Automobilista 2 Custom Force Feedback - Overview & Recommendations

Wheel base strength/gain should always be set to max on the Windows side and reduced ingame to suit. Failing to do this and you will experience clipping. And why would you spend the money on a 20nm wheel only to turn it down? On my wheel for example which is 26nm I only use about 20 gain ingame and zero everything else. I just add damping to suit.

 

Hey Supa!

It's not just about how much power your base has.This is completely irrelevant.The very large dynamic range of a DD is much more important than having 20NM force.If you tell me that you use 20% gain in the game and 100% on the base, I can tell you that it is the wrong way.You will lose almost all FFB feedback with 20% gain in the game.You will mainly only feel your pure power on the base, but that has nothing or very little to do with FFB.Anyone can test this themselves with 100% to 20% and then vice versa (then take a look at the FFB graph). It also becomes dangerous when there are impacts or contact with opponents.

 

Hey, your file works great with a Moza R9! The only thing is that the wheel friction/resistance is a bit too high in the corners, what lines could be adjusted to lower that effect a little bit? Thanks!

EDIT: I switched from the 5-8Nm file to the 8-25Nm and it feels better now

 

What car and which Moza settings are you using?

In response to your edit:
I just got the R12 and it seems ok, but I went directly to the 25Nm file
It's supposed to be the more neutral, not trying to compensate for anything...

 

I could be wrong and someone with more knowledge can correct me but what I understand by far is:

Base Gain (or Strenght if you prefer) is a percentage of the Nominal Base Power or Nominal Base Torque (I really don't know and I suppose it doesn't matter in this case).

In-game Gain express how much of the Base Gain is transferred to the game.
Combining those two Gains you'll have the Final Gain witch naturally is also a percentage.

And the Final Gain equals Base Gain versus In-game Gain.

Setting up your Base at 100% you could be happy because you're "using" all the Wheel Base can offer.
Setting up In-game Gain at 20%, you'll effectively using 20% of that Base Gain or merely 20% of the Total of 100%.

It is evident that many combinations of Base versus In-game game gain will lead to the very same result.
And to a extreme, setting Base gain at 20% and In-game at 100% the resulting gain will also be 20%.

Conlusion is doesn't matter the individual gain setting of Base and Game but the combined gain.

Note that even if the wheel base isn't using the high forces it can deliver, if you set up them too high in some bases and specially the lesser powered one's, that forces will be still electrically available and generating internal forces to the motors (and internal gears) stressing the base without any exploitation.

So, an aditional conclusion is don't stress too much your base.

 

C

can I raise in game and lower the base side instead? Kinda to add to the gain conversation. If I have the game end lower than 100% wouldn’t I be getting less than the 17-18nm you’re saying? I figured both cut the max power as it averages out but I’m not totally certain how it all works.

 

Wow V300. I've been away for far to long.

 

The problem is clipping. Every stage should be clipping-free or else the signal gets squared/topped out, no matter what you do in the next stage.

Reiza has to ensure car-specific FFB is not clipping. Everything else is up to us.

This explains some factors:
Automobilista 2 V1.4 Force Feedback Overview & Recommendations

And yes, we can buy a 20Nm wheel but limit it to 10 or 15N not to get hurt. No game setting can unsure that. We will still get benefits from the overall system.

 

Hey Theo!
I'll try to explain it as simply as possible. Sorry for my English, it's not my native language. The gain in the game scales everything from the game and the custom file, the gain at the base is just the pure power of the base and that's why it's a very big difference. Not so much in terms of forces but even more so in terms of the FFB. Either you lose a lot of FFB or you get too much FFB (clipping) if you set the gain incorrectly in the game. Now in Ams2 there is a good sweet spot for gain in the game (60-65) to be able to use the full potential. Of course, this also depends on the Custom File, and this value is specific to my Customs Files. Other Reiza customs or defaults can have completely different values. If you have now set this value to 60-65 then you can adjust the gain on the base accordingly. This gain works very well with around 40-60 gain depending on the car and wheel and personal taste. That means you have forces of 10-15Nm, which should be enough for all cars. There is an easy way to test it yourself. Test the 100% and 20% gain and then vice versa with 20% and 100% gain in the game and in the base. Observe the FFB graph in the game and how different the curves and deflections will be. I hope I could help you a little

 

Yes, you should also remain realistic. What's the point of driving such cars with 20Nm of power? This is not realistic and makes no sense to me. The only place it makes sense is with go-karts. Increasing or decreasing the vehicle-specific FFB is not important to me and I don't need it because it can be bad for the FFB. It changes the gain in the game and therefore the whole FFB feedback. For me it is always at 100%, I only change the gain on the base depending on the car. Yes, you're right Reiza has a force feedback thread where you can learn and read a lot. However, in my opinion this is no longer current and should be revised. There are things that have changed a lot and no longer work the way they originally did.

 

I'm realistic, I guess.
Why buy a car capable of 230Km/h if you only make 120Km/h in some (a minority) of the roads?
Because it will accelerate faster, brake better, turn better, have better safety features, usually have more/better confort equipment, etc.

Why buy a wheel base capable of 20Nm if your are only going to use 10Nm?
The same reasons!
Torque is not the same as power, not the same as rotational speed or acceleration, usually are built for life (DD motors are present in most of our home appliances for decades!).

I bought a 12Nm wheelbase and I'm using it at 50% torque (I want to gradually increase it, but keep inside my safety margins).
I already notice your files works very well with it, mine don't. With my TS-PC was exactly the opposite and I'm (theoretically) using the same 6Nm max torque...
If I hadn't more torque available, I would be happy already.

 

As I said, it makes no sense for me personally to drive with such forces of 20Nm or even more. Except for go-karts, because they also experience such forces in reality. And I won't write a scientific study about it or create tables. In the end, I just offered people an additional way to have a good experience with different bases and different customs. And above all, it should be easier. Base is given then select Custom and have fun without having to deal with complicated settings. Whether people like it or not is up to each individual

 

I spent a couple hrs tonight writing a little "dependency analyzer" for `ffb_custom_settings.txt`, that helps make sense of how a variable is directly & transitively used across the ffb file. Written in python, call it as a cli specifying the ffb file name / path + the variable of interest. It spits out all lines that either directly or transitively use the variable, and displays the "dependency chain" of variables leading back to the input variable of interest (highlighted in yellow). Just a fun little thing I'm using to help me understand how a given change might impact things. Would anyone find this useful / interesting? Not sure how useful this is in and of itself, but I think tools to make sense of / help understand how modifications would impact ffb would be neat. If anyone who has spent much time modifying FFB files and has a gut feel for them has feedback on what *would* be useful, that'd be interesting as well.

`python3 ams2_find_var_depencency_chains.py ffb_custom_settings.txt steerPct`

 

It sounds good and would be nice to have

 

Hey PAV!
There are definitely people who find your work excellent to understand it better.And above all, it brings you something personally - that's the most important thing.I personally don't need it, but I still really appreciate your work.Maybe you could write a custom and publish it in this thread.It would be a pleasure for me

 

could be quite useful, it’s something that a lot of people don’t understand how the custom file flows , I’ve described like this before and will do again.
The file is like a backwards tree , it has a main trunk that feeds to the roots which is the final output, but it’s variables are the branches, and they are ment to be free of each other and not tangled , so a good tune is a well pruned tree.

I see a lot of files that have important bits of branches lopped off and others that have extra branches that have no physical connection to the truck , thus have no impact on output , yet placebo dictates that it does .

The main problem I see with this identification code on quick inspection on face value , is some branches that are completely separate can share the same terms , which makes it appear to be a dependency , but it’s not necessarily the correct one . It depends on how high up the trunk they blend in .

 

It would be nice to have a watch window for some variables (values, graphs, etc.)
It is a pity AMS2/Madness engine doesn't offer this - let's not forget what "CARS" stood for (Community-Assisted Race Simulator). Right now, it's only possible violating the law and/or ethic/moral concerns.

We leave the game, change the file, run it again, test, repeat... this is not a method. It's a sacrifice .
I had to transpose variables to histogram to see the changes in real time, use in-game dials to adjust several parameters at once and it's still a pain.

It's a pain to create/adjust custom files and they are essential to AMS2 success.

 

This is super useful. I will need to digest this a little bit. As I mentioned, I threw this together pretty quickly precisely because I am new to tinkering with this file and don’t understand it well. The bit about using a tool to validate that variables are actually having an effect on final output is interesting as well.

Question for you all… numerically, does anyone know what the actual values of final output are, concretely? I also don’t quite yet know these files well enough to understand if there is more than 1 final output, and if so, how do I know what this final output is (just variable names?). I told you I’m a novice .

 

Totally. Any details on the transpose thing? Also, yeah, it would be the dream to not need to quit the game .

 

Yes - it definitely brings me something personally! I have a file that I’m having fun with but is likely way too aggressive for most people’s tastes lol. I might share it at some point.