bug physique suspension

[lucarace]

mod-edit:

 

[Joao Alves]

Translation for those of you who, like myself, don't speak the language the video is subtitled in, and would still like to understand what the author is trying to share, but couldn't be bothered to translate himself.

Caution: Google Translate, but I think I got the gist of it.

• Subtitle 1 - The following video may hurt the feelings of some of our viewers.
• Subtitle 2 - Suspension.
  
• Subtitle 3 - So a car in Assetto Corsa reacts to a big bump at 100 km/h.
  
• Subtitle 4 - Curious result. I for one was expecting something like this.
  
• Subtitle 5 - Which is what happens in a real simulator.
  
• Subtitle 6 - Let us see what happens at a lower speed.
  
• Subtitle 7 - No, it's not the same video as before. Its another at 50 km/h less. The car even jumps a little more.
• Subtitle 8 - Now let's try 25 km/h.
  
• Subtitle 9 - We can go over a bump at 100 km/h and the car is not affected, but at 25 km/h it almost falls apart.
  
• Subtitle 10 - Let's see what happens at 15 km/h.
  
• Subtitle 11 - Good, not bad, suspensions look very natural. A car of more than 1000 kg rising almost half a meter off of the ground looks almost realistic.
  
• Subtitle 12 - Now let's see how the shocks react to compression and gravity.
  
• Subtitle 13 - Strange, nothing happens! Maybe from the low speed physics?
  
• Subtitle 13 - That is how well modeled suspension works.
  
• Subtitle 14 - Is the suspension model better in Assetto Corsa or rFactor 2? Everyone judge for yourself.
  
• Subtitle 15 - For me the answer is clear.

 

[Joao Alves]

I think its a valid point, and after trying if for myself, I think that it does in fact need some further discussion. Unless it happens to be like that because we're using a pre-damage suspension model. Since when I tried it for myself I noticed that the body work of the car can pass through the road surface when the car lands sideways. The tires are bound the the physics mesh, but the body work seems to ignore it as of now.

That being said, I think the presentation could have been more respectful and without the ISI fan-boy-ism at the end...

 

[Chris]

You cannot accuse the Assetto Corsa suspension model of being broken simply by comparing it to some grainy YouTube footage, and some rFactor 2 gameplay using a completely different car.


@Joao M Alves
Thank you for translating the captions in the video.

@lucarace
In future, please remember that when posting here at RD to post in English. This includes videos.
If something requires you to speak anything other than English, please do so via the Private Message system.

 

[Chronus]

That being said, I think the presentation could have been more respectful and without the ISI fan-boy-ism at the end...

Leave the "AC fan-boy-ism" bandwagon and look at it objectively - the video author is either right or wrong (dispense with sidebars). From what I know from Aris, he's perfectly able to do just that and have a in-depth discussion of the work he is involved in, so perhaps you and others could too.

As for the issue the video author noticed: it is odd and certainly Kunos knows about it or will look into it.

 

[Joao Alves]

Leave the "AC fan-boy-ism" bandwagon and look at it objectively - the video author is either right or wrong (dispense with sidebars). From what I know from Aris, he's perfectly able to do just that and have a in-depth discussion of the work he is involved in, so perhaps you and others could too.

I'm not on a 'fan boy bandwagon'. You yourself quite literally just said to look at it objectively. So why does the OP get to have slams to AC and "ISI 4 EVER" in his video, but when I say to drop fan-boy-ism I'm told to not defend Kunos when I wasn't?

I'm very capable of discussion and all for it, as I posted at the beginning of my comment.

"I think its a valid point, and after trying if for myself, I think that it does in fact need some further discussion."

 

[lucarace]

This video aims to show the problems in the future, with modeled by moddeleurs cars when physics has serious shortcoming.I hope it will be resolved by kunis!

 

[Chronus]

lucarace, two things:
- as Chris said, posting the video in English would help in fueling discussion
- if the intention is to help Kunos (I am not saying it is not), maybe establishing a comparison to a known competitor (also under development) is not well advised

At least, you raise a interesting point. Anyway, as I said, the developer must be on top of this already and either an explanation will be provided or the anomaly will be squashed in due time.

 

[Joao Alves]

To clarify how I feel about this before it gets heated from my posts.

I like that this was posted and brought up. Its not a small anomaly and should be brought to attention. I stand by what I've said up towards the top though.

If Aris would be able to post on this it would be great as his posts usually are, and to shed some light on the situation.

This is also the first I'm hearing/seeing of this, so I'm not entirely sure if the devs are aware. If not, even better, and if they are, that insight would be great to have.

 

[lucarace]

the translation was made by,Joao M Alves,the video speaks for itself,each problem has its correction, the future will tell.

 

[Rupe Wilson]

A picture says a thousand words correct
I would say even with all its so called flaws rf2 is a more complete sim at this point in time.
So comparing the 2 is not really fair and can always leave to issues but its still good to see a side by side comparison.
lets not forget both are still in development. but from the footage shown you can see a difference in the 2 sims. And as pointed out by others we are looking a 2 different cars.
The problems with vid is that we have no clue how the suspension has been set up
and we all know how easy it is to manipulate vids.(not that im saying this has been done.)

If you asked me to test something, this is what id be doing not just driving round a track to see how fast i can lap.
i drive all over the track over every bump looking for such strange reactions, so for me the vid gets a big thumbs up. and this is what i would be looking for.
And looking at the rf2 footage on first looks i would say rf2 does a better job, going over that kurb, but that car should soak up the bumps better.
But i would also hope to see the BMW doing the same good job. at some point in its development
We all know AC has lots of work to be done on it in many different aspects, its far from perfect and far from finished.

Give it a year and we should see some big improvements.
We only have to look were rf2 was this time last year and were it is now, its like a different sim.

Well at least you did not use Mario kart to compare AC with

 

[Georg Siebert]

I mostly agree with Rupe. My take is,

• These are video games, both still in development
  
• These are different cars, not comparable
• The default setups for AC cars are not optimal/good/whatever adjective with positive connotation
• We don't know the bound/rebound, etc settings for each cars, not comparable
• The physics may be different for each game
• These are racing games for cars driving on tarmac at high speeds, not 15 kph, not Rally games
• Maybe the programmes at Kunos had to make a compromise between high speed and low speed driving physics or with the physics engine in general, we don't know

Bottom line, currently the car behavior at racing speeds and car setups make sense and follow logic to the best ability of the simulation. Modifying bound/rebound by one values has tangible effects.

Edit:
I want to reiterate bullet point #1. I think there are many people, who would like to drive iRL Formula 1, GT3, Endurance series, you name it, however some have no time, money, opportunity, talent, etc to follow through on that, myself included. Would I like to race Formula 1 with Vettel & Alonso? Put me through the proper training and here we go. The 24h at the Nordschleife? Hell, yeah - I'll bring coffee and a blanket.
In order to get as close as possible we play video games, different games from different developers. RF2 from Image Space Incorporated or Assetto Corsa from Kunos. I like Dallmayr coffee over Starbucks. Pick one, it's a matter of taste.

 

[Stereo]

The main difference I see which can definitely be attributed to differences in the game (rather than setup changes) is the way the 2 games determine wheel-ground collisions with the sausage.

If you look at very low speed footage, RF2's tires never clip through it - whereas AC's do.

AC has a single point of contact with the ground for each tire, which is the bottom outside edge of the tire. I'm not at all sure what RF2 does, but clearly something is different - the result is that in RF2 it's a much shallower bump for the tire, and doesn't throw the car around in the same way.

On the left is tarmac, on the right, grass - which is offset a little bit downward from the kerb, and makes AC's tires bump very suddenly. The upper lines show the path the wheel takes - RF2's is rounded, AC's jolts upward.

Just based on one video it's not possible to determine what's different in RF2 - is the tire picking up multiple contact points around its circumference, is the physical model of the sausage more rounded, etc.

In terms of making the cars act silly, running across that kerb from the grass side is going to showcase the worst of AC's assumption - the tire suddenly bounces upward a couple inches in a completely unnatural manner, which loads the springs suddenly and springs the car into the air even at very low closing speeds.

So yes, AC's got a flaw in it. Will it affect racing? Only if the track authors ignore the issue and make the racing surface too bumpy.

 

[cerebus23]

The italia s3 on snoopys nord is a shining example of something wrong with the suspension on that car or the physics in general. A road car simply cannot have so stiff a suspension that a bump will throw it off track. That car simply launches off the bumps in that track.

If there is a flaw with the actual physics and how the tire "surface" reacts with bumps i hope the devs do address it.

 

[Dinca Andrei]

The italia s3 on snoopys nord is a shining example of something wrong with the suspension on that car or the physics in general. A road car simply cannot have so stiff a suspension that a bump will throw it off track. That car simply launches off the bumps in that track.

If there is a flaw with the actual physics and how the tire "surface" reacts with bumps i hope the devs do address it.

Hahahahaha..you really give us as a shining example a track that its not laser scanned (no offence to Snoopy,its a great effort and great work) to point out the suspension problem?...amaizing....the problem, like Stereo posted,is totaly different...collision

 

[cerebus23]

I am just saying something seems incredibly off with how that car encounters unexpected bumps. official track or not the s3 italia is the car that has the most problem with that to a redicilous degree.

 

[Georg Siebert]

I wouldn't have put it so bluntly, Dinca, however I have to agree - modded or converted tracks can't be the benchmark, especially not the Nordschleife, which is notoriously bumpy.

Additionally, we already know that the track mesh is something different from the surface we see - there was a useful thread somewhere....

 

[Dinca Andrei]

Sorry for the "blunt" hehehehe

 

[cerebus23]

Be as blunt as you want, easy enough to test take a few cars get an official track turn off penalties and start ramping stuff off curbs.

See how different cars react then we can compare. maybe my project later this eve.

 

[Dinca Andrei]

I understand the problem,i didnt say its not there,but i also understand its comming from something else then suspension,and of course the cars handle differently..i`ll really question it if didnt act different