Xgen Archive Issue

[Anonymous]

Hi,

I'm trying to create some feathers on a model using xgen archives. But I'm having a problem where the archives are going in different directions. There are no triangles on the model, which seems to have solved the issue for other people having the same problem. The UVs are plain and simple all within UV space 0-1. 

Example picture below. No settings touched except for scale and density.

Another example where the same archive is attached to a sphere(with its top and bottom triangles removed). Nothing else done to the model, UVs or xgen settings.

Any idea of what might be causing this or how to solve it?

[sean.heasley]

Hi @Anonymous and welcome to the community

If I'm understanding this correctly, it looks like its working fine on a sphere but not on your custom model right?

If that's the case, you'll want to check your model and make sure there's no history on it and that it has been placed at the origin with the transforms frozen (Modify->Freeze Transformations).

Also, did you set it to distribute randomly or evenly?

Let me know if anything changes!

[Anonymous]

Yes that's right. On the sphere it is working correctly but not on my model. The model has no history, frozen transforms and placed at origo. It is also of a reasonable scale. Neither too big or small.

It is set to distribute randomly, but so it is on the sphere. Xgen settings are all at default values on both of them.

It seems to be the rotation along the surface normals that is messing up. The direction of the archives changes per face. I've check all normals of the model and everything checks out just fine.

Furthermore, I tried to apply the same xgen archive on a smoothed cube, and it gave me the same messed up result as on my model. 

I'm not at work atm so I'll try to distribute evenly tomorrow, but I can't imagine that would change anything as the sphere is also set to distribute randomly.

Thank you.

[sean.heasley]

Hi @Anonymous

Hmm ok. Out of curiosity can you try moving the gizmo/pivot to a different spot around the model and see if that changes the orientation of the feathers at all?

If that doesn't help at all, when you get a chance can you please zip and attach the scene file here or via dropbox/google drive or another file sharing program so I can take a look at it? If the file is confidential or under NDA please send it to me in a private message instead of posting it here.

[sean.heasley]

Hi @Anonymous

Just wanted to check in to see how things were going. Are you still having this issue?

If your issue is resolved, please click Accept as Solution on the posts that helped you so others in the community can find them easily.

[sean.heasley]

Hi @Anonymous

Thanks for sending me the file!

I took a look at the file and see the issue. I had to use this expression and set Generate Primitives to 'In uniform rows and columns'

 

Obviously this isn't perfect but with some custom maps I think you could get this to work.

That said, is there any reason why you aren't using cards to generate your feathers?  This tends to be the industry standard and would probably give you better results!

[sean.heasley]

Hi @Anonymous

I'm just checking in again to see if you need more help with this. Did the suggestion I provided yesterday work for you?

If so, please click Accept as Solution on the posts that helped you so others in the community can find them easily.

[henrik.linnes · Accepted solution]

Hi Sean,

Thanks for you suggestions, unfortunately they were more of a workaround than a solution. 

I managed to solve it using this expression which a colleague came across. 

$mult = 1.0000;#.0,1.0

$proj = [0,1,0] - $N[1]*$N;

$rad = angle(norm($proj),norm(-$dPdv));

$angle = deg($rad)*$mult;

$cross = cross($proj, -$dPdv);

dot($N, $cross) < 0 ? $angle : -$angle

The reason I'm not using cards is because they are too limiting. I'm creating my own feather stand-ins from xgen splines.