"Tor Olav Kristensen" <tor### [at] TOBEREMOVEDgmailcom> wrote:

> If you calculate the two partial derivatives (d/dU and d/dV) for each of the
> bivariate functions for a Bezier patch you can use these to reorient objects to
> align with the U-, V- and normal directions of the surface.

One more thing to ponder the full import of.
Just keep raising the bar.   :P

> This article contains some useful information about that:
>
>
https://www.scratchapixel.com/lessons/advanced-rendering/bezier-curve-rendering-utah-teapot/bezier-patch-normal

I do believe I've seen that one, but haven't had the time to fullly dissect and
digest it.

> The attached image shows cubes that are reoriented to align with the directions
> I mentioned above. It also shows a lot of splines that creates a grid on the
> surface of the patch.

Well, assuming the patch is bounded by a unit sphere, then those are ---
subcubes.

:D

I have a vague notion that some of this might be accomplished via matrix
transforms.   But that's only a tenuous supposition at this point.



....
Care to make a MEDIA Bezier patch?  :)

Post a reply to this message