Discussion in 'Automobilista 2 - General Discussion' started by Karsten Hvidberg, May 30, 2020.
How much gain are you using in-game? Why not just use lower gain if it is too strong?
@Danielkart if some feel it and others don't, you will surely find a solution somewhere with DD friends ... the fact remains, however, that having explored so many roads has served me a lot
For now , when you can, try this with your advice ... i think it is much smoother and more manageable ...
power_steering_angle slightly higher than 0.15
I use 60Gain in game. No, I don't want to go down because I'll lose ffb then. I described that on page 241 with the gain. But thanks, I have my own custom file, but it would have been nice if yours or Stakanov's file would work for everyone
@Karsten Hvidberg the idea of wanting to bring my file closer to @Danielkart 's needs is mine, he is fine with his file, he was just trying to make this possible for me and my file ... and then is him that help me!!!
When others with DD1 said they didn't have Daniel's problems then I hoped there was a simpler solution.
Why do I insist or pursue this need?
Because I suspect that by solving the oscillations on DD at high GAIN we can improve something on the whole range of steering wheels ... we will see, for sure I am experimenting with solutions and values out of the ordinary.
You will not lose FFB going to lower gain. On the contrary you will get more headroom for the downforce gain as well as make it clip less in the file, possibly.
i’ve not had time to keep up with the latest experiments, but i’ve been trying to find a solution for hands off the wheel oscillation for quite some time.
as i think it’s the number one problem. everything else is just a matter of taste.
It became a problem with the 5.0 series of file’s, the main value that introduced the oscillation was the higher understeer scale ,
but lowering that now doesn’t feel good now .
it generally comes on in severity with increased speed ( like above 200kph) and even more so with high downforce cars.
what seems to be happening just from a feeling point of view is that the centre is tightening or becoming narrow as the speed builds.
i also feel it’s something to do with the way LFB is being utilised. as one way of reducing the effect of the oscillation wave , (which you can see in telemetry, ) is to lower the Lfb strength (in file ) to low levels. which makes the lfb curve a gentle slope but this also comes at a cost of sensations.
Another band aid fix is to use a older understeer relax .
so perhaps it’s something that needs to be altered in the deeper code that controls how the LFB behaves at higher velocity. and also a look at what can be done in terms of relaxation of understeer and centre.
@Karsten Hvidberg, @Stakanov and @Danielkart.
I like all your files, the reason I preferred Karsten´s is that that was the one that gave me the best feedback on when I was pushing (too) hard and the front tires were about to lose grip.
The other files were pretty close, and I could probably live happily with either one. All three are very, very good I guess this comes down to personal preference as well. I like to feel the front tires losing grip above other feedback, so Karsten´s file provided me with that.
I have oscillations on all three files, I have just learned to live with it and not let go of the wheel.
Settings currently are in-game gain at 50, LFB at 10, damping at 10. Fanatec base settings gain at 75, damping at 30. No other filters.
On my DD1 in game damping actually make the oscillations worse, that´s why I run damping at the wheel base as well. Need to experiment more with damping and LFB values.
Will test all your files further when I´m back from vacation
I find the self centering is far too fast for me on my simplicity SW20 , Danielkarts last but one file has worked best for me as there seems to be more dampening built in by default .
I also said today that the damping and lfb amplifies the oscillation. Anything above 0 amplifies it. And suddenly people come and say the same problem with the oscillations. I still say that it has gotten worse since 5.01. Stakanov I can't find a way to eliminate these extreme oscillations, sorry. I don't have this problem with my custom file, that's a fact. I'm at a loss with your and Karsten's file, I can't get any further
@Karsten Hvidberg sorry if I disturb, I know that on these formulas you have already told others that everything is as it should be .... but these two changes would be plausible and above all functional fewer oscillations .... is possible only here delete this in bold and red ... with dt now set in static way and equal to 1/300 i have difficult to understand this multiplication and division in red which is practically a little less than 1 and always equal to 0,900900900900 .... and if i set or use dt =1/333 is always = to 1 ... even if i not mistake
(sa_delta_rack (* (/ 1 (* dt 333)) (- steering_angle_rack old_sa_rack)))
(sa_delta_rack (hard_clip sa_delta_rack delta_sa_max))
(sa_delta_l (* (/ 1 (* dt 333)) (- steering_angle_l old_sa_l)))
(sa_delta_r (* (/ 1 (* dt 333)) (- steering_angle_r old_sa_r)))
What is the need for this calculation if now dt is static and not variable???
Sorry in advance do not hate me
EDIT: Try deleted the red code ... or in alternative try neutralized the code (=1) for dimostration or for test with game .... many oscillation go away.
enable this on top
(dt (/ 1 300))
(dt2 (/ 1 333))
and now modify the code like this for result = 1 (and not 0.900900900) !!!
(sa_delta_rack (* (/ 1 (* dt2 333)) (- steering_angle_rack old_sa_rack)))
(sa_delta_rack (hard_clip sa_delta_rack delta_sa_max))
(sa_delta_l (* (/ 1 (* dt2 333)) (- steering_angle_l old_sa_l)))
(sa_delta_r (* (/ 1 (* dt2 333)) (- steering_angle_r old_sa_r)))
If i right for now or even if the dt is static the code in red can be removed or neutralized as in this example
Center_full_ gemischte neue Werte. Relax_us Start wurde reduziert. Kerbs_surface etwas reduziert. Large_Track und Large_offroad reduziert. Federung etwas reduziert. Eine sehr homogene sanfte Handhabung. Sehr empfehlenswert für DD-Basen
Nachtrag: Die Datei kann auch für schwächere Laufräder getestet werden. Reduzieren Sie einfach den Wert dafür
(power_steering_angle 0,15) oder 0,1
I am completely Kukus opinion. Priority is to eliminate these oscillations
I like where this file is going, and I have a CSW 2.5. Just a couple of small changes to make it feel how I like it. A bit less understeery, a little less scrub, and a couple of other small bits and pieces. Feeling good, and basically no oscillating.
@Danielkart you have a wrong code and you need to replace it ... this line
(grip_feel_rear (min 1 (/ (+ (* grip_rl rel_weight_RL) (* grip_rr rel_weight_RR)) (max 0.0001 (+ rel_weight_RL rel_weight_RR)))))
you should change it with this one below, because without the calculation of the rear variables you have to simulate them with the front variables
(grip_feel_rear (min 1 (/ (+ (* grip_rl rel_weight_FL) (* grip_rr rel_weight_FR)) (max 0.0001 (+ rel_weight_FL rel_weight_FR)))))
In your file they are not calculated "rel_weight_RL and rel_weight_RR" ... you are only this for front ...
(rel_weight_FL (smooth FL_load_norm 0.1))
(rel_weight_FR (smooth FR_load_norm 0.1))
ohhh thanks a lot Stakanov for the info will change it as soon as possible
Yes michael, everyone should furnish according to their taste, just as you like it The main thing is that the file works really well
Sorry if I continue but ... the mistake above with rel_weight_rl and rr to 0 change the result of grip_feel_rear now with error equal always to 1
(grip_feel_rear (min 1 (/ (+ (* grip_rl rel_weight_rl) (* grip_rr rel_weight_rr)) (max 0.0001 (+ re_weight_rl rel_weight_rr)))))
when you correct it (and replace the correct variables) you will change the result of the item grip_feel_rear and consequently all that is connected ... like what was remained the yaw of the momentum rack that you had already strongly abolished in your file (no yaw scale, no yaw front, no yaw rear, no relax yaw .... ).
(yaw_l (* (/ wanted_sa_l (* ratio_scale 3.14)) kmh -1 (+ yaw_scale (* FL_stretch yaw_stretch_front) (* RL_stretch yaw_stretch_rear)) 0.05 grip_fl grip_feel_rear))
(yaw_r (* (/ wanted_sa_r (* ratio_scale 3.14)) kmh -1 (+ yaw_scale (* FR_stretch yaw_stretch_front) (* RR_stretch yaw_stretch_rear)) 0.05 grip_fr grip_feel_rear))
This is why your file is the only one that does not oscillate ... it is only a case due to an error ... At this point you can also totally eliminate or delete the yaw from the momentum rack you will get 0 oscillations and it seems physically obvious but it is wrong for physics, yaw exists and must be there ... i think!!!
I know a lot of time has gone into the scrub and tear items in any of these files. But for me I just get a much better sense of grip feel, tire feel and surface feel with turning them both off in the file. I got this idea from Danielkart as he brought it up a few days back to turn off one of them, but I tried turning them both off and the tire feel is really, really good.
The scrub and tear FX add this extra muddiness right at the edge of grip. This muddy feeling changes how the FFB feels when you do the following:
Trail braking or threshold braking - Because the tires are close to sliding or sliding just a bit so tear and scrub FX is added in here.
During mid corner grip- Again because your trying for the best slip angle and the tires are sliding ever so slightly so tear and scrub FX is added in here.
Corner exit if the wheels spin at all - With wheel spin then scrub and tear FX is added in here.
If you don’t think that much scrub and tear FX is added in during the above, then I have a way for you to test and feel the scrub and tear FX while the car isn’t moving. This way you will be able to feel this extra muddiness that is added into the FFB in all the above situations.
Find a RWD car and turn TC off
Go to a track that has a wall on tarmac (Not off road because that will add extra FX to the feel)
Place the front bumper against the wall
Rev up the engine and drop the clutch which will make the rears spin and hold the rears spinning.
You can then take your hands off the wheel and watch the wheel oscillate back and forth even though the front tires are stopped, and the car isn’t moving.
Put your hands back on the wheel and feel the oscillations along with the extra tear and scrub FX from the rears spinning and the front tires are stationary.
All of this extra tear and scrub FX creates this muddy feeling in the above area’s that I listed. The oscillations during mid corner and on the edge of grip is the one I dislike the most as these oscillations are telling me the wheel is pulling back and forth ever so slightly mid corner, so you really loose the feeling of being on the edge.
Once you disable scrub and tear in the file then all this extra muddiness is gone. To disable you need to change the following two lines in the file to the following:
You can then complete the same test above after both have been disabled and you can clearly feel the difference. The wheel no longer oscillates back and forth telling you the rears are spinning. This then clears out the muddiness and makes the FFB so much cleaner during the above listed edge of grip times.
I don’t want to do away with the scrub and tear FX I just think it needs to be provided through the FFB a different way. In a way that doesn’t create this muddy feeling right when it’s the most important time to just feel the tire grip and tarmac under.
That's why the yaw always bothered me. I noticed that it interfered with the ffb so I deleted all yaw_werte. They think then your custom would work without yaw. Very interesting. For me, these Yaw values are not important, because for me they do not fit into the feeling of the steering wheel anyway. Buttkicker or motion platform. But of course it's a matter of taste again. For me it doesn't change anything for my file, but I'm happy that you found the error stakanov
Hey Michael, I too have a CSW 2.5, is there any chance you can upload your tuned file please so that I can compare it to what I am currently using (some tweaks).
Separate names with a comma.