After two weeks of failed attempts to get a reasonable force feedback (including several posts to this forum), I decided to change my approach and try to sort this out by myself. Having no idea of what each parameter actually does, I adopted an hands-on investigation method.
I won't bore you with the details, because most probably they won't be of any use to anyone else, and also because what I've got so far isn't even remotely comparable to the out-of-the-box FFB in other PC sims, although it finally allows me to play Project Cars without breaking my wrists. Anyway, if you are curious, I can send you the spreadsheet (which may also be useful to anyone who is interested in following a similar procedure). What I am posting here are a few findings and some doubts I had along the way. Perhaps someone can help me clarify those.
I used my Fanatec CSR wheel with everything set to zero, except force feedback at 100. The same with PCars: no car parameter tweak file or anything of the sort. Since I wanted to have a general base to start building on, I worked only with global FFB parameters, not with car-specific values. For the tests, I drove the Catterham 500 at the Coulton speedway.
Initially, I set everything to zero in the menu and then went down methodically from Tire Force all the way to Soft Clipping Full, testing every parameter by itself and then together with a few other parameters. What I found out:
1. With everything at zero, a Tire Force of 20 (!) was sufficient to have some feel of the road and the trackside, although the wheel was still too heavy. With more than that, the wheel started clipping on turns.
2. The only parameter that seems to reduce that heaviness without killing all road feel altogether is Deadzone Falloff. There are others that make the wheel lighter, but they also produce all sorts of weird effects (clipping on turns, flatlining on straights, what you will). In my case, a value of 0.01 was enough to achieve that. With 0.02 the wheel begun to feel too light.
3. Looseness of the wheel center can be fixed with Deadzone Removal. The procedure here is to increase the value one notch until the wheel starts to shake on the track, then go back to the last value that worked.
4. I felt almost no difference with the Linkage group. Linkage Scale doesn't seem to do anything by itself; probably it is a multiplier of other parameters. If it is so, it probably shouldn't be placed as first in the list. On the other hand, both the Scoop and Soft Clipping groups seem to have strong and wildly fluctuating effects. Since I couldn't discern a consistent pattern and had no patience to try every single combination, I finally left everything at zero as well.
In the end, what worked best for me (so far) was Tire Force at 20, Deadzone Removal at 0,06 and Deadzone Removal Falloff at 0,01, with everything else at zero. The wheel is light to drive and now I have a reasonable feel of the road. I could have gone on testing, but I stopped when I noticed some strange things happening (see below). I really wish I had more information about the other parameters and what they are supposed to do.
Now to some weird findings during this experiment:
1. The arrows beside the parameters in the FFB tuning page and the bar to roll down the list, although clickable, do nothing ľ at least not in my PC. To roll down the list you must use the mouse wheel or the relevant keyboard keys. The same with the values: you must use the mouse directly on the slider for small adjustments or the arrows for larger steps. Is this only me or there is a bug in the interface?
2. When you turn on the engine, the force feedback line in the virtual dashboard remains flat (as it should be, since the car is stationary) but immediately jumps from the halfway vertical position to the middle of the upper scale. Nothing is felt on the wheel, but this limits the area you have to work with before clipping starts. Is there a reason for this odd behavior?
3. The behavior of the FFB line during sharp turns appears to depend on the direction of the turn. I tested this extensively at the first sharp turn and at the chicane in Oulton. If you turn right, the line goes up almost to the limit or starts clipping, depending on other parameters. If you turn left with the same parameters, it remains in the middle region and you only see spikes. In fact, at the first bend in Oulton, the line begins to go up even before you reach the turn. At first I thought it could have something to do with the car going up or down the road, but this happens even while entering and leaving the same chicane. I also tried to take the same curves at the same speed in the reverse direction, and the strange behavior persisted. Is there a reasonable explanation for that?
4. Finally, something that may have to do with the loss of FFB adjustments some people are experiencing. When I finished testing and dialed back the best values I had found so far (see above), leaving everything else at zero, the car was impossible to drive. After trying several possibilities, I finally left the game and launched it again. Bingo! The car was drivable again. After further testing, I found out that the Soft Clipping Value, when moved out from zero, never actually returns to zero. You see zero when you move the slider, but when you remove the mouse the value is actually 0.10. This seems to produce a heavy amplification of all sorts of FFB effects. The only way to reset it is to restart the game. Has anyone else noticed this?
My conclusion is that the FFB system in PCars, in addition to being too obscure, is also fairly unpredictable and possibly prone to bugs. I have several other points I'd like to discuss (like similar trackside obstacles that produce different FFB effects), but I will spare you that for now. In the meantime, I would appreciate any help you guys can give me to proceed. Sorry for the long post!
- ffb.jpg (100.2 KB)
Original Writen by aleph99 in PC - Technical Help & Support Category, the date of 24-06-2015 05:49.