"Straightening" quads

I’m testing out QuadRemesh for a production process–sorry keeping vague for reasons–and it’s…pretty sweet, it’ll obviously save hours over how I was doing this. The results are almost perfect for what I need, except that I’ll need to straighten out rows like shown below. Maybe adding more guide curves can control this, but you know, there’s a cost/benefit analysis there, it’d be better to just knock this out for a concept then tweak at a later stage.

So I’ve got kind of a philosophical question here. I haven’t really gotten into Grasshopper yet, so I guess the question is am I better off figuring out how to adjust the QuadRemesh output to my particular needs or does Kangaroo or PanellingTools or something have the solution already?

yeah would be welcome to have a command for this, @gustojunk also asked for it.

it takes 5 keystrokes, including selection, to do this with the right tools:

Step1: Press “2” for edge mode selection (so only edges will get selected, nothing else)
Step 2: Select one edge on the direction you want to straighten
Step 3: Press “L” Propagate to loops
Step 4: Press “Alt + L” to propagate to all rings
Step 5: Press “Ctrl + T” to invoke SuperTaut tool, release on the type of straightening you want.

In Rhino you can select even quicker in this case by run _SelEdgeRing and double-clicking on one edge on that direction. But then it’s impossible to straighten all your loops.

G

I’ve never gotten very deep into modeling inside modo, but I remember that many tools I used were made by scripts. Is this one of those?
I suspect that if we get script access to underlying subd geometry things like these will start popping up quickly once v7 is released.
Besides that, there are so many good modeling tools in Modo that are applicable to subd. Should we start listing them all? And how to prioritize?
For example what I’ve always found very powerful are the (live) falloff types in modo, but also their way of stacking tools to customize your own tools for a particular job.

yes that was a script. I agree that the entire subD/Mesh pipeline should be opened up for scripting. I also think anything that works on mesh should work on a polygon cage of a subD. And any script or plugin, or GH definition that anyone makes for a mesh, should work for subD [cages] too.

I would not recommend replicating the toolsets of Modo here. YOu can already buy Modo if you can handle the learning curve and pain. Or even get free Blender if you want even steeper learning curve, and even more pain.

But design-centric tools? absolutely! Especially falloffs, meshfusion and action centers. Gumbal without action centers is so dumb IMO.

G

Related: circularizing edge loops and aligning vertices to guide curves and guide surfaces or planes.

These are all things that would seem to be possible using some of the “closest point” grasshopper components and someone already made a script to snap edit point to a Cplane. But they should just be commands.

I don’t think this has the outcome that @JimCarruthers intended.
This flattens the surface curvature in one direction, I think Jim wants to straighten the edges maintaining the double curvature of the surface.
I could be wrong though…

Can modo also circularize vertices on edge loops?
Can modo snap vertices to the closest point on a guide curve, a guide surface or a plane?

I agree, especially when working with SubD. We should be able to align the gumball to an edge, vertex or face independent of the selection. Relocate Gumball takes too much time.

2 Likes

Yep.

Philip

1 Like

Also, we should be able to align it normal to the overall SubD, or normal to any chosen point on the on the SubD, regardless of the subselection. I’ve noticed Rhino does this automatically with individual vertex selections, but it seems to not do it with even slightly more complicated selections, like two edges in a row on the same loop.

Well in that specific case, no, I would just want to make one direction ruled. There are others where I’d want to keep the curvature I guess, yes.

My question was really more, is QuadRemesh the state-of-the-art in Rhino for “Panelizing” polysurfaces the way it does?

1 Like

Yes, it was a script, got popular, then Modo team finally built it-in maybe 8 years ago or so. I think pretty much every SubD modeling app has it by now.

To curve no, curves in Modo as a total shit-show.
To surface (meaning polys, because there are no Nurbs) and plane (also a flat poly), you can do it with background projection that uses a closed point approach, but it’s not accurate for technical/hard surface work.

Rhino has an amazing potential to do so many more things better, especially for accurate design modeling.

I don’t know what others might think, but I don’t think the current implementation of QuadRemesh is good for this, nor it was ever meant for this.

You can use guide curves to give ‘a bit of flow’ but you cannot force it to follow your U and V input flows, spacing, let alone exploring panel count on each direction.

Take a look at ‘Parameter Plane Mesh’ in Putterfish (GH plugin) it does a great job at this. They also have frames and curves options of this tool if you prefer that.

G

Right now it’s super tedious but not a problem to manually move vertices to a curve. The only thing that’s confusing me about it is that it’s accurate in facet mode but in smooth mode when you move one vertex the adjacent vertices are affected.

And that leads me to something else I’d like to see: more tutorials about the underlying theory of SubD. There’s a lot about curves; degrees, spans, number of points… all manner of things to prepare for lofting and other surface generation. But I haven’t seen much about SubD. Like, for instance, I had to measure on my own how many edges I needed to use on an Ngon or how many faces around a pipe to get within a certain tolerance of a true circle. (10 sides gets you close enough for almost anything. Less than 8 is almost always visibly off)… this sort of info should be in an explanatory document somewhere.

2 Likes

+1 to this. People experienced in Rhino are likely to be experienced mainly in NURBS modelling (me anyway). Apart from some dabbling in Blender and ZBrush, that’s all. Theory/good practice/limitations/combinations are a bit abstract to me at the moemnt. That’s why I feel coy about even making a suggestion for improvements. How should I know what industry standard SubD tools should/shouldn’t be in the package? It would be the blind leading the enlightened.

I filed RH-58782 for Align > ToCurve to have a 2D option.