Message-Id: <mailto:199408191333.IAA03993@library.wustl.edu> Date: Fri, 19 Aug 1994 08:18:00 CST From: Paul R Montague <mailto:Montague-Paul@OPHTHALMOLOGY-PO.OPHTH.UIOWA.EDU> Subject: Archival Color... To: Multiple recipients of list IMAGELIB <mailto:IMAGELIB@ARIZVM1.BITNET>
Hi imagelib...
I monitor this list in digest mode. The following comments came from
yesterday's discussion.....
"Firstly, the "slides" provide a second archival medium for a
record. Secod/Second, they are inevitable a higher resolution
than anything that you could dream of digitizing. The particale
size /resolution of even 35mm film far exceeds the 4000 pixel
resolution of "high-res " digitization. If you made 4x5 slides.
The comparison would be so extreme that the comparison is sort of
ridiculous. Third, the portability and ease of use. There are
many circumstances in which 35mm slides are simply easier to use,
ie., submissions for publication, presentations at conferences,
etc. Fourth, 35mm slides are easier to get digitized than
odd-sized paper. There are all sorts of slide scanners as was
pointed out in previous discussions.
Ektachrome and other E-6 films (Fujichrome, Agfachrome, etc) will last
around 30 years before the dyes begin to degrade. Kodachrome (K-14)
will last slightly longer, perhaps in the 50 year range. All color
film dyes fade with time, so I believe it is difficult to ever call a
color photograph "archival".
We all know that properly processed and stored black and white silver
images can survive for over one hundred fifty years, since we have
images which are so old. Relatively permanent color images can be
made with three black and white separation negatives instead of one
color negative (or positive). The old Technicolor movies were done
this way... Look at a new print of the Wizard of Oz, which was filmed
with the old Technicolor system on three black and white separation
negatives. The colors are as vibrant as ever.
The resolving power of Kodachrome 25 is in the range of 125 line pairs
per millimeter with a 1000:1 test object. Mathematically, that works
out to about 6000x9000 pixels. However, film resolution is not
measured in a way that can be related to pixels directly. The more
important question is "At what pixel resolution can an observer no
longer differentiate between digital and film resolution". A study
has been devised to address that question. A collection of images are
being distributed among 600 observers. *Preliminary* findings
indicate observers loose their ability to distinguish with digitized
resolutions beyond 1024x1536 pixels.
This isn't an argument for or against anything. Just some thoughts
that might be useful in making decisions...
mailto:paul-montague@uiowa.edu