colourvis - University of Western Ontarioinstruct.uwo.ca/psychology/386b/colourvis.pdfelse than a certain Power and Disposition to stir up a Sensation of this or that Colour” Sir

Slide 1

“… the Rays to speak properly arenot coloured. In them there is nothingelse than a certain Power and Dispositionto stir up a Sensation of this or thatColour”

Sir Isaac Newton (1730)

Slide 2Light and Colour

• Visible light occupies the electromagneticspectrum from approx. 400-700 nm

• The wavelength of the light is correlated with the colour we experience

Slide 3 Sir Isaac NewtonThe Founder of Colour Science

• Before Newton, colour was thought of as afundamental property of objects

• Newton made several crucial discoveries:

– that colour was a subjective experience

– that white light was made of a mixture of manydifferent wavelengths of light

– that the colour experience is determined by thecombination of wavelengths that reach the eye

Slide 4Newton’s Prism Experiments (1)

• Newton’s classic experiment was to show thatwhite light could be broken down into its spectralcomponents by passing it through a prism

Newton’s Original Experiment

Slide 5

Slide 6Newton’s Prism Experiments (2)

• He also showed that:

– once broken downinto a spectrum,single componentscould not bebroken downfurther

– it was possible torecombine severalwavelengths toproduce white

Slide 7Seeing Rainbows

• The colours of a rainbow are determined byprismatic refraction in rain drops

Slide 8

Slide 9 Things to know about colour vision

• What are the phenomena of colour?• How do we describe colours? Colour specification

• How do we produce colours? Colour mixing• Colour matching. The psychophysics of colour• Colour vision theory. Trichromacy vs opponent

processing

• How is wavelength information processed by thevisual system?

• Why do some people not see colours normally?Colour deficiencies

• Is colour experience universal across species?Comparative colour vision

Slide 10The Phenomena of Colour

Coloured afterimages

+

Slide 11

+


Coloured afterimages

Slide 13

+


Colour assimilation


Colour assimilation


Colour Contrast


Colour Contrast


+

Saturation Adaptation


+



+



Colour Deficiencies

Slide 22Things to know about colour vision



processing




Slide 23

How do we describe colours?

Colour Specification

Slide 24The Dimensions of Colour

• Although we tend to think of colour interms of colour names, colour is a multi-dimensional experience.

• Each of these dimensions is associated witha different physical property of light

• There is a need for a system that allows forcolours to be described accurately andreproduced reliably

Slide 25Dimensions of Colour

• Hue– the “colour” of the

target• Saturation

– the degree ofwhiteness in thetarget

• Brightness– the perceived

intensity of thetarget

• Wavelength

• Spectral purity

• Luminance

Subjective Physical

Slide 26Dimensions of Colour

Slide 27 What colours do we see?Hue

• All discriminable colours can be describedin terms of 4 colour names:

Blue; Yellow; Green; Red

e.g. purple = red + blue brown = dark yellow cyan = blue + green

etc.

Slide 28How many colours can we see?

• Can calculate the theoretical maximumbased on the number of jnds for each colourdimension

• Wavelength Discrimination - 200 jnds• Saturation - 20 jnds• Brightness - 500 jnds• Therefore total range of possible colours

200*20*500 = 2 million

Slide 29Saturation and Brightness

• All colours can vary in saturation andbrightness

Saturation

Brightness

Slide 30What is colour for?

• Some form of colour vision is almostuniversal across species

• Colour vision capacity varies a great dealacross species

• Colour vision capacity is related to thevisual environment

• No definitive answer as to why colourvision evolved, but it seems likely that itprovided an advantage in the identificationof food sources or in mate selection

Slide 31The Specification of Colour

• The Colour Wheel

– Only givesinformation abouthue


• The Colour Disk

– Gives informationabout hue andsaturation


• The Colour Solid

– Gives informationabout hue, saturationand brightness


• The colour shapes provide a qualitativedescription of colours

• There is a great need for a precisequantitative system to ensure consistency inpaints, dyes, inks, etc.

• Several systems in use

Slide 35 The CIE System• The CIE system wasdeveloped to provide adescription of any givencolour using a set of“primary” wavelengths and a“standard” observer.

• Based on the fact thatdifferent wavelength mixturesproduce different coloursensations

• A colour is defined by therelative amounts of each ofthe primaries needed and canbe plotted as shown




processing




Slide 37

How do we produce colours?

Colour Mixing

Slide 38Producing Colours

• Although hue is determined by wavelength,we are very rarely exposed to singlewavelengths

• Most of the time, what we see is a mixtureof many different wavelengths

Slide 39Colour and Wavelength

• Typical wavelength mixtures for colours we see

Shades of Grey Broad-band Band-pass

Slide 40Determining the colours we see

• Ultimately, the wavelength composition ofthe light that strikes the retina determinesthe colour we see.

BUT

• The wavelengths that reach the eye dependon several factors.

Slide 411. The spectral composition of the source

Different light sources have very different spectral compositions

HeNe Laser

Slide 42 2. The spectral reflectance of the surface

Achromatic surfaces Common pigments

Slide 43Reflectance of some common objects

Slide 44 Determining the colours we seeThe wavelengths that reach the eye represent the productof the source and the object wavelength distributions

Source

Surface

Product

Slide 45Metamers

• It is possible to produce the same coloursensation using a variety of wavelengthcombinations

• When two colours with differentwavelength compositions generate identicalcolour sensations they are said to bemetameric

Slide 46

Slide 47Colour Mixtures

• Colour mixing refers to the way in whichwavelength combinations may be deliveredto the eye

• Most of the time we are aware simple of theend result and are not concerned with theprocess of producing specific colours

• However, sometimes we wish to ensure thatwe can create a specific colour by mixingwavelengths

Slide 48Mixing Colours

• There are two ways in which we can alterthe wavelength composition of lightreaching the eye

• For subtractive mixtures, the sourceproduces a wide range of wavelengths,some of which are eliminated

• For additive mixtures, wavelengths fromdifferent sources are combined

Slide 49Primary and Complementary Colours

• Primary Colours– are those colours that will give the widest range

of colours when mixed together.– these are different for additive and subtractive

mixtures

• Complementary Colours– colour “opposites”– have different effects when mixed additively or

subtractively

Slide 50Subtractive Colour Mixing

• Applies to paints, dyes, inks, etc., and to lightpassing through filters

• Begin with broad range of wavelengths then takeaway some away– resultant colour is always darker than the

components

• Primaries are blue, yellow and red

• Mixing complements produces black

Slide 51Subtractive mixtures

Through selective reflection Through selective absorption

Slide 52

Slide 53 Colours from subtractive mixtures

ReflectiveSurface

Single filter Multiple filters

Slide 54Additive Colour Mixing

• Applies when light is coming from more than onesource; e.g. spotlights, TVs, magazine images

• Light reaching the eye is the sum of thewavelengths of the sources

– final result is brighter than the components

• Primaries are blue, green and red

• Mixing complements produces white

Slide 55Additive colour mixing

Slide 56

Print and TV coloursare produced additivelybecause the individualcolour elements are toosmall to be resolved




processing




Slide 58

Colour matchingThe psychophysics of colour

Slide 59

• Colour mixing experiments showed that manycolours could be produced by varying therelative proportions of the componentwavelengths

• Colour matching experiments were designed toquantify the component mixtures

• These data were then used to infer somethingabout the underlying mechanisms

Slide 60Colour matching experiments

• A metameric match means that two different setsof wavelengths are having identical effects on thevisual system

• To understand this, we need to understand thecondtions under which metamerism occurs

• But natural metamersp have complex spectrald distributions

Slide 61Metameric matches

Need to use a simplerarrangement with onlya limited range of wavelengths

Slide 62Two important findings from colour matching:

1. All spectral lights could be matched by mixingseveral other wavelengths (primaries) together invarying proportions

2. A maximum of three primaries was needed tomatch all spectral lights

This result led to the conclusion that there must bethree classes of receptor responding to light ofdifferent wavelengths.




processing




Slide 64

Colour vision theory. Trichromacy vsopponent processing

Slide 65Trichromatic Vision

• Why should the fact that we need threewavelengths to make all spectral matchesmean that we have three receptors?

• Can understand this if we understand whyrods are colour blind

Slide 66The Purkinje effect: Rods are colour blind

• The Purkinje effect shows that the transitionfrom cone to rod vision results in a loss ofcolour sensation

• As light levels decrease:– colours begin to fade and eventually everything

looks grey

– reds and yellows look very dark, while bluesand greens look relatively bright

Slide 67 Because the rods are more sensitive at shorter wavelengths, these look relatively bright in dimlight

Slide 68Why are the rods colour-blind?

• The spectral luminosity functions show only thatthe rods are more sensitive at shorter wavelengths.

Question: Why do the rods not permit colourvision?Answer: There is only one photopigment in therods.

Explanation: Receptors can only signal that theyhave been stimulated by light, not whatwavelength has stimulated them (The Principle ofUnivariance )

Slide 69 The response of a single receptor systemto light of different wavelengths:

• The output of a photoreceptor is a product ofthe intensity of the light and the sensitivity ofthe receptor to that particular wavelength

Output = Intensity * Relative Sensitivity

• This means that two lights of differentintensity can produce the same effect if theirintensities are adjusted appropriately

Slide 70

• If a photoreceptor can onlysignal the number of quantathat it has absorbed, then:

• for a single photoreceptor, itwould be possible to adjust therelative intensities of any pairof wavelengths so that theyproduce exactly the same effectof the photoreceptor andtherefore would look identical.

• Such a system would beconsidered monochromatic

Slide 71 The response of a dual photoreceptorsystem to light of different wavelengths

• If there are two photopigments with overlappingspectral sensitivities then it is impossible to adjustthe relative intensities of two single wavelengths toproduce a match

Slide 72 The response of a dual photoreceptor system to lightof different wavelengths - single primary

Slide 73 The response of a dual photoreceptorsystem to light of different wavelengths

• To produce identical outputs from two detectorswith different, but overlapping, sensitivityfunctions, one has to to adjust the intensities oftwo different wavelengths simultaneously tomatch any other wavelength

• Because two primaries are required to make amatch, the system os said to be dichromatic

Slide 74 The response of a dual photoreceptor system to lightof different wavelengths - two primaries

Slide 75• It follows from the results of the colour matching

experiments that if three primaries are necessary tomake a match, then there must be three receptors

• For colour matching, the number of primaries neededto match all spectral lights implies the number ofunderlying receptor systems

• By plotting the relative intensities of the primarywavelengths needed to make a spectral match, it ispossible to derive the shape of the underlying receptorsensitivity functions

• These data provided evidence in favour of thetrichromatic theory of colour vision

Trichromacy

Slide 76 Trichromatic TheoryYoung-Helmholtz Theory

• The original suggestion that we have only alimited number of photoreceptive mechanismswas based on logic rather than experiment

• Thomas Young (1802) realised that we could notindividual receptors for all the colours we see.

• He proposed that were only three types ofreceptor, each responding to a wide range ofwavelengths

Slide 77 Trichromatic TheoryYoung-Helmholtz Theory

• The original suggestion that we have only alimited number of photoreceptive mechanismswas based on logic rather than experiment

• Thomas Young (1802) realised that we could notindividual receptors for all the colours we see.

• He proposed that were only three types ofreceptor, each responding to a wide range ofwavelengths

• Helmholtz provided the psychophysical evidenceto support this theory with his colour matchingexperiments

Slide 78Opponent-Process Theory

• Although colour matching experiments could beexplained easily by trichromatic theory, there werea number of colour phenomena that did not seemto fit with this theory

– afterimages

Slide 79

+

Slide 80

+

Slide 81

+

Slide 82

+

Slide 83

+

Slide 84

+


• Although colour matching experiments could beexplained easily by trichromatic theory, there werea number of colour phenomena that did not seemto fit with this theory

– afterimages

– simultaneous colour contrast

– “fundamental” character of blue, green, red,and yellow

• Hering suggested that these colours were linked insome way


• Hering proposed that red-green, blue-yellow, andblack-white were organised in some opponentfashion so that the activation of one would supressthe other

• On this basis it was possible to explain manycolour phenomena

• Immediate difficulty was that there was nocandidate mechanism as there was for theTrichromatic Theory




processing




Slide 88The Physiological Basis of Colour Vision

• For many years there was disgreement aboutwhich colour vision theory was correct

• Until the 1960s, virtually all of the available datawere psychophysical and they supported Y-H

• A new technique, micropspectrophotometry(MSP) allowed for direct measurement of conephotopigments

• First measurements of cone absorption curvesmade by Brown and Wald in1964. Showedpresence of three cone pigments

Slide 89Microspectrophotometry

Test Beam

ReferenceBeam

Slide 90 Absorption spectra of the cones

Slide 91The Molecular Basis of Colour Vision

• Research over the past 15 years has shown that theabsorption spectrum of a photopigment isdetermined by the sequence of amino acids in theopsin protein

• Small differences in the sequence shift the peak ofthe absorption curve along the spectrum

Slide 92 Trichromatic vs Opponent Process TheoryWas Helmholtz right?

• MSP appears to vindicate Y-H theoryBut

• Some psychophysical evidence in favour ofopponent process (mainly from colourcancellation experiment)Then

• Physiological data began to appear that showedneurons responding in an opponent fashion

Slide 93

Svaetichin recorded fromthe horizontal cells of fishand found some cells that responded by hyperpolarizing to some wavelengths and depolarizing to others.

This was the first evidencefor an opponent system

Slide 94Neural processing beyond the receptors

• Advances in technology allowed for recordingfrom single neurons

• One characteristic of neurons is that they have aspontaneous rate of firing. This means that theycan respond by increasing or decreasing theirfiring rate

• Recordings from the lateral geniculate nucleusshowed spectrally opponent responses

Slide 95

Slide 96

Excitation and Inhibition

Slide 97Receptive Fields

• A receptive field is the area on the retinathat feeds into a single neuron.

• If this area is stimulated by light the neuronwill change its firing rate

Slide 98 Receptive fields

Slide 99

R+G- R+G-

Opponent Process Receptive Field

Red light on Green light on

Slide 100Responses of a colour opponent cell

Slide 101Was Helmholtz right?

• Both thetrichromatic andopponent processtheories arenecessary toexplain most of thephenomena ofcolour vision

Slide 102How do trichromatic cones become opponent processingganglion cells?

Slide 103• While Y-H and opponent theories together explain

most colour phenomena, Land’s experiments showthat other factors need to be taken into account

• Opponent cells can signal the wavelength of anobject efficiently but don’t account for spatialeffects like such as simultaneous contrast,coloured shadows, etc.

• One class of neuron that might be involved is thedouble-colour-opponent cell

Slide 104Receptive fields

Slide 105A double opponentreceptive field

G+R -

R+G-

A cell with this receptivefield arrangement wouldrespond well to a red objecton a green background

Slide 106 Colour Processing in the Visual Cortex

•Double colour opponent cellsare found in specific areas of the visual cortex

•Because of their appearance they are known as “blobs”

Slide 107The Physiological Basis of Colour Vision

• For many years there was disgreement aboutwhich colour vision theory was correct

• Until the 1960s, virtually all of the available datawere psychophysical and they supported Y-H

• A new technique, micropspectrophotometry(MSP) allowed for direct measurement of conephotopigments

• First measurements of cone absorption curvesmade by Brown and Wald in1964. Showedpresence of three cone pigments

Documents

colourvis - University of Western Ontarioinstruct.uwo.ca/psychology/386b/colourvis.pdfelse than a certain Power and Disposition to stir up a Sensation of this or that Colour” Sir