Automobilista 2 Custom Force Feedback - Overview & Recommendations

Yes, correct.
If you want to adjust them, do so in the graph first, so you see it. Then take those values and insert in the file.
It's the first 4 lines of values in the desmos graph in the left side & same order. Just desmos does not allow full names(or maybe I just didn't look up how to do that), so it's 1 letter names instead.

 

(relax_os_line2_start_y (if isKart 0.0024 relax_os_line2_start_y))
This should represent when the curve of the relax function begins with respect to the starting point ie at zero forces (zero over_steer)
Start

(relax_os_pos (if isKart 0.00144 relax_os_pos))
This should represent where the relax is found along the curve, the higher it is the more relax occurs away from the zero force
When relax over_steer

(relax_os_width (if isKart 0.0105 relax_os_width))
This should represent how large is the relax is in the above position, the higher it is the longer the relax itself
How much width relax over_steer

(relax_os_coef (if isKart -0.015 relax_os_coef))
This should be the coefficient that is the amount of relax
Quantity of relax over_steer

Depending on the value if you exaggerate it can amplify the relax itself and / or turn it upside down in amplification depending on the combinations you have to be careful about the quantity of positive and / or negative values because you could overturn what you are looking for and amplify rather than relax ...

I hope I didn't say stupid ... only @Karsten Hvidberg can confirm we wait

 

I can add a little explanation:

There are 2 lines:
1) The relax line, that's based on x^(x^relax_coef)
2) relax_os_line2_start_y is the starting point on y of line 2 going from that point (0,starty) to (1,1).

Then there are 2 other values: relax_pos & relax_width
They control the position and width of an interpolation function that creates a smooth transition over the given width.

The final line is using that interpolation function to combine line 1 & line 2.

I hope this helps a little.

 

Thank you very much to Karsten and Tumi for your feedback.
i have downloaded and installed the 3.5 experimental version i found a few posts back. hopefully this gives me the feedback im looking for. (I will test tonight)
i believe i was using an older file from a few months back but admittedly i struggled with my TD settings and in game settings.
this is the 1st time i get a sense of direction on where i should at least start my baseline and adjust from there.

 

If it helps, here's my True Drive settings (got them directly from Karsten himself over a month ago):

http://www.dropbox.com/s/ofcaukn0fo..._09-04-21_ffb_settings_true_drive_1.png?raw=1

http://www.dropbox.com/s/brslgsmene..._09-04-21_ffb_settings_true_drive_2.png?raw=1

http://www.dropbox.com/s/p5xfgfr41h..._09-04-21_ffb_settings_true_drive_3.png?raw=1

 

Ok so I tried tumi's settings I copied everything except the reconstruction filter which I left at 1 as per karstens post from earlier and I can say it's a world of a difference I am definitely getting more response and control from the car.... despite the fact that I need like 5 years worth of practice lol I'm able to push the car and control it much better....you guys are amazing...now I just need to find the same equally effective settings for my other titles

 

Yes this is the problem, nothing else feels good anymore, I’ve even tried blending recent code into older Pc2 silver files , but it doesn’t seem compatible.
Not sure @Karsten Hvidberg has time to update that side these days and it’s fast becoming obsolete as a title.

 

Shame but understandable ams2 is a newer tile with more promising updates likely on the horizon... That being said I like pc2 and just like having a wide variety of sim games to use with my rig... The more options the better but that's just my opinion... For right now I can say I'm loving and 2 so the more refined it gets the better

 

Hi @Karsten Hvidberg et al

Really loving Rfuktor3.4, but I do have to chime in with a bug report which I've seen reported a few pages ago: when the weather transitions from dry to a fully wet track, the FFB gets increasingly violent. It starts as though the road noise and bumps are massively amplified, and ends up with the wheel wrenching violently in the direction I turn it. Switching to default stops it, switching back to Rfuktor and it's violent again.

This has happened with several cars on various tracks, on test day and quick races, but it only happens when a session starts dry, and transitions to fully wet. Races that start and stay wet or start wet and transition to dry do not exhibit the behaviour.

Hope this helps, please let me know if I can provide any other info

 

Yes, the last experimental 3.5 posted here fixes it to a large degree.
The problem is that some of the madness tyre variables start acting weird for some reason.
I will have to add a custom calcvulation for tyre deformation to fix it to a level I would be hapopy about for the final 3.5 file.

 

Hey @Karsten Hvidberg,
I have tuned your 3.5 EXP file for use with Logitech bases (according to my personal tastes) ... I preferred to use the trail, I have activated the tightening in the general part and the values that manage the center of the wheel leaving the curves relaxation like you did, and modify only the curve of slip_crossover ... in the specific karts part, on the other hand, I disabled the tightening, used high values of center and centering and modified the relaxation of os and ds curves ... if it can be useful before the final 3.5 ( even just out of curiosity) do not hesitate to ask ... I can publish it here or even in pvt ... otherwise I will wait for your final file for which I am very curious.

Curious or stupid question, if I open the links (in the FFB file) of the functions and curves (relx, yaw, etc.) the starting values are a little different from those of the file ... they are different because they are the values you are using now in the Final 3.5???

Have a good day

 

Does the wet weather bug only affect the latest files or is it all of them?

 

I think the later rFuktor files, but no other files.

 

Yes, def. post your file as it will make it simpler to see if there's some preference that makes sense for the high end file as well.

The links in the file have values from rFuktor 3.4 or so, I did not make any change since the 3.5 exp. file. Also the links are static, so the values won't change even if I change something & you can adjust them exactly as you like & not worry about overwriting anything.

 

@Karsten Hvidberg
This is the tuned file for Logitech ... I find it very coherent for karts (although many things need to be fixed by Reiza) ... as you can see, leaving aside the usual quite personal values, I reduced a little the slip_crossover in the general part ... while in karts I disabled the tightening, I modified the curves of relx "os" and "ds", starting the relaxation as soon as possible, widening the relaxation as much as possible and setting a quantity consistent with the shape of the curve so as not to create jolts ... without tightening to make the center full I had to raise the centering and relax_center a lot

(over_steer_scale (if isKart 0.85 over_steer_scale))
(relax_os_line2_start_y (if isKart 0.0001 relax_os_line2_start_y))
(relax_os_pos (if isKart 0.00400 relax_os_pos))
(relax_os_width (if isKart 0.0400 relax_os_width))
(relax_os_coef (if isKart -0.030 relax_os_coef))

(under_steer_scale (if isKart 0.035 under_steer_scale))
(relax_us_line2_start_y (if isKart 0.0001 relax_us_line2_start_y))
(relax_us_pos (if isKart 0.00100 relax_us_pos))
(relax_us_width (if isKart 0.0100 relax_us_width))
(relax_us_coef (if isKart -0.010 relax_us_coef))

I tested everything with my usual in-game FFB values of 65-50-55-55, but I noticed that I could still raise something, I will set it precisely in your next final file 3.5

 

nice work @Stakanov , these curve values help a lot with the karts.

 

Thanks @Kuku Maddog, the idea came to me when I understood how it worked after Karsten's explanations ... the problem for me was to find a curve that relaxed as much as possible the immediate tendency of the Karts to spin when cornering (both in entrance, in distance and in exit) or for example when the kart turned over passing over the highest curbs, ... I found the solution by inverting the way to tune the file, i.e. first I create a relaxation curve that makes me drivable the car and then I tune all the remaining values to make all the other aspects coherent ...

The same thing I want to do in Karsten's final 3.5 also in the general part, first I will create a relaxation curve to make the Caterans and the other more swaying cars (Vintage, OldStock, SuperV8, etc.) drivable, then I will set the remaining parameters to make the 'FFB more informative about the same cars, ... finally, only later, I will make the remaining adjustments (in order) to make driving even more rigid cars like the Formula Reiza, Ultimate, Gen 2, the other Formula, the GT3-2-1, the P1-2-3-4, etc. .... let's see what comes out!!!

 

when I look at the graph, I can see what it does but have no reference to what is ideal , so I don’t really get it, then again I failed math

 

If this can help you, and if I interpret it correctly, do not look at it with a mathematical cut, but look at it in a graphic way ... you have two functions (the first OS_Scale and the second OS_Relax that relaxes), speaking of the second you have to mentally start from a borderline case that is this:



- Assume that the graph of the OS_Relax function is perfectly linear and inclined by 45 degrees (1: 1) as in the photo above ... in this condition (given the first starting OS_Scale function whatever its shape, linear or non-linear) the relaxation implemented by the second OS_Relax is applied in a uniform and linear way, and therefore by both increasing and decreasing the relaxation values in a fixed manner. Practically the OS_Relax function would be unnoticeable on the steering wheel when the force changes, because the final result will always be "OS_Scale - Relax_OS (linear relaxation)" ... you would get the same result by directly lowering the value of the OS_Scale function without relaxation.

If instead OS_Relax is not linear (therefore different from 1: 1), then the relaxation is applied in a variable way and in an exponential + or - way, depending on whether OS_Relax is concave or convex, and whether above or below line 1: 1 and this depends on the two functions or lines contained in OS_Relax (and joined by interpolation) and on the 4 variables we discussed earlier.

 

Ragazzi Buongiorno
Scusate ma sono poco pratico.
Il file aggiornato per simcube 2 è quello nel primo post?
Ho letto di vari aggiornamenti fino al 3.5 ma non riesco a capire se è sempre quello nel primo post aggiornato mano mano.
Grazie mille