Automobilista 2 Custom Force Feedback - Overview & Recommendations

I look forward to trying 3.7 later.
while some cars felt marvellous on 3.6 , some others felt rough.
I just updated my thrustmaster firmware to 2021 3.2 as well so will be interesting if that changes anything more than just compatibility with new products.
it felt a little smoother in idle state

 

Hi all,
I must say I have only just discovered this custom force feedback option from a YouTube video that I watched the other day. And it has brought this sim alive for me to the point that I want to play it daily rather than once a month.

Is there a guide anywhere as to what settings in the file do? I apologise if it's in the forum thread somewhere but 130+ pages I got lost searching. I want to do some reading up before I was to go digging into the file.

The things that I want to change:
how heavy the steering gets when I'm on the brakes. It's a little off putting when I transition from brakes on with heavy steering and then getting off the brakes the steering feels suddenly very light.

I am also having trouble picking how much steering lock I need to apply in certain corners. It seems like there is no drop off in force when I push the front tyre past the ideal slip angle. The limited real world experience I have of racing cars I was taught is you were steering and then the wheel felt lighter you had gone past that ideal amount of steering lock for a given speed.

If I were to be able to tweak those two factors a bit then it would be the best ffb I have ever used

 

Hi Ben,

Which wheel base do you have and what are your in-game ffb settings?

The brake is simple to adjust, just lower "brake_feel" in the file to maybe 1.3 or 1.2.
The reason for having the brake feel stronger is @Shadak pointing out it is one of, if not the most violent force when racing irl, so it adds to the immersion.

 

I have a dd1 set to Max torque of 18nm on the wheelbase itself
In game I'm running 40/2/15/30

Yea I get the immersion rationale for the brake force it's just a bit too much of a jolt to the system if I want to touch the brake a little to help get the nose into the apex I end up missing the apex anyway because the steering suddenly was twice as heavy and I didn't turn the wheel enough.

 

Have only briefly tried stock file but feels good . a little different, but good.

i find i need to increase lfb to get the wheel more free moving , which is reverse to previous for me. but then this does have the other centering method .
At low setting it’s more rubbery in centre.

it seems to fill the wheel well and don’t seem to get the looseness some cars used to have and it almost physics defying the way it resists tank slap like peter mentions.

It’s probably the best the camaro ss has felt and been manageable.
i usually really struggle with it .

Much more testing and settings optimising to do later .
track bumps seem low .
I also feel like the TM firmware update has smoothed something on my wheel , as default ffb is now a little less buzzing with rod noises.

 

Ben schippers
If you've just discovered these custom files, you might not know that with them (only for the called RFUKTOR file series):

- The FX slider is independent and no longer depends on the gain.
- The damping slider, now serves as power steering, the higher it is set, the less force there is when cornering.

 

Is it pure placebo or am I feeling more from the rear end with this compared to 3.5? Running both 3.5 and 3.7 as they came at the moment.

 

@Karsten Hvidberg @Shadak - The 3.7 file is very nice I'm taking a little more time because the tuning is very subjective and talking only about value preferences seems repetitive to me and (this time) I want to analyze very deeply (step by step) and check all the evolutions from 3.5, to 3.6 and finally to this 3.7 to be able to do (hopefully) a good mix for the low end wheelbases that enhances the pleasure that users have had with the 3.5 but with the characteristics had up to this last file!!!!
But let's talk about 3.7:
01) Regarding the new items it can be said that they are very interesting, that is, the (bump_feel_follow_weight 0.75) and the (brake_feel_follow_weight 0.75), while the second does not create particular discrepancies feeling good both on cars/light forces and on cars/heavy forces, the first creates strong jumps with the car with more downforce and weight that is the F. Ultimate and also in the straight, in fact in the presence of the bumps, the jolts are perhaps amplified too much ... for what has been said above I think that I will bring the second item to 0.80 and the first at 0.40 (I just have to ask you a question that I will ask at the end of the post);
02) The slip_crossover_pos curve has been very close and from 0.1 it has been reduced to 0.06 in the rFuctor 3.6 and 3.7, perhaps this is very useful for DDs that are connected to the shaft, but with gear wheels I think it is better to 0.1 by starting the effect a little later since we already have a lot of friction and softness on the center or 0 force (we'll see);
03) Another thing I want to verify is the reduction of damping from 0.15 of file 3.5 to 0.10 of file 3.7, it may seem little but we are talking about a reduction of 33% which is no longer modifiable by the cursors as DP now manages Power Steering;
04) I have little to say about the effects, I have seen the changes in the scales and the one pronounced in the tear but they are subjective and easily modifiable.
WE WILL SEE WHAT WILL COME OUT

Clarifications required:

01) What value can I modify in this code to reduce only the scale without reducing the relative effects on which they apply??
(bump_weight_scale (/ 500 (/ (+ corner_FL corner_FR) 9.8)))
(bump_weight_scale (+ 1 (* bump_feel_follow_weight (- bump_weight_scale 1))))

EDIT: I would like to reduce only this relationship with the micro bumps ... the rest I feel good
(micro_bumps_scale (* micro_bumps_scale bump_weight_scale))

02) Only for test, is it always possible to insert the trail both as centering and as force_scale???

Now is:
(centering (blend (- 1 trail_scale) 0.22 0.5))
(force_scale (* 1.5 (/ 0.22 centering)))

Is always possible this for test??
centering (blend (- 1 trail_scale) 0.22 0.5))
(force_scale (blend (- 1 trail_scale) 1.3 0.8))

I notice a sequence of values familiar to me despite not having a DD ... that is the LFB at 2, and a DP (power steering) reduced by 10 compared to the GAIN ... if I may ask, keep the LFB at 2 to avoid oscillations??

 

@Stakanov I tried a bunch of different combinations including the power steering in increments of 10 all the way from 10 to 90 and it wasn't able to change the heaviness on brakes or me not being able to feel when I have gone too far on steering lock.
So one thing I did try was to copy your recommendation to rule out me using a silly set of numbers. I wasn't having any oscillation issues at lfb up to about 10

 

@Ben Schippers thanks for info ... same here with lowend G920 ... in this last 3.7 the LFB the range tollerance for the oscillation is 0-10 and especially with cars like Formula Ultimate 2019 and F10 Gen2 i go down to 0-2 ... instead with other cars I can easily raise but you are DD and is correct low value

 

@Karsten Hvidberg @Shadak

I think this is the only change I will make of much note. Its a pretty special file on the cars ive tried so far.

# Trail adjustments
#

(centering (blend (- 1 trail_scale) 0.22 0.5))
#(force_scale (* 1.5 (/ 0.22 centering)))
(force_scale (blend (- 1 trail_scale) 1.3 0.8))

This way feels more lively and immersive, It gives me more front end and tire feel and more sense of the slip angle.
It does make power oversteer a little more alive and slightly more caution is needed , but the other way feels too docile and almost like cheating, such is the way it self controls the oversteer.

But even with this force scale method, ive now got control and can actually enjoy driving things like the Camaro SS.

edit..
I may change my mind on this yet, some cars are a little loose..
And I’m find more lfb again is improving things , so still need to play around.

 

I'm an idiot ... don't laugh at the outburst, but I'm pretending to be in the mirror and tell myself to my face ... the beauty is that I'm at work and I can't test how stupid I am!!!

@Peter Stefani, since you also have the G920, I need your confirmation later ... not if i am stupid or not ... but on a parameter that I was wrong to convert ... even if I were a child at asylum!!!!

At least I hope....

 

Ask away anytime my friend. I hope there aren't any wrong parameters, but if there is, it didn't break anything so far. LOL

 

Thanks @Peter Stefani ... no it's all correct now I just want a confirmation ... I know that you don't use the tightening now and I don't want you to change your mind or judgment ... I don't even ask you to know the satisfaction ... . let's get to the point my simple question is this:

With these 2 lines and with these small values change your emotion with your file without changing anything else ??

(tighten_range 0.0050) #Use for lower end wheel bases.
(tighten_falloff 0.0005) #Use for lower end wheel bases.

The question is extended to all (DDs included) ... not just to Logitech owners ... the answers I would like are:
- I don't notice any differences;
- there is just something different but maybe it is a placebo;
- the difference is minimal, nothing that cannot be realigned with a different little tuning.

Try it with all the cars you want, even the most unpleasant ones to you (except the karts, for now, leave it alone) .... on all the cars I have tried my answer is among those proposals ... I am only interested in making sure if someone is creates problems otherwise I have to change direction ... then I'll explain let me know

I leave you just a clue:
(output (if output (if (+ tighten_range tighten_falloff) (tighten output tighten_range tighten_falloff) output) output))

 

@Stakanov
I haven't used TR/TF since 01/23/2021 in version 6-6-5-43_5-2.
And the values I used were TR_0.04 and TF_0.03

So, I'll do your request for experiment and see if I can determine any plus or minus opinions, and let you know ASAP.

 

Thank you dear I know my purpose is not to convince ... I am only interested if you feel differences and if they are non-invasive to be ignored ... take it easy

 

@Stakanov
I check wheel check graph with these values below, and graph is almost perfect straight line down to zero axis points. Wheel used for test is G920.
(tighten_range 0.002
(tighten_falloff 0.0002
In-Game feeling seems to to have no bad effects to any feelings. It did feel like there was a very slight change of improvement in center tight at top dead zero position of wheel.
I think it's is safe to use these values in the Logitech Series. And most likely ok to use in other low end wheels if they may need it. I don't have access to a DD, so I can't say if any effect is to be made, but values are so small, I imagine that it may make no difference.

 

In regards to the centering and force scales .
i think it’s probably best to just keep as original in file instead of my variant above.

It reduces inconsistencies between vehicles. It helps keep the weight in the wheel and centre tighter.
while the trail scale force scale adds some details it also opens up the centre a bit an some cars like the 99 stock car that have loose centre become a bit erratic and a mix of too light but with strong peaks in bumps stuff.

I’m finding lfb in the range of 60 is good and the higher i go over say 30 the more dampened the knocking from jolts feel and centre feels more balanced.

 

@Peter Stefani perfect, I agree and confirm ... I tested this reduced value 0.002 and 0.0002 and also on the karts (which are very sensitive) I felt (as you say) that there was a slight improvement change in the narrow center in the top dead zero position of the wheel ... but the feeling does not change and the logic of Karsten's basic settings at all and everything works as the original would like ... I also agree that the value is so small that no low-end wheel and/or med-end wheel may have a TR / TF lower than this ... is activated only the mechanism!!!

 

@Karsten Hvidberg ... many months ago, there was talk of Tighten and you wondered if there could be a value that was unique for all steering wheels ... obviously if you want to cover the real dead zones this value does not exist or rather it is different from steering wheel flying (therefore impossible) and then using a single random value there would be flying with the dead zone recovered and others not because the zone itself is much higher.

But here comes my "stupid" idea: if the purpose was, not to cover the dead zone, but to activate only the tighten mechanism, regardless of whether it is audible or not, to have a different behavior on the center (or rather on the position of force equal to 0), then I thought that it would be enough to enter the lowest value among all the belt or gear steering wheels on the market (and perhaps also DD as they are also equipped with a micro dead zone even if hooked directly to the shaft) ... and since the output is this:

(output (if output (if (+ tighten_range tighten_falloff) (tighten output tighten_range tighten_falloff) output) output))

practically all the forces (whatever they are and if they are lower than the sensitive minimum of the steering wheel) will be increased by the tighten range and at the speed of the falloff ... but it would not change anything in the file, it would be inserted but not felt (it increases the forces only in a decimal way) , but it would activate a different attitude on the zero force position with an imperceptible stabilizing effect that can't bother anyone ... and this applies to all steering wheels ... to be ascertained on the DD but I think that such a low value is inaudible there too !!!

With these premises on tighten, I have practically finished my tuning on the 3.7 file which is a collection of the values present in your original 3.7, but also some values relating to 3.6 and especially to 3.5 very pleasant to many ... it is a very ordinary tuning and not overwhelming but it will be suitable not only for Logitech but possibly it will be usable by almost everyone ... to see the DDs!!!

Thanks to @Peter Stefani for the support in chat ...