It might be known to some of you that I despise(d) ISI physics. Since 1999 (F1 2000) not a single release or mod convinced me. Oversteer behaviour was so wrong, countersteering was.. counterproductive. And if, with a random combination of panic inputs, you thought you came out of a slide alive, you would be thrown off the track in the other direction. Driving a car needs confidence and I never had any with ISI games, even though I could 'learn' to be 'quick', doing top 50 or top 100 laptimes. It just never had anything to do with driving a race car.
LFS has always been a lot better at simulating a car 'over the limit'..
The last two days have been very strange. After reading the discussion on rFactor 'drifting' at RSC, I asked JTBo what his rFactor Volvo project was about and - Thanks! JTB! - I got access to the beta, which made my jaw drop. Normal car inputs work, car 'balance' (throttle / power) works.. A pretty good dream, except it was for real!
I liked the idea of the 79 mod for rFactor but also couldn't drive it. While they changed the lateral tyre curve, they also kept lots of 'bad' stuff in there. I simply turned off most non linear effects (load sensitivity / camber effect on grip / speed effects / heat / wear) to see what a 'pure' more realistic set of grip curves would do to the 1979 Ferrari. I also added 10..15% inertia (values seemed low) and put the COG 2 centimeters higher.
I know there are a few ISI .. dislikers here on the LFS forum, although I must've been up there in the top 3.. I made a short video of just some outtakes, showing proper car behaviour and control. I only spun a few times, and even that looks pretty good. I crashed only when really being very eager. I *never* disagreed! How is that..
I'm baffled / stunned / amazed.
It is still HORROR to actually build a car for rFactor, there are no proper physics tools to analyze suspensions, 'check' tyres etc. Its all very very bad in this respect. But guess what.. The physics engine works if you provide realistic, even though very basic, data!
I hope the video is very un-ISI like, as that sure is how it drives!
http://www.youtube.com/watch?v=cevxP5GtD68
Edit, it is 5:44 at night. That hasn't happened since Grand Prix Legends!
LFS has always been a lot better at simulating a car 'over the limit'..
The last two days have been very strange. After reading the discussion on rFactor 'drifting' at RSC, I asked JTBo what his rFactor Volvo project was about and - Thanks! JTB! - I got access to the beta, which made my jaw drop. Normal car inputs work, car 'balance' (throttle / power) works.. A pretty good dream, except it was for real!
I liked the idea of the 79 mod for rFactor but also couldn't drive it. While they changed the lateral tyre curve, they also kept lots of 'bad' stuff in there. I simply turned off most non linear effects (load sensitivity / camber effect on grip / speed effects / heat / wear) to see what a 'pure' more realistic set of grip curves would do to the 1979 Ferrari. I also added 10..15% inertia (values seemed low) and put the COG 2 centimeters higher.
I know there are a few ISI .. dislikers here on the LFS forum, although I must've been up there in the top 3.. I made a short video of just some outtakes, showing proper car behaviour and control. I only spun a few times, and even that looks pretty good. I crashed only when really being very eager. I *never* disagreed! How is that..
I'm baffled / stunned / amazed.
It is still HORROR to actually build a car for rFactor, there are no proper physics tools to analyze suspensions, 'check' tyres etc. Its all very very bad in this respect. But guess what.. The physics engine works if you provide realistic, even though very basic, data!
I hope the video is very un-ISI like, as that sure is how it drives!
http://www.youtube.com/watch?v=cevxP5GtD68
Edit, it is 5:44 at night. That hasn't happened since Grand Prix Legends!