digital images of type specimens

[Peter Rauch]

>Date:         Tue, 22 Aug 1995 10:53:39 -0400
>From: "Shawn Landry (BIO)" <landry at CHUMA.CAS.USF.EDU>

>  I am in need of specific product info as well as information
>regarding the best (or most appropriate) techniques.

>The main conflict (as usual) is between resolution and affordability and
>usefulness of the images to the scientific community.

Shawn,
Several others have indicated that video imagery isn't going to give you
the same quality images as digital cameras or digitized photographs.
In general, for a given amount of money, and depending on what level
of quality you need, they are correct.

So, I'll toss in two cents on the Kodak PhotoCD as a relatively cheap,
and in my experience (contrary to another reporter's) excellent method.

First, you would have to decide that you can take photographs (color
35mm transparencies) that would meet your requirements. If the photos
won't, then no other technique has much chance of satisfying your needs
either.

If the photos work, then given that you have not taken them yet, the
cheapest way for you to get digital images, at about $1/each over the
price of the film, its development, and your time to photograph, is
to take a hundred pictures at a time (3 36-exposure rolls, more or less)
and have them developed and Kodak PhotoCd-processed all at one time.
That'll be about $115 out of pocket. For it, you get one cd-rom with
100 photos stored in each of 5 resolutions from about 100 dpi to 2000 dpi.
2000 dpi is quite good resolution for most 35mm images that you might take
through a scope or of a pinned specimen (you didn't say _what_ you will
be photographing --glass slide mounts? dissecting scope preps, pinned
specimens? big specimens? tiny specimens? ...).

The advantages of the Kodak PhotoCd over alternative photo digitizers
is convenience, compact storage of high-resolution images (e.g., you can
fit more images [not PhotoCD] on a cdrom, but not at the same
high resolution; lossy compression of images is not a fair choice for
archival storage of the digital images), and low cost.

The amount of time you will spend digitizing your own images in your
lab is simply not worth the cost of your time (unless you truly don't
value your time).

You are fortunate to not have begun the project yet, so you can digitize
the photos when the film is freshly mounted and still relative clean.

The secret is to locate a processing lab that does a good consistent job.
(And, if they do a bad job, a reputable lab will make "adjustments" by
re-doing the job. That's another advantage of film; if you do a good
job of photographing, your photographs can always be digitized again.

If you are concerned about digitizing photos that you haven't looked at
first, then get them developed, inspect and select the good ones, and
have just those digitized. The cost, if you collect together a 100,
is still going to be around $1/each. But, in this case, you need to
make sure you keep the dust/fingerprints off of them, and orient them the
way you want them digitized (really doesn't matter, since you can always
rotate the image inside the computer).

With respect to color balance (another complaint of another reporter),
usually, for specimen preps, true color is not an issue (except on
pinned, untreated specimens perhaps). In addition, there is a lot of
color balancing you can do in the computer, sometime adding special
effects to highlight certain characteristics of interest.

The alternative to high quality PhotoCd is possibly a _high_ resolution
digital camera (1000-2000 pixels/inch) for which you need other computer
equipment in addition to the camera (that you wouldn't need with the
PhotoCD, or other type of film digitizer). I think you'll find the
initial cost outlay to be quite a committment compared to film (if
you already have a good camera, essentially your cost is the film,
microscope adapter(?), and photo lab digitizing fees. Even if you don't
have a camera, it'll be a cheaper hardware outlay than what you need
for an "equivalent" digital camera.

In addition, regardless of which technique you use, you need good
image manipulating software, and enough computer horsepower to manipulate
the images to your satisfaction in this century.

Try the PhotoCD. If you don't like it, you've invested about as little
as you could and still learn something about your specific needs.

Is "USF" San Francisco or South Florida, or ??? If San Francisco, you
are in luck. There are good PhotoCd processors in the Bay Area.
Peter