[filmscanners] Re: Color Misregistration (Especially PolaroidSprintscan 120)

["Arthur Entlich" <artistic-1@shaw.ca>]

The design of the Nikon LS-8000 does avoid color misregistration, but it
is that same design that causes the banding problem, since all three
color separation scans of any one "line" of the image are done by the
exact same CCD sensor line, so any misregistration due to stepper motor
problems, slippage, or film distortion can cause a banding problem.

The SS120 overlays the color separations from different CCD sensor lines
taken at slightly differing times of scanning, so this "blends" the very
slight (usually under one pixel) variation in movement.

  Once again, I must explain about the use of collimated lighting of the
Nikon (although apparently Nikon does slightly diffuse the lighting in
the LS-8000) versus the very good use of diffused lighting in the
Polaroid scanners. Polaroid's design vastly reduces the visibility of
grain edges, grain aliasing, dirt, dust and scratches, making dICE
pretty much unnecessary, where it is mandatory with the Nikon scanners.
  Unfortunately, when it comes to film that dICE cannot be used with
(films which have high levels of residual silver, such as true black and
white films and some Kodachromes) you have no dICE to help you, but you
still have the same problem with the lighting causing significant grain
exaggeration and the same for dust, dirt and scratches.

To appease the people who have been brainwashed into thinking they "must
have" dICE, mainly due to people using scanners that really require it
warning people how necessary it is, Polaroid developed a Dust and
Scratch filter which is quite effective, but usually completely
unnecessary.  It is available free for PC users, I don't know if a Mac
version will exist or not.

Yes, the Polaroids also can show up some dust and such. but it is
usually pretty obvious stuff, which should be cleaned off to begin with.
  A few minutes of cloning usually fixes it all.  If you have enough
man-handled images, with scratching, or fungus or fingerprints, then
dICE is a real godsend.

The other side of the LS-8000 is that the poor depth of focus pretty
much requires a glass carrier be used for medium format images, another
considerable expense, AND a producer of Newton rings (which I happened
to see quite obviously, smack in the blue sky of one of the few image I
looked at on Anthony's website some time back).

Don't make any decision before you have your SS120 serviced by Polaroid.
  What you are seeing now is simply NOT normal, and I'd expect was the
result of some severe mishandling during shipment.

Art


Anthony Atkielski wrote:

>>I'm interested in hearing about how good color
>>registration is on filmscanners in general, especially
>>on the Polaroid Sprintcscan 120, which I just bought,
>>and the Nikon 8000, which I apparently should have
>>bought.
>>
>
> I've never seen any error in color registration on the 8000 at all.
>
> I'm not sure of how either of these scanners handles colors, but some
> designs would not permit misregistration.  I don't know if the Nikon design
> is in that category, however.
>
>
>>On my Sprintscan I am getting color misregistration
>>of up to 8 pixels (at 4000dpi) in the scan direction.
>>
>
> My guess is a stepper motor problem, or the equivalent.
>
>
>>The Infrared dust/scratch feature on the Nikon
>>holds little charm for me; multi-sampling is cool,
>>but not $1000 worth of cool.
>>
>
> Digital ICE can be a godsend for dusty or dirty slides or negatives.  When
> you need it, you need it badly.  If it saves you a couple of hours
> retouching on just one assignment, it has paid for itself.  The normal
> setting of ICE seems to have no effect on resolution.
>
> Multisampling buys very little, as the Nikon already sees pretty well into
> the shadows.  I usually leave it set at 4x.
>
> The worst issue with the 8000 is banding, which is sometimes visible in very
> dark areas adjacent to very light areas.  You can eliminate it by switching
> to a "superfine" scan, which uses only one CCD line instead of all three,
> but the scan takes three times as long.  I've only found it worthwhile to do
> this for occasional night shots where banding is obvious enough that I'm
> willing to wait longer for the scan.
>
>
>
>
>


----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 
body