We have measured sensitivity for discriminating the spatial phase of m
ulti-harmonic and two-harmonic patterns modulated either in luminance
or in chromaticity (red-green). The multiharmonic patterns were either
highpass squarewaves, lines or ramps. For all patterns, contrast thre
sholds for discriminating 0 from 180 deg phase were similar to those f
or discriminating -90 from 90 deg, for luminance or chromatic modulati
on (or both). For all types of multi-harmonic patterns, the ratio of c
ontrast thresholds for the phase discrimination to that for pattern de
tection was the same for luminance and chromatic modulation, and for c
ombinations of both. Similarly, phase thresholds, the minimum detectab
le differences in phase (about a mean 0 deg), were the same for chroma
tic and luminance patterns, provided that contrast was scaled to equat
e detection thresholds of the patterns. Similar results were observed
for simple three-harmonic patterns (f+ 2f+ 3f), and for (f+ 2f) two-ha
rmonic patterns, Strangely, however, two-harmonic patterns of f+ 3f (f
irst two terms of square-wave) of moderate to high spatial frequency d
id show a two-fold advantage for luminance over colour, as Troscianko
and Harris (1988) have previously reported (Vision Research, 28, 1041-
1049), possibly because the two harmonics have a greater separation in
frequency. However, for most classes of patterns, sensitivity for spa
tial phase is as good for chromatic as for luminance modulation, sugge
sting that similar sorts of mechanisms operate under these two conditi
ons.