Sensitivity and Resolution of the Eye

The previous two lessons treated the eye purely as an optical system — a set of refracting surfaces bringing light to a focus. But a focused image is useless without something to detect it. This lesson turns to the retina itself, treating it the way any detector is treated in physics: by asking how sensitive it is, and how finely it can resolve detail.

The retina as a photodetector: spectral response

Any device that detects light can be characterised by its spectral response — how its sensitivity varies with the wavelength (colour) of the incoming light. The retina is no exception. It is not equally sensitive to every wavelength across the visible spectrum: its overall response peaks somewhere near the middle of the visible range, corresponding to green light, and falls away on both sides — toward red at the long-wavelength end, and toward violet and blue at the short-wavelength end.

This means that, for two sources emitting the same amount of light power but at different wavelengths, the eye will generally perceive the one closer to the peak of this spectral response as brighter, even though the physical power reaching the eye is identical. The practical consequence is that the eye is a genuinely wavelength-dependent detector, just like a photodiode or photographic film with its own characteristic response curve — it cannot be treated as equally sensitive across the whole visible spectrum.

Two types of receptor: rods and cones

The retina's light sensitivity comes from two distinct types of photoreceptor cell, each suited to a different job.

• Rods. Far more numerous than cones across the retina as a whole. Rods are highly sensitive, able to respond even at very low light levels, which makes them responsible for vision in dim conditions. Their major limitation is that they provide no colour information at all — a scene viewed purely by rod vision is effectively monochrome.
• Cones. Considerably less sensitive than rods, requiring brighter light to respond usefully, but responsible for colour vision. There are three distinct types of cone, each with peak sensitivity to a different part of the visible spectrum (roughly corresponding to red, green, and blue regions); the combined, relative response of all three types across a scene is what the visual system interprets as colour.

This division explains a familiar everyday observation: in low light, colours appear washed out or disappear almost entirely, and vision becomes dominated by shape and brightness rather than colour — the cones are simply not receiving enough light to respond usefully, leaving vision reliant on the colour-blind rods.

Spatial resolution: receptor density and neural wiring

Spatial resolution is the eye's ability to perceive two closely spaced points as genuinely separate, rather than blurring them into one. Two distinct physical factors determine how good this resolution is in any given part of the retina.

• Receptor density. Resolution is best wherever photoreceptors are most tightly packed together. This packing is greatest at the fovea, a small central region of the retina consisting almost entirely of densely packed cones. Because the receptors here are so close together, two points imaged onto neighbouring receptors can still be told apart, giving the sharpest, most detailed vision available anywhere on the retina — which is exactly why looking directly at something (placing its image on the fovea) gives the clearest possible view of it.
• Neural wiring. Resolution also depends on how receptors are connected onward to the nerve fibres carrying signals toward the brain. In the fovea, each cone is connected to something close to its own individual nerve fibre, so the fine spatial detail captured by densely packed receptors is preserved rather than lost on the way to the brain. Elsewhere in the retina, by contrast, many rods commonly share, or pool, their connections into a single shared nerve fibre — this pooling increases overall sensitivity to faint light (since the combined signal from many rods is stronger than any one rod's signal alone) but sacrifices resolution, since the brain can no longer tell exactly which of the pooled receptors detected the light.

Mark as complete