Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uwm.edu!rpi!batcomputer!eric
From: eric@batcomputer.tn.cornell.edu (Eric Fielding)
Newsgroups: comp.graphics
Subject: Re: Fractals, and Philosophy of Science
Message-ID: <9515@batcomputer.tn.cornell.edu>
Date: 9 Jan 90 02:35:48 GMT
Reply-To: fielding@geology.tn.cornell.edu
Distribution: na
Organization: Department of Geological Sciences, Cornell University
Lines: 83

Just a few comments on this (probably already too long discussion):

[sorry if the attributions are mangled]
In a recent article thomson@cs.utah.edu (Rich Thomson) wrote:
[...]
>In response, markv@gauss.Princeton.EDU (Mark VandeWettering) writes:
>>	Some graphics programmers are botanists, and geologists, and
>>	biologists.  And physicists and artists and students and.....

Well, I happen to be a geologist graphics programmer.

>Right.  That is why not every image need be generated from radiosity or
>ray-tracing techniques -- many times wire-frames will do quite nicely.

I find that moderately advanced rendering of topography shows me much more 
information than a wire frame.  Synthetic lighting makes much more subtle 
features visible.  My business is to analyze real data.  I have not tried any 
very advanced rendering, but I would guess there is a point of diminishing
returns.  As a matter of fact, I am mostly CPU limited as it is.  
(Anyone know how to render a 4000 x 5000 mesh of data ;-)

>:- If you want to completely and accurately
>:- model a mountain, you might as well sketch it by hand because the compute
>:- time necessary to model the complete developmental cycle of a terrain is
>:- prohibitive; 

Actually we are trying to model the development of terrains, but the problem
is more a lack of understanding of how mountains are formed and eroded to
formulate equations than a lack of computer time.

>>	Well, sketching it by hand implies that you have the terrain in 
>>	front of you, which is certainly not a certain matter.

I have the data in front of me, but it just looks like a computer tape to me.

>The point I was trying to make is that there are trade-offs between
>accuracy and time when modelling objects of the complexity of terrains.
>For cinematic or theatrical purposes, going to highly accurate geophysical
>models for terrains aren't worth it.

This trade-off is why we end up sub-sampling the data before rendering it.  I 
agree that it would be nice but not necessary to use real data to create 
terrains for movies or synthetic images.

>>	you are saying that fractals are good at creating cheap effects
>>	at a high level of visual complexity, then I agree.

Unfortunately, I am rather upset at the number of people taking fractals out of
the pretty-picture or effective animation graphics domain and applying them
to real data in my field.  People can find a 'fractal' in almost every
geophysical phenomenon.  This does not mean that it is necessarily a new 
contribution to the science.  In most cases, it amounts more to renaming 
a relationship that was already known, which seems a little silly. 

>>	If you try to claim that there is some subjective comparison between
>>	fractal models and the objects they mimic, then I might even yield.

There are some cases where a fractal model does a good job of describing 
reality.  For instance faults are really made up of fractal levels of smaller
and smaller scale faults.  My particular peeve is about fractals applied to
topography.  Somehow it just has not impressed me as a useful measurement,
other than the observation that most real topography has the same fractal
dimension.  So what.  How can we tell a recently created mountain from an old
one?  The fractal dimension is about the same.

>>	There are perhaps two aims to computer graphics:
>>		1. to "describe" a scene visually
>>		2. to "simulate" a scene visually

There are also two types of scientists:
  1. those who look at real data and try to observe how things work
  2. those who work from known principles and theorize how other things work

I put myself in category 1 for both graphics and science.  It takes all types.

>	  More Colombians are killed by American cigarettes
>		 than Americans by Colombian cocaine


Great quote.  
And how many Colombians are killed by American-made guns...

				++Eric Fielding