Re: filmscanners: SS120 & Nikon 8000 ... how do they work?

[Arthur Entlich <artistic@ampsc.com>]

Lynn Allen wrote:
> 
> Art wrote:
> 
> >It seems to me for some reason that most of the newer medium format
> >scanners manufacturers decided to forego the zoom lens approach
> >that Minolta has and continues to use with their Multi models,
> >and just basically use the same optics for all the film formats.
> 
> I seem to be missing something, here. All my scanners have "zoom"
> capability, but on the screen, only--they can make the image bigger, just as
> I can with my own imaging software. Can a scanner actually enlarge an area
> with an enlarging lens? Well, why not--I can do it with a loupe, can't I?
> 
> So let's say I select a 1/4" x 1/4" section of a slide that would normally
> be scanned at 2400ppi on my Scanwit--if a 2x scanning lens were available,
> would this then appear like a 1/2" x 1/2" section of the slide, at 2400ppi?
> Given this, it would seem that the resolution of the section would become
> 4800ppi.
> 

Yes, this is in fact exactly what I am speaking of.  Minolta does this
on a small scale with their Multi scanner line.  You might think of the
CCD as a projection screen (a very narrow one, but that isn't
relevant).  If you could project a cropped portion of the frame upon the
whole CCD, you could still maintain the "native resolution" of the CCD.

Many moons ago, I was working on the concept of a system to allow a 35mm
frame to be projected on a flatbed scanner surface.  This could, in
theory, allow for even a 600 dpi scanner to record a 35mm frame at about
4800 x 7200 ppi, optically.  The problem was dynamic range and figuring
out a way to create a evenly distributed illumination without using some
type of back projection media over the scanner bed to allow for a focal
point. Anyway, I never got very far with it, ultimately.


> Somehow, I don't see this as likely, tempering it with a bit of skepticism
> for mfgr specs. Can a 4000ppi scanner in fact boost its effective scanning
> resolution to 8000ppi, or did it boost it to 4000 in the first place with a
> magnifying lens, from which point it cannot improve?  Inquiring minds seem
> to want to know. :-)

No, the issue is always film size versus number of CCD elements.  This
(in theory) is no different that the idea of converting an image in
Photoshop to a printer.  Just count pixels.

If you could take a 1" wide film frame and distribute it across 4000
sensor elements, you get a 4000 ppi scan.  Now, if you instead use only
.5" frame across that same 4000 sensor elements, you could indeed end up
with that portion having 4000 ppi,, and since it was .5 inch, that means
effectively the scan was at 80000 ppi, as you stated.

The problem is no scanners currently offer the optics or control to
allow for this to occur.  Minolta's Multi comes closest, using set
magnifications based upon the the film size.  I imagine it would require
some pretty sophisticated focusing to do what we are discussing above.

Art