MoI discussion forum
MoI discussion forum

Full Version: Death of SUB-D Artist)! (Closed)

Show messages:  1-8  9-28  29-48  49-68  69-88  89-108  109-119

From: Andrei Samardac
31 Jul 2013   [#49] In reply to [#47]
BurrMan,

1. WarrenM says - "No, you can't go back once you commit the booleans" is it true Sub-D?

2. So any way if it is true sub-d modeler is it possible to import obj of any complexity without modification like retopology and modify them ?
From: Andrei Samardac
31 Jul 2013   [#50] In reply to [#48]
PaQ,
As I can see there is some kind of misunderstood, noticed it just now...

I mean death not for sub-D like software or method of modelling. I mean MAN - MODELER. I mean this figure I made will kill him! I meant death for modeler/Artist not method!!!
Method is good I know a lot of GREAT stuff that was made with Sub-D.
Once again not method SUB-D but MAN who model it)))))! I think it takes a lot of time to model it in Sub-D, so it's "death" for man that will think - My God I have to model it...
Hope now you understand me)
From: Andrei Samardac
31 Jul 2013   [#51] In reply to [#50]
Just changed in first post Modeler to Artist.
And once again it was joke!!! As you can see in my first post, so that is why I wrote - Joke.

And once again I convinced this forum is not for jokes most people here are very serious :D
From: WarrenM
31 Jul 2013   [#52] In reply to [#49]
You can't go back to MeshFusion once you commit but you CAN still continue to model with the mesh it produces - If Groboto is any indication, the resulting mesh will have decent topology and should be easy-ish to work with.

However, the same is true of MoI ... once you do your boolean operation, you can't go back and change your mind. If you want to do something different, you have to rebuild the surfaces and try again. Right?
From: BurrMan
31 Jul 2013   [#53] In reply to [#49]
"""""""No, you can't go back once you commit the booleans" is it true Sub-D?"""""""

All booleans are fully editable and interactive. Both originating objects remain, always.

"""""""""2. So any way if it is true sub-d modeler it is possible to import obj of any complexity without modification like retopology and modify them ?"""""""""

Interesting how you jump around to make a point....
From: Andrei Samardac
31 Jul 2013   [#54] In reply to [#52]
WarrenM,
I can repair any booleans I made in MOI, to do it you have to know some repair techniques.
But what you can do with mesh that you commit? I think only sculpt.
From: Andrei Samardac
31 Jul 2013   [#55] In reply to [#53]
BurrMan,
I'm not jumping I just think logically. All you say have to be proven.
From: Michael Gibson
31 Jul 2013   [#56]
The new Groboto fusion stuff looks very cool! That would be a major step forward for incorporating booleans into sub-d modeling work while before they were like oil and water.

From what I can see it has to be changing the topology of the existing sub-d objects on the fly and it is fairly easy for any automatic retopology method to introduce kind of stretched or squashed areas in the result, so it will be interesting to see if there tends to be any shape perturbations in the result, and it's also a big question whether any input mesh can be used or whether there are restrictions on the input topologies.

One of the strengths of NURBS modeling is not just that you can do booleans but also that you can form large parts of your model from 2D curves. It's actually the 2D curve focus that tends to make NURBS modeling easier for novices to get started with, and it's also where advanced users gain a lot of modeling speed from as well. The new Groboto Modo plug-in does not really bring that particular same aspect to the table, but it looks like it should be really useful for already experienced sub-d modelers to be able to merge forms together a lot easier than they could previously.

One thing that might be kind of weird is to only be able to apply fillets as the result of doing a boolean, it's not unusual for something that you want to fillet to be an edge of the natural outline of a shape for example rather than only something coming from a boolean.

It will be cool to see more demos of it, it's certainly going to be a lot more useful to people than the stand-alone Groboto was.

- Michael
From: Andrei Samardac
31 Jul 2013   [#57] In reply to [#56]
Michael,

>and it's also a big question whether any input mesh can be used or whether there are restrictions on the input topologies.

Yep I asked the same question... If it is true SUB-D it have to accept any kind of poly model without retopology...


>One of the strengths of NURBS modeling is not just that you can do booleans but also that you can form large parts of your model from 2D curves.

True!
From: WarrenM
31 Jul 2013   [#58] In reply to [#54]
"I can repair any booleans I made in MOI, to do it you have to know some repair techniques."

Right, just like subd modeling.

"But what you can do with mesh that you commit? I think only sculpt. "

If the generated mesh is better than the Groboto meshes (which they have said is the case), you will have a good mesh for further subd modeling after you are done with your booleans. So, no, not only sculpting. In fact, you could even put the mesh BACK into MeshFusion and do some more boolean work on it.
From: WarrenM
31 Jul 2013   [#59] In reply to [#56]
"It will be cool to see more demos of it, it's certainly going to be a lot more useful to people than the stand-alone Groboto was."

HUGE understatement. :P I loved the tech aspects of Groboto but the app was exceedingly blah.
From: Michael Gibson
31 Jul 2013   [#60] In reply to [#58]
Hi Warren,

>> "I can repair any booleans I made in MOI, to do it you have to know some repair techniques."
>
> Right, just like subd modeling.

No, the repair options are pretty different with NURBS modeling compared to sub-d because NURBS has the concept of an "Underlying surface". When you do a boolean operation on a NURBS object new trim curves are created but the original full surface is still underneath those trim curves and if you erase the trim curves it will restore the full surface.

Sub-d modeling does not have any concept similar to that, if you eliminate some portion of a sub-d mesh you can't then just remove the boundary areas of where you cut it and restore it.

- Michael
From: WarrenM
31 Jul 2013   [#61] In reply to [#60]
Yes, my point was that you can repair the booleans in subd, not that you can repair them in the same way. Clearly the techniques would differ.
From: PaQ
31 Jul 2013   [#62] In reply to [#54]
// But what you can do with mesh that you commit? I think only sculpt.

Hi Andrei,

Further modeling, sds modeling is allways a destructive process.
For instance you often freeze a sub-d shape just to add further details, so you minimize the distortion the new detail
can create on the original surface.

It's definitivelly not something that scare us :)

(and indeed I did completely miss understood the meaning of your first video, sorry)
From: Michael Gibson
31 Jul 2013   [#63] In reply to [#58]
Hi Warren, here's an example of the kind of repair that Andrei is referring to. If you start with a surface like this:



If you then cut the surface with a 2D profile like this:



What you now have is a "trimmed surface":



Trimmed surfaces are part of how NURBS models are structured, the trim edges basically mark which areas of a surface are active and which are cut away. If you then select all those boundaries and hit delete the boundaries will be removed and the full underlying surface will be restored again:




NURBS modeling is based on this concept of a trimmed surface, while sub-d modeling is built on a bunch of individual small elements at all boundary areas, there isn't a kind of larger "underlying" element in sub-d modeling like there is in the NURBS object structure.

- Michael

Image Attachments:
untrim1.jpg  untrim2.jpg  untrim3.jpg  untrim4.jpg 


From: Andrei Samardac
31 Jul 2013   [#64]
WarrenM and PaQ,
As I can see the cage is pretty dense is it good for modeling? With this dense it is more comfortable to sculpt not to model, or I'm wrong?

PaQ,
> (and indeed I did completely miss understood the meaning of your first video, sorry)

No problem)) May be I have to name it a bit differently :)
From: WarrenM
31 Jul 2013   [#65] In reply to [#64]
Michael

I think we're crossing streams. I appreciate the explanation, but I totally get that. I know basically how NURBs work. All I was saying to Andrei was that both methods are repairable if you know the techniques. Is subd harder to repair? Sure. But it IS possible. That's my point.

Andrei

Groboto meshes were fairly dense. I believe the MeshFusion ones are going to be lighter weight. We'll see! If they aren't, well, that'll be a problem...
From: Michael Gibson
31 Jul 2013   [#66]
Another question is - does it _only_ do filleted booleans? What if you want to have sharp crisp edges in a particular cut area?

If you can't actually make sharp edged booleans and can't fillet already existing edges of a model and only ones where you are booleaning something, that seems like it could make for some kind of weird modeling constraints when you're trying to do a larger and more complex project all using that construction method.

I guess for a hard edged boolean you would freeze the result and then use regular Modo non-fusion mesh booleans on it, but then you have a non-sub model after that so what if you then want to do another fusion boolean on it again after that?

- Michael
From: Michael Gibson
31 Jul 2013   [#67] In reply to [#65]
Hi Warren,

> Is subd harder to repair? Sure. But it IS possible. That's my point.

I don't get this point - I mean if it's only whether something is "possible" that makes the difference, it's totally possible to do what the new Groboto fusion is doing by just editing a mesh by moving individual points around. But because that's way way harder it's not part of what we're discussing...

If something is easier to the point of it taking 2 seconds to do it versus a long time, that's a major difference.

Surface repair in NURBS modeling is very different than sub-d modeling repair, if you are talking about recovering a portion of a surface that has been cut away.

- Michael
From: WarrenM
31 Jul 2013   [#68] In reply to [#67]
Michael

Andrei said:

"I can repair any booleans I made in MOI, to do it you have to know some repair techniques.
But what you can do with mesh that you commit? I think only sculpt."

My point was to refute the idea that a subd mesh is only good for sculpting after using a boolean on it. However, NURBs are easier to repair, you're correct on that front.

Show messages:  1-8  9-28  29-48  49-68  69-88  89-108  109-119