|
 |
Jan Danielsson <Jan### [at] falun mailteliacom> wrote in article
<wnaqnavryffbasnyhaznvygryvnpbz.eud1ei1.pminews@news.povray.org>...
> :>A problem of coinciding surfaces?
>
> You are absolutely correct!..
>
> :>Try making your "liquid" sphere a bit smaller than 0.98 (like 0.979)
> :>
> :>Just a guess,
>
> And a good one! Thanks!
The only problem that you may get now is that the extra surfaces drive the
trace level way up; and you may get reflections from the thin film of
"vacuum" that this introduces. It may help to make one or both of the
interface surfaces completely nonreflective - have one "liquid" texture for
the liquid-glass interface and another, complete with reflection (and
ripples?) for the top.
IWBNI (say) POV4.X had an additonal union-like CSG operation as well as
union and merge, which indicated that a ray that left one body and entered
the other within some threshold distance was to be treated as having only a
single interface. Other possible ways of doing the same thing would be a
modifier for union; changing union itself on the grounds that nobody ever
seems to *want* the messy ambiguity; or an "asymmetric union" that clips
the second body at the edge of the first, so that you just need to make the
liquid a tiny bit larger than the hole it occupies.
-Robert Dawson
Post a reply to this message
|
|
