AC Modding Questions Thread

[Ghoults]

I have questions about the aeromap feature of shaders patch

1. How do you actually use the aeromaps to add yaw sensitivity to a rear wing in a car using the map? If car has:
wheelbase 2.5m
cog 45% front, f-cog is 1.375m
cog height 0.44
wing 1.5 meters behind cog, 0.5m above

Would this be correct?

Spoiler

[MAP_0]
NAME = rear_wing
MAP_CL_RH = CL_RH.2Dlut ;cl*area versus meters (ride heights at each axle)
CL_RH.2Dlut:

       0.015 0.020 0.025
0.015    1    0.95 0.95
0.020    1.05 1    0.95
0.025    1.1  1.05 1

(wing aoa changes as front changes relative to rear. Wing height is 0.44m + 0.5m = 0.94m)

MAP_CD_RH = CD_RH.2Dlut ;cd*area versus meters (ride heights at each axle)
CD_RH.2Dlut

       0.84 0.94 1.04
0.84    1    0.95 0.95
0.94    1.05 1    0.95
1.04    1.1  1.05 1

(wing aoa changes as front changes relative to rear. Wing height is 0.44m + 0.5m = 0.94m)

MAP_BALANCE_RH = filename.2Dlut ;ratio of front:rear cl*area versus meters (ride heights at each axle)

       0.84 0.94 1.04
0.84   0    0    0
0.94   0    0    0
1.04   0    0    0

(map is one wing. 1 is at front axle, 0 is at rear? Would -0.3 work, rear wing being behind rear axle?)

MAP_YAW_CL = filename.lut ;degrees|multiplier
lut contents:
0|1
5|1
10|1
30|0.7

Do numbers need to go to 90? Asymmetric works, as such?:
-30|0.67
-10|0.98
-5|0.98
0|1
5|1
10|1
30|0.7

MAP_YAW_CD = filename.lut ;degrees|multiplier
lut contents:
0|1
5|1
10|1
30|0.7

FRONT_RH_OFFSET = 0.0 ;meters (static ride height offset - it adds to the ride height input that's sent to the 2D-LUT lookup)
REAR_RH_OFFSET = 0.0 ;same as above
(0 because earlier we set our 2d cd and cl maps for the correct wing height)

CL_MULT = 1.00
CD_MULT = 0.3
BALANCE_OFFSET = 0.00 ;%/100
INTERPOLATION = LINEAR ; LINEAR, CUBIC
DRAG_OFFSET = 0.00, 0.00 ;meters, meters - left/right, up/down
REAR_FORCE_DELAY=0.04 ;delay for forces at the rear of the car - higher numbers mean less lag
DRAG_FORCE_DELAY=0.045 ;delay for drag forces - higher numbers mean less lag

2. DRAG_FORCE_DELAY=0.045 ;delay for drag forces - higher numbers mean less lag.
How much time is 0.045, how much is 1?

 

[ben32star]

Someone knows if is there any Driver Rig template for steer animation? Specially for blender (I don't have 3d max) and preferible that isn't that Driver Rig 1.7 from Assetto Corsa forum - and neither the one from the developer pipeline-

 

[McGherkin]

What's the best way to simulate an ATB style diff? I know there's the usual power, coast, preload setup which determines locking across the axle in percentages for a typical clutch type LSD but ATBs are not that simple. Is it best to try and simulate it as an EDL which detects wheel spin and brakes that wheel by anything up to a maximum amount?

 

[Kyuubeey]

What's the best way to simulate an ATB style diff? I know there's the usual power, coast, preload setup which determines locking across the axle in percentages for a typical clutch type LSD but ATBs are not that simple. Is it best to try and simulate it as an EDL which detects wheel spin and brakes that wheel by anything up to a maximum amount?

Set PRELOAD to 0 and use POWER and COAST appropriate to your TBR per direction. It's not 100% proper because it's in rather than out torque based, but you won't notice the difference in typical use.

I've talked to JPG about it but there is no useful data out there. I wonder if it can simply be done as a torque bias ratio thing and that's it; fully linear TBR.

 

[McGherkin]

So how would you intepret, for example, a 2.5:1 TBR in that figure? From what I can tell the amount of locking is variable so it could be anything from 0 to 1.00 depending on the load from the spinning wheel and the total torque input.

Even worse the diff I want to simulate is a Wavetrac so it effectively has some dynamic preload too.

 

[Kyuubeey]

So how would you intepret, for example, a 2.5:1 TBR in that figure? From what I can tell the amount of locking is variable so it could be anything from 0 to 1.00 depending on the load from the spinning wheel and the total torque input.

Even worse the diff I want to simulate is a Wavetrac so it effectively has some dynamic preload too.

2.5TBR is 0.43~ so 0.345 after break-in. You generally won't actually be able to tell the difference from a genuine one, although the more complex ones like yours might be a problem. Helicals also won't lock 1.0 ever, it'd need an infinite TBR.

If you can provide some data, JPG might be able to make an implementation.

 

[McGherkin]

How do you get to 0.43!? I don't doubt your answer but could do with seeing your workings!

Trying to get data on a wavetrac diff is very difficult, but whatever I can do to help simulate one accurately I'd be happy to help with.

 

[Kyuubeey]

How do you get to 0.43!? I don't doubt your answer but could do with seeing your workings!

Trying to get data on a wavetrac diff is very difficult, but whatever I can do to help simulate one accurately I'd be happy to help with.

The image is from Gleason's 1988 paper so I don't exactly doubt its accuracy. They kind of made the Torsen.

 

[McGherkin]

Awesome, cheers for that. About tallies up with the numbers I've been experimenting with. When I asked initially I was wondering if anyone had done anything fancy using the EDL parameters with wheel speed etc. but that should do for now I guess.

 

[Kyuubeey]

Awesome, cheers for that. About tallies up with the numbers I've been experimenting with. When I asked initially I was wondering if anyone had done anything fancy using the EDL parameters with wheel speed etc. but that should do for now I guess.

You can't use EDL for that, there is no L/R inputs.

Cphys did add some LSD controllers, though. Actually, I forgot about them for a long time, I could maybe use it to improve my R34 too.

AXLES_DIFFERENCE_ABSOLUTE
AXLES_DIFFERENCE_RELATIVE
AXLES_DIFFERENCE_SIGNED
FRONT_AXLE_DIFFERENCE_ABSOLUTE
FRONT_AXLE_DIFFERENCE_RELATIVE
FRONT_AXLE_DIFFERENCE_SIGNED
REAR_AXLE_DIFFERENCE_ABSOLUTE
REAR_AXLE_DIFFERENCE_RELATIVE
REAR_AXLE_DIFFERENCE_SIGNED

In rad/s

No documentation, I do not know how they work. You can figure it out. It is kind of in the name. I assume they are controller inputs for the AC inbuilt controllers. Requires Cphys.

This is what Abbo told me when we talked whenever it was:

"Well those output were known in AC for several calculations and, not latest, for AWD2 but there weren't input for us modders...
Signed gives a - or a + is delta rear/right is smaller/bigger than front/left... or the opposite... I don't remember now. I use ABS mainly because normally its soymmetrycal..."

They work on rad/s though, and not torque, but it is something. I don't know if torque ever got added.

Btw, if you want to make a V-LSD with these, don't, there is an actual implementation via Cphys.

VISCOUS_FACTOR_1=
VISCOUS_FACTOR_2=
VISCOUS_FACTOR_EXP_K=
VISCOUS_FACTOR_EXP=

It works in parallel with POWER/COAST/PRELOAD ie: add the open-diff friction locking of 5-10% as well.

Refer to my pipeline and tools for a solver for it in Car Analyzer.

 

[McGherkin]

Awesome, I'll get fiddling. Thanks for the help.

 

[p3rtan]

Hello everyone. I'm pretty new to AC modding and I'm mostly doing some personal tweaks to cars at this point.
I'm using CM showroom to unpack KN5 files. Then I'm importing the FBX into Blender, doing my tweaks, exporting the FBX and packing with the KN5 packer. This seems to work well.

The only problem I have is I don't understand how the materials are stored trough this process. They seem to transfer fine from KN5 to FBX, into blender, back into KN5 and into AC.

The problem here is that I see a workflow issue. I do several trial and error exports from Blender to get results I want. The material editing seems to be done when the FBX is converted to KN5 in the KN5 converter which means I would have to either do the material changes over and over again, or keep unpacking and packing the FBX file each time.

So in short my questions are:
How are the materials stored? Is there any way to save the progress I've done in the KN5 converter while re-importing the FBX?
Is there any way to make use of the shader editor in Blender to work on the materials?
Are there perhaps any addons for blender that are useful for AC modding, the materials in particular?

Thanks!

 

[Kyuubeey]

Hello everyone. I'm pretty new to AC modding and I'm mostly doing some personal tweaks to cars at this point.
I'm using CM showroom to unpack KN5 files. Then I'm importing the FBX into Blender, doing my tweaks, exporting the FBX and packing with the KN5 packer. This seems to work well.

The only problem I have is I don't understand how the materials are stored trough this process. They seem to transfer fine from KN5 to FBX, into blender, back into KN5 and into AC.

The problem here is that I see a workflow issue. I do several trial and error exports from Blender to get results I want. The material editing seems to be done when the FBX is converted to KN5 in the KN5 converter which means I would have to either do the material changes over and over again, or keep unpacking and packing the FBX file each time.

So in short my questions are:
How are the materials stored? Is there any way to save the progress I've done in the KN5 converter while re-importing the FBX?
Is there any way to make use of the shader editor in Blender to work on the materials?
Are there perhaps any addons for blender that are useful for AC modding, the materials in particular?

Thanks!

Re-do the materials properly from the start with ksEditor, then make a persistence file to save them.

 

[p3rtan]

Re-do the materials properly from the start with ksEditor, then make a persistence file to save them.

That's very useful info! Thanks

 

[p3rtan]

Follow up question. When I save the persistence materials in KsEditor it doesn't save as a file. Is the save / load of this linked to the KN5 file name? Say for example I work on a different car and load the persistence, it will automatically differentiate between the two? Thanks

 

[Kyuubeey]

Follow up question. When I save the persistence materials in KsEditor it doesn't save as a file. Is the save / load of this linked to the KN5 file name? Say for example I work on a different car and load the persistence, it will automatically differentiate between the two? Thanks

It saves an .ini file. It's linked to the .kn5 name. Read the Pipeline.

 

[p3rtan]

Great, thanks!

 

[ohyeah2389]

How and where do people get textures for tree walls and Y trees?

 

[Fat-Alfie]

I use flat renders using SpeedTree, adjusted in Photoshop.

 

[Mitch9]

How and where do people get textures for tree walls and Y trees?

you can render custom stuff, there´s is also a nice free collection; google textures4ever vol. 8