Guide: Optimal FFB settings for rFactor 2 - The key to being in the "Zone" :D

Technically they are the same thing however software clipping is the cause and hardware clipping the resultant output.

Your hardware will clip when the software is causing the ffb signal to clip. This can occur from too high a "car-specific ffb multiplier" in rf2 and/or too high an "overall effects strength" in your specific ffb wheel profiler through the windows game controller.

Software clipping from rf2 is easily identifiable. All you need is TechAde's plugin (link to it in the first post) which will allow you to see in real-time if the ffb signal is trying to go above 100% (which isn't possible and therefore will cause software induced ffb clipping). To tell if your profiler is causing any possible undesirable clipping, you need to use the wheelcheck.exe software (also linked to in the first post with instructions on how to use it) and run the step_log 2 test. And those are the only 2 tests you need to avoid ffb clipping.

There is no need to think or fiddle (if you even can) with the "nominal max steering torque" per car. The "nominal max steering torque" for each car is the maximum virtual steering wheel torque value that rf2 will use to correspond to the maximum (i.e. 100%) ffb signal.

 

It's the G25 letting you down. Upgrade the wheel.

Of course other games feel fine on the G25. They don't provide the dynamic range that rF2 does.

 

I don't think this is right. STC only takes effect if it's greater than the car specific "nominal max steering torque" value.

So if for example you have a g25 which has a maximum real world torque output of say 2.5Nm and you set the STC to 2.5Nm to reflect that, if the car specific "nominal max steering torque" was 10Nm, then rf2 would ignore the STC = 2.5Nm value. The game would then scale 0-10Nm of virtual steering wheel torque to the 0-2.5Nm torque output range available from your g25.

 

thank you

for the first just to know, ordinary curiosity.
second, (hope I understand things well) I can imagine scenario with powerful hardware and nominal max steering torque in mod set low to cover usual situations (e.g.conservative/normal setup, tracks without high Gs) and better exploit the potential of weak hardware but unable to cover situations at the limit (e.g. very high caster, non oval vehicle on oval...). Of course, also in this scenario, changing nominal max steering torque makes sense only for powerful hardware because increasing NMST causes low forces from physical engine even lower in FFB signal (lower % of the whole FFB signal range).

I caught some posts with special accent that techade plugin shows only software clipping so probably this made me confused. Do you know why this phrase is used when with current FFB technology (signal is not raw torque value but only %) is impossible to experience pure hardware clipping?

 

For Axe:

It's called software ffb clipping because it's software induced. Hardware induced ffb clipping does not technically exist because any hardware clipping is only due to a direct result of software clipping. After all, hardware cannot do anything without software. Therefore it technically all comes down to software problems.

I think i'm the cause of the confusion because I used "software clipping" to differentiate sources of ffb clipping from within rf2 compared to sources of ffb clipping caused by the ffb wheel profiler which i (think i) referred to as "hardware clipping". When in actuality they are both sources of software (induced) clipping that ultimately manifest themselves on the hardware.

 

Well, if the car your driving has a "nominal max steering torque" less than your 13Nm STC value (for your AF Pro) then it would make sense for your AF Pro to clip above 13Nm of virtual steering wheel torque.

My comment was in relation to where you told Ozzy that "even if you can't use the full strength, if you have the STC set accurately and use a FFB multi of 1, then the torque will be accurate up to the max torque of your wheel, and then it'll obviously clip beyond that." Sounded to me like you were saying that setting the STC to the correct value of his g25 would make his wheel produce 1:1 torque outputs to the virtual torques up to the limit of his g25 (which is ~2.5Nm). If so, it shouldn't because STC is ignored if it's lower than the car-specific "nominal max steering torque" value and i'm pretty sure there isn't a car like that has a NMST value less than ~7.5Nm.

Sorry, that was not meant for Axe.

 

Just want to say I can drive most cars with the default or slightly lower FFB multiplier the Palatov is the only car where I have to go down to 0.5-0.6 multi.

 

I have a t300 gte.

anymore excuses?

should rename the thread

Guide: Optimal FFB settings for rFactor 2 - Lets blame everything but rf2

 

My assumption:

The enginge calculates the force acting on the steering column. The value will be scaled to ether the nominal steering torque of the car or the steering torque capability of your wheel, whichever value is higher.
Example: Physics engine calculates 8Nm of torque, the nominal max. steering torque is 16Nm, the steering torque capability is 2.5Nm. rFactor will output 50% FFB. In case you have a wheel with 32Nm torque rFactor will output 25% FFB.

My assumption would lead to the conclusion, that the Palatov's nominal steering torque is set too low (it is set to 11 Nm with "New rFactor 2 System")?

EDIT: Or the suspension geometry / steering system settings are messed up and putting out to much force (provided that rFactor calculates everything correct)

 

Correct. I'm pretty sure it wouldn't even matter for a CSWv2 either because the lower NMST ("nominal max steering torque") value of any car is ~7.5Nm (if not higher) and the correct STC value for a CSWv2 is ~7Nm.

The ffb multiplier does not factor into the ffb equation until after the STC value is compared with the NMST value of the specific car.

I think the best way to explain how it fits into the whole process is to just explain the entire ffb calculation process.

First, rf2 checks to see if your STC value is greater than the car-specific NMST value. If so, it will use your STC value instead of the car-specific NMST value. This value represents the maximum virtual steering wheel torque value that is to correspond to the maximum (i.e. 100%) ffb signal that can be sent to your ffb wheel (i.e. demanding maximum torque output from your ffb motor). So if your STC value correctly matches what your ffb wheel is capable of and the STC is greater than the NMST value, then your ffb wheel will produce 1:1 torque outputs to those occurring on the virtual steering wheel in rf2 (up to the maximum torque possible with your ffb motor ofc).

However this 1:1 torque matching is only true if the car-specific ffb multiplier is also set to the default value of 1.0. Changing it from 1.0 does not affect which value (STC or NMST) rf2 will choose. The FFB multiplier takes affect afterwards like so...

If there is a virtual steering wheel torque of (say) 5Nm occurring at some moment in time in rf2…

- An FFB multiplier of 1.0 will convert 5Nm into a 5Nm reading

- An FFB multiplier of 2.0 will convert 5Nm into a 10Nm reading

- An FFB multiplier of 0.5 will convert 5Nm into a 2.5Nm reading.

The ffb multiplier quite literally multiplies itself to the virtual steering wheel torque measurements before being fed into the ffb engine. But again, it does so without affecting the chosen STC/NMST value. So for example lets say your ffb wheel can go up to 10Nm and that rf2 used your STC or NMST value of 10Nm. With the default FFB multi of 1.0, a maximum 10Nm of physical torque output from your ffb wheel will occur when the virtual steering wheel torque is also exactly 10Nm. But if you raise the FFB multi to 2.0, the maximum 10Nm of physical torque output from the ffb wheel will now occur when the virtual steering wheel torque reaches only 5Nm. Conversely, lower the FFB multi to 0.5 and the maximum 10Nm of physical torque output from your ffb wheel will occur when the virtual steering wheel torque reaches 20Nm.

So the FFB multiplier is simply a means of increasing/decreasing the range of virtual steering wheel torques that corresponds to 0-100% torque output from your ffb wheel.

 

Sweet.

I'm a bit confused. Is/Has Ozzy said he can get 1:1 torque with his g25 or am i misunderstanding something?

Either way, there is no way to get 1:1 torque with any car in rf2 on a g25 with the default ffb multi of 1.0. However, you could make it 1:1 up the limit of the g25 by increasing the ffb multiplier. For example, if the g25 is physically limited to 2.5Nm and the car-specific NMST was 10Nm, then increasing the FFB multiplier to 4.0 would give you 1:1 torque matching with the g25. But the clipping threshold would now be lowered to 2.5Nm of virtual steering wheel torque which would result in a giant clip-fest.

 

Correct.

I don't know but maybe someone does. I would check for you if i had my pc.

 

thank you
it seems all clear for me. just to be sure:
clipping does not depend in any manner on hardware capability but on ratio of virtual steering forces an NMST (if we exclude possibility of profiler or FFB multiplier settings). Is this statement correct?

and still this question is not answered, hope somebody knows:
3. where exactly is "nominal max steering torque", how to find out exact value for each car, in which file it is and what is the exact name of the parameter?

 

If by "scaled to within the range of the wheel" you mean giving it 1:1 torque output, then no. FFB multi of 1 will not give you that unless the chosen STC/NMST value is also 2.5Nm for a g25.

You can still get ffb clipping if the chosen STC/NMST value is below the maximum virtual steering wheel torque you experience on track under normal racing/driving condition (i.e. ignoring curbs and crashes which always cause clipping). Even if the NMST was set to (say) 20Nm by the mod developer, it could be that in practice you experience up to 25Nm of maximum virtual steering wheel torque during the highest speed corners. In which case you will definitely experience clipping in that corner with the default FFB of 1. You would need to lower the FFB multiplier to 0.8 (calculated from 20Nm/25Nm) to avoid the clipping in this hypothetical example. Or you could mess with your STC and set it to 25Nm so that you can leave the FFB multi at 1 but that's both messy and unnecessary. Simply adjusting the in game FFB multiplier as and if needed is best.

 

No problem.

Correct.

Not that it's super helpful and I could but wrong but i think it's located within a text file in each cars rfmod package/file.

 

@DrR1pper:

May I suggest? You may consider write up a white paper on subject of relations among: STC, FFB mulltiplier, STS, NMST, and HW wheel max torque, etc. Give couple examples too to help. You may post as new thread.

Just a thought.

 

Sorry, Dan, misread between your comments and Ozzy's. If you filled out your online profile about hardware, then I'd have less confusion.

 

That's correct.

I can't see a contradiction. "An FFB multiplier of 1.0 will convert 5Nm (of virtual steering wheel torque) into a 5Nm reading" that is then used to compute the final ffb output. It does this by working out what percentage of the STC/NMST value = to the virtual steering wheel torque reading. That percentage is then converted into an ffb signal that tells your ffb wheel to output a corresponding percentage of maximum torque.

 

Ah ok, yeah, that's not what i meant. I can see why you were confused now. By "reading" i meant the adjusted virtual steering wheel reading (from the FFb multiplier) that then goes into the ffb engine to compute the final ffb output.

 

Remember! The signal the game sends to your wheel (telling your wheel how much power/force/torque to output) is simply a 0-100% signal, not Nm of torque or anything else.


Hardware Clipping
Hardware clipping is if your wheel outputs 100% of it's power while only being asked to output, lets say, 92%. This can potentially happen (but not necessarily) if you raise the overall power of your wheel too high in it's control panel settings. It depends on the wheel (motor), firmware, etc.


Software Clipping
Software clipping is much more common and easier to detect. The game sends a percent to your wheel and can't go over 100%. Let's say that the game-output-signal of 100% equals 20 Nm of torque (20 Nm in the game-car's virtual world). If the game-car ever comes across a corner, or accident (or whatever) that is more than 20 Nm, you will get clipping because the game's 100% FFB point was set to a virtual 20 Nm therefore you can't feel anything more than the virtual car's 20 Nm because the game can't send a signal of over 100% to your wheel.


Nominal Max Steering Torque (NMST)
This number determines what torque in the virtual car equals 100% FFB output signal when running a FFB Multi of 1.00. Eg. If a car has it's NMST set to 18, then 18 Nm of the virtual car torque will be when your real-life FFB is told to output 100% power. The NMST of each car is individually set by the car-creator/modder. We can essentially adjust this ourselves by raising/lowering the in-game FFB Multi or by adjusting the Steering Torque Capability (STC).

Note: If the car clips too easily with a FFB Multi of 1.00, then the NMST of that car was set too low and you need to lower your FFB Multi. For eg. The ISI 60s Eve F3 has to be run around a FFB Multi of 0.72 (or maybe a touch more). That means the car creator/modder set the car's NMST too low.


Steering Torque Capability (STC)
This allows us to globally (not per car) change - well, more like override - the car's NMST and set our own. For eg. If we set STC to 25 (and FFB Multi at 1.00), the game will tell your wheel to output 100% force when the virtual car hits 25 Nm.

Note: Any cars that happen to have a NMST value higher than your STC will ignore your STC setting. In other words, your STC setting will only affect cars that have a NMST lower than your STC. On the other hand, there was at-least 1 person on this forum who said he noticed a difference when changing his STC even though the new value was most likely below any cars's NMST. Maybe it was placebo, maybe a bug/glitch, but ISI themselves said the STC only comes into affect if it is higher than the particular car's NMST.