Opponent-processing

The trichromatic and centre-surround processing described above creates a time-coded, high fidelity, redundancy-free stream of information ready to be passed forward to ganglion cells where opponent-processing takes place and colour information is even more rigorously and economically encoded.

Opponent-processing is believed to represent a shift in emphasis from the physiologically orientated mechanisms of trichromatic processing towards data-types more adapted to the broader-ranging demands of fully conscious perceptual discrimination.

But whilst opponent-processing is another extraordinarily complex undertaking, even together, trichromatic and opponent-processing fall far short of supporting the plethora of deductions, inferences and anticipatory functions that underpin human visual sensibilities.

So before attempting a brief description of opponent-processing, let’s be clear about what lies ahead. To produce the everyday experience of the visual world that enables human beings to undertake yet more complex sense-making tasks, the stream of colour information considered so far needs substantial embellishment. The forms of colour information assembled within the retina will need to:

Feed into and support the entire range of distinct attributes of visual perception and be available to be resolved within the composition of our immediate present visual reality as and when needed.
Correspond with our extraordinary discriminatory capacities that interrogate a palette that can include millions of potentially observable variations in colour.
Provide colour information about the outside world that is highly detailed, specifically relevant, spatially and temporally sequenced and capable of being cross-referenced with higher mental functions.
Be subjectively orientated in a way that provides the impeccably accurate detail needed to deal with our more insane intentions. Think for example of leaping from rock to rock across a raging waterfall or diving off a mountain-top wearing only a wingsuit and negotiating the topography of craggy landscapes speeding past at 260km per hour.

Within the context of such extremely broad demands, much more is needed than the efficient transmission of the cleanly encode colour information that trichromatic phototransduction provides. The next stage involves the addition of contextualising detail and off more spacial and temporal information relevant to everyday circumstances. And as well as all this, the output must be capable of being handled within the data-carrying capacity of the optic nerves embedded in the back of each eye.

Opponent-processing of colour involves a sort of either/or interrogation of the infinitesimally small electrical impulses received from the three-channel output of the trichromatic investigation undertaken by bipolar calls and involves a biological equivalent of algorithms. Put simply, each signal received by a ganglion cell must pass three simultaneous tests that go something like this:

Are you carrying information about the presence of more or less light than at other locations in your immediate area?
Are you carrying information about the presence of a kind of red-green colour-contrast than at other locations in your immediate area?
Are you carrying information about the presence of a kind of blue-yellow colour-contrast than at other locations in your immediate area?

Now, remember that the three cone types respond to a range of wavelengths within different bands of the visual spectrum with peak sensitivity centred on a very narrow range of wavelengths. With this in mind, it is reasonably easy to visualise that the incoming trichromatic traffic can be represented as a plot on a graph showing the intensity of response at every wavelength being surveyed by each cone type. If the original stimulus was produced by sunlight, containing wavelengths from right across the spectrum, then the graph for each cone type will show a curve across its range of sensitivity with a spike at peak sensitivities of L, M and S cones.

It is also clear that the simultaneous interrogation by ganglion cells of all three curves, and by cross-referencing each impression at multiple scales across the entire encoded retinal image, causes a quantity and quality of useful data to emerge that puts even the most sophisticated camera to shame. How? Because as this process proceeds in parallel with the broader discriminatory processes at work, every bit of colour information is already being invested with rudimentary forms of meaning!

Once the required level of organisation has been achieved, the encoded information exits the retina and is channelled along its respective optic nerve.

The two optic nerves contain around a million nerve fibres each and independently transmit composited information about every location within the visual field of their respective eye. At subsequent processing centres, all these separate packets of information will somehow be assembled into an image, with each signal in its proper spatial and temporal location, so that we see a stereoscopic picture of the world, updated many times a second. More than this, our impressions will seamlessly include information about the sound, feel, smell or tastes that relate to each matter of interest.

Imagine the amount of processing required to build coherent and fault-free images of our surroundings at this level of detail, all resolved into a continuous movie-like stream packed with meaning precisely relevant to our immediate needs! Imagine what it means to maintain this level of intelligence-saturated acuity, every day through the entire course of our lives!

Don’t forget, that as all this is going on, the images we will see are constantly being stabilized as we move our heads from side to side. Again, as our eyes scan left and right or up and down in their sockets, a single impression is sustained. And at any moment we may blink, focus on something else, or recompose everything we see before us as priorities change.

Let’s map some of these ideas out as bullet points.

- The description of human vision as a trichromatic colour system or as an opponent-process colour system refers to two complementary theories that help to explain aspects of colour perception that occur in the retina.
- Our understanding of trichromatic and opponent processes are called theories because our human understanding of exactly what happens at these infinitesimally small cellular scales can only be fully observed empirically and scientific enquiry is the circuitry is ongoing.
- There are four different photosensitive receptors commonly associated with the human retina, three types of cone cells and one type of rod cell. Cones are responsible for the perception of colour whilst rods function mainly in dim light.
- Cones work at their best at high levels of illumination.
- When lighting is low, rods come into their own. Because the rods are more sensitive to short wavelengths than to long ones, rods respond to the blue end of the visible spectrum. At night the scenes we see are registered as being relatively colourless.
- Too much light causes rods to saturate. Saturation in this context means that once the amount of light reaches a certain level the rods max-out and fail to register light levels rising further. This is where the cones take over.
- A distinction can be made between types of vision that result from adaption to light levels by the photosensitive rod and cone receptors:
  - Photopic vision: This is where cone cells are most active. It refers to adaptation to higher levels of overall brightness within the field of view that an observer has adjusted to.
  - Scotopic vision: This is where the rod cells are most active. It refers to adaptation to low levels of overall brightness within the field of view that an observer has adjusted to. Under these conditions, the ability to discern colours is sharply reduced. In addition, since there are no rods in the fovea and the cones there are not receiving enough light to be stimulated, the ability to discern fine detail is also compromised.
  - Mesopic vision: This term refers to a combination of photopic vision and scotopic vision during which both rods and cones are active. There is no sharp transition at the top and bottom of the range of mesopic vision.
- Adaption refers to the fact that it can take as long as 20-30 minutes for our eyes to fully adjust to new light levels. This is reflected in the way we squint when going out into bright sunshine and by the fact that we may initially see little or nothing when indoor lights are turned out at night.
- Each of the millions of cone cells in the human eye can at any moment respond separately as photons strike them at light-speed. This is why a vast array of different colours can be perceived simultaneously at a single glance.
- Colour vision depends on bipolar and ganglion cells receiving input from at least two types of cone cells to produce trichromatic vision.
- People who have only two types of cone cells have dichromatic vision and are able to distinguish thousands rather than millions of different colours.