Get both! Each covers a different part of technology. And no, modeling techniques are no transferable.
Cheers
Lemo

Yes, polygon and nurbs modeling are very different. Skills in one don't transfer well to the other.

Hi Chris,

You wrote: "maybe not as efficent in certain areas"

I think you may be getting a kind of mistaken idea of the degree of difference in efficiency - for mechanical shapes it is not like MoI is just a little bit more efficient, it is more like for shapes that involve one piece cutting another the poly modeling approach can be so extremely less efficient that it can become just not really very feasible to get stuff done without spending a really huge amount of time.

In general the boolean tools in poly modeling programs are just not very reliable, here are a couple of threads discussing this on the silo forum:
http://silo3d.com/forum/showthread.php?t=14401
http://silo3d.com/forum/showthread.php?t=13986

Don't get me wrong - Silo is a really great polygon modeler. But if you need to do mechanical shapes, it is not just slightly less efficient, it is really not the proper tool for that kind of a job at all.


Also I think that you will find that poly modeling actually has a higher learning curve than MoI - one of the nice things about MoI is that you can get a lot of stuff done by working in 2D by drawing curves. In a lot of ways this will probably be more familiar to you coming from a graphic design background, it is generally more similar to stuff like Adobe Illustrator / Photoshop than poly modeling is.

In poly modeling you tend to be working more in a "sculpting" type manner manipulating points in a 3D cage directly, it is kind of a special skill to develop on its own, not really a lot in common with 2D type drawing.


If you only need to model things occasionally and don't really have a whole lot of time to spend refining your skills in a new toolset, then I would actually think that MoI is really the most suitable thing for you.

If you need to do highly organic stuff like human faces, creatures, characters, stuff like that - that stuff is way better to do in a poly modeler than in MoI. However, that doesn't mean that stuff is "easy" to do by any means, if you need to do those kinds of models you should plan to spend quite a bit of time to learn how to do it in the poly modeler as well. This type of modeling is just not an easy task even when you are using the right tools for it.

But certainly a lot of this is subjective as well, different people can feel more comfortable with different things, there really isn't any substitute for trying things out yourself to see what suits your particular needs best.

- Michael

Chris,

Funny - I narrowed my search the same as you did. Both MOI and Silo seem indespensible.

Definitely indispensible when they finish the new surface tool which will allow you to retopologize MOI objects into a more game/animation friendly form - that's what I'm hoping and waiting for!

-Will

Hoping I don't send like an advert :)

3dcoat has just done a new beta which has retopo-whatever tools, might be worth a look ?

edit - video link

http://207.44.140.6/~dcoat/Tutorials/3dc_retopo.swf

this was done by one of the users from an earlier beta

3D Coat is the new "Jewel" with this hugely affordabable, brilliant, 3D Paint/3D Manupulation app.
(A great partner to MoI!)

My problem with 3D-Coat, similar to ZBrush is that it's not all that friendly to hard surface model imports (NGons are typically spit out in distaste.)

Being a scifi / space-nutt kinda guy, my modeling leans more towards hard surfaces than organic...

I'll take another gander at 3DC and its new retopo since it's a function I really need to make MOI more prominent within my modeling pipeline which is something I truly desire!

-Will

The Current update info on 3D Coat.

ok, it seems beta update postponed again... Too much changes and too raw still. The upcoming list of changes:

1) selecting packs of faces/verts/edges and operating them
2) subdividing individual faces, incremental subdivision (if you have subdivided some face and want to subdivide neighbour face later, it will be subdivided like they was subdivided at once)
3) splitting any configuration of edges, nor only edge rings
4) merging vertices together
5) making retopo mesh from current low-poly mesh
6) import retopo mesh
7) interface adjusting
8) cursor flicking fixed
9) GMA945 issue fixed (i hope)
10) importing mesh with 3 types of uv-smoothing (smooth but keep edges, smooth all, smooth but keep corners)
11) baking tool improvement (speed, bugfixes and possibility to work with input meshes with many uv-sets)
12) maybe something forgotten....

Will.
Most apps seem to dislike ngons. The quality of MoI .obj saves without ngons seems just perfect to me.
Are Ngons something that really are of value?----

Brian

.

Hi Brian, nope, in lieu of "Export Zen" (all quads), my models typically have NGons.

From past experience, I just try to avoid triangles like the plague since they always caused ugly facets in my surfaces when trying to render.

Granted, this avoidance was developed in "earlier generation" applications, Rhino v2 for example - the exports from there were so ugly going into lightwave at the time it made it unusable for me. (Of course now we know that LW has it's own issues with how it's not all that great with surface smoothing or working with surface normals)

However, majority of times I don't need to subdivide my models, just load, perhaps a bit of retopo, and paint.

Local subdivision and sculpting is nice (dents and dings for example) but not always necessary for me. 3D-Coat immediately wants to subdivide your model upon import (unless I'm missing something on that import dialog...)

I'll take your advice and see what triangles and quads will do for me (ultimately everything reduces down to tris anyway...)

I'll pop in and ask if there are any plans to better support hard surfaces - no import subdivision, just working at the poly level; retopo, and some minimal poly edting (which seems to be in there already.)

-Will