American Cinematographer - December 1970 - 18

r

CINEMA
By ANTON WILSON
FLUORESCENT LIGHTING
The principle of an incandescent
light source is relatively simple. An
element is heated until it glows or
incandesces. From previous discussions
it should be clear that the light emitted
from an incandescent source is "con¬
tinuous". This is reflected in FIGURE 1
where it can be seen that the energy
level for all colors is approximately
equal. This smooth even color response
makes the incandescent source an ideal
form of illumination for color cinema¬
tography.
Unfortunately one cannot always
film by incandescent illumination. With
"available light" cinematography be¬
coming increasingly popular, many cam¬
eramen find themselves facing the cruel
world of fluorescent lighting that exists
in almost every public, industrial and
office building. The fluorescent lamp
does not exhibit the smooth color re¬
sponse of the incandescent source due
to the fact that it operates on a totally
different and more complex principle
called fluorescence.
Before discussing practical methods
for filming by fluorescent illumination,
a short analysis of the fluorescent prin¬
ciple may prove beneficial. The fluores¬
cent tube contains a gas composed
essentially of mercury vapor through
which an electric current is passed via
electrodes at either end. This current
causes the electrons in the mercury
atoms to jump between several energy
levels. Each time an electron jumps to a
lower level, it releases a "photon" of
light energy whose color is determined
by the difference in energy between the

350 400 450 500 550 600 650 700 750
WAVELENGTH-MILLIMICRONS
INCANDESCENT 6000 KELVIN

1164

two levels. For a given atom, such as
mercury, there are only certain energy
levels that the electrons can occupy and
therefore there are only certain specific
colors that will be emitted. (This princi¬
ple is most familiar in a neon sign where
it happens that the only energy level
available in the neon atom corresponds
to pure red-orange light.) The mercury
vapor will emit essentially 5 distinct
wavelengths or colors of light which are
represented in FIGURE 2 by the five
vertical "spikes" at approximately 363,
405, 436, 546 and 578 millimicrons
respectively.
It would be impossible to film with
this limited spectrum, just as it is
impossible to film by illumination from
a neon sign. Unlike a neon sign, how¬
ever, the fluorescent lamp has a phosphore coating on its inner surface. When
the lines ("spikes") of mercury vapor
radiation strike the phosphore coating,
the identical principle occurs again, that
is, the electrons in the phosphore begin
to jump to higher energy levels and
when they jump down, a photon of
light is emitted whose color corresponds
to the loss of energy due to the down¬
ward jump. The only difference is that
instead of 4 or 5 distinct "spikes" of
color, as was the case with the mercury
vapor, the phosphore is an exact blend
of chemicals with a multitude of avail¬
able energy levels and thus the light it
emits consists of a myriad of colors that
approximates a smooth spectral curve.
But not quite. For one thing, an appre¬
ciable portion of the light still emanates
from the mercury vapor spikes. Second¬
ly, even though the phosphores are
blended very carefully, they cannot
duplicate the smooth, even response of
the incandescent source. (Compare FIG¬
URE 1 and FIGURE 2) As a result the
fluorescent source exhibits significant
inconsistencies in its color response
which make it generally unsuitable for
color photography. Furthermore these
inconsistencies are random and vary
from one type of phosphore to another.
Thus a "deluxe warm white" bulb may
cause a brown tint while a "cool white"
may turn everything green-blue on the
film. The term "color temperature" is

absolutely meaningless when discussing
fluorescent tubes.
There are several ways to circumvent
the fluorescent problems. Most obvious
is to turn them off. If it is possible (and
practical) to illuminate the area with
tungsten halogen, simply turn off all
fluorescent illumination in the area. In
many cases this is not practical, as with
large factories or office areas.
The second alternative is a filter
pack. By analyzing the curve of the
"soft white" lamp in FIGURE 2, it can
be seen that it suffers from a strong
peak in the red region, and a mild peak
with a large spike in the blue region. By
employing a cyan filter (minus red) and
a yellow filter (minus blue) the red and
blue peaks can be respectively reduced
sufficiently to produce a relatively
smooth response curve. Sylvania and
GE have analyzed just about every
fluorescent bulb in the above manner
and publish a fluorescent filter chart
which specifically recommends a given
combination of filters for each type
bulb and film stock.
This system has several serious draw¬
backs of which you must be aware. First
it requires the cameraman to carry a
complete set of cyan, magenta and
yellow filters. Secondly, both GE and
Sylvania stress the fact that their recom¬
mendations are just "starting points"
for individual experimentation since
bulbs change significantly with age.
Most cameramen cannot afford to run a
test prior to the actual shooting. In
addition many rooms are illuminated
with several different types of bulbs
Continued on Page 1223

350 400 450 500 550 600 650 700 750
WAVELENGTH-MILLIMICRONS
SOFT WHITE FLUORESCENT

AMERICAN CINEMATOGRAPHER, DECEMBER 1970

<
American Cinematographer - December 1970

Table of Contents for the Digital Edition of American Cinematographer - December 1970
