"Christopher James Huff" <cja### [at] earthlinknet> wrote in message
news:cja### [at] netplexaussieorg...
> In article <3f871c85@news.povray.org>,
>  "Hughes, B." <omn### [at] charternet> wrote:
>
> > Also, only has self-shadowing on a per primitive basis too. In a
> > union of sphere and cone, for example, each doesn't cast a shadow
> > onto  the other.
>
> I'm not sure if you're describing what it does or what it should
> do...are you saying look_like objects cast shadows on themselves but not
> on other objects?

Themselves, yes; on a per individual primitive basis, having only a shadowed
side versus lit side. So what I was saying is that if a sphere and cone are
next to each other, in a union, they don't cast a shadow onto each other. I
guess your explanation about bounding below is the reason why.

> Unions are kind of tricky...they are internally split up into separate
> objects to make bounding more efficient. Manually bounding the union
> will disable this behavior unless the split unions option
> (Split_Unions=on or +SU for non-Mac versions) is turned on, in which
> case even bounded unions will be split. This burned me when I was
> experimenting with the proximity and curvature patterns...I have no idea
> how it interacts with looks_like, I'd need to take a close look at the
> code.
>
> Anyway, I'd suggest you just use a union of the light source with the
> object representing the light. This gives you most control over what
> actually happens...you can make specific parts shadowless, etc. If we're
> going to talk about what looks_like *should* do, I think it should
> behave in exactly the same way as a union with the light source, but
> changing it at this point would probably cause more problems than it
> solves.

I'll agree with that. Many times I've passed on using a looks_like and used
union with a light source instead, with the idea being to get more control
of the object appearance. I know people must've suggested a 'shadow on',  or
'no_shadow off', method before, for doing things like this, instead of being
stuck with the shadowless invocation caused by looks_like. Therefore being
able to force shadowing within an otherwise no_shadow CSG object. I can
imagine it being of some use. Due to the way all other light sources fail to
interact properly with the current way looks_like works.

Just to give it a try, I checked what you suggested about manual bounding
within the CSG union, except still for a looks_like object, and still aren't
any shadows. I think it's been said that looks_like is only a shortcut to
creating a light plus object, not the only way available and perhaps that's
why no one has deemed it necessary to change. Or, as you said, there might
be some difficulties is allowing shadowing via the internal bounding
hierarchy?

-- 
Bob H.
3digitaleyes.com

Post a reply to this message