REQUEST: FFB thread

John.Persson

John.Persson

Registered
I just removed my old install and moved to steam.

And by doing that I lost all my configurations. Haven't been playing the last couple of builds, but I really need help getting a good FFB feeling in my old G25 again. Everything feels flat and dead now.

Didn't we have a thread about a plugin to find the clipping values in FFB and so on?

I really want to feel my muscles fighting the wheel again, so can anyone give me a brief introduction on how to find good values for FFB in the game settings.
 
Would you care to share your findings here, after you have something that works for you? I've been looking for a usable configuration for my G25 for a long time now, and I'm always open for suggestions.
 
Gevatter said:
Would you care to share your findings here, after you have something that works for you? I've been looking for a usable configuration for my G25 for a long time now, and I'm always open for suggestions.
Click to expand...

To just keep it short. Download the plugin for pedal overlay, the yellow one should never go red when it reaches 100%.

And to avoid that from happening, adjust the setting for car specific FFB multi. And try to keep the FFB smoothing as low as possible. But our g25 gets the rattling at zero usually. So play around with those values. And keep logitech profiler value arund 98 or 104 and see hhow it feels.
 
Just did some quick laps at nordshleife with nissan gtr.

Put:
FFB Smooting 0
Car specific multi FFB 1.08
Minimum FFB 8.5

And feel how that feels. I got some clipping occuring, but that was mostly on the kerbs and only when turning in to em.


I want to get more force from the wheel though, so going to play around a bit more. But not tonight.
 
The only way you will be getting a "forceful" feeling from a G25 is by setting the forces clipping, in my experience. I have compared rigs with T300/T500 and G25/G27 and the latter simply feel like toy wheels after driving with the Thrustmasters, which basically produce double the torque. There is no way to produce muscle fighting torque by editing the software when the wheel itself is too weak.
 
wpthayer said:
Here you go..Should take about a month to get through it.
http://isiforums.net/f/showthread.p...-for-rFactor-2-The-key-to-being-in-the-Zone-D
Click to expand...
I would like to ask if what it is recommended in this thread is also supported by ISI.

I simply do not conceive how there isn't an official FFB section where ISI explains to the newcomers how should they setup their wheel.


Even the most basic thing as it is the profiler is unexplained and it greatly impacts the feeling.

So please Christopher if you can put some light in this area it would be highly appreciated.

Enviado desde mi ONE A2001 mediante Tapatalk
 
Euskotracks said:
I would like to ask if what it is recommended in this thread is also supported by ISI.

I simply do not conceive how there isn't an official FFB section where ISI explains to the newcomers how should they setup their wheel.


Even the most basic thing as it is the profiler is unexplained and it greatly impacts the feeling.

So please Christopher if you can put some light in this area it would be highly appreciated.

Enviado desde mi ONE A2001 mediante Tapatalk
Click to expand...

We support the idea that advanced users may choose to use an in-depth guide like this one. It's great that fellow community members take time to share knowledge and help customize and improve the experience for those that want to dig a bit deeper.

That said we do try and make sure FFB just works out of the box, that's our goal anyway when testing. We have pretty much all current wheel brands covered by our test team, but some things like FFB are partly based on personal taste. Even "profiler" settings, testers user different values, so it's hard to say what's ideal for everyone.
 
Christopher Elliott said:
We support the idea that advanced users may choose to use an in-depth guide like this one. It's great that fellow community members take time to share knowledge and help customize and improve the experience for those that want to dig a bit deeper.

That said we do try and make sure FFB just works out of the box, that's our goal anyway when testing. We have pretty much all current wheel brands covered by our test team, but some things like FFB are partly based on personal taste. Even "profiler" settings, testers user different values, so it's hard to say what's ideal for everyone.
Click to expand...
Also dd wheel in testmembers?
 
Christopher Elliott said:
We support the idea that advanced users may choose to use an in-depth guide like this one. It's great that fellow community members take time to share knowledge and help customize and improve the experience for those that want to dig a bit deeper.

That said we do try and make sure FFB just works out of the box, that's our goal anyway when testing. We have pretty much all current wheel brands covered by our test team, but some things like FFB are partly based on personal taste. Even "profiler" settings, testers user different values, so it's hard to say what's ideal for everyone.
Click to expand...

Thanks for the reply Christopher.

However, I disagree in some points.
FFB isn't something subjective based on personal taste. If you ride a real car you get what you get in terms of FFB. Considering that rF2 is stimulating the forces in the steering column, the wheel should be configured to translates those as accurately as possible to the driver arms.

It makes no sense to me that you define specific controller.json settings for several wheel models but you are not providing the recommended profiler settings. What is one without the other? Saying that different testers use different values for the same wheel is actually more worrying than relieving. I have to understand that the settings are very different between each other and that is the reason not to provide a recommended setting.

This calibrations should be done based on torque measurements and not based on opinions.

However, for some reasons I don't get to understand, it seems that many prefer that FFB remains as something subjective.

At least I tried.

Enviado desde mi ONE A2001 mediante Tapatalk
 
Exactly. A preset without a profiler is incomplete. In fact, when I asked if ISI is supporting what it is stated in the FFB thread I should have remarked especifically about how the profiler should be set which is defined and justified in the first post of the thread.
There it says everything should be turned off expect the first setting which is set to 100%.

For the controller.json it is clear ISI is already providing a recommendation which is the preset the game provides after detecting the wheel.

Enviado desde mi ONE A2001 mediante Tapatalk
 
I don't get what the fuss is about. There are 2 or 3 values in rF2 FFB that needs configuring, rF1 had about 20-30. From those key FFB values in rF2, one is already pre-tuned by ISI (car-specific multiplier). ISI cars are tuned to deliver optimum compromise at 1.0 FFB multiplier, i.e. enough forces for the wheel to feel heavy but not reach clipping in normal cases. Then you have the "Steering Torque Minimum" value which is wheel-specific and has to be set by the user. Even when the user only changes rim of his wheel this value needs to be re-tuned (a bigger rim usually has more rotational inertia), so it's a bit difficult for ISI to come up with a profile for this. Then there is filtering, which should not be used at all unless necessary to avoid rattling sound. Again this is subjective and depends if you have neighbors to the wall next to you or not :).

The rest of the values more or less need no editing, rF2 already delivers the raw forces to the wheel, so editing anything else will probably compromise the original intended forces.
 
Paul Loatman said:
I think the issue is that most wheels simply can't produce the forces accurately regardless of the settings, so what settings feel closer to real can be very subjective.

If you had the perfect FFB wheel that could produce the forces exactly like the real car, then i'm sure it would work out of the box with enough effort from ISI, but pretty much all FFB wheels have some problems reproducing the forces accurately, even if very small, it's enough to make some people have different points of view regarding what is considered to be 'better' feedback.



In my opinion, it's not as simple as getting the torque outputs correct, torque doesn't define position or speed, at least not in the way you would want for steering feedback with current hardware. There's a lot of odd things going on in a real car's steering feedback which is surprisingly very difficult to reproduce with current hardware.
Click to expand...

Sorry to say Paul but what you are stating makes absolutely no sense.

It is clear that many mid end wheels may have problems reproducing several dynamic FFB patterns. For what it is being suggested, the error must so big, that it is not possible to define a default profiler to go with the predefined controller.json.

I wonder then, why ISI provides a predefined preset for those wheels. I understand that all the subjectivity resides in the profiler but is objective for the controller.json. This makes absolutely no sense.

Furthermore, with high end DD wheels which have a much superior dynamic response, the controversy about how the wheels needs to be setup does not end. It is still subjective to the different users.

I do agree FFB is a complex item. Too complex IMO to be addressed in such a subjective way as it seems it is being done. I must seriously think after all this, that the FFB (steering column torque) cannot be correctly calculated by the physics engine since it seems impossible to validate it objectively.

Regarding your comment on position and speed. I think that old paper regarding position FB made too much hurt. I already responded to it when it was arised several years ago to show that what was said there, showed a big misunderstanding about how FFB should be calculated. It basically said that the user provided a force and the system responded to it which is conceptually incorrect. Any system reacts to displacement and not viceversa.

Enviado desde mi ONE A2001 mediante Tapatalk
 
Paul Loatman said:
ISI uses essentially the same basic settings for every wheel in terms of FFB, and every wheel will feel slightly different with those settings because of the difference in hardware. If you really look at the JSON settings of different wheels, the differences are usually button map related. The FFB settings are generally the same among different wheels, other than some very obviously necessary things like smoothing and effects strength + or -. The settings that drastically change the feedback are kept the same between all wheels.

So, i'm not sure what doesn't make sense to you. The fact that ISI uses similar FFB settings for every wheel, or the fact that every wheel will feel different regardless of the settings?

In terms of speed and position, i never mentioned Bodnar's paper, i wasn't even talking about that. I was pointing out the obvious fact that a certain amount of motor torque does not equal a certain speed or position, since that would depend largely on the exact motor being used. Therefore if you define everything in terms of torque without taking into consideration those other forces, then you'll end up with vastly different feedback among different wheels, even if the sim is giving it the same levels of torque.
Click to expand...
So if all the rF2 preset configurations, which have lots of parameters affecting FFB, are pretty much identical between different wheels with very different performances, then no real customization is being done by ISI.

It is clear for me that this task requires a knowledge that seems to be beyond ISIs capabilities for what it can be observed.

I will write down a procedure of how this should be done from a theoretical point of view. It is actually not so difficult if you know what you are doing. It requires some controlled testing for sure but I think it can be quite easily done with a minor investment.

I seem to be not fully understanding your last statement. In any case I think you are missing the simple fact that the game should calculate the torque at the end of the steering column and provide that signal to the driver of the motor so that it provides that torque. Full stop.

The wheel rotation and inertias do not need to be simulated since they are real. They become boundary conditions that need to be monitored in terms of wheel rotation and its derivatives. The easily achievable point is as if the driver had his own sim wheel installed in the real car. To account for the sensation with the real wheel, its differential friction inertia and stiffness should be implemented into the profiler (considering that this worked correctly which should also be verified)

Enviado desde mi ONE A2001 mediante Tapatalk
 
Euskotracks said:
So if all the rF2 preset configurations, which have lots of parameters affecting FFB, are pretty much identical between different wheels with very different performances, then no real customization is being done by ISI.

It is clear for me that this task requires a knowledge that seems to be beyond ISIs capabilities for what it can be observed.
Click to expand...

I still don't understand what is the "customization" you are after? Which parameters? As I said, there are only 3 base parameters to adjust. I personally think you should only adjust the obvious parameters from the sim and leave the rest to be adjusted from your wheel control panel, otherwise you will not achieve the forces that ISI intended. The other parameters in Controller.JSON are mostly leftover stuff from rF1.

ISI provides custom profiles because people want plug-and-play functionality without remapping XX buttons every time they switch controllers. Only a minority is interested in fine-tuning the FFB, which is already superb and state of art if you compare to any other sim.
 
stonec said:
I still don't understand what is the "customization" you are after? Which parameters? As I said, there are only 3 base parameters to adjust. I personally think you should only adjust the obvious parameters from the sim and leave the rest to be adjusted from your wheel control panel, otherwise you will not achieve the forces that ISI intended. The other parameters in Controller.JSON are mostly leftover stuff from rF1.

ISI provides custom profiles because people want plug-and-play functionality without remapping XX buttons every time they switch controllers. Only a minority is interested in fine-tuning the FFB, which is already superb and state of art if you compare to any other sim.
Click to expand...
I would like ISI to define the recommended profiler settings for each of the wheels supported by the game. A picture of the profiler should be enough for G27 but maybe other wheels require more pictures.

Enviado desde mi ONE A2001 mediante Tapatalk
 
You must log in or register to reply here.
Back
Top