Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!olivea!samsung!spool.mu.edu!mips!zaphod.mps.ohio-state.edu!wuarchive!mit-eddie!media-lab.media.mit.edu!halazar
From: halazar@media-lab.media.mit.edu.MEDIA.MIT.EDU (Michael Halle)
Newsgroups: comp.graphics
Subject: Re: Digital Holography
Message-ID: <5817@media-lab.media.mit.edu.MEDIA.MIT.EDU>
Date: 10 May 91 21:31:16 GMT
References: <1991May7.215514.6676@csl.dl.nec.com> <261@rins.ryukoku.ac.jp> <1991May9.153446.21742@leland.Stanford.EDU> <1991May10.165256.12414@nas.nasa.gov>
Organization: MIT Media Lab, Cambridge, MA
Lines: 41
In-reply-to: uselton@nas.nasa.gov's message of 10 May 91 16:52:56 GMT


"Rendering with/like holography" (more correctly, rendering at the
light wave level) is a little like building macroscopic structures out
of individual molecules: sure, you could imagine doing it, and with
enough effort you could, but in the general case you might be able to
think of a better way.  In a similar way, natural phenomena can be
considered solely in terms of quantum effects, but classical mechanics
generally work pretty well and are much less painful.

What level of realism are you trying to achieve that would require
such accuracy?  It's a lot of work to go to just to spatially quantize
to x by y (by z?) pixels.  The display (and the eye itself) usually
define the reasonable limits.  Perceptual restrictions are *always* an
issue in display.

Sure, computing holograms can be expensive.  Holograms (usually)
contain more information than do two-dimensional images, so computing
them *should* take longer.  And the expense of the calculation is
proportional to the stupidity of the approach (and not linearly,
either), especially if you're a purist and insist on diffraction
limited three-dimensional images.  But if you make those tradeoffs
like they taught ya in engineerin' school and don't do more work than
you have to, computation time for 3-D images might be only, say, ten
times that for 2-D images.  And you might be able to actually make
holograms instead of dreaming about them.  (And a previous poster was
right; unless there's some physics that we don't know about, mid-air
projection of 3-D images, with no display material in front or behind,
is right out.)

Here's a little thought question that may shead some light on the
comutation question:  What is the intrinsic information content of a
pure sine wave oscillating at an arbitrarily high frequency?  Think
about the relative costs of the different ways of producing such a
sine wave.  Would your answer differ if the signal were specified to
be analog or digital?


					Michael Halle
					Spatial Imaging Group
					MIT Media Laboratory
					mhalle@media-lab.media.mit.edu