As far as i was aware the Mazda was the only car to have the new tyre/downforce physics. I can see what Dave is getting @ with the rfactor feel to the Vette, it maybe setup based or something we haven't figured out yet, but i dont think its quite as bad as rfactor.
Having said that i hope they put some effort into getting it to feel right.
You are right. Grant Reeve confirmed it in a post about the increased grip on grass in the Mazda:
"The grass is different on the Mazda (and C6R and Dallara) because they use the slightly updated tire model. The model is still aimed at solid surfaces, and whatever happens on grass/dirt/etc is not completely correct either way, it's just that grass just happens to feel nicer."
I really really like the new tyre physics, I really liked the Mazda before, now it's just wonderful... Dave, instead of taking the unreasonable setups of some aliens, have you tried recreating RL-setups and went from there?
The default setup of the corvette had a certain unrecoverable snappiness to it, but a few tweaks to steering, brakes and diff made it much better in my opinion... Also, can we really expect radial racing slicks to have a very wide limit like road tyres? I concede though that it can be better - I suspect tyre frequency, its dampening and flexing is just missing which would explain the need for those incredibly soft setups the fast guys are running...
Of course I don't know, but I can be pretty sure. We've had many threads on these forums with Todd Wasson's expertise, who has clearly explained how LFS is pretty close to real tyres regarding slip angles and lateral force and such. I'm well aware things can be improved, but at the moment it feels to me like iRacing is at one side of the realism line, while LFS is an equal amount the other side, which makes it difficult if you want to race on both sims equally.
Thats the thing. I find I have to stick in crazy values in order to get rid of the natural understeer most of the cars display. I'm looking forward to seeing what the community can come up with setupwise.
And now I do. Having driven a couple more hours in the thing around Barber, my opinion has changed. It's a complete and utter unrealistic piece of shit to drive anywhere near the limit. No really, I can't stand it anymore.
The server that I was racing on pretty much confirmed my thoughts for this car, and many others (Mazda, Corvette, to name some). The server was spin central. I span many many times myself. Since when does a car result in an instant spin the minute you have some oversteer? OK maybe you can save it 20% of the time, but normally it feels like luck alone. I'm in disbelief at the number of people who think it's realistic to have an unsavable spin so easily so often.
I really am starting to doubt how much "real world" data iRacing have available to them. Recently they reduced the engine braking effect in a few cars considerably. If they had lots of data available why would they get something so vital completely wrong first time round? I wouldn't be suprised if quite a lot of physical information is just made up TBH. One thing in particular though which stands out for me ATM is that everything in iRacing seems to have far too little inertia. Now I could be talking complete bollocks here, but I'll carry on anyway. To start with the engines rev from idle to redline in neutral extremely quickly, take the Jetta for example, a lazy diesel engine which probably has a fairly hefty flywheel. Then there is the general feel to some of the slick tyre shod cars, such as the Mazda and the Dallara. It's as if they have very very low reluctancy to suddenly switch direction, especially in rotation. To me it seems the moment of inertia for the cars are far too low. Would this explain the apparant snappiness in a lot of the cars? Maybe if the cars had higher moment of inertia in the yaw direction they wouldn't swap ends so violently and quickly? Do physics engines even have specified inertia values or are they calculated based on the spread of mass etc throughout the vehicle?
In the meantime I'm sticking to the Skippy since that's currently the most exciting car in the sim which feels even remotely realistic to me.
I never agreed with you on the mazda Dave, as I never liked the thing. But now that you look at all the cars the same, I'm 100% agreed with you (if that makes sense). The inertia thought is a good one, but I think that its just sucky tire physics.
The amount of data they get probably varies a lot, but I don't think engine braking is such a vital issue and educated guesses will do fine if there's not enough data to calculate it accurately. One of their engineers admitted that they had calculated the engine intake wrong in the SRF, the only car of the few with engine braking changes that had a significant change in handling. But even with SRF the true effect of the changed engine braking remains to be seen, since Grant Reeve said that he had accidentally included some work-in-progress tire physics for that car in the update, and according to him the car behaves very much like it used to after switching back to the "correct" tire physics.
The fact that they have such accidents might not increase anyone's confidence in them, either, but at least they admitted it right away and a fix is likely coming as early as Monday...
Corvette is definately more forgiving and easier to slide around than Dallara, but both are very much recoverable from slides. I wouldn't expect them to be easy as Skippy to slide around considering differential is completely different and having both rear wheels spin at the same time will result in much quicker snap-oversteer.
"There is an issue with contrast and color balance. We're working on graphics updates that deal with that, and true HDR lighting, DOF blurr, etc. We've had higher priorities right now, but all of that stuff is on our todo list. I've been able to play around with the contrast and color balance and it makes a huge difference.
Our track surface shaders are very lame right now. :-) The only objects with shaders doing anything complicated are the cars.
I don't have to be Will Power to be able to tell you an opinion of whether or not a virtual car that I've never driven IRL is realistic or not. More often than not the "feedback" from real drivers regarding how realistic a sim car is is often not very accurate at all. Most real drivers are often not experienced enough with sims to know the subtle differences between ones which are currently available, as they don't have the time I guess to really sit down and get fully connected with the difference in feel at a basic level, and really get close to the limit. Regardless of how realistic a sim is, your brain will not be relying on the same feedback you get in RL in the sim. Obviously there are exceptions, say Dale Jr, who has got to the level where he is just as capable in a sim as in real life, but obviously he has a reason to be careful of what opinions he has of how the cars drive. Most real racers try a sim and as long as the feel is vaguely correct, the sound is accurate, the graphics are good and it will go round the track, it's therefore brilliant and as accurate as could be on a PC.
To me, given my experience trying various different sims and noticing the subtle differences each one has, comparing that to my limited real life experience, as well as studying many many onboard videos of various forms of racing, I think my opinion is just as valid as Will Powers.
P.S. I did some testing in rFactor which is obviously moddable. In the hdv files where all the physics data is stored, there are 3 values for rotational inertia, presumably each one assigned to a rotation axis (pitch, yaw and roll). These are independant (I think) from the overal vehicle mass, which only affects non-angular acceleration of the car. I tried reducing the values, which would indicate most of the mass of the car is very near the center of mass, and understandably the car became very very responsive to changes in direction. When pushing near to the limit, the car was very very tricky to balance as you had very little time to react. Sounds like iRacing to me. Likewise when I tried larger values, which would simulate the mass of the car being spread out away from the center of mass, the car was less responsive, but much much more controllable and less snappy. I'd like to know how realistic values of rotational inertia from real cars are obtained? Because I'm getting more and more convinced iRacing has this way off currently...
If you want to test this for yourself I'll explain. Basically take any iRacing car, say the Mazda, to the skid pad as it loads quickly and is completely flat. Then bring the car to 100 mph or so, and start weaving from side to side (drive lengths across the diameter of the skid pad not around it) until you find the cars natural "weaving" frequency. You should feel the weight of the car moving with you. In the Mazda you'll find this is at a very very high rate, the car changes direction extremely fast.
Now try in the Skippy. You should find that the natural side to side motion the car naturally falls in is much slower. Why is that? Does the Skippy have significantly higher rotational inertia IRL than the Mazda? I doubt it, but there is no doubt that in iRacing the Mazda has extremely little. Try the Corvette out as well, a pretty hefty GT car with a huge V8 out at the front. Surely this means it would have a high moment of inertia? Obviously not in iRacing, as it changes direction almost on par with the Mazda. The Solstice on the other hand, a smaller car, smaller engine, yet this has a calm whale like weaving action in comparison. Compare with other sims if you want.
The cars that seemed more responsive all have wings at the back and much stiffer suspension than the Skippy or Solstice, as well as slick tires unlike the two. I think inertia has only a small part to play in the end result. This article has a couple of quotes about inertia values: http://issuu.com/autosimsportm ... t_vol4_num4_irmastersmass
Good thoughts, although suspension will play a huge part in that too. The solstice weighs almost 2900lbs, how much does the C6R weigh? I would assume from common sense that the C6R would be a hell of a lot more responsive than a roadgoing Solstice regardless, so I'm not sure about that part of your thesis... :doh:
But in terms of moment of inertia of the 2 cars, I can't see them being too different. The Corvette is lighter, but at the same time it has a longer wheelbase. At the moment the Corvette in iRacing is more responsive in direction change than a Kart for me. Of course it's a terribly crude way to test the theory as like you say suspension plays a part etc. It would be interesting to know how values of rotational inertia are obtained for sims, because TBH I wouldn't be suprised if iRacing were making do with poorly education guesses ATM, and as I said it makes a huge difference in rFactor for example.
Based on the article I linked to, for cars which don't have pre-measured inertia values provided by the manufacturer they probably have a basic equation that uses all the known masses and weight distribution, and the resulting inertia value is refined by comparing it to the inertia and masses of those vehicles for which they have accurately measured data.