© 2007 The Authors
Journal compilation © 2007 Optometrists Association Australia
Clinical and Experimental Optometry 90.4 July 2007
272
CLINICAL AND EXPERIMENTAL
OPTOMETRY
Clin Exp Optom 2007; 90: 4: 272–281 DOI:10.1111/j.1444-0938.2007.00162.x
Key words: age, contrast constancy, contrast matching, contrast sensitivity, contrast threshold
The use of achromatic sine-wave gratings
with varying spatial frequencies is an effi-
cient tool to study contrast perception. It
is known that contrast perception be-
haves differently at threshold and supra-
threshold levels.
1
Research on contrast
threshold with varying spatial frequencies
has shown that the maximum contrast sen-
sitivity is at two to five cycles per degree
(cpd).
2–6
Factors such as grating area,
duration time, mean luminance and reti-
nal location have a significant effect on
contrast threshold.
6–13
It could be argued
that studies of supra-threshold contrast
perception give us more information
about our day to day visual perception.
Supra-threshold contrast matching
functions (CMFs) have decreasing spatial
frequency dependency with increasing
contrast. At higher contrast they take the
shape of a wide band-pass or become
flat.
1,6,14–17
This phenomenon was termed
‘contrast constancy’
1
and holds for a wide
range of spatial frequencies and retinal
eccentricities. Because of this contrast
constancy, we perceive objects at a con-
stant contrast regardless of the size and
distance. Several studies have demon-
strated conditions under which contrast
constancy breaks down. Peli and co-
workers
6
tested CMFs under natural view-
ing conditions, showing that below 8 cd/
m
2
, contrast matching is dependent on
luminance. Legge and Foley
18
agree that
contrast processing above 10 per cent con-
trast is the invariant with number of grat-
ing cycles. With dichoptic presentation
and a long duration of light adaptation,
Hess
19
showed that at very low luminance
(less than 0.02 cd/m
2
) and higher spatial
frequencies (10 cycles/degree), there
were substantial deviations from contrast
Ming Mei MD MSc
Susan J Leat PhD FCOptom FAAO
Jeffery Hovis OD PhD FAAO
School of Optometry, University of
Waterloo, Waterloo, Canada
E-mail: m2mei@uwaterloo.ca
Background: The effects of age on contrast threshold are well known but little is known
about its effect on supra-threshold contrast perception. This study examines supra-
threshold contrast matching and the effects of age in naïve observers.
Methods: Two age groups (from 20 to 50 years with 14 subjects and 51 years and older
with 15 subjects) participated in the study. Contrast threshold and supra-threshold
contrast matching up to 8.53 cycles per degree were measured.
Results: Both age groups demonstrated some degree of contrast constancy at medium
and higher contrasts but this was not perfect even at the highest contrast tested (55.9
per cent). There was no overall effect of age on supra-threshold contrast matching
(p = 0.086) but there was an interaction between age and spatial frequency (p < 0.001).
The plots of matched contrast against standard contrast showed that for some spatial
frequencies, the slope was significantly different from unity, indicating a gain in the visual
system for supra-threshold perception. This was still true when corrected for threshold
differences.
Conclusion: Contrast constancy exists in a larger group of naïve subjects of different
ages but does not perfectly compensate for the differences in thresholds. The results are
discussed in terms of the currently proposed models of contrast perception.
Supra-threshold contrast matching and the
effects of contrast threshold and age
Submitted: 30 October 2006
Revised: 8 February 2007
Accepted for publication: 2 March 2007