RE: [APML]: Second HP v. Kodak Test

[Benjamin Gomes-Casseres <ben@alliancerevolution.com>]

Dave's comments sent me doing further tests. I also contacted HP tech 
support and they, too, had me do tons of things, including running 
"regular" negatives and slides to see if it still occurs. Yes, in every 
case where there is a bright white area on a dark area, the left edge of 
the white field has one or two rows of reddish pixels and the right edge 
has one or two rows of blue pixels. Of course, this shows up more (and 
matters more) on astrophotos, but it does occur in every kind of photo. It 
also shows up very clearly on moon pictures: the left edge of the moon is 
reddish, the right edge is bluish (with magnification).

The effect shows clearly on my M20/M8 image. I've now placed a horizontal 
"slice" of this image on my site, so you can scan across the whole 
negative. You'll see that the pattern is the same all across....except for 
one puzzling (vertical) area in which the Red-White-Blue pattern reverses 
direction (it's on the left 1/3 of my "slice")! Go to the bottom of the   
page at www.alliancerevolution.com/astro/scanner.htm


I am not an expert by any means. But it does appear to me like some sort of 
misalignment of either the sources or the recorders of the lightbeams. The 
HP tech told me there are no lenses as in a flatbed scanner, just 
"mirrors". Does anyone know how these machines work and what might be 
causing these rainbow edges? Is it my unit, or the design?

Also, it would be very useful to know if other HP scanner owners see the 
same effect on their units (if they look for it). I know there are many 
users out there. Please check and let us know.

Thanks for your help,

Ben

-----Original Message-----
From:   Dave Kodama [SMTP:kodama@alumni.caltech.edu]
Sent:   Thursday, January 22, 1998 8:51 PM
To:     astro-photo@nightsky.com
Subject:        Re: [APML]: Second HP v. Kodak Test


Ben,

Yes, very interesting

(snip)

Now the question in your case is whether or not it's a simple
mis-registration of the R,G,and B images (possibly a light
source positioning effect?), or actual optical chromatic
aberration caused by a lens system, assuming one is used
in this scanner when scanning film (it's almost certainly
necessary to scan the prints).

If it were chromatic aberration in the optics of the scanner,
I would guess that the middle of the scan would show little
or no effect, whereas the edges of the scan would show the
most pronounced effect.  By middle and edge, I mean along
any given scan row, which I would guess is horizontal in the
pictures you've posted.  Can you draw and post a diagram
showing what part of the 35 mm frame we are looking at and
how much of the frame each image covers?

This is very interesting.  We might be seeing a hint of how
the lower cost of a scanner is achieved.  Now I want to take
one apart myself!

Dave
---------------------------------