RE: scanner quality when profiling?

["Steve Jacobson" <stevelists@jjworld.com>]

These are some really good questions.  Here are some of my observations.

Scanners typically use either cold cathode lighting or a modified
fluorescent tube lighting.  The nature of the light generation process in
these tubes produces large amounts of UV.  What the tube manufacturers try
to do is to convert that UV into visible light using the phosphors in the
tube.  However, that conversion isn't 100%, and a certain amount of UV
remains.  UV is generally considered to be wavelengths below 400 nm.
Looking at the spectral characteristics of the C50 and SPX50 fluorescent
tubes, GE makes a good attempt to reduce emissions below 400 nm.  However,
it doesn't go to 0 until about 375 nm.

Fortunately, glass is not a very good transmitter of UV, and I would expect
the scanner glass to reduce the UV component even further (but probably not
eliminate the higher-frequency UV).  So I would expect that a certain small
percentage of UV (400 nm - 375 nm) gets through and might cause some small
amount of fluorescence of the scanned material.  I don't believe UV below
375 nm even gets through the glass.  I don't have a UV meter, otherwise I
would measure the UV from my scanner.  If someone else has one, I would be
interested in having them measure the UV.  I don't think fluorescence is a
big problem, but I haven't tested it.

I believe that the spectral characteristics of the scanner light source are
fairly well known, and calibrated to a standard color temperataure.  When
you go to replace the tube, the scanned image will look somewhat different
with the new tube than with the old.  However, scanner profiles should take
the tube spectral characteristics into consideration and compensate for
them.

The question gets even more interesting because all the 5000K light sources
discussed on this list and those I saw at the Seybold show (with the
exception of a few halogens) were fluorescent, and these will emit some UV
(even the sun emits UV, as those who lay out in the sun too long will
testify).  Thus, the viewing characteristics under either artificial of
natural daylight will probably cause some fluorescence of the inks, and
might be similar to the scanner.  Viewing though a piece of window glass
will probably reduce the UV to what the scanner produces, and "look" the
same to the scanner as it does to the viewer.

Just my $.02

-----Original Message-----
From: owner-epson-inkjet@leben.com
[mailto:owner-epson-inkjet@leben.com]On Behalf Of Laurie Solomon
Sent: Friday, February 18, 2000 4:23 PM
To: epson-inkjet@leben.com
Subject: RE: scanner quality when profiling?


One issue I haven't seen brought up in relation to scanner quality is the
light source used. Since all mixed color printing processes are metemaric to
some extent it would seem that the light source and the exact CCD filter
responses could have a considerable effect on the accuracy of profiles. For
example if the light source contained a large UV component then the
fluorescent whiteners in many bright white papers would be activated to a
greater extent than they would be under incandescent lights.

Even profiling packages that generate scanner profiles from known targets
would be affected since the exact dyes and spectral characteristics of the
target would be different from the prints being profiled so that a print and
a target that looked identical under D50 viewing conditions might appear
quite different to the scanner.

Do scanner manufacturers specify what viewing conditions the scanner
emulates?


-
Please turn off HTML mail features. Keep quoted material short. Use
accurate subject lines. http://www.leben.com/lists for instructions.