Measurement of military helmet and head-mounted display (HMD)

visor optical properties

Dean F. Kocian and H. Lee Task
Air Force Research Laboratory
2255 H Street, Room 300
Wright-Patterson AFB OH 45433-7022

ABSTRACT
This paper examines the light transmission, absorption, reflection, and scattering characteristics ofmilitary helmet visors
used for see-through helmet-mounted displays (HMDs). HMDs used for the within-visual-range (WVR) counter-air
mission normally use the inner surface ofthe helmet visor to reflect the HMD image to the pilot's eye. This approach is
popular because it minimizes any optical structures that interfere with the pilot's vision, while also maximizing see-through
to the ambient scene. In most cases, a reflective coating, which increases the cost ofthe helmet visor significantly, must be
applied to the inner surfaces in order to achieve enough contrast between the HMD image and the external light passing
through the visor. Recently, with the development ofhigh luminance miniature cathode-ray-tubes (CRTs), it has been
possible to eliminate the reflective coatings on neutral density helmet visors having a see-through range of 13-35%. This
paper examines the light management properties of both types of visors. The paper stresses measurement techniques that
produce repeatable results and what these results might imply about visual performance under operational lighting
conditions.

Keywords: helmet, visor, flight visor, helmet-mounted display

1. INTRODUCTION

With the advent of see-through helmet mounted displays in military aircraft cockpits, the normal visual environment for the
pilot has changed dramatically. Prior to their introduction, a pilot, particularly a fighter pilot would fly "bare" eyeball.
That is, he would fly with his helmet visor stowed out of his line-of-sight to gain maximum visual acuity for seeing hostile
air or ground targets at the maximum distance possible. The introduction of high off-boresight angle weapons that are cued
by the pilot to a desired line-of-sight using cueing information imaged onto the flight visor forces the pilot to have the visor
in front of his eyes. This situation is known to reduce the maximum distance at which he can "see" or achieve a "tally-ho"
on a given target. The added use of reflective coatings that maintain the needed contrast for viewing helmet mounted
display symbology can further reduce the pilot's ability to discern targets that are threats to his survival, particularly if the
contrast between the target and its background is low to begin with. A reflective coatings will not only transmit "wanted"
light from the helmet display to the pilot's eye, but "unwanted" light as well, such as light reflected from the pilot's face to
the reflective coating and back to the eye.

To minimize weight and maximize safety the helmet visor is made ofplastic. The use ofplastic, usually polycarbonate, in
military helmet visors has also resulted in the addition of several optical degradation effects not normally associated with
glass optical transparencies. Among these are rainbowing (birefringence), multiple imaging, distortion, and haze. The
purpose ofthis paper is to explore the effects ofvisor transparency reflective coatings on vision using the basic conceptual
approach developed for haze by Task and Genco' as a starting point.

2. TEST METHOD BACKGROUND

The effect of haze in a transparency is primarily a reduction in the contrast of objects viewed through the transparency due
to the veiling luminance effect ofthe haze (Task & Genco, 1985). Figure 1 shows how an illuminating light source (Es) is
scattered into an observer's line ofvision resulting in an apparent veiling luminance (LH) due to the haze ofthe transparency
(visor, in this case) and defines the quantities of interest. The quantitative effect on contrast is determined by equation 1:

In Helmet- and Head-Mounted Displays V, Ronald J. Lewandowski, Loran A. Haworth,

120 Henry J. Girolamo, Editors, Proceedings of SPIE Vol. 4021 (2000) e 0277-786X/0O/$15.O0

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 C.
CIi = (I)
I 2— (L
(L, L11)T

Where contrast is defiiied as:

-
—
L- L13 (2)
L.+ L1

Helmet
Visor

Target
Es
LHt
LB

Blue Sky
Background

T = visor transmission (uncoated) L11 = veiling luminance Uue to naze

Li- = target luminance LB = bacLground turn inance
Es = illuminance at visor from sun
Figure 1
Schematic for haze relationships

The procedures outlined in ASTM Standard lest Method 1)1003 were used to measure haze values listed in this paper
(which, unfortunately, cannot be directly related to the veiling luminance effects of haze delined in equation 1, but should in
general correlate with these effects). ['he haze measurenients were performed to determine whether haze was different for
the different visor types. which might indicate the problem with the coated visors wasn't really the reflective aspect of the
coatings but rather haze that covaried with the coatings (the data indicated this was not the case: haze was about the same
for all new, unused visors). In equation I it should he rioted that the visor transmission term does not cancel out of the
equation.

Using the same conceptual approach. it is possible to derive an equation which describes the veiling luniinance effects of
reflected light from the visor on the contrast of outside scene objects perceived through the visor. Figure 2 depicts these
retatioiiships and provides a definition of the quantities of interest for developing the empirical relationships shown in
equations three through ten..

121

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 Helmet
Visor

Target
LT Es

LB

Blue Sky
Background

T = visor transmission (uncoated) T\(• = visor transmission (with coating)

LT = target luminance LB = background luminance
= visor absorption = visor reflection coefficient with coating
= visor reflection coefficient with no coatino = face diffuse reflectance coefficient
E5 = iliuminance at visor from sun EF = illuniinance at face from sun
Figure 2
Schematic defining relationships of primary variables for visor contrast calculations

For the see-through visor the amount of light reaching the eye is a combination of the transmitted light froni the background
and targets of interest (that form an image on the retina) arid scattered or reflected light that has no useful image forming
information. The extraneous light can come from either haze as defined in equation I or from light reflected from the
diffuse surface of the skin that finds its way back to the reflective surfaces of the visor arid into the eye (which causes a
contrast loss and visual performance degradation). It is for this second source of extraneous light, hereafter referred to as
veiling luminance from the visor reflection (LR), that we now modify equation 2 to account fur its effect. It is easily seen
(see figure 2) that 1R is derived from the terms shown in equation 3. The terms in equation 2 must he modified for the
effects of visor transmission and the additive veiling luminance term as shown in equations 4 and 5. For an uncoated visor
the following relationships are derived.

LR = E5 * T * Rr * = Veiling luminance from visor reflection (3)

Where: E5 * T = E1

L1 = ( L1 * T ) + LB = Target luminance (4)

LBV = ( LB * T ) + LR = Background luminance (5)

Substituting the relationships for equations 4 and 5 into equation 2 gives the result shown in equation 6

c= Lrv - LBV
(6)
L+

Ill
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx
 Substituting for LR a defined in equation 3 into equation 6 and arranging terms produces equation 7.

_— Tv*jLT-LB
CU T * (LT + LB) + 2(E * T * RF * R) (7)

Collecting terms and multiplying the numerator and denominator by the quantity 1/(L1 + LB) produces equation 8.

LT LB I
(LT+
' LB1
C=
U (8)
2(Es*RF*R,)
( LT+ LB)

Recognizing that the numerator ofequation 8 is now just equation 2 above we have the final result in equation 9.

C= (9)
2(E*RF*iT
(LT+ LB)

For the case ofthe coated visor the derivation follows the same process and steps and results in a similar relationship for Cc
except that Rv in the denominator has been replaced by Rvc as shown in equation 10.

Cc= (10)
2(Es*RF*R)
(LT + LB)

Itis worth noting that the coefficient for visor transmission drops out of both equations 9 and 1 0. Thus, the effects of lower
visor transmission coefficients must be attributed to either haze or the loss of visual acuity at lower light levels. Using the
relationships derived in equations 9 and 10 and some typical values for key variables, it is possible to get a quantitative feel
for the role that helmet visor reflective properties may play in reducing overall contrast for targets located in the ambient
scene. Some typical values are listed below with results shown in table I.

T = 25% or 35% visor transmission (uncoated)

Tyc 21% visor transmission (with coating)
LT 300 ft-Lamberts (target, in sky) I 100 ft-Lamberts (target in shadow of terrain)
LB 600 ft-Lamberts (blue sky) I 160 ft-Lamberts ground shadowed from sun
Ryc 6%, 9%, 13% visor reflection coefficient (with coating)
R visor reflection coefficient with no coating (3.5%)
RF face diffuse reflectance coefficient
Es 10,000 ft-candles illuminance at visor from sun/3850 ft-candles on terrain surface with 33° Solar inclination
angle

Table 1 tabulates results for two general viewing conditions where the pilot sees the airborne target silhouetted against either
a sky or ground background. The ambient illuminations agree reasonably well with measured data. For the airborne target
against a ground background, the illumination level represents a sun angle of-30 degrees to the horizon in Dayton, Ohio on
28 Feb 00. Results are tabulated for each condition for both an uncoated and coated visor and then the reduction in contrast
is compared on a percentage basis. It is interesting to note a change in contrast is obtained as the surface reflectivity
increases, and the percentage loss does not match the percentage increase in the surface reflectivity.

123

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 Table 1
Comparison of representative results for airborne targets set against a sky and ground background
Airbome Target Against Sky Background Airborne Target Against Ground Background
Input Value Units Contrast % Loss of % Loss of Input Value Units Contrast % Loss of % Loss of
Condition Contrast Contrast Condition Contrast Contrast
Lt= 300 ft-L Lt= 100 ft-L
Lb= 600 ft-L Lb= 160 ft-L
Es= 10000 ft-c Es= 3850 ft-c
Rf= 0.3 no Rf= 0.3 no
units units
Rv= 0.035 no Rv= 0.035 no
units units
Rvc= 0.13 no Rvc= 0.13 no
units
Result Result
C= 0.33 no Baseline Baseline C= 0.23 no Baseline Baseline
units units
Cu= 0.27 no Uncoated 18.9 Baseline Cu= 0.18 '° Uncoated 23.7 Baseline
units visor units visor
Cc= 0.19 no Coated 46.4 33.9 Cc= 0.11 no Coated 53.6 39.2
units visor units visor

3. TEST SETUP AND PROCEDURE

In order to substantiate that the actual visor surface reflectivities for both standard uncoated and coated flight visors were
close to manufacturer claims, two different test setups were used to obtain laboratory-based quantitative measurements. The
idea being that the use oftwo independent measurement techniques would help resolve errors in the measurement
procedures ifthey differed from each other or ifthey produced nearly the same results raise the confidence level that the
values obtained were reasonably accurate. The first method involved measurements ofthe visor reflectivity using an
extended light source and photometer. The second involved making spectral transmissivity measurements ofthe HMD
optical chain including the relay optics, visor, and visor coatings.

METHOD I : Minolta Spotmeter and 8' x 8' light box

The first measurement technique employed the use of an eight foot by eight foot light box on the reflective coating side of
the visor under test with a black low reflectance surface on the opposite side as shown in figure 3. The visor in figure 3 is
shown mounted to the visor frame and its associated removable optical-mechanical assembly (ROMA) with universal
connector that allows the entire assembly to be attached to an appropriately modified flight helmet. The ROMA/visor was
mounted to an optical rail and then leveled side to side and front to back. The ROMA/visor test asset was centered in front
ofthe 8' x 8' light box at a distance ofapproximately five feet. The light box was illuminated and allowed to electronically
settle for a period ofabout 30 minutes while the ROMA/visor was tilted back approximately 30 to 45degrees and then tilted
left to right to allow the full field ofthe light box to fill the left or right side ofthe visor. This repositioned the inside ofthe
visor approximately four feet from the light box. A small piece ofblack masking tape was placed on the light box as a
locating spot in the center ofthe field to be measured in the visor. The tape provided adequate contrast in the reflected
image to be seen clearly when making measurements. A partition was placed 48" directly behind the visor, parallel to the
visor and a flat black cloth placed over the partition for a consistent background. The Minolta Spotmeter was then focused
on the reflection in the visor and a reading was taken. The location ofthe area in relation to the black tape (Center) was
noted. Another reading was taken from a position between the visor and partition along the same axis as closely as
possible. The Minolta was focused on the light box and the same area that was reflected previously was taken as a direct
reading above, below, and beside the visor, whichever kept as close as possible the original axis ofmeasurement. The
spotmeter measurements were repeated for 5 trials for the left and right sides ofthe inside ofthe visor. The amount of
reflected light coming offthe low reflectance black cloth and back through the areas measured on the reflective surface was
also measured. The reading obtained from the reflective surface minus the reading obtained from the black cloth times the
visor transmittance was divided by the direct reading from the light box to obtain the surface reflectance value for each
reading. This procedure was then repeated for each visor tested.

124

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 nsmitted light
ming off black cloth

Diffuse light
from lightbox

I
Figure 3
Depiction of test setup for Minolta Spotmeter and !ightbox measurement method
I
METHOD 2: Visor HMD test using spectradiometer and dummy light source

The second method makes use of a spectradiometer to measure the amount of light being reflected off the visor to the eye
using a dummy light source (DLS) to provide the light stimulus instead ofa cathode-ray (CRT) as shown iii figure 4. To
determine the reflectivity of the visor surface it was first necessary to measure the light output from the I)LS and to also
determine the light losses in the relay lens. The method used to do this was to construct a helmet mounted display (HMI))
simulator whose light transmission characteristics were known and which could then he used to measure losses through the
VCATS relay lens assembly optics.

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 Figure 4
Depiction of test setup for visor reflectivity using spectradiorneter and durnrn light source

IJsing the procedures outlined in the operator's manual for the Radoma Spectradiometer. the calibration was verified for the
radiometer. A green HeNe laser (543nm). red HeNe laser (632.8nrn), and a laser diode (S7Orim) were individually scanned
off a 2' x 2" reflectance standard to confirm wavelength accuracy. E)ue to the intensity ol the laser eiiiissioil the exposure
was reduced to 0.03 Seconds, all the other parameters were the same. The aperture setting and spot sue measured were
found to be much smaller than the light images to be measured, which is a required condition for accurate measurements.
The HeNe laser was translated the length ofan optical table (48 inches) and a reference axis was established. [he laser was
translated and aligned over the length of this axis. The laser output was then blocked and the ROMA/visor system was
aligned perpendicular to the LASER AXIS. The HMD simulator was mounted to the ROMA/visor with D1,S imaging its
light through the VCA1S relay lens assembly and into the HMD simulator. [he mount was translated to center the I IUD
image on the laser axis as a point of reference. Then the laser was removed from the path and the spectradiometer was
directed dowii the reference axis and translated to center the measurement rcticle on the output of the [)1.S (see figure 5 --
frame A). The output was scanned from 517 nanometers to 574 nanometers and saved. It should he noted that the room
was darkened for all scans, which figure 5 does not show. The DLS was then connected to the HMD simulator assembly
and its output was measured across the same 517-574 nm spectral bandwidth(see figure 5 — frame B). The output was again
scanned and saved. The output from the DLS-HMD simulator combination was then divided h the output from the DLS to
find the percent losses across the measured spectrum for the HMD simulator optics. Ihe I)l,S and I IMI) simulator were
then connected to the VCATS ROMA/visor assembly and a scan of the light output coming froni the [-IMI) simulator was
taken (see figure 5 — frame C). With the data from this scan and knowing the light output of the DLS and losses in the
IIMD simulator assembly it was then possible to calculate the losses in the VCAIS relay optics assemhl. This
measurenient multiplied by the reciprocal of the loss computed for the HMD simulator optics and divided by the measured
output form the DLS gives the loss for the relay optics. Finally the IIMD simulator was removed from the VCAIS
ROMA/visor assembly and a radiometer scan was taken off the visor using the DLS as the stimulus (see figure 5 franie
D). This data when multiplied by the reciprocal of the light loss in the relay lens and divided by the known 1)1.5 output
gives the percent reflectivity for the visor.

.26

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 DLS mounted in . .. S relay optics wiin HMD
Srnu'.!or rnour!ed on other end ready tor radorneter scar

Figure 5
Step-by-step device setups for radiometric measurements of visor reflectivity

4. RESULTS

Figure 6 shows the appearance of the JHMCS clear and yellow (shooter's) visors as they were received Rr test andthe
mounting apparatus used to test their reflection, transmission and haze properties. t)SN personnel who performed tests prior
to their receipt by AFRL, applied tape markers shown on the clear JHMCS visor (figures 6 and8). Figure 7 depicts the
approximate appearance of the reflected light from thevisors after they had been properly aligned. These photos clearly
show the portion of the visors that have been coated.

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 Figure 6
Appearance of JHMCS visors viewed with the 8' by 8'lightbox in the background and
showing extent/position of highest density reflective coatings as received for test

Figure 7
Characteristic reflected image of 8' by 8' light box in JHMCS coated visor

Figure 8 depicts the haze values obtained for the JHMCS clear and yellow visors, which ran between 0.4% and 0.9%. The
data show normal results for polycarbonate visors that are almost new and have experienced little human handling or use.
The haze values also do not show any trends upward or otherwise with respect to whether the area nieasured had a
significant change in its reflective properties due to the coatings that had been applied. This is as it should he since haze is
primarily a result of surface or volumetric imperfections from stresses or exterior scratches, which were not visually
apparent with these sample visors. Figure 9 depicts the locations from which reflectivity measurements were taken using
method l.only.

28

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 CLEAR YELLOW

LEFT RIGHT
Tape delineating edges
of coating
Figure 8
Haze values (%) and measurement locations for the JHMCS visors

YELLOW

LEFT RIGHT
Tape delineating edges
of coating
Figure 9
Reflectivity measurement locations for the JHMCS visors

Table 2 shows the results obtained in percent for the reflectivity of the Jl-IMCS visors. Measured values were somewhat
higher than the nominal l2% to l4°/ range claimed by the manufacturer. The measurement fir location 2 on theclear
visor matches closely with what was obtained for other uncoated polycarbonate visors. I lowever. location 3 for the yellow
visor is significantly higher than would be expected and may infer that a lesser amount of the material used to form the
reflective coatings during the deposition process "bled" into this area. For the yellow visor reflectivity values appeared
remarkably similar at all three locations measured, which included those ftr the MMD (2), the right upper aiming reticle
area (4). and the left upper aiming reticle area (I).

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 Table 2
Reflectivity measurements for the JHMCS clear and yellow coated visors
Numbered to correspond to test points shown in figure 9

Visor Type Reflectance (%)

JIHMCS Clear Visor Position Right Eye Left Eye
1 15.0
2 4.2
3
JHMCS Yellow Visor Position Right Eye Left Eye
1 19.5
2 19.6
3 6.5
4 19.9

The VCATS 25% transmission neutral density gray visor with 6.5% reflective coating and the 25%and 35% neutral density
gray uncoated visors were mounted in similar apparatus and tested using the same procedures. It should be noted that the
25% and 35% nomenclature refers in a general manner to the transmission characteristics claimed by the manufacturer.
Figure 1 0 shows the approximate areas from which haze and reflectivity measurements were taken using test method I.
Test method 2 scanned one portion ofthe right side ofthe visor located in the upper potion ofcircular area 3 where the
HMD image falls.

from light box

Figure 10
Haze and reflectivity measurement locations for the VCATS 6.5% reflective(left), and 25%35% uncoated visors

Table 3 shows the results for haze and surface reflectivity using methods 1 and 2 for all three visor types. The haze
measurements were higher for the visor with the 6.5% reflective coating than the 25% and 35% uncoated visors. This can
be attributed to the fact that the coated visor had been used in approximately 15 test sorties in an F-15C over a period of
approximately 6 months while the other two visors were brand new. Fine scratches and abrasions could be observed on this
visor. The reflectance values obtained using method I correlated quite closely with those claimed by the coating vendor
and as would be expected for uncoated polycarbonate visors. The reflectance values obtained using method 2 were
consistently lower. This is probably due to the use of a spectrally selective filter in the HMD optical train (for the P53
phosphor) which restricted the spectral range for the radiometer measurements compared method 1 (see
Discussion/Conclusions section).

130

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 Table 3
Haze, transmissivity and reflectivity measurements for the VCATS 6.5% reflective and 25%/35% uncoated visors
Numbered to corresnond
& to test—points shown in figure 10 —
Visor Haze % Reflectance % Reflectance %
Light Box Method Spectradiometer
Method (averaged
over 517-574 nmL
25% Transmission Position Right Eye Left Eye Right Eye Left Eye Right Eye Left Eye
6.5% Reflective Upper
Coating Visor Portion 3
1 1.05 0.73 6.9 6.8
2 0.71 0.72 6.9 6.7
3 1.06 0.73 6.8 6.9 4.56
4 0.88 0.91 7.3 7.1
5 0.71 0.90 7.4 7.0
25% Transmission Position Right Eye Left Eye Right Eye Left Eye Right Eye Left Eye
Uncoated Visor
1 0.48 0.49 3.9 3.7
2 0.65 0.48 3.7 3.9
3 0.49 0.49 3.8 3.8 2.69
4 0.49 0.49 3.6 3.7
5 0.49 0.48 3.7 3.7
35% Transmission Position Right Eye Left Eye Right Eye Left Eye Right Eye Left Eye
Uncoated Visor
1 0.42 0.41 3.9 4.0
2 0.57 0.54 3.8 3.8
3 0.28 0.27 3.6 3.8 2.70
4 0.29 0.27 3.7 3.9
5 1.01 0.27 3.7 3.7

5. DISCUSSION/CONCLUSIONS

The theoretical analysis of contrast loss due to visor reflections developed in the first part ofthis paper (equations 9 and 10
and Table 1) provides some insight into why visor reflectivity is an important issue. However, this analysis understates the
problem by assuming the only effect is a loss of contrast from the veiling (reflection-induced) luminance. In reality, the
visor reflections are not uniform veiling luminances but have some structure since they are reflections from different facial
features. This structure may further degrade vision by serving as a masking pattern, a distraction, and/or an accommodative
trapping mechanism. While these other vision degradation issues are beyond the scope of-this paper, it should be noted that
they do exist and will induce a greater visual performance loss than would be predicted by simple contrast loss alone.

Table I provides two examples ofcontrast loss calculations for two (hypothetical but typical) sets of viewing conditions. In
each case the uncoated visor (3.5% reflection in the calculations) by itselfproduces a certain amount ofunwanted contrast
loss (18.9% and 23.7% contrast loss for the two examples) but these are dwarfed by the contrast loss from a coated (13%
reflection) visor example (46.4% and 53.6%). This provides some insight into why pilots, using uncoated, non-HMD visors
push their visors out ofthe way to improve their "talley-ho" target acquisition distance (keep in mind this analysis does
NOT include the structure effects discussed earlier or further contrast loss due to haze). Note that ifthe 19+% reflection
coefficient measured for the JHMCS yellow visor were inserted in equation I 0 the reduction in contrast would have been
significantly higher.

Measuring the reflection coefficient ofa curved surface is not a simple matter. Coatings, in particular, may exhibit different
reflection properties depending on the incident and reflected angle ofthe light and the spectral content ofthe light. In
addition, with a highly curved shape like the visors tested, the reflected image could be either real or virtual depending on
which side (concave or convex) ofthe surface is measured and how close the light source and detector are to the surface.

131

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

 With all ofthese issues in mind, two measurement procedures were used as described earlier in this paper. The first method
used an approximate 2980K color temperature light box (with flourescent lamps) as the light source and a photometer as the
detector. This provided a simple, single number for the reflection coefficient that was weighted for the human eye response
(by the photometer) and represents the visor reflectivity for a broad-band, white light source (the 2980K light box).
However, using this method it was difficult to accurately determine the incidence angle ofthe light and no information was
obtained regarding reflectivity as a function ofwavelength. It should further be noted that the reflection coefficient included
both the front surface and the back surface ofthe visor (although the second surface reflection contribution to the
measurement should be small for the tinted visors due to the double pass through the light absorbing dye in the visor). In
order to determine ifthere was any effect due to the real versus virtual image measurement for the geometry used, the two
uncoated, tinted visors (25% and 35%) were measured for reflectivity from both the convex and the concave sides of the
visor. These results were essentially identical indicating that for the specific distances and geometries used the
measurements from the concave side (which produces a real image ofthe light box) were valid.

The second measurement technique used a spectral scanning radiometer to obtain reflection coefficients by wavelength and
used the HMD optical system to control light incidence angle. However, the HMD optical train included a filter designed to
pass primarily the green peak ofthe P43 phosphor (at about 545 nanometers). Thus when a white light "dummy" CRT was
inserted into the optical system the light that came out was primarily green. This provided a good configuration to assess
the efficiency ofthe visor at relaying the HMD image to the eye but didn't necessarily provide good information as to the
unwanted reflections coming from solar illumination on the face which is, in turn, reflected back by the visor. Still, the
results indicated the reflection percentage was about the same as measured by the first method.

Figures 6 and 7 demonstrate in interesting aspect of coated visors. In Figure 6 the visors are set in front ofthe 8' by 8' light
box and the coated patches are barely visible; which makes sense since this photo depicts transmission effects and there is
very little difference in transmission coefficient between the coated portions and the uncoated portions. By contrast, Figure
7 was taken with a black cloth as a backdrop and the light box was set in a position so that it would reflect in the thevisors.
In this photo the coated portions ofthe visors reflect the image ofthe light box much stronger than the uncoated portions as
would be predicted by the measured reflection coefficients. Viewed in this geometry one obtains a better appreciation of the
probable impact ofthese coatings on vision caused by a veiling luminance reflection.

The bottom line for this effort is that the effects of visor coatings can be modeled and that, for typical conditions, veiling
luminance caused by reflections can have a significant effect on contrast ofobjects viewed through the visor.

6. REFERENCES
1. H. L. Task and L. V. Genco, The measurement of aircraft windscreen haze and its effect on visual performance, Air
Force Aerospace Medial Research Laboratory, AFAMR:-TR-85-016, 1985.

2. RadOMA System And Applications Software Preliminary Instruction Manual 1.04, September 11,1992, Copyright
1992, EG&G Gamma Scientific

132

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/20/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx