Adding to Cart…

Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.You currently have no notifications.
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.
Comments
Alright, back from Blender Forum.
No help, as usual (which is why I hate going there).
But in the spirit of "Push Buttons, see what happens" I GOT IT!!
I GOT IT!!I GOT IT!!I GOT IT!!I GOT IT!!I GOT IT!!I GOT IT!!I GOT IT!!
I still don't know what it is or how it happens but, here's the fix:
In Geometry Editor, IF simply flipping faces FAILS, the one two punch is:
1-ROTATE TRIANGULATION
2-Flip Normal
Using the Flip Normal worked for MOST, but not all.
For the stuck on stains, Rotate Triangulation/Flip Normals makes it go away.
(**Sometimes just the rotation will be the fix, if not add the second punch)
I really hope this helps someone else.
Thank you everyone for putting up with me and this Flipped Normals insanity.
You guys ROCK!!!
mirjagirl :
I was going to post last night after I saw your Blender wires in post #15 -- but, I was too busy.
It confirmed one of my theories, and also touches on the hypotenuse question by Hanabi and now the concave question by jestmart.
When I first saw your original wires in post #4 -- I thought these were quad-dominant meshes, but you confirmed that they are all quads, so that, along with your new wires from post #15 proved what I believed to be true.
Someone sold you a bill of sale stating only quads are good, so you believe the cookie-cutter mesh based on quad mesh nets that yield all quads is good, but they have two edges divided by a single vert on a triangle hypotenuse plane. This will create render/viewport artefacts even if the polygon is planar.
This is what jestmart was referring to as concave -- All quad polygons must be convex, and if two edges line up at 180 degrees, they can be considered concave. These are easy to deal with on an interior (just move offending vert to slightly convex), but on a border this can not be done without destroying the design.
Those polygons should actually be triangles, and yield a quad-dominant mesh.
There is nothing wrong with having tris in your mesh, and this will eliminate the artefacting issues.
If I were to guess -- I would say you converted imported image shape files to curves or mesh and used either a quad-net mesh boolean or some kind of other cookie-cutter method.
I was writing an addon in the 2.7x series that does this with the Carve Boolean function (not available in 2.8x series), and working on this issue of auto-deleting all hypotenuse/concave verts to yield clean quad-dominant meshes.
I see someone has created something similar for 2.8x series, but it was a while ago that I saw this, so I forget the name of it -- Perhaps it is what you are using?
It looks like you are creating some kind of wing (Insect? Fairy?), and looks like you are using a two-sided technique (with inverse normals for back side?) with cylindrical spines on the borders. Years ago, I created a dragonfly pendant with this technique and hand-corrected the quad-net mesh to erradicate the artefacts.
If you would like an extra pair of eyes on the mesh, I would be willing to take a look for you. If not, no worries...
DaremoK3: You'll have to excuse me if I only understand the nouns.
I did read somewhere that quads were simply two tris, and how that works with SubD. And I think your explanation of concave makes sense, I just have to re read it about 10 more times.
The wings are hand drawn, but the membranes have been giving me migraines for a while now. Specifically because of the Normals issue.
I tried using the boolean but still haven't figure out how to use it. Yes, I fail at Shrink Wrapping.
Everything is drawn, old school. The wing using the Bezier curve and a self made 'ruler' so that each segment is exactly the same length.
The Membrane started with a Plane and that same 'ruler'. Very slow and I'm sure there are much easier ways, but that was the way that I could understand.
I think I'm good on the mesh, but I would love to be able to contact you if I find myself in this position again.
I've been in this position so many times, I don't want to celebrate too loud, or too long.
I still have the other 2 wings to clean up. It's a set of 3.
Once again, I'm almost done.
Looks like I missed the excitement, but glad you got it sorted :)
Looking good in the test render :)
Your test render looks great!
Just curious why you went with double planes for the wing mesh as opposed to single-sided -- Is it to catch light better?
I like the look of the wings, and I am glad you got it to a usable state.
I am always around if you need any help, and you can PM me here if I miss a thread.
Take care...
Ken
Thank you again for everything!!!
Not really. The double sided came out of my 'research' into retopology and the issues with the faces.
Now that I have them double sided, I do like the look better.
Had everything gone well with the retopo and the faces, single sided would have been fine.
As far as 'retopology' goes... don't ask. Please don't ask.
I'd say no, but what can happen is that DS smoothes the mesh differently when there is some geometry next to that poly than when it is an outer edge. In that case inserting an edge loop can help.
Good to know.
Thank you.
Sooooo......
Triangulated Faces are just......bad. Pulling and tearing at the mesh that was not there with my single sided, pre retopo (or what I thought was retopo) membranes.
If you're still offering an extra pair of eyes, which mesh file would you like to look at?
For the Top, Mid and Bottom Wing, I have the
UV file (pre join) (3 files)
Joined (3 files)
Vertex Group (3 files)
and Positioned for DAZ (6 files)
I'm going to have a drink (or ten) and maybe cry a little.
"I dreamed a dream in time gone by,
When hope was high
And life worth living."
I thought you fixed your problems and were all set...
What issue are you having with your mesh that you would like me to look at?
I can look at your Blender (.blend) mesh file or files that contain your pre and post join meshes to check for mesh problems, then we can go from there.
PM (private message) me here, and we can set up a way for you to send me the files.
But, your post above is a little confusing -- A UV file is contained within an object file whether joined or not as long as you created the UV's. Same goes for Vertex Groups - contained in object file as long as you created them -- Both should remain upon export and written into the OBJ file..
Just for redundancy sake, I created a wing part in Blender with the technique I believe you used and crossed checked the work in Daz Studio -- Uncorrected mesh yielded the exact same results as you first shown with what you thought were flipped normals, but were infact viewport/render anomalies created by the concave quad polygons on the mesh borders.
Correcting these by either making these polygons tris, or making the offending vertex slightly convex fixed the issue.
* Technique used was as follows (may or may not be what/or part of what you did):
1. Created a Bezier curve flat plane shape
2. Converted curve object to mesh object
3. Added Solidify Modifier
4. Added Remesh Modifier
5. Applied both modifiers
6. Edit Mode -- Selected rim edges/faces and deleted them - leaving me a front and back quad meshed membrane shape (no flipping of any normals necessary)
7. Created Bezier Curves on wing shape border for spines -- added bevel, sized as needed, and converted to meshes
8. Exported as OBJ (no pre-join) with 'Selected Objects Only' checked -- OBJ file auto-merges all objects as one object
9. Imported into Daz Studio -- Came in as expected with no flipped normals or viewport issues
Images :
* DS viewport issues on mesh
* Blender mesh adding color to correspond with DS issue for testing
* Final corrected mesh with spines added (just quick work - would perfect if going to use seriously)
I will PM you, but thought an answer would also help anyone else watching this thread.
The problem was 'fixed' on the surface. However during rigging, my twist showed me the triangulation was a much deeper issue.
The files I listed are my 'Flow' / backup steps. I keep each step just in case I need to go back 1 step (or in this case 5). Less keystrokes if I don't have to disassemble for a quick fix.
The process you listed above, I'll have to try that one day.
I'll pm you.
Okay, I see the issues you are talking about -- replied to PM...