Re: filmscanners: Silverfast and LS1000

["Alan Tyson" <AlanTyson@bknowl.freeserve.co.uk>]

Frank,

I bet you're right, and this explains the wide divergence of
view here on Scanwit multiscanning accuracy which we've
discussed at least twice over the last year or so.

If using Herm's subtraction method to identify
discrepancies, the experiment to do is to compare several
different single pass images with a single multipass
composite. The smallest differences found would be the
nearest to the errors introduced by multiscanning, but this
still might exaggerate them, unless one of the single-pass
images happened to start in the same place on the frame as
the multipass one.

Does this make sense?

Alan T

----- Original Message -----
From: Frank Nichols <frank@theNichols.net>
To: <filmscanners@halftone.co.uk>
Sent: Tuesday, July 10, 2001 1:24 AM
Subject: RE: filmscanners: Silverfast and LS1000


> Actually scanning the same image twice is not exactly the
same as multipass
> scanning in Vuescan - I wish Ed were here to answer this,
but I recall a
> post he made where he described using the "course"
positioning stepper to
> position to the frame and the a "fine" stepper to
scan/multi-scan.
>
> So, If you actually scan twice the course stepper must be
used to position
> twice which is not as accurate.
>
> (maybe?)
>
> /fn
>
> -----Original Message-----
> From: owner-filmscanners@halftone.co.uk
> [mailto:owner-filmscanners@halftone.co.uk]On Behalf Of
Herm
> Sent: Monday, July 09, 2001 3:49 PM
> To: filmscanners@halftone.co.uk
> Subject: Re: filmscanners: Silverfast and LS1000
>
>
> the way to test is to scan the same image twice, then
subtract them from
> each
> other in photoshop.. if you get all black then its ok,
usually you will see
> a
> few defects. I will include an example so you can see it.
I inverted the
> image
> so you could see the differences.
>