Hi Michael,

You seem to have had no chance to react to this edited part of Jesse's message...



Petr

Oops, I didn't see that edited-in later part.

Jesse, you wrote:
> Can it be used to move points as well as scale and rotate?

To move points, you just drag directly on the points same as v1... Are you asking about a certain kind of movement different than that?


> *Could you also explain how scaling the solid object
> rebuilds the amount of control points? It makes a big
> difference in terms of the potential of the object to remain
> smooth and workable, if that makes any sense. -JDK

Scaling a solid actually should not change the number of control points in the surfaces in the solid...

Maybe I'm not quite following what you are referring to here though, could you go into some more detail?

(PS - I'll be out for a couple of days)

- Michael

>hehe... have you started learning Czech, Burr?


Sorry no! When I was in Italy I met a Czech girl and all I can remember is jak se maj, ja se mum dobre je. or something like that.

I always wanted to learn czech or russian though never invested the time I needed. Just one of those dreams I have. And always running into Czech guy's playing hockey!

THat czech girl was the most beautiful woman ever!

Burr

Hi Michael,


Well, I'm a little confused about this...but it looks like when a solid sphere is drawn in MoI and scaled 2d in a particular way, it changes from a 3 degree object to a 2 degree object, decreasing the number of control points...however, when I saved the sphere and opened it in Rhino, it read as a 2 degree object.

My original point was that if you want to control point edit a sphere into another shape, scaling it 2d also reduces the number of control points, making it less shape-able.

Jesse

Hi Jesse - I see what you mean.

The part that is confusing here is the "automatic deformable points" feature.

This is what makes a degree 2 sphere that is made up of 4 quadrants instead show the control points of a deformable degree 3 surface approximation. The sphere is actually degree 2 the whole time but if you move one of the points, it will get automatically converted into a rebuilt degree 3 surface approximation so that it can be squished smoothly (you have probably already seen the kind of quadrant segmentation that happens if you move the points on a degree 2 surface that has fully multiple knots in it).

This automatic refitting is only active for surfaces of revolution. If you scale a sphere in the direction that its seam is running, it will remain a surface of revolution, but if you scale it along one of the other directions it will no longer be a surface of revolution and won't have the automatic refitting applied to it.

I think it may be possible to me to extend this to work better in the future instead of only working on surfaces of revolution.

- Michael

The basic idea is that if you draw a sphere, I want to keep it as degree 2 if you don't need to mess with its control points since the degree 2 version is a totally precise exact sphere.

But if you want to turn on control points and squish it, you get the points for a rebuilt degree 3 version, but the sphere itself does not get changed to use that degree 3 rebuilt surface until you actually drag one of those points around.

- Michael

Hehe... the "automatic refit with deformable object" feature seems to have a lot of hidden secrets -Petr