Misc RasmusP's LUTs for G27/29 and DFGT 2.0

[darrian]

I tried the g920 lut and for me it seems super numb and light compared to the recommended lut. Should I try turning the gain in cm beyond 100%?

 

[Cmavridis]

ok i cannot make up my mind here.

first of all congrats for a very nice guide. i used to play assetto corsa on a g27, now 5 years of absense, i got a g29 and competizione.

i have logitech gaming software driver on windows 10 and the settings mentioned here, i have tried these different setups:

1. no lut and aris settings for g29 (50-70% gain, 12% min force, 100% dynamic, 20% road)
2. G25+G27+G29_NoClipButStronger_recommended.lut with 50% gain, 0% min force, 100% dynamic, 20% road
3. created my own lut with wheelcheck and 70% gain, 0% min force, 100% dynamic, 20% road
4. my own lut and 70% gain, 12% min force, 100% dynamic, 20% road

I can unfortunately feel no significant difference between the 4. i mean its different, but i am unable to understand if one is better than the other.

its strange because also look how different the noclipbutstronger lut is compared to my own:

i really dont know what to do from here. i also cannot exactly feel over/under steering with any setting of the 4.

so what would you recommend guys? the game is certainly playable i could do some laps, but i fear i am missing out on something. should i just drive, get better, and compare again in some months? and if so, which setup sounds better in theory?

 

[Interslice]

I had a mess about with these over the weekend having used noclipbutstronger pretty much since i got the into sim racing with the g29. I'm still undecided if i'll stick with the wheelcheck style curve. One thing in it's favour is the consistancy across games and the nicer feeling at low speeds, with no rattle.

I've made noclipbutstronger with 30% less value 1, 20% less value 2 and 10% less for the rest - as rasmus suggested for the g29. I've yet to test it but it could be a winner.

It could also depend on the cars you are mainly using too. Mods such as the rss formula cars or shutoko revival have totally different base ffb setups and are definitely nicer with the wheelcheck curve.

 

[RasmusP]

Sorry to all you guys! I somehow didn't get alerts for your posts!

Gonna answer them bit by bit now

What is LUT? Forgive me for sounding stupid, but I feel normally it's a good idea to always explain abbreviations, at least in the first post. I use a G25, and don't experience any sort of shaking, but am always up for improved setup for better 'feel.'

It's for Look Up Table. I know your struggle.. I searched so long to find out what the heck everyone was talking about.
Somehow I didn't explain it in my download... However if you click on the link in the overview post near the top, it will lead to my old thread, where my sources are linked.

But yeah.. quite messy..

Anyway, it's just a table containing numbers from 0-1 on the left and 0-1 (or more here) on the right.
Left side is the ffb percentage the game is putting out, right side is what gets sent to the wheel driver.
So if you have a wheel where the first 15% are "dead" like with a G27, you need:
0.01 | 0.15
The first of the hundred steps will instead be 15% now.
I also added some "white ffb noise" for 0.00 | x.xx which fills the last little hole when passing the center. It gives some slight vibrations though when not driving. As there is ffb created out of nothing, basically..

All my LUTs are hand-tuned by looking at the graphs in excel. Sadly I don't know how to use curvature smoothing equations to make it more scientific hehe.

Hope that explains it?

 

[RasmusP]

I tried the g920 lut and for me it seems super numb and light compared to the recommended lut. Should I try turning the gain in cm beyond 100%?

Yes, that's normal if you don't have a g920 or you don't have one of the "weird" g920's.
Some have too high internal gain, resulting in the deadzone being only 1-2% instead of the usual Logitech 15%.

If the g920 lut feels way too weak and numb, use the recommended one

Don't go beyond 100%, this will clip too much. There won't be any difference between 80% ffb output and 100% ffb output from the game if you put it to 120 gain.
So feeling the grip maximum in faster corners will be impossible. You'll only feel the wheel going lighter or turn into opposite lock when you lose the rear but you won't actually feel the level of grip.

 

[RasmusP]

ok i cannot make up my mind here.

first of all congrats for a very nice guide. i used to play assetto corsa on a g27, now 5 years of absense, i got a g29 and competizione.

i have logitech gaming software driver on windows 10 and the settings mentioned here, i have tried these different setups:

1. no lut and aris settings for g29 (50-70% gain, 12% min force, 100% dynamic, 20% road)
2. G25+G27+G29_NoClipButStronger_recommended.lut with 50% gain, 0% min force, 100% dynamic, 20% road
3. created my own lut with wheelcheck and 70% gain, 0% min force, 100% dynamic, 20% road
4. my own lut and 70% gain, 12% min force, 100% dynamic, 20% road

I can unfortunately feel no significant difference between the 4. i mean its different, but i am unable to understand if one is better than the other.

its strange because also look how different the noclipbutstronger lut is compared to my own:

View attachment 406300

View attachment 406301

i really dont know what to do from here. i also cannot exactly feel over/under steering with any setting of the 4.

so what would you recommend guys? the game is certainly playable i could do some laps, but i fear i am missing out on something. should i just drive, get better, and compare again in some months? and if so, which setup sounds better in theory?

Hi,
are you sure the LUTs are working?
The recommended lut vs your lut-generator one should feel completely different. Like not subtle, but really night and day difference.
You can check this by deleting the lut file and the ff_post_process.ini from the folder and leaving the minimum force at 0%.
You should have a big dead zone around the center.

Is this there?
My recommended LUT at 50% should feel completely tight around the center in comparison. If you leave the gain at 100%, the wheel should be shaking massively. Same for if you put the minimum force to 12% when using my recommended LUT.

Do these checks to know if my LUT is actually working. FFB settings can have massive placebo effects. You think it's different but in reality your settings didn't even apply and didn't do anything hehe.

So summary:
Put the gain to 100%, minimum force to 0%.
You should have nice and strong ffb but a big fat deadzone around the center.
Put minimum force to 12%. Should feel okay now around the center.

Then apply my recommended LUT but keep the gain at 100% and the min force at 12%.
The wheel should go absolutely crazy on the straights and rotate left and right out of control.
If that's not happening, the LUT isn't working and you need to check the settings.

Report back your results!

 

[Cmavridis]

...

thanks a lot for the reply this is a nice procedure to follow. if i remember correctly i can tell the difference with and without LUT, but i will try to pay attention to what you said and will report tonight.

 

[Cmavridis]

Ok back with some results!

Tried 5 different setups, all with nissan gt3 2018 at SPA, with dynamic damping 100% and road 20% always.

1. No lut, minimum force 0%, , gain 50% gain 50%
Big deadzone in center (5cm approx total left to right), very light wheel in general

2. No lut, minimum force 20%
Big deadzone again, a tad stronger wheel, nothing too significant.

3. 50% lut file, minimum force 12% , gain 50%
Deadzone eliminated. Stronger overall wheel, felt the kerbs lightly also, left-right movement on straights without holding the wheel.

4. Wheelcheck.exe generated lut file, minimum force 12% , gain 50%
Same as 3 but with lighter wheel. No deadzone, stronger than 1-2 but nothing fancy. Kerbs not like 4

5. Wheelcheck.exe generated lut file, minimum force 12% , upped gain 70%
Same as 4 but with a tad stronger wheel. No deadzone, kerbs again not like 4, did not feel like a 50->70% increase in gain.

So the lut file selection works, and i feel like i should stick to the 50% lut file that you posted with min force 12%-gain 50%-dynamic 100%-road 20%. Although i feel like i will get tired easily, its kind of strong like this. Do you think i should drop the gain ingame? Or remove road-min force?

I still feel like the correct thing to do theoretically is use the generated lut as its made for my specific wheel, but it doesnt feel as good as the 50% lut

edit. did my first 10 laps with settings #3 and i think i have a winner. thanks a lot @RasmusP

 

[RasmusP]

Ok back with some results!

Tried 5 different setups, all with nissan gt3 2018 at SPA, with dynamic damping 100% and road 20% always.

1. No lut, minimum force 0%, , gain 50% gain 50%
Big deadzone in center (5cm approx total left to right), very light wheel in general

2. No lut, minimum force 20%
Big deadzone again, a tad stronger wheel, nothing too significant.

3. 50% lut file, minimum force 12% , gain 50%
Deadzone eliminated. Stronger overall wheel, felt the kerbs lightly also, left-right movement on straights without holding the wheel.

4. Wheelcheck.exe generated lut file, minimum force 12% , gain 50%
Same as 3 but with lighter wheel. No deadzone, stronger than 1-2 but nothing fancy. Kerbs not like 4

5. Wheelcheck.exe generated lut file, minimum force 12% , upped gain 70%
Same as 4 but with a tad stronger wheel. No deadzone, kerbs again not like 4, did not feel like a 50->70% increase in gain.

So the lut file selection works, and i feel like i should stick to the 50% lut file that you posted with min force 12%-gain 50%-dynamic 100%-road 20%. Although i feel like i will get tired easily, its kind of strong like this. Do you think i should drop the gain ingame? Or remove road-min force?

I still feel like the correct thing to do theoretically is use the generated lut as its made for my specific wheel, but it doesnt feel as good as the 50% lut

As I wrote a few times in multiple, imo the generated luts are not usable. It measures how far the wheel will move with a certain amount of "push".
This doesn't really take into consideration that the friction might not be linear. If you push the wheel twice as strong, will it move twice as far?
And does the "throwing" actually has any relation at all to how your body perceives ffb forced?
I highly doubt this all...
It's a cool thing to play around with and honestly the coding is absolutely awesome!
But I just don't think that it's usable for us...

I had a little laugh though at your "100% is quite strong". My sim racing buddies all use thrustmaster ts-pc, fanatec csl or csw or even a direct drive wheel.
The Logitech has a maximum of 2.5-3.0 Nm of torque. My fanatec wheel has 8.0 Nm maximum.

When you want to really feel the grip, you need to feel how the steering gets heavier when the aero starts to work above 100 kp/h, for example.

Anyway, you could just use my lut and reduce the gain. If the dead zone comes back, raise the minimum force a bit.

About my lut and minimum force though:
It has about 18% min force inside the lut. So you should put min force to 0%!
When gain is at 50% ofc. If you reduce the gain, raise the min force.

My lut + 50 gain + 12 min force must be shaking a lot on the straights?

 

[Cmavridis]

My lut + 50 gain + 12 min force must be shaking a lot on the straights?

ah i didnt see your reply and edited my own. i kept your lut and 12% minimum as well, i didnt mind the rattling on the straights, its ok when i hold the wheel, i believe i feel more kerbs etc at 12%min. what you say about lut generation is 100% correct i have no comment whatsoever.

i think i ll keep the 12% minimum - i managed to consistently drive at spa and got a first 2.26 time without much effort. perhaps after 100 laps i will try with 0% min to see how it goes laptime-wise... so i guess i am good to go thanks again, you made one sim racer happy today.

 

[RasmusP]

ah i didnt see your reply and edited my own. i kept your lut and 12% minimum as well, i didnt mind the rattling on the straights, its ok when i hold the wheel, i believe i feel more kerbs etc at 12%min. what you say about lut generation is 100% correct i have no comment whatsoever.

i think i ll keep the 12% minimum - i managed to consistently drive at spa and got a first 2.26 time without much effort. perhaps after 100 laps i will try with 0% min to see how it goes laptime-wise... so i guess i am good to go thanks again, you made one sim racer happy today.

Nice to hear!

To give some further explanation:
I created my luts, because I was unhappy with how the minimum force setting felt.
It's not smooth.. It simply puts the beginning of the ffb to 12% for example and from there everything is linear.
You can see this with the graph of the lut from wheelcheck too, massive bump at the very beginning.

This leads to either a dead zone around the center or it leads to a "bump" around the center of you set too much minimum force.

So I manually tried values in excel that looked like a smooth curve and were somewhere around to 10-18% force at the beginning.
But all my luts end close to 0%!

After testing around a lot, I put in a value for 0%. I call it "white ffb noise". It's random ffb, created out of nothing.
But it gives some little resistance at the very center when going straight or driving through chicanes.
I manually dialed this very first value in. Too high and the wheel goes crazy while not even driving!
Too low and the little hole at the center comes back.

Anyway...

The only purpose of my luts is to have a smoothly shaped curve for the minimum force instead of a hard slider that isn't smooth.

My luts don't do anything else.. No shaping, nothing.
Apart from the recommended lut, which kinda tricks the clipping meter and allows higher forces sent to the wheel.
But this might be placebo.. Somehow the starting curve feels the best with the recommended lut though so I never changed it..
I couldn't recreate this feeling with a lut that scales to 100% gain...

Eh... Anyway again?

So using anything else than 0% minimum force in combination with my luts is against the purpose of them hehe.

Not that I would care, everyone should just set things up how they feel best! Absolutely fine for me if you like the 12% min force!

But I wanted to explain this

 

[Cmavridis]

But I wanted to explain this

oh i see now... thanks for further explaining, it all matches of course with the graph i see in excel.

so actually its what aris the greek dev suggests for g27-29 (min 12%) but implemented in a better way.

i cant help by trying anoher 10 laps wih your lut and 0% min. i never fight a good theory/implementation.

 

[RasmusP]

oh i see now... thanks for further explaining, it all matches of course with the graph i see in excel.

so actually its what aris the greek dev suggests for g27-29 (min 12%) but implemented in a better way.

i cant help by trying anoher 10 laps wih your lut and 0% min. i never fight a good theory/implementation.

Hehe yeah the excel graphs are extremely helpful. I basically just tested around while looking at them until I was like "yeah that looks kinda smooth, let's take it for a spin".

Interestingly project cars 1, that isn't known for the best ffb has a very nice setting:
"dead zone removal" and "dead zone removal - fall off".
One is the minimum force basically and the "falloff" slider sets how close to 0% it drops close to the center.

With these 2 sliders you can dial it in pretty smoothly!

I started my sim racing journey as a kid with a thrust master belt driven wheel. Only 180° and poor ffb quality but it had no dead zone at all!

Didn't drive much with it from 11 years old until I turned 23 and project cars got hyped.
Bought a g27 as my thrust master didn't have drivers for anything past windows XP.

Deadzone was annoying but with the 2 pcars 1 settings I got it dialed in nicely.

Then it got more serious after 2 years and while AC drove a lot nicer, the Deadzone annoyed me massively...

 

[Yeeters]

Saw someone say this worked for ACC, was that a typo or does the LUT work?

 

[RasmusP]

Saw someone say this worked for ACC, was that a typo or does the LUT work?

It works
There's a guide on steam.
Sadly I didn't have my g27 anymore when acc updated to LUTs being possible so I could never test it myself.

But the steam guide creator approached me here via pm for permission to link and copy everything + change for acc.
Gave him some feedback after it was done

In the end you simply need to throw in the files into the acc config folder in the c: documents.
Here's the link at steam:

Steam Community :: Guide :: Logitech G29/G27/DGFT Force Feedback LUT Settings

After the v1.0.2 update , force feedback post-processing is now available for Assetto Corsa Competizione, just like in the OG Assetto Corsa. Making it possible...

steamcommunity.com

 

[Yeeters]

It works
There's a guide on steam.
Sadly I didn't have my g27 anymore when acc updated to LUTs being possible so I could never test it myself.

But the steam guide creator approached me here via pm for permission to link and copy everything + change for acc.
Gave him some feedback after it was done

In the end you simply need to throw in the files into the acc config folder in the c: documents.
Here's the link at steam:

Steam Community :: Guide :: Logitech G29/G27/DGFT Force Feedback LUT Settings

After the v1.0.2 update , force feedback post-processing is now available for Assetto Corsa Competizione, just like in the OG Assetto Corsa. Making it possible...

steamcommunity.com

Cool, thanks so much! I used your guide for AC and it helped a lot, so thanks for making it!

 

[RasmusP]

Cool, thanks so much! I used your guide for AC and it helped a lot, so thanks for making it!

I created the luts etc for myself and simply shared the results haha.
Writing all this stuff down took quite some hours though...
Happy that others like it too!

 

[darrian]

Yes, that's normal if you don't have a g920 or you don't have one of the "weird" g920's.
Some have too high internal gain, resulting in the deadzone being only 1-2% instead of the usual Logitech 15%.

If the g920 lut feels way too weak and numb, use the recommended one

Don't go beyond 100%, this will clip too much. There won't be any difference between 80% ffb output and 100% ffb output from the game if you put it to 120 gain.
So feeling the grip maximum in faster corners will be impossible. You'll only feel the wheel going lighter or turn into opposite lock when you lose the rear but you won't actually feel the level of grip.

Awesome thanks so much for the response

 

[Deankenny]

hey @RasmusP me again Had quite a break from the game aand now have a new system.
Still got the custom LUT you made me and forgot everything on how to use it, don't even remember what FFB is meant to feel like lol.

What do you usually set your in game % too, the one you change the the numpad + and - keys?
I remember the tiny vibration/shake you get when sat still, not an issue, but I get some quite violent shudder, shaking of the wheel when slowing down, is this normal.

Im using a G29 still, only difference is no longer using LGS, but using the newer Ghub instead.

As always, nice to see you still helping everybody.

 

[RasmusP]

hey @RasmusP me again Had quite a break from the game aand now have a new system.
Still got the custom LUT you made me and forgot everything on how to use it, don't even remember what FFB is meant to feel like lol.

What do you usually set your in game % too, the one you change the the numpad + and - keys?
I remember the tiny vibration/shake you get when sat still, not an issue, but I get some quite violent shudder, shaking of the wheel when slowing down, is this normal.

Im using a G29 still, only difference is no longer using LGS, but using the newer Ghub instead.

As always, nice to see you still helping everybody.

Hey

Depending on the LUT, I used either 50% or 100% for the overall gain in the menu.
Then when in the car, using numpad +/- , I start with 100% and then go up and down until the center of the wheel feels perfect.

Too high and you'll get a "notch", too low and you'll get a deadzone/hole in the center. Just dial it in until it's smooth.

My lowest car was around 60% I think and the highest car about 140%. So feel free to dial in whatever you want! I wouldn't go below 50 or above 150 though.. Then there's something wrong somewhere