Page 1 of 10 123 ... LastLast
Results 1 to 10 of 93

Thread: My FFB testing blog - or what does what

  1. #1
    WMD Member
    Join Date
    May 2015
    Location
    G'day mate
    Posts
    862
    Platform
    PC

    My FFB testing blog - or what does what

    I'm loving Project Cars, finally a real sim on a console, but it certainly has given me some headaches. When it's working, it's absolutely fantastic, when it's not, well I guess you could imagine.

    I suppose the biggest change from other serious driving games (not that arcade rubbish) is the ability to fine tune/personalise how the wheel works with the game. This has left me with a very steep learning curve.

    The explanations in the menus haven't been as helpful as I'd like, and trying to find an explanation anywhere on the net has been impossible. But with lots of reading, and finding a blog in these forums somewhere (can't remember where now), have finally led me to a method of testing what bits do what, and how to change them for improvements.

    So I though I'd do a blog about my experience, and hopefully gain some knowledge about how the FFB works and can be adjusted for me. I hope this makes it clearer for those in my predicament.
    XB GT - Steel Jockey. Steam GT - STEELJOCKEY.Project CARS 2 GOLD Member
    FB AUS/NZ Combined Racing League YouTube Mrsteeljockey
    Xbox Scorpio,TS-XW, Sparco P310 rim, Protosimtech PT2 pedals, racing cockpit, Onkyo 7.1, Panasonic 55"
    Gigabyte Z390 M Gaming, i9-9900k, 32gb DDR4-2133, Aorus GTX 1080Ti Xtreme, Samsung SSD 850 EVO 500gb, WD Blue 2tb HDD, Oculus Rift S
    The following 8 users likes this Post: 911 C4S, gruzzlebeard, joking_aura, KkDrummer, nissan4ever, Outlaw Peet, poirqc, Schnizz58


  2. #2
    WMD Member
    Join Date
    May 2015
    Location
    G'day mate
    Posts
    862
    Platform
    PC
    Day 1

    So, to start, I'm using Thrustmaster's TX 458 Italia wheel. What I'm doing I would think would be the same for other wheels with FFB, results might be slightly different on other wheels.

    Following the testing blog I found here (maybe someone else has seen it and can link it in), reset my wheel, re calibrated, grabbed a Formula B, reset it's FFB setting, and went for a spin on Brands Hatch GP. After the first couple of attempts, I went back and turned off damage. Hitting walls on the first lap and bending the car weren't going to help me test things. I needed nearly a lap of taking it easy to start warming up the tyres and get some grip.

    NOTE:- when changing settings in the wheel FFB area, just backing out is slow, hitting the menu button on the wheel, going down to options and hitting A to take me back out is much quicker. Otherwise I'd be here all day trying to get out of the wheel menu instead of testing settings.

    To me, the wheel felt vague around the middle, not a lot of reaction to my initial input on turn in, and no feeling of straight line stability. Soft, you might say.

    So I was guided to look into the wheel deadzone area.

    So what I've done, is set dead zone to zero, tested dead zone removal range starting from zero and increasing it. By increasing the range I got a better feeling around the middle and better response on turn in. But as I increased the rangep the wheel starts to get notchy/sticky around the centre. Too much and on straights the wheel jumps from side to side. So I started to increase wheel deadzone removal falloff. This seems to smooth the centre spot. It seems to need to be about half what the range is. Ie if range is 0.05, falloff needs to be 0.025. I'm currently running range at 0.1, and falloff at 0.055-0.06. This makes the wheel tighter around the middle, giving good feel, but not sticky or notchy at the centre. Settings above 0.1 on the range meant I was running out of falloff to smooth out the centre, and I found didn't give much more improvement.

    One thing I did notice was as I increased range, when I started loading a practice session, and the menus loaded, the wheel would spin full lock. The more range I added, the faster this would happen. As I was adding falloff, this wheel spinning would lessen. The balance I found between range and falloff for driving seemed to just stop the wheel spinning as the menus started, so this gave me a good idea the balance was right before I tested out the settings. Not sure if this happens on all platforms though.

    So far, happy with the wheel, but more improve to come, and understanding too I hope.

    To me, the deadzone area is a bit like looking at a canyon at the centre of the steering wheel's rotation. The deadzone is the canyon width, the range being the the height of the canyon wall, that is at the bottom no strength in the forces and the top have force, the greater the height the stronger the force. The falloff is the slope of the walls. A steep wall makes a big change between no force and lots of force, a shallow wall smoothing out the forces from one side of the valley to the next.

    Summary
    Deadzone 0
    Deadzone removal range 0.1
    Deadzone removal falloff 0.055

    So, onto testing the next bits.
    Last edited by STEELJOCKEY; 11-07-2015 at 07:55. Reason: Corrected deadzone names to match game menus
    XB GT - Steel Jockey. Steam GT - STEELJOCKEY.Project CARS 2 GOLD Member
    FB AUS/NZ Combined Racing League YouTube Mrsteeljockey
    Xbox Scorpio,TS-XW, Sparco P310 rim, Protosimtech PT2 pedals, racing cockpit, Onkyo 7.1, Panasonic 55"
    Gigabyte Z390 M Gaming, i9-9900k, 32gb DDR4-2133, Aorus GTX 1080Ti Xtreme, Samsung SSD 850 EVO 500gb, WD Blue 2tb HDD, Oculus Rift S
    The following 6 users likes this Post: aleph99, Apoc112, gruzzlebeard, kingleer, poirqc, Schnizz58


  3. #3
    GT3 Pilot Schnizz58's Avatar
    Join Date
    May 2015
    Location
    The Duchy of Grand Fenwick
    Posts
    4,449
    Platform
    PC
    Outstanding! Looking forward to future installments.
    97.8% of mass public shootings in the US between 1950 and 2018 occurred in so-called "gun-free zones".
    ASUS GL702VI-MH72 | Thrustmaster TX Wheel | Thrustmaster VG Ferrari 599XX EVO Wheel Alcantara Edition
    Thrustmaster T3PA Pro Pedals with Ricmotech Load Cell Mod | Wheelstand Pro v2

  4. #4
    WMD Member
    Join Date
    May 2015
    Location
    G'day mate
    Posts
    862
    Platform
    PC
    Day 2

    There are so many bits to this FFB system, time to decipher them properly.

    Restarted, setting deadzone at zero, deadzone removal range and falloff to zero, all relative adjust settings to zero, scoop and soft clipping to zero, and start again.

    After that, I've gone back to those deadzone settings from day 1, removal at 0.1 and falloff at 0.055.

    Now to look at soft clipping, as I can see in the telemetry HUD the signal appears to be clipping (bit hard to read with line only using the top half of the box).

    The wheel feels pretty good, apart from the clipping. Turn in is nice, but signal just settles to something a bit wooden. Must be running out of FFB signal, so that's the clipping?

    So I lift soft clipping half input to 0.5, and now I can see the wheel is not clipping at all anymore, feels better. I'll try 1.0, wheel feels so much softer, I'm not getting the feed back I'm looking for. Back to 0.5, what does 0.1 do? The wheel setting doesn't go back down to zero unless I reset and recalibrate, start again, so 0.1 it is for testing.

    Yuk, FFB much stronger, but no depth the minute you turn off centre. Also the wheel notch is back. FFB in the meter doesn't appear to clip, but the line likes to sit either near the top or bottom depending on which way I'm turning.

    Looks like 0.5 for soft clipping it is.

    One thing I have noticed is the wheel is back to spinning as I enter the menu, small soft clipping half input giving the wheel a good jolt, larger numbers giving a slower spin. I'm sure this will sort itself out as I get everything together.

    Ok, so this setting looks like it is compressing the signal into something useful on this wheel, stopping the clipping and losing feeling. The HUD shows this too, but I can see by the wavy line, there is room to get more FFB.
    Last edited by STEELJOCKEY; 09-07-2015 at 05:42.
    XB GT - Steel Jockey. Steam GT - STEELJOCKEY.Project CARS 2 GOLD Member
    FB AUS/NZ Combined Racing League YouTube Mrsteeljockey
    Xbox Scorpio,TS-XW, Sparco P310 rim, Protosimtech PT2 pedals, racing cockpit, Onkyo 7.1, Panasonic 55"
    Gigabyte Z390 M Gaming, i9-9900k, 32gb DDR4-2133, Aorus GTX 1080Ti Xtreme, Samsung SSD 850 EVO 500gb, WD Blue 2tb HDD, Oculus Rift S

  5. #5
    WMD Member
    Join Date
    May 2015
    Location
    G'day mate
    Posts
    862
    Platform
    PC
    On to soft clipping full output.

    Going from 0.0 to 0.5 makes a noticeable difference, wheel feels really strong, but I can see clipping is back, 1.0 is much better, still a bit strong, and I can see in the big right sweeper on the front straight when the left side is loaded up that the FFB line just sits at the top. 1.5 keeps it just under, I can see definition in the FFB line but it's not clipping but using pretty much all useful forces. 2.0 and the wheel is starting to get soft again.

    Alright, so 1.5 for the full output. FFB forces have been boosted but still feel useful, signal is not clipping, at this track and with this car at least, but I can feel turn in, little bumps, and such.

    So the clipping half input feature appears to be compressing the signal to stop clipping, while full output boosts it back to using as much signal as you can before clipping again, fine tuning the top bit.

    Summary
    Deadzone 0
    Deadzone removal range 0.1
    Deadzone removal falloff 0.055
    Soft clipping (half input) 0.5
    Soft clipping (full output) 1.5

    Relative adjust and scoop are currently 0
    Last edited by STEELJOCKEY; 10-07-2015 at 02:36.
    XB GT - Steel Jockey. Steam GT - STEELJOCKEY.Project CARS 2 GOLD Member
    FB AUS/NZ Combined Racing League YouTube Mrsteeljockey
    Xbox Scorpio,TS-XW, Sparco P310 rim, Protosimtech PT2 pedals, racing cockpit, Onkyo 7.1, Panasonic 55"
    Gigabyte Z390 M Gaming, i9-9900k, 32gb DDR4-2133, Aorus GTX 1080Ti Xtreme, Samsung SSD 850 EVO 500gb, WD Blue 2tb HDD, Oculus Rift S

  6. #6
    GT3 Pilot Schnizz58's Avatar
    Join Date
    May 2015
    Location
    The Duchy of Grand Fenwick
    Posts
    4,449
    Platform
    PC
    Quote Originally Posted by STEELJOCKEY View Post
    So the clipping half input feature appears to be compressing the signal to stop clipping, while full output boosts it back to using as much signal as you can before clipping again, fine tuning the top bit.
    That's not exactly right. The half input is the value that will become 0.5 (on a scale of 0.0 to 1.0, where anything above 1.0 is clipping) after the compression. The full input is the value that will become 1.0 after compression. So if you set them to 0.5 and 1.0 respectively, it will change nothing. If you set them to 0.5 and 1.5, low level signals below 0.5 will not be compressed at all. Signals in the 0.5 to 1.5 range will be compressed depending on the signal level with 1.5 being compressed to 1.0. Signals above 1.5 will still be clipped. They are not independent of each other, you have to set them in conjunction.

    Incidentally, I just figured out why soft clipping causes the "notchy" feeling. The UI bug (fixed in 2.0) that prevents SC (half) from being set to 0.0 causes it, sort of. The lowest value you can enter is 0.1. That means you'll actually be amplifying low level signals rather than compressing them. A signal of 0.1 will become 0.5 and so what should be weak FFB signals when you're mostly driving straight will get multiplied by 5 and thus become much stronger. If the wheel is perfectly straight you won't get anything but the moment you turn it even slightly you get a very strong FFB signal. So until 2.0 comes out the way to disable soft clipping is by setting half to 0.5 and full to 1.0.
    97.8% of mass public shootings in the US between 1950 and 2018 occurred in so-called "gun-free zones".
    ASUS GL702VI-MH72 | Thrustmaster TX Wheel | Thrustmaster VG Ferrari 599XX EVO Wheel Alcantara Edition
    Thrustmaster T3PA Pro Pedals with Ricmotech Load Cell Mod | Wheelstand Pro v2
    The following user likes this Post: poirqc


  7. #7
    WMD Member
    Join Date
    May 2015
    Location
    G'day mate
    Posts
    862
    Platform
    PC
    Hmmm, still trying to get my head around what it all means, and what I'm feeling or seeing changing as I change things.

    That bug with only being able to drop half clipping back to 0.1 makes sense. When it was zero when wheel is first reset/calibrated, it's not there, but minute you lift the slider, you can't get it back to zero, only 0.1, unless you completely reset wheel again.

    As there have been no simple explanations of what all this stuff does, I've originally just been following the blind and applying other people's settings. Now I'm trying to understand it, and hopefully put it simply enough maybe others can too. If what I'm explaining is not right, it's because I still don't fully understand what the bits mean or are doing. These are just my observations of trail and error, and trying to work out what changes what.

    Maybe some of these things will be cleared up in 2.0, and I'll be back to the drawing board, until then I'll carry on with 1.4. What is becoming obvious though is that only a small window of settings appear to be useful, so having a full range is really a waste of time, and quite confusing. Maybe a smaller window of adjustments could have been available, or if everyone comes to the same conclusion that just one set becomes useable, maybe the game should have been simpler just to have wheels calibrated with that one set. It probably would have been easier for the masses. Something to look at maybe for PC2.
    XB GT - Steel Jockey. Steam GT - STEELJOCKEY.Project CARS 2 GOLD Member
    FB AUS/NZ Combined Racing League YouTube Mrsteeljockey
    Xbox Scorpio,TS-XW, Sparco P310 rim, Protosimtech PT2 pedals, racing cockpit, Onkyo 7.1, Panasonic 55"
    Gigabyte Z390 M Gaming, i9-9900k, 32gb DDR4-2133, Aorus GTX 1080Ti Xtreme, Samsung SSD 850 EVO 500gb, WD Blue 2tb HDD, Oculus Rift S

  8. #8
    GT3 Pilot Schnizz58's Avatar
    Join Date
    May 2015
    Location
    The Duchy of Grand Fenwick
    Posts
    4,449
    Platform
    PC
    By all means carry on. I don't think it will be wasted effort following 2.0.

    As for the range of values, only a small range of values appear to be useful for a TX wheel. But there are other wheels out there that might benefit from values that would be ludicrous for our wheels.
    97.8% of mass public shootings in the US between 1950 and 2018 occurred in so-called "gun-free zones".
    ASUS GL702VI-MH72 | Thrustmaster TX Wheel | Thrustmaster VG Ferrari 599XX EVO Wheel Alcantara Edition
    Thrustmaster T3PA Pro Pedals with Ricmotech Load Cell Mod | Wheelstand Pro v2

  9. #9
    WMD Member
    Join Date
    May 2015
    Location
    G'day mate
    Posts
    862
    Platform
    PC
    True, and a different wheel might need its own set, and as each wheel is separately supported and recognised, their own setup could be applied. Simple, and no need to have a million adjustments. Don't get me wrong, I love the option to personalise feeling for my wheel, but if everyone with this wheel comes to the same conclusion that one setup works, and a similar situation arises with other brand wheels, maybe this whole thing can be simplified.

    But I digress from what I'm attempting here - to understand what all these settings mean and do, and how to apply them. Whether these same settings work for other wheels is another discussion, but maybe the method works, certainly the understanding what it's all about can help all around, as I've yet to find the hidden manual.
    Last edited by STEELJOCKEY; 09-07-2015 at 06:02.
    XB GT - Steel Jockey. Steam GT - STEELJOCKEY.Project CARS 2 GOLD Member
    FB AUS/NZ Combined Racing League YouTube Mrsteeljockey
    Xbox Scorpio,TS-XW, Sparco P310 rim, Protosimtech PT2 pedals, racing cockpit, Onkyo 7.1, Panasonic 55"
    Gigabyte Z390 M Gaming, i9-9900k, 32gb DDR4-2133, Aorus GTX 1080Ti Xtreme, Samsung SSD 850 EVO 500gb, WD Blue 2tb HDD, Oculus Rift S
    The following user likes this Post: ashasha


  10. #10
    GT5 Pilot inthebagbud's Avatar
    Join Date
    May 2015
    Location
    all over the place like my driving!
    Posts
    1,912
    Platform
    PC
    Quote Originally Posted by Schnizz58 View Post
    That's not exactly right. The half input is the value that will become 0.5 (on a scale of 0.0 to 1.0, where anything above 1.0 is clipping) after the compression. The full input is the value that will become 1.0 after compression. So if you set them to 0.5 and 1.0 respectively, it will change nothing. If you set them to 0.5 and 1.5, low level signals below 0.5 will not be compressed at all. Signals in the 0.5 to 1.5 range will be compressed depending on the signal level with 1.5 being compressed to 1.0. Signals above 1.5 will still be clipped. They are not independent of each other, you have to set them in conjunction.

    Incidentally, I just figured out why soft clipping causes the "notchy" feeling. The UI bug (fixed in 2.0) that prevents SC (half) from being set to 0.0 causes it, sort of. The lowest value you can enter is 0.1. That means you'll actually be amplifying low level signals rather than compressing them. A signal of 0.1 will become 0.5 and so what should be weak FFB signals when you're mostly driving straight will get multiplied by 5 and thus become much stronger. If the wheel is perfectly straight you won't get anything but the moment you turn it even slightly you get a very strong FFB signal. So until 2.0 comes out the way to disable soft clipping is by setting half to 0.5 and full to 1.0.
    Schnizz58 - I actually understand that concept now, needs putting in the manual - Thankyou

    Keeping in the easy speak tone, when you talk about compressing what does that actually mean - is it reducing the FFB feel by (softening) compressing or is the FFB feel actually increased as it is compressing a larger signal into a smaller area and it's fighting to get out (sorry trying to think of an analogy - so not sure if this comes across as to what I mean)

    Sorry to jump in as I do not want to fill the thread with needless talk as it has been informative and I am enjoying reading the posts which are in a language I actually understand
    CPU:i7 8700K | RAM:Corsair 32GB 3000MHz | GPU:Asus ROG GTX1080 Ti 11GB | Board:Asus Rog Strix Z370-E | Win 10 x64 |
    SEN:AUT | FF:to suit | LIN:On | NDP:50 | NFR:Off | NIN:50 | INT:6 | FEI:100 | Rest :100
    Sim Lab P1 Rig| Fanatec DD2, Wheel Rim GT, Pedals V3 ,Shifter SQ

Similar Threads

  1. Requesting Blog post manual on setup and explanation of Force Feedback
    By dresified in forum PC - Technical Help & Support
    Replies: 0
    Last Post: 16-06-2015, 21:27
  2. TESTING SETUPS
    By emessem in forum Project CARS on Playstation 4
    Replies: 2
    Last Post: 31-05-2015, 12:45
  3. In-game Blog?
    By Arkymedes in forum Feature Suggestions
    Replies: 4
    Last Post: 28-05-2015, 13:16
  4. Why is testing done on a separate forum?
    By Francorchamps in forum General Discussion
    Replies: 6
    Last Post: 23-05-2015, 11:39
  5. After 11 days testing and testing...
    By onepueblo in forum Project CARS on Playstation 4
    Replies: 4
    Last Post: 18-05-2015, 07:10

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •