PC1 bManic's FFB experiment - Thrustmaster TX / T300

[bmanic]

UPDATE! I changed Relative Adjust Bleed to 0.3 instead of 0.45 (which could cause some oddities)

UPDATE2!: More and more people with other wheel brands have tried the settings and it's starting to look as if these may work for quite a few wheels. T500 is confirmed as working pretty well, Fanatec GT2/GT3RS v2 also confirmed as working better than defaults and one Fanatec CSW v2 also happy.

UPDATE3!: There seems to be some confusion about Steering Arm Angle and what it is. It is NOT the same as Steering Ratio, which is a VEHICLE SETUP parameter. Arm Angle can be found in the car garage menu, under the Force Feedback Tab (it's the middle tab, between Vehicle and Summary tabs), the last option on the first page is called Arm Angle. This is the one I mean. Everything in this thread is about Force Feedback, not about vehicle setups!

Update4!: Brandon Wright noticed quite a bit of spikes and odd behavior which seems to be related to the Relative Adjust Bleed parameter. It is possible that my value of 0.3 is still too large so if any of you guys are suffering from these spikes then try a lower setting. You can go as low as 0.05 until you lose the benefit of the Relative Adjust processing block.


Hi guys,

So I finally had some time to play the game and I've spent the whole day experimenting with a new way to squeeze the massive FFB dynamic range into a range that our consumer wheels can tolerate while getting as much information as possible from the FFB (meaning: no hard clipping).

I would love it if you Thrustmaster TX and T300 owners would try these settings. NOTE: You must follow these instructions TO THE LETTER or the experience will be different to mine. Feedback that is provided without following every single point to the letter will be useless and thus ignored.

Why? Because the FFB system in pCars is extremely flexible and thus complex. It is also highly dynamic which means that a single tiny tweak of a crucial parameter (like tire force / master scale multiplier, spindle arm angle or the ratio of Fx, Fy, Fz and Mz) will alter the system. Thus, if you want stronger or weaker forces, use your Thrustmaster Control Panel program to control this! Do NOT change the in-game settings to achieve this!

These are my Thrustmaster Control Panel gain settings:

Overall strength of all forces = 75%
Constant = 100%
Periodic = 0%
Spring = 0%
Damper = 0%

Auto-center settings = by the game (recommended)

Step 1)

Set your Thrustmaster Control Panel settings. You can use my settings to start with. These result in fairly heavy forces. A good workout but not impossible to drive with. If you are a masochist, set the strength to 100%. I often drive longer races with the strength set to 60%.

Step 2)

Start Project CARS and reset your wheel then do the wheel calibration again as follows:

a) turn wheel FULLY LEFT AND RIGHT.. all the way! My TX wheel requires me to truly wrestle the wheel to get it 100%, otherwise it stops at 96%.

b) follow the 90 degree rotation to the letter.. this means you set your wheel so that it is physically pointing exactly 90 degrees to either left or right. Do NOT set it so that you get perfectly 900 or 1080 degrees here. Set it so that it is truly pointing 90 degrees to either side. For example, my wheel when set so that it is physically pointing 90 degrees results in a steering rotation of 882 degrees.

c) calibrate your pedals

d) configure your buttons

Step 3)

Make sure your FFB Strength in-game is set to 100/maximum! This is very important! It defaults to 75. Also make sure the FFB damping saturation setting is set to ZERO. It defaults to 25. (you find these options under the CONFIGURATION tab in the controls preferences)

Step 4)

Go into the "calibrate FFB" menu and set the following (parameters not mentioned should be left at default!):

Tire Force = 100

Deadzone Removal Range = 0.03
Deadzone Removal Falloff = 0.02

Relative Adjust Gain = 1.37 (sometimes it shows 137 and leaves the " . " out)
Relative Adjust Bleed = 0.30
Relative Adjust Clamp = 1.33 (sometimes it shows 133 and leaves the " . " out)

Scoop Knee = 0.12
Scoop Reduction = 0.08

Soft Clipping (Half Input) = 0.60
Soft Clipping (Full Output) = 1.79

Step 5)

Individual car setup FFB settings! I've only had time to test 3 cars but from these you can quickly extrapolate and get the general idea. The main idea is to have a Mz heavy setup and to set a good spindle arm angle and then to set a good Master Scale value so that the FFB "pushes" into the Relative Gain and Soft Clipping at proper levels. These usually end up in the 26 to 36 range, depending on how much down force the car produces or how heavy it is. Here are the settings for the 3 cars that tried:

<--- CAR SETUP - FFB TAB - --->

Lykan Hypersport:

Master Scale = 32

Fx = 68
Fy = 36
Fz = 74
Mz = 100

(note!! set all Fx, Fy, Fz, Mz smoothing to zero!!)

Steering Arm Angle = 2200

BMW Z4 GT3:

Master Scale = 28

Fx = 66
Fy = 44
Fz = 82
Mz = 100

(note!! set all Fx, Fy, Fz, Mz smoothing to zero!!)

Steering Arm Angle = 2000

Formula B:

Master Scale = 30

Fx = 68
Fy = 36
Fz = 74
Mz = 100

(note!! set all Fx, Fy, Fz, Mz smoothing to zero!!)

Steering Arm Angle = 2500
----------------------------------------------------------------------

That's it. You are done. Go into the game and enjoy pretty detailed FFB. To appreciate the full range of FFB I highly recommend turning off all driving aids. I'll add more cars to this thread once I have a chance to test some more.

NOTE!: I will NOT be providing FFB .xml files because they do not seem to work as expected. This is a complete system where every individual setting feeds into one another and thus becomes a quite complex thing. The general signal flow is like this (at least it used to be like this right before release):

Tire Force * Master Scale -> Fx, Fy, Fz, Mz -> Spindle Arm Angle -> Relative Gain -> Soft Clipping -> Scoop -> Tighten Center

As I've understood it, Relative Gain and Soft Clipping are non-linear functions and thus quite complex in nature. Even small input variations can cause quite a different feel in the wheel. Because there are two of these functions after one another and they feed an inherently non-linear device (your consumer wheel) you can probably appreciate how quickly things get complicated. You'll easily notice this by simply tweaking the Master Scale of each car or change the Fx, Fy, Fz, Mz relative ratios. Like I said in the beginning of the post, the main point is to squeeze as much "useful" information into the limited dynamic range of our consumer wheels and this means compressing the heck out of it.


You lucky s.o.b's with direct drive wheels can ignore all this and simply set pCars as fully linear and have a ridiculously awesome experience. You may still want to adjust the Fx, Fy, Fz and Mz settings like I have them above but disregard everything else (remove all soft clip and all relative gain and scoop stuff. Heck, I'd even remove tighten center).

Cheers!
bManic

 

[jimortality]

im sorry but basing FFB on how quick you go is wrong.. imo..

You basically using the ffb to make the game easier for you to make you go faster.. real life drivers do not get this luxury..

I recommend making it feel as real as possible then try and go fast after that. otherwise its your making the sim even more fake.

Seriously Dan, the guy is spending a lot of time helping us to get a better experience and you try and be negative about it. Use your own settings for goodness sake but don't be a dick for the sake of it.

 

[Jezza819]

I partially tried your settings on the BMW 1M last night and took the car out and it had no brakes and I mean NO brakes. But I discovered I forgot to change the arm angle and still had it on 1000 which I was running on the old setup. Brought the car back in and moved it to 2000 and I had brakes again. Does Arm Angle have anything to do with stopping power on the brakes?

 

[bmanic]

Seriously Dan, the guy is spending a lot of time helping us to get a better experience and you try and be negative about it. Use your own settings for goodness sake but don't be a dick for the sake of it.

No it's okay, I've heard that complaint before and here's my answer yet again:

No, Dan, you are wrong. I am not making the FFB do anything unrealistic. On the contrary. I'm striving for ultimate realism. A real driver actually has a huge advantage over us simracers.

First of all they have true seat of the pants (G-forces acting on their body, they can feel the car moving in all 3 dimensions which is a huge help).

Second, they have no latency. The wheel inputs they get, that is their FFB, is instantaneous. This is another HUGE benefit they have over us simracers. When you are in a nasty oversteer situation in a real race car (had a few of those myself) you simply "follow the wheel", kind of let go of it, and it rotates lighting fast to almost the correct lock position to get out of the oversteer moment (due to the self aligning torque of the tires). You can almost do this in pCars due to it's awesomely fast FFB/input implementation (faster than any other sim.. try doing the same in Assetto Corsa for instance) but we are still way behind reality.

Third, the real life racing driver doesn't have to deal with a massively inferior, completely non-linear, cheap motorized steering device. Only a small fraction of simracers have or can afford a direct drive wheel where you can almost get 1 to 1 forces with a real race cars wheel. Thus us mere mortals need to figure out ways to squeeze the huge dynamic range of the FFB output into our crappy toy wheels. There's just no other way around it and the road is full of compromises at every turn. Tweak this or that and one thing gets better while another gets worse. It's like an audio recording of a gunshot.. for us humans in reality it's loud, punchy, sudden and violent. On a recording it's a pathetic little "snap".. unless you clip and saturate the hell out of it (which is why professional sound designers still record gun shot samples on to tape machines).

This is why my method of tweaking the basic tire forces + steering rack and basing it on laptimes is actually a pretty good method that gets us relatively objectively towards more realistic FFB settings. I'm not going for subjective "feel".. I'm not going for subjective fun immersion "road noise". I'm going for an input-output system that tries to mimic real life through the measurement of laptimes. If you guys have any other ways of measuring this objectively then please let me know. My system is of course not at all fool proof. It requires hundreds if not thousands of laps and to learn a few different tracks very well (luckily I've got over 20 years of simracing behind me so that helps a little). The key is consistency, not ultimate laptimes. There's no point in going for a single fast lap.. it gives you no data. However if you do 10 laps and they are all within +/- 0.1 seconds then you know you have solid data.

Cheers!

 

[bmanic]

I partially tried your settings on the BMW 1M last night and took the car out and it had no brakes and I mean NO brakes. But I discovered I forgot to change the arm angle and still had it on 1000 which I was running on the old setup. Brought the car back in and moved it to 2000 and I had brakes again. Does Arm Angle have anything to do with stopping power on the brakes?

Nope. FFB settings have absolutely nothing to do with anything else. However interestingly a friend of mine lost his pedal calibration while fooling around with the FFB settings. I think there may be a bug that kills the pedal calibration randomly.

 

[bmanic]

I stand corrected, arm angle IS available from the pit menu. I had been trying to use the mouse to scroll down and it wouldn't scroll so I figured there were no more options, but if I hit the arrow pad it does scroll down.

Indeed! A fantastic find! Saves me a ton of time. Thanks Chadd for the discovery!

 

[jimortality]

As promised, here are settings for the Clio Cup (what an awesome car!). Note! I did change the default setup a bit.. it was a bit too tricky for me to drive so I had to change a few things. In the end I managed to consistently lap Donington National at 1:17.7xx +/- 0.1 seconds.

Renault Clio Cup

Master Scale = 34

Fx = 82
Fy = 60
Fz = 68
Mz = 100

(note!! set all Fx, Fy, Fz, Mz smoothing to zero!!)

Arm Angle = 700

Cheers for that, I'm at work all night but I'll have a blast tomorrow. This is in preparation for the club race at Brno on Tuesday.

 

[Brandon Wright]

Nope. FFB settings have absolutely nothing to do with anything else. However interestingly a friend of mine lost his pedal calibration while fooling around with the FFB settings. I think there may be a bug that kills the pedal calibration randomly.

I've had this happen a few times myself, I think if you recalibrate the wheel it loses the pedal calibration.

I tried the settings and the Clio Cup settings, too light for my preference. I tweaked the arm angle, Fy, and master scale a bit and got it to where it has enough weight for me. Feels good, but even with smoothing I still get big spikes. For example, the wheel will jerk violently from 2 o'clock to 3 o'clock in a slow turn. Everything else feels good but I really dislike these spikes.

 

[Dave McMillan]

When you are in a nasty oversteer situation in a real race car (had a few of those myself) you simply "follow the wheel", kind of let go of it, and it rotates lighting fast to almost the correct lock position to get out of the oversteer moment (due to the self aligning torque of the tires). You can almost do this in pCars due to it's awesomely fast FFB/input implementation (faster than any other sim.. try doing the same in Assetto Corsa for instance) but we are still way behind reality.

I agree with most of what you said, except about AC.....

FFB nowhere near full (without clipping) and I can let the wheel do all (most of) the work while drifting 900 degrees lock to lock no problem in AC.

I dunno if I could do this in pCARS though, don't really have many drifting cars to try it with, ill maybe try the 1M later.

 

[Chadd]

Indeed! A fantastic find! Saves me a ton of time. Thanks Chadd for the discovery!

Glad I could help

 

[Kurupt CDN]

im sorry but basing FFB on how quick you go is wrong.. imo..

You basically using the ffb to make the game easier for you to make you go faster.. real life drivers do not get this luxury..

I recommend making it feel as real as possible then try and go fast after that. otherwise its your making the sim even more fake.

Sorry for the dislike but I have to disagree with your opinion

I tried all the cars he posted and group them by class and they all work very well and are extremely detailed with subtle effects Compared to the bas setting....too many people set there wheels with too much ffb which makes it heavy less detailed and result in fighting the wheel......the while purpose of ffb is to get as much detail as possible of what the car is doing with out clipping so you can react faster which results in going faster which these settings help do....may not be for everyone but it's an excellent start!

 

[SnappyDEe]

Do real race drivers tweak there set ups to help them go faster?

 

[Msportdan]

delete

 

[SnappyDEe]

@bmanic - I used your Z4 GT3 settings on my RUF GT3. FFB feels awesome. Only difference is my over all strength of forces on the wheel is set at 85 and the master scale is at 44 (as opposed to 28 for your Z4)

Question - is that master scale number too high? Am I loosing some finer details? should I turn it down and raise the wheel up to 90?

I mean, the FFB is very communicative where it is so I don't think I am losing any small details.

Also, in the SOP section there is a setting (can't remember the name of it, sorry) that really helps with feeling the vertical forces acting on the wheel but I find it interesting that it causes two things:

1. I can feel road details a bit more and gear shifts translate much more through the wheel.
(I really liked that so, not a bad thing)

2. Under hard breaking where I am also downshifting I can feel the front wheels starting to lock up as more and more load/compression is placed on the front wheels. Very interesting. It was keeping me very honest in the RUF at Catalunya (especially turn 1).

My Equipment: Thrustmaster TX (Ferrari F1 add-on wheel), Thrustmaster T3PA Pro Pedal Set, TH8RS shifter.

 

[Msportdan]

sorry but you can set tyre and ffb to 100 as long as you bring down the Master gain of each cars It makes no difference... you don't loose details as long as you bring one of these options down..

I agree Sop values help greatly with the feel of the car.

my sop settinsg are

30/30/30/0

 

[jimortality]

I've had this happen a few times myself, I think if you recalibrate the wheel it loses the pedal calibration.

I tried the settings and the Clio Cup settings, too light for my preference. I tweaked the arm angle, Fy, and master scale a bit and got it to where it has enough weight for me. Feels good, but even with smoothing I still get big spikes. For example, the wheel will jerk violently from 2 o'clock to 3 o'clock in a slow turn. Everything else feels good but I really dislike these spikes.

I've had this happen a few times myself, I think if you recalibrate the wheel it loses the pedal calibration.

I tried the settings and the Clio Cup settings, too light for my preference. I tweaked the arm angle, Fy, and master scale a bit and got it to where it has enough weight for me. Feels good, but even with smoothing I still get big spikes. For example, the wheel will jerk violently from 2 o'clock to 3 o'clock in a slow turn. Everything else feels good but I really dislike these spikes.

What wheel are you on Brandon?

 

[SnappyDEe]

Well, your FFB settings are the best. Thanks so much for your hard work. I am going to lower the in-car master gain/scale to something closer to what you have and then try different FFB levels on the wheel. Thanks again.

 

[Brandon Wright]

What wheel are you on Brandon?

The T300.

 

[bmanic]

I've had this happen a few times myself, I think if you recalibrate the wheel it loses the pedal calibration.

I tried the settings and the Clio Cup settings, too light for my preference. I tweaked the arm angle, Fy, and master scale a bit and got it to where it has enough weight for me. Feels good, but even with smoothing I still get big spikes. For example, the wheel will jerk violently from 2 o'clock to 3 o'clock in a slow turn. Everything else feels good but I really dislike these spikes.

I think I might have figured it out.. give it a go: It may be that you are experiencing buggy behavior from the Relative Adjust Bleed setting. Sometimes it has some really weird snappy spikes and they seem to happen more often if you run higher tire forces or master scale multiplier. Try setting Relative Adjust Bleed (in the controller settings) to a much lower value like 0.05. Does it help? I originally had it at 0.45 for my new settings but noticed odd spikes, especially over curbs and bumpy sections of the track so I dialed it down to 0.3. It is very possible this is not enough.. I suspect something is a bit buggy with it and I've experienced it before (at one point during the alpha it defaulted to 1.0 which was very weird and very buggy). So yeah, give it a go. Set it to a low value (original FFB settings default to 0.08 I think).

Cheers!

 

[bmanic]

Question - is that master scale number too high? Am I loosing some finer details? should I turn it down and raise the wheel up to 90?

If your aim is to get higher FFB forces then yes, you should first set the wheel as high as it can go, even to 100 if that is how you like it. Then try to up the Master Scale.. a value of 44 is quite high and is probably pushing very hard into the Relative Adjust Gain and Soft Clip. In the end it's completely subjective, there are no "wrong" settings per say.. but too high Master Scale settings will not work optimally with the kind of system I had in mind (= detailed FFB).

There are other ways to get more FFB strength that you could try. You could for instance set Relative Adjust Gain to 1.50 and Relative Adjust Clamp to 1.48 or something close to this (just keep it a tiny bit below 1.50). This will mean that the Relative Adjust system pushes harder into the Soft Clip.

It's good to remember though that if you want truly accurate, highly detailed but heavy FFB forces there really is no shortcut other than buying a better/more expensive wheel with a properly strong motor. The more you push our relatively "cheap" consumer wheels the less detailed they become. It's easy to get addicted to too high FFB instead of trying to learn to drive with a lighter wheel that has more information.

 

[bmanic]

my sop settinsg are

30/30/30/0

Those are extremely high SoP settings and will heavily mask any FFB coming from the tires. It's important to understand that SoP is a completely separate system and is NOT getting it's input from the tires. It gets it from g-forces. So yeah, you can definitely drive with heavy SoP settings but you'll never learn the absolute optimum slip angle that way.. and it's kind of important.

I do use SoP sometimes but then I set it to 2/20/10/0.