Re: Medium format scanners
Re: Medium format scanners
- Subject: Re: Medium format scanners
- From: David Scharf <email@hidden>
- Date: Thu, 27 Oct 2005 06:00:58 -0700
Hi Ernst,
Epson does list their optical and hardware resolution at 4800; the
stepper motor for the array is 9600 lpi. More than enough for medium and
large format neg scanning. The SilverFast software makes a big
difference in realizing the capability of the 4990. I can see every
grain in my negs and every scan line from the film recorder they were
recorded on. And as your probably aware, according to Nyquist theorem it
takes at least a 2x oversampling to properly define any feature (but in
my experience, I prefer at least 3X). Anyway the point is not to
quibble, but just that if the optical dpi measured by reviewers or
testers differs from the stated 4800, it won't be by much according to
what I'm seeing.
Regards,
David
DAVID SCHARF PHOTOGRAPHY
Scanning Electron Microscopy
Ernst Dinkla wrote:
The 4990 has a 4800 ppi optical resolution, not 2400. For $599 with
SilverFast Ai software, its one of the best deals out there (for my
purposes anyway). If your scanning 35mm, get a dedicated scanner. If
its 127, 120, 6x9, 4x5, up to 8x10, this is a great little scanner.
Talking about negs and transparencies here. You may occasionally have
a problem with a very dense neg, since the Dmax is only 4.0, but it
handles all of my S.E.M. images on Plus-X or FP-4 so far; and these
can have very troublesome highlites. Unfortunately, I'm still
scanning and doing a lot of my graphics in OS 9. (New G5 some day).
David,
The 4990 has a 4800 SPI = sampling rate per inch. The true
optical resolution in PPI has to be measured in tests. That
Epson mixes up the terms in their specs isn't new, almost all
the scanner manufacturers do and it isn't unusual to see even
DPI mentioned in scanner specs or tests.
With (non-matrix) linear CCDs where the wells of the sensor
are packed to one another it is simple to find the sampling
rate when the sensor specs are known:
Horizontal axis: 1 / sensor well pitch in mils x enlarging
factor.
Vertical axis: stepping rate per inch x enlarging factor.
With the right optics and good mechanical components the true
optical resolution could be close to the SPI then, it usually
is lower.
The Microtek Sigma Six sensor and the Epson matrix sensors
however use more rows of linear CCDs shifted to one another by
half a well pitch or 1/3 a well pitch so the well size is
bigger than the pitch. Other scanners use a stepping rate
that's half the well size. In both cases the SPI is 2x or 3x
higher but the resulting PPI will not make a similar jump. On
the other hand this delivers a better Dmax - Dynamic range as
the wells are bigger compared to the higher resolution sensors
with the same total outer dimensions, so it is a sound concept.
I have both the Epson 3200 and the Nikon 8000. Slightly older
models but with the same internal designs. The Nikon specs are
much closer to the actual optical resolution than the Epson
specs. As tested by some German magazines (actually tests by
Dietmar Wueller of Image Engineering). The Epson 3200 scored
1200 PPI despite the much higher Epson specs, the Nikon 8000
was rated something like 3800 PPI. Don't expect of an Epson
4990 that its true optical resolution will be better than what
the Nikon 9000 can deliver. With a lot of work you can bring
it quite close though.
--
Ernst Dinkla
www.pigment-print.com
( unvollendet )
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden