Direct digitizing

[Kenelm Philip]

	Having now made a number of photographs of pinned butterfly
specimens with a Sony Mavica FD-7 digital camera, I am prepared to
evaluate the capabilities of the camera for this application. I have
also sent copies of the raw images from the camera to a couple of other
lepidopterists, one of whom is actively engaged in a large-scale digit-
iazing project of his own. I believe they would agree with my comments.

	My conclusions are as follows:

1) Macro capability: The 10 to 1 zoom lens provided can be used to fill
the CCD with the image of any butterfly down to the size of a typical
Lycaenid.

2) Resolution: Filling the CCD with an image gives nearly 640x480 pixels
resolution in the resulting image. When the image is reduced to near life-
size you have approximately 300 dpi for an average-sized butterfly, and
up to 600 dpi for a small one.

In practice, if you are careful about focusing, the image will show con-
trasting scales as individual dots of color. In many cases the individual
scales in wing fringes are resolved. As far as the level of detail that
is shown, these images are certainly suitable for use in a field guide,
_provided_ that enough of this detail survives the printing process. I
hope to have information on that topic later.

3) Color: When using incandescent light, the images are, of course, defi-
cient in blue and over-sufficient in red. That situation is easily ad-
justed in many color-image editing programs (and I have recently heard
about a Photoshop plugin designed specifically for the task of correcting
for incandescent lighting color-shift). Once corrected for this shift,
the images are fairly good except for some degree of 'flatness'--a slight
loss of contrast which many digital cameras exhibit (and which can, at
least to some extent, be corrected in Photoshop). I have not yet checked
the full range of available species, but right now my guess is that these
images, when corrected in Photoshop, are good enough to be used for
species ID.

4) Major problem: When photographing small specimens (Lycaenids) the
specimen must be rather close to the camera, which makes it difficult
to illuminate the specimen properly. At present I am using very low-angle
lighting (nearly parallel to the wings) which is not optimum. I plan to
modify the lighting setup for such conditions to produce something closer
to 45-degree illumination (from both sides). For average and larger
specimens there is no lighting problem at all.

5) Major advantages: This is nearly interactive photography. You can see
the picture on your monitor within seconds after taking it--and if lighting,
focus, or orientation need to be changed you can do so instantly while
the specimen is still before the lens. There is only an infinitesimal cost
in taking more pictures--you can store about 1500 images on one $13 Zip
disk. The images, being JPEG files, are capable of being shipped to all
and sundry via e-mail (although I cannot yet do that here, so please don't
deluge me with requests).

	There has been a lot of talk lately about the (admitted) fact
that affordable digital photography is inherently inferior to film photo-
graphy, especially in resolution. This is true--but in my opinion the
quality of these images is now good _enough_ for many applications, even
if not up to film. Any color image you see in a book, for example, has been
halftoned at a resolution far below the limit of color film--yet one
seldom hears complaints that books do not all use direct color prints!

	I'll have more to say, pro or con, after I have tried printing
these images.

							Ken Philip
fnkwp at aurora.alaska.edu