Wasn't it Rafal 'Raf256' Maj who wrote:
>nos### [at] econymdemoncouk news:$0EsSAAl+jCA### [at] econymdemoncouk
>
>> Hint: If you can't see your groove, try temporarily changing
>> "difference" into "union" so you can see which direction you have to
>> translate it in order to get it to intersect the correct part of your
>> isosurface.
>> 
>> It's probably not a good idea to try to make your text or heightfield
>> into an isosurface function if it has sharp edges since you're likely to
>> get infinite function gradients.
>> 
>
>I want the height-field to be a part of isosurface function, so I can i.e. 
>displace or wrap it, like I can with any isosurface function

I presume that you've tried making a function from your heightfield
image and adding it to your isosurface function and found problems with
that. Here are a few alternative approaches.

1. If your only problem with that approach is due to the infinite
gradients caused by sharp edges in your image, then you could try pre-
processing your heightfield image to apply some form of blurring or
smoothing.

2. Don't knock it until you've tried it: Use your heightfield image as a
bump_map. I know it's not true 3D, but it can often look pretty good.

3. Since you're using the isosurface to generate a simple modification
from a box, you might try approaching the problem from the other
direction. Use the function that you were perturbing your isosurface
with to create a pigment. Use that pigment to create an image. Use that
image as six intersecting heightfields to create a warped box. Blend
your groove heightfield with a copy of this structural heightfield.

-- 
Mike Williams
Gentleman of Leisure

Post a reply to this message