[REL] Celica gto IMSA + semi-fictional track

Discussion in 'Vehicles' started by 32_Pixels, Nov 24, 2014.

  1. P.S.R.

    P.S.R. Registered

    Joined:
    Jun 15, 2014
    Messages:
    1,794
    Likes Received:
    4
    So is this first update to original or did I miss something? Jeurgen will your overall package include this? Feel like I'm missing something.
     
  2. redapg

    redapg Registered

    Joined:
    Jan 16, 2012
    Messages:
    4,027
    Likes Received:
    2,886
    This is just a small Update for the Track, with a new AIW and 3 to it matching Rubber files.
    It's meant to give you the possibility to test it and make suggestions or report bugs.

    I've created it as an Update, because many people can't or don't want to play around with loose files.

    And i thought it would be good to separate the track from the car, because each original component is a "standalone" also
     
  3. Juergen-BY

    Juergen-BY Registered

    Joined:
    Jun 16, 2012
    Messages:
    3,089
    Likes Received:
    440
    As redapg allready said, this is just for testing the AIW`s, pls give us feedback as soon as possible. At the end, we w`ll package the Track and the Car in two separate components.
     
  4. P.S.R.

    P.S.R. Registered

    Joined:
    Jun 15, 2014
    Messages:
    1,794
    Likes Received:
    4
    Tested with AC Cobra (Comp) and T280. Observations:


    • With both vehicles, AI are taking T2 way wide which is keeping them from being able to thread the needle in the subsequent esses

    • With AC I am > 2 seconds faster than AI @ 100%. With T280 I am > 1 second faster than (six fastest) AI @ 100%. Likely due to poor line mentioned above and elsewhere.
    • At least with T280, AI are swerving/taking extreme blocking on starting straight leading up to T1
    • AI had several wall contacts in both vehicles and T280 swerved left (like blocking but to outside toward tire stacks) before split where track goes to the right just before passing over train rails). Be happy to run replays at be more specific if helpful. This is what T280 incident metrics looked like...

    Pos Driver Vehicle Contacts Other Contacts Overall Contacts
    1 Charles Stoll 8 35 43
    2 Kevin Knorpp 8 26 34
    3 Piers Mike 28 28
    4 Paul Rush 24 24
    5 Alessio Romito 23 23
    6 Michael Hogson 22 22

    ...pretty messy

    Player.json:

    View attachment 17551

    Let me know if need more info or want me to test under certain conditions or different vehicles, etc...
     
  5. Juergen-BY

    Juergen-BY Registered

    Joined:
    Jun 16, 2012
    Messages:
    3,089
    Likes Received:
    440
    with standard setup? If not, please test again with standard setup, thanks.
     
  6. P.S.R.

    P.S.R. Registered

    Joined:
    Jun 15, 2014
    Messages:
    1,794
    Likes Received:
    4
    New AC PB with standard setup lol. Took more than 2 seconds of my time :p Now I'm over 5 seconds faster than 2nd position and more than 6 seconds faster than the rest. HOWEVER, it was a bloodbath...

    Pos Driver Vehicle Contacts Other Contacts Overall Contacts
    1 Piers Mike 6 29 35
    2 Michael Hogson 4 26 30
    3 Alessio Romito 4 21 25
    4 Charles Stoll 4 19 23
    5 Paul Rush 2 11 13
    6 Kevin Knorpp 8 5 13

    ... so no idea if AI even got a clean lap as they were bouncing off walls and spinning terribly, especially in the esses. And you can see their fastest lap times were in earlier laps...

    Lap K.Knor A.Romi M.Hogs P.Mike P.Rush C.Stol
    1 1:50.622 1:49.798 1:48.937 1:52.092 1:53.036 1:52.496
    2 1:53.213 1:44.396 1:43.305 1:44.865 1:51.294 1:47.432
    3 1:39.937 1:44.172 1:48.656 1:44.503 1:44.502 1:48.480
    4 1:46.133 1:48.294 1:48.255 1:52.905 1:46.122 1:48.716
    5 1:39.966 1:48.704 1:49.073 1:47.083 1:47.402 1:48.094
    6 1:37.598 1:47.751 1:45.385 1:44.936 1:44.727 1:48.259
     
  7. Marc Collins

    Marc Collins Registered

    Joined:
    Jan 11, 2012
    Messages:
    3,159
    Likes Received:
    162
    This whole update testing procedure is great--and wonderful detailed feedback from P.S.R. However, I would respectfully suggest using a car more suited to tight confines (pretty much anything other than the Cobra) for testing. I am away on business and can't do it myself.
     
  8. Juergen-BY

    Juergen-BY Registered

    Joined:
    Jun 16, 2012
    Messages:
    3,089
    Likes Received:
    440
    This comes up to my mind, too...

    From my experience i could say, not each car isn`t fitting to each track. Lester is a track, with fast turns and fast load changes, where you need cars with enough downforce to handle these turns. I did my testings with the Megane and the URD T5, wich offers a lot of racing fun ;)
     
  9. redapg

    redapg Registered

    Joined:
    Jan 16, 2012
    Messages:
    4,027
    Likes Received:
    2,886
    Like in other threads already mentioned, creating AIW files is a hard and ungrateful job. :)

    And it's almost impossible to create one that fits for every possible car.

    Just to call some influencing things: engine power, basic setup (includes gear-settings), tires, talent files, ....

    A clio takes another line than a form ula 1.

    And on a track like this, with all the walls bordering the track, things don't get easier.

    And if you are faster than the AI is, raise the AI strength. ;)
    If you are still faster if the AI strength is set to 120%, the AIW needs an overwork.

    And please, always test with the basic setup or assign your setup to the AI.

    But i will try with the Cobra too, now.
     
    Last edited by a moderator: Aug 5, 2015
  10. redapg

    redapg Registered

    Joined:
    Jan 16, 2012
    Messages:
    4,027
    Likes Received:
    2,886
    OK, the Cobra is not suitable as AIW Testcar. :)

    I could create an AIW file especially for this car with that basic setup, but that would fit for only this car then.
    I think i don't have to explain the Video.

     
  11. Juergen-BY

    Juergen-BY Registered

    Joined:
    Jun 16, 2012
    Messages:
    3,089
    Likes Received:
    440
    I may found a small gfx bug with the tires. When I`m looking from the side, while the AI are driving, it seems, that the tire bumping into the chassis and remove partially the chassis. Is it just me?

    //Edit: seems, the Tire diameter is a little bit to large. Would be fixed in the Update.
     
    Last edited by a moderator: Aug 5, 2015
  12. Juergen-BY

    Juergen-BY Registered

    Joined:
    Jun 16, 2012
    Messages:
    3,089
    Likes Received:
    440
    Guys...i need some Referenz Laptimes for the Celica on Lester, driven - not AI...please...gogogo ;)
     
  13. P.S.R.

    P.S.R. Registered

    Joined:
    Jun 15, 2014
    Messages:
    1,794
    Likes Received:
    4
    NSX-R ; default setup

    Pos Driver Time Gap Possible Lap Best laps Team Vehicle Aids
    1 Kevin Knorpp 1:47.805 - 1:45.010 6 2 NSX Type-R NSX
    2 Astro Krandle 1:47.976 0.171 1:47.894 3 2 NSX Type-R NSX AIControl
    3 Arthur D'Angelo 1:48.366 0.561 1:48.081 3 1 NSX Type-R NSX AIControl
    4 John Steinberg 1:48.531 0.726 1:48.324 6 1 NSX Type-R NSX AIControl
    5 Martin van Nostrand 1:49.725 1.919 1:48.392 2 NSX Type-R NSX AIControl
    6 Jorgan Continanza 1:49.744 1.938 1:48.407 3 NSX Type-R NSX AIControl

    Pos Driver Vehicle Contacts Other Contacts Overall Contacts
    1 Kevin Knorpp 10 11 21
    2 Arthur D'Angelo 6 9 15
    3 Astro Krandle 13 13
    4 John Steinberg 4 9 13
    5 Martin van Nostrand 6 6
    6 Jorgan Continanza 1 1
     
  14. alpha-bravo

    alpha-bravo Registered

    Joined:
    Dec 22, 2012
    Messages:
    1,374
    Likes Received:
    31
    I agree.
    Unfortunately I'm at the moment short on time. But I did yesterday some laps (with update1.01) and the APEX GT3 Mod. AI strenght 100 % AI Agression 28 % AI Limiter 3%.
    I had some fantastic laps with the AI two three things I noticed some of them I think are already mentioned:
    - AI crossed the whole track from left to right immediately after leaving the pitline. Not sure if it was the intention to block my line.
    - AI block very aggresive especially in front of the corners. It was better after I reduced the aggression value from 38% to 28 % but still noticeable.28 % is still to high ?
    - Around turn 6 I got some cut warnings but was not able to reproduce the exact conditions ^^?

    Many thanks to redapg for this AI update. I had not often so close und exciting matches with the AI :cool:
    And before this update the track was not usable for me because of the bad AI behavior.

    Keep up the good and appreciated work folks and again many thanks :)

    cu alex
     
  15. Juergen-BY

    Juergen-BY Registered

    Joined:
    Jun 16, 2012
    Messages:
    3,089
    Likes Received:
    440
    This wasn`t one version earlier, maybe this would be fixed again

    I`m using 20, wich is nearly perfect for me

    Same here, i can reproduce this cut warning nearly 100%...

    Same here ;)
     
  16. P.S.R.

    P.S.R. Registered

    Joined:
    Jun 15, 2014
    Messages:
    1,794
    Likes Received:
    4
    Haven't driven this car for a long time. Only able to do 6 laps before losing concentration. Need food. Default setup. Using clutch. 1:27.605.



    I'll try more later if I can find time...or if someone else posts a faster time ;)
     
  17. redapg

    redapg Registered

    Joined:
    Jan 16, 2012
    Messages:
    4,027
    Likes Received:
    2,886
    Thank you P.S.R and alpha-bravo for your tests and statements (also with the Cobra :)).

    @P.S.R.: can you tell me if you had a good or a bad race with the AI? Contacts on such kind of tracks happen, that's normal. It's more important to know if you did enjoy the racing.
    Or did you experience special problems, like the reported "cut track" message?

    @alpha-bravo: the immediately crossing of the track after leaving the pits, should be solved in V 1.01. The first version, i offered as loose files, did it very badly. But i will have another look at it.

    And i also will investigate the turn 6 "cut track" warning.

    The extreme blocking in front of some turns can be a result of a mixture between the lines of the Fast/Block pathes and that i tried to force the cars to pass tight turns one behind the other, by editing the corridors, to try to avoid too many "into the wall"-crashes.

    I will see what i can do and offer a new version then.
     
  18. P.S.R.

    P.S.R. Registered

    Joined:
    Jun 15, 2014
    Messages:
    1,794
    Likes Received:
    4
    FYI getting weird timing in delta best near final turn where it inaccurately shows positive delta before dropping back to acurate minus delta similar to Zandvoort.
     
  19. Juergen-BY

    Juergen-BY Registered

    Joined:
    Jun 16, 2012
    Messages:
    3,089
    Likes Received:
    440
    CelicaGTO @ Lester - 1:26.338...but tbh, with a little bit different tire :p

    ...and i really have not to much experience with the car and this track. But I`m working on my pace :D

    The (for me) difficult part is the fast S-Turn with the following right hander... :mad:
     
  20. P.S.R.

    P.S.R. Registered

    Joined:
    Jun 15, 2014
    Messages:
    1,794
    Likes Received:
    4
    TBH it's distracting when the cars make the same mistakes over and over and sliding almost exactly into the same position waiting to be T-boned. I decided to try the Celica GTO with 14 AI @100% and 20% aggression (was 25% before) and the AI are literally driving straight into walls (not even out of control but just straight into the walls and not even when having to thread needle like in esses but just the outside walls) as if on purpose and then right afterward driving straight into the next wall. Watch the Tide car and PNewman in this video...



    Edit - Drove the Palatov Track Day here and lol several of the turns are flat out. No where near as fun here as a high power, lower grip vehicle like the Celica GTO. The line is much closer to the path the AIW are following so presumably it's the age-old trade-off of low vs high grip vehicle AIW...
     
    Last edited by a moderator: Aug 6, 2015

Share This Page