Colour vision Flashcards
What was Newton’s theory of colour vision?
Before Newton it was believed that colour was produced from a mixture of light and dark.
In 1666 Newton observed sunlight shining through a slit in a shuttered window producing a white image on the wall with a blue upper edge and red lower edge.
He then found that introducing a prism into the light revealed the full spectrum of light in the projected image.
Introducing a second prism into the path of the spectral beam combines the coloured rays to form white light again.
This white light is composed from a mixture of wavelengths of light ranging from 400nm-700nm.
When light passes through a prism, the shorter wavelengths are refracted more than the longer ones, spreading the constituent wavelengths out across the image plane.
Each wavelength is associated with a characteristic hue which we recognise as the colours of the rainbow.
A rainbow is a natural prism due to the sun’s rays being refracted by a band of rain.
It is a reversible process as the separated rays can be recombined to reproduce the original mixture which is perceived as white light.
What is colour vision?
the ability to discriminate amongst the different wavelengths of light
helps animals to separate objects from their background and conveys potentially useful information about the surface properties of those objects.
Many animals and plants exploit this by using colour signals to manipulate the behaviour of other animals.
What is colour mixing.
Additive mixing involves shining different colour lights onto the same part of a screen, will ultimately produce white.
Subtractive mixing refers to mixing paints and will ultimately produce dark grey.
A paint is colour because it absorbs some wavelengths and reflects others
if two colours mix the range of wavelengths reflected decreases as only the wavelengths common to neither of the mixed colours will be reflected e.g wavelengths common to not blue and not yellow.
Additive colour mixing is our point of interest.
When two coloured lights are projected onto the same part of the screen the sum of their wavelengths are reflected, producing a new hue.
This is the principle that underlies colour rendering on luminous screens like TV and computers.
what was Young’s (1802) colour matching experiment?
Thomas Young (1802) Discovered that if you shine a target light onto a screen it is possible to match the appearance of the light with a mixture of 3 testing lights as long as their relative intensities could be adjusted and are primaries in the sense that they cannot be matched by a combination of the other two. From this observation, Young inferred that wavelength discrimination relies on three types of particles in the eye. Each corresponds to a particular range of visible light and are what we now refer to as cone photoreceptors.
How did Helmholtz contribute to the theory of trichromacy?
Proposed that the three types of photoreceptor could be classified as short, medium and long wavelength preferring.
The perceived hue of a light is determined by the relative strengths of the signals elicited by the light in the three types of receptors.
Forms the basis for the Young-Helmholtz theory of trichromacy.
What evidence to MacNicohol and colleagues (1964) contribute to the Young-Helmholtz theory of trichromacy?
MacNicohol and colleagues directly measured the absorption spectra of photopigments of single cone cells obtained from dissected human retinas.
Sucked indiviudal cone cells from freshly dissected retina into the tip of a glass micro pipette
shone beams of light through the cone cell and measured how much of that light was absorbed
Confirmed the existence of three different types of cone receptors, each with a different photopigment.
Combining this with empirical observations of monochromats enables the accurate estimation of the relative sensitivities of each type of cone to different wavelengths of light.
Short wavelength cones peak at 419nm, medium at 532nm and long at 558nm.
What is the principle of univariance and why is it relevant to colour vision?
According to Rushton’s Principle of Univariance, a single cone mechanism which signals photon capture by hyperpolarisation confounds wavelength with light intensity.
This means that fewer photons of the optimal wavelenght could generate the same response as more photons at a less optimal wavelength, thus rendering the output ambiguous.
It is an issue of metamerism, two lights producing identical photoreceptor responses.
Adding a second cone cell with an overlapping range of sensitivity and a different peak sensitivity helps to disambiguate the response so metamers are less likely.
Increasing the number of cone cells increases spectral resolution as the visual system is better at discriminating between two close wavelengths but at the cost of spatial acuity as the distance between different cones increases.
Three cones is the best compromise between spatial acuity and spectral resolution.
The fovea is dichromatic as there are no short wavelength cones because spatial acuity is the priority.
Some people are completely dichromatic and are partially colour blind as they can discriminate colour but are more prone to metamerism and tend to confound hue with saturation.
Monochromats have only one type of cone cell.
What is colour opponency?
Hering (1872) proposed that colour is encoded by colour opponent processes.
Observed that some coloured lights blend when mixed but others cancel each other out e.g red and yellow make orange light but red and green make yellow which is not red-green.
Proposed that the pairs of colours that cancel each other out were represented by opponent channels in which the responses elicited by each colour were mutually inhibitory.
What are colour after images and what do they tell us?
staring at a black dot surrounded by 4 colours which are then removed leaves the illusion of seeing after images which correspond to the complementary colours of the original patches e.g removing red dot induces a green after image.
If the black dot is instead surrounded by four large black circles, removing the larger circles results in bright white images in the corresponding positions.
Based on these observed pairings from colour mixing experiments and colour after images Hering proposed that colour is coded for by 3 opponent channels: red-green, blue-yellow and light-dark.
Does the circuitry needed for colour opponency exist?
In 1958 the neural basis of colour opponency was discovered in monkeys.
Some retinal ganglion cells and relay neurons in the LGN have on-centre, off-surround concentric RFs and vice versa.
These RFs form the basis of the light-dark channel.
There are analogous cells in the retina and LGN of the monkey that are colour opponent.
The most abundant of these are red-green opponent cells.
All possible permutations of red-green ‘on-offness’ are represented in the centre and surround of different RFS e.g red on-centre red off-centre green-on centre green-off centre.
More correct to say long wavelength on centre, medium wavelength off surround, as inputs are wavelengths of light not colours.
If the whole RF is filled with white light the excitation and inhibition will cancel out.
If the uniform light has a reddish cast then the excitation due to the long wavelengths of light will exceed the inhibition due to the medium wavelengths of light and the cell will signal red.
How does colour opponency explain colour appearance.
The channels are independent and can be envisaged as an orthogonal axis in 3D space.
Blue-yellow axis, red-green axis and light-dark axis.
Characterise colour appearance in terms of hue, saturation and brightness which can be represented as polar coordinates.
Hue traces the colour wheel and is parallel to the colour plane.
Saturation is increasing distance from the origin which represents white.
Brightness is the height above the colour plane
Any given colour specified in terms of hue, saturation, and brightness corresponds to a specific balance of excitation and inhibition in the red-green, blue-yellow and light-dark dimensions
The trichromatic system provides the inputs to the colour opponency channels so the Young-Helholtz model is compatible with Hering’s theory.
What is colour constancy according to Newton?
The surface of objects are not coloured but have characteristic reflectance for some wavelengths more than others
The perceived colour is an attribution of the brain based on the reflected wavelengths
The ability of the visual system to attribute the correct reflectance properties or colour to a surface regardless of the changes in the spectral composition of colour cast of the incident light is referred to as colour constancy
What is Land;s Colour Mondrain demonstration?
relies on the visual system pooling information from across the entire visual field as exemplified by Land’s first retinex theory.
Demonstrated by colour constancy: Covered a large board with variagated overlapping squares of coloured paper placed in front of an array of 3 projectors.
Light was projected through either a short, medium or long wavelength filter.
A spot photometer was used to measure the amount of short, medium and long wavelengths of light reflected at any patch.
The observer could vary the intensity of each projector but the colours did not change.
This demonstrates the importance of light surrounding the patch for the target patch to be perceived correclty regardless of changes in the incident light .
Covered the entire board with a mask of black card apart from a small aperature over any given patch.
observers varied the intensity of the short, medium and long wavelenghts of light until the patch appeared grey, which was possible.
When the black card was removed the patch reverted immediately to its original colour despite no change in the balance of wavelengths of light.
Shows that the light reflected from the surround is takne into account when correcting for the colour cast in the incident light.
What is the theory of local adaption?
Relies on local mechanisms which involve small, independent RFs such as those found in the retina and LGN
Illusion of seeing two apparently identical images. Stare at white dot in centre, image replace with one green and one red square mask for several seconds. When reverted it appears the images have changed as if the visual system has corrected for the colour cast of the mask. Moreover different corrections were induced in different parts of the visual field which means that any corrections applied could not have been global.
Colours appear different as a result of local adaptation
Can test whether this occurs centrally in the cerebral cortex or peripherally in the retina or LGN by interocular transfer test
This is because the information from the eyes is kept separate in the LGN and only combined in the visual cortex
If the illusion is repeated with one eye shut. The images appear unchanged through the unadapted eye which means that the adaptation must have occurred in the retina or LGN
Not a dichotomy - both different adaptation/cellular
However this does not explain the phenomenon in Land’s experiment.
comparing apples and pears - comparing two different things that shouldn’t be compared.
What can we conclude about colour vision?
In order to explain colour vision we need to explain 3 phenomena: colour matching, colour appearance and colour constancy.
Colour matching is explained by the Young-Helmholtz theory of trichromacy that the spectral composition of light is encoded by the relative strengths of the signals elicited by the light in three types of cone receptor
Colour appearance is explained by Hering’s opponent theory in which the colour of light is encoded by signals in three colour opponent channels
Colour constancy may depend on global processes or local processes or a combination of both
Global processes would require large RFs and so must occur in the brain
Local processes could rely on small, independent RFs which could be in the retina or LGN.
