uv point values

Hi All,

When you divide a surface you get the points P (x,y,z coordinates) and the sampled uv location on the surface.

This is a very basic question but where are these uv values coming from ?

I thought it was based of the isocurve length that the points were evaluating but I’ve tested it and thats not correct.

Thanks as always

Replies to This Discussion

Permalink Reply by Danny Boyes on July 17, 2012 at 4:20am

A Point is made up of a list of co-ordinates.

We typically think of them in terms of 3D space and therefore need 3 co-ordinates to represent them. X, Y and Z.

But they can also be expressed in 2D and 1D when it comes to a location on a surface or a curve. If a 3D point shares a location on a surface we can represent it by means of the U and V co-ordinates of that surface.

In your example above the 4 surface corners are {2,2.5,0}, {17, 2.5, 0}, {17, 19, 0} and {2,19,0}. Unless you reparametrise the domains they will typically take the same domains as the curves that constructed them in this particular case the lengths (but these curves are only that length at the edges and only when you created the surface).

So the U domain is 0 to 15 (17-2) and the V domain is 0 to 16.5 (19-2.5). Even if you transformed the surface to another location or another shape these domains will not change and therefore the UV co-ordinate will not change. If you reparemterise the surface then the domains are set to 0 and 1 in both directions and this might be easier to work with. You can think of them as a percentage then, a UV location of {0.5, 0.5} of a reparameterised surface will always be in the middle of the 2D space.

All points on a surface in 2D have a 3D space co-ordinate as well, but not all 3D points have a 2D co-ordinate. This is why we need to use the Surface CP to get a UV value to evaluate a surface at a given point.

Incidently the 1D co-ordinate of a curve is represented by the parameter t

Permalink Reply by Madu Mohan on July 17, 2012 at 5:34am

Hi Danny,

Thanks for prompt the reply.

If I understand correctly your saying that it is related to the initial way that you construct your surface (related to the actual dimensions). In the example I created a plane, then rebuilt it and moved up the cv points to give some curvature.

I guess I'm just still a bit confused because I've extracted the isocurves from the sufaces and measured them at the u/v intersections and the lengths are not corresponding with the uv values (but were pretty close) using the evaluate surface component in gh. So it seems it is more of a percentage somehow ?

Permalink Reply by Danny Boyes on July 17, 2012 at 5:56am

Yes when you construct the surface it takes on the properties of the two edges of the Plane which are lines and typically lines have the domain of their length.

If you take an iso curve at the edge and one in the middle of your example surface, physically they cannot have the same length, but at even spaced points along the curve they will have the same parameter.

RSS