Automobilista 2 V1.4 Force Feedback Overview & Recommendations

Whats your base settings, and in-game settings ?
Did you use custom file ?

 

is it possible to have the .txt of the DEFAULT+ profile?
Thanks

 

Question:
Do you use 100% FFB force nm in base and have lower ffb strenght in AMS2 ?
(max possible inbase)
or
Do you use lower FFB in base and have AMS2 ffb strenght close to clipping?
(max possible ingame)

(I couldn't find how to create a poll)

 

I use moderate settings on both. I've tested 100% on wheelbase but I really think any perceived benefit is placebo.

 

I think it depends on 2 factors. You can safely set a base up to 12Nm to 100% and use the full power of the base. For a base with 15Nm or even higher, I would reduce the force but adjust the gain in the game to the optimal "range" (60-65% Customs)(approz. 50% Defaults) to get as much detail as possible but as little clipping as possible. Ams2 is different than most other simulations where they are better with 100% gain on the base and then adjust the gain in the game. Yes, you also have the same powers in Ams2 if you take this:

100% Gain (Base)
30% Gain (Game)

or you take:

100% Gain (Game)
30% Gain (Base)

in both cases you will get 130% gain and there will be no difference in pure power (or no difference you will feel)
But the big difference will be felt and seen in the clipping or loss of details (in the FFB graph)
Do this test and you will see and feel what the difference is.
And one last point is the risk of injury when driving such high Nm on the base

Addendum: This test is recommended with (100% vehicle-specific FFB). It is also irrelevant whether you test it with Reiza defaults or a custom FFB, the result will remain the same

 

I used to have like in most of my other sims and mainly because I saw jardier do this in ACC and I just got used to doing it always like this which was.

Not really 100% ingame but as high as possible without clipping
(if i use 100 force it clips in telemetry there will be lots of red line, so I will use max gain possible until i see a sliver of clipping on extreme conditions like 64 in my settings in game)

then In fanalab I would use the gain while driving and find the max gain I can handle while still being confortable so i try to feel all the details .


then recently i saw a jaaames video where he used his simcube base at 100% all the time and only adjusted smaller levels of gain in the sim.

i tried it, i also asked in AMS2 facebook post, and it turned out to be actually what most people do, and it seems to work better, which I'm quite surprised literally!

It's all trial and error, but this kind of gain adjustment never crossed my mind and it's like i'm rediscovering FFB all over again and ams2 feels even better.
it's not "that different" nothing too crazy but feels different for sure.

thanks for the replies anyway

 

Imho the correct way is:

• Wheel Software - As high, as you think is the upper limit for the force the wheel should EVER produce. For 15nm you can keep it at 100%, or lower it to 85% (this will limit it to 12.75nm).
• Game - As high as you feel comfortable. Generally avoiding clipping (shouldn't be much issue with a 15nm wheel).

Why this way? Becasue of clipping. There is a lot of misconception as to what clipping is and why it happens. At a core, it is not a wheel strength limitation, but an API limitation. Two different wheels, with the same in-game FFB settings, will experience clipping at the same exact situation (even if one is 5nm, and the other is 50nm).

The thing is, the game doesn't know how strong your wheel, or what its limit is. It doesn't specify the effect strength by nm.

The api has a range, for simplicity let's say 0 is no ffb and 100 is maximum ffb (be it 5nm or 50nm, the game doesn't know).... The game does its calculation of force acting on the wheel, and comes with a number, the number is then multiplied with gain (0-1). This resulting number can be bigger than 100, if it is.. it will get clipped to 100, becasue that's the API limit. That's why lowering gain solves clipping.

This is the number that gets sent to the wheel driver. This is where the wheel SW settings comes in, and the number gets multiplied with the settings in the wheel SW. But at this point, the signal is already clipped.

That's why you want to keep the number ingame relatively low, because the lower it is, the more headroom you have in the api range. In theory you don't want it super low, like 10%, because then you're might be limiting the fidelity of the FFB, but this shouldn't be an issue for a 15nm wheel.

 

Posting this useful video here, originally posted by Marc Collins in the custom force feedback thread.

 

This is incorrect, you do not get 130% ffb.
Think of your wheelbase as your main amplifier feeding your speakers and the game as your pre-amp.
Once you have set your main amplifier gain to say 50% and set your preamp to 100% you are still only going to be sending 50% gain out of your main amp to the speakers.
It's all about gain staging.
Set your wheelbase to near maximum and then adjust the amount of signal being sent from the game using the individual sliders and overall gain controls.
Sending too much signal to your preamp will cause distortion.
Sending too much signal to your speakers will cause distortion i.e. Clipping

The same is true for FFB

(The game settings are your preamp with tone controls, and your wheelbase is your amp to the speakers)

 

I won't start any more discussions about this. If you like setting 100% on the base (I have no idea what kind of base you personally have), then go for it. If you like adjusting the various controls on the base, then go for it. To each his own

 

This is incorrect. If you set the in-game gain to 100% and the wheelbase to 50%, you'll end up with a high-amplitude signal, resulting in significant clipping because the wheelbase doesn't have enough torque headroom to reproduce the maximum output from the game.

On the other hand, if you set the in-game gain to 50% and the wheelbase to 100%, you'll have a lower-amplitude signal from the game, giving your wheelbase more headroom to deliver maximum torque without clipping.

Therefore, 100% game gain with 50% wheelbase torque is not the same as 50% game gain with 100% wheelbase torque.

Also, having a higher in-game gain (e.g., 100%) provides a more detailed and amplified signal to the wheelbase. This makes subtle forces and details, such as road textures or minor vibrations, more pronounced compared to the inverse scenario.

 

That's exactly how it is

 

I can see how people experienced in audio signal engineering would think that clipping is only at the output, as this is how a signal chain works in a digital DAW for example, but the case for sim games is that it can be clipped before reaching the output, and so the wheelbase actually can be simply amplifying an "already clipped" signal. As an audio engineer myself, this took a while for me to get my head around, that the wheelbase itself and its software are simply amplifying the signal from the game, - ie: If you have clipped the signal already and caused a flat topped compressed wave, then you will have less detail in the signal no matter how many nm wheelbase you purchased.
I used to look at the FFB Graph in R3E and think that the clipping it was showing me was at the base, but no, the game has no idea if my base is clipping.

Yeah so if you are an audio guy who uses DAWs and your thinking in terms of individual tracks peak meters going red from hitting 0.0dB and have the old DAW forums argument "It's not really clipping until we get to the master bus" - this doesn't apply in the sim racing software realm.

Matus Celko explains this well because he describes how there is an API in between the softwares which translate everything and so it can clip here before reaching your base.

High in base software + lower FFB gain in Racing Sim Game = better detail, less clipping

 

Correct, and actually the fact that there is such a thing as "Low Force Boost" kinda proves this, because it tells us that under a certain signal level detail will be completely lost. (This does not happen in digital audio, because even if the signal is -80dB it is still there and can be amplified back up without lose of detail) It tells me that LFB is a kind of compression of the signal so it can be raised to contain smaller parts of the signal without clipping the top by simply increasing the overall gain. If the signal "could not clip within the game" then there would be no need for a LFB slider, as we would simply increase the in game gain to feel the small details on weaker wheels. The LFB slider control kinda shows that we are working within a window of sorts, and beneath a certain level, your wheel will never even "hear" that signal to begin with.

 

@Domagoj Lovric so which is it? 100% in game and adjust base to taste or 100% on base and adjust game to taste? I was always under the impression it was 100% base then game??

 

LFB in AMS2 essentially amplifies the weakest part of the FFB signal, similar to how it worked in AMS1:


source: Automobilista - How the FFB works and what is new about it

The "Basic/Reiza Custom/barebones" FFB file in AMS2 includes the following lines:

##(LOW FORCE BOOST)
################

(output ( lfb output 0.15)) [my emphasis]

The AMS1 LFB graph indicates the 0.15 value specified in the AMS2 custom FFB file acts as a pivot, for lack of a better word, between the initial, more-than-double, signal boost (note: when LFB slider is set to 60) and the subsequent roll-off to preserve a relative degree of linearity for all values above the specified LFB pivot. Just for the fun of it, it is possible to specify different LFB values in the Reiza custom file to test how different lfb output values interact with different LFB slider values.

 

There is snake oil and leech doctors a plenty in both SIM racing and audio engineering. It's most amusing watching this 100% game Vs 100% wheel base discussion.

This is the best opinion here, and the true answer is that for every wheel base and user the answer will be different. What feels (sounds for the audio metaphor) right for you is what is right! It may be 100% wrong for another. There is a ffb meter in the game, use it! For the best fidelity have absolutely no clipping. The wheel base software may also, by the settings used in the driver, cause clipping, but no wheelbase driver I've ever used aside from Simucube, has had any indicator of clipping. And even with the Simucube driver, I'm not convinced it could detect a cllipped signal from the game! I sususpet this is only detecting clipping within the driver. Moreover if your power supply to the wheel base is substandard or failing it too could introduce clipping or distortion, as could any cable, digital or analouge... It's a deadly minefield!


Best practice I'd recommend, limit the output of your base to say 50/75% if it's powerful enough to cause injury, anything above 10nm perhaps, remove all filters/settings in the wheelbase driver so it's as linear and natural as possible. Use the ffb meter in game with all the settings in game to provide the best feeling, increase/decrease gain so there is absolutely no clipping and leaving a little headroom. That is to say using using 80 to 90 of the availble range before clipping. This will give some leeway in the driver software to prevent clipping there. Refine said feeling with wheelbase drive settings. Go back in game, make the signal clip on the meter, access whether you like distortion today (maybe tomorrow will be different )... Finally adjust wheelbase gain until your happy happy happy . Every car will need to be checked for clipping, use the per car adjustment.

Bottom line, tldr, the best feeling/sound (to the individual) may indeed have clipping, a distorted guitar/synth sounds good to an individual sometimes, and others not (both 'times and 'vidual). Filter, gating, eq, revamping, compression, limiting, clipping, saturation, reverb, chorus, you name it, is all a distortion of the original!

Final note, as the great Salvador Dahli said, "my aim is to spread confusion, not eliminate it." Just do what feels best for you, that can't be wrong, but it may well change...

EDIT for typos and clarify.

 

While it is common to use every wheelbase and specially powerful direct drivers like my DD1 at 100% and limit the gain in the game, for a few reasons:
Game signal never clips
You are using all of your wheelbase potential to amplify even the smallest details
This can work in many games, and it is usually the way to go.
Seems to extract "more detail" from a smaller signal (better on other games IMHO)

But in AMS2 I figured out that it is better to have the FFB not so low in the game settings, and to limit a bit the gain to your liking in the wheelbase software. (instead of using 100%)

If you can't handle big spikes at 100% (20nm) go lower, even if for you that would be 50%.

The reason is that even at lower FFB gain in AMS2, you can experience dangerous spikes that can hurt you by suddenly using more power than what you would normally expect.

At the same time, the FFB here seems to be layered by 20% 40% 60% 80% 100% and above 100% in red (clipping) inside AMS2. Using very small amounts of gain seems to "compress" fidelity and makes the whole signal work inside the 20% ffb telemetry bar, or maybe just the first 2 bars (20% and 40%) which can be limiting since then your wheelbase will amplify smaller forces together with big forces into too high of signal for such details.

Then people tend to over dampen the FFB settings and I've seen the most ludicrous settings with tons of damper friction inertia and all sorts of filters.

I would not recommend this for every game, but AMS2 seems to be a special case where a healthy compromise makes more sense. The ffb from the madness engine seems to have been designed to work better if you explore it's resolution (still without clipping) in game, and then limit the force on the steering wheelbase software.
(I've tested this both with Default+ and Custom files)

If you think otherwise you're welcome to pickup the McLaren 570S GT4 at Road Atlanta and beat my time of 1.22.9 there using your version of the settings, and if you do please share them with me so that I can get faster as well I would be very pleased if you would provide me with better ffb.

 

Thanks for this forum link John, so like I said, LFB is adding a y = sqrt(x) shape, by I guess adding this factor in that equation you listed.



I understand this to mean that in order to have a linear curve such that you would get the same details down low without using LFB (Red Line), you would need to increase gain to a point where even pure force level 0.4 and above would start clipping the output.

Thus, (and something cannot be graphed here as every wheel different) is that I assume the motors in my R5 wheel have like a minimum signal they can even "hear" or better said "reproduce" .. so whilst it looks on this graph that all start from the same 0 point on the x line, if we graphed this from the actual final wheel output, I assume that stronger nm wheel like R12 will start actually making FFB before weaker ones like mine.

 

I think there are many good explanations from various people on the last page of this thread. I also tried for a long time to understand everything, including logic, graphics, and code. In the end, I had to realize that there are no "perfect explanations" or precise operating instructions like you would find for assembling an Ikea piece of furniture. The requirements and possibilities are too complex and diverse to provide an operating manual for a FFB. Physics, the Madness Engine, the variety of cars and tracks, the different car setups, the different base systems, the different forces, the different hardware, and much more... everything will change your FFB, and that's why, for me personally, it's always a compromise (in every simulation, by the way). For these reasons, 90% of the time, the only thing that counts for me is the feeling I ultimately feel on my steering axis and the various bases, and thus on the wheel