Red Edge Problem

  • Thread starter Thread starter Mike - EMAIL IGNORED
  • Start date Start date
M

Mike - EMAIL IGNORED

Using my Nikon 5000ED scanner, and then examining
the result with Photoshop, the short edges
sometimes have a reddish "border" most intense
at the very edge and fading as distance from
the edge increases. I cannot see these on the
film, but I have yet to examine the film with
good optics. I considered light contamination,
so I scanned with the room lights off. No
difference. Any idea what this might be?

Thanks for your help.
Mike.
 
Using my Nikon 5000ED scanner, and then examining
the result with Photoshop, the short edges
sometimes have a reddish "border" most intense
at the very edge and fading as distance from
the edge increases. I cannot see these on the
film, but I have yet to examine the film with
good optics. I considered light contamination,
so I scanned with the room lights off. No
difference. Any idea what this might be?

The LEDs are not perfectly aligned which - in a way - is to be
expected as they can't occupy the same space. I've noticed this
misalignment (LS 50 here) a while back by comparing individual
channels in Photoshop and was somewhat surprised. Kennedy - a well
known expert in these parts - explained why that is.

However, unlike you, I do not see the reddish border as such when
viewing the image with all channels on. I only notice the misalignment
(in the sub-pixel range!) when flipping between individual channels
and this is throughout the image. At 100+ magnification the image
clearly "moves" when the channels are flipped.

On a tangent, I spotted something similar on my flatbed, especially on
thick photographs. The leading edge has a distinct blue tint while the
trailing edge has a distinct red tint. This, again, is apparently due
to the three sensors/filters/etc not being in the same place so the
light reflects at a slightly different angle for each one.

Are you noticing a similar blue edge at the other end? If so and/or
the effect is really objectionable it may very well be that the
scanner needs to be serviced by Nikon.

BTW, one good test is to scan B&W film in color. That's how I spotted
the misalignment in the first place. I never noticed it with color
film.

Don.
 
The LEDs are not perfectly aligned which - in a way - is to be
expected as they can't occupy the same space. I've noticed this
misalignment (LS 50 here) a while back by comparing individual
channels in Photoshop and was somewhat surprised. Kennedy - a well
known expert in these parts - explained why that is.

However, unlike you, I do not see the reddish border as such when
viewing the image with all channels on. I only notice the misalignment
(in the sub-pixel range!) when flipping between individual channels
and this is throughout the image. At 100+ magnification the image
clearly "moves" when the channels are flipped.

On a tangent, I spotted something similar on my flatbed, especially on
thick photographs. The leading edge has a distinct blue tint while the
trailing edge has a distinct red tint. This, again, is apparently due
to the three sensors/filters/etc not being in the same place so the
light reflects at a slightly different angle for each one.

Are you noticing a similar blue edge at the other end? If so and/or
the effect is really objectionable it may very well be that the
scanner needs to be serviced by Nikon.

BTW, one good test is to scan B&W film in color. That's how I spotted
the misalignment in the first place. I never noticed it with color
film.

Don.

Thanks for these leads.

Because of imminent travel, I cannot spent much time on this now.
In October, I'll dig out some old B&W and try your suggestion.

For now:

I have never seen a blue border.

I just measured a border, it covered about 4% of
the length of the frame. On several occasions, I
was forced to use an otherwise suboptimal crop.

Mike.
 
Thanks for these leads.

Because of imminent travel, I cannot spent much time on this now.
In October, I'll dig out some old B&W and try your suggestion.

For now:

I have never seen a blue border.

I just measured a border, it covered about 4% of
the length of the frame. On several occasions, I
was forced to use an otherwise suboptimal crop.

Mike.

In my case it was not a function of image size and definitely far
below the 4% mark. It was only about 10-15 pixels deep, so it may very
well be you're dealing with something else. Especially since in my
case that was on a flatbed and not a film scanner. Anyway, good luck!

Don.
 
Back
Top