AC FFB Implementation, Improvement Suggestions

Well. What we need is some compression to the FFB signal, I think. Since we don't have wheels that are mathematically accurate, provisions must be made to approximate. Our wheels just aren't very dynamic, so a gamma slider or adjustable compression of some kind would come in handy. The range of wheels and capabilities is equivalent to trying to make something sound good on a phone speaker and high end stereo at the same time. Pretty much impossible, and to please more people you need some processing to cater for the lower end guys.
Or you could just say "screw you guys, get a Bodnar wheel". Which is fine, but not many are going to afford one....

 

The overall concept is very interesting and already very promising but the Steering with the t500rs is atrocious. Massive input lag, too slowly, lacks dynamic and feels dull.

It feels simply like driving a vehicle with a flat tire but the flexing work not to feel while the kinematic is apparently completely eliminated by something else. The patch feeling is great but tire wall forces and kinematics simply not exist. You guys talking about no bump steer and feeling of modern power steering cars but no lateral forces to transfer to the handlebars (track rod ) is simply wrong. In this case also the load changes are not clear enough.

A mix of netkar pro's dynamics with this tire patch could be nice.

 

I think FFB software in general, is a "young version" and will be implemented. I think some software mods we are testing will be default in early future.
The FFB now seems to me to be too simple, and too weak (too weak for all force settings). I have an old DFP and loved the FFB system of lot different games like netcar and gtr2, and in recent days I think project cars FFB (which is not good by default and need to be modded) is really FANTASTIC. The physics of this game is FANTASTIC, so it need ad adeguate FFB by default. IMHO.

 

I have tried many things with the t500 and now ffb is not bad.
But one thing is still missing : sensation of road grip is not strong enough. It leads to a lack of precision is slow turns for instance. GT2,is ally hard to drive, but other cars too.
We clearly feel this lack of grip when we do a fast right-left or left-right in a chicane for example. There is a kind of "dead zone" when you pass from right to left (or left to right).
G25 and g27 owners in my league seem to have more feeling than I have.
Apart from this, excellent game.

 

I'm using the wheel setup recommended by KS at their forum site and I still find the FFB a bit weak and the tracks feel very smooth in AC.

I was watching the AC pedal meter bar and IIRC someone said the gray bar on the right is for measuring FFB clipping? As I understand it, we should try to avoid the bar reaching the top and turning red (not sure what the bar turning red indicates...saturation? losing data?)?

Thanks!

 

Exactly. A little bit of red is OK, but if you see too much you need to lower your FFB gain in the controller options, or your FFB will be maxxed out and won't be able to deliver any more subtle effects. The gain needs to be checked for each different car; they all clip differently.

 

I have forgotten to say that my lack of road feeling is not due to a clipping or a saturation. All stay grey.

 

Yes, I have this deadzone in the FFB on changes of direction too, the wheel doesn't load up until far too late.

 

It would be nice to read someone from Kunos about this. I strongly believe in this game but I think I have to be reassured on this point
Do you work on this ?

 

Has anyone ever thought about implementing a FFB gamma slider? So you could increase the overall level of FFB without risk of clipping as the signal increases. Kind of like adding compression to an audio signal.

 

I've suggested this in numerous FFB threads on here. We pretty much get laughed at by the devs if we don't like their perfect FFB...

 

Well, FFB is best when linear.

 

I agree if we have an idealized actuator that will convey every force level and frequency in a completely linear fashion. Our toy wheels are nothing of the kind, and each model behave VERY differently in terms of linearity. Since we have no way of compensating for thin non-linearity in drivers, this is down to the simulator developers to provide for. In some cases they do. iRacing have the linearity scale thing which basically is a dynamic compression.

Ideally we should have an "equalizer" and a set of test forces to fine-tune the ffb for each wheel, but I have no hopes of this since it's not an acknowledged problem. and most users are happy with the occational jolt and vibration in their wheel, regardless of fidelity.
On top of that the developers have the attitude of "not our problem, buy a better wheel or nag at the wheel manufactureres to fix it. Our physics are perfect", which make me less than optimistic.

 

To bad. It makes me sad when i read all this. Full of one's own importance that is nothing else. Such people usually do not learn anything more. Shame on me and the money I have invested.

 

If they would have to listen to every user request and implement in AC, we will have Multiplayer in 2017 and AI in 2023.

A very very large majority of the users is extremely happy with the current FFB in AC. Why would the opinions of maybe 3 or 4 vocal users matter so much to them?

 

It shouldn't. What should matter was the simple logic of the issue. They listened to those who wanted to drive a high end sim with a MOUSE. But they won't make a small alteration to compensate for flaws in the controllers MOST of their customers use? They listened to those that wanted canned effects for the FFB, but not those who want a physics based ffb?
Germans voted for the nazis too, the majority isn't always right....

 

Because of the same reason as not everybody is a champion or the best in anything, apart from consumer service.

 

It was just an idea

Clearly all brake pedals respond differently, hence the inclusion of brake pedal gamma. I just thought FFB gamma might work as well, as all wheels respond differently.

And Requiem84, it wasn't a request - merely a suggestion

 

As long as you turn off damping and whatever other on-wheel effects you can test this pretty easily (I say this with no idea how to write actual code; I imagine the directInput is actually somewhat tricky to implement)
Physics says F=ma - you can determine acceleration using the wheel's position signal, and mass is constant (weight, or rather rotational inertia, of the wheel). So what you do is apply the range of force signals (10%, 20%, ...) and see how fast the wheel accelerates. Gives you a nice demonstration of the linearity (or not) of your wheel, and will also reveal what the actual max-force is, and maybe friction too (which affects how much it feels like the wheel's torquing obviously)

It would be nice if a game could use this sort of principle to calibrate your wheel automatically; then let people who care tweak it, but it'll work predictably for everyone.

 

It's not just about slew rate, it's also about frequency response. A T500 is much faster than a MOMO for instance, and also faster to reverse direction, so it will have a higher slew rate and reach a higher frequency (better step response).
Thing is, we have all these settings for everything and their grandmother for graphics to accomodate different configurations, but wheels and force feedback haven't been given a second thought, except "throw them some canned effects, that'll shut them idjits up"....
A gamma slider would be an easy way to amplify the smaller forces, and a high frequency boost to the signal will make it come alive and in effect compensate for the inherent lowpass filter of a mechanical wheel. You only need to take a 2 minute drive in the classic formula cars in rF2 to know how it SHOULD feel to drive a high performance machine.