Apex Modding GT3 - 2012 (v1.09 )

[David O'Reilly]

I think I improved that set again.
Dampers, rear camber, diff pre-load +1. (edit and rear toe in).
Brake balance needs to be 51.5 front (not sure I saved that).
Engine brake mapping can be savage if you don't watch your revs downshifting to lowest gears, but if you do its ok.
NB I have done little to nothing on brake or engine cooling or brake pressure. My task was really chassis balance.

https://www.dropbox.com/s/o91mzb2c3ne36zp/Apex458v930SUZ_BEST.svm?dl=0

 

[Lgel]

I think I improved that set again.
Dampers, rear camber, diff pre-load +1.
Brake balance needs to be 51.5 front (not sure I saved that).
Engine brake mapping can be savage if you don't watch your revs downshifting to lowest gears, but if you do its ok.
NB I have done little to nothing on brake or engine cooling or brake pressure. My task was really chassis balance.

https://www.dropbox.com/s/o91mzb2c3ne36zp/Apex458v930SUZ_BEST.svm?dl=0

Thanks David for your setup which drives well as is, at Sebring for instance.

Suzuka would be the perfect test track (in fact it is a test track also used as a racetrack), I used it at some point but it has three peculiarities that
make it less than perfect in my view:
- it is very technical which is both very good and not so good, because it introduces a lot more variability in lap times than a simpler track,
in the weakest link of the test chain, the driver.
- at least the version I know, is very smooth, and not perfect for damper settings.
- and the biggest inconvenient in my view is the fact that it is one of the few tracks with a cross over, so left and right turns are compensated, which is good,
but is not the case of almost all other race tracks. So, for instance and understeery setup will balance equally wear between left and right front tire, so you must take into account, that this unbalance in wear between front and rear end will be exacerbated on a standard track

An experienced tuner like you knows these facts and takes them into account, my remark is just a warning to other less experienced tuners.

Cheers.

 

[David O'Reilly]

LGEL,
First Point: Agreed that there are pros and cons for Suzuka.
Lately I have used it a lot due to it providing such a strong mix of Slow (hairpin) , Med (S-Curves)and High speed corners (Dunlop, Degna 1, 130R, T12) and the need for a stable car under brakes. Direction change is so critical at the entry to the S-Curves and Dunlop. It also provides a lot of cornering in a lap, the handling cant hide.
Your second point is very true in that from driver to driver it creates a fairly big lap time delta. (in old codemasters leagues we used a wet suzuka to rank drivers). However as a solo experience that is not an issue.
The tyre wear balance will not be so apparent left to right front but the wear balance front rear is still very clear, as are temp differentials. With this car I was seeing front temps as high as 200 deg and rear temps of 80 deg. One positive is that if you do see excessive front wear on one side only it is most probably from locking under brakes (which happened to me).
Finally fast bump dampers. I confess to not spending enough time on these on "Base set ups".
I usually only play with them at a "final track", the actual race track if taking kerbs is an issue. But I'm starting to work with them more.

Second Point: Default set. I suspect something funky was happening with the defaults. When I started at Suzuka I saw front camber -3.7, rear -1.9. The car felt nasty (very unstable entry understeer-exit oversteer) and started working downwards In my next session when checking against defaults I saw -2.4. (by this time I was at -3.0).
So I have a hunch that somehow it might have been picking up an old default set and this might be a part of the concerns some have raised. The default set that has -2.4 is by comparison, pretty good. Understeer of course as that is by design to make it accessible (easy to drive).

At any rate by the time I finished the car by my opinion no longer felt broken. It was driving ok and so the set up variables can achieve that.

Third Point Diff Lock I would be tempted to suggest this adjustment is left open. I know IRL it isn't free, but set as it is might not agree with all drivers and its a great adjustment and learning tool. (edit: Diff lock should be used to control behaviour in transitional phases of the turn, get chassis balance right first).
Thanks

 

[Lgel]

LGEL,
First Point: Agreed that there are pros and cons for Suzuka.
Lately I have used it a lot due to it providing such a strong mix of Slow (hairpin) , Med (S-Curves)and High speed corners (Dunlop, Degna 1, 130R, T12) and the need for a stable car under brakes. Direction change is so critical at the entry to the S-Curves and Dunlop. It also provides a lot of cornering in a lap, the handling cant hide.
Your second point is very true in that from driver to driver it creates a fairly big lap time delta. (in old codemasters leagues we used a wet suzuka to rank drivers). However as a solo experience that is not an issue.
The tyre wear balance will not be so apparent left to right front but the wear balance front rear is still very clear, as are temp differentials. With this car I was seeing front temps as high as 200 deg and rear temps of 80 deg. One positive is that if you do see excessive front wear on one side only it is most probably from locking under brakes (which happened to me).
Finally fast bump dampers. I confess to not spending enough time on these on "Base set ups".
I usually only play with them at a "final track", the actual race track if taking kerbs is an issue. But I'm starting to work with them more.

Second Point: Default set. I suspect something funky was happening with the defaults. When I started at Suzuka I saw front camber -3.7, rear -1.9. The car felt nasty (very unstable entry understeer-exit oversteer) and started working downwards In my next session when checking against defaults I saw -2.4. (by this time I was at -3.0).
So I have a hunch that somehow it might have been picking up an old default set and this might be a part of the concerns some have raised. The default set that has -2.4 is by comparison, pretty good. Understeer of course as that is by design to make it accessible (easy to drive).

At any rate by the time I finished the car by my opinion no longer felt broken. It was driving ok and so the set up variables can achieve that.

Third Point Diff Lock I would be tempted to suggest this adjustment is left open. I know IRL it isn't free, but set as it is might not agree with all drivers and its a great adjustment and learning tool. (edit: Diff lock should be used to control behaviour in transitional phases of the turn, get chassis balance right first).
Thanks

Yes, I think you had been using a setup from 0.92 with 0.93 version.

In 0.93 default values for camber are front 2,4 rear 1,8.

I remind everyone that setups for 0.92 don't work as intended with 0.93 as was specifed in the release notes.

Opening diff would allow more customization, but I feel would add complexity to the tuning of the cars that is may be not desirable, and not available on the real cars.

Cheers.

P.S.

I know you know that, but for less experienced readers, fast bumper settings don't only affect taking kerbs, they affect braking efficiency and stability on a bumpy track.

 

[Guimengo]

First of all cars have differential locked as in real life GT3

Exactly. If it's locked in real life, and those values in the setup page are the real life values, we gotta adapt to them. Either that, or make the unlocked Gear Ratio update unlock the diffs.

 

[cosimo]

Thanks Apex team, this mod is stellar! Car behaviour is fantastically deep and believable.

Thanks!

 

[Lgel]

Physics are open, if you want you can easily change those values.

Cheers and have fun racing.

 

[FuNK!]

Stupid question maybe: Do you guy have/use a bug tracker!?

 

[Led566]

...

I have to partly agree with what some others are saying here about sliding around as-well as general oversteer of everything but I think we cannot point the finger solely at the mod-makers...

I believe the issues ppl are experiencing are only 50% a function of the mod but also 50% a function of the core ISI engine. In any car, even pre-RF2, you can approach a corner, add quite small amounts of braking, add small amounts of steering lock, and that technique makes the car turn into the corner and want to continue turning. Sometimes, especially if you're inexperienced, you can find yourself all of a sudden on the very inside of the corner/road as if you have the quickest steering ratio ever (but it has nothing to do with steering ratios...

In short, in real-life, everything wants to naturally go straight, in ISI physics engines, everything wants to keep turning and turning (which can then lead to actual oversteer sliding instead of just turning-in more and more).

All you need to do is add just a small amount of steering lock, hold or dab small amounts of brake and you'll see that you can get any car around any turn without hardly turning the virtual tyres, without hardly applying any steering lock. You can get (eg.) Corvettes and Porsches around the final hairpin turn @ Adelaide, or the tightest turns @ Bathurst without hardly turning the tyres of the car or your steering wheel due to this core behavior of all iterations of the ISI physics engine. It's a far cry from reality. ...

This natural, very long-standing behavior in any ISI physics engine, combined with a car, setup, or driving style that is just naturally more pointy and/or oversteery (as Lgel stated he personally likes his cars [post 1675]), and...well, the issue becomes even more magnified.
...
It's a far cry from not only real-life but most other sims too....

The assumption you are repeating here (flaw in the ISI physic engine) has already been demonstrated FALSE by REAL race engineers.
At least in three different studies I know.
This is one of the most complete:
https://drracing.wordpress.com/2014/01/03/how-close-is-close-enough/
Scroll down to the steering lock comparison graph.
This graph completely falsifies your assumption.

 

[David O'Reilly]

The assumption you are repeating here (flaw in the ISI physic engine) has already been demonstrated FALSE by REAL race engineers.
At least in three different studies I know.
This is one of the most complete:
https://drracing.wordpress.com/2014/01/03/how-close-is-close-enough/
Scroll down to the steering lock comparison graph.
This graph completely falsifies your assumption.

+1
I have no beef with Spinelli but this hobby horse is geting long in the tooth now. The stated concerns pay no account to the physical impact of rotational inertia or polar inertia. IRL a car once it starts to pivot has two sets of inertia. Longitudinal inertia (making the centre of the car want to keep going where it is going) PLUS rotational inertia which wants it to keep rotating. The masses at the extremities play a big role here. For exampls a real life Porsche 911 once turned in ever so slightly under brakes then has a (rear) engined an gearbox carrying a lot of inertia that is outside of the perpendicular to the forces retarding the car. Plain language the rear keeps coming around). You can subtly reduce your turn radius by managing the weight transfer-the slightest braking effort will increase tun in. Once on throttle, to maintain a constant turn the driver is actually pointing his tyres the wrong way (counter steer). This is not ISI physics its real physics.
It also becomes very apparent that a car with substantial weight at both ends has greater polar inertia. (put two bricks on a broom stick near the middle, 50/50 weight dist, swing it around stopping and reversing your direction, easy yes, now move them to the ends. Same 50/50 weight distribution. Swing it around again-much harder that is polar inertia. A front engined rear gearbox car (like the ISI Nissan GTR GT1 is a good example). Once rotating it has a lot of rotational inertia. The cars with more rotational ineria need to be introduced to the corner with more planning. Once turning they like to keep turning. Mid engined cars less so.

This is how many drivers use throttle and weight shift to steer and also why in some tight turns at Bathurst (esp in a Porsche 911) you dont just turn your steering wheel to turn a tight corner. You make the slightest of introductions of turn in to the corner by getting the car out of perpendicular then you need to manage the oversteer as the arse tries to overtake the driver. Keeping that engine gearbox on the track with throttle and brake management is 50% of getting that turn done. Steering is just the fine tuning.

(Edit)..the grip of the tyres then determine how much you can challenge the inertia with where the car is now pointing. Your direction of travel will be somewhere between the two. (slip angle).

(Edit #2) and these above factors are why IRL and in RF2 the 911 needs far greater diff lock on coast than on power.

and yes I have driven real life front, mid end rear engined cars of 400hp on track in competitive events.
Just my kitchen sink science views thats all.
Thanks, Peace Out.

 

[Spinelli]

Sorry, it's highly incorrect and flawed and apparent in any ISI engine and not apparent in 99% of real life slides and spins and much less apparent in most other sims.

In real-life, you don't turn the wheelrim of a Porsche or Corvette steering wheel only 40 degrees to get around hairpin turns like the Adelaide hairpin, the tight turns at Barhurst, etc. What a joke. In real-life, you actually have to turn the steering wheel - and hence tyres - to make a car turn and go through a corner. I was also turning the steering wheel of the Stock Car about 1/4th as much as real-life Stock Cars turn their wheelrims at Indy, not only that, but I was using the absolute slowest steering ratio of all - "20:1 Slow". Wow.

In-game, you can hold tiny amounts of brake with tiny amounts of throttle and that wants to almost magically make the car want to continue turning more into, and more and more into, a corner as if every car has an inherintly programmed behavior of wanting to keep on turning more and more, sharper and sharper, the second any sort of minute amount of steering lock is applied. In real life, you actually have to turn the wheel, and therefore the tyres, to get around corners.

A real car hardly wants to (almost instantaneously) drive left (or right) extremely sharply across the track - it doesn't want to veer off across the track - just becauase you slipped the rear a bit.

Real cars require holding a steering angle for a half-second or second or so to be corrected, not instantaneously correct-then-return-back-to-centre like in ISI physics. There is supposed to be a moment in-between where you hold the wheel and hold the slide in a static-like state as the rear-rotarion stops, but hardly in ISI physics. If you try to apply realistic technique - rather than how the physics engine wants you to drive - then the car will most likely snap-overcorrect the other way.

This is easly visible from watching , you don't even have to play. It's a far cry from reality and the differences can also be seen/felt/experienced in other sims let-alone real-life.

 

[Marc Collins]

+1
For exampls a real life Porsche 911 once turned in ever so slightly under brakes then has a (rear) engined an gearbox carrying a lot of inertia that is outside of the perpendicular to the forces retarding the car. Plain language the rear keeps coming around).

I agree with this and the rest that you wrote. I also have some limited experience in all forms of cars, including AWD vs. RWD and FWD. I also don't want to further a discussion if people think it's annoying, but..................

For me, the Apex Porsche responds very well--exactly as I would expect it to in this regard. However, several of the other cars do not and despite NOT being rear-engined, actually respond and feel (FFB) more as though they are rear-engined than the Porsche. Camaro is a good example. The Porsche can be balanced on edge through a corner. It doesn't immediately loose control and grip with a small error. It can be MANAGED, the way any race car must be. If a car is not manageable at or near the limit, it will be left in the pits, unless the driver is suicidal. So my only frustration is why is the car that should be the scariest to drive, the old "doctor killer" 911, be by far the least scary?

And by the way, it is not at all fake, or too easy. The Porsche is deliciously challenging and you have to be on top of it at all times or it will bite you. That's the fun part--to tame it. I don't know what differences there are in the detailed physics, but from a pure blind-folded feel perspective, some of the other cars feel like they have a much higher centre of gravity and so little weight over the rear, they might be a pickup truck instead of a race car. Or, perhaps they have a rock solid broomstick sticking out the back about 3 metres with a huge weight attached to the end of it that only comes into affect AFTER you have approached the edge of adhesion...not before. It's hard to describe, but they have tendencies for the rear to suddenly lose it the way you expect only a rear-engined vehicle to.

It's strange to me--and as I have mentioned affects several ISI cars, so it has nothing specific to do with this mod. But if I could transplant the 799 "feel" (but applied that each car had it own proper weight distribution and handling characteristics) to every other car in rF2, I'd do it in a moment. I'd pay extra to do it.

I helped a tiny bit to test these cars. And has been pointed out, we used the Portugal track--for many good reasons. It has always been my personal test track since Day 1 of rF2. I have thousands of laps there. One of my favourite tests/spots that show what a huge difference there is in car sim physics is when you run just a bit wide on the first turn after the huge breaking required from the main straight. This is easy to do since you are coming down from top speed and have to hit your braking spot perfectly. Watch and feel very carefully what happens to each car as it runs up and possibly even a bit past that curb at the left side of the track. Try accelerating (from the gentlest amount all the way to pushing it as hard as possible on various runs to compare) while this is happening. This is one area that separates the good modded cars from the not-so-good. I will only say that the 799 nails it for me in terms of authenticity. Most other cars in this mod and any others do not.

 

[hexagramme]

+1
I have no beef with Spinelli but this hobby horse is geting long in the tooth now. The stated concerns pay no account to the physical impact of rotational inertia or polar inertia. IRL a car once it starts to pivot has two sets of inertia. Longitudinal inertia (making the centre of the car want to keep going where it is going) PLUS rotational inertia which wants it to keep rotating. The masses at the extremities play a big role here. For exampls a real life Porsche 911 once turned in ever so slightly under brakes then has a (rear) engined an gearbox carrying a lot of inertia that is outside of the perpendicular to the forces retarding the car. Plain language the rear keeps coming around). You can subtly reduce your turn radius by managing the weight transfer-the slightest braking effort will increase tun in. Once on throttle, to maintain a constant turn the driver is actually pointing his tyres the wrong way (counter steer). This is not ISI physics its real physics.
It also becomes very apparent that a car with substantial weight at both ends has greater polar inertia. (put two bricks on a broom stick near the middle, 50/50 weight dist, swing it around stopping and reversing your direction, easy yes, now move them to the ends. Same 50/50 weight distribution. Swing it around again-much harder that is polar inertia. A front engined rear gearbox car (like the ISI Nissan GTR GT1 is a good example). Once rotating it has a lot of rotational inertia. The cars with more rotational ineria need to be introduced to the corner with more planning. Once turning they like to keep turning. Mid engined cars less so.

This is how many drivers use throttle and weight shift to steer and also why in some tight turns at Bathurst (esp in a Porsche 911) you dont just turn your steering wheel to turn a tight corner. You make the slightest of introductions of turn in to the corner by getting the car out of perpendicular then you need to manage the oversteer as the arse tries to overtake the driver. Keeping that engine gearbox on the track with throttle and brake management is 50% of getting that turn done. Steering is just the fine tuning.

(Edit)..the grip of the tyres then determine how much you can challenge the inertia with where the car is now pointing. Your direction of travel will be somewhere between the two. (slip angle).

(Edit #2) and these above factors are why IRL and in RF2 the 911 needs far greater diff lock on coast than on power.

and yes I have driven real life front, mid end rear engined cars of 400hp on track in competitive events.
Just my kitchen sink science views thats all.
Thanks, Peace Out.

A very nice and detailed explanation, where even the layman can comprehend what's actually going on.

Looking at what engineers and physics experts has written about the ISI physics, I always figured that Spinelli's theory was quite a long way off from being correct.
When data backs up the claims of there not being any sort of fundamental flaw in the engine, I guess that's that.

 

[Led566]

sorry, it's highly incorrect and flawed and apparent in any ISI engine and not apparent in 99% of real life slides and spins and much less apparent in most other sims.
....

Is it apparent or is it real?
http://www.collinsdictionary.com/dictionary/english-thesaurus/apparent

If it is real can be measured and compared to real data.

Sensations, camera cars, "once the second cousin of my friend Joe drove a Real porsche" and all these sort of "apparent" things don't apply.

Please.show.data.

 

[andrea_274]

Well, i think mod is really a great job. Just a little thing. The dashboards of some cars are worse than the .928 version. Expected some improvement in future releases?

Thank for your passion and your hard work

 

[Flatspotter]

Fantastic mod, really enjoying it! Thanks for all your hard work!

A couple of bugs: The telemetry for the tachometer lights (bar graph) doesn't work on the 854 and the 799. This causes the rev lights not to work on my Fanatec wheel. Also, the right side wing mirror image is backwards on the 854.

 

[Hash]

Awesome, awesome mod, seen a couple complaints about "ICE Grip" which I'm always suspicious of people's driving skills when they talk about "ice skating", the cars feel extremely responsive and for good or bad respond very well to user input, and unlike the vast amount of GT3 content in sim world, these are by far the most "fun", can really get them wild if want to, as opposed to the self driving gt3's found elsewhere.
https://www.youtube.com/watch?v=a_pMzsqGnpc

 

[deBorgo83]

Only tried the McLaren so far but just wanted to say that I'm much happier with it than I was with v.928. Not sure if that's the new default setup, physics changes or both but I drove the car at Nords and thought it was excellent. A bit of understeer on the power (as intended, I believe) but I felt very comfortable to push the car. And Tosch's graphics updates are wonderful!

Must try other cars but so far I must say, great work guys. Thanks very much.

 

[Spinelli]

Is it apparent or is it real?
http://www.collinsdictionary.com/dictionary/english-thesaurus/apparent

If it is real can be measured and compared to real data.

Sensations, camera cars, "once the second cousin of my friend Joe drove a Real porsche" and all these sort of "apparent" things don't apply.

Please.show.data.

It's completely different from real-life as-well as most sims and has been since at-least F1 2002. If one can't see the long-standing, 15-ish year weirdness in some aspects of vehicle behaviour in the ISI engine then I guess I am more perceptive than him/her. Not that I'm the only one, I've talked to many who have agreed with me online in rooms, my friend in real-life, and I used to see lots of stuff about it back in the mid 2000s during the RaceSimCentral days.

Obviously people are in denial and want to come to try to think of anything they can to dismiss what i'm saying due to holding on firmly to what they believe and/or their belief system. It's called cognitive disonance. I advise some people to look up that term.

 

[Ernie]

Could someone give me an advise how to drive these cars fast, but without ruining the front tyres? I had a fun event yesterday at Mosport, and the front tyre wear on every car i tried was much higher than at the rear axle.

For a 30min race i was driving the Z06 (using the default setup with slightly stiffer rear ARB and ABS+TCS on low). I have to admit that during the test my driving style wasn't really careful on the tyres. But even with a lot of oversteer and drifting around corners the rear tyres were fine after 30min, but the front tyres were nearly dead.