Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!wasatch!helios.ee.lbl.gov!ncis.tis.llnl.gov!blackbird!news
From: news@blackbird.afit.af.mil (News System Account)
Newsgroups: comp.graphics
Subject: Re: What's a hologram?
Message-ID: <1311@blackbird.afit.af.mil>
Date: 26 Aug 89 01:51:43 GMT
References: <4791@portia.Stanford.EDU> <1746@dover.sps.mot.com> <591@mit-amt.MEDIA.MIT.EDU> <681@berlioz.nsc.com>
Reply-To: tmouser@blackbird.afit.af.mil (Tommy A. Mouser)
Organization: Air Force Institute of Technology; WPAFB, OH
Lines: 35
Summary: What takes so long.

In article <681@berlioz.nsc.com> andrew@berlioz (Lord Snooty @ The Giant Poisoned Electric Head ) writes:
>
>The article by Michael was great. I wonder if the slowness of the computer-
>generated method is limited by compute MIPs or by the electron beam-writing
>speed?

I think the problem is with computer MIPs.

example:
you wish to make a 1" by 1" CGH.  you choose to illuminate it with a
HeNe laser (wavelength about .5 microns).  During the calculation of
the intensity values you would use a 2-D array consisting of 50,800
elements on a side.  Thats a total of 2,580,640,000 elements in the array.
(The array represents the hologram plane)

BTW, 50,800 * .5 microns = 1".  Now, for each point on the object you
must calculate the light intensity from that point to every point on
the hologram plane.  The calculation involves taking a Sine, Cosine,
several multiplies, etc.

So, if there are 1000 points describing the object, you've got some
2 trillion very complicated calculations to make.

end example.

I've calculated it would take a Sun 4 about 280 days to make this
calculation.  A Cray2 about 6 days.

All of these numbers change drastically if you choose to convert
the problem such that Fourier analysis is applicable.  But, this
doesn't seem to work too well for 3-D holography.


tom
.