Lummox JR wrote:
> 
> Mike Williams wrote:
> > Rather than using sin waves, you might find it better to start with an
> > isosurface of the y plane [ function {abs(y)} ] and then add some random
> > hilly bumps by subtracting a noise function or a pigment function, like
> >
> >     function { abs(y) - noise3d(x,0,z)*0.5 }
> >
> > It's possible to add a suggestion of ground cover (unfortunately, not
> > particularly grassy) foliage by subtracting another noise or pigment
> > function at a much smaller scale, like
> >
> >     function { abs(y) - noise3d(x,0,z)
> >                -noise3d(x*100,0,z*100)*0.02
> >              }
> 
> I took your advice on the basic hill structure (haven't gotten as far as
> ground cover yet) and used a formula that was basically
> y-noise3d(x,0,z), with a little noise added to x and z as well to
> turbulate it a bit.
> Unfortunately, although the basic shape is right, I've run into an
> interesting problem: Because the noise function is clipped to values of
> 0 to 1, most of the resulting hills have flat surfaces.

I vaguely remember someone complaining about a similar problem, and
noting that using dents as a function eliminated the plateaus.

-- 
Xplo Eristotle
http://start.at/xplosion/

Post a reply to this message