Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!zaphod.mps.ohio-state.edu!rpi!crdgw1!ge-dab.ge.com!tarpit!ucf-cs!wilson
From: wilson@ucf-cs.UCF.EDU (tom wilson)
Newsgroups: comp.graphics
Subject: Re: A Ray Tracer Accelerator: ???
Message-ID: <2016@ucf-cs.UCF.EDU>
Date: 30 Nov 90 07:35:39 GMT
References: <2010@ucf-cs.UCF.EDU> <4381@idunno.Princeton.EDU>
Organization: Univ. of Central Florida, Orlando
Lines: 24

In article <4381@idunno.Princeton.EDU> markv@acm.Princeton.EDU (Mark VandeWettering) writes:
[comments and analysis of value of inner bounding volume for shadows]
>But in general?  I dunno.

In general? I doubt it. Consider the "standard" bounding volume for a general
object. Do you make a bounding volume for a sphere? Na. How about a polygon?
Well, you can but I doubt you would use a sphere because there's too much
empty space.

My point: it's not any more straightforward to choose an inner bounding volume
than it is to choose the normal bounding volume. It may even be harder.
I have considered a special case of generating an inner bounding volume for
a convex object given that an outer bounding volume can be generated. It is
quick but not necessarily an efficient volume. Given the outer volume, you
can scale it so that it fits entirely within the object itself. This should
work since the object is convex. I have been unable to think of an example
where it wouldn't work (but that doesn't mean there isn't one, of course).

What I'm hoping is that someone already has a scene with a complicated object
that is really expensive to intersect a ray with since that is where the
idea would prove to be most useful. Obviously, finding a good inner volume
for this object may be a task.

Tom


Brought to you by Super Global Mega Corp .com