Released Caterham Academy

Discussion in 'News & Notifications' started by Paul Jeffrey, Aug 4, 2022.

  1. Cubits

    Cubits Registered

    Joined:
    Jan 22, 2012
    Messages:
    7
    Likes Received:
    1
    I was lapping the car to build some sense of the limits when i noticed the AI was massively outbraking me, which was weird. In controller configs i made sure my pedals were able to hit 100% easily then hit the track and not even the fronts would lock up.

    I then went into the setup to see how much more brake force i could add, maxxed that out, then on track the fronts BARELY left a mark under 100% braking. The balance felt good though, the car was wiggling about nicely, but it should be possible to lock them properly.

    It feels like the maximum brake force in the setup should be set as the new minimum, to give people who want to just mash their logitech pedals a good base level, and the new maximum should be significantly higher than that to accomodate for how people have their loadcell brakes set up. As is, there is braking performance left on the table and i have to alter my brake profile for this one car to minimise that loss.
     
    Last edited: Aug 12, 2022
  2. MileSeven

    MileSeven Registered

    Joined:
    Apr 1, 2014
    Messages:
    650
    Likes Received:
    244
    Thanks - not just me then!

    My real-life Caterham had a different engine with the exhaust on the passenger side. It wasn’t the most tuneful exhaust (disappointingly bland really) but you could definitely hear it better than the one in the sim that’s right next to you.
     
    RaceNut likes this.
  3. Bernat

    Bernat Registered

    Joined:
    Jun 15, 2020
    Messages:
    747
    Likes Received:
    581
    I've just tried the Caterham and it feels just right. I don't understand all the talking about the FFB. It's weaker than other cars but I think it's understandable and confirmed by real drivers. Someone said it lacked detail, my wheel is very weak and it's not very good reproducing details but since this car produces less max force than other cars I can increase the ffb multiplier and get more detail than I would get with other cars without much clipping.

    I think maybe there should be some article explaining how FFB works, not in the physics engine level but in the steering devices. I have developed a Linux app that does several FFB tests to measure wheel performance, and I have also worked improving the FFB in the Logitech Linux driver. I think I know a bit about the subject so I will try to explain it.

    The rFactor 2 physics engine calculates the forces that have to be applied to the wheel to be the most accurate possible compared to the forces that would be generated IRL. These forces are very simple, rotate the wheel left or right with a given strength. That's all, but it's being calculated and updated very very fast.

    These forces aren't directly applied to the wheel. They go through a filter that adapts the forces to the limitations of the wheel. All wheels have limitations, but some are more limited than others.

    The main limitations of any wheel is the minimum and maximum forces it can apply. Other limitations are how fast they can react to changes in those forces, and the precision and linearity of the strength.

    For the precision and linearity you can think of stairs. Precision and linearity would be about how homogeneous and small are the steps in the stairs. Tall and/or wide steps would mean low precision while small/narrow steps would mean more precision. Irregular step width or height would mean bad linearity.

    The filter rFactor applies can't do much about precision and linearity, They're just qualities of the steering wheel. But it tries to workaround the limited range of forces, the minimum and maximum, and also the max update frequency. My Logitech G29 can reproduce forces between roughly 0.2Nm and 2Nmm and with time, due to increased friction from usage, this range will get smaller.

    The min torque sets the 0 calibration point for the wheel. It's an offset added to the calculated force. For my wheel it'll be around 0.2nM. If this is not correctly adjusted, a value too low will makes us loose detail in the low end of the forces and a value too high will make the wheel shake without reason.

    Dealing with the maximum is done though the multiplier. We need to max out the forces without sending forces stronger than the wheel can handle. When the forces are higher than the wheel can reproduce the result is clipping and loosing detail in the high end of the forces. If it's set too low we're only using a small range of what the wheel can reproduce.

    The other setting, smoothing, reduces the number of updates. When the wheel can't keep up with the update rate, setting a higher smooth value can improve the FFB. It can also simulate some flexibility in the steering column that would do the same kind of smoothing.

    OK, so what?

    Setting these values correctly is important to get the best from our wheels. If you adjust these values to your liking for a car, a track or a group of cars and tracks, probably it won't work well with other cars that are very different.

    Some users with the strongest wheels are complaining about weak FFB. If they have tuned down the force in their wheels because they think at full is too much, then lighter cars will have very light FFB. The min torque setting that may seem useless for other cars might be important for the lighter cars so that no detail is lost in the low end. It's important for any car, but you'll notice the lack of detail more pronounced in the lighter cars.

    I have to tune very well these settings in my weak Logitech wheel because most racing cars have a very broad force range, much bigger and stronger than my wheel. When I've tried this car, of course, it felt very weak, but I had a lot of room because this car produces a narrower range of forces. I've upped the multiplier until I started getting some clipping and it feels really well. I'd say I can feel more detail than with other cars.

    I don't know if they use some kind of internal multiplier that adjusts the forces to tone down the heavier cars and tone up the lighter ones, it wouldn't be realistic but it would help those that don't understand how it works. In any case, having your wheel deliver as much force as it can, or just as much as you want to deal with, then increasing a bit the range for those lighter cars is a good solution because there's no perfect solution unless S397 take the realistic approach and your wheel is able to deliver brutal forces like a real wheel would. And don't forget the importance of the other settings.

    If you want to find a simile, it's like with audio. Some will complain that HI-FI recordings, or recordings from classical orchestras have a very low but also unstable volume. Some sections sound very low and need to up the volume while in some sections you need to take down the volume to not take the house down with the sound. It's because they want to use a richer and broader dynamic range for realism. It results in a better and higher fidelity recording but with bad amplifiers/speakers or in a room where the volume can't be set very high because of the acoustics or neighbors, you will never hear it as if the orchestra was in front of you. You will only see the inconveniences.

    In contrast, modern music is recorded to sound good on the radio and any cheap set of speakers. That means a lot of compression, detail is lost, dynamic range is lost, but it will sound pretty good on your car stereo, not realistic.

    The settings I've just discussed let you adapt the force feedback to any situation thanks to the implementation of said filter. If you have a very strong wheel but don't have it set to the highest force all the time, you need to tweak the forces depending on the car, that's why there's a ffb car multiplier. If you have a strong wheel set at the max level, you're fortunate, but you must realize that you should feel big differences from car to car, you just have to get used to it. If you have a weak wheel, play with the settings to have the maximum force without clipping on all cars, it won't be realistic because all cars will have barely the same force but it's all you can do with those weak wheels.
     
    Last edited: Aug 12, 2022
    Comante, mantasisg, pkelly and 2 others like this.
  4. MileSeven

    MileSeven Registered

    Joined:
    Apr 1, 2014
    Messages:
    650
    Likes Received:
    244
    Apologies for being a bit of a stuck record, but I do run my SC2 Pro at the full available headroom of 26Nm (and am still nursing a sore wrist after crashing the 7 despite having lowered the impact forces across the board in the JSON).

    And I still maintain as a long term owner and trackday driver of a real Caterham that I’m not getting the detail and feedback I was used to (that is - with realistic levels of caster at least). Maybe the characteristics of my real-world car were unusually good. It had the quick rack (but @Seven Smiles does too) but mine was raised around 4mm to influence bump steer. My chassis was a 2001 race chassis with the under-dash handbrake (fun-rumour/fact - the early Academy chassis allegedly had the under-dash handbrake instead of the tunnel-mounted one as it was believed the tunnel mounting stiffened the tunnel too much and could result in increased injuries in the event of a side impact… the under-dash version let the tunnel deform around your hips more easily. Yay! Later cars had better side impact protection and adopted full cages and - partly for production reasons I guess - moved back to the conventional handbrake position (you couldn’t reach the under-dash one in a 6-point harness anyway). As a result, the early Academy chassis were quite poor in torsional terms - they didn’t have full cages - I could jack the front of my car from the towing eye (front left corner of the chassis) and the front left wheel would be well up in the air before the front right left the ground. Later cars would lift both pretty evenly…

    Anyway, I’m still enjoying the mod - but I can’t unlearn my experiences.

     
    Havner, Travis and mixer61 like this.
  5. Bernat

    Bernat Registered

    Joined:
    Jun 15, 2020
    Messages:
    747
    Likes Received:
    581
    I think your comparison and experience is interesting. Obviously, there can be differences from your car to the car being modeled, and also Studio 397 relies on data supplied by manufacturers and teams. But this is nowhere near the more dramatic reports that the car had no FFB or that it could be driven with one finger. I certainly don't see that when using the weakest wheel in the market and neither are you.

    That's why I thought the wheel setup might play a role here. If we add to the equation that everyone likes to setup the wheel to their liking, and that everyone's expectations are different and sometimes very far from real life, the combination of possible outcomes increases substantially.
     
    Last edited: Aug 12, 2022
    atomed likes this.
  6. RaceNut

    RaceNut Registered

    Joined:
    May 21, 2013
    Messages:
    659
    Likes Received:
    539
    And yet, the Caterham ffb could simply not be as the dev's intended, just as I'm sure they intended it to have front wheels and fenders (from driver view) upon release.

    From my perspective, the ffb makes no sense to be that weak. There is not another car in rF2 that has such weak ffb. By comparison, it is about half the strength as cars with the weakest ffb (Mini / Radicals (not the GT3, though). I can't think of any car I've ever driven, whether from the 1960's to current, power steering or not, that has so little steering resistance.

    One possible factor in the varied reports (based on hardware), is the very substantial difference in ffb dynamic range. Perhaps that difference is progressive in how it affects the ffb signal.

    Let's wait and see if they make changes for an update.
     
    Last edited: Aug 31, 2022
    Havner likes this.
  7. Bernat

    Bernat Registered

    Joined:
    Jun 15, 2020
    Messages:
    747
    Likes Received:
    581
    Just one setting, the other settings would benefit all cars and have to be set only once for all.
     
    Last edited: Aug 12, 2022
  8. RaceNut

    RaceNut Registered

    Joined:
    May 21, 2013
    Messages:
    659
    Likes Received:
    539
    Ideally, yes. In practice, not in my experience because the effects in this case, do not seem to scale as expected. Changes to filtering / Friction / Damping are needed. That it itself, is unusual. That's another reason I think there's more at play that makes the resulting Caterham ffb strength as "unintended" by the dev's.
     
  9. Paul Jeffrey

    Paul Jeffrey

    Joined:
    May 21, 2016
    Messages:
    660
    Likes Received:
    4,336
  10. Scoopscat7

    Scoopscat7 Registered

    Joined:
    Jun 21, 2015
    Messages:
    157
    Likes Received:
    87
    • Made gauges visible at night

    My sigma engined Caterham I can hardly see the gauges at night That’s not realism :D
     
    Seven Smiles likes this.
  11. MileSeven

    MileSeven Registered

    Joined:
    Apr 1, 2014
    Messages:
    650
    Likes Received:
    244
  12. Ricky Law

    Ricky Law Registered

    Joined:
    Nov 27, 2019
    Messages:
    98
    Likes Received:
    31
    Hello Bernat. Myself, being a G29 user would love to see your FFB settings.
     
  13. Bernat

    Bernat Registered

    Joined:
    Jun 15, 2020
    Messages:
    747
    Likes Received:
    581
    Global multiplier is set to 100%, min torque is 11% and smoothing around 5.
    I don't really notice a difference with smoothing values from 0 to 10.
    Min torque was set to 9%, now to 11% since I don't notice any shaking but I may change it back to 9-10% if it shakes with some cars. Depending on the condition of your device you might want to try a slightly higher value or lower if the wheel shakes.
    For the Caterham the specific car multiplier is set to 150% so it uses the full range of the wheel.

    I've been thinking about trying a 150% multiplier for all cars, like some Logitech users seem to do. It would give more detail when reproducing weaker forces but would loose detail in the stronger ones because of heavy clipping.

    The Logitech Windows software might have some custom settings. I would try to set them to off or neutral values as much as possible so it doesn't interfere.
     
    Last edited: Aug 12, 2022
  14. GeraArg

    GeraArg Registered

    Joined:
    Oct 4, 2010
    Messages:
    604
    Likes Received:
    1,066

    Attached Files:

  15. Travis

    Travis Registered

    Joined:
    Dec 7, 2012
    Messages:
    245
    Likes Received:
    255
    Thank you to the forum users who own and provided feedback in this thread with their real world experience with Caterhams vs. the rF2 version.
     
    Last edited: Aug 13, 2022
  16. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,926
    Likes Received:
    3,876
    I wonder what "New tires" mean in the changelog. Guess will find out next week.

    Probably graphically, because next point is "stiffer sidewalls". I think we might expect 1.05 to drive slightly less drifty, and little bit more edgy, twitchy. Perhaps also tire to generate little bit more cornering force before it will slip, anyway curious to learn how it will be.
     
    mesfigas and coasting&cruising like this.
  17. LikeMike

    LikeMike Registered

    Joined:
    Dec 18, 2021
    Messages:
    14
    Likes Received:
    0
    hhmmm, just drove the caterham on a g27, and i have i guess "normal" FFB; especially over curbs.
     
  18. wrff.tiger@gmail

    wrff.tiger@gmail Registered

    Joined:
    May 14, 2021
    Messages:
    178
    Likes Received:
    237
    I was so hoping for the Caterham fix. Anyways, hopefully this was just some honest mistake and someone uploaded the wrong file or something. Regardless, I'm loving the Q3 content. Never expected to enjoy the BTCC cars as much as I do. They are fantastic. Some of the best racing in any sim I have experienced. So much fun. Patiently hoping that the Caterham gets it's wheels and really reaches it's full potential. It's really fun to drive regardless. But leaning over and looking out the side of the windshield in VR and not seeing any wheels really ruins the "immersion" as they say. I have full confidence that this will get fixed though. Regardless, kudos too Studio-397 for a great 2022 Q3 release!
     
    RaceNut and Rui Santos like this.
  19. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    Paul said on Discord a few hours ago he didn't know why it hasn't updated, as it was actually updated first on their end. So while it's not here yet, it will get here.
     
  20. Haris1977

    Haris1977 Registered

    Joined:
    Nov 25, 2019
    Messages:
    401
    Likes Received:
    195
    Still no thumbnail image for the car in steam?

    Untitled.png
     
    Balazs Magyar likes this.

Share This Page