Automobilista 2 Custom Force Feedback - Overview & Recommendations

@Rob Davies @Karsten Hvidberg and to All for testing... suspension topic ... I don't know if this feelings are just mine!!

Karsten this is the your last code for suspension in the file 5.0.1.4....
(load_f (+ FL_load FR_load))
(load_f (* 650 (/ (- load_f (smooth load_f 0.2)) (+ FL_start_load FR_start_load))))
(load_r (+ RL_load RR_load))
(load_r (* 650 (/ (- load_r (smooth load_r 0.2)) (+ RL_start_load RR_start_load))))
(load (- load_f load_r))
(suspension (- load (smooth load 0.03)))
(suspension (smooth suspension 0.2))
(suspension (* 2 suspension 0.002 suspension_scale))
(as (blend (min 1 (* (abs steering_angle_rack) 10)) 1 0))
(as (power as 0.7))
(as (smooth as 0.1))
(suspension (* suspension as))

take a very informative car in the suspension, for example the V10 Gen2 (for Formula) or a GT3 of your choice, try some tracks like Kansai (right-left and kerbs), Spa (long curves with kerbs) and Nurburing (many bump and kerb) and memorize the sensations of the suspension

Now try this
(load_fl (- FL_load (smooth FL_load 0.1)))
(load_fr (- FR_load (smooth FR_load 0.1)))
(load_f (+ load_fl load_fr))
(load_f (* 650 (/ (- load_f (smooth load_f 0.1)) (+ FL_start_load FR_start_load))))
(load_rl (- RL_load (smooth RL_load 0.1)))
(load_rr (- RR_load (smooth RR_load 0.1)))
(load_r (+ load_rl load_rr))
(load_r (* 650 (/ (- load_r (smooth load_r 0.1)) (+ RL_start_load RR_start_load))))
(load (- load_f load_r))
(suspension (- load (smooth load 0.03)))
(suspension (smooth suspension 0.2))
(suspension (* 2 suspension 0.002 suspension_scale))
(as (blend (min 1 (* (abs steering_angle_rack) 10)) 1 0))
(as (power as 0.7))
(as (smooth as 0.1))
(suspension (* suspension as))

leaving aside the necessary adjustments on the intensity of the effect still to be done at a later time, I feel something very consistent and improvement not only in the movement of the suspensions but also on the bumps, on the curbs and in the sensation of weight that the used car represents!!! ... it is possible or is it just me on the low end and on you it feels differently???

 

1st point tested.... Smoothness = 0 did the trick

(kerbs_surface_smooth 0.00)

dt1 300 or 333 didn't matter.

 

@Rob Davies and now use the alternative code above for suspension ... now the kerb is here even with smooth enable but little (0.004-0.003 or 0.002 for me the best) ... and the suspension is ok very very informative on kerbs and on bumps and on street

For the dt is ok even here then i don't touch for the moment .... i think the noise that @Michael Enright feel are the texture_vibes ... try disable but if you don't disturb is ok

 

Evening @Stakanov et. al

So... I've had another hour playing.... a few notes below and my working file attached (NOTE: NOT READY FOR PRODUCTION - Use at your peril!)

- I like what you're trying to do with a base file that spans all bases with Belt/DD paramters - great goal and I'm supportive - simpler and easier for everyone!
- As such, I'm trying to map my existing favourite file (based on @Kuku Maddog 's DD tune - also attached) into it.... but I need to understand more about how the file works to do that.... master scales vs. the effects values etc. - so that I can retain the same master scales that you have in your 'master' and adapt the effect scales accordingly (that makes sense? right approach?) - I've attached it to this post - suffix "The Best.txt"
- The WIP file attached suffix "(RD DD Edit - WIP) is based on your No-Yaw file with the following changes:
-- No Relax. I notice this works well for oversteer, but you seem to lose the wheel going light under braking/lock-up which would be great to keep - the way the code works?
- I've added #RD to the lines I've changed from the recommended DD Values (more to do here, hence WIP)
-- I prefer more suspension feel than your DD tune value (the rest of supsension/track values need to be played with)

Code:

(suspension_scale             0.60)      0.35    #How much the suspension is felt. #RD

- The kerb_smoothness 0 fixes the lack of kerbs.
- Your other suspension code above (also pasted in and commented out original) - to me (after 15 mins) - I can't say it's better or worse.... sorry! - Perhaps more time or other opinins on this one... or once the effect intensities are better balanced.
- *** HOWEVER *** .... one thing I am settling on is that the slip_crossover values are FUNDAMENTAL to how the car feels/drives and how much you can read/control it.... I've noticed from the 4.6 files to the 5.0.1 files... to the latest... there's lots of changes in these values which result in the 'driveability' of the cars being completely different. My limited read of these values and the file seems to suggest that the rest of 'what, how and when' you feel the effects hangs off these values so they transform these custom files. When I put the 4.6 values back into @Danielkart 's file it felt so much better. A few people I race with when going back from 5.0.1.3 to my 5.0.1 tune have said it solves their 'snap oversteer' problem, which they descibe exactly like I described the 'lag/latency' in what/when you feel the car start to go... which makes me think...
-- What are the right values? How do we test that the slip is felt when the tyres are actually slipping rather than when the alogrithm suggests it? Telemetry? How can we be sure. It feels pivotal to get these right.... sorry - my knowledge here is so limited I'll bow to the experts!
-- I've added all 3 sets of crossover values (4.6, my preferred 5.0.1 tune & yours) into the working file so you can comment each out (there's 5 lines in each) and see what I mean / what you think. Again, I think a good one for others to test/comment on as the difference is night and day.

Code:

# --- Original Crossover ---
#(slip_crossover_pos           0.05)            #Slip effect graph: https://www.desmos.com/calculator/iskij6snwe
#(slip_crossover_width         0.60)
#(slip_crossover_line1_slope   0.30)
#(slip_crossover_line2_start_y 0.18)            #Increase this value to increase or prolong the grip feel of the tire on the edge before starting to slide/slip.
#(slip_crossover_line2_end_y   0.95)
# --- Crossover based on 5.0.1 KukuMaddog --- #
#(slip_crossover_pos           0.125)#RD    #slip effect graph: https://www.desmos.com/calculator/iskij6snwe
#(slip_crossover_width         0.54) #RD
#(slip_crossover_line1_slope   0.2)  #RD
#(slip_crossover_line2_start_y 0.1)  #RD
#(slip_crossover_line2_end_y   0.95) #RD
# --- Crossover based on 4.6 DX2 ---
(slip_crossover_pos           0.245)    #slip effect graph: https://www.desmos.com/calculator/uf9w9q170h
(slip_crossover_width         0.50)
(slip_crossover_line1_slope   0.2)
(slip_crossover_line2_start_y 0.15)
(slip_crossover_line2_end_y   1.15)

- I've also attached @Danielkart 's file I've made a few changes too, which again I find great... best tyre feeling - if we can get this into your file somehow I think we're onto a winner

I can see how you can lose hours on this!

Hope this is again helpful.... will continue to test (between racing of course!)

Keen to hear thoughts and views on the crossover values though....

Cheers
Rob

 

Wow @Rob Davies it is right what you are wondering about the slip_crossover differences and the sliding sensation you seek depends on these 5 settings, however these settings do not work in isolation they depend on how they are drowned in the file and their success depends on how all the other parts of the code relate to each other and with the slip ... the fact that you really like the sliding of the kuku 4.6 file is a certainty, but this pleasure depends above all on how everything else is composed ...
Now the files have a very different code from 4.6 and duplicating this for example on a file 5.0.1.4 mine or Daniel or Mix is very difficult, moreover, even the heart of the file that is the Momentum Rack and all others parts are different from one version to another, but nothing is impossible.
You could act in two ways, or starting from the last 5.0.1.4 by inserting the slip values of the 4.6 of kuku and then, on each part of the file, check step by step if the old code works better than the new one but you will understand that the logics are different and it is very difficult indeed impossible for the slip of 4.6 to adapt to that of 5.0.1.4 .... or reverse the process, starting from your favorite 4.6 of kuku tuned as you like best (your the best to understand), in this way above indicate.

Initially leave the heart of the file 4.6 unchanged (i.e. the Momentum Rack of 4.6 and the initial part of the code called Helper vars & prep), try to change it starting from the periphery of the file, i.e. from the "effects", one by one replace it and check if like the old one more, 5.0.1 or 5.0.1.4 and keeping the version you like best (it doesn't matter if the oldest or the newest, if mine, kuku or Daniel's) ... once the effects are over, move on to do the same operations on the code relating to the "road" and leave or replace the parts according to your tastes ... after all this step you would already have your 4.6 with the same sliding feel but with all the effects and road updated to your best perspective with only the old-fashioned heart (Momentum Rack and Helper vars & prep of 4.6).
ALL THIS MUST BE DONE KEEPING THE SAME SLIDING FEELING THAT ONLY YOU ON YOUR WHEEL AND YOUR BRAIN CAN VERIFY

Once this is done you could move on to the rest (The center of file, the hurt - Mom.Rack+Helper) with the same principle, and I would like to help you a lot, even if necessary long time, but how can I understand your tastes, your emotion, your sensation and your brain ... and how and what to replace, step by step according to all this??? Any changes made here by me could cause you to lose the feel and sliding sensation you aim so much for.

So far I have practically done this, and I started from other previous files, to pass also from 4.6 (I recognize the beauty of the original 4.6 file or that of kuku or others), but while doing an incremental work that tries not to betray these preferences, these tastes and those feelings trying to just take steps forward, as you can see, now the file has become something else than before and whatever improvement you bring in my head there is always someone who has feelings or a different feeling ... in this situation I have long ago decided to adapt as much as possible to the latest Karsten file, otherwise sooner or later the project will become something other than Karsten's and the two realities will no longer be able to compare ... you are in this situation and you want shorten distances but i think is very difficult!!!

But I am here to help you as best I can and if you want to start this replacement I am available to help you or advise you step by step, but you will have to be the one to do the tests ... I could only duplicate them on my low end to give you a personal advice.
...................................................................
If, on the other hand, I have misunderstood your request and you would just like to approach the feeling of slip of the 4.6 in the RD file then the thing is simpler and if you want I can explain to you what I understood about the 5 items and how to move

 

Wow, thank you for your long and detailed reply - much appreciated.

I understood from your explanation that everything is connected and that it’s not as simple as replacing the slip over values in isolation.

The thought of systematically trying bits to rebuild the file blows my mind - I fear I’ve no real idea of how it all works and comes together to produce the magic it does… so it’s be a long journey!

however, I am keen to help the community get a good feel as I think it transforms the enjoyment of the game when done right.

there’s so many questions I have to try and understand all the values in the file - but don’t want to waste your time going through it if it’s more productive for me to try and explain what I’m feeling and test changes. Im just about getting the hang of changing effect scales, but suspect I’d have to try and dissect the code to work out how it all comes together (if I’m even capable of that!)

Let me think about how best I can help. It’s why I like Kuku’s tunes as he gets a feel that’s 95% right for my tastes… so perhaps I sit tight and see if I can improve from there.

but yeah, that crossover momentum thing, makes a big diff.
Thanks again. Will sleep on it.

 

I’ve posted in feature request about the ability to reload custom FFB file from within game here - The Feature Request Thread...

perhaps a few likes or agrees from us all might help get it on the roadmap

 

Hey @Rob Davies yes replicating the whole file and slip 4.6 is difficult but if you want to try to recreate a similar feeling then it's different .... i think and I hope that this can help you.... imagine GRIP go to SLIP and return to GRIP

slip_crossover_pos how soon starts the whole process , low value anticipate the process, high value retard the process (when is the process)

slip_crossover_width how large is the procese or the effect, high values widen the effect, low values tighten it. (how large is the process)

slip_crossover_line1_slope how distant is the slip from grip, low for distance and high for near, then how mutch is distant (how slope )

slip_crossover_line2_start_y ------> Increase this value to increase or prolong the grip feel of the tire on the edge before starting to slide/slip. Low value decrease or reduced the time of grip on the edge. (when lose grip on the edge and slip)

slip_crossover_line2_end_y -----> The same of above but when finish the slip and return on the edge of grip. Low for anticipate the grip and High for prolong the slip (when lose slip and grip on the edge again)

In red are quantitative variables (how mutch), the other are timing (when)

-------------------------------------
EDIT: I changed in the post above the suspension code ... if you try again maybe you have a surprise ... i dont now if you like

 

Hi @Stakanov I tried with the Vibes lines taken out, but unfortunately for me it is still the same. It is just like continual Road Features. Strange that I am the only one. I will stick with your 4W Grip Mod file for now. Thanks for the suggestion though.

 

Strange but the solution must be under the eyes, the basis is the same and by exclusion comparing the files you should find what generates it.
How much of FEI do you use 100% ... if I'm not mistaken Daniel was warning of weird effects or noise with values above 30, so try 30 or lower values

BTW the different other voices between the files that could generate more noise or lateral effect are:
- master_road 1.35 ----> before it was 1.25
- road_rolling_slip_feel 0.85 ---> before it was 0.75
- rear_end_feel 0.25 ----> was not present before so try 0.00

As a code, nothing comes to mind after dt and texture_vibe, maybe only the smooth of the scrub is different and disable, try to invert the # symbols to return or to activate the smooth ... the same of the GRIP file!!
(FL_scrub (* (if FL_bump_offroad 0 1) (smooth (abs FL_scrub) 0.01)))
(FR_scrub (* (if FR_bump_offroad 0 1) (smooth (abs FR_scrub) 0.01)))
(RL_scrub (* (if RL_bump_offroad 0 1) (smooth (abs RL_scrub) 0.01)))
(RR_scrub (* (if RR_bump_offroad 0 1) (smooth (abs RR_scrub) 0.01)))
#(FL_scrub (* (if FL_bump_offroad 0 1) (abs FL_scrub)))
#(FR_scrub (* (if FR_bump_offroad 0 1) (abs FR_scrub)))
#(RL_scrub (* (if RL_bump_offroad 0 1) (abs RL_scrub)))
#(RR_scrub (* (if RR_bump_offroad 0 1) (abs RR_scrub)))

EDIT: Or disable with # this scrub effect vibration

#(scrub_effect_vibration (* 1.5 (spring scrub_effect 3000 0.001 1)))
(scrub_effect (spring scrub_effect 75000 1 1))
#(scrub_effect (+ scrub_effect (* 0.2 scrub_effect_vibration)))


Excuse me if I insist I don't want to bother you and it's okay to use the GRIP or others file ... it's just that I'd like to solve everyone's needs ... so I learn more information!!
Good days @Michael Enright

 

Thanks, I'll try a couple of things later. It is strange that none of this is in the Grip only Mod file. It happens in a straight line, so I wouldn't think scrub should effect it, but who knows. I have tried changing those other settings also, but doesn't help. I do run FEI at 60, but no other files are effected in the same way as your Grip & Stretch is. I still think that the std 4WGripMod is good. Gives me all the feel I like and need. On a different note, I don't know if this actually worked, but I added these 6 lines to the Grip Mod file, and it seemed to work, because, with (tear_scale_rear 0.85) I got a lot of wheel movement (Wheelspin) when I accelerate off the line, so I dropped it to 0.75 and was as I would expect, slightly less.

(tear_scale_front 1.00)
(tear_scale_rear 0.85)
(FL_tear (* FL_tear tear_scale_front))
(FR_tear (* FR_tear tear_scale_front))
(RL_tear (* RL_tear tear_scale_rear))
(RR_tear (* RR_tear tear_scale_rear))

 

Very useful thank you... couple of Q's if I may to see if I've understood correctly...

slip_crossover_pos - the grip vs. slip threshold I guess... so ideally the right position would be 'on the edge' or slighly before? (I think this is key in my observations around 'lag' on feeling)

slip_crossover_width - what is meant by 'how large' here... the size of the difference in forces felt (high on grip) vs. low on slip?

slip_crossover_line1_slope - is this kinda like - how wide is the feeling of being on the edge? So like the width of the ramp between grip and slip feeling? (larger ramp, more gradual feed-in of slip effect?) Higher means there's a longer transition from grip<>slip, lower is shorter? (If so it feels tighter here should give a sharper 'on-edge' feel?)

slip_crossover_line2_start_y - so does higher value here keep the forces applied longer as the tyres are still 'gripping' giving me the feel that the tyres are biting back? (which maybe the thing I like in @Danielkart's file) - or is it not about the tyre grip forces and more about when the slip effects (srub/tear?) start?

slip_crossover_line2_end_y - I'm not sure I understand the difference and relationship between this and the start.... do the two combined define the distance between the 'edge' on the grip side and the slip side?

Do the values fall in range of what's posted in the chart? slip2 - and next question of course, is how should I read that chart

Thanks again for the explanation.... very helpful and hope I'm interpreting it right.

Cheers
Rob

 

Depending on the file used, I noticed some files ago that the way tear was set up caused it to always be felt. So I think that might be it?

 

No problem. I would like you to solve it as well. Have a great night.

 

I have a file in the oven that just needs some magic applied to make Karts semi drivable, not that I have any interest in the karts myself.
It’s still essentially based on 5.0.1 , but the most connected to the car I’ve ever felt. The momentum rack and the physics of each car shine through in a beautifully simplistic way ( by custom file standards) , It’s almost like the steering column is coming through the desk. I’ve done countless sessions of testing different new options but keep coming back to the good old original.
I’m not sure it will be something non DD wheels will run , it will ask a lot of the bases power range, time will tell.
Hopefully just another day away.
Good night all.

 

Yes and No .... I know I don't help you but the right explanation would be the mathematical one and if you open the link on the first entry of the slip in the file you can graphically check how the single variables move the function that describes the slip, or its shape, or its width, and then choose the two positions on it ...
I try to simplify it, imagine that the transition from grip to slip is described as a function that passes from one straight line to another of different inclination and therefore that starting close they move away ... this whole procedure taken as a whole (and therefore all the function that describes it) you can move it without changing absolutely anything forward or backward in time (pos), so I am not referring to any detail (tyre, edge or other) but I refer to the whole procedure whatever it is.

Then within this space in the time we are going to define it :
- you can increase or decrease the weight felt or lost on the steering wheel as it passes from grip to sliding (windth);
- you can move the effect away or closer on the other hand by increasing or decreasing the inclination and therefore the length of the overall passage regardless of the weight chosen (slope);
- and finally within this space so anticipated or postponed, with this weight variation and with this distance of jump, you establish where is the limit of grip on the edge before and where is the limit of the end of the slip after and these moments on the function are defined by the two positions "start y" and "end y"

I don't know if I confuse you or not

 

@Karsten Hvidberg

Want just to say THANKS for the ---- rFuktor 5.0.1.3 ---- !
It has changed for me AMS2 completly and my shame that i didnt seen that before.
Using a TS-PC wheel and i cant stop driving! :Good work!

 

Thank you - that's actually really helped

Will continue to play when I can find time.

 

Just one comment for the day.... for Fanatec users and the FEI (force effect) setting (I’m on a DD1 btw)

There seems to be different schools of thought on what this should be set to. I've always had it at 100%, until recently a couple of people have suggested 30% max as it drowns out some FFB signals.

With FEI set to 30%, I lose all of the granular detail in the feedback due to the smoothing... all of the road texture, micro-bumps, scrub/tear when over and understeering, vibrations etc.... I've just done 15 mins forum reading and to me, FEI is another damper setting - 'smoothing' the input FFB from the game. My wheel when you set FEI to 100 says 'take pure FFB from the game'.

I tried at 80%, which does reduce a bit of the 'graininess and harshness' sometimes, but from my limited understanding, I'd rather any smoothing of the signal be done in the custom FFB file/game and the wheel translate this in a 'pure' fashion to what's output to the wheel forces.

So back to 100% for me.... interested in others thoughts on this one?

 

According to Fanatec you should never go below 80% at FEI because of the latency. If you want to smooth it more, you should increase the INT value.