Hi Michael ,Jonah , end Friends ;
today I have sent to the tspline forum this work ;
end so it is like imported in Moi one tspline surface-
best
al
Sorry for my English

That looks really cool speedy.

lovely pictures!

and a really good example of t-splines :)

Fredrik

Just wondering, because of the parenting link between MoI and Rhino, is it really a lot of work to build a Tspline for MoI version ?
I have no idea the amount of work required to do a MoI flavour of this technology.

+1 perhaps just the tsImportOBJ comand for starters?

Hi PaQ,

> Just wondering, because of the parenting link between
> MoI and Rhino, is it really a lot of work to build a
> Tspline for MoI version ?

Basically the internal guts of MoI and Rhino are pretty different.

I mean there are general similarities like MoI and a point picker mechanism and Rhino has a point picker mechanism.

But the details of them are really different. When I was doing the architecture for MoI I was focused a lot more on doing a kind of "next generation" start-from-clean-slate improvement of things rather than trying to maintain exact compatibility with Rhino's internal architecture.

Especially anything having to do with UI is very different, for example Rhino commands expect to interact with the user through the command line, and tell Rhino to show text options on the command line - MoI's UI does not have a command line, MoI instead expects for commands to interact with the user by the command options area where it shows buttons, checkboxes, and GUI controls like that, and these are defined by HTML documents which is a much different structure than Rhino's command line options.

So anything involving UI would certainly involve a lot of work and cannot just be automatically moved over.

Things which do not involve any UI are always a lot easier to do, but a lot of the value of T-splines revolves around using a sub-d modeling approach which has a lot of UI to go along with it...

- Michael

Hi yannada,

> +1 perhaps just the tsImportOBJ comand for starters?

Doing just that part would certainly be a lot more feasible from a "time of work involved" perspective.

But it wouldn't deliver too much "new" value - I mean it would just be replicating something that you can do already with Rhino.

Basically it just does not fit in well with the overall dev strategy of MoI to kind of hack partial things for the short term that don't quite fit in, especially if they are already covered well by other software...

- Michael

Michael Gibson wrote: "Especially anything having to do with UI is very different, for example Rhino commands expect to interact with the user through the command line, and tell Rhino to show text options on the command line - MoI's UI does not have a command line, MoI instead expects for commands to interact with the user by the command options area where it shows buttons, checkboxes, and GUI controls like that, and these are defined by HTML documents which is a much different structure than Rhino's command line options... So anything involving UI would certainly involve a lot of work and cannot just be automatically moved over....

This probably becomes even more true in the case of version 2.0, as they've now disengaged major parts of the Rhino UI to enable a more subd friendly workflow... I would think the best hope for non-Rhino users is for ts to build a stand-alone conversion app which eats OBJ and spits out 3DM...

jonah

>I would think the best hope for non-Rhino users is for ts to build a stand-alone conversion app which eats OBJ and spits out 3DM...

jonah

hmmm that's what the should have done on the fist place but is never late...

Is there any other solutions similar to Tspines?

your first page = expired
second want a login ;)

Hi jonah,

> I would think the best hope for non-Rhino users is
> for ts to build a stand-alone conversion app which
> eats OBJ and spits out 3DM...

That's a really good product idea for the T-splines guys, I forwarded that to one of their developers who I was talking to recently.

That would work well with MoI but there are also all kinds of other scenarios where a stand-alone simple conversion app could be helpful, like someone who wanted to use Modo and SolidWorks in combination with each other, etc....

It would possibly not take too much effort for them to set up a product like that, it would be able to have a very minimal UI, just something like a viewer window to view the imported mesh.

- Michael

I'm just a bit confused. I thought TS gave the abilty to take a MoI model and set it up like subd with point pulling type stuff.

I guess it sounds like also a conversion to get your subd model into MoI?

Hi Burr,

> I'm just a bit confused. I thought TS gave the abilty
> to take a MoI model and set it up like subd with point
> pulling type stuff.

It's sort of the reverse of that - it lets you make a sub-d model by using a polygon hull + point pulling type stuff, but then can output a NURBS model as the final result rather than only exporting a polygon model as the final result.

So you can then do NURBS operations on the result that you just sculpted, like boolean holes in it, etc...

It's not really something that you use to modify an existing NURBS model that you have already created, it's an alternate way to create shapes from scratch (or from imported polygon models that have been created in a poly modeling program).

- Michael

>It's not really something that you use to modify an existing NURBS model that you have already created,
>it's an alternate way to create shapes from scratch (or from
>imported polygon models that have been created in a poly modeling program).

I'm glad you cleared that up as I thought tsplines allowed you to model with nurbs using sub'd techniques. To me this would be the single most leap forward for nurbs over the very limiting traditional nurbs tools, such as loft, sweep and network which tend to warp, twist & buckle nurbs surfaces even when used on minor changes of shape or direction.

I hope they do make a standalone product for tsplines that writes 3dm. That way it could be used with MoI.

Hi Kevin, one thing to keep in mind is that sub-d surfaces can also have lumps and buckling in them in various situations as well, your sub-d control cage can't just be a random pile of points and still needs to have well placed points and topology in it.

See here for some examples of the kinds of problems you can run into:
http://www.guerrillacg.org/home/3d-polygon-modeling/subdivision-topology-artifacts

But it does tend to make you constantly aware of the control point structure and encourages you to keep a minimal point count which simplifies things.

If you try to do the same kind of minimal point spacing with your input curves for lofting or sweeping it can kind of have a similar effect of reducing buckling in many situations there as well. If you try to use a whole bunch of curves or a whole bunch of points it usually tends to exacerbate it.

- Michael

< http://www.guerrillacg.org/home/3d-polygon-modeling/subdivision-topology-artifacts
Cool site!

Someday, when Michaels done working a bit and taking a break, Maybe we can see a page with links of "Michaels Favorites". I bet it's a wealth of Knowledge!

>one thing to keep in mind is that sub-d surfaces can also have lumps and buckling in them in various situations as well

Very true. I guess the difference is that you can add some more edges or rearrange your sub'd cage and do some point tweaking to remove the lumps and buckles. With nurbs there are just too many points on the surface to do this. Nurbs however are certainly great for the more mechanical type parts.

Hi Kevin,

> I guess the difference is that you can add some more
> edges or rearrange your sub'd cage and do some point
> tweaking to remove the lumps and buckles.

The equivalent for NURBS is more about doing a similar process on your original curves to smooth them out or simplify them and recalculating the loft or sweep, rather than working on the NURBS surface directly.

But one thing that can make this difficult is if you're trying to do a "reverse engineering" type task where you're trying to very specifically control the curves to match some existing blueprint. That has a tendency to lead you to put in more points or more curves, it can kind of lead you away from a simple curve setup...


> Nurbs however are certainly great for the more mechanical type parts.

Yeah for those things you use the "even more traditional" and fundamental NURBS tools such as extrusions, revolves, and booleans which are pretty much immune to buckling or wiggling type problems.

- Michael