Re: ? for image digitizers

[John Roberts <roberts@xxxxxxxxxxxxxxxxx>]

>Date: Thu, 7 Mar 1996 20:28:26 -0600
>From: telscope@xxxxxxxxxx (Peter Abrahams)
>Subject: ? for image digitizers

>Recent questions on the possibility of stereo pairs of galaxies truly
>stimulate the
>imagination.  Those who have seen a large, high-contrast, well focused
>photograph
>of a spiral galaxy can imagine how impressive a stereo would be.
>Most recent astronomical images have been digitized, either by exposure on a
>CCD chip or by scanning a photograph into software such as PhotoShop.

Note: CCD imagers are analog - you still need an A/D conversion at some point.

>I know that there have been paintings & other images made into stereos via
>the computer.
>If I were to obtain a digitized version of a photo, probably a very large
>digital file, and 
>also copyrighted, what expense would be involved in making it into a stereo
>pair?

It depends on how "true to life" you want it. Some people make a "stereo pair"
of a flat photo by sticking a window in front, an arbitrary background behind,
or both - an image of M31 floating above second base in Yankee Stadium wouldn't
be too hard to produce, but it wouldn't be a very convincing photo either. :-)

A truly plausible stereo pair that places the galactic features where they
are believed to be in 3-space requires extensive analysis of the original photo
to produce a 3D model - redshift/blueshift of individual regions, distance
estimates to variable stars, etc. I expect this has already been done for
some galaxies. But I don't have an estimate of how much it costs to convert
a 3D computer model into a stereo pair. A wild guess would be several thousand
dollars for the first one, including labor. But once the technique is proven,
costs for subsequent images may go down considerably. This sounds like 
something that might be worthwhile for a school of astronomy to attempt.

John R.


------------------------------