Can Rhino model any object which polygonal packages like Maya are able to create?


(Alex) #1

Hi,
Can Rhino without Tspline or Clayoo have the power to model any kind of objects that Maya or 3ds max are able to create?
For example the following image is a polygonal model which has been rendered. Can Rhino model such a thing without Tspline or Clayoo?


Almost T-Joint/Star-Joint how would you model this?
(Nathan 'jesterKing' Letwory) #2

Check out the replies to Microphone Mesh


(Alex) #3

Ok,
So, If Nurbs is powerful why do some people use polygonal packages for complex product modelings, like digital camera models? Why don’t they use Nurbs?


(Nathan 'jesterKing' Letwory) #4

To each their own? Using a specific package doesn’t necessarily rule products out.


#5

One mayor reason is convention. People tend to use the tools they already know for whatever problem they have. People also use Rhino for tasks which really would be a lot faster to do inside a mesh package…

Further reasons:

Certain topologies are indeed a lot easier to model with meshes. I would not mention a detailed digital camera here, but a half way good looking (render only) car body concept is in comparison easy to model with Subdivision Surfaces, as used in Maya.

Subdivision surfaces models – when mesh based – are very light geometry, things render fast.

Mesh modelling packages traditionally come with greatly superior tools to texture and render models, compared to (any) Nurbs modeller. It is easier to add a certain degree of imperfection and wear and to apply micro fillets and displacement which make (at least the rendered model) look very convincing.

One of the major reasons for the development of mesh based Subdivision Surfaces tools is abiltity to deform finished models. This plays a major role in character modeling and can’t be done effectively with Nurbs.


#6

If you don’t need to care about production-ready data or control of many interlinked dimensions or surface quality, you can work well with polygonal modelling software. Maybe your clients are doing all proper CAD work in-house and you only need to supply concepts, renderings and so-so 3D-prints.

If you need to also render your products in Arnold, Maxwell, etc. you can convert the NURBS model into a polygonal model; often handled by the rendering software’s plug-in, if one is available. You can also save all components as OBJ files from Rhino and take it from there.

You could of course do it, but in a commercial studio/client environment, you would rather not want to accurately model a mostly G2 continuous consumer product like the one below in a polygonal modelling software, if tolerances, draft angles, tooling requirements are your goal.


(Michael Pryor) #7

You can consider a comparison to images where:
Mesh = pixels (approximations and blends to fool the eye)
Nurbs = vectors (mathematical and infinitely scale-able)

Meshes are straight patches so the more resolution of curvature you need to more pixels you need, This can result is very heavy geometry if the object is needed for the real world and can also mess with reflection like a disco ball made of many square reflective surfaces. Mesh however is much better for texturing and mapping in things like videogames.

Nurbs are actually curved and therefore more accurate in the real world but harder to model some things with since complex geometry requires a working of many patches and managing the continuity of curvature across seams.

Personally I prefer to sketch with mesh because the workflow to get a shape and idea out is just faster, but then translate to Nurbs for the real object.

As usual though it really depends on what the model will be used for and how / if it will be manufactured in the real world.

So I would say any model software can make what the other makes, it is up to the skill of the person modelling really (unless the software package is missing some very basic functionality)


#8

As a distillation of the thorough, detailed responses above, I typically describe the distinction to the uninitiated simply as:

Does it need to look right, or be right?

If the creation only needs to ‘look right’ on screen, paper, etc., then the polygonal mesh approach may be appropriate.

Conversely, if the creation needs to both ‘look right’ and ‘be right’ incarnate - for manufacturing, fabrication, performance, etc., then traditional CAD tools are typically required.

There are gray areas, however.


(Gustavo Fontana) #9

After many years of using both SubD and Nurbs for me is absolutely incomprehensible why people would only use one topology/tool. Just like we sketch before modeling anything. We also SubD before we Nurb anything. Unless is a very mechanical form with zero surfacing challenges.

Also we combine both. Main forms in SubD. Details in Nurbs after conversion. Since each of the two topologies excel or absolutely such for those two types of modeling.


#10

Indeed, Gustavo, the lines should continue to increasingly blur going forward. Spot on!

Beyond ideation and rapid digital viz, we know that careful subD construction can find its way direct to manufacture for some forms; or nurbs elements may be leveraged as snap templates for regions of subD forms; or subD forms may serve as bridge connections between nurbs elements.

The import of such grows daily. Any introduction of new methods for ‘accuracy’ of design intent (think nurbs) applied to any form of subD construction will move the needle.


(Alex) #11

I find mesh modeling very difficult in many cases especially if you need to add more details. For example if you want to create a free-form shape which contains lots of custom-shaped holes it’d be very difficult to maintain a good topology and avoid any artifacts after smoothing. But in Nurbs surfacing softwares like Rhino or Alias, this is done simply.
I also found out that in Rhino 5, “blendsrf” command has G0, G1, and G2 continuity options while Rhino 6 has G0, G1, G2, G3 and G4 continuity options. Personally, I’d recommend Rhino 6, not Rhino 5.


(Gustavo Fontana) #12

This is why I said:

And that was an autocorrect typo. I meant ‘absolutely suck’


#13

Somewhere it is more convenient to use NURBS and where SubD is more correct. I hope that the near future (by the beginning of 2019) the SubD system will be introduced in Rhino (and somewhere at the level of T-Splines). Then it will be possible to switch easily for faster, high-quality modeling. :grinning:


#14

About Rhino a few cool free videos about modeling NURBS - it does the program more difficult.
And to SubD it is possible to find all and free.
The choice is simple!


#15

I am sorry, I missed something. Which camera was modeled, designed, and brought to market using a polygonal modeler?


(David Cockey) #16

Perhaps he meant the creation of an object which looks like a camera, not the actual design of a camera.


(Alex) #17

I think Rhino 6 has SubD tool now. Of course I saw it in Youtube but I couldn’t find its command in my Rhino 6.
Does anyone know whether there is any command about SubD in Rhino 6?


#18

Nobody even implied that.

This is one of those classic discussions. Some say that Designers should use Nurbs all the way – as seemingly the outcome can by principle no more make sense when utilizing less precise digital tools somewhere in the process (your question implies that you sympatize with this group).

Others say that it can be quick, fun and extremely inspiring to use all sorts of modeling paradigms: Possibly throughout the project for stuff which generally doesn’t need micron level detailing – or at least in project states where superb precision ain’t the main goal yet.

I have learned building furniture at a time when pretty much anyone still drew with ink. We used rulers to measure distances from these drawings. Every furniture drawing done in Maya or Sketchup is more precise.
Somehow things still worked out…


#19

And through SubD the camera in general in any way it is impossible to model that? Or I haven’t understood something?


#20