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

[Mikke73]

I did some more testing with new LUTs, and I'll stick with 80 LUT at least for now. Although I liked the heavier feel of 100 gain LUT, I noticed that I was able to feel slides better with the 80. And that is more important to me now.

 

[RasmusP]

Sounds like a plan and makes a lot of sense!

I've found out that at about 65% of gain (cars being 100%), the clipping level is right at the edge of the grip and kerbs are peaking into clipping.
Calculating at what percent of the physics engine this would be, since the clipping meter always clips at 100% output:
100% physics * 100% gain = 100% output

X % physics * 65% gain = 100% output

=> 100% / 65% = 154% physics

Physics aren't really in percent though. But the calculation for the clipping limit makes sense like this.

It's more like "with 65% gain, your physics output, right before clipping happens, can be 154% of the physics output that you have with 100% gain right before clipping happens".

The 80 LUT clips at 124% physics output (100/80 = 1.24), so you gain at least some detail at the limit compared to 100% gain.

 

[Mikke73]

Just some thoughts:

Looks like I'll be going between new 80 and 100 LUTs for some time, it's hard to decide when both have their own positive sides.. I've tried maybe a dozen very different cars with both, and the choice between the two feels impossible.
Of course the 100 LUT feels closer to a real car being heavier, but I really doubt if there's any sense in caring about this when using G29.

Actual question:
Do you maybe have any recommendations of car/track combinations that are in your opinion good ones to test with?

 

[RasmusP]

Looks like I'll be going between new 80 and 100 LUTs for some time, it's hard to decide when both have their own positive sides.. I've tried maybe a dozen very different cars with both, and the choice between the two feels impossible.
Of course the 100 LUT feels closer to a real car being heavier, but I really doubt if there's any sense in caring about this when using G29.

Looks like I should create a 90 LUT for you

Actual question:
Do you maybe have any recommendations of car/track combinations that are in your opinion good ones to test with?

From my experience:
Mugello and Silverstone! Both have very flat and calm parts, lots of left-right-left-right parts to check if the center is smooth and some wilder parts.
Mugello was my test track, but it's quite a wild ride. Silverstone gives you more time to feel the grip limit.

For the cars:
Tatuus FA01 is a great car to test things in general. Not too fast, but very agile, stiff and quick.
Or any of the GT3's, with the McLaren 650S being my favourite to test LUTs with.

 

[Mikke73]

Thank you for the recommendations! Maybe they'll help me to decide sooner.

There's always the possibility to change car-specific FFB if I really need to, so there's really no need for any additional LUTs

 

[RasmusP]

There's always the possibility to change car-specific FFB if I really need to, so there's really no need for any additional LUTs

Yeah, but the gain also changes the first part around the center too.
Since you were using a higher gain with the recommended lut, you're not too picky about it, but I've got the pattern for the CORRECT lut scaling now in Excel so it's a sub-5-min-job to create any lut with identical center area but different clipping level

 

[Mikke73]

Yeah, but the gain also changes the first part around the center too.
Since you were using a higher gain with the recommended lut, you're not too picky about it, but I've got the pattern for the CORRECT lut scaling now in Excel so it's a sub-5-min-job to create any lut with identical center area but different clipping level

I don't think a 90 LUT would help me, with some cars even the 100 LUT felt a bit on the light side - not surprising, considering what I used before
I assume that all I need is just more testing to get used to these settings, but so far the differences haven't been big enough to make me decide which one to use. And that's why I asked what you'd recommend to test with

 

[matticiao]

Sounds like a plan and makes a lot of sense!

I've found out that at about 65% of gain (cars being 100%), the clipping level is right at the edge of the grip and kerbs are peaking into clipping.
Calculating at what percent of the physics engine this would be, since the clipping meter always clips at 100% output:
100% physics * 100% gain = 100% output

X % physics * 65% gain = 100% output

=> 100% / 65% = 154% physics

Physics aren't really in percent though. But the calculation for the clipping limit makes sense like this.

It's more like "with 65% gain, your physics output, right before clipping happens, can be 154% of the physics output that you have with 100% gain right before clipping happens".

The 80 LUT clips at 124% physics output (100/80 = 1.24), so you gain at least some detail at the limit compared to 100% gain.

Does the new LUTS work with DFGT? They're good with it?

 

[RasmusP]

Does the new LUTS work with DFGT? They're good with it?

Sadly not. The DFGT has way smaller Deadzone in the center than the Gxx wheels. If you'd try the new LUTs, your wheel would stark shaking like crazy.

The DFGT LUT is pretty much the best I could do though.

Maybe I'll try to put it through the same scaling algorithm like the new Gxx LUTs to give the DFGT a bit less clipping (also lower ffb strength on average though!).
I'm not sure how easy that would be to do.

 

[matticiao]

Sadly not. The DFGT has way smaller Deadzone in the center than the Gxx wheels. If you'd try the new LUTs, your wheel would stark shaking like crazy.

The DFGT LUT is pretty much the best I could do though.

Maybe I'll try to put it through the same scaling algorithm like the new Gxx LUTs to give the DFGT a bit less clipping (also lower ffb strength on average though!).
I'm not sure how easy that would be to do.

Aye, I find DFGT force not that hard. Often I feel it's lightern than what could be. I am comparing BeamNG and Assetto Corsa, on BeamNG it's heavier, in Assetto it's hard to find a good compromise. Your LUT work well doing so.

 

[Mikke73]

Choice is made, RealNew100Gain LUT was much better.

I actually couldn't find significant difference with McLaren 650S GT3, maybe because my setup for it eliminates slides a lot. But then I tested the GPL mod at Fuji 1968, and the difference was huge. Those cars don't want to go in a straight line at all, and with gain 100 I could feel everything that the car tried to do, in a way I've never experienced before. With gain 80 LUT I lost a lot of that feel, so it was actually an easy choice after all.

Have to say this once more: great job!!

 

[RasmusP]

Nice!
That makes sense, although it's clipping earlier, you get more ffb strength a bit earlier so you notice when the grip goes off also a bit earlier.
You basically trade details vs. overall grip information.

Here's the 90% gain LUT though

RealNew90LUT.zip

Shared with Dropbox

www.dropbox.com

 

[freddyhung]

is there a way to get a 90LUT for the G920 ? i've tested the 90LUT of your last reply and the wheel shakes with the cars standing still , i´ve tried lowering the gain and also adding minimum force to no avail!. Thanks in advance

 

[RasmusP]

is there a way to get a 90LUT for the G920 ? i've tested the 90LUT of your last reply and the wheel shakes with the cars standing still , i´ve tried lowering the gain and also adding minimum force to no avail!. Thanks in advance

In theory the G920 and G25/27/29 should behave identically but for some reason some G920 have a different gain setting internally and you can't adjust that.

The first value of the LUTs is basically "white noise" for FFB at 0%. If your internal gain is higher than that of the G27, you get a crazy wheel.
The AC gain doesn't scale that value, since it's at 0%, so you can't adjust it.
Looks like this:

What you might need:

That's just the very first line of the LUT changed from 0.100 to 0.000

Just open the .lut file with any text editor, change the first line to 0.000|0.000 and the shaking while standing still should be off.

The next step is to reduce the gain until it feels good for you.
Can you test this and then tell me the car, the AC gain and the gain for that specific car that feel good? Preferably a standard AC car, not a mod.
(Numpad +/- changes the per-car-gain or "car specific multiplier").

 

[Alphamox]

Hello Rasmus,

I recently uploaded your recommended LUT for my Logitech G25 wheel.

I recently changed the stock wheel to a 14 inch cheapo/faker Nardi wheel.

What adjustments would need to be made now that I'm using a bigger and slightly heavier wheel??

Look forward to your feedback.

 

[RasmusP]

Hello Rasmus,

I recently uploaded your recommended LUT for my Logitech G25 wheel.

I recently changed the stock wheel to a 14 inch cheapo/faker Nardi wheel.

What adjustments would need to be made now that I'm using a bigger and slightly heavier wheel??

Look forward to your feedback.

ouff, that's a big diameter for the little Logitech motor

You can only trade "more ffb earlier on" against linearity and clipping.
Basically, you can make 35% AC output = 90% ffb output, but then 36-100% ffb output will be squashed into only 10%.
I'll ask ChatGPT to create a curve that does that but keeps the same first 20%, since they are to counter the deadzone.

I'll reply again tonight with the new "Super Strong but Squashed like Heck" LUT

 

[Alphamox]

Hello Rasmus,

Appreciate your quick response, I'm quite surprised to be honest.

Ideally using a bigger diameter wheel isn't the way to go with such a small unit, but I like it as it replicates driving a real car. I've also ensured that the weight differences in the wheels are marginal and not too extreme. The adapter and wheel themselves came in a kit specifically for the G25 - this was purchased from a racing sim shop on eBay.

I did have the opportunity to look back on your previous posts and saw you released other LUT files. Could using the 100LUT file make a difference or should I simply wait for that "Super Strong but Squashed like Heck" LUT ??

Really what I'm after is the same strength/feeling I had with your legendary recommended LUT, on the bigger wheel. I acknowledge that there are limitations on this, and some trade-offs will need to be made.

Again, appreciate your quick response and passion for us dinosaurs in sim racing

 

[RasmusP]

Hello Rasmus,

Appreciate your quick response, I'm quite surprised to be honest.

Ideally using a bigger diameter wheel isn't the way to go with such a small unit, but I like it as it replicates driving a real car. I've also ensured that the weight differences in the wheels are marginal and not too extreme. The adapter and wheel themselves came in a kit specifically for the G25 - this was purchased from a racing sim shop on eBay.

I did have the opportunity to look back on your previous posts and saw you released other LUT files. Could using the 100LUT file make a difference or should I simply wait for that "Super Strong but Squashed like Heck" LUT ??

Really what I'm after is the same strength/feeling I had with your legendary recommended LUT, on the bigger wheel. I acknowledge that there are limitations on this, and some trade-offs will need to be made.

Again, appreciate your quick response and passion for us dinosaurs in sim racing

Thanks man

Yes, the 100 LUT is already a step in the right direction. It starts clipping a bit earlier compared to the 80 LUT (which is identical to the "recommended", just 80 gain instead of 50 gain).

But I doubt that it'll be enough.
You probably need to go from the 80 lut to the 100 lut and 3 times further into that direction

 

[Alphamox]

Thanks man

Yes, the 100 LUT is already a step in the right direction. It starts clipping a bit earlier compared to the 80 LUT (which is identical to the "recommended", just 80 gain instead of 50 gain).

But I doubt that it'll be enough.
You probably need to go from the 80 lut to the 100 lut and 3 times further into that direction

Lmfao.

I'll just upload the 100lut file and be done with it.

So all I need to do is download it, place it in the appropriate directory and select it through the AC controller ffb menu.

 

[RasmusP]

Lmfao.

I'll just upload the 100lut file and be done with it.

So all I need to do is download it, place it in the appropriate directory and select it through the AC controller ffb menu.

Yep
Please report back if the 100 lut is strong enough or if you need more. Didn't create the "super strong" yet, I'll wait for your feedback.