THE VISUAL SYSTEM
1. What the visual system is
The visual system is the set of structures that capture light, turn it into neural signals, and interpret those signals as a visual world.
Seeing is not something the eye does alone. The eye collects information; the brain builds perception.
2. Anatomical basics of the eye
The eye works like a biological camera.
Light enters the eye through the cornea and then passes through the lens. Together, these bend (refract) light so that it is focused onto the back of the eye. Good focus is essential—blurred input limits how much detail the brain can recover later.
At the back of the eye sits the retina, which is actually neural tissue, not just a screen. This is where light is converted into electrical signals the brain can use.
The eye itself does not “see.” It samples light and encodes it. Perception happens later, in the brain.
3. The retina: more than a detector
The retina is layered and highly organised.
Photoreceptors (rods and cones) detect light.
Bipolar cells pass signals onward.
Retinal ganglion cells send the output of the retina to the brain via the optic nerve.
Signals mostly flow vertically through these layers, but there are also horizontal connections that allow the retina to compare nearby signals. This means the retina already does some processing before information ever reaches the brain.
There is a blind spot where the optic nerve leaves the eye. No photoreceptors exist there, so no light is detected. You don’t usually notice this because:
you have two eyes with overlapping views, and
the brain fills in missing information automatically.
4. Photoreceptors: rods and cones
Photoreceptors convert light into neural signals using light-sensitive pigments.
Rods
Extremely sensitive to light
Used in dim conditions (night vision)
No colour vision
Poor spatial detail
Most sensitive to greenish wavelengths, which is why green light seems brighter in the dark
Cones
Work best in bright light
Support colour vision and fine detail
Three types, sensitive to different wavelength ranges (roughly red, green, blue)
Colour comes from comparing activity across cone types, not from one cone alone
Densely packed in the centre of the retina
5. Visual processing starts in the retina
When light hits rods and cones, it triggers chemical changes that alter the cell’s electrical activity. These changes are passed through retinal circuits to ganglion cells.
Crucially, the retina does not send a simple pixel-by-pixel image to the brain. Instead, it:
enhances contrast
emphasises edges
reduces redundancy
This makes visual information more efficient to transmit, because the optic nerve has limited capacity.
6. Central vision and the fovea
At the very centre of the retina is the fovea.
Packed with cones
Almost no rods
Each cone connects to very few downstream neurons
This setup preserves fine detail, giving you high visual acuity at the point you are directly looking at.
This is why:
reading and recognising faces require direct fixation
your eyes constantly move to place important objects onto the fovea
In low light, the fovea performs poorly because it lacks rods. Looking slightly away from a dim object places it on rod-rich peripheral retina, often making it easier to see.
7. From eye to brain: post-retinal pathways
Signals leave the eye through the optic nerve.
At the optic chiasm, fibres from the nasal half of each retina cross to the opposite side, while fibres from the temporal half stay on the same side. This ensures:
the left visual field goes to the right hemisphere
the right visual field goes to the left hemisphere
Most fibres then travel to the lateral geniculate nucleus (LGN) in the thalamus.
The LGN:
has six layers
keeps information from the two eyes partly separate
preserves different signal types (contrast, motion, detail)
From the LGN, signals travel to primary visual cortex (V1) in the occipital lobe.
8. Primary visual cortex (V1): ocular dominance
In V1, inputs from the two eyes are still not fully mixed.
Neurons are organised into ocular dominance columns, where some respond more strongly to the left eye and others to the right. These columns form alternating stripes across cortex.
This organisation develops through competition between the eyes early in life. Balanced input supports normal binocular vision. Unequal input (for example, if one eye is blurred early on) can bias cortical organisation and impair vision in the weaker eye.
9. Retinotopic maps
V1 contains a retinotopic map:
neighbouring points in visual space map to neighbouring neurons in cortex
The map is distorted in a useful way. The fovea takes up a huge amount of cortical space, a phenomenon called cortical magnification. Peripheral vision takes up much less.
This explains why:
central vision is so detailed
damage to specific parts of visual cortex causes specific visual field losses
Retinotopic organisation continues beyond V1 into multiple visual areas.
10. Vision beyond V1
After V1, visual information spreads into many specialised cortical areas.
Two broad pathways are often described:
The ventral stream supports object identity: colour, shape, and recognition.
It answers: “What am I looking at?”The dorsal stream supports spatial processing, motion, and visually guided action.
It answers: “Where is it?” and “How do I interact with it?”
These streams are not isolated. They constantly exchange information so perception and action stay coordinated.
11. Take-home ideas
Vision is built through multiple processing stages, starting in the retina and extending across the brain.
The system is organised around retinotopic maps that are largely in place very early in life.
This early structure shapes how experience refines vision and sets limits on how plastic the system can be later on.
The strange but elegant result is that what feels like a seamless visual world is actually the product of many specialised circuits working together behind the scenes.