I was amazed by the scanning light photomacrography
images Jim Gerakaris showed in his Inter/Micro-84 presentation a “Second
Look at Scanning Light Photomacrography”. My own talk on “Method for Calculating
Relative Apertures for Optimizing Diffraction- Limited Depth of Field
in Photomacrography” complemented Jim’s talk by demonstrating the limitations
of conventional photomacrography even with the aperture diaphragm stopped-down
so much that the loss in image detail started to become visible.1  
Jim’s talk was my first exposure to the apparently
unlimited depth of field with the Dynaphot scanning light photomacrography
system.2  My article on
photography of fractured parts and fracture surfaces in Volume 12 of the
ASM Handbook includes a description of the Dynaphot and comparison fractographs
taken with and without scanning.3  This article also includes a table and graph based upon my optimum
aperture analysis. The front cover of the Dynaphot sales brochure has
a spectacular scanning light photomacrograph by Darwin Dale of the head
of a fly, shown in Figure 1. The Dynaphot and its operating principles
are well explained on the second page of the brochure, shown in Figures
2 & 3. There is some confusion caused by the brochure’s diagram in
Figure 3, which claims that the specimen is scanned vertically through
a thin sheet of light. The patent drawing by the inventor of the method,
Dan McLachman Jr, clearly shows in Figure 4 that the beam of light from
the slit lamps converges to a minimum thickness where it intersects the
optical axis of the camera objective. 



Figure 1


Figure 2

Figure
3

Figure
4

 

The effect of the illuminating beam converging and
diverging at the edges of the field being recorded and the diffraction
limited minimum beam thickness at the center of the field was the subject
for my later publication of a mathematical analysis of the scanning method
including a photograph and description of a scanning illumination system
of my own design to be used on my very robust and precise photomacrography
stand.4  This stand is
intended for use in photographing parts and fracture surfaces of significant
weight and is based upon restored components of an early twentieth century
bench-lathe. The stand with the completed scanning light system is shown
in Figure 5. A microscope eyepiece adapter with a 10X high eyepoint eyepiece
is shown mounted in place of a parfocal Olympus 35 mm camera back. This
eyepiece adapter was initially intended as an aid in viewing the fine
details and as a relay lens for an eyepiece-mounted Nikon Coolpix 995
digital camera to be used for test exposures prior to final recording
on film. I was pleasantly surprised to find that the test exposures with
the digital camera were of high quality, even with the shutter held open
on B (bulb) setting for 12 second scans. Figure 6 is a scanning light
photomacrograph of the head of a house fly taken with this now completed
system. The 35 mm film camera would record a much larger field area at
somewhat higher resolution than the digital camera.

 


Figure
5

Figure
6

My photomacrography stand initially used the Olympus
Auto Bellows, except for the bellows rail. The gray cast iron, double
dovetail sided bellows rail for my stand is somewhat longer than the Olympus
aluminum rail and held at both ends in sliders. These sliders mate with
the guiding surfaces of the lathe bed vertical column for coarse adjustment
of the bellows rail position using the long feed screw shown in Figure
7. These sliders are locked to the lathe bed before the adjustment for
final focus is made with the micrometer head at the end of the bellows
rail. The bellows rail dovetail and mating surfaces of the sliders were
hand scraped for a very precise fit and alignment with the lathe bed axis. The
all-metal camera and lens mounting boards shown in Figure 7, replaced
the earlier Olympus components to provide much higher rigidity, and to
eliminate the crack prone plastic inserts mating with the bellows rail
in the Olympus system. The X-Y feed slides from the lathe are shown attached
to a jackscrew driven knee slide in Figure 8. This slide provides the
vertical feed for scanning when motor driven. The cast iron slide for
the knee, made in my home shop, was precision hand scraped and fitted
with a tapered gib for maximum rigidity. The dial indicator shown in Figure
8 is used to determine when to open and close the camera shutter.  In
order to assure the slide is moving at a uniform speed upward, I allow
an initial 0.100” of scan travel before reaching the position where the
shutter is opened with a cable release.  An adjustable micro switch shuts
off the scanning drive motor if it is not first switched off based upon
the dial indicator reading for the end of the scan. The jackscrew, 0.025”
elevation per screw revolution, is driven through a 20:1 gear reduction
box salvaged from an electric drill. A 0.1 HP AC-DC motor with a belt
reduction is used to drive the gear box through a flexible shaft. The
motor speed is governed by a variable speed controller. The scanning exposure
is controlled by suitable combinations of scanning speed and slit illumination
intensity. 

 


Figure
7

Figure
8

 

The illumination system described in my earlier article
is shown in Figure 9 configured to scan the head of a fly, shown in Figure
6. Figure 10 is a close-up showing the fly between the illuminating lenses. The
blue cables at the sides of the Figure 9 are portions of a bifurcated
fiber-optic light-guide connected to adjustable sliders containing the
slits.  The ends of the light-guides are linear fiber arrays measuring
0.50 mm x 14 mm. These light-guide ends are positioned 6 mm behind the
10 mm long by 0.025 mm slit openings formed between two razor blade segments.
The inner two sliders contain Spiratone macro lenses with a 35 mm focal
length. Color balancing 80A filters are mounted in caps on the outer ends
of these lenses. The system is configured to produce a 5 mm wide beam
at 0.5X magnification of the slit sources to illuminate the fly head for
the image in Figure 6 obtained with the Olympus 38 mm focal length macro
lens set at f/4 with the bellows length set for 5X magnification. The
illuminating lenses are set at f/4 giving an illumination NA of 80% of
the imaging numerical aperture (NA). The numerical apertures are calculated
from the following equation using magnification Mi = 2X for
the cone of light illuminating the specimen. The lens relative aperture
setting (f/no) is the focal length of the lens divided by twice the lens
opening diameter.

 

 


Figure
9

Figure
10

 

Eyepiece inspection of the portions of the field
illuminated by the slit system revealed objectionable diffraction artifacts
when the illumination NA was reduced much below that of the imaging NA. This
same condition applies to brightfield illumination with the light microscope. 

 

The sliders containing the slit sources and the lenses
mount on a 610 mm long dovetail slide obtained from Edmund Industrial
Optics. These sliders incorporate vertical dovetail mounts permitting
the lenses and slits to be adjusted vertically for beam alignment. The
aligning operation begins by rotating the caps containing the slits on
the mating tubes containing the fiber-optic linear arrays until an image
of the slit formed by the adjacent macro lens exhibits a full length image
of the slit with uniform brightness. These angular positions are then
locked with the thumb screws in the caps.   The next operation is to align
the slits so they are the same distance above the stage and parallel to
the stage. A right angle monocular microscope with a 5X objective and
15X graticule eyepiece was fabricated for this operation shown being done
in Figure 11. The next operation is to align the illumination lenses so
that both beams are coaxial. The slit sources are moved to near the opposite
ends of the horizontal slide for this operation with one of the illumination
lenses removed from its mount. The remaining lens is then used to form
the image of the adjacent slit on the end of the cap containing the other
slit. The illuminating lens is then adjusted up or down until the image
of the illuminated slit is centered relative to the opposite slit. The
other illuminating lens is then installed for the final part of the aligning
operation shown in Figure 12.  The bellows lens with the 10X viewing eyepiece
is used to center and focus the slit images at one of the sharp edges
of an aluminum pyramid test target.The right angle microscope is also
used for this operation as an aid for establishing precise focus of both
slits on the pyramid edge. The illumination lens not previously aligned
vertically is then adjusted vertically so that both slit images are exactly
coincident on the edge of the pyramid when viewed with the right angle
microscope.

 


Figure
11

Figure
12

I expected the high magnification images to be the
most difficult for this system to achieve with uniform high resolution
and even illumination. My initial tests of the scanning system were with
a 45 degree inclined flat target covered with a patch of mm graph paper
and 5X magnification for the bellows lens set at f/4 for an intended 50X
final magnification with an NA of 0.1. This target would reveal uneven
illumination as well as variation in the resolution of the matted paper
fibers. The graph paper was oriented so that one set of lines was parallel
to the stage and facing one of the slit sources. A circle was drawn at
the center of the target with a graphite pencil and the system aligned
with the slit and camera lens both focused on the horizontal graph line
passing through the circle. Figure 13 shows the very narrow scan line
for this condition.  The stage was manually raised and lowered for Figures
14 and 15. The scan line, indicated between arrows, greatly broadens near
the edges of the field along the axis of the slit illumination lens so
that it falls just within the high resolution portion of the depth of
field. The system was covered with a light-proof cloth tent for the scanning
light image of this target shown in Figure 16. Note that the illumination
and paper fiber resolution are uniform across the entire field.  The graphite
coated circle becomes an ellipse with the graphite coating giving rise
to specular reflection of the scanning light beam. This test needs to
be repeated with recording on 35 mm film. The edges of the field in the
direction of the short axis of the ellipse (highest and lowest portions
of the field) would be expected to be blurred by the further broadening
of the illuminating beam thickness with the field width 1.5 times wider
on the film image.

 


Figure
13

Figure
14

Figure
15

Figure
16

 

The fly head was photographed with the digital camera
using broad area lighting from the side in addition to the stationary
ring of light from the two opposed slit light sources for the image in
Figure 17. This was done for comparison with a scanning light image of
the same field of view shown in Figure 18. Comparison of these images
demonstrates that the scanning method does not faithfully record the black
hair patterns. Another problem with the scanning image is the lack of
clues to judge depth of the features because the images are isometric
projections and lack out-of-focus regions. This missing information is
evident in the low magnification side view of the fly head shown in Figure
19. This conventional photomacrograph was recorded with the Nikon Coolpix
lens at maximum magnification and broad area lighting. Engineering drawings
typically contain front, top, and end views of a subject to aid in three
dimensional visualization. Jim Gerakaris showed that the best way of obtaining
the missing depth perception is to record stereo pairs with the scanning
light method. 

 


Figure
17

Figure
18

Figure
19

 

This article is the first progress report for my
now functional scanning light photomacrography system. A large capacity
eucentric stage has been built so that stereo pairs can be easily
recorded. This stage was described in my article Eucentric Stage for Recording Stereo Pair Photomicrographs, originally publshed in the September 2005 issue of Microscopy Today.
My original calculations of the field size limitations of the
scanning light method need to be revised now that I know that the NA of
the illumination beam must be significantly greater than anticipated. The
theoretical field size limits need to be verified by experiments using
35 mm film recording covering a wide range in magnification. These results
can be the subjects for future articles.