In article <ZTq### [at] econymdemoncouk>, Mike Williams
<mik### [at] econymdemoncouk> wrote:
> Using a negative threshold isn't a solution, it just makes the surface
> that you're interested in larger. MegaPOV still treats locations where
> the function is numerically less than the threshold as "inside".
Er, right...what was I thinking?
Just using the negative of the function should work though.
> Experiments also reveal that "open inverse" produces a rather different
> effect from "sign -1". In CSG, the "open" seems to be ignored most of
> the time, so the objects intersect with just the "inverse", i.e. the
> positive copy of the target surface inside an inverse contained_by
Yes, this is right. It should work fine in media, but not with CSG.
And I don't remember why I thought "open" was necessary(I must have
written that in a moment of caffeine deprivation). "inverse" should work
the same as "sign -1", though. They both have the effect of flipping the
inside and outside, right? Or is my brain still short-circuiting?
> The problematic surfaces seem to be those in the i_algbr library where
> the origin lies in the region that we would intuitively consider to be
> the "outside" of the object. The author of the i_algbr library seems to
> have rigged the code so that the point <0,0,0> is always in the "inside"
> region, even in surfaces like Torus_gumdrop and Dupin_Cyclid (which are
> rather like torusses, with <0,0,0> in the middle of what we would
> consider to be the hole) .
This is certainly non-intuitive. I have noticed a lot of odd design
decisions in the isosurface syntax(the use of <> brackets for holding
parameters, a different meaning for "bounded_by"(no longer the a
problem, now that we have "contained_by"), functions using the bounding
box for data, etc.), hopefully all of these will be cleared up when the
official version is released.
Christopher James Huff - Personal e-mail: chr### [at] yahoocom
TAG(Technical Assistance Group) e-mail: chr### [at] tagpovrayorg
Personal Web page: http://chrishuff.dhs.org/
TAG Web page: http://tag.povray.org/
Post a reply to this message