“The biggest advance came when we began looking into coherent
light, and many of the problems just melted away,” Leith recalled.
an unclassified project, making Gabor-style holograms. Despite
Upatnieks’ lack of optics experience, he succeeded. The reconstructed images were fascinating, but they suffered from the
same twin-image problem as Gabor’s originals.
Leith had developed a theory of holography based on the
concept of a signal modulating a carrier frequency that is a
basis for radio communications. The coherent input light that
did not illuminate the object was the carrier, and the light
scattered by the object was the signal. In a simple modulation system, combining a low-frequency signal with a higher-frequency carrier produces two side bands, the sum and
difference frequencies, one above the carrier frequency and
one below. In holography, the twin images were the two side
bands produced by in-line Gabor holography.
The challenge in holography was to eliminate one of the
side bands, leaving a single clear reconstructed image. Leith
thought he could do that by separating the object and reference beams, so they reached the photographic plate recording the hologram at different angles. But separating them in
practice was difficult, and Upatnieks’ experiments did not yield
reasonable holograms of the transparencies they were using
until he came up with an optical trick in the spring of 1961.
He directed a single line from a mercury-vapor lamp at a
diffraction grating and selected two different orders of diffracted light. One became the reference beam, aimed directly
at the plate recording the hologram. The other became the
object beam, aimed at the object so it would scatter light
toward the plate.
The trick worked, yielding the first off-axis hologram.
Further experiments showed that separating the two beams by
a large enough angle produced results that confirmed a refined
version of Leith’s theory. Leith described the results at the Optical Society’s October 1961 meeting in Los Angeles and submitted a paper to the Journal of the Optical Society of America.
By then the project was on hiatus. Upatnieks, who had
enlisted in the Reserve Office Training Corps (ROTC) in college, was called to fulfill his military obligations in May 1961.
He didn’t return to Willow Run until November 1962. As
soon as he did, he returned to holography.
The new round of experiments started with a mercury lamp,
but much had happened in the young world of lasers in the 18
months Upatnieks spent on duty. Bell Labs had developed the
red helium-neon laser, and Spectra-Physics and Perkin Elmer
had teamed to produce the first commercial version. The lasers
were just starting to reach labs, but another optical specialist
at Willow Run, Anthony VanderLugt, had a three-milliwatt
model he was using in image-recognition experiments. It was
a wonderful new toy, VanderLugt was working close by, and
Emmett Leith and
Courtesy of Juris Upatnieks
inevitably, as Upatnieks says, “We kind of talked him into
letting us borrow his beam. We put a mirror in his room, and
bounced the beam off to our setup.”
Their simple setup was based on a standard optical bench.
They expanded the laser beam and split it by passing it through
a wedge prism resting vertically on top of their object, a pho-
tographic transparency. The object-beam light went straight
through the transparency, but the prism bent the reference beam
down at 5 to 10 degrees, so the two overlapped, recording a
hologram on a photographic plate. Recording good holograms
required extra-flat glass photographic plates that Eastman
Kodak had developed for spectroscopy. Exposure was very slow,
making the higher intensity of the laser beam an important
advantage over the fainter light from the mercury lamp.
The results presented at the March 1963 OSA meeting in
Jacksonville, Fla., showed a dramatic improvement over the
earlier mercury-lamp holograms. In the version published in
the December 1963 issue of JOSA, it’s hard to tell an original
1.5-cm slide with clear lettering from a holographic reconstruction. Holographic reconstructions of slides of a child in an
outdoors scene and an adult portrait are speckled but clear.
Initially, switching to the laser had disadvantages as well
as advantages. The speckle inevitable with lasers is noise in
the reconstructed images. Mercury lamps could also record
holograms, but the higher power and longer coherence length