AC Modding Questions Thread

[Cimmerian_Iter]

that's hierarchy
yours does look different than the screenshot below (which i believe is from the PDF guide from kunos ?)

so basically, (not sure why you have 2 "face") but it needs to be under (child of) driver:driver, just like the bottom one
in other words, driver:driver is the parent
View attachment 595570

No the screenshot below is what I get in ks editor.

I have 2 face because that's how Blender works, I just have only one object.
I will try what you said.



On a side note, I have my two Driver model (2010 and 2011 helmet) that works on my Ferrari F10.

But when I try to load the driver model in my F150 (and it shows in CM showroom that it's the 2011 driver model) it reverts back to kunos driver model in game. Do you know why?

 

[Ben O'Bro]

in the data of the cars, in driver3d.ini, you choose which driver model you use (driver_no_HANS for example)
Maybe that's the reason, the F40 may use a different type than your f1 ?

 

[Cimmerian_Iter]

in the data of the cars, in driver3d.ini, you choose which driver model you use (driver_no_HANS for example)
Maybe that's the reason, the F40 may use a different type than your f1 ?

No it's driver_FVA11 like for my F10, I even swapped the F150 driver3D.ini with the one I use with my F10 which is why I don't understand. CM Showroom actually show the FVA11 driver in the car when I load the driver

 

[Cimmerian_Iter]

that's hierarchy
yours does look different than the screenshot below (which i believe is from the PDF guide from kunos ?)

so basically, (not sure why you have 2 "face") but it needs to be under (child of) driver:driver, just like the bottom one
in other words, driver:driver is the parent
View attachment 595570

Ok Blender keep the link of the parent inside the object, while not being under DRIVERRIVER, by changing it I broke the link.

I've done it by "replacing the object" and moving it and now it works, thanks

 

[Cimmerian_Iter]

Ok one last question, I have mapped and all the helmet like kunos one, but now the 3D model of the helmet is "weird"

Before it was all round

Now we can kind of see the mapping

How can i get ride of that and get a very round helmet?

Thanks

 

[AccAkut]

Ok one last question, I have mapped and all the helmet like kunos one, but now the 3D model of the helmet is "weird"

Before it was all round
View attachment 595900


Now we can kind of see the mapping
View attachment 595901
How can i get ride of that and get a very round helmet?

Thanks

you somehow got some vertices set to sharp Normals I'd guess. Try setting the whole thing to smooth again

 

[Cimmerian_Iter]

you somehow got some vertices set to sharp Normals I'd guess. Try setting the whole thing to smooth again

Do you know what's the option in blender?

 

[EttieApple]

I am trying to create an AI line for my point to point track, but the fastlane.ai file does not get saved inside the track folder after i complete the course and close Assetto corsa with esc>exit. Did anyone have this before?

It is probably an issue with my track, since I am able to record AI splines for other tracks.

Clicking "save AI" does create a payloads file, the pitlane.ai also gets saved normally. The track has a start and finish line working, pits and starting positions set up, physics mesh with invalid track sections set up.

 

[AccAkut]

Do you know what's the option in blender?

Object -> Shade Smooth. or in Edit Mode it's under Mesh - Shading - Smooth Faces

 

[Ryno917]

Ok one last question, I have mapped and all the helmet like kunos one, but now the 3D model of the helmet is "weird"

Before it was all round
View attachment 595900


Now we can kind of see the mapping
View attachment 595901
How can i get ride of that and get a very round helmet?

Thanks

I'm guessing something's up with the FBX export; I assume those seams are where the UV seams are, yeah? Might be splitting the edges at those seams on export. I can't recall the fix off the top of my head, but some googling around UV seams breaking normals might bear some fruit, or lead you in a promising direction.

 

[Cimmerian_Iter]

Object -> Shade Smooth. or in Edit Mode it's under Mesh - Shading - Smooth Faces

Did not work

I'm guessing something's up with the FBX export; I assume those seams are where the UV seams are, yeah? Might be splitting the edges at those seams on export. I can't recall the fix off the top of my head, but some googling around UV seams breaking normals might bear some fruit, or lead you in a promising direction.

"
The results are correct. The reason it appears wrong is because the normal map is using tangent space (which is what you usually want) - tangent direction is based on the UV map, so there will usually be a difference in colour where there is a UV seam.

However the colour difference is not an artefact. When used as a normal map (not as a colour map), everything will display correctly:"

That's an explanation I found about a user that had the exact same problem as me. But I don't really find the solution. Like ok I understand the problem behind it, but I don't really know what to do to fix it

Edit : Ok so I double checked when the helmet seams appears, when I export in collada the helmet from 3DSIMED and import it in Blender, it already shows the seams, and there's no normal map applied yet. And I don't think 3DSIMED broke it since when I load the collada back in 3DSIMED the helmet looks fine, but when I export the fbx from blender it's not round anymore


Edit 2 : Okay, so with 3Dsimed, I converted the helmet into kn5 and applied the same materiel (ksmultimap with same texture as before) and the line effect doesn't appear. Even with the normal map applied

EDIT 3 : I tried unpacking the kn5 to fbx with showroom, and imported it in blender.

Great it doesn't mess with the seams anymore

And then I exported it as fbx then kn5 with ks editor and again those lines are there...

EDIT 5 : Ok it's when I apply the normal map again. But then I don't understand why the first helmet (directly converted from 3Dsimed to kn5) has no normal seams

 

[Fat-Alfie]

I am trying to create an AI line for my point to point track, but the fastlane.ai file does not get saved inside the track folder after i complete the course and close Assetto corsa with esc>exit. Did anyone have this before?

It is probably an issue with my track, since I am able to record AI splines for other tracks.

Clicking "save AI" does create a payloads file, the pitlane.ai also gets saved normally. The track has a start and finish line working, pits and starting positions set up, physics mesh with invalid track sections set up.

Did you set up your AC timing objects correctly for a P2P track?

 

[EttieApple]

Did you set up your AC timing objects correctly for a P2P track?

These are my objects, placed slightly above the track surface, on the rights sides. Timing also works correctly in AC. The laps are not invalidated, I made sure not to even touch the illegal surfaces.

When I go back to the pits after a lap, there is a purple line of my previous lap. Should I also directly see the track limits maybe?When closing AC, the program also freezes for a few seconds indicating that something is happening. Setting another lap directly afterwards does not help.

EDIT SOLVED:
my physical track mesh was quite untidy due to a lot of boolean operations for carving road lines and track borders. Going back to a clean version of the physical mesh immediately solved the problem

 

[Stereo]

Did not work

"
The results are correct. The reason it appears wrong is because the normal map is using tangent space (which is what you usually want) - tangent direction is based on the UV map, so there will usually be a difference in colour where there is a UV seam.

However the colour difference is not an artefact. When used as a normal map (not as a colour map), everything will display correctly:"

That's an explanation I found about a user that had the exact same problem as me. But I don't really find the solution. Like ok I understand the problem behind it, but I don't really know what to do to fix it

Not the most clear description but the issue they're describing is that if the normalmap isn't "flat", then at any seam in the UV map, it'll be angling the normals slightly different directions. The gotcha here is that there is no such thing as a flat normalmap in most of AC's shaders. The closest you can get is 127,127,255 and one of the "flat_NM.dds" files provided with the game is not even this close, it's 130,128,255 or something.

Usually that would show up as a small reflection error though, like in your more recent post of "applying the normal map". The earlier pics where the actual shading has a sharp edge would be something wrong in the mesh. It is worth double checking that your normalmap is that 127,127,255 colour though. If it's far enough off then it could cause pretty noticeable problems. If you can avoid using a shader with a normalmap then that's probably a good idea.

Looking at the Kunos helmets some of them are using object space normalmaps, if you've got one of those then you probably just need to get rid of it if you've edited the mesh/uvs at all.

 

[Cimmerian_Iter]

Not the most clear description but the issue they're describing is that if the normalmap isn't "flat", then at any seam in the UV map, it'll be angling the normals slightly different directions. The gotcha here is that there is no such thing as a flat normalmap in most of AC's shaders. The closest you can get is 127,127,255 and one of the "flat_NM.dds" files provided with the game is not even this close, it's 130,128,255 or something.

Usually that would show up as a small reflection error though, like in your more recent post of "applying the normal map". The earlier pics where the actual shading has a sharp edge would be something wrong in the mesh. It is worth double checking that your normalmap is that 127,127,255 colour though. If it's far enough off then it could cause pretty noticeable problems. If you can avoid using a shader with a normalmap then that's probably a good idea.

Looking at the Kunos helmets some of them are using object space normalmaps, if you've got one of those then you probably just need to get rid of it if you've edited the mesh/uvs at all.

Thank you I understand now, also damn flat_nm isn't flat they lied to us (also why i was wondering myself why using flat_nm still didn't solve the issue)

For now I'm avoiding normal shader until I create a very flat normal map

 

[Cimmerian_Iter]

Hey, I'm trying to do suspension animation, but apart of the empties method (that I don't particularly like) the DIR_ seems better but only available on 3DSmax with the community script, is there a way to use DIR_ on blender? Thanks

 

[Stereo]

Hey, I'm trying to do suspension animation, but apart of the empties method (that I don't particularly like) the DIR_ seems better but only available on 3DSmax with the community script, is there a way to use DIR_ on blender? Thanks

I'm not sure what the script does but the way DIR_ works you put an object, and then you put an object named DIR_object, and ingame it rotates the object so its x-axis points at DIR_object. (eg. you have a lower suspension arm named LCA_LF, and then an object named DIR_LCA_LF)

Getting it to match up with suspension means using the hierarchy of objects that Blender controls directly, which is WHEEL_LF, SUSP_LF, and DISC_LF. Defining these, WHEEL is rotating parts that are removed during pit stop tire changes, SUSP is non-rotating parts (brake caliper, suspension upright+knuckle), DISC is rotating parts that are not removed (brake disc, hub). If something attaches to the upright, you put the DIR_LCA_LF as a child of SUSP_LF.

 

[Cimmerian_Iter]

I'm not sure what the script does but the way DIR_ works you put an object, and then you put an object named DIR_object, and ingame it rotates the object so its x-axis points at DIR_object. (eg. you have a lower suspension arm named LCA_LF, and then an object named DIR_LCA_LF)

Getting it to match up with suspension means using the hierarchy of objects that Blender controls directly, which is WHEEL_LF, SUSP_LF, and DISC_LF. Defining these, WHEEL is rotating parts that are removed during pit stop tire changes, SUSP is non-rotating parts (brake caliper, suspension upright+knuckle), DISC is rotating parts that are not removed (brake disc, hub). If something attaches to the upright, you put the DIR_LCA_LF as a child of SUSP_LF.

Ok, so I understand about the concept of DIR_

So let's imagine that this square is a DIR_ object

If I want the Triangle_UP_RF to move according to the suspension hub, would I need to name this object DIR_Triangle_UP_RF and put it in the SUSP_RF family?

 

[Stereo]

Yes, and make sure its origin is the place you want it to pivot around. Usually I do this by placing new empties, rotating them so +X points at the DIR, and then making them the parent of the mesh.

There's no limit on which axes the empty will turn on, so if it's a control arm that rotates on an axis, you want it to specifically be placed so it's at the point on that axis closest to the DIR (thus making the axis perpendicular to the direction of the DIR)

 

[Cimmerian_Iter]

Yes, and make sure its origin is the place you want it to pivot around. Usually I do this by placing new empties, rotating them so +X points at the DIR, and then making them the parent of the mesh.

There's no limit on which axes the empty will turn on, so if it's a control arm that rotates on an axis, you want it to specifically be placed so it's at the point on that axis closest to the DIR (thus making the axis perpendicular to the direction of the DIR)

OK so here's what I have done

This is the empty that will follow the tyre up and down,

This is the DIR_ empty with y on top and x pointed toward the origin of the empty

I hope this isn't a problem (rotation settings for the DIR_ to point toward the origin)


With that in mind, this is the hierarchy

SUSP_RF -> susp1 (the empty on the first picture)

DIR_susp1 -> Triangle_rf (the suspension object)

The result ?

The suspension isn't moving still.


I know DIR_ is correctly parented to the suspension, if I rotate it through the Y axis it moves up and down