PSYC10006 Chap.7 Colour Vision: Physics, Theories and Constancy
Colour Vision: Physics, Theories and Constancy
Colour is not a property of light. Light has wavelength; colour is what a visual system makes of it, which is why the module opens with what colour vision is for before describing what it is made of. The physics section separates two mixing rules that give different answers for the same pair of colour names, so the rule has to be selected before any prediction is made.
Then come two accounts of colour coding that are usually taught as rivals and are better read as describing different stages: three cone types whose relative activity codes colour, and opponent channels that predict which colour experiences are impossible as well as explaining afterimages. The module ends on a genuinely hard problem.
The light reaching your eye is the product of a surface's reflectance and the illumination falling on it, and recovering the stable term from that product is underdetermined.
What this chapter covers
- 01
Four jobs colour does, and why only one of them explains why an eye can see colour
- 02
Wavelength and the visible band, with ultraviolet and infrared outside it
- 03
Reflection against transmission: where an opaque and a transparent object get their colour
- 04
Subtractive mixing with paints, worked through absorption bands
- 05
Additive mixing with lights, and why primaries cannot be made from each other
- 06
Rods and three cone types, with the peak the module states for each
- 07
The ambiguity argument: why one cone type can never report a wavelength
- 08
Monochromatism and dichromatism, and what each confirms about comparative coding
- 09
Three opponent channels, and the two lines of psychophysical evidence for them
- 10
Afterimages by desensitisation, and colours that cannot be experienced
- 11
Reflectance times illumination, and why habituation is not sufficient
- 12
Discounting the illuminant, and why one photograph divides its viewers
Reason from a missing cone type to a real consequence
- 1Name the loss in comparison terms. Colour is coded by ratios across cone types, so losing the medium-wavelength type removes the medium-against-long comparison.
- 1Predict the hard cases. Anything resting on that comparison: ripe against unripe fruit, red against green indicators, a red object against green foliage. These are hard because the colours look the same, not because they look wrong.
- 1Predict the easy cases. Discriminations resting on the short-wavelength cone against the others are unaffected, so blues and yellows remain distinguishable, and every brightness difference remains available.
- 1Explain the plate. It prints a figure in one hue against a background in the confusable hue, matched so brightness gives nothing away. Someone with the full set reads the figure from the colour difference; someone missing a type has no signal separating them.
- 1State the design principle. The plate works by removing every cue except the one being tested. Matching the brightness is not a refinement; it is what makes the plate diagnostic rather than merely difficult.
Key terms
- Additive mixing
- Combining lights, so that the mixture contains every component's wavelengths. It gives different answers from combining paints for the same pair of colour names, which is why the rule has to be chosen before predicting.
- Subtractive mixing
- Combining paints or filters, so that only the wavelengths every component reflects survive. Each component subtracts what it absorbs, and what remains is what is seen.
- Trichromatic theory
- The account in which colour is read from the relative activity of three cone types with different peak sensitivities. It is a claim about the receptors, and the comparison is essential because no single cone can separate wavelength from intensity.
- Dichromatism
- Lacking one of the three cone types, which removes one comparison rather than dimming vision. Two of the three forms are stated to affect about one per cent of males each.
- Colour constancy
- Perceiving a surface's colour as stable across a wide range of illuminations, achieved by estimating the colour of the light and subtracting it out. The subject states what the system achieves and not how it estimates the light.
- Opponent channel
- A pathway carrying two colours in opposition, so it can signal one or the other but not both at once. It explains afterimages and predicts which hue combinations cannot be experienced at all.
Colour Vision: Physics, Theories and Constancy FAQ
Are the trichromatic and opponent-process accounts rivals?
The objective asks you to contrast them, and the natural reading is that they describe different stages of one system: three receptor types, recombined into three opponent channels afterwards. That reading is consistent with everything the subject says and the subject never states it, so present it as an inference rather than as taught content.
Why can a single cone type not report a wavelength?
Because its output confounds wavelength with intensity. A given response level could come from a wavelength near its peak at low intensity or one further from its peak at high intensity, and one number cannot separate two variables. The ratio across cone types differs between those cases, which is why coding is comparative.
Which way round does the dress illusion work?
Derive it instead of recalling it. You subtract the illuminant you assume, so assuming a warm yellow light removes yellow and leaves blue, while assuming a cool blue light removes blue and leaves the warmer reading. Deriving it also covers the other constancy demonstrations without extra memorising.
Is the long-wavelength cone the red cone?
It is conventionally nicknamed that, and the subject gives its peak as a wavelength it labels yellow. Both statements can stand: the cone peaks in the yellow and still responds furthest into the red, which is what makes the long-against-medium comparison the red-green channel. Reproduce the number with the label the subject gives it.
How much of the opponent-process material is examinable?
All of it, even though it has no text page of its own in the subject and lives in a video. That is worth knowing because a text-only revision pass will miss the afterimage mechanism and the impossible-colours argument entirely, and those are the two lines of evidence the objective asks you to discuss.
Assessment move
Do the physics first and do it as two rules rather than as a set of colour facts, because a question giving you blue and yellow can have either answer depending on whether lights or paints are being combined. Learn the four peak wavelengths as numbers with their labels attached, and notice how close the medium and long cones sit, since that proximity is the reason the comparison between them carries red and green.
For the two theories, do not try to hold both accounts in full: hold what each is a claim about, receptors in one case and recombination in the other, and the evidence unique to each. Colour constancy is best revised as a single operation you can perform, subtracting the illuminant you assume, because that one move covers every demonstration in the module and cannot be remembered backwards.
Working through Colour Vision: Physics, Theories and Constancy in PSYC10006? Sia is AskSia’s AI Psychology tutor — ask any PSYC10006 Colour Vision: Physics, Theories and Constancy question and get a clear, step-by-step explanation grounded in how PSYC10006 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
