Discussion in 'Automobilista 2 - General Discussion' started by Karsten Hvidberg, May 30, 2020.
If I have not read it wrong, the differences between the three files are only these highlighted line in the photo below that I am attaching ... if correct, I think you could simplify the development work by making only 2 variants, unify the HIGH with the MED (perhaps adding in the notes the values, if cleared in the med, to be inserted as a possibility) and leave the LOW.
Almost yes as components and hardware (and therefore as performance) but a little different as firmware and software, the sensations, although obviously almost comparable, have slight differences even if of little importance, however, in short, if we start from the G27 as a basis, we have a G29 that compared to the first is a little more informative and grainy in the sensations, and a G920 that is superimposable in all to the G29 but which strangely has (don't ask me why ...) a different (minor) dead zone respect for his twin brother ... for example, on the RD forum these slight differences have been addressed, especially in the sub-forums of the AC and ACC games ... in these games it is possible to upload customized *.lut files calibrated on the specific steering wheel ... who created the *.lut files, received different feedback between the owners of the 3 steering wheels, well, to optimize them properly they created different *.lut files specific for each model. So, although very slight, there are still some differences that can be significant such as the different dead zone between the G29 and the G920 (it is significant but not so much).
What values in game for Logitech? Obviously it is subjective but I think that two approaches can be implemented:
1) High Gain 80-90 (very hard up to the clipping limit), low LFB 20-30 (enough to fill the center), medium FX 50 (not too much to not add to the Gain) and Damping 0 (or if you like max 50 to dampen the peaks due to high gain);
2) Low Gain 40 (the minimum to feel a bit of hardness behind the wheel), high LFB 80-100 (we fill the wheel to the maximum, or almost, the wheel in the center so that it is constant), high FX 75-85 (with gain low do not disturb and season) and Damping 0 (there is nothing to dampen).
For now, with Karsten's files post version 6.6.5.* (as already mentioned above), I prefer the second approach .... we never have clipping, very little pendulum effect, minor variations of ffb settings between one car and another (with 40-90 -85-0 I run 80% of cars without changing ffb settings), and finally, I get a light and information-rich steering wheel ... in the first approach, however, the steering, although more decisive, has more clipping complications, of pendulum effect, and moreover, require greater deviations of ffb values between the various cars which must often be adapted.
Which version do I prefer of * 22.214.171.124 between LOW, MED and HIGH ??? I have always preferred the MED, but this last time, the LOW amazed me (it was a long time since I tried the LOW) ... very rich and tuned to my steering wheel and above all to my brain ... thank to all ... good job!!!
Great info above ! I used to work for a motor control company, and prior to that for a company that made high end/ hand wound multitrack tape recorder motors, and one thing that became obvious was that no two motors are ever the same. So, doing the Wheel Check on your wheel and looking at the response curves will help you understand what your particular wheel is doing, and what you need to compensate for.
Eventually everyone should be able to at least get a good default set up for their wheel/s, and those who have the know-how can custom tune for their own likes. Lots of good work going on here and other places with FFB, and when they get the other bugs tweaked, we should have a great SIM.
I hope that day comes sooner than later, until then I'm just sort of hanging and waiting.
Thanks, Karsten! That helped
Just following Karsten's advice, high force in the wheel settings, low force in the in-game settings And Gain 50 / FF 75% is actually quite light in the McLaren GT3 car.
It's not that extreme force in the Reiza Ultimate either, here's a couple of graphs from the telemetry HUD:
And here's me holding the wheel very lightly, letting it start to get out of contol. This is at the end of the Kemmel Straight. In game gain at 25, Fanatec force turned down to 60:
Oscillations easy enough to control at that strength, but I think something is off causing the FFB get out of control like that... If I had let go of the wheel completely, there is no way I would be able to grab it again, the kill switch would be the only way to stop it.
torsteinvh Try lowering the (balance scale That should reduce the tendency to oscillate.
done some work with lut files once but mostly haven’t delved much into AC series.
still can’t find a nice feel with them. and there so limited in tuning in game. i been trying the last couple of days to get back into figuring Acc out but i just feels average.
I have had some great success using wheel check based force curve modifier that someone maded for pc1 , i have created scoop figures with it to make my wheel very linear in PC2 with some custom files. i haven’t tried it with karstens files though.
I will post a link to the FCM later. it’s very good.
FCM 1.20 tool (google > 'Force Curve Modifier pcars' for download) is really useful tool. You can remove deadzone and correct wheel response curve (make it more linear).
For me, these two lines of code
(output (scoop output 0.50 0.52))
(output (tighten output 0.17 0.012))
transform FFB completely, for the better of course. They are like the lut file in AC/ACC.
(output (scoop output 0.50 0.52))
(output (tighten output 0.17 0.012))
Latest revision just posted: https://forum.reizastudios.com/attachments/silver-raw-6-6-5-38-zip.12084/
* Works well for all cars. This by auto detection of max lateral force per car & then scaling.
* Dual lfb, one for each tyre.
* Better feel of grip loss on higher lfb(@Kuku Maddog , this should help you).
* Experimental Low End file. More differentiated & includes @TomP 's lines just posted in his prior posting to this.
Project CARS Official Forum
i was at
(output (scoop output 0.81 0.44) on my wheel at 75/100/ 100/0/0 in thrusmaster control panel.
Nice work as usaul
Mid file feels good ( testing with JCW mini tonight) out of the box with one slight change of reducing scrub scale as those pesky little rough vibes were creeping back in to files lately.
I think you had already reduced it , but i went a point or two further .
Its overall a bit less stiff or on the lite side compared to previous , so need to adjust settings up for that, no problem for me, but wonder if others who like stronger wheel have the range.
Ok, maybe you can adjust "master_rack" to fit?
If so, which value works well for your base?
Also curious, which version had the best scrub for you so far?
Is it possible to get a version list with the downloads of the last versions after 126.96.36.199? In the past you could find quite a few downloads in the beginning thread, now unfortunately not. I am currently trying to find the best for me for the G27, I think I can say that the "tighten" function is essential. I have some time ago here
Automobilista 2 Custom Force Feedback - Overview & Recommendations
already posted. In V188.8.131.52 "tighten" is built in for the first time (to my knowledge) and that is a quantum leap for low end steering wheels like the G27.
I would ideally need all subsequent versions after ...26 to test more.
Translated with www.DeepL.com/Translator (free version)
yeah i figured master rack was the reason, i dont know just yet, but im enjoying it, I think scrub come in a little in the 30's. , but maybe ive just picked up on it in the mini.
The tighten has been added in the version just posted today again, not as values at the top, but directly at the bottom of the file instead.
I was hoping one size would fit all?
Is it that you don't want the scrub at all or can you describe how it is wrong?
Next revision should be tuning of:
1) Low end & mid version files
2) Tighten for low end
4) High end only if something is somewhat off in that one for some reason.
So any feedback as to how those values make sense is appreciated.
It would be important that the feedback comes in relation to the latest version .38 just posted today.
I might post some interim revisions if some fine tuning values are important for added adjustment options, but nothing that will change the overall settings until we have agreed on some good compromises & file settings.
It’s not that bad, not like when we had issues in pc2, it’s subtle, but not all cars show it. But it’s just a underlying oscillation during understeer, it’s more a personal preference to not experience it . I find it ok at lower scale of 0.1
I have run other files ( Pc2) without any scrub code just fine in the past, so I don’t think it’s meaning I’m missing out on something.
I’m liking the file just fine as it is, but need to test some rwd cars more, as just did a race in tail happy camaro gt4 and wasn’t sure about the oversteer feeling at - value.
Ok, great, I'm simply trying to find some middle-grounds that works for all.
Reg. the over-steer there is also this line down in the file:
(over_steer (- 1 (max 0 (- grip_feel_front (power (- 1 grip_feel_rear) 0.80)))))
The 0.80 is the coefficient for how it should kick in. Linear would be 1.0, of course, so at 0.80 it starts a little faster. Just fyi in case you want to adjust the "attack" of the over-steer as well.
I found the fcm tool, but I don’t know how to use it. Is it used in conjunction with wheelcheck as per the pc2 forums?
Quick instruction for FCM tool
Separate names with a comma.