Daz closing polygon holes

Anyone knows what can I do to prevent this issue? When I import this 3D model, Daz automatically closes the geometry hole. 

issue 00.png
1269 x 736 - 202K
issue 01.png
983 x 875 - 94K

Comments

  • ebergerlyebergerly Posts: 3,255

    Well that's interesting...don't recall ever seeing something like that before.

    Is it possible there's actually a polygon there, but its normal is flipped, making it invisible in whatever 3D package you're exporting from, but when imported to DAZ it ignores the normal or something?

    My other thought is maybe there's something going on with the export format, maybe an export setting or something. Can you try exporting in a different format (like OBJ) and seeing if that helps? 

    I do a lot of export/import from Blender to DAZ using .obj format and have never seen anything like that. 

  • Richard HaseltineRichard Haseltine Posts: 102,252

    That's certainly possible, but I think it more likely that therea re n-gons making the wall bases - perhaps a single polygon for the whole open base, perhaps two, or perhaps a set of "arrowheads". In any event, DS and Iray/3Delight will usually try to fill in "indented" polygons, giving the appearance you see. Try triangulating that area in the modeller, or triangulate the whole thing in DS.

  • ebergerlyebergerly Posts: 3,255

    Wow, you're right. I extruded a quick cube with some "E" shaped N-gons on the base, and it tried to insert polys. Didn't quite fill the entire face, but maybe with more sides to the N-gon it might. Wow, Iray reallly hates N-gons, huh?

    NGON.JPG
    816 x 706 - 24K
  • ebergerlyebergerly Posts: 3,255

    And here's that object in Blender. You can see there's like a 12-sided N-gon forming the "E" shaped base on the right side, and DAZ tries to fill that in. Maybe that entire base in the OP's model is one big N-gon with like 40 sides or something.

    NGON1.JPG
    744 x 701 - 45K
  • ebergerly said:

    Well that's interesting...don't recall ever seeing something like that before.

    Is it possible there's actually a polygon there, but its normal is flipped, making it invisible in whatever 3D package you're exporting from, but when imported to DAZ it ignores the normal or something?

    My other thought is maybe there's something going on with the export format, maybe an export setting or something. Can you try exporting in a different format (like OBJ) and seeing if that helps? 

    I do a lot of export/import from Blender to DAZ using .obj format and have never seen anything like that. 

    No, there's not flipped normals or invisible polygon there. :( 

  • jestmartjestmart Posts: 4,449

    It is easier to diagnose mesh problems if you switch to wire shaded view.

  • BejaymacBejaymac Posts: 1,897

    Odds are it was done with Boolean, so you now have a bunch of concave polygons around the edge, and DS is generating artifacts as a result.

  • ebergerlyebergerly Posts: 3,255

    I'm not sure what is meant by "indented" and "concave" polygons. Is that referring to non-planar, where one or more vertices are not in the same plane as the others in the polygon? I'm still trying to figure what exactly is going on here. Looks like n-gons are, as expected, a problem, which is why I (and many/most other modellers) take great care to avoid them at all costs, as well as triangles, due to weird shading and other nasty effects.

    But I can't see how a boolean would cause non-planar polygins. Triangles, yeah, but non-planar? I suppose a lot also depends on the particular software doing the booleans. And I'm also not sure how a many sided n-gon could be generated in this case. Took me significant extra effort to model an n-gon in this case, since most of the procedures I (and many others) use in modelling make n-gons pretty much non-existent. 

    A real head-scratcher, IMO. I'll be interested to see what caused this.

  • Richard HaseltineRichard Haseltine Posts: 102,252

    Consider puttin an elastic band around the polygon, if the actual edges (and at least one vertex) dip inside the outline formed by the band then the polygon is concave. A chevron made up of four vertices would be an example, the band would define a triangle and there would be an empty triangle within that.

  • ebergerlyebergerly Posts: 3,255

    Okay, sounds like you're referring to non-planar. Like if you added a plane with 4 vertices into DAZ, and all the vertices were located at 0 on the Y axis. All vertices on the floor, effectively. Except, say, one of them was at Y = 0.1, so it's no longer on the floor (the same plane) like all the others. But I believe most renderers convert quads to triangles anyway to do the rendering (since triangles are, by definition, planar), so I'm not sure why concave/non-planar/indented would be an issue.

    Interesting... 

  • AscaniaAscania Posts: 1,855

    ebergerly said:

    I'm not sure what is meant by "indented" and "concave" polygons. Is that referring to non-planar, where one or more vertices are not in the same plane as the others in the polygon? I'm still trying to figure what exactly is going on here. Looks like n-gons are, as expected, a problem, which is why I (and many/most other modellers) take great care to avoid them at all costs, as well as triangles, due to weird shading and other nasty effects.

    Go on Google image search and search for "concave polygon". I has nothing to do with non-planarity and all with cavities in the outline.

  • ebergerlyebergerly Posts: 3,255

    Here's an example of non-planar...

    Added a plane to DAZ Studio as the floor, and a second plane (a square with 4 vertices) as the non-planar polygon. I used Mesh Grabber to pull one of the vertices of the non-planar up so it was, well, non-planar ("concave"?). And as you can see Iray automatically throws a diagonal edge in there to convert the 4-vertex square into two triangles, which are, by definition, planar, and all is well.  

    NonPlanar.JPG
    1023 x 663 - 67K
  • ebergerlyebergerly Posts: 3,255

    Ascania said:

    ebergerly said:

    I'm not sure what is meant by "indented" and "concave" polygons. Is that referring to non-planar, where one or more vertices are not in the same plane as the others in the polygon? I'm still trying to figure what exactly is going on here. Looks like n-gons are, as expected, a problem, which is why I (and many/most other modellers) take great care to avoid them at all costs, as well as triangles, due to weird shading and other nasty effects.

    Go on Google image search and search for "concave polygon". I has nothing to do with non-planarity and all with cavities in the outline.

    Ahh, okay. Basically we're back to an n-gon, where the sides can be concave or convex depending on the internal angle. I kinda categorize those as different n-gon shapes, and I'm still trying to understand how the different n-gon shapes come into play here, and how they could result in a clean plane covering the opening. And does the concave or convex-ness of the n-gon shapes make a difference?   

  • Hello! I solved. It was a N-gon problem. I just used "remove n-gons" option on my 3D modeling software (Cinema 4D). Thank you all! 

  • Richard HaseltineRichard Haseltine Posts: 102,252
    edited March 2021

    ebergerly said:

    Ascania said:

    ebergerly said:

    I'm not sure what is meant by "indented" and "concave" polygons. Is that referring to non-planar, where one or more vertices are not in the same plane as the others in the polygon? I'm still trying to figure what exactly is going on here. Looks like n-gons are, as expected, a problem, which is why I (and many/most other modellers) take great care to avoid them at all costs, as well as triangles, due to weird shading and other nasty effects.

    Go on Google image search and search for "concave polygon". I has nothing to do with non-planarity and all with cavities in the outline.

    Ahh, okay. Basically we're back to an n-gon, where the sides can be concave or convex depending on the internal angle. I kinda categorize those as different n-gon shapes, and I'm still trying to understand how the different n-gon shapes come into play here, and how they could result in a clean plane covering the opening. And does the concave or convex-ness of the n-gon shapes make a difference?   

    It doesn't have to be an n-gon - my chevron example was a quade. As I understand it the issue is the way that the mesh is broken into triangles. Consider my chevron: it could be split from the apex of the outer triangle to the apex of the inner, and that would give two long, thin triangles that fitted inside the desired shape; however, it could also be split between the other two vertices, across the line the "elastic band" follows while the actual shape edges dip towards the fourth vertex, and that would leave a big triangle that covered both the desired shape and the inset that should be empty, plus another triangle that covered the empty space - so that would fill in the shape, but with overlapping geoemtry which would often make the result look messy (and it would also throw the UV mapping).

    Concave quad split.jpg
    800 x 800 - 63K
    Post edited by Richard Haseltine on
Sign In or Register to comment.