Hi Michael,

Maya can only convert what they called 'subdiv surface' into nurbs, so in the image posted, it's the middle model. This model is a conversion from
the polygonal one. The subdiv shape allready shows some complex area in the front of the model, and this seems to be translated into the nurbs model at the end.

I have just no experience about how subdiv actually work, it seems to be way more heavier than catmull scheme. Maybe I should first learn this part to have a better control of the nurbs conversion.

I didn't remember having this kind of result with the t-spline. With the t-spline plugin, the nurbs flow was more looking like the maya subdiv one. (I'm not sure if this sentence is clear enough :P).

Hi PaQ, I see - I'm not really familiar with some of these details on how Maya works, but it appears to me that the middle "subdiv surface" has already been divided down in those valence 5 areas.

So my guess is that a "subdiv surface" in Maya is some kind of hybrid object, which has already been prepared in its structure for converting to NURBS? Something like that anyway.

But if you continue that edge loop that I showed above to the end of the shape in the original polygon model before conversion to "subdiv surface", it will probably simplify that area that you were pointing at with the red arrow above.

You can't always avoid such areas, but in areas that you can it will probably make the result more simplified.

- Michael

Yes I'll give a try with a pure quad model, ... at the moment I'm still learning all the selection and objets organisations tools ... (in other words, I'm nowhere :o)).