S Afr Optom 2007 66 (1) 3 − 11

A comparison of the effect of reduced illumina-

tion and tinted lenses on stereopsis at near

M Mehta * , P Ramkissoon ** and AM Bhagwanjee***

* Discipline of Optometry, University of KwaZulu-Natal, PO Box 54001, Durban 4000 South Africa

** PO Box 1097, Newcastle, 2940 South Africa

*** School of Psychology, University of KwaZulu-Natal, PO Box 54001, Durban, 4000 South Africa

* <mehtam@ukzn.ac.za>
** < pauleyes@mweb.co.za>
*** < bhagwanjeea@ukzn.ac.za>

Abstract used to measure stereoacuity. Participants (n =

Relative depth may be appreciated with the 60) between the ages of 17 - 29 years (mean =
use of one eye using linear perspective, shad- 23.58; sd = 3.14) were purposively sampled
ows, parallax and texture as monocular cues to from a clinical practice to participate in this re-
depth. Stereopsis, on the other hand, is the direct search study.
appreciation of relative depth that requires the Using repeated measures ANOVA and ap-
use of both eyes to construct a three-dimension- propriate post-hoc multivariate analysis, it was
al percept from disparate two-dimensional reti- evident that there was a significant decline in
nal images. The advantage of stereopsis is with stereopsis as the level of illumination decreased,
respect to complex visual tasks especially that regardless of tint condition; also there was no
requiring accurate hand-eye coordination. statistically significant difference in stereopsis
Tinted lenses are prescribed for a variety of between the no lens and white lens conditions
reasons, including but not limited to photopho- at each level of illumination; and stereopsis was
bia, asthenopia, improving colour perception in significantly superior with the no lens condition
colour deficient individuals, enhancing cosmesis compared to all six other tint conditions (grade
and protection against glare or harmful radiation B), at each level of illumination.
and enhancing visual performance as in sports. These results indicate that stereoacuity, as
The aim of this study was to investigate the com- measured by the TFS, is adversely affected by a
parative effects of six specific CR39 tinted spec- decline in retinal illuminance and by the use of
tacle lenses (grade B), and a white CR39 lens, tinted lenses. This information could be utilised
against a no lens condition, on stereoacuity over to advise patients on the performance implica-
a range of illumination levels. Illumination was tions of the six tinted lenses tested with respect
varied with the use of neutral density (ND) fil- to their effects on stereoacuity under different
ters, while the Titmus Fly Stereotest (TFS) was illumination levels.

*BOptom (UDW) CAS (NewEnCo) MOptom (UK)

**BOptom (UDW) CAS (NewEnCo) MPhil (RAU) DPhil (RAU)
***BA (UDW) BA Hons (UDW) MA (Clin. Psy) (UDW)

Received 15 September 2006; revised version accepted 14 December 2006

3 The South African Optometrist − March 2007

A comparison of the effect of reduced illumination and tinted lenses on stereopsis at near

Introduction that there are at least two independent stereopsis

Even though the retinal image is two-dimen- mechanisms, one sensitive to chromatic contrast
sional, the world one looks out on is three-di- and the other to luminance contrast10. Although
mensional1. The three-dimensional shape can the information from each mechanism may be
be determined by estimating relative depths. combined into a unified percept, in comparing
Depth perception can be appreciated using mon- chromatic stereopsis to luminance stereopsis,
ocular cues such as linear perspective, shadows, the former mechanism is less sensitive to con-
parallax and texture amongst others2 - 5 as well trast, has a more limited disparity range, poorer
as binocular cues such as stereopsis6. Stere- stereoacuity and poorer ability to encode ste-
opsis contributes to the judgment of depth and reoscopically defined shape than the luminance
distance while facilitating the recognition of stereopsis mechanism11. Empirical evidence on
solid objects, with localisation playing a valu- the effects of illumination on stereopsis indi-
able role in a variety of daily tasks2, 3. Although cates that stereoacuity decreases at low levels of
we have two eyes, we usually have only one vi- retinal illumination12 - 14. A significant decrease
sual world7. This is made possible by the use of in stereoacuity was noted when the value of the
egocentric (knowing the distance and position ND filter was 1.4 ND (4% transmission) with
of an object relative to our body) and oculocen- TFS12. Stereopsis had also been measured under
tric (quantification of the location of the object scotopic conditions and was found to be possi-
within the field of view) axes of localisation1, 2. ble in dark adaptation even though it decreased
This stereoscopic function is due to the frontal steeply15.
positioning of the eyes through the process of The clinical value of stereopsis testing is that
evolution in humans1 - 3. The horizontal separa- it is a good indicator of the overall functioning
tion of the two eyes results in relatively small of both the sensory and motor aspects of the vi-
retinal positional differences, giving rise to dif- sual system3, 5. The unit of measure of stereopsis
ferent principal visual directions for the same is seconds of arc and the stereoacuity is obtained
object2, 4, 6, 8. If this positional difference (retinal when the least horizontal disparity evokes the
disparity) is within Panum’s fusional space, it perception of depth2, 3. Elkington and Frank16
provides significant information about three-di- report that 60 seconds of arc is considered nor-
mensional scene structures, giving rise to stere- mal stereoacuity measured at 40 cm, though 15
opsis4, 6. While similar objects stimulating non- seconds of arc or better can be measured at 80
corresponding (disparate) retinal points within cm using the Frisby test. Good stereopsis at near
Panum’s fusional area may be fused to give rise is required for accurate hand-eye coordination
to single binocular vision, very dissimilar ob- when using tools, threading a needle, performing
jects cannot be fused, resulting in suppression, surgery or even using a computer17 - 19. Reduced
superimposition, binocular rivalry or diplopia6. stereopsis may cause symptoms of discomfort
Fusion is the sensory neural process whereby such as eyestrain, headaches and diplopia3, 20.
these two possibly disparate retinal images are Tests of stereopsis can be broadly divided into
associated to produce a single percept in the two categories: contour stereotests and random-
higher cortical centers, that is, binocular single dot stereotests6. Random-dot stereotests have no
vision6. However, if the horizontal disparity ex- monocular cues so that stereoscopic depth per-
ceeds two degrees and if there is any vertical ception can only occur when binocular fusion
disparity of over a few minutes of arc, it leads to has occurred. In this case, a process of global
diplopia1. Vergence is the eye movement most stereopsis is used as evaluation, with correlation
commonly associated with fusion1, 5. Fusion to- of corresponding retinal and disparate points oc-
gether with vergence eye movements has evolved curring over a large retinal area6. On the other
to support stereopsis8, 9. It has been reported hand, in the presence of monocular cues, local

The South African Optometrist − March 2007 4

M Mehta, P Ramkissoon and A M Bhagwanjee

stereopsis is used to evaluate horizontally dis- nisms11, it is important to investigate the effect of
parate images in contour stereotests such as the these tinted lenses on stereopsis over a range of illu-
Titmus Fly Stereotest (TFS)6, 21. Stereopsis may mination levels. Tinted lenses are commonly avail-
be measured at distance and at near. In a study able in a range of colours as either gradient or solid
by Wong et al where monocular cues were elimi- tints from grade A (light) to grade D (much darker).
nated and where they used the same tests at all Illumination levels may be consistently decreased
distances, they found no significant change in by using Gulden neutral density filters (ND) to al-
stereoacuity with viewing distance21. As this ste- low for measuring stereopsis ranging from room il-
reoacuity is a threshold value for the distance at lumination of 300 lux to measuring stereopsis under
which it is measured, stereopsis is not effective decreased light levels (<300 lux)24, 25.
beyond a certain critical distance3, 21. The aim of the study was to investigate the ef-
An important aspect of visual information re- fects of reduced illumination and tinted lenses on
lates to the perception of colour provided by the stereopsis at near. The specific objectives were as
three types of photoreceptors in the retina. Light follows:
is interpreted as colour according to the wave-
lengths that strike the retina18. When light reaches i. to compare stereoacuity values across five lev-
a surface, it undergoes three main changes. Some els of illumination (No ND filter, 0.3 ND, 0.6
of the light is reflected, some absorbed and the re- ND, 0.9 ND and 1.2 ND) without any lenses.
mainder transmitted18 - 20. A tinted lens possesses ii. to compare stereoacuity values using a white
a definite colour and acts as a filter that alters the lens against a no lens condition across the five
intensity and the spectral distribution of light that levels of illumination.
passes through it17 - 19. The colour of the tint is as iii. to compare stereoacuity values using six tinted
a result of chemicals added to the spectacle lens lenses at grade B against the no lens condition
to alter the transmission and absorption of the dif- across the five levels of illumination.
ferent wavelengths of visible light19, 23 - 25.
Optometrists frequently prescribe tinted lens- Methodology
es for their patients. Tinted lenses are prescribed, Selection of participants
inter alia, in the following circumstances: for the A purposive sampling procedure was used to
relief of photophobia26 - 28, to reduce asthenopic select sixty participants from the private practice
symptoms17, 26, to decrease light scatter in condi- of the clinician researcher (P Ramkissoon). This
tions such as albinism, retinitis pigmentosa (RP) purposive sample comprised a non-probability
and cataracts29, to provide protection for people cohort of participants who satisfied the inclusion
exposed to high levels of invisible radiation such criteria for this study. Adult participants ranging
as ultra-violet and infra-radiation19, to alleviate in age from 17 - 29 years were selected. Older
sensitivity to sunlight23, 24, to improve the cosme- adult participants were excluded in order to con-
sis in situations of disfigurements18, to enhance trol for age-related changes in vision. A compre-
visual ergonomics17, to reduce glare from reflec- hensive eye examination (including measures of
tive surfaces such as snow, sand and water19, 23, binocularity, that is, the cover test and fusional
as a placebo28, to improve colour perception in vergences) was used to screen all eligible par-
colour deficient individuals30 and to decrease ticipants to ensure that they satisfied the inclu-
light sensitivity in patients taking photosensitis- sion criteria for this study. Participants who had
ing drugs3. < 0.50 D of astigmatism and < 0.50 D sphere and
Since different colour tints are generally used monocular and binocular visual acuities of 6/6 or
by patients under varying levels of illumination, better at distance and near were included, thus
ranging from bright sunlight to very poor light excluding bias accruing from compromised visu-
levels and in keeping with the more recent infor- al status. Participants with eye diseases and those
mation regarding chromatic stereopsis mecha- who failed the Ishihara colour test were similarly

5 The South African Optometrist − March 2007

A comparison of the effect of reduced illumination and tinted lenses on stereopsis at near

excluded. Other exclusion criteria were poor favourably to two randot tests under compro-
general health status and intra-ocular pressure mised levels of binocularity in young partici-
above 21 mmHg. There was no evidence in the pants32. Provided that the page is inverted to re-
literature suggesting the need to control for race verse the disparity, the TFS controls adequately
and gender of participants. for the presence of monocular cues31. Further,
the TFS is a widely used clinical test that allows
Lenses used for rapid testing without significant subject fa-
A white CR 39 lens and six tinted lenses tigue, thereby yielding data that could inform
(CR39) were used on all participants in the clinical management decisions31. In the TFS,
study, namely, pink, blue, brown, grey, yellow horizontal disparity is presented via the vecto-
and green. The depth of the solid tints used in graphic technique33. This test comprises three
this study was grade B. All the lenses were of subtests, with the level of disparity decreasing
zero power (plano) with a 2 mm standard centre progressively33. When tested at 40 cm the first
thickness and equal transmittance, as measured subtest (the fly) has a disparity of 3600 sec of
on a spectrophotometer. Thus, all pertinent lens arc, the second subtest (the rows of animals) has
characteristics, including lens material, depth of a disparity ranging from 400 - 100 sec of arc and
tint, power, thickness and transmittance were the third subtest (the Wirt rings) has a disparity
controlled for in order to eliminate secondary ranging from 800 - 40 sec of arc6. The ultimate
variance. Stereopsis was assessed, for each measure of stereopsis is the participant’s ability
participant, using a no lens condition and seven to discern the finest possible level of disparity.
other lens conditions. Given these considerations, the TFS was con-
sidered the measure of choice for this sample.
Illumination levels
Neutral density (ND) filters have been wide- Procedure
ly used in photography, cinematography and Phase 1: Stereopsis measured under room illu-
vision research to assess visual function under mination of 300 lux with no ND filter
controlled levels of illumination24, 25. Stereopsis Stereopsis was measured initially without any
was initially assessed under bright room illumi- lenses (base line measure) followed by measure-
nation of 300 lux measured using a light meter, ments through the white lens and then through
with ND filters (0.3 ND, 0.6 ND, 0.9 ND, and each of the six tinted lenses, under normal room
1.2 ND) being subsequently used to decrease illumination, using the TFS. The order of pre-
the level of illumination. Thus, stereopsis was sentation of tint condition was randomised to
assessed, for each participant, at five different avoid treatment interaction effects. As recom-
levels of illumination. mended, this test was held at a test distance of
40 cm measured using a RAF rule. The partici-
Stereoacuity test used pants viewed the targets binocularly using a pair
Standard clinical tests for stereopsis include of polarising spectacle lenses.
the TNO random-dot stereotest that does not The first subtest, the fly, was shown to the
have monocular cues based on the presence of participants who were asked to pinch the edge
contours, as well as the Titmus Fly Stereotest of the wings of the fly. The normal response
(TFS) that utilizes real contours, resulting in was that the pinching fingers should be off the
the presence of monocular cues. While the use page by several centimetres.
of randot stimuli are generally considered the In the second subtest comprising the three
“gold standard” for measuring stereopsis, they rows of animals, the participants were informed
are prone to false negative errors31. The TFS that all except one of the figures in each row re-
on the other hand has been shown to compare mains flat. The participants were asked to in-
dicate the animal that appeared raised in each

The South African Optometrist − March 2007 6

M Mehta, P Ramkissoon and A M Bhagwanjee

row. of the study. The Bonferroni test was used for

In the third subtest (a series of nine diamonds post-hoc multivariate analysis. In this multiple
each of which contains four Wirt rings), one of the comparisons procedure, the familywise error (at
rings would be viewed at a different depth from α = 5%) was divided by the number of compari-
the other three. The participants were asked to sons, thus controlling for Type I error by setting
identify the ring that appeared to stand out from alpha at a more stringent level (the Bonferroni
the page. The results were recorded in seconds of correction).
arc, corresponding to the last correctly identified Table 1 demonstrates the progressive increase
circle/ animal. in mean stereoacuity scores (that is decline in ste-
As recommended by Garnham and Sloper31, reopsis), for the no lens and each of the seven
the page was inverted in order to reverse the dis- tint conditions, as the level of illuminance de-
parity, thus controlling for the effects of monoc- creased from room illumination (0 ND) towards
ular cues. Each participant was allowed a five- scotopic levels (1.2 ND). This pattern of results
minute rest interval between measures in order to is confirmed in the multivariate analysis (Table
minimise fatigue effects. While the order of pre- 2), where a statistically significant difference
sentation of the three stereo subtests was main- was evident for the main effect of illumination
tained from subtest one to subtest three, the order (F = 119.706; df = 4; p = .000). Inspection of the
of presentation within subtest two and three was means in Table 1 also reveals that, at each level
randomised to avoid errors of habituation and ex- of illumination, the stereoacuity scores for both
pectancy, thus precluding guessing. the no lens and white lens conditions were con-
sistently lower, that is better, than for each of the
Phase 2: Stereopsis measured under decreasing six other tint conditions.
levels of illumination Table 2 demonstrates a statistically significant
The second part of this study assessed stere- difference for the main effect of tint (F = 8.996;
opsis under decreasing levels of retinal illumi- df = 7; p = .000) as well as the interaction of il-
nation through the seven lens conditions. ND lumination by tint (F = 4.760; df = 28; p = .000).
filters were placed binocularly over the no lens The graphical illustration of these results (Figure
condition as well as with the seven tinted lenses 1), as well as the post-hoc analysis (Table 3), dem-
and stereopsis was measured as per the procedure onstrates specifically where these differences lie.
outlined in phase 1. Four ND filters (0.3, 0.6, 0.9 Table 3 shows no statistically significant dif-
and 1.2) were used to decrease the luminance lev- ference in stereoacuity scores between the no
els. A counterbalance design was used to control lens and the white lens condition for each of the
for possible confounding resulting from treatment five levels of illumination (df = 59; p = .321; p
interaction effects, where the order of presenta- = .616; p = .047; p = .048; p = .167). Statisti-
tion of luminance levels was randomised across cally significant differences were found between
the participants34. the no lens condition and each of the other six
tints (pink, blue, brown, grey, yellow and green)
Data analysis at every level of illumination, thereby accounting
Data was entered onto the Statistical Package for the statistically significant interaction effect
for Social Scientists (SPSS), and a file audit was reported in Table 2.
conducted to eliminate entry errors. Using gener- The main findings arising from the analysis of
al linear modelling, a repeated measures ANOVA the data are that:
was run, incorporating within-subjects factors at 1. there was a significant decline in stereopsis
two levels viz. tint and illumination. This analy- as the level of illumination decreased, regard-
sis rendered statistical comparison across tint by less of tint condition.
level of illumination, in line with the objectives 2. there was no statistically significant differ-

7 The South African Optometrist − March 2007

A comparison of the effect of reduced illumination and tinted lenses on stereopsis at near

Table 1: Modified Population Marginal Means for Illumination by Tint

Illumination Tint Mean Std. Error 95% Confidence Interval
Lower Bound Upper Bound
0 ND No lens 42.721 .906 40.890 44.553
White Lens 42.888 .965 40.938 44.839
Pink 46.667 1.227 44.187 49.146
Blue 46.982 1.223 44.509 49.455
Brown 45.881 1.381 43.090 48.672
Grey 47.552 1.539 44.441 50.664
Yellow 46.379 1.347 41.050 48.707
Green 46.137 1.790 42.519 49.755
0.3 ND No lens 67.339 2.108 63.079 71.600
White Lens 67.274 2.252 62.723 71.824
Pink 75.329 2.461 70.354 80.303
Blue 77.582 2.316 72.901 82.263
Brown 75.031 2.526 69.925 80.137
Grey 74.862 2.047 70.725 78.999
Yellow 72.862 3.246 66.302 79.422
Green 74.465 2.828 68.749 80.182
0.6 ND No lens 123.519 4.874 113.667 133.371
White Lens 125.057 5.455 114.033 136.081
Pink 172.845 9.266 154.119 191.572
Blue 151.543 10.818 129.679 173.407
Brown 154.767 9.160 136.254 173.280
Grey 155.624 11.913 131.546 179.702
Yellow 157.510 5.928 145.528 169.491
Green 180.712 9.303 161.911 199.513
0.9 ND No lens 189.205 13.519 161.883 216.527
White Lens 205.483 17.545 170.024 240.943
Pink 280.893 29.499 221.273 340.512
Blue 347.019 23.916 298.683 395.356
Brown 403.940 28.822 345.690 462.191
Grey 284.395 27.985 227.836 340.955
Yellow 267.898 24.464 218.454 317.341
Green 295.514 29.266 236.365 354.664
1.2 ND No lens 598.000 62.781 471.116 724.884
White Lens 610.333 63.776 481.436 739.230
Pink 860.119 98.672 660.695 1059.543
Blue 1002.095 141.615 715.880 1288.310
Brown 1157.262 131.507 891.477 1423.047
Grey 1022.619 153.342 712.704 1332.534
Yellow 887.238 114.713 655.394 1119.083
Green 1014.095 124.517 762.437 1265.753

ence in stereopsis between the no lens and variations in the level of illumination, but varia-
white lens conditions at each level of illumi- tions in the spectral composition of light35. The
nation. response of the visual system (spectral sensitivity)
3. stereopsis was significantly superior with the to different light levels and colours is an inherent
no lens condition compared to all six other property of the type and distribution of the photo-
tint conditions (grade B), at each level of il- receptors and neurons across the retina. Based on
lumination. the difficulty experienced in seeing depth in isolu-
minant random-dot stereograms it was postulated
Discussion and Conclusions that stereopsis is “colour-blind” and that therefore
Information regarding the visual world is de- the information is most probably carried by the
duced based on the changes in the quantity and magnocellular system35. However more current
quality of light from a source and its interaction information suggests that there are two stereopsis
with matter such that one has to consider not only mechanisms, one sensitive to luminance contrast

The South African Optometrist − March 2007 8

M Mehta, P Ramkissoon and A M Bhagwanjee

Table 2: Multivariate Analysis of Variance of Main Effects (Tint an Illumination) and Interaction Effects on Stereoacuity
Source (Sphericity Type III Sum of df Mean Square F Sig.
Assumed) Squares
illum 202129922.333 4 50532480.583 119.706 .000
Error (illum) 99624752.667 236 422138.782

tint 5535304.500 7 790757.786 8.996 .000

Error(tint) 36304320.500 413 87903.924

illum * tint 10308889.667 28 368174.631 4.760 .000

Error(illum*tint) 127772035.333 1652 77343.847

the parvocellular system to the visual centers of

the cortex, scotopic vision carries no information
regarding the colour of objects1, 6. Further, maxi-
mum visual acuity is possible in bright room illu-
mination (300 lux) for central fixation when the
cones alone respond6.
Figure 1 represents a composite illuminance
response function of stereoacuity scores when
various tinted lenses were used under varied reti-
nal illuminance conditions. As shown in Figure
1, neutral density filters worsened retinal illu-
minance and consequently the stereoacuity (that
is, the threshold values increased) for all lenses
tested as well as when no (tinted) lens was used.
The greater the density of the filter, the greater
was the reduction in stereoacuity. This meant that
and the other to chromatic contrast11. It has been as illumination was reduced, stereopsis worsened
suggested that even though the disparity range of similar to other studies which compared the rapid
the luminance-contrast-sensitive mechanism is decline in stereopsis as opposed to visual acuity
larger than that for the chromatic-contrast-sensi- under poor light conditions15, 36. While the decline
tive mechanism, these two mechanisms must in- was relatively steady up to the level of illumina-
teract before the extraction of stereoscopic depth11. tion decreased by the 0.9 ND, the illumination
In fact, under certain conditions of poor stereopsis level between 0.9 ND and 1.2 ND resulted in a
with luminance contrast alone, added isoluminant dramatic increase in the stereoacuity scores sug-
chromatic contrast does improve stereopsis signif- gesting very poor stereopsis. It is highly likely
icantly (probability summation) 11. These findings that whilst stereopsis under scotopic conditions is
formed the basis for this investigation into whether possible15, rod vision does not adequately support
tinted lenses impacted negatively on stereopsis, as stereopsis in the absence of the additive effect of
compared to no lens conditions, under decreasing the chromatic mechanism of stereopsis11. These
levels of illumination. results therefore support the luminance contrast
The Duplicity Theory in vision states that, sensitive stereopsis mechanisms. Inspection of
depending on the level of illumination, either the marginal means reveals that at 1.2 log units,
the rods respond (scotopic light levels), both the yellow and pink perform better than the other
rods and cones respond (mesopic light levels), or four tints. Even though it has been shown that
mainly the cones respond (photopic light levels)6. yellow tinted lenses do not improve visual acuity,
Since only the cones are responsible for chromat- stereopsis or contrast sensitivity, they are found
ic sensitivity and this information is carried via to enhance brightness based on rod signals car-

9 The South African Optometrist − March 2007

A comparison of the effect of reduced illumination and tinted lenses on stereopsis at near

Table 3: Post-hoc Analysis: Familywise paired t-tests for Illumination by Tint(Mean Stereoacuity)*
Pairings Mean Std. Deviation Std. Error t df Sig. (2-tailed)*
Illumination Tint
No ND no lens - white -.333 2.582 .333 -1.000 59 .321
no lens - pink -3.667 7.584 .979 -3.745 59 .000**
no lens - blue -4.167 5.612 .725 -5.751 59 .000**
no lens - brown -3.500 6.846 .884 -3.960 59 .000**
no lens - grey -4.667 7.471 .965 -4.838 59 .000**
no lens - yellow -3.509 6.932 .902 -3.908 59 .000**
no lens - green -3.500 6.915 .897 -3.927 59 .000**
0.3 ND no lens - white -.500 7.686 .992 -.504 59 .616
no lens - pink -9.000 12.849 1.659 -5.426 59 .000**
no lens - blue -9.333 12.604 1.627 -5.736 59 .000**
no lens - brown -8.667 13.957 1.802 -4.810 59 .000**
no lens - grey -8.500 14.001 1.807 -4.703 59 .000**
no lens - yellow -7.333 16.351 2.111 -3.474 59 .001**
no lens - green -8.333 13.550 1.749 -4.764 59 .000**
0.6 ND no lens - white -3.667 14.018 1.810 -2.026 59 .047
no lens - pink -35.667 64.318 8.303 -4.295 59 .000**
no lens - blue -35.333 52.253 6.746 -5.238 59 .000**
no lens - brown -34.333 52.863 6.825 -5.031 59 .000**
no lens - grey -36.667 63.691 8.222 -4.459 59 .000**
no lens - yellow -30.667 31.346 4.047 -7.578 59 .000**
no lens - green -41.667 68.325 8.821 -4.724 59 .000**
0.9 ND no lens - white -15.333 58.729 7.582 -2.022 59 .048
no lens - pink -109.667 165.887 21.416 -5.121 59 .000**
no lens - blue -128.000 172.173 22.227 -5.759 59 .000**
no lens - brown -131.333 177.761 24.769 -5.693 59 .000**
no lens - grey -104.667 168.719 21.781 -4.805 59 .000**
no lens - yellow -98.667 142.988 18.460 -5.345 59 .000**
no lens - green -114.333 135.651 17.512 -6.529 59 .000**
1.2 ND no lens - white -16.667 92.364 11.924 -1.398 59 .167
no lens - pink -283.333 446.898 57.694 -4.911 59 .000**
no lens - blue -443.333 802.616 103.617 -4.279 59 .000**
no lens - brown -443.333 793.270 102.411 -4.329 59 .000**
no lens - grey -425.000 840.223 108.472 -3.918 59 .000**
no lens - yellow -270.000 550.901 71.121 -3.796 59 .000**
no lens - green -430.000 714.807 92.281 -4.660 59 .000**
* α = .001 (.05 divided by 35 paired t-tests)
** t-test values that meet the significance criterion

rying the information along the chromatic chan- change in stereoacuity induced by tinted lenses
nels and amongst other optical explanations, a will influence visually guided performance.
selective reduction of short-wavelength light has Luria39 also found that depth perception was di-
been posited37-39. This study indictates the need minished with tinted lenses; his proposal that the
for further research to investigate between-colour reduced transmittance produced by tinted lenses
comparisons using a range of tints and depths of decreased the stereoacuity is corroborated by the
tints as well as photochromic lenses under differ- findings of this study. Therefore, the practitioner
ent levels of illumination. should ensure that tinted lenses prescribed do not
Since the white lens performed as well as the retard stereoacuity in circumstances where good
no-lens condition under the different levels of il- stereoacuity is required3. In advising a patient
lumination, but both differed statistically signifi- concerning tinted lenses, the practitioner should
cantly from all the other six tinted lenses at each question the patient not only on his sensitivity to
of the levels of illumination, it is evident that the light, but also as to the activities and levels of il-
tinted lenses adversely affected stereopsis at near lumination for which the lenses are to be used27.
under reduced levels of illumination. Pitts and Chou3 advised that care must be tak-
As it is necessary to constantly distinguish ob- en to ensure that the tint does not impair vision
jects in the environment at varying distances, any if industrial workers move quickly or frequently

The South African Optometrist − March 2007 10

M Mehta, P Ramkissoon and A M Bhagwanjee

from brilliant sunlight into a dimly lighted envi- pia, accommodation, retinal illuminance, and pupil size on stere-
ronment. The optometrist must have available a opsis. Inves Ophthal Vis Sci 1985 26 741-750.
14. Lit A, Dwyer WO and Morandi AJ. Effect of background wave-
range of tinted lenses that would be suitable for length on stereoscopic acuity at scotopic and photopic illumina-
the environment in which the wearer works or tion levels. Am J Optom 1968 45 195-203.
15. Livingston MS and Hubel DH. Stereopsis and positional acuity
intends to use the spectacles25, 27. When stereop- under dark adaptation. Vision Res 1994 34 799-802.
sis is affected by tinted lenses under poor light 16. Elkington AR, Frank HJ. Clinical Optics. 2nd ed. London: Black-
well Scientific Publications, 1984.
levels, depth perception may still be appreciated 17. Edwards K, Llewellyn R. Optometry. 1st ed. London: Butter-
using monocular cues and luminance contrast worths, 1988.
mechanisms and the individual’s performance 18. Brooks CW, Borish IM. System for Ophthalmic Dispensing: Lon-
don: Butterworth-Heineman, 1979.
may not be significantly compromised. Howev- 19. Jalie M. The Principles of Ophthalmic Lenses. 4th ed. London:
er, certain vocations require good depth percep- Association of Dispensing Opticians, 1984.
20. Lehmann OJ, Verity DH, Coombes AGA, Ah-Fat FG, Francis PJ,
tion and since stereopsis has been found to be the Ionides ACW. Clinical Optics and Refraction. 1st ed. Oxford:
most superior cue to depth perception up to 450 Butterworth-Heinemann, 1998.
metres, it is important that optometrists offer the 21. Wong BPH, Woods RL, Peli E. Stereoacuity at distance and near.
Optom Vis Sci 2002 79 771-778.
best possible advice to these patients. 22. Bhola R. Binocular Vision. Available from: URL: http://webeye.
It would be valuable to compare the levels of ophth.uiowa.edu/eyeforum/tutorials/Bhola-BinocularVision.htm.
23. Borish IM. Clinical Refraction. 3rd ed. Chicago: Professional
stereopsis using a range of solid tints from A to Press, 1970.
D as well comparing the effects of gradient tints 24. Drew R. Professional Ophthalmic Dispensing. Chicago: Profes-
to solid tints. Clinically important information sional Press Inc. 1970.
25. Drew R. Ophthalmic Dispensing. The present-day realities. Lon-
would be available if the stereopsis could be eval- don: Butterworths, 1990.
uated while performing dynamic critical tasks to 26. Fannin TE, Grosvenor TG. Clinical Optics. Boston: Butterworths,
1987.
assess the severity of the effects of the decrease 27. Hickson-Curran SB, Nasson RJ, Becherer PD. Clinical evalua-
in stereopsis on performance under different light tion of Acuvue Contact Lenses with UV blocking characteristics.
Optom Vis Sci 1997 74 632-637.
levels. 28. Kirkner R. Using contrast sensitivity testing in practice. Rev Op-
tom 133 1996 139-140.
References 29. Eperjesi F, Fowler CW and Evans BJW. Do tinted lenses or filters
1. Hubel DH. Eye, Brain and Vision. 2nd ed. New York: WH improve visual performance in low vision? A review of literature.
Freeman and Company, 1995. Ophthal Physio Optics 2002 22 68-77.
2. Barton RA. Binocularity and brain evolution in primates. 30. Peli E. Treating with spectacle lenses: A novel idea? Optom Vis
Proc Natl Acad Sci 2004 101 10113-10115. Sc 2002 79 569-580.
3. Benjamin WJ. Clinical Refraction. 1st ed. Philadelphia: WB 31. Garnham L and Sloper JJ. Effect of age on adult stereoacuity as mea-
Saunders Company, 1998. sured by different types of stereotest. Br J Ophthal 2006 90 91-95.
4. DeAngelis GC. Seeing in three dimensions: the neurophysi- 32. Fawcett SL and Birch E. Validity of the Titmus and Randot cir-
ology of stereopsis. Trends Cogn Sci 2000 4 80-90. cles tasks in children with known binocular vision disorders. J
5. Griffin J. Binocular Anomalies, Procedures for Vision Thera- APPOS 2003 7 333-8.
py, 2nd ed. New York: Professional Press Books, 1988. 33. Fricke TR and Siderov J. Stereopsis, stereotest and their relation
6. Webvision. Available from: URL: http://webvision.med.utah. to vision screening and clinical practice. Clin Exp Optom 1997
edu/KallDepth.html. 80 165 – 172.
7. Vision and Colour. Available from URL: http://www.du.edu/ 34. Robinson, P.W. Fundamentals of Experimental Psychology, 2nd
~jcalvert/optics/colour.htm#Ster ed., Prentice-Hall, New Jersey 1981.
8. Long J, Siu C. Randot stereoacuity does not accurately pre- 35. Spillman L and Werner JS. Visual Perception: The Neurophysi-
dict ability to perform two practical tests of depth perception ological Foundations. London: Academic Press, 1990.
at a near distance. Optom Vis Sci 2005 82 912-915. 36. Rabin J. Luminance effects on visual acuity and small letter con-
9. Daw NW. Visual Development. 1st ed. New York: Plenum trast sensitivity. Optom Vis Sci 1994 71 685-688.
Press, 1995. 37. Kelly S. Effect of yellow-tinted lenses on brightness. J Opt Soc
10. Simmons DR and Kingdom FAA. On the independence of Am 1990 7 1905-1911.
chromatic and achromatic stereopsis mechanisms. Vision Res 38. Wolffsohn JS, Cochrane AL, Khoo H, Yoshimitsu Y, Wu S. Con-
1997 37 1271–1280. trast is enhanced by yellow lenses because of selective reduction
11. Simmons DR and Kingdom FAA. Interactions between chro- of short-wavelength light. Optom Vis Sci 2000 77 73-81.
matic and luminance contrast sensitive stereopsis mecha- 39. Luria SM. Vision with chromatic filters. Am J Optom 1972 49
nisms. Vision Res 2002 42 1535-45. 818-829.
12. Chang YH, Lee JB, Kim NS, Lee DW, Chang JH and Han SH.
The effects of interocular differences in retinal illuminance on vi-
sion and binocularity. Clin Exp Ophthalmol 2006 244 1083-8.
13. Lovasik JV and M Szymkiw. Effects of aniseikonia, anisometro-

11 The South African Optometrist − March 2007