What if you could set the size of the minimum area of a polygon in viewport. And the wireframe get denser when zooming in and less dense when zooming out. Of course this should work on organic models too. I just used a fast simple model to illustrate.
If you think about it. You need to zoom in anyways if you want to move a vertex anyways.
No. Not the same. The polygons won’t be less . It will only render it on the viewport as fewer polygons. You keep the high polygon data.
yeah, tessellation does not modify the base mesh, it’s just how its GLSL shaders decides to draw it, the base geometry still remains,
But the render of the tessellated mesh is actually low poly until you zoom in.
This should not be the case.
The mesh is always high poly even if it shows on viewport that it’s not. Also you should be able to determine how big the fake polygons will be in the viewport.
Forget tessellation.
what your talking about could be done if the base geometry was low poly, and you use subdivide (simple) based on the camera distance.
LOD -> could do the same,
but that is not generated in real time, and if it were it would be slow as you get above X vertex.
You’re still thinking of polygons and I’m talking about what the viewport is showing.
Rather it is more like patches. A group of polygons show as one patch. The lines still follow the several polygons. Like wrapping of fake lines over the mesh but the mesh is still in it’s original state.
BPR, this isn’t about tessellation. What Bigbad is essentially proposing is hiding edges as you zoom out so the new ‘faces’ are in reality a group of faces.
When you select the group, then you’ll find that every face within that spot was selected as soon as you zoom back in and the edges are revealed.
The face count remains completely constant, and in my opinion it actually does not sound like a bad idea at all.
but would it handle non quads?
I can imagine this being annoying as hell for ngons
It should be handled as normal. Most meshes strive to be quads. If there are ngons they will become a part of the fake quad polygon, it might even show as a tri. What is important here is the flow of the polygons when generating the fake wireframe.
Personally, I think the more user-friendly way to go about this would be for Blender to automatically draw groups of faces as one unit rather than trying to do anything that could cause Ngons to appear as quads or tris (in other words, the drawing code would simply be hiding or revealing graphics when zooming in and out while drawing a new selection dot for each set of grouped faces).
This should also be done as an optional workflow option as well, otherwise we’ll have a tidal wave of questions from users wondering if the Blender viewport code is on the fritz.
Yes. This should be an optional feature. But I thought of one more feature. You could select a vertex in zoomed out view and when you move it; it should behave as a proportional editing for the nearby vertices of the group.
This won’t happen. Sorry.
It’s a little like the optimize-mesh function on the subdivision modifier, however what you’re proposing is to temporarily store mesh data as you zoom in and then display it as a varying level of density. This has absolutely no use what so ever other than for creating viewport renders which people hardly ever use or even know of.
As a Python developer myself I’d say this is a bunch of work that’s unnecessary and could be put towards more useful development in Blender.
You have to remember that Blender has a limited development fund and core development decisions are based on whether it’s an alteration of an existing tool or a complete re-write of code, the latter of which is usually avoided due to the time and money it costs to develop versus the number of modifications and improvements that could be made to existing code.
With this in mind, it’s a nice idea but unfortunately one that’s likely to go on the pile of “Here’s a cool idea…” posts.
I was thinking this would be a good idea for overview perspective on really dense meshes. Where you could follow the flow of the polys and also alter the faces/edges/vertices with a quick hack (like auto proportional editing because it’s groups of polys).
But yes of course. It’s just a cool idea that is not necessary and would take time.
I was just thinking of what Ton was thinking about the wireframes. Because he didn’t like them. This would be like middle ground. Still wireframe but not ugly.
I know it’s not the same (it’s object mode not edit mode, not zoom related) but is a similar principle of visualization. Draw edges vs. Draw all edges. One possible intermediate step is to maybe have the same setting for vertex groups. So the part of the mesh is not being worked on currently could be drawn like that.
What I was thinking here was that such a thing could help with tasks like selecting faces along with certain modeling operations (extrusion for instance).
The only real flaw with this idea was that it also came with the idea of having faces masquerade as something they are not (ie. Ngons being drawn in a way that resembles a quad, because I can see it being very frustrating for users if the viewing started lying to them). There should also be a slightly different coloration in the editmode graphics for the face groups to indicate that they are a group of faces with central edges hidden rather than a single face (the hidden edges could even be drawn with a very low opacity so you have at least some idea that faces were simply not dissolved without permission).
This is a great idea. To be visible but barely.
