Rhino SubD - new fixes and features in 9 Sep 2017 WIP

There are new Rhino SubD fixes and features in the Rhino WIP for Windows we published today.


Details are in Rhino SubD status document.


Keep it coming…thanks!

Hopefully the Mac cohort will see this in a WIP relatively sooner rather than later?

…and the ‘Rhino SubD Vision’ in your Google doc sounds promising and spot on for moving the goal post. Cheers on intent!

1 Like

And I have to ask: Did you just make a Nurbs->SubD tool to convert the Rhino logo into SubD? Or did you rebuild it?

BTW! I am so happy you let some of this great stuff make it into V6, Thanks!!!

To make the Rhino SubD Logo, I extracted wireframe control polygon from the NURBS logo faces and 3d NURBS edge curves and used them to make a mesh object SubD logo control polygon, then use SubDFromMesh to make the Rhino SubD logo. It was educational and pointed out places were we need better ordinary mesh editing tools.


Limit surface. The result you would get if you applied the subdivision method an infinite number of times.
Note: Rhino uses sophisticated techniques to create the limit surface without applying global subdivision.

As extracted from the status document, i think those lines are important enough to be highlighted here once again, something which unfastened my interest pretty well after i had read about it somewhere in the depth of this forum. It is a unique approach compared to other SubD projects and will keep up the high quality we are used to from NURBS if i understood that correct?

and i am also thrilled and urged to see this coming to the mac version sooooon…
any estimate when this can happen?

Precisely @dalelear !

And it would be my opinion that there be a mesh tool (retopo) specifically ‘tuned’ for SubDFromMesh.

Getting under the hood and messing with mesh creation, for advanced and specific purposes, would still be present, however, the ‘preset’ would be targeted towards ‘general’ successful mesh-subD-nurbs trips.

For example:

In tests, had difficulty getting Grasshopper generated mesh through mesh-subD-nurbs trips, with the goal being a reasonably editable subD, and successful nubrs patch solid.

One of the more consistently successful paths was to run the Grasshopper mesh (first image below) through AD Remake (now called ReCap Photo) and select the Fusion 360 T-Splines preset. (Hint - at present can’t do this inside of Fusion.)

Could not figure out how to get a mesh out of Rhino that would convert at all, or convert well enough for further process.

Instant Meshes algo was useful as well, but as you infer, should all be self contained in Rhino for full and proper value.

Getting the hands dirty while on a mission is the best (and only) way to discover what’s needed! And it looks like you did a great job too!

And I’d add recognition that ‘direct’ Sub-D design, via Rhino tools, not mesh conversion, is of course of primary interest as outlined in the ‘vision’ text pasted below:

"The Rhino SubD object is designed to quickly model and edit complex organic shapes.
Unlike traditional mesh-based SubD implementations, Rhino SubD objects are NOT a subdivided mesh object."

That said, in leveraging and massaging Grasshopper generations to Sub-D, it is assumed we’d need improved mesh tools, unless you guys have some new future tricks up your sleeves?

Hi, Richard,

It is our intent that presenting and managing the Rhino SubD object as a surface object with a clear and mathematically precise surface location and surface normal vectors will “feel” like the Rhino NURBS surfaces and polysurfaces you are used to modeling with.

– Dale Lear

1 Like

Rhino SubD will NEVER be in any Rhino 5.x product on any OS.

Rhino SubD will start working on the Rhino WIP for Mac (destined to be Rhino 6 for Mac) within month or so.

Rhino WIP for Mac is currently not available to the public. It will be made available to the public sometime after commercial Rhino 6 for Windows is released.

– Dale Lear


hi Dale, just one last question/wish for now if i may:

why not designating it Limit Surface instead of naming it with a catchy, infamous Sub-Division, which does its real capability in Rhino no justice? I can understand the probable marketing idea behind it, inviting professionals, fulfilling what was long wished for, but it really has a remarkable distinction and this should be well pronounced in my view. And by all means Limit Surface just sounds far catchier anyway.

Limitless Surface


this anyway :wink:

@RichardZ to be fair – SubD with a limit surface based approach isn’t something entirely new and unique to Rhino. Plugins for Rhino do use similar methods and MCAD program with a SubD-worksspace (such as Autodesk Inventor Fusion, Dassault Catia or Siemens NX…). Apart from Fusion all of these programs are pretty expensive though.

what is MCAD? i cant find anything other than some link to solidworks. manufacturing CAD?

anyway my apologies for my unreflected statement in that case, i of course have no first hand experience with any of those declared packages, so i quote from the weaknesses i have read up about t-splines regarding its surface precision. its well known that subdivision surfaces are being used for a long time already, but that any of these packages use a converging approach is really new to me.

“researching” once more about Limit Surface i stumble over Jos Stam, quoting from wikipedia:

Exact evaluation
The limit surface of Catmull–Clark subdivision surfaces can also be evaluated directly, without any recursive refinement. This can be accomplished by means of the technique of Jos Stam. This method reformulates the recursive refinement process into a matrix exponential problem, which can be solved directly by means of matrix diagonalization.

and since he had worked for alias being acquired by the all swallowing cow (sorry if this sounds rude) Autodesk, i am now wondering why they needed to buy t-splines at all :wink: making me even more eager to see this coming as a more or less, sooner or later, native instrument in Rhino.

Just did first test of SubD in recent wip and I’m more than happy with the result. Obviously it’s far from perfect yet but still…deeply impressed. Thank you McNeel team :slight_smile:

On the left my standard “render” geometry on the right reduced mesh with SubD


Sorry – I meant mechanical CAD programs. Some of them have SubD inbuilt for quite some time already.

I don’t think there’s proprietary bits involved in representing control cages with infinite subdivision. Tsplines deviated insofar that they offered a unique approach to local refinement. The latter is a ubiquitous problem in SubD modeling: One can’t simply add to or cut away stuff away from the model without ruining it’s topology.

It’s also worth stating that the makers of SubD programs which work with mesh based approximations very deliberately do so. This choice has many advantages for the intended area of use.

Hi @dalelear,

How about option described in RH-40392? When to expect?


None of the final user interface terms have been determined. Ideally, “surface” with no adjectives would work, but we need to see when and where users without a working knowledge of early subdivision product workflows and behaviors get confused. That said, your opinion is shared by some people in our office.

The long term goal is to choose terminology that supports having clear, concise, and accurate help files, tutorials, and teaching materials.


Hi Dale!
It’s great to see some progress on this, thanks!

But I have two issues with the current build.

1: EDIT: This is not true: SubD doesn’t shade or show in rendered
EDIT: It inherited the SetObjectDisplayMode from the mother-mesh.

2: I can not find the control cage option in either SubD nor SubDivide and History doesn’t work, so I wish I could have the old “keep input as control cage” option.