Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
LIsters,
Replying specifically to Julio's questions about printing: the
highest screen mesh used by journals -- including scientific journals
publishing micrographs -- is typically 171 lpi ("lines per inch"),
but can be up to 300 lpi. Two (of many) sites about this issue are:
http://desktoppub.about.com/library/weekly/aa101800e.htm
http://www.sos.com.au/info/dpi%5Flpi.html
Any computer book on Photoshop or desktop publishing or imaging with
cover this as well. The basic rule of thumb is to size images at
twice the resolution at which they are to be printed, but no bigger.
That is, an image to be printed at 171 lpi should be sized at a
resolution of 300 dpi. The next step is 600 dpi, and that's just too
big -- lots of wasted file space and slower printing. 600 dpi may be
useful for publication sometimes, but we do OK with 300 dpi in
Microscopy Today.
This is for journals and the like. Printing for posters, hard-copy
images, etc. is different, and if you have the time and high-quality
glossy photo-print paper, then the higher the resolution (more dpi),
the better the print. Paper is really the limiting factor here, and
high-quality photo paper can be expensive, but worth the price.
Images for web pages are displayed at 72 dpi, so there's no value in
saving such images at over 150 dpi, unless you want visitors madly
scrolling about -- like for the high-res Space Telescope images at
the Hubble site.
Finally, there's another problem with color images: color space or
gamut. Since printing uses a four-color process to simulate the full
palette, the colors available for printing are not the same at the
colors displayed on computer monitors, on silk-screened t-shirts,
photographs, etc. This is also true for colors printed with different
printing technologies. So a wonderful color photograph that displays
well on a web page may look horrible when printed, unless this color
balance problem is allowed for.
But this is irrelevant to dpi, lpi, and the like. It is in Photoshop
and publishing books.
But! It's always best to collect the images at the maximum resolution
possible. This resolution can be reduced to meet printing needs, but
the more resolution, the more data. This also allows for printing
subareas of an image at the maximum printer resolution.
Phil
Jeff:
It is frustrating indeed that the prices of dedicated micrograph
systems should be so high while their performance generally lags far
behind that of much cheaper consumer digital cameras.
I have used my Nikon Coolpix 950 (< $ 1,000) for photomicrography.
Just jack up the illumination on the microscope a bit, rest the
camera's lens lightly on the scope's ocular (not even the trinocular
tube), let the autofocus do it's thing, and shoot. Surprisingly good
results with no mod at all. There's the occasional blurry micrograph
from shaky hands, but it's easy enough to check the pic (digital
magnifier available on the camera screen) and reshoot. Will be glad
to send you sample micrographs; You tell me how compressed you want
them, at normal compression average size is 400Kb. Resolution is 2
megapixels on that model, I think current models run 3 megapixels
and above at comparable prices. At any rate, running the 2 Mp
micrographs through Photoshop's (or similar software) contrast /
color / range adjustments will often bring out detail that was not
obvious by eyeball. You can even obtain Nomarsky-like enhancements
from a plain vanilla micrograph, using Photoshop's "layering" and
"embossing" features. It would seem that a 2 or 3 Mp camera will
capture about as much information as you can eyeball, save for the
inability to up/downfocus.
I've heard a few other personal cameras can do the same job;
Seems the effective optical diameter of camera's front lens element
should be a certain size to get a good chunk of the scope field with
little wasted area. And the Nikon's jointed-body design
does facilitate preview.
I can think of several ways to improve the arrangement if it's to be
used more frequently:
1) Make a proper stand for the camera. An old stand from a darkroom
enlarger might fit the bill. Use the camera's timer to reduce
jerkiness from hand-pressing the trigger, or get a remote trigger if
available on current models.
2) These Nikon models have a thread on the lens mounting (most other
brands don't have it). You can get adapters at camera shops. Have a
metal tube made to slide snugly around the scope's ocular (or
trinoc), bond the tube to the adapter, thread the adapter on the
camera and you should have a fairly stable arrangement that can be
set up quickly. Not sure if camera's weight may slightly distort
optical path or render the scope a bit unstable physically. An extra,
regular ocular placed on the trinoc tube should be much better than
loading up one of the binocular tubes, and handier in many ways.
3) I once located on the web a German (or Austrian??) firm supplying
an adapter for the trinocular tube - $ 750, I think. Sorry,
didn't save web address, it came up after a google search, probable
keywords were: nikon coolpix microscope adapter.
Cameras are more usefully rated by megapixels rather than dpi, as the
former give you a true measure of the amount of information that may
be captured. At time of printing or displaying, Mp are translated to
image size by choice of dpi resolution. A 2 megapixel image at full
resolution on your computer screen (about 80 dpi at the 800 x 600
setting on a 10" by 7" screen) would be about four times the screen's
area and so must be seen by scrolling or displaying at reduced
resolution.
As to printing: If you print at 600 dpi, a 2 Mp image would be
sized roughly 2 x 3 inches. However, I don't think actual resolution
of journal & magazine grayscale or color pics even approach 300 dpi,
so a much larger image might possibly be printed without any blurring
above what's inherent in the printing process. I **think** this
depends heavily on the printers' particular color separation /
"screen" process, with which I'm not familiar. Perhaps a journal
editor or fellow crust-ler can enlighten us in this regard.
As to computer interface: I've seen a Kodak model that has a much
more versatile (real-time preview and remote control!) interface than
my particular Nikon - which downloads quite slowly, via serial cable.
Alas, that Kodak model did not capture the microscope field
satisfactorily. Unless Nikon has changed it's interface, I'd
recommend getting an accessory that will allow you to plug the
camera's memory device (flash card on mine, perhaps memory stick on
current models) into the diskette drive or USB port on the computer -
much faster downloads. Might be easier to find in computer places
than camera shops. Also, my cam had only an 8Mb flash card standard;
Recommend buying a 32 (or 48?) Mb extra card, much more like what
you'll need.
Hope this helps, and would appreciate learning of any new findings.
Regards
Julio
----- Original Message -----
From: <mailto:[log in to unmask]>Jeff Shields
To: <mailto:[log in to unmask]>[log in to unmask] ;
<mailto:[log in to unmask]>[log in to unmask]
Sent: Monday, March 11, 2002 5:20 PM
Subject: [CRUST-L:1485] digital photomicroscopy
Crust-Lers, Marine_pathols,
I'm interested in obtaining a digital photomicrography system that
shoots at 600 dpi or better and that preferably has a good computer
interface. I currently have a ccd that gives about 150 dpi at best
and am looking for better resolution. And of course I'd prefer this
at <$5,000 :-) Any suggestions?
Cheers, Jeff
--
Philip Oshel
Technical Editor, Microscopy Today
P.O. Box 620068
Middleton, WI 53562-0068
Daytime phone (608) 263-4162
Fax (608) 836-1969 (please make sure my name is on any fax)
email: [log in to unmask]
Address for UPS, FedEx, or other couriers:
Department of Animal Sciences
University of Wisconsin
1675 Observatory Drive
Madison, WI 53706 - 1284
|
