Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!seismo!rutgers!ucla-cs!zen!ucbvax!jade!saturn!skinner
From: skinner@saturn.ucsc.edu (Robert Skinner)
Newsgroups: comp.graphics
Subject: Re: Rainbows (was Re: Ray tracing and caustics.)
Message-ID: <583@saturn.ucsc.edu>
Date: Fri, 7-Aug-87 17:06:35 EDT
Article-I.D.: saturn.583
Posted: Fri Aug  7 17:06:35 1987
Date-Received: Sun, 9-Aug-87 20:42:34 EDT
References: <219@sugar.UUCP> <543@saturn.ucsc.edu> <1827@vax135.UUCP> <439@sugar.UUCP>
Organization: U.C. Santa Cruz, CIS/CE.
Lines: 25
Keywords: ray-tracing caustics algorithm reality
Summary: Rainbows have been done

In article <439@sugar.UUCP>, peter@sugar.UUCP (Peter da Silva) writes:
> OK, I think I have come up with something the rendering equation won't
> cover. Chromatic abberation. You know... the stuff that makes rainbows
> colorful and diamonds sparkly. 
> 
> I don't remember seeing this one described anywhere. Does this get me
> a citation :->?

Rainbows were done by Ken Musgrave here at UCSC this past spring.
He basicly extended the distributed ray tracing concept to include sampling
the frequency of light.  Rays were tagged with a frequency when they
hit a refracting object, and the resulting illumination values were filtered
by the color of the given frequency.  This works for prisms and diamonds,
etc.  The rainbow position is easily caluculated by a dot-product and table 
look-up.  

This does not handle rainbows cast on a wall from a prism...
That is really a multicolored caustic....

.... and the caustic debate continues ...


----------------
Robert Skinner
skinner@saturn.ucsc.edu@csnet-relay