Fanatec CSL Elite + and BMW GT2 wheel

[glen64]

Anyone recommend some good settings for this wheel and base please? currently only using an AC setup, but prefer options , so if somone has a tried and testing group of setting would appreciate it

 

[Ace King]

Do you run FFB strength at 100% in the Thrustmaster software Martin? I prefer it at 60-70% and then I think I run 100% gain in the sim and adjust each car to suit.

Denis.....just wanted to add that his article was from a different firmware version thus his "100% overall strength of all forces" suggested in the Thrusmaster Control Panel felt that forces were too strong so i ended up using 60% recommended by Thrustmaster as they claim "60%" is like 100% previously and it does feel much better.

 

[RasmusP]

Denis.....just wanted to add that his article was from a different firmware version thus his "100% overall strength of all forces" suggested in the Thrusmaster Control Panel felt that forces were too strong so i ended up using 60% recommended by Thrustmaster as they claim "60%" is like 100% previously and it does feel better.

If thrustmaster didn't implement some un-linear amplifier to boost some forces it shouldn't make a difference apart from giving more dynamic.
Lowering the game ffb vs lowering the driver setting both lowers the average ffb to your hands. While lowering game ffb will still allow for dynamic spikes to 100%, these spikes will be hard-limited when lowering at the base.

This was a whole new world for me when upgrading from my logitech. That you can actually limit the dynamic for more fun instead of never having enough dynamic anyway

Additionally: I love the clipping meter in raceroom! One of the Sims where I prefer to have some clipping and therefore I lower the ffb on my base and run quite high ffb levels in the game.
When hitting a wall or a massive sausage kerb you'll see a spike waaaaay above the clipping level!
Running 100% on the base and lower levels in the game will send these massive spikes into your hands! Not pleasant at all..
(although the most global ffb setting in raceroom actually sits after the clipping and is therefore doing the same the wheel base ffb setting does. That's different to assetto or rF2 where in car and global settings in the game are before the clipping)

Anyway, you're probably all annoyed from my many posts by now so good night. I'll leave you in peace now

Great discussion btw! No fights started, just a good, hard but fair discussion

 

[Case_]

But the most dynamic range will be with the base at 100 and the multiplier adjusted?
If you pick just 100 steps with 1%-100% output then a ingame multiplier of 0.5 in rf2 will make the 1% be 0.5% and 100% be 50%. Still the same range. But everything that was clipping before will now come through.

Well, it depends. See above my comparison to audio - you don't want to run your amplifier at 100 % all the time and adjust the volume by lowering the input signal. You want the input signal to always be the best possible and adjust the volume on the amplifier. I realize FFB might work slightly differently, but I still think it will give you a better result to treat it the same way - maximize the output (and its resolution) from the software, and then "amplify" it on the wheel according to how you want it. Maybe you will notice a difference, maybe you won't. But I do believe that's how it should be done, and it's kind of a big deal for me to do things the way I believe they should be done (possibly a bigger deal than it should be at times, but that's a different topic). But again, I won't argue about this too much - I am willing to accept this as just one of several approaches to FFB, like I've said.

I only use the ffb logging via damplugin and Motec and with the mclaren 720s gt3 at 0.45 in the game I still reach 100% and therefore get some clipping. Sure, only every few laps when I hit a sausage kerb in a bad angle but it's still there.

I don't think it's reasonable to completely avoid clipping on our wheels. It's perfectly OK for the signal to clip running over some kerbs in a high downforce corner or in a similar situation. The wheels are too weak to be able to deal with that, and if you try to avoid clipping altogether, you end up with a low FFB level that's likely to lack detail that could be there if it was set higher. So I always set my FFB output so I don't get clipping in general driving situations but the signal is as high as possible, and I just ignore the occasional momentary spikes. It's fine to get spikes in extreme situations - as long as they are momentary and not too frequent. I try to follow the example of Reiza in AMS - the FFB in that game is definitely set high enough so it gives you some clipping through the lap, but it's never excessive (IMO, at least).

So for that 720S (and the S397 GT3 in general), I very rarely go below a multiplier of around 0.75, because the default 1.0 is clipping a bit too much for my tastes, but going lower would be counterproductive and I might start losing detail (and strength, on my wheel). But on some tracks, I even raise it slightly from the default 1.0.

 

[Ace King]

Input/output signals is one of my expertise as i ran a electronics repair business for 40+ years. When i get some spare time i would be curious hooking up my dual trace scope and observe ffb wave signals.

 

[Case_]

BTW, that part about avoiding clipping and ending up with a general signal that's too low - that's exactly the main reason I have issues with the MakCorp Group C mod and it's FFB. The bumps and downforce and wheel weight are very strong in that mod, so you're clipping hard all the time in corners, under braking or on bumps, but if you try to tame these, you end up lowering the general "driving" FFB too much. It's not that the FFB is bad in that mod - but the dynamic range it tries to cover is (IMO, again) way too high for a generic consumer wheel, and therefore could use some adjustment to compress the range. It's probably much better on a DD wheel (then again the bumps might break your wrists on one ).

 

[RasmusP]

Well, it depends. See above my comparison to audio - you don't want to run your amplifier at 100 % all the time and adjust the volume by lowering the input signal. You want the input signal to always be the best possible and adjust the volume on the amplifier. I realize FFB might work slightly differently, but I still think it will give you a better result to treat it the same way - maximize the output (and its resolution) from the software, and then "amplify" it on the wheel according to how you want it. Maybe you will notice a difference, maybe you won't. But I do believe that's how it should be done, and it's kind of a big deal for me to do things the way I believe they should be done (possibly a bigger deal than it should be at times, but that's a different topic). But again, I won't argue about this too much - I am willing to accept this as just one of several approaches to FFB, like I've said.



I don't think it's reasonable to completely avoid clipping on our wheels. It's perfectly OK for the signal to clip running over some kerbs in a high downforce corner or in a similar situation. The wheels are too weak to be able to deal with that, and if you try to avoid clipping altogether, you end up with a low FFB level that's likely to lack detail that could be there if it was set higher. So I always set my FFB output so I don't get clipping in general driving situations but the signal is as high as possible, and I just ignore the occasional momentary spikes. It's fine to get spikes in extreme situations - as long as they are momentary and not too frequent. I try to follow the example of Reiza in AMS - the FFB in that game is definitely set high enough so it gives you some clipping through the lap, but it's never excessive (IMO, at least).

So for that 720S (and the S397 GT3 in general), I very rarely go below a multiplier of around 0.75, because the default 1.0 is clipping a bit too much for my tastes, but going lower would be counterproductive and I might start losing detail (and strength, on my wheel). But on some tracks, I even raise it slightly from the default 1.0.

I guess you were busy typing while I flooded the thread. I'm looking forward to yoh quoting the rest of my posts
Btw my "no clipping" setup works very well for rF2 and assetto. Not for raceroom and dirt rally.
I don't think in modern Sims ffb you actually "lose resolution". Imagine you'd have perfect audio setup. Perfect means no noise in the signals etc. Then, apart from hurting your ears when someone pulls the cable roughly or puts the input volume too high, you won't lose details at all. You would however gain dynamic for stuff like reproducing someone playing an unplugged e-guitar and a rock concert in actual, real volume levels.
Is such dynamic always good? Depends!

The wheels are too weak to be able to deal with that, and if you try to avoid clipping altogether, you end up with a low FFB level that's likely to lack detail that could be there if it was set higher

Argh I wanted to stop and go to bed. Too much fun to discuss things with you guys!
My csw is able to make my hands slip on the rim when using gloves. Alcantara rim + suede gloves. Not as sticky like rubber on rubber but still very sticky!
I'd say it exactly on a level before it would hurt me.
I don't like weak ffb so I doubt that I'm "ending up with a too low average ffb level and will miss details".

I can tell you that the ts-pc is not capable of reaches this level of barely not hurting me. 100% was still pretty manageable for me. Still lowered it in most cases though!
Perfect consumer level I'd call that. Csw 2.5 might scare some inexperienced people...

I'm looking forward to you having a play with a csw or DD wheel at some point Martin!
I absolutely agree that your way of doing things is totally fine and making sense. However I think it would be great to increase your horizon a little at some point.
I remember being at a professional recording studio (small, but professional) via my uni and when they recorded a guitar I was like "what the... That signal input is so low it will be crap quality!"
And then they boosted the signal digitally in the DAW (logic) and it was totally fine!
If I'd do that with my guitar recordings at home it would be a noisy mess...

 

[Marc Collins]

I believe the "steering torque capability" setting really does nothing outside of direct drive wheels, so unless you have one, it should be safe to ignore the value completely. As far as I know, that setting should only be relevant in case your wheel is capable of producing more torque than what the nominal steering torque of the car in question is, and I believe pretty much all cars exceed this even for a CSW/CSL wheel.

It functions for all wheels. As mentioned above, if you increase the value, you will eventually end up with no FFB. What interests me is why the profiles supplied by S397 almost all use a value of 2.5--for everything from a Logi G25 to a very powerful Fanatec or Thrustmaster top-of-the-line wheel. Huge variation in wheel torque outputs; same default setting.

Would love to understand how to optimize this setting for a particular wheel, or, if 2.5 somehow works out as the best setting for all...

 

[Case_]

@RasmusP Well, you are in a position of someone with a wheel capable of producing twice (or so) the torque of mine, so that might certainly have something to do with it. If we both set our wheels to feel the same in strength, you will still have a lot of headroom, while I will likely be near the maximum. So you have more leeway in the settings no matter which approach you take.

 

[RasmusP]

It functions for all wheels. As mentioned above, if you increase the value, you will eventually end up with no FFB. What interests me is why the profiles supplied by S397 almost all use a value of 2.5--for everything from a Logi G25 to a very powerful Fanatec or Thrustmaster top-of-the-line wheel. Huge variation in wheel torque outputs; same default setting.

Would love to understand how to optimize this setting for a particular wheel, or, if 2.5 somehow works out as the best setting for all...

Don't spend too much thought about it. My guess would be that s397 simply doesn't care for it. Someone of the team who has a DD wheel might have pushed for this "simple calculation correction" to be implemented but then the guys who do the profiles couldn't care less..
I mean in the end you can just lower the per car multiplier to your liking no matter what and he happy.
This setting only lowers the ffb if you would feel higher forces than the physics engine calculates for. It only takes away from you to manually lower the ffb for cars you are sure about having "too much ffb".
But as I said the question is: how trustworthy are the steering force levels that the physics engine comes up with? More trustworthy than your opinion on them? Who knows...

 

[Marc Collins]

Don't spend too much thought about it. My guess would be that s397 simply doesn't care for it. Someone of the team who has a DD wheel might have pushed for this "simple calculation correction" to be implemented but then the guys who do the profiles couldn't care less..
I mean in the end you can just lower the per car multiplier to your liking no matter what and he happy.
This setting only lowers the ffb if you would feel higher forces than the physics engine calculates for. It only takes away from you to manually lower the ffb for cars you are sure about having "too much ffb".
But as I said the question is: how trustworthy are the steering force levels that the physics engine comes up with? More trustworthy than your opinion on them? Who knows...

So, since you can eliminate the FFB by setting the value too high, we wouldn't want to do that. What is the effect of setting it too low? Seems like it matters and many people may be experiencing un-optimized FFB as a result. At 2.5, the FFB feels perfectly weighted at a multiplier of 1.0 for almost every car. I rarely have to touch the multiplier. Would that be true if I had a Logi? If not, I'd rather make one master adjustment to account for my wheel than fiddle with the multiplier on every individual car.

 

[RasmusP]

Btw last word from me:
I'd love to have ffb plug-ins similar to what is available for sound editing.
How awesome would it be to run a "maximizer" on a logitech wheel, boosting lower forces but keeping higher forces with full details, resulting in a really beefy ffb feeling for road cars while keeping all details!
Or a compressor for dirt rally 2, letting you ramp up the suspension forces etc but compressing the kattle grids!
Currently we only have "clipping", which feels like Joseph Wright sounds when shouting in his microphone while commentating. God awful

 

[RasmusP]

So, since you can eliminate the FFB by setting the value too high, we wouldn't want to do that. What is the effect of setting it too low? Seems like it matters and many people may be experiencing un-optimized FFB as a result. At 2.5, the FFB feels perfectly weighted at a multiplier of 1.0 for almost every car. I rarely have to touch the multiplier. Would that be true if I had a Logi? If not, I'd rather make one master adjustment to account for my wheel than fiddle with the multiplier on every individual car.

To answer the question quickly:
If you set it too low the car will just use the normal 0-100% range without getting multiplied at all! The game would go like "physics engine = 3Nm, wheel capable of 0.1Nm, don't do anything and just use the per car multiplier"

It's not just a multiplier. It only blocks the ffb from going higher than it would be realistic.
Spot on indy car mod for example would go up to 20 Nm on a direct drive. A spot on fiat 500 mod would only go up to 3 Nm on a direct drive. (fantasy numbers!)
Now on a logitech you basically need the full 2.5 Nm for both cases to make it enjoyable at all. So these spot on mods would be done in a way that they both reach "100%" more or less with a multiplier of 1.0.
Resulting in the fist 500 being as heavy as an indy car...
The problem is you can't really use it as a global multiplier as you mentioned though.
If you put in "15" for example the fiat 500 ffb would be lowered by factor 7. The indy car ffb won't be touched at all, since the physics engine comes up with higher forces than these 15 Nm.
So you would only "globally multiply" some cars but some not.
There really is only one useful setting and that is putting in the real number. So with my csw I would have exactly the real ffb from the fiat 500 but the indy car would simply use all force available on my wheel base.
If you put in a custom value to multiply the ffb to your liking you'll end up with a harsh border where some cars are extremely light and some cars extremely strong.
You would still need to raise the too light cars and lower the too strong cars!

With the default of 2.5 they make sure that basically every damn car in rf2 will exceed that level and therefore will just always use the full force you have available (or whatever you set in the wheel driver software).
You just put every car to your liking via the multiplier.

What you end up with is basically the same ffb strength for all cars. They just have a different shaping from the actual physics but not a vastly different peak torque level!

 

[Marc Collins]

It's not just a multiplier. It only blocks the ffb from going higher than it would be realistic.
Spot in indy car mod would go up to 20 Nm on a direct drive. A spot on fiat 500 mod would only go up to 2-3 Nm on a direct drive.
Now on a logitech you basically need the full 2.5 Nm for both cases to make it enjoyable at all. So these spot on mods would be done in a way that they both reach "100%" more or less with a multiplier of 1.0.
Resulting in the fist 500 being as heavy as an indy car...
The problem is you can't really use it as a global multiplier as you mentioned though.
If you put in "15" for example the fiat 500 ffb would be lowered by factor 7. The indy car ffb won't be touched at all, since the physics engine comes up with higher forces than these 15 Nm.
So you would only "globally multiply" some cars but some not.
There really is only one useful setting and that is putting in the real number. So with my csw I would have exactly the real ffb from the fiat 500 but the indy car would simply use all force available on my wheel base.
If you put in a custom value to multiply the ffb to your liking you'll end up with a harsh border where some cars are extremely light and some cars extremely strong.
You would still need to raise the too light cars and lower the too strong cars!

With the default of 2.5 they make sure that basically every damn car in rf2 will exceed that level and therefore will just always use the full force you have available (or whatever you set in the wheel driver software).
You just put every car to your liking via the multiplier.

What you end up with is basically the same ffb strength for all cars. They just have a different shaping from the actual physics but not a vastly different peak torque level!

OK, so what's your best guess for a CSW if you wanted some aspect of differences and to feel an Indycar being at least a bit heavier than a Fiat 500?

 

[RasmusP]

OK, so what's your best guess for a CSW if you wanted some aspect of differences and to feel an Indycar being at least a bit heavier than a Fiat 500?

Put in the real value, which for csw 2.5 afaik would be 8.5.
Or leave it at default (or put it even lower so it never does anything) and simply use the multiplier.

Sorry, I edited the first lines in my post as I recognized I didn't answer your question at all really...

 

[Marc Collins]

Put in the real value, which for csw 2.5 afaik would be 8.5.
Or leave it at default (or put it even lower so it never does anything) and simply use the multiplier.

Sorry, I edited the first lines in my post as I recognized I didn't answer your question at all really...

Is the 2.5 appropriate for a Logi?

 

[RasmusP]

Is the 2.5 appropriate for a Logi?

I think so. I did some measurements myself by putting a stick on the rim, measuring the length/leverage etc, grabbing a kitchen scale, putting the centering spring in the driver to on and maximum and then turning the wheel close to the maximum.
Then I put the scale beneath the stick's end and noted the weight it showed. Then I pushed the wheel further to the end with the scale and noted the constant weight is showed again.

Calculated the base torque from all this and l came up with something like 2.5 Nm for the wheel pushing against something and 3.5 Nm for you turning the wheel when it's at maximum force output.
No guarantee for this calculation though

I'd say with anything other than a DD wheel you can just let it at default and use the multiplier. I just put in the value for the sake of putting it in.
Today I found out that it got overwritten and I didn't notice it when driving the 720s gt3 and then jumping into the btcc mod. No difference...
I'm yet to find a car that has a lower physics engine output than my csw. Question is how trustworthy that car will be..

Honestly I'm thinking I will go back to the default of 2.5

So, since you can eliminate the FFB by setting the value too high, we wouldn't want to do that

You won't eliminate the ffb though. The car specific multiplier goes up to... Good question. I know I once used x2.5 with my g27 to get some feel for the car at all. Clipped like hell over kerbs though

 

[Denis Betty]

I tried the AC setting from the BoxThisLap link, except I left my overall strength setting at 70% and damper at 100% in wheel settings and it's miles better in game. So, thanks. I don't wanna go higher with my wheel strength setting, because I don't see an overall gain setting in rF2. The mak-Corp cars, I'm aready running at The 1 or 2 in the tuning menu - I guess I could reduce the car specific multiplier, but... well I'll have a play with it and see how I feel.
In AC do you guys run FFBclip in AC?

 

[Case_]

BTW, out of curiosity, I just did some comparisons with the steering torque setting set at default 2.5, at 0.5 and at 20. As far as I am concerned (on a T300), I can't feel any difference, nor can I see one on the FFB monitor I use. I tried several different cars from road cars to Indycar. This is obviously not very scientific, but still.

@Denis Betty Regarding the MakCorp cars - you can lower the multiplier, but like I mentioned, it won't really solve much, since the spikes are so high (the dynamic range of the FFB) that as you try to get those under control, you lose the useful FFB for driving the car, so that's kinda counterproductive.

And yes, I use FFBClip, it's quite useful for my approach to FFB (=maximizing per car FFB with some mild clipping).

 

[Marc Collins]

BTW, out of curiosity, I just did some comparisons with the steering torque setting set at default 2.5, at 0.5 and at 20. As far as I am concerned (on a T300), I can't feel any difference, nor can I see one on the FFB monitor I use. I tried several different cars from road cars to Indycar. This is obviously not very scientific, but still.

And yes, I use FFBClip, it's quite useful for my approach to FFB (=maximizing per car FFB with some mild clipping).

Interesting. When I do the same thing on my Fanatec CSW, at 20 there is just the faintest trace of FFB left. At 0.5, it's unnaturally heavy. I had already mentioned this effect earlier, but then re-did my test using the same numbers as you.

How do we explain that?

 

[RasmusP]

I tried the AC setting from the BoxThisLap link, except I left my overall strength setting at 70% and damper at 100% in wheel settings and it's miles better in game. So, thanks. I don't wanna go higher with my wheel strength setting, because I don't see an overall gain setting in rF2. The mak-Corp cars, I'm aready running at The 1 or 2 in the tuning menu - I guess I could reduce the car specific multiplier, but... well I'll have a play with it and see how I feel.
In AC do you guys run FFBclip in AC?

If the slight gyro change is "miles better", what settings did you use before? Is that the only thing you changed or did you change more?

Interesting. When I do the same thing on my Fanatec CSW, at 20 there is just the faintest trace of FFB left. At 0.5, it's unnaturally heavy. I had already mentioned this effect earlier, but then re-did my test using the same numbers as you.

How do we explain that?

Interesting... 0.5 shouldn't do anything while 20 makes sense but somehow doesn't become active for Martin...
Gonna do some testing when I'll come home!