Automobilista 2 Force Feedback Overview & Recommendations

Thanks Daniel, by the way, I sometimes use your V7000 file. Is the LFB slider in game for that file still a LFB slider? I feel that my wheel shakes too much from engine when at low speeds pulling out of the pits. Should I lower something in your file? or Lower the LFB or FX?

 

Please send me a private message with your in-game settings and your base

 

In my oppinion, @Matus Celko and @ChasteWand explained very well how FFB works and why I always suggested (in my FullFFB "documentation" ) using wheelbase driver to get the max force you're confortable to use and then using GAIN and vehicle-specific gain to control clipping on telemetry HUD.

My FullFFB deals with values that are more than 3x the output "clipping limit", although without any clipping at all...
In the end, I do multiply the internal FullFFB calculations by 0.43 (43%), which in turn is multiplied by GAIN (65, meaning 65%). That make the output only 27,95% of my processed signal.

Code:

(output (* output 0.43 gain))

There is no way to bend physics, no way to surpass a physics clipping. But we can make systems that set a "max" to be half of what is physically possible - that system would accept the "max" to be surpassed (the "max", the "100%", the "0dB" are just relative measures on many processing stages) and still get the signal corrected in the the end avoiding any clipping at all (in the master track volume/gain on the audio analogy).

I hope I had contribute to Salvador Dahli's great work (@ChasteWand ...and no, the Simucube driver - or any driver, for that matter - cannot detect game clipping).

 

No one seems to mention, or think about, "Dynamic Range" - well I will call it that.

Just say you log a race lap, the FFB graph (How? Well, say you could). Around the lap you will have various 'highest' FFB points for each of those areas. Let's say 8 of them (corners?) are '80', and then just a couple of areas ever hit '100' and those could very well be curb hits or something more dramatic than when you are on the 'normal', optimal, racing path. Or they could even just be a 'dip' - some valid place you race on/over etc.

So you saw '100's' and if you set the IN GAME FFB Strength to make sure those were the maximums ever 'seen', and not clipped, that means you have limited the Dynamic Range that you COULD have had for the 0 to 80 of all the more common 'normal' driving.
If you, instead, chose to LET it have clipping for anything over 80, that means you will get the maximum dynamic range for all the 'normal' driving range (the 0-80 range/values)... and WHO CARES about the clipping and 'inaccuracy' of those other cases. At those places the detail 'lost' is likely to be irrelevant anyway!

But to do this properly, correctly, you would need to check VALID LAPS and then work out which areas were the really valid highs, versus the highs that were 'excessive cases'. Then you can say that the '80'... or '70'.. or '90'... were the only truly valid highs that matter. Adjust the in game FFB strength so that THOSE become the '100' - thus you have the maximum Dynamic Range of the signals that MATTER.

You could do this via the in-game FFB graph - you would need to keep LOOKING at it as you drive (or maybe can do it in a Replay?). Or rather, record it to watch back later.

An example of this would be Spa and Eau Rouge, which is the most FFB 'flooding' place of the lap. Running the FFB Strength higher, and thus letting that whole zone clip 'heaps' is no issue as the information/detail lost is irrelevant there anyway. If you were blindfolded, or not told it was missing, you would never even know!
But then you get the other 99% of the lap in HIGHER detail..... better Dynamic Range, to differentiate the span of all FFB factors.

-----
You always set the GAME FFB Strength to whatever clipping amount you want to allow it.
THEN you set the WHEEL to whatever final maximum strength you would like.

If you have a 'weak' wheel, and want higher strength, then you could choose to run even more clipping (Game FFB strength higher) so that you get that greater wheel strength - at the cost of a loss of detail for SOME situations (the clipped ones).
The reason you get "more strength" is that the most felt 'force' is the normal turning force, which is well under what curb hits and high stress situations give - so you are raising the FFB Strength so that those 'steering' forces are higher - they are now high level things not lower level - and let the higher level ones lose detail due to the fact they are now clipping off much more and more often. That is all you can do with a weaker wheel... OR, settle for the weaker steering feel.

 

I would like to know this too, does it still amplify lower forces for belt driven wheels like in default/default+ or does it have a different function?

 

Everything is exactly the same as with Reiza's defaults. LFB strengthens the center and can increase subtle forces in the lower range of gear or belt bases. With the V7000 files, however, this is only necessary for gear bases. But even here, only about 20% is perfectly sufficient. Using too much LFB will lead to anomalies in the steering axis

 

I have a Fanatec DD2 and have tested the peak vs. linear setting back and forth. I feel the weight transfer with the linear setting better. It removes the feeling that the wheel gets lighter mid-to-end corner. This is with the Default FFB setting.

 

'All possible settings should be linear, neutral, default' - Marc collins, internal tester for Reiza.

 

There are lots of config recommendations still use peak including the one in the Fanatec website

 

Peak is the stronger setting, Linear is the softer setting. Linear Cuts every Peak which could go beyond the Holding Torque capability of the base.

Fanatec probably recommended peak because it makes the base feel stronger, it's like when tv makers ship their products set with exaggerated picture modes to give the illusion that their tv is the brightest most colorful tv whilst actually destroying the picture we were intended to see.

DD wheel software should come with a 'Gamemaker mode', like how good tv's have a 'Filmmaker mode.'

 

Yes, that's right.Fanatec simply cuts off the peak and thus achieves fewer Nm in Linear mode in order to only achieve a constant holding torque.The logical conclusion from this is that with Linear you can feel more subtle details than with Peak because higher forces can mask this.I think with a dd1 it's around 15 Nm instead of 20 Nm.There are many opinions about what is better and what isn't.For me, after many comparisons to other bases, the problem with Fanatec is the motor in these old dd1/dd2 bases and the poor internal FFB algorithms.This means that the in-game signals of a simulation are not processed and output well.Nevertheless, I personally still favor "Peak" so that the full highest torque can be used in certain simulation situations.But to use this peak effectively, certain settings have to be adjusted.In concrete terms, this means that "FEI" must be significantly lowered to feel the more subtle FFB details of the game again. Otherwise, everything will be masked too much, or there will be extreme vibrations or high-frequency noises, or a very coarse and grainy. "FEI" and "INT" are key for me to achieve peak performance with Fanatec. If these 2 values are kept high then I would also set to Linear .
I'm also someone who tends to set everything to linear and neutral, but it can't be applied across the board because there will always be special cases. If these 2 values are kept high then I would also set to Linear. If these 2 values are kept low, I would bet on Peak

 

Guys this can sound weird… But have you tried driving some laps with the chase cam? I agree with @Joaquim Pereira that there is something weird with the visuals… And I notice some differences between external cameras and cockpit camera… External ones feel more aligned with what I feel on the wheel with Default FFB setting. I use the same camera settings as Joaquim BTW.

And I also agree that iRacing cameras are impressive and are truly aligned with what is happening with the car.

 

Definitely disagree. The car's settings are only a small part of the improvement. The other, bigger problem lies in physics. We'll probably never know, but I'm fine with that

 

BTW, Damping at 50 is quite a lot, have you tried lower settings?

 

I will check that. I don't use Default+ so not much effort making it work right. Before I start doing FullFFB, I used Default+ and I have it much lower (10-20?), if I remember correctly!

 

I've recently looked into this "boat feel" issue more intensively.And I've made a clear trend and observed changes.Just as I recommend lowering the brake and throttle sensitivity on my customs, I've noticed the same thing with steering sensitivity.Reducing it brings clear advantages.I've already tested several cars, and I've had better results with lower steering sensitivity on all of them.I'll continue testing and checking, as I've only tested it with the VNM Base so far.But if that's really the case, I'll change my recommendations for steering sensitivity as well.Please no comments about linear or non-linear; for me, in the end, only the feeling counts

 

I will try this by myself, but I can imagine lowering the steering sensitivity has a mild damping effect?

 

Yes, it's best to test it with extreme values. That means with 50% standard steering sensitivity and then with 1% steering sensitivity. This way, you'll feel the difference better. Of course, you should visually hide the steering wheel before doing this test

 

After further testing, I was able to confirm this strange behavior of the steering axis (boat feel) and stability on all DD bases I tested. The interaction between haptic and visual feedback is not optimal in my opinion. This means that there is a delay between the steering axis inputs and the visual image. This delay leads to incorrect rotation of the vehicle or, more precisely, to an incorrect rotation feel. Your steering inputs do not match the visual image on your monitor.

Furthermore, reducing the steering sensitivity brings much more stability and directness to your vehicle.More stability means more control and more feedback, and more connection between the car and the track. I'm currently at 25% steering sensitivity, which makes it much better.This improved directness of the steering axis is a huge advantage. It is completely irrelevant whether you are testing with the standard, basic, or custom file.

still need to test whether it's less pronounced with the belt or gear bases, or whether it will have the same impact. That would also better explain people's different opinions and feelings. By the way, you can also test it without FFB, i.e., set all in-game values for Force Feedback to 0% as indicated below. This will give you even better feedback for the steering axis, because no other details (FFB) can be felt.Test it with 50% steering sensitivity, then with 1% steering sensitivity, and then with 25% steering sensitivity. For me, it makes a big difference in the cars' handling, and the less pronounced boat feel.

Gain: 0%
LFB: 0%
FX: 0%
Damping: 0%

 

Why is a similar effect not achieved by a change in the steering lock angle?