Automobilista 2 Custom Force Feedback - Overview & Recommendations

Just tried with the original coding in, no comments. It now doesn't crash when loading V8's and Bathurst, but instead crashes when you try to Exit the track. Thought we were on a winner.

 

Until yesterday it was ok, I will try again tonight at home, last night they made an update of a few Mb that I did not test, but it is clear that it is the game that behaves strangely in this combination of car/track, how can affect the exit of the race a ffb file?? And whatever it is,...
Anyway no problem we will find what does not digest that goes into the output of the file, provided that I also encounter this problem!! Always available...

 

Here I do not detect anything abnormal, even coming out of the race
I also try to do other tests

 

I know that I could have taken the screenshot at the exit of any race or just entering the game (if you have any doubts I can only make you a video), but after launching the game 15 times by selecting the race in combination Super V8 / Bursturst , at the exit I had no problems.

This is the screenshot at the exit of the race from which you can also view the version of the game that I have on my pc (after the further update and using the ATMOS 1.4 FIX file).


At this point I would not be strangled that in a few days the file can be restored to the previous conditions, but since I am not the creator, this result is enough for me for now.

I have the 5 ATMOS files ready in this 1.4 FIX configuration (from 5.0.1 to 5.0.1.4) I will wait to be sure that the game and the file are stable and then I will publish them, it makes no sense if I have to rearrange something.
I also don't forget that I also promised you a post about errors not detected by users on this forum, which do not create crushes to the game, but which are sold like gold in the market.

 

Here is all ok, same version game.

 

Well you guys are lucky. I have turned everything off that is normally on, and still get the CTD. Oh well, I'll just stick with the modded 1.4 or the No Yaw versions. Still great versions.

 

Yes @Michael Enright you are right, everything works with modded version 1.4 (not the original where still CTD) so I'm talking about Manolo's ATMOS 1.4 FIX both with YAW and without YAW

 

That's the strange thing. The original 1.4 Without Yaw works without CTD, it's only the one With Yaw.

 

I've been trying The "ffb_custom_settings - ATMOS 5.0.1.4 - MOD V1.4 FIX (WITH YAW)" and like it, except for the "front bite". At this point, I guess, my problems has more to do with FFB perceptions than physics engine.

What I'm seeking: more FFB around the center - example: turning the wheel in a long straight should produce more FFB strenght, I mean, the FFB should depend more on speed and less on wheel angle (effective lateral forces). As a matter of fact, it could be lighter, that it is now, in slow turns (<100Km/h, for instance).
I want the FFB to have a more logarithmic curve than an exponencial one (it appears to have now).

What I may need: Where (if possible...) can I get some descriptions of custom file variables?

 

Yes, I think it is a subjective feature to be adjusted to your liking, moreover, for now the bite you are missing I think it is also due in a small part to the correction I made to the file to avoid the conflict, I had created a perfectly functional FIX 2, but yesterday the game has been updated to version 1.4.1.2 and the lines of code I modified have become incompatible again ... I am close to the solution but Reiza changes the cards on the table from time to time, I only know that in this version 1.4 FIX the file is stable.
But let's get to your requests, in addition to a bit of gain and lfb I would try to vary the values of these lines of code a little.....

(velocity_res_scale 0.34) #How much velocity affects resistance on steering (turn up to reach more easily the target angle_d_l & angle_d_r during the curve).
(velocity_res_offset 0.66) #How much is the resistance on steering at velocity 0.
(vel_rot_scale 1.04) #How much velocity affects lateral force (high for more travel_l & travel_r or down for less). Use for filler wheel during rotation of steering wheel.
(vel_rot_limit_coef 1.08)

This may help you whatever file you use.
------------------------------------------------------------------------

For the others tomorrow I will publish the 5 ATMOS FIX (from 5.01 to 5.0.1.4) ... for now I don't want to go after the "Reiza variable" for a single track/car combination.

 

Thanks!
In fact, I started changing what I think leads me to what I want and had already changed vel_rot_scale to 1.5 (from1.04) and power_steering_angle to 0.1 (0.3) to great effect!
The wheel (TS-PC) is much tighter around center and the lack of precision (one of my complains) is greatly reduced.

What is angle_d_l/r and travel_l/r?

 

I tried this on my TGT- II it feels quite good but a bit too strong, for the vet_rot_scale i have it at 1.20. Currently only tried one car and track so far, further testing required.

 

I'll test my version a little further... I don't want to arrive tomorrow and find out I was biased.
I'm trying to get to the felling of front end I get in RR and rF2, and I could have it!!
Tested: Porsche GT3, MacLaren and Mercedes GT1 and Formula Ultimate Gen 2. They are all fabulous!!

FFB in automobilista has the potential to be the best. Custom versions are the way to go (not even rF2, the most modable of current sims, has this).

I'll post my changes if they proven to be good.

 

First, Travel_l/r is the work or force for turn in velocity relation and with vel_rot_scale fill the center, but this work is affected from vel_rot_limit_coef if you low this second voice you can reduce the first, if you high the second you have help with high first value to fill. Is ok 1.5 and 1.25 or 1.26 and 1.14 for most car (i like it), but try at Kansai the formula Cart in the last short but angled and fast curve before the last chicane, the front of car slips without grip in an anomalous way (the travel and the force slip and disappear), it is highlighted only there because there are few sharp curves at this speed (320 Kmh). I used 1.20-26 before, but I preferred to recalibrate the balance with lower values 1.08-1.04 and then modify others voice, Karsten set this to 1.00 and 1.00 but if you prefer 1.5 this is your taste then ok. This are the code for velocity and travel_l/r....

## limit rot velocity
(a (abs kmh))
(f (* vel_rot_limit_coef 1.1 0.004))
(a (- 1 (/ 1 (+ 1 (* f a)))))
(a (/ a (- 1 (/ 1 (+ 1 f)))))
(a (/ a (max 0.000000001 (abs kmh))))
(n 1)
(pn (/ (- 1 (/ 1 (+ 1 (* f n)))) (- 1 (/ 1 (+ 1 f)))))
(kmh_l (* kmh a (/ n pn)))

(travel_l (* vel_rot_scale 0.001 kmh_l))
(travel_r (* vel_rot_scale 0.001 kmh_l))

-------------

Second, angle_d_l/r, search "velocity_res_scale" and watch the code under ... you stop here...

(angle_d_l (- wanted_sa_l steering_angle_l)) # diff from where it would have gone byitself to actual steering
(angle_d_r (- wanted_sa_r steering_angle_r))
(angle_d_l (* angle_d_l 2.3))
(angle_d_r (* angle_d_r 2.3))
(vel_res (+ velocity_res_offset (* velocity_res_scale (min 1 (* kmh 0.004)))))
(angle_d_l (* angle_d_l vel_res))
(angle_d_r (* angle_d_r vel_res))

Then angle_d/r is where it would have gone byitself to actual steering, so if for example you want to keep the steering at 30° during the curve it represents the force you have to keep in order not to return the steering to its natural position ... in this the speed takes over (velocity_res_scale), the faster you go (vel_res), the stronger your strength because the variable angle_d / l is multiplied by vel_res ... more speed means more resistance to keep the steering curved at the target angle.

 

I continue the crusade against philosophy, as I said I have prepared 5 files all in ATMOS 1.4 FIX format and all with the same harmony and let's see why.

I remind newbies that one thing is the code of the file that normally we not touch (which follows the development done by Karsten 5.0.1-5.0.1.3-5.0.1.4-etc ....), one thing is change only determinateds coefficient in the code of Karsten (change this is equivalent many time to change the phisic or modify it) and another is change the items to be set according to your tastes (many variables normally to be set at the beginning of the file from users), furthermore these variables can be varied up or down by aggregate and to be precise we have 3 aggregate macros, the master_rack (all items for the rack and road) , the master_road (all the items concerning the road texture, bumps, details, kerbs, rear and suspension) and the Effects scale (all the items concerning the engine, scrub, tear and flatspot), and after we have the cursor in game to define and how you set your hardwares.

Without being an expert you will understand that comparing one file with another makes no sense if you do not equal the starting point, then the solutions are two:
- use the same file version and therefore the same underlying code, varying only the many variables (this is a job exclusively for tuners and only with this you can achieve very different results);
- or follow the reverse approach, use the same tuning with the same values in the variables and change the underlying code using the version that suits you best to get the result you want (to do this, however, you need to know well the code and know how to modify it).
The ideal, therefore, would be not only knowing how to read the code to choose the most suitable version, but it would also be useful to know how to modify it, and also to know how to tune it.

Only with this premise do you understand that when you say a preference you have not explained anything, because the files are many, the variables involved in the files are so many, moreover, you have different brains, different sensations and different equipment....

I follow Stakanov's project not because it is the most beautiful one file but because it can become whatever it wants, whatever version or file you like, moreover, on the chosen version you can apply any tuning of your liking and this leads to a clash with the jealousies ... unfortunately the users cannot distinguish all this background, they fall in love with a version, they make up their minds and the stadium cheer is nice and ready. Saying Stakanov's File is better than X doesn't make sense, saying that X is better than Stakanov doesn't make sense in the same way ... but it can become whatever you want, others no because it does not have the extra items that can be deactivated or activated or modified, it just has more elasticity and possibilities.

It's only a game or test, here we have ALL ATMOS 1.4 FIX with the same tuning, same voices, same values, and only the progressively different code, which starts from 5.0.1, passes through 2 hybrids, arrives at 5.0.1.4, ending with the latest ATMOS. If you insert the files one at a time and test with the same machines and tracks, you will compare the little difference and different codes without the variable of the different tuning ... the funny thing is that I can't wait to take the next most famous file, copy the tune and propose it in these 5 forms ... I forgot after correcting the errors!!!

 

The very slight differences are felt (on the wheels, on the movement, on the suspension, on the center, on the effects and everything else) and practically if I like 5.0.1 or 5.0.1.4 in general, but I want the 5.0.1 or 5.0.1.4 yaw, i just replace the relative lines of code, if I want the 5.0.1 suspensions in 5.0.1.4 I do the same, on the contrary, if I generally prefer 5.0.1 but I want the rear 5.0.1.4, I do the opposite.
Before, if I was comparing two files of two different guys, I didn't understand it (because too many variables changed between two differently tuned files) now I understand it because the change is for single steps and the comparison between the files is not affected by the different tuning.

 

Keep an eye on FFB clipping (tyre/suspension HUD). It requires lowering car specific FFB.
I'm only have clipping in high-speed turn on braking hard over kerbs - and that's ok for me as we don't need detail in kerbs, what we need is a really good shake as "get out of where!"
BTW, I have my wheel driver @75% max, so it may change thinks a little.

Thanks for your time putting this info together.
I surelly have not achieved what I want (we tend to overcompensate what's missing), but my FFB has changed a lot to a point my main concernes about having nice multi-class races are not physics or FFB - is about perfecting Ai

 

Let's change the subject, some of you will remember the user RaceRa**Ra**, he left us some time ago on this forum with the promise (threat) to come back over and over again and with more accounts, he had attacked practically everyone on the forum to the point of being banned from Reiza for the violence and reactions he had shown on this platform.
Now I have a list of at least 10-15 accounts (this time almost all peaceful and experienced after a few accesses to the forum), who say the same things, use the same words, have the same needs, do even the same things and use the game with the same finality and with the same physics necessity with the same values. Every now and then a new one comes up already trained to put winners or likes.
Sometimes they don't write all together and disappear (maybe they even go on vacation together), then they come back and one to one they all write little by little (they probably also work together) and maybe they even go to bed together.
It's really amazing how much people like each other sometimes, I come to think they could even be the same person. To me in 3 years it has happened here only with 2-4 people and in general we use the same file but a lot of different settings, for the rest it was more arguments than loves.

On top of that, here we have tuners or modifiers who are also friends of youtubers but we will tackle this as well with a "crisp" review, but not for now, only after the review of the corrections code not detected for ignorance and shared by all oblivious to this.

 

Forget it, it's not worth it, they all love them equally and most moderate people are silent and don't take part in these supporters, don't fall into their mistake, ignore them !!!

Reviewing the files I find the movement of the 5.0.1, hybrid 1 and hybrid 2 versions very beautiful, but I like the feeling of 5.0.1.4, the first 3 files have different suspensions, have you tried to move this code to the final ATMOS, the result is not bad.

 

@Manolo I would let it go, if it is really as it seems, I am convinced that this is of no use, indeed do not correct anything anymore, let them build their castle of untruths and leave them convinced of what they claim.
Anyone who knows the code a little, knows and is silent, I can only feel sorry for the user who cannot objectively evaluate the situation because he can not help but upload a file and give an emotional and hasty judgment.

@AntoAntoDD I'm going to test this combination.