Automobilista 2 Custom Force Feedback - Overview & Recommendations

Discussion in 'Automobilista 2 - General Discussion' started by Karsten Hvidberg, May 30, 2020.

  1. NuScorpii

    NuScorpii Well-Known Member

    Joined:
    Apr 7, 2020
    Messages:
    433
    Likes Received:
    472
    LOL, thought you didn't want to go to "the ugly place" Karsten? Yet here you are claiming your new function is superior than the Sigmoid based one, yet you clearly don't even understand how that even worked. not only that you release 3 attempts each with bugs that give a lower output than default. Again.

    If you had understood my Sigma based function you would have noticed that it actually maintains a perfect 1:1 ratio to a preset point when the sigmoid takes over. Not "almost perfect" like yours. The curves actually look quite similar up to a certain point when you have to cut yours off otherwise it gives a decrease.
     
  2. Karsten Hvidberg

    Karsten Hvidberg Well-Known Member

    Joined:
    Feb 27, 2020
    Messages:
    1,060
    Likes Received:
    1,457
    I never looked in to the Sigmoid function other than knowing how a general Sigmoid looks & the solution felt exactly like that to me, so are you sure?

    I would not have gone the custom route if it didn't(to me).

    Sorry to have worded it like that, just rushing. Not too much time.

    I would be happy to take a look at the details of it to make sure if I am mistaken.
    As said, went by feel & how they generally look & it matched. I prob. am wrong then.

    Cheers
     
    • Like Like x 1
  3. F.Santos

    F.Santos Well-Known Member AMS2 Club Member

    Joined:
    Mar 3, 2020
    Messages:
    405
    Likes Received:
    311
    Dont think so, using this one.
     

    Attached Files:

  4. Karsten Hvidberg

    Karsten Hvidberg Well-Known Member

    Joined:
    Feb 27, 2020
    Messages:
    1,060
    Likes Received:
    1,457
    oh, ok, did the last "fixed" one feel bad & can not be adjusted good?
    (post #429)
     
  5. NuScorpii

    NuScorpii Well-Known Member

    Joined:
    Apr 7, 2020
    Messages:
    433
    Likes Received:
    472
    "Felt exactly like that to me" - Still karsten shade huh? :p

    Yes, here is a comparison scaled to match the maximum:
    upload_2020-6-12_15-47-22.png

    Yours in red. The dotted green line is the linear part before the solid green sigmoid part. You seem to be aware of the other issue with your function so I won't comment on it further.
     
  6. Karsten Hvidberg

    Karsten Hvidberg Well-Known Member

    Joined:
    Feb 27, 2020
    Messages:
    1,060
    Likes Received:
    1,457
    That sounds very weird?
    I use pretty normal settings, maybe except for recon 1.
    I also use a low LFB of maybe ~10.
     
  7. Karsten Hvidberg

    Karsten Hvidberg Well-Known Member

    Joined:
    Feb 27, 2020
    Messages:
    1,060
    Likes Received:
    1,457
    Yes, it's just weird.. I must have felt wrong...
    Anyways, will plug in a new function later.

    Maybe when I pasted over the Sigmoid function I made a mistake, did not really look in to it, as I said. Just felt very weak at the low end.

    Sorry, I can see I did no better than yourself from the other day.
     
    • Like Like x 2
  8. F.Santos

    F.Santos Well-Known Member AMS2 Club Member

    Joined:
    Mar 3, 2020
    Messages:
    405
    Likes Received:
    311
    The thing is, out of the box this one was really good, easy to set up in-game and recognise what was missing in the file. So I think it took me about 5min to have it ready to play.

    The second file feels similar but weaker, which makes the pendulum sensation stronger and all cars light when in the pits.

    The third one never felt right to me, and I put some time going through the file and changing things and couldn't make it work, not better or to the level of the first file.
     
  9. WarmRed

    WarmRed Well-Known Member AMS2 Club Member

    Joined:
    Jun 30, 2019
    Messages:
    743
    Likes Received:
    393
    my settings in the game are (55/0/0) DD1 profiles selected. The AMS2 default FFB setting is therefore much more pronounced.
     
  10. Karsten Hvidberg

    Karsten Hvidberg Well-Known Member

    Joined:
    Feb 27, 2020
    Messages:
    1,060
    Likes Received:
    1,457
    Ok, in the new fixed file, did you try to "up" the center_momentum variable?
    I think it is at 1.5 in the file.
     
    • Like Like x 1
    • Informative Informative x 1
  11. Karsten Hvidberg

    Karsten Hvidberg Well-Known Member

    Joined:
    Feb 27, 2020
    Messages:
    1,060
    Likes Received:
    1,457
    Ahhh, the LFB Reiza problem.

    LFB at 0 causes an error, please try & set it to at least 1.
    It will change any custom file dramatically:)

    Also, I would use at least some FX. I have it at 50%.

    Cheers
     
    Last edited: Jun 12, 2020
    • Informative Informative x 2
  12. F.Santos

    F.Santos Well-Known Member AMS2 Club Member

    Joined:
    Mar 3, 2020
    Messages:
    405
    Likes Received:
    311
    Gonna try to turn it up to 3.0 later. Been following your conversation with Nus and what you said earlier about being a quick fix (the last file), so I'm gonna wait for your next one and probably stick with SilverBullet1.0 fo now. Thanks mate.
     
    • Like Like x 2
  13. Pixel

    Pixel Active Member

    Joined:
    Apr 25, 2020
    Messages:
    248
    Likes Received:
    103
    Tried both files and they are great. I wished Reiza allowed more than 1 named file to have multiple setups, and custom variables exposed in the UI. Great effort though, hopefully Reiza will take some queues from these, at least for some more 'standard' flavors.
     
    • Like Like x 1
  14. NuScorpii

    NuScorpii Well-Known Member

    Joined:
    Apr 7, 2020
    Messages:
    433
    Likes Received:
    472
    This seems to work to get around the lfb 0 issue:

    (<output> (if (smooth (max 0.0 (- 0.99 (abs (lfb 1.0 0.1)))) 0.5) <value if lfb 0> <value if lfb gt 0>))

    You get the odd tick if FX is set very high but it will certainly stop the bug from being noticed.
     
    • Like Like x 2
  15. Karsten Hvidberg

    Karsten Hvidberg Well-Known Member

    Joined:
    Feb 27, 2020
    Messages:
    1,060
    Likes Received:
    1,457
    Update:


    * Fixed custom ffb scaling function.

    * Adjusted turning momentum to better fit the cars, tested on 4 very different ones.

    * Adjusted Friction. Friction fills the bottom, making cars feel more real & works against oscillations.


    For a more heavy feel when driving faster: Increase "turning_momentum". For less heavy, lower it

    For less friction, if you think that is too much: Lower "friction".


    Attached custom ffb file "Silver Bullet" v4:
     

    Attached Files:

    Last edited: Jun 12, 2020
    • Like Like x 5
  16. NuScorpii

    NuScorpii Well-Known Member

    Joined:
    Apr 7, 2020
    Messages:
    433
    Likes Received:
    472
    Edit: Latest file can be found here

    ---

    Updated

    V2.1a Bug Fix:
    * Reset experimental variable that was causing cornering forces to be too high.

    Heavy Messing V2.1:

    Continuing the WR pedigree - new WR set at Jerez Standard in the Stock Car Brazil V8 using Heavy Messing V2.1:
    upload_2020-6-12_20-6-17.png

    Main aim for this release was to get a slightly more rf2 feel to the ffb but keep the details. Steering now feels tighter without any centre notchiness that doesn't move with weight shifts. Big differences between cars, especially high down force.

    * Added tyre adhesion based grip feedback. Makes the steering feel lighter when the front loses traction e.g. understeer and lock ups. Builds on the previous version for brake locking.
    * Adhesion effect strength can be controlled using adhesion_scale.
    * Revised details balance so road texture and bump details are not as harsh.
    * Increased brake effect slightly on the rack to provide better feedback on braking.
    * Added LFB 0 work around "fix". FFB won't spaz out with LFB set to 0.
    * Added a damping parameter so users can make the wheel give more resistance to turning, can also help with oscillations: overall_damping.
    * Histogram now outputs full range all the time.
    * Revised bump stop damping to decrease bounce back.
    * Increased the affect that the FX slider has on the road texture and bump details.

    *** For gear and belt driven wheels try increasing the min_force value to 0.06 or higher.
     
    Last edited: Jun 26, 2020
    • Like Like x 10
  17. F.Santos

    F.Santos Well-Known Member AMS2 Club Member

    Joined:
    Mar 3, 2020
    Messages:
    405
    Likes Received:
    311
    Enjoyed this update, messing around now to reduce the heaviness of cars like the sigma. I also added more scrub from 0.25 to 0.30 and in scale to 1.5atm, need more testing. Really good file.
     
    • Like Like x 1
  18. Karsten Hvidberg

    Karsten Hvidberg Well-Known Member

    Joined:
    Feb 27, 2020
    Messages:
    1,060
    Likes Received:
    1,457
    Deleted
     
    Last edited: Jun 29, 2020
  19. F.Santos

    F.Santos Well-Known Member AMS2 Club Member

    Joined:
    Mar 3, 2020
    Messages:
    405
    Likes Received:
    311
    Now it's back to where it was before for me, really easy to set didn't change a thing in the file only in-game. I noticed now fiddling with @NuScorpii latest file the overall strength difference. Yours has a more subtle approach to all cars even the heavy ones, you notice the difference but its not quite dramatic.

    Anyways, I guess we have another update from Reiza soon right, we are not quite there with v1.0 of AMS2. But good stuff mate, have a good problem now to choose which file to use, you guys are almost producing the same custom ffb.
     
  20. NuScorpii

    NuScorpii Well-Known Member

    Joined:
    Apr 7, 2020
    Messages:
    433
    Likes Received:
    472
    .
     
    Last edited: Jun 29, 2020

Share This Page