Visual Perception in Low-Light
Levels
Implications for Shooting Incidents
By Paul Michel, O.D.
While on evening patrol, officers discovered two men lurking
near a closed gas station in a high-crime area. In the
confrontation that followed, the officers fired on the suspects,
one of whom appeared to be holding a shotgun. The officers
believed that the other man had pulled a chrome-plated handgun
from his waistband. Later investigation revealed that the man
was, in fact, holding a beer can. He sued the officer who shot
him. During the trial, expert testimony centered on the nature
of human vision, the low level of light at the time of the
incident, and the results of a research study that demonstrated
the ability of healthy subjects to identify lethal versus
non-lethal items under a range of low levels of light, the type
of conditions officers often face when working at night.1 The
results of this study can assist investigators when determining
what an officer can identify under certain environmental
circumstances.
Procedure
This experiment used 12 police cadets as subjects. Prior to
admission to the academy, a general physician had prescreened
the cadets for corrected 20/20 distance visual acuity. Each
cadet was reexamined individually for corrected 20/20 vision and
measured for hidden refractive error—the cause of
nearsightedness, farsightedness, and astigmatism—by observing
how parallel beams of light reflect off the retina of the eye.
The examination detected no eye disease among the cadets. The
cadets were taken from their classroom, which was at a standard
office lighting level, and brought to the research room. A
research assistant wore a black jacket, consistent with clothing
often worn by crime suspects, and showed each cadet three
non-lethal objects and a large-frame handgun under each of four
incrementally increasing levels of low light. The black jacket
served as a background for the object, and the assistant stood
behind an opaque partition that was quickly shifted down for 1
second. The assistant did not point the object in the direction
of the cadet but held each object in a clenched fist close to
his body, similar to the physical circumstance of many shooting
incidents. Specifically, the non-lethal objects consisted of a
6-inch piece of green garden hose, an 8-inch piece of black
pipe, and a 6-inch chrome-plated screw driver.
According to police documents, officers had misidentified
similar objects as lethal during the past 10 years. Academy
regulations prescribed only the use of academy-deactivated
firearms in this study; therefore, a blue steel model 59 Smith &
Wesson handgun was chosen as the lethal object because it has a
large and distinctive shape. The experiment used several
lighting levels. These levels ranged from .04 foot-candles to
.45 foot-candles. For comparison, a bright, full moon on a clear
night exhibits illumination comparable to a .01 foot-candle
lighting level. A person standing 30 to 40 feet from the direct
beam of a vehicle’s headlights at night compares to a .45
footcandle lighting level. Each cadet viewed each object
individually for 1 second. After the presentation of the object,
the cadet’s attempt to identify the object was recorded.
Results
Each cadet viewed one lethal and three non-lethal objects at
each lighting level. Therefore, 48 responses were recorded at
each level. At .04 foot-candles, cadets correctly identified an
object only 4 times; they misidentified or said they could not
identify an object 44 times. This represented a 9 percent rate
of correct identification. At .10 foot-candles, cadets correctly
identified an object only 8 times; they misidentified or said
they could not identify an object 40 times. This represented an
18 percent rate of correct identification. At .25 footcandles,
cadets correctly identified an object only 15 times; they
misidentified or said they could not identify an object 33
times. This represented a 34 percent rate of correct
identification. At .45 footcandles, cadets correctly identified
an object 37 times; they misidentified or said they could not
identify an object 11 times. This represented an 84 percent rate
of correct identification.
Cadets most frequently identified the handgun correctly. At .25
foot-candles, 10 of the 12 cadets identified the handgun
correctly, but 2 cadets still incorrectly identified it or
stated they could not identify the object. The gun used in the
experiment is one of the largest handguns usually encountered on
the street. Had a smaller pocket handgun been used, a higher
rate of incorrect identification might have occurred in the
low-lighting levels due to the smaller size. Cadets most
frequently misidentified the 6-inch piece of green garden hose.
Even at the .45-foot-candle level, cadets most frequently
identified the hose as a gun. Only one cadet identified the hose
as a pipe or cylindrical object.
Findings
This study demonstrated the difficulty officers have
distinguishing between lethal and non-lethal objects under
low-lighting levels. Most of the cadets voiced uncertainty about
their responses, even when they were correct. When asked to make
a determination of the object in less than .25 foot-candles of
light, cadets most frequently responded “I cannot tell.” During
an exit interview, 80 percent stated they relied upon the
positioning of the assistant’s hands to help make their
determination of the objects. Yet, the assistant purposely had
held the object in a neutral manner.
Practical Applications
Officer-involved shootings rarely occur exactly as constructed
in this experiment. The time frame involved from when an officer
perceives danger to the time deadly force is considered
necessary is more realistically only a fraction of a second and
not the full second allotted in this experiment. Additionally,
the suspect or the officer is frequently in motion before and
during the shooting. This movement decreases visual ability. The
psychological and physiological effects of fear also decrease
the level of visual functioning. The human body instantly
undergoes profound physiological changes in response to
perceived threatening circumstances. Visual functioning
dramatically decreases in response to hormones secreted during
acute fear. Without sufficient lighting, the retina of the eye
cannot form an accurate image of the external environment. An
ambiguous image is created at the retina and transmitted to the
brain. The brain then integrates this ambiguity with cognitive,
memory, and emotional elements to form a perception. The officer
uses this perception to evaluate the suspect’s actions and to
react.
Based on the factors that affect officers’ visual perceptions
during confrontations in low-light levels, officers need at
least 2.5 to 3 foot-candles of light to accurately identify an
object. Shining a halogen flashlight on an object from a
distance of 20 feet would create the level necessary for proper
visual functioning. Furthermore, lighting conditions that
officers face prior to an incident may significantly affect
their ability to discern shapes and details in a darker
environment. During the 20 minutes prior to the incident, if an
officer is exposed to lighting higher than when the incident
occurs, a transient disability affects the officer’s vision. The
retina experiences chemical and neurological changes as lighting
levels change. After 40 minutes, a person’s eyes adapt to
low-level lighting conditions. If that lighting condition
changes to a higher level of light for even a fraction of a
second, the dark adaptation is lost. For example, if an officer
seated in a darkened patrol car uses the car’s interior light to
check an address, the lighting level immediately changes and the
officer loses the dark adaptation. An overwhelming majority of
officer-involved shooting incidents have demonstrated this
often-overlooked disability. A thorough history of the officer’s
activities prior to the incident will aid a complete
understanding of the visual environment under which the incident
occurred.
Recommendations
This research study demonstrates that police officers have
difficulty differentiating between lethal and non-lethal objects
illuminated by less than .50 foot-candles of light, the level
frequently encountered during routine police working conditions
at night. Officers should recognize this disadvantage and adapt
their procedures accordingly, by increasing lighting levels
using their car’s headlights and flashlights with halogen bulbs.
A weapon-mounted lighting device, attached to the muzzle of an
officer’s weapon, can increase visual ability in low-level
lighting conditions. This device is designed for short-barrel
defensive shotguns, uses a small halogen bulb, and has an on/off
switch. In addition, officers should train in low-level lighting
conditions.
Finally, during the investigation of an officer-involved
shooting, police administrators should document the lighting
level at the time of the incident with an illuminometer and
consult a vision expert with experience in this area.
Conclusion
The investigation of an officer-involved shooting is never an
easy undertaking. A number of variables complicate the process,
including the officer’s visual perceptions prior to and during
the incident. Prior shooting incidents and this research study
indicate that officers have difficulty differentiating between
lethal and non-lethal objects in low levels of light. By
understanding the nature of human vision and the implications of
this research, administrators and officers alike can prepare for
the inevitable encounter with the suspect in the dark alley.
Perhaps more important, by properly documenting shooting
incidents, conducting thorough investigations, and preparing
expert testimony on the nature of these encounters, law
enforcement agencies can avoid legal liability while reminding
the public of the dangers associated with protecting the
community.
Endnote
The author of this article conducted the study and testified in
court.
Dr. Michel, a board-certified therapeutic
optometrist, serves as a specialist reserve police officer for
the Los Angeles, California, Police Department’s
officer-involved shooting investigations unit.
©2004 The Police Policy Studies Council. All rights reserved.
|