and knowledgeable physics programmer literally said 'Don't worry about that tire model' (regarding the combined forces).

There are plenty of things NOT modeled in the engine, and who knows if there are some slight bugs in there somewhere. But all the evidence we have, and that also includes lots of real drivers, point in the direction that AMS can be pretty much amongst the top of available realistic sims when it comes to driver training.

Of course I would like to revisit all the cars in AMS, but the reality is that they were made over the years and tweaked with our tongue at the right angle. If we were to redo them all, I'm sure some cars would benefit as the numbers are just as vital as the engine the game runs on.

Edit: this is about AMS 1 of course, as that was where the handling discussion was about, even though we are in the AMS 2 forum!

Thanks for taking the time to reply!
I wasn't overlooking your superior driver skill as a contributing factor of course
I'm using a g27, how much available tyre slip information am I losing in AMS with respect to a direct drive setup?

 

[Niels_at_home]

I wish I was joking when I say with a G27 you loose about 90% of the feel compared to a Direct Drive wheel but I'm not..

The immediacy and strength (even at non strongman settings) just help you so much to catch slides, its almost cheating.

 

[Marc Collins]

Hey look at that, indeed the lap a bit after 25 minutes has some very calm and collected (pun intended) slides!

This has been years, it probably was pre-release physics, and while that lap went well, it wasn't easy. Then again when Michael Schumacher slides such a car on a very low grip Argentine track, doesn't mean it is easy, the guy knew how to drive..

Spinelli must have written megabytes worth of text over the years.
Whether that is commendable or worrying, well... ;-)

I'll just repeat what I've kinda said before.
To my knowledge there is nothing inherently wrong with the physics engine. Over the years most of the times we thought there was something wrong, it was just our lack of understanding. More recently I've figured out more of the source code. The tire model when it comes to combined cornering / braking / acceleration forces behaves as you would expect a good tire model.
An experienced and knowledgeable physics programmer literally said 'Don't worry about that tire model' (regarding the combined forces).

There are plenty of things NOT modeled in the engine, and who knows if there are some slight bugs in there somewhere. But all the evidence we have, and that also includes lots of real drivers, point in the direction that AMS can be pretty much amongst the top of available realistic sims when it comes to driver training.

Of course I would like to revisit all the cars in AMS, but the reality is that they were made over the years and tweaked with our tongue at the right angle. If we were to redo them all, I'm sure some cars would benefit as the numbers are just as vital as the engine the game runs on, but its also subjective so in the eyes of some people we'd make them worse..

Edit: this is about AMS 1 of course, as that was where the handling discussion was about, even though we are in the AMS 2 forum!

Of course I agree with you because I have been arguing with @Spinelli about this for years.

But intended in a non-confrontational manner and yes, even more off-topic than AMS in an AMS 2 forum, has anyone tried ACC? I can only assume @Spinelli absolutely hates that title because the issues of lack of rear grip and spins that seem to happen too frequently and with too-little to no ability to recover are at least 10 times worse in ACC than in AMS (for me, at least). Even the most skittish cars in AMS feel more planted and are more controllable than the cars in ACC.

How do we explain that this totally unrelated sim has the same problems that concern @Spinelli about ISI-based sims, but to a much greater extent?

 

[passenger]

I can only assume @Spinelli absolutely hates that title because the issues of lack of rear grip and spins that seem to happen too frequently and with too-little to no ability to recover are at least 10 times worse in ACC than in AMS (for me, at least). Even the most skittish cars in AMS feel more planted and are more controllable than the cars in ACC.

How do we explain that this totally unrelated sim has the same problems that concern @Spinelli about ISI-based sims, but to a much greater extent?

AMS cars are more controllable than ACC´s. But, has the new ACC 1.0.7 update, the new tire model,
fixed this problem? I read someone said it is now easy to ride curbs with this new model... apparently
it offers more grip.

 

[Salvatore Sirignano]

I wish I was joking when I say with a G27 you loose about 90% of the feel compared to a Direct Drive wheel but I'm not..

The immediacy and strength (even at non strongman settings) just help you so much to catch slides, its almost cheating.

Thanks Niels, I'll write a letter to Santa Claus!

 

[Marc Collins]

AMS cars are more controllable than ACC´s. But, has the new ACC 1.0.7 update, the new tire model,
fixed this problem? I read someone said it is now easy to ride curbs with this new model... apparently
it offers more grip.

No, I am referring to the latest post-update situation. I hadn't bothered with ACC for quite a while, but thought I should try it after the latest update to the tire model (which sounded exciting). I always hated AC and ACC curb behaviour. Curbs are better than before, but I was immediately struck by the similarity to @Spinelli's complaints about ISI sims and what I was feeling in ACC. It was like the direct translation in a sim of what he was describing he always feels in ISI-based sims (and that I have felt, too, in poorly-implemented cars in ISI-based sims).

 

[passenger]

Yeah, I didn´t like it either when I tried it a while back, but I sense they are making good progress on this
subject. ACC is still a vey 'beta' Sim, so a lot of room for improvement. I´ll stick to AMS and AMS 2 when
it comes out, but ACC is a vey nice looking Sim, though very limited -just one Series and no mods. Let´s
keep an eye on it for future updates though; it could surprise us, you never know...

 

[Niels_at_home]

Again this is about AMS1. The tire model also only uses one contact point, but some of the issues I saw users have in ACC, where they even span out going straight over kerbs are odd. I also heard with AC really stiff cars could be physically unstable. This could be down to the numbers or more perhaps some physics instability at sudden large spikes of force or something..

The AMS1 (all rF1 derived sims) don't just look straight underneath the tire though, they look for the 3d polygon where a simplified tire shape has penetrated the most, and this becomes the contact point. So if you hit a high kerb ridge with the outside of the tire, it will probably become the contact point.

 

[Ghoults]

I think there is a small difference in how much of the physics is displayed visually via the car movement which is the camera movement. In non-isi based sims I can attach the camera rigidly to the car (disable camera motion completely) and drive without ffb and still see (feel) the tires flexing in corners in the view. It almost feels like all isi based sims have something missing or simplified in the way the view is connected to the physics. Like the tire vertical and/or lateral deflection is not part of the cam movement (so car and cam roll in corners is slightly wrong) or is somehow simplified. In other sims like lfs, iracing, ac I can see more of the car roll in corners for example as it is not just suspension deflecting causing the view rotate but tires also flex which adds a little bit to the view rotation and changes it too. And it makes it easier to sense the limit of the tires as well. In isi based sims (rf1, ams, rf2) that for me has always felt that something is missing.

This also meant that driving those other games was possible without ffb but in isi sims the ffb simply gives you more information as some of the info is not coming through the screen (even if the ffb itself could be worse). Considering the isi engine is very old I would not be surprised if it was overly optimized and stuff like tire deformation was not part of the cam movement whereas it is part of it in the newer engines. This same issue also makes catching slides a bit harder in all isi based sims because you can't see the tires loading up on your screen. You don't see how the tires start gripping and then the suspension loads up and the car rolls little more before the car goes into tankslapper. In isi sims that is all about ffb whereas in other engines you can see it coming as well as feeling it in ffb. It is not ffb issue as the same issue can be felt if you turn ffb off.

 

[Niels_at_home]

Only in replays is there a difference between physics and graphics, when driving it really isn't hard, the physics tell the graphics what state the car is in, and this gets drawn as far as I know.

We are derailing this thread...

There is a great risk of placebo and belief in simracing. People saying they can feel the tires flex and when the coffee is done through the view or force feedback, in my opinion has a degree of unlikelyness to it. Sure you are in better control of the cars in one sim versus the other, but describing this feeling is unlikely to be correct. there is so much code and physics going on, not to mention the parameters chosen by the physics guys. You can't just pinpoint things when driving, at least not specifically enough.

It can be a bit frustrating for both sides. There is probably no way you can be specific enough to point us towards parts of the source code that could be wrong. I suspect it is more likely a matter of your expectation of how a car should respond versus how it responds in game, plus driving skills, FFB wheel, pedals, input lag etc.

I'm pretty sure the engine would allow cars that handle way more the way you want them, it is so flexible depending on the paramters you put in.

 

[Timberwind]

I think there is a small difference in how much of the physics is displayed visually via the car movement which is the camera movement.

I will give Niels the benefit of doubt that physics are largely correct. But your explanation makes sense where the "sudden snap" feeling I get, despite being able to drift some cars extensively around the corner. Thanks for this idea.