Photometry, contrast, colour and adaptation

Why the measurement conditions matter

An EBOD optics problem can ask what changes when a lamp moves; a clinical case can ask why a patient with good chart acuity struggles at dusk. Both require separating the physical stimulus from the visual system's response. Radiometry measures radiant power without weighting it for vision. Photometry weights radiation by a specified human spectral-sensitivity function. Equal radiant power at different wavelengths does not necessarily produce equal photometric output or apparent brightness.

Photometric quantities

These distinctions follow the BIPM photometry definitions. A lamp's lumen rating alone does not tell you the illuminance on a near task; distance, beam distribution and obstruction matter. Illuminance on a chart also does not equal its luminance: reflectance and viewing geometry contribute.

For an ideal point source of intensity $I$, a small surface at distance $d$ receives $E = I\cos\theta/d^2$, where $\theta$ is the angle between the incident ray and surface normal. With perpendicular incidence, a 100 cd source gives 25 lux at 2 m and 100 lux at 1 m. Halving distance gives four times the illuminance under these assumptions. Do not extend that calculation unchanged to a nearby large panel, focused slit-lamp beam or a room with multiple reflected sources. The NIST photometric calibration reference distinguishes the quantities used in real measurements.

Filters and transmission

A neutral-density filter reduces transmission relatively evenly across its specified spectral range; a coloured/selective filter changes spectral composition. Optical density is $OD=-\log_{10}T$, with transmission $T$ expressed as a fraction. Thus OD 1 transmits 10%, and OD 2 transmits 1%. In the ideal case stacked filter densities add because transmissions multiply. State the wavelength/range: a filter that appears neutral to the eye may not be neutral outside visible light.

Polarising filters preferentially transmit one plane of polarisation and can reduce suitable reflected glare. Visible darkness does not prove ultraviolet protection. Choose protective lenses by their specified transmission and task requirements; a very dark tint can reduce useful low-light vision. In a colour assessment, use the prescribed illumination and avoid tinted lenses that change the stimulus being tested.

Acuity and contrast

Minimum angle of resolution (MAR) describes the angular size of critical optotype detail at threshold. For 6/6 Snellen vision, MAR is 1 arcminute and logMAR is 0; at 6/12, MAR is 2 and logMAR is approximately 0.30. Whole-letter size is larger than the critical detail. Record test distance, correction, chart type and whether an age-appropriate matching or detection task was used.

Contrast sensitivity is the reciprocal of contrast threshold. Weber contrast suits a target against a comparatively uniform background; Michelson contrast suits alternating bright/dark patterns:

$$C_W=\frac{L_{target}-L_{background}}{L_{background}},\qquad C_M=\frac{L_{max}-L_{min}}{L_{max}+L_{min}}.$$

Specify the convention and, where relevant, use contrast magnitude rather than interpreting a dark target's negative sign as negative sensitivity. A patient can resolve small high-contrast letters yet fail larger low-contrast targets. Cataract, retinal disease and optic neuropathy can affect contrast differently; the primary clinical comparison demonstrates that acuity and contrast provide different information.

Colour signals and testing

Three cone classes are relatively sensitive to short, medium and long wavelengths (S, M and L cones); their sensitivity curves overlap rather than each detecting one pure colour. Downstream comparison of cone signals supports red–green and blue–yellow opponent processing. Protan defects concern L-cone function, deutan M-cone function and tritan S-cone pathways. Inherited red–green defects are common, while a newly acquired asymmetrical change requires assessment of ocular or neurological disease and medication exposure. The NEI classification distinguishes these patterns.

Ishihara pseudoisochromatic plates primarily screen congenital red–green deficiency; they do not exclude every acquired blue–yellow deficit. Use the specified lighting and plate protocol, checking that the patient can understand the task. Arrangement tests such as Farnsworth D-15 assess colour ordering and characteristic confusion axes; more extensive hue testing explores subtler discrimination. An anomaloscope quantitatively examines colour matching under its defined protocol. Test each eye when investigating an acquired defect, and integrate acuity, red desaturation, pupils, fields and retinal/optic-nerve findings. A failed plate is not itself a lesion-localisation diagnosis. The original plate instructions explain the screening scope.

Light and dark adaptation

Photopic vision is cone dominated; scotopic vision is rod dominated; mesopic vision involves both. After a substantial bleach, cone sensitivity initially recovers more rapidly, while rod sensitivity improves more slowly and eventually becomes greater under an appropriate peripheral test. The crossover is the rod–cone break. Recovery reflects photopigment regeneration and neural adaptation; timing depends on bleaching, wavelength, retinal location and protocol. A foveal or long-wavelength stimulus need not produce the same biphasic curve. Primary human rod-adaptation measurements demonstrate the rod phase after the crossover.

The Purkinje shift is a shift toward shorter-wavelength relative sensitivity in dim, rod-dominated conditions; it is distinct from Purkinje optical reflections. A red object may appear relatively darker at dusk. Nyctalopia prompts history about inherited rod disease, nutrition, medicines and media opacity, followed by appropriate retinal examination and testing. Do not infer a particular genotype from delayed adaptation alone.

In a patient reporting dusk difficulty despite 6/6 vision, repeat functional assessment with proper correction and controlled lighting, assess contrast and colour if indicated, and consider retinal testing. Increasing illumination may help a task, but glare, adaptation and underlying pathology determine whether that intervention actually improves function.