Attached a NURBS curly cord and its mesh. I am trying to reduce the poly count when running Mesh on this type of object, and I ran into the situation where no matter what I set, I cannot reduce it further - although it seems like I should be able to. It would be nice to have only 8 facets around the circumference instead of the current 16, and perhaps only half the longitudinal segments as well - that would reduce the size to 25% of the 16+K polys now in the file. 4K would be cool.
I know I can use ReduceMesh to lower it further - except that it triangulates everything, so the poly count is still double what it could be - but it seems I should be able to do this even better just with normal mesh parameters. The one in the file has been meshed with everything set to 0… I tried a bunch of different things, nothing got better than that… Anyone else want to play?
For the size of the objects and the level of detail I need in this case, the following isn’t too bad:
Rebuild the surface degree 1 in both directions, 6 in the circular direction (make the cross section a hexagon) and ~700 in the other… Mesh with “simple planes” checked. Makes a 4200 poly mesh.
My point is that I should be able to do this with Mesh alone…
Ah, OK, that works, I had tried with Max angle, but only up to 45, I see that going above that it does reduce the poly count… and setting a min edge length also helps. I can get it down to around 4K as well that way, thanks! However, in this case only, the mesh resulting from my rebuild procedure is cleaner and more even…
How about this one…? (yes, it’s a mess) I tried all of the above and a few more, can’t get it below 10K polys or so… Unless I rebuild the surfaces themselves. --Mitch
The MESH tool is just as crap as the Make2D tool. They both are WIP developments that made it into the product and has only gotten really lousy updates. It is a shame, and it is sad that we have to learn workaronds to solve simple tasks like this. Meshes are important.
The thing with MESH is that it is very buggy and badly written, so for some settings to work they requires other settings to have a value higher than 0. This might be “logical” in math, but it is FAR from logical for any user.
That is, Rhino sais it is 980 polygons and 1314 vertices if you look at the object properties details, BUT 1958 polygons if you start reduce mesh… (If you unweld it with 0 degrees and explode it then it becomes 980 polygons though… so who knows how reducemesh counts)
Update:
You can skip the maxedge length for both of these objects.
I sometimes set the min and max edge length to the same value to be sure I get an even mesh, to avoid artefacts.
Yeah, the thing is it’s just wrong in this case (second one)…
If you set just a minimum edge length, it should pay attention to that, which it doesn’t.
If you change the max dist edge to srf (in addition to or instead of the above), nothing changes.
I wouldn’t have thought that setting the max aspect ratio higher would have the effect of making bigger quads; I assumed it would do just the opposite… But it seems to make Rhino pay attention to the minimum edge value, which it otherwise doesn’t…
I think we would see better results if we didn’t ask McNeel to go and rip off someone else’s software.
Phrasing a feature request as: I am a [Naval Architect] and I would like to see Rhino [create better Meshes by being able to divide faces once they reach a threshold size compared to their neighbours] as this would be beneficial to me and other [Naval Architects] when [Exporting geometry to FEA Packages like ANSYS]
Fulfilling user requests has much more plausible deniability than going all out on an organised expedition to Plagiarismstan
I agree, but after ten years of asking politely it is about time to try something new
Meshing has been high on the wish list for both V3 and V4, it has improved, but not vastly.
And by the way, improvements also comes a long way from looking at others. It’s called reusing the wheel
