Path: utzoo!attcan!uunet!husc6!rutgers!att!alberta!calgary!cpsc!jevans
From: jevans@cpsc.ucalgary.ca (David Jevans)
Newsgroups: comp.graphics
Subject: Re: Ray tracing vs. Scanline algorithms
Summary: that depends
Message-ID: <78@cs-spool.calgary.UUCP>
Date: 19 Oct 88 10:44:57 GMT
References: <3023@mipos3.intel.com>
Sender: news@calgary.UUCP
Distribution: na
Lines: 36


At the University of Calgary we have used an abuffer scanline renderer
for several years.  Before that we used a pretty much bog standard
zbuffer algorithm.  Scanline algorithms are used because they are
usually faster than things like ray tracing.  The reason people use
raytracing is not for its ease of coding.  In fact, to write a fast
ray tracer requires a lot of thought, research, and often some
difficult coding.

Ray tracing is used primarily to get shadows, reflections and refractions.
Of course these can be faked pretty well for scanline algorithms though
the use of light/shadow buffers and environment maps, although it is my
opinion that ray tracing is more elegant and can be made faster than
a scanline method which uses all these tricks.

Recently I have been working on new spatial subdivision methods to
speed up ray tracing and I have several test images which do not take
much longer to ray trace than to render with our abuffer renderer.  These
images consist of large(?) numbers of polygons ( > 20000) and are rendered
with shadows and multiple light sources while the abuffer renderer has
no shadows at all.

You will get varying arguments from everyone.  I think scanline methods
are probably still faster for generating images with textures and that
use environment maps, although you will need to ray trace to get nice
refractions.  Of course my big argument for ray tracing is that it
lends itself to parallelization so easily.  Try arguing with a dedicated
scanline freak as to the usefullness of this though...

"we're dancing down church aisles, dancing on holy books..."

David Jevans, U of Calgary Computer Science, Calgary AB  T2N 1N4  Canada
uucp: ...{ubc-cs,utai,alberta}!calgary!jevans

David Jevans, U of Calgary Computer Science, Calgary AB  T2N 1N4  Canada
uucp: ...{ubc-cs,utai,alberta}!calgary!jevans