PSY280 - 6. Colour Flashcards
Color
not physical property, but it’s related to a physical property (wavelength)
diff wavelengths of light are interpreted by system as diff colour
Color
words we use to describe colour evolves over time
every culture/language, 1 colour emerges first - red
as language evolves, more colours identified
blue last colour, not much in nature is blue - we just collectively call the sky blue
Color
colour not actually a physical property, just related to one
result of interaction betw stimulus + nervous system
rats just interpreting wavelengths differently we do
wavelength of light reflected is a property
function of color: signaling
tell diff betw safe food/rotten food - food safety
illness can be expressed as changes in colour of skin
traffic lights
The function of color: perceptual organization
allows us to perceptually separate objects
facilitate finding fruit
The function of color: object recognition
to identify objects
association betw colour + object
changing colour messes with object identification
typical colours/atypical colours/black + white: effects for reaction time + accuracy for object recognition
Achromatic colors
experienced when light is reflected equally across the spectrum.
Chromatic colors (hues)
experienced when light is selectively reflected – when some
wavelengths are reflected more than others
experience green, some wavelengths of blue reflected
yellow: represented at 570 - yellow, actually has long wavelengths
translate physical wavelengths - perception of colour
3 problems:
- Detection
- Discrimination: we have to be able to tell diff betw wavelengths
- Appearance (constancy): assignment need to go with certain objects + not change in different lighting conditions
Rods
one kind of photopigment (rhodopsin protein + retinal)
detection of light in the eye by wavelength
Cones
1 of 3 kinds of photopigments (“opsins” + retinal).
they vary based on the opsid
rhodopsin, diff opsin but same retinal
Opsin
determines spectral sensitivity of photoreceptor
dictates which wavelengths is strongly activating
relative proportion of light absorbed vs wavelength
It’s combo of sensitivities that gives
us the visible spectrum
diff absorption spectrums - diff wavelengths absorbed effectively
peak sensitivities of 3 cones roughly correspond to blue, green, red
in terms of short, medium, and long wavelengths
S-cones
5–10%
none in fovea- fovea not sensitive to this wavelength
M-cones
~30% more in fovea 531 nm sensitive to range of wavelengths green: each stimulated to some degree, M is just stimulated the most
L-cones
long
~60%
more in fovea
L-cones: 558 nm
discrimination of color
Each photopigment sensitive to a range of wavelengths.
response strength varies for diff wavelengths of the same intensity
discrimination of color
pattern of activation important in discriminating colour
response curve for single photoreceptor
light presenting to photopigment same intensity, only thing varies is wavelength
The problem of univariance
output of a single photoreceptor is
completely ambiguous
any mix with properly adjusted wavelengths
The problem of univariance
H
Trichromatic Theory
it’s all relative
Color vision depends on 3 different receptor mechanisms (cones)
Trichromatic Theory
s cone response: 450 nm big/625 nm absent
m cone response: 450 nm moderate/625 nm moderate
l cone response: 450 nm smaller/625 nm big
Trichromatic Theory
Diff wavelengths of light produce a unique pattern of activation for 3 cones.
mess with intensity, response size will change, but relationships will not.
relative proportion will stay the same regardless of the intensity
Trichromatic Theory
light we perceive as orange 625
stimulating L cone, M-cone 70% of response of L cone
blue: s cone response big, m-cone moderate, l-cone small
each colour elicits 3 levels of activation for each cone
Trichromatic Theory
Newton (1666): color not a physical property of an object, but in the light.
sunlight can be split into discrete components
diff components make up the diff colours
white light contained all wavelengths
Young-Helmholtz Theory
Using color matching technique, Young (and Helmholtz) found that 3 mixing lights are needed to match any
reference light.
Metamers
Physically diff mixtures of wavelengths that look identical.
M-Cones L-cones
red light 40 units 80 units
green light 80 units 40 units
total 120 units 120 units
yellow light 120 units 120 unitsMetamers
relative pattern of activity that cause perception of colour
now we know the mechanisms are the cones
perception is combo of activation across conetypes
level of activation from right combo is going to be the same as when you present 1 wavelength
if pattern of activity is the same = experience is going to be the same
Metamers
TV only emits blue, red & green light. What about yellow?: as long as you can match the pattern of activation then you can match the perception
Metamers
Pass white light through a yellow” filter: filter that absorbs short wavelengths.
TVs use trichromatic pixels: each pixel has g, b, r - get yellow by activating g+b
additive color mixing: mixing light
Perceived color depends on wavelengths reflected
S - blue, M - green, L - red, M+L - yellow, L+M+S - white
when combined together, all wavelengths are reflected
we add together wavelengths to predict final perception
additive color mixing: mixing light
colour wheel: combo of 2 wavelengths
perception can be predicted by colour in the middle
nonspectral colours: don’t appear in wavelengths - purple + brown
subtractive color mixing: mixing paints
mixing blue + yellow paints
blue paint: S reflects all, M reflects some, L absorbs all
yellow paint: S absorbs all, M reflects some, L reflects some
blue + yellow paint: S absorbs all, M reflects some, L absorbs all
subtractive color mixing: mixing paints
when you mix paints, fewer wavelengths get reflected
so only ones reflected are those reflected by both
subtract out what isn’t reflected by both
Reflectance Curve
produce additive colour mixing with paint through pointilism
could produce small points of paint + produce colours not on face, but at a distance would produce flesh colour
appearance of color
Using relative activity of 3 types
of cones, we can discriminate more than 26000 colors.
Physically, colors are distinguished based on hue, saturation & brightness.
appearance of color
saturation: amount of hue present - varies from outside to inside
diff hues outside boundary of wheel
brightness: physical intensity of light
reddish green is not possible - more like brown
after images
due to adaptation
Herring made 2 observations
1.When we assign objects or surfances a color, some combinations of colors are never used.
after images
Adaptation to colors produces
afterimages which reflects polarity of colour combination
suggested alternative interpretation
Opponent Color Theory
Perception based on 3 mechanisms (receptors) each based on opponency betw colors: b+w, r+g, b+y
observing how ppl behaved- phenomenology
1700s, not until 1960s that we found proof
Opponent neurons
(retina, LGN,V1) show an excitatory response to one color and an inhibitory response to its opponent color.
Opponent neurons
at the level of the retina, LGN + V1
centre surround organization
a) single-opponent: suppressed activity as green surround is activated + no red centre
b) red surround reduces inhibition + excites centre
best: when whole cell is bathed in red - excites whole centre + none of surround
Opponent neurons
Activity in opponent neurons is relative to baseline: compare R/G activity over whole visible spectrum.
organization can change, but colour combo doesn’t change
gradual change instead of step function
Opponent neurons
blue stimulates g a little bit, but green stimulates it maximally
yellow stimulates some of red which decreases activation, but red suppresses it (negative threshold)
changes as smooth curve as you vary wavelength presented
Adaptation
reduction in firing rates of neurons with continued application of a stimulus: adapt to green light
white has all wavelengths
Adaptation
green response” is reduced due to adaptation red is ok, so white minus green looks red neuronal fatigue: neurons get tired B+/Y-: high response for S, small response for M+L
So is color perception based on the trichromatic theory or opponent process theory?
It’s both!
both but operate on diff levels
Convergence
Convergence of excitatory & inhibitory inputs from photoreceptors produces opponency.
M + L connected to cell can create opponency
Y is combo of M+L wavelengths: M+L feed into Amecrine cell which feeds into retinal excitation of Amacrine inhibits retinal cell
constancy
A purple cup looks purple under different lighting conditions despite interactions between illumination & properties of the cup!
constancy
If only one wavelength is available for reflection, everything will appear that color.
interaction betw illumination + properties of object (wavelength most reflected)
most lighting conditions include more than 1 wavelength
constancy breaks down because only violet light is reflected
color constancy
tendency of a surface to appear the same color under a fairly wide range of illuminants.
purple = short + long wavelengths reflected
how the cup appears also a function of illuminance
sunlight is pretty equal reflection
skylight: lots of S, less of M + L
curves generated by purple cup are diff, output is diff, but we still perceive it as purple
Chromatic Adaptation
Skylight is rich in short wavelengths, your eye adapts to short wavelengths, decreasing your sensitivity to short wavelengths.
neuronal fatigue specific to illumination present
Skylight: S wavelengths are going to adapt because lots of S
pattern of activation becomes more similar, not identical (not a perfect system)
Effects of the surrounding
Illumination of colored
objects changes in a relative
manner.
none of surfaces will reflect L because it’s not there to reflect
apples will reflect the most of those very little L wavelengths presented
nonred are reflecting none of the few L wavelengths
Brightness
changes across shadow boundaries, hue does not.
as long as hue doesn’t change across brightness changes
because side is a shadow, 2 cubes interpreted differently in diff illumination
Memory Color
People know the color of familiar objects.
asked to reduce the chromaticity until it matched background (gray)
participants made banana slightly bluish
needed to overcompensate
fight the memory colour to produce the opponency
Light Constancy
we see whites, greys & blacks as staying about the same
shade under different illuminations.
ratio principle solves this problem
Light Constancy
Of the total light striking an object, an object’s reflectance is the proportion of light that the object reflects into our eyes.
‣black: 5-9% (9/100 or 900/10,000) ‣gray: 10-70%
‣white: 80-95% (90/100 or 9,000/10,000)
Light Constancy
under low lighting: less light reflected, but ratio of reflection is still the same with high key lighting
proportion stays the same in both illumination
Ratio Principle
As long as the ratio of reflectance for the object and its surrounding remains the same, the perceived lightness will remain the same.
works best in uniform lighting condition
our typical experience is variation in illumination
(a) vs. (c) = reflectance edge
(a) vs. (b) = illumination edge
a is reflecting more light than c - a is lighter than c
edge betw 2 areas where reflectance betw 2 surfaces changes
border betw 2 areas created by diff light intensity in the 2 areas
information in shadows
penumbra is the fuzzy border of shadows.
bottom up reasons we can tell diff betw painting + real shadow - penumbra
on right image, silhouette is too sharp, cues that it’s a painting
Top-down contribution
If we know a piece of paper is folded, our visual system uses that information to interpret change in illumination.
if we eliminate top down contribution: look at it through tiny hole
ability for colour + light constancy is eliminated
Top-down contribution
Lightness can be affected by the way elements are perceptually organized.
gestalt property
overlay 4 circle,
see them as darker if overlay constituted as light
vice versa
infant color vision
Children can see color within the first 3–4 months of life.
Habituation
when an animal learns to ignore an uninformative repetitive 50 stimulus. percentage of response from 1st response how it changes with repetitions with repetition animal learns to ignore
looking paradigm
present colour patches to infants
as soon as they stop paying attention you take it away
with each repetition decrement in looking time
present new colour to see how it compares with test condition
looking paradigm
present a diff green/a blue colour patch
babies categorize green the same way we do because treat it similarly as 510
480: still some habituation but significantly less - diff category of stimulus
looking paradigm
results were replicated even when researchers controlled for the confound.
ifferent wavelengths, different brightness
different brightness (same wavelengths)
different wavelength (same brightness)
in original experiment: failed to control for brightness
replicated even when controlled
