Am 24.11.2016 um 01:20 schrieb Mike Horvath:
> On 11/23/2016 6:39 PM, clipka wrote:
>> And do you have any idea what it actually does? ;)
>> Because I have a hunch you're still struggling with the basic concepts
>> you're trying to employ.
>>
>> My recommendation would be to toy around with the isosurface tutorial a
>> bit, to get a better understanding of what isosurfaces actually do,
>> before you try to apply them to your problem.
>>
> 
> No I don't. In fact I don't understand why a parametric wouldn't work
> better. Except that it is limited to u and v, whereas I need a third
> parameter L.
> 
> But you said isosurfaces were the way to go.

Yes. Because parametrics are kind of the other way round.

In isosurfaces, you use a function that maps 3D space to something.

In parametrics, you use a set of functions that map something to 3D space.


In the isosurface approach to your problem, you basically design a
function that performs the mapping

    3D cartesian -> Lch -> RGB -> "distance" to boundaries

and let the isosurface iterate over 3D space to find the points that
fall inside valid RGB space.


In the parametrics approach to your problem, you basically design a
family of functions that perform the mapping

    RGB -> Lch -> 3D cartesian

and let a set of six(!) parametrics iterate over two dimensions of RGB
space (RG, RB and GB, respectively), with the third dimension set to one
of the limits (0 or 1, respectively) to find out where those faces end
up in 3D space.


You may note that the isosurface approach will give you a volume, and is
therefore fully suited for use in CSG, whereas the parametrics approach
will give you surfaces only, and therefore will only work properly in
unions.

Post a reply to this message