Automobilista 2 Custom Force Feedback - Overview & Recommendations

Then you have a dd rixx no belt base. Then you can also try my latest version 35, with the values of "neurotic" it also has a csl dd. The values are in the file. And by the way, there is no such thing as a bad file, you can try and test everything. You just take what you like best

 

I know it's a DD ^^ I was just assuming most files are for DD, so it'd be easier to just name those for non DD. I thought the files are specifically tuned/developed for DD or belt driven, that's why I'm asking. Thanks anyway!

Indeed I need to try some things on the wheel base itself. I'm using Fanatec's recommended settings but they must be for default and default+. I think I will have to reduce FEI as Fanatec recommends 100%, it's probably a bit too much.

They also recommend NFR and NIN at 0, which I'll increase and see how it goes.

 

Indeed I need to try some things on the wheel base itself. I'm using Fanatec's recommended settings but they must be for default and default+. I think I will have to add

The values for the csl dd are in my custom file from a member (neurotic) in this forum. I'm sure he can give you more tips. Regards

 

Just sharing my settings, because I am always happy to find a recommendation specific to my wheel base to have a start. I don't have the time to dive into the custom files myself, so I am happy and thankful for everyones contribution. If there is only one person who finds my settings helpful, well here is my little contribution.

In the end FFB is a very subjective topic and my settings might not be for everyones taste. I am also still testing to find my own favourite settings. I will keep posting settings that worked for me when I give feedback to a custom.

With my settings I am trying to find good balance between dampening and FFB details. For Fanatec wheels you can use NDP and NFR for additional dampening that helps reduce/eliminate the oscillations. I tend to always have some on the wheel (NDP: 15 - 30, NFR: 0-10) but too much makes the wheel too heavy, so I try to keep them low. I rarely use NIN to add weight to the wheel, but sometimes a car with a very light wheel just needs a little bit of it and it feels better. Here in AMS2 also FEI has a big impact on how the customs feel. I ususally stick to the recommendation from the author and try to adapt to the CSL DD 8nm wheel base, but I also want to test higher FEI with some INT to counter the additional noise in the wheel. That's what I am going to test tonight.

 

I can agree with you about the fanatec settings. That's why I have 2 variants for the dd1/dd2. There are people who like to have more damping and lfb, which works very well in Ams2. Personally, I prefer to have almost no damping or no damping at all (for testing) only then do I have this feeling for the ffb and I can feel the finest differences

 

I get the shakes on the wheel only off the line or from a standstill if I spin. If I give it more throttle it just gets worse. I will try your earlier, non Atmos, files again, to try and pinpoint when it first started happening and let you know. Thanks for all your efforts.

 

My wheel started doing that at a stand still even with the Default or default + after the new .1 update to the game. I am about to test the Atmos, but I was still using the original rfuktor

 

so there are 3 variants in this thread?
DanielKart, mdecker79, both based off the latest file in Stakanov signature. Is that correct?
just trying to get the correct ones.

 

@Stakanov I tried the 4W Grip Mod file and there was very little steering wheel shake. Then I tried the 4W Grip And Stretch file and that one was much worse, but not quite as bad as the Atmos file. So it seems the problem has got progressively worse each change. For me anyway. I have used basically the same settings in each file. Hope this helps.

 

I have been testing the @Stakanov file tonight with various settings in the wheel. I find the feed back is good for the most part, and helps my lap times. The test for me of max feed back effort by the wheel is Turn 1 at Brands hatch. In EVERY other sim and both the default settings in AMS2 when you hit the bottom outside of the turn and nearly bottom out the car, you feel a strong outside force push the wheel out to the left.
With this file, no matter the settings, the only thing I feel is the curb. there seems to be little outside loads in the turns, even with the FFB in the game turned up to 90.
I am not an expert, but not sure where the outside load has gone?

 

Thanks to everyone for the constructive help, I have two days before resigning, I have memorized your information and I know that the things to improve are the mobility of the load to the outside and if we do not have to solve it (with this next update) the steering shake.
They will be my points to face but first know this .....

I am testing the ATMOS V.1.3 file for posting on the forum, it is dedicated to skaters, gliders, haters of both the perennial Grip and the Tear, and has these characteristics:
- Further corrections of some lines of code especially in the bumps but also in the part of the code dedicated to the preparation of the road (now we should have finished correcting and have a stable code base to develop in detail);
- New variable and new code dedicated to deactivating the sense of grip when the sliding of the wheels is maximum or better than 85% and up to 100% ... so the Grip is not present in this interval ... but on the contrary when slide is from 0 to 85, the grip is present from 100% to 15% (you can change the deactivation level of the Grip from 15, down to 10 or up to 20 as you wish .... or restore it as before by setting 0.0 or 100% Grip if you like !!);
- New variable and new code for the deactivation of the tear, useful to lessen the sensations on the wheels when the tear effect has not yet arrived (for now the deactivation will be present only for values of tear lower than 0.01, therefore at values lower than the tear will not be felt while at higher values it will (you can increase or decrease the amount of deactivation or remove it completely, returning it to the original effect by setting 0.00);
- With the mobility changes indicated above (of the Grip and Tear) the sliding is no longer flat and characterized by the ON-OFF or 0-1 or grip-release effect of the Grip, but it is very modular, therefore, it was possible to bring the coefficient back of sliding to the original levels of the Karsten file 0.85-0.90 and it is not necessary to reshape the curves as before (you are always free to do so);
- Dynamic adjustment to accel feel based on car weight (thanks to Daniel);
- Only one settings for both DD and for belts or gears, few are the different values proposed and these have become only as ALTERNATIVES no longer dedicated to DD but to anyone who likes them ... so all the voices or item are without alternatives value (only tyres resistence and Os Ds have), if you don't like them, will be from set as you desired;
- Tuning updated to new changes.

I hope I am not the only one to notice these improvements and if confirmed by next weekend we can devote ourselves to the details!!! See you tonight for the file!!!

 

As usual, two Atmos 1.3 files (with and without YAW) ... and as usual I attach the last original Karsten's file ....

Name: MOD ATMOS 4.1 (4xFX.1WHEEL) V.1.3 (with & without "Yaw") based on 5.0.1.4
(GRIP&SLIP REVISITED, TEAR, CORRECTIONS TO CODE AND RETUNED)

Name: rFuctor 5.0.1.4 Original Karsten Files

Use whatever you want, the last habit of the user wants that the file without yaw is for DD, while the other with the Yaw is for the belts and for the gears, but this obligation is not true, for DD it is only harder to balance a file with Yaw (maybe) and Karsten's original to date has it.
Start with the tests by finding the right sliders in the game (this is very important), initially use the original values of the files and try more cars and more tracks (it often happens to me too to change the values because you don't like a car but then the modification is not good for another one, if I change more values I don't understand what to correct anymore, so test well I recommend) thus applying only if necessary your taste by raising or lowering the values.

New item....
(tear_deadzone 0.01)
(grip_deadzone 0.15)

Item moved for yours comodity .....
(slide_factor 0.85)
(scrub_deadzone 0.00) -----> Not used

Spoiler: Grip&Slip Deadzone

The Grip is the opposite of the Slip....... then when Grip is 100%, Slip is 0% and the contrary...
(grip_fl (- 1 slip_fl2))
(grip_fr (- 1 slip_fr2))
(grip_rl (- 1 slip_rl2))
(grip_rr (- 1 slip_rr2))

but is rare that the Slip is 1 or 100% (maybe only when fly) then the Grip is always here in the tyres ... then if we set a deadzone of 0.15 on the grip this is present only after this value and when the Slip are from 0.0-0.85 (even if the scale are 0-1) ... then

(grip_deadzone 0.15)
(grip_fl (max 0 (- grip_fl grip_deadzone)))
(grip_fr (max 0 (- grip_fr grip_deadzone)))
(grip_rl (max 0 (- grip_rl grip_deadzone)))
(grip_rr (max 0 (- grip_rr grip_deadzone)))
(grip_fl (/ grip_fl (- 1 grip_deadzone)))
(grip_fr (/ grip_fr (- 1 grip_deadzone)))
(grip_rl (/ grip_rl (- 1 grip_deadzone)))
(grip_rr (/ grip_rr (- 1 grip_deadzone)))

Then is not always present here in the stretch ....
(tyre_fl_stretch (* front_tyre_stretch_feel grip_fl FL_stretch angle_d_l wheel_diam_FL 0.5))
(tyre_fr_stretch (* front_tyre_stretch_feel grip_fr FR_stretch angle_d_r wheel_diam_FR 0.5))
(tyre_rl_stretch (* rear_tyre_stretch_feel grip_rl RL_stretch angle_d_l wheel_diam_RL 0.5))
(tyre_rr_stretch (* rear_tyre_stretch_feel grip_rr RR_stretch angle_d_r wheel_diam_RR 0.5))

and then is not always present here in the tyres ....
(tyre_l (+ tyre_l_init tyre_l_res1 tyre_l_res2 tyre_l_res3 tyre_fl_stretch tyre_rl_stretch))
(tyre_r (+ tyre_r_init tyre_r_res1 tyre_r_res2 tyre_r_res3 tyre_fr_stretch tyre_rr_stretch))

If you want the Grip at 100 you can set this to 0 and all is as it was before....
(grip_deadzone 0.0)

Spoiler: Tear Deadzone

The same for Grip are applied for the Tear, but the deadzone are activated only for reduced or stop the effect on the tyres when the Tear are very little or not present ... the value 0.01 is the limit for not consider the Tear in the tyres .... then

(tear_deadzone 0.01)
(FL_tear (max 0 (- FL_tear tear_deadzone)))
(FR_tear (max 0 (- FR_tear tear_deadzone)))
(RL_tear (max 0 (- RL_tear tear_deadzone)))
(RR_tear (max 0 (- RR_tear tear_deadzone)))
(FL_tear (/ FL_tear (- 1 tear_deadzone)))
(FR_tear (/ FR_tear (- 1 tear_deadzone)))
(RL_tear (/ RL_tear (- 1 tear_deadzone)))
(RR_tear (/ RR_tear (- 1 tear_deadzone)))

If you want the Tear without this you can set this to 0 and all is original
(tear_deadzone 0.0)

Everything should be smoother, less pumped up, less impetuous and very informative ... personally I live very well the conditions of the road, the information coming from the wheels and the general games of the forces ... I hope you feel it too, but for sure, this is a finally stabilized version ... from here on I will only follow your advice to develop it slowly and without haste.
-------––--------------------
EDIT: for now, in this 1.3, i have low the lateral load for stabilized the rest of file ... if you want restore the lateral Load sensation of 1.2 you try to modify this line which is at the bottom of the file .....

(as (power as 0.6)) ---> set to 0.65 (the same of 1.2) or, if you want more, set to 0.7 that is the original value for lateral load of suspension

For all, I communicate that it is possible to restore the 1.3 file to the state of the previous 1.2 just set the following new items as indicated below:

(steer_spring_strength_feel 0.50) ---> set to 0.65 (the same of 1.2)
(grip_deadzone 0.15) set to 0.00 to restore the different grip sensation
(slide_factor 0.85) set to 0.6 to reduce the slide
(tear_deadzone 0.01) set to 0.00 to remove the deadzone

with these changes the file 1.2 and 1.3 is practically identical ... only the correction code of bump and of the road are the difference ... remember this if you take the tests.
-----------------------------------------------------
EDIT 2: The file was update to version 1.3.1 ... in this post the download ....
Automobilista 2 Custom Force Feedback - Overview & Recommendations

 

Awesome... Will try it tonight. Question: in an odd case I have a DD pro (CSL DD) base and AMS2 wont recognize it as anything but a CSL elite in the profile name.
So do you suggest the with or without Yaw version of the file?

 

You can start with what you want, let's say with the No Yaw one, first find the right sliders and modify the values that you think are most suitable for you ... try more machines ... once you have found your balance you can replace the values on the other files with Yaw, they are interchangeable, and so you will have the exact comparison between Yaw and without Yaw ....
Regarding the lateral forces, the setting is a bit wanted down to be able to easily use all the cars, even the most swaying ones with the same file ... the difference between a Formula and a V8, or a Chateran with a GT3 is huge in terms of strength .... then if you want some advice I'm here for you as soon as possible ...

 

thanks. I tried the with Yaw yesterday, so I will try that again for a fair comparison.

 

Same here Yaw version

 

File the 1.3 very different from the previous 1.0-1.1-1.2 and despite being the same base has a final effect that is completely different, we note the stability and coherence (even the previous ones were not terrible despite the errors), but above all the wheels are different,
You feel the grip all the same, but now also the lightening, the slide and the final tear and scrub effects, which I see again reversed in order, which you haven't done for a long time, a symptom that perhaps the tear does not give more bother ??
The effects on tyres and the "road" are awesome and I think it just needs to balance the load only on very few cars.... for the rest I am amazed ... thanks as always @Stakanov

 

(Son)
It is amazing how a few lines of modified code, two more correct lines and a perfected tuning can upset the result.
Here on T300 everything seems alive and less impetuous and it looks like another file than before!!!

I will use it a few days before highlighting what I don't like, if I find it ... congratulations!!!!
I like it very much !!!

@Stakanov I'm starting to study the code of the FFB file a bit, how did you train on it? I would like to deepen ...

 

thanks for the effort and sharing. Will hopefully test it this weekend.

 

This time the changes modify the instructions on the tires and the Grip component was and is very intrusive, inserting it or not inserting it put many people on not marrying or marrying the variable, or they loved it or hated it, so the difference is like this noticeable (positive or negative), then the bumps have been corrected and some values revised ... the summation is easily noticeable compared to before, I hope positive, we'll see.

Regarding the code, I did everything by myself, there is no specific language, indeed you write the code following certain rules, only you need to be aware of the product to be achieved and therefore know how to instruct, aggregate, shape and calculate what you want. it serves to reach the final goal (in the case of the FFB to self-learn mathematics, geometry, physics and dynamics). Even today I don't know how to do it from start to finish, let's say that I only took the red pill that allows me to see the "Matrix" code that I didn't see before, now I can read it, copy it and shape it respecting its rules ... but absolute compiler is abundantly light years away Karsten, who is not only a good compiler but certainly a good connoisseur of mathematics, geometry and physics.

To help you I can tell you the "General form of code", which is the only thing you need to understand it.....

#(A B ....)
If A is known function name, then that function is done, with B ... as the arguments.
If an argument is itself in (), then that is resolved first.
If A is not a function name, it is taken as a variable name, and is assigned B, and ... is ignored.

So for example this....
#(output (relative output 1.0 0.5 2.0))
becomes in effect something like in C ish form:
#output = relative (output, 1.0, 0.5, 2.0);

or

#(Function (+ A B))
becomes
# Function = A + B

Then we have "variables", "argments" and "function" ... native or generated/calculated ... practically a castle is built up to the final output.