World Sleep Industry ENCYCLOPEDIA

Photobiological Safety Testing

6.3.2

Definition, protected tissues, and assessment boundary

Photobiological safety testing evaluates the potential for optical radiation emitted by an artificial source or source-containing product to cause photochemical or thermal injury to the eye or skin under specified viewing, distance, and time conditions. In WSIC it is classified under 6.3.2 Third-Party Testing and Evaluation Services. It can apply to night lights, circadian lighting, reading lamps, display or indicator sources, luminous wearables, ultraviolet products, and other optical devices in sleep settings, but the applicable method must follow source type and product purpose.

Broadband incoherent sources, lasers, products whose primary purpose is ultraviolet emission, image projectors, and therapeutic, diagnostic, cosmetic, or monitoring sources may fall under different standards or special requirements. The base scope of IEC 62471 includes many electrically powered incoherent broadband sources and LEDs but generally excludes lasers; laser safety follows the appropriate laser-product framework. A shared ability to emit light does not justify one classification method, and a component-source result cannot simply be applied to every end product.

Photobiological safety addresses tissue-injury risk. It is not equivalent to visual comfort, glare, flicker, lighting quality, color quality, circadian effect, or sleep benefit. A product with low photobiological risk under defined conditions can still cause discomfort, affect sleep onset, trigger symptoms in a susceptible person, or fail lighting-performance requirements. Conversely, a high correlated color temperature, blue content, or nighttime use alone does not establish retinal injury.

Hazard types, spectral weighting, and exposure time

The assessment identifies relevant hazards from spectral coverage and accessible radiation, such as photochemical ultraviolet effects on skin and the ocular surface, near-ultraviolet effects on the eye, blue-light retinal photochemical risk, retinal thermal risk, infrared thermal effects on the cornea or lens, and thermal risk to skin. Each hazard uses different action spectra, measurement quantities, and exposure rules. One visible luminance, illuminance, or total optical-power value cannot replace the full evaluation.

Spectral weighting incorporates the relative effectiveness of different wavelengths for a biological effect, so spectral resolution, response correction, stray-light control, wavelength and amplitude calibration, and integration ranges must be preserved. A broad peak, narrow peak, ultraviolet leak, or infrared tail can be underestimated by an ordinary illuminance meter. The laboratory should justify the use of radiance, irradiance, or another quantity and report measurement uncertainty and its decision rule near a classification boundary.

Risk depends on more than instantaneous intensity. Exposure duration, repetition, apparent source size, viewing direction, distance, aperture, and field of view also matter. Pulsed, dimmed, or scanning sources may require applicable treatment of peak values, pulse trains, and time averages. Long close use in a sleep setting, child-accessible placement, direct bedside viewing, and device faults should enter reasonably foreseeable-use analysis, but an arbitrary assumption of infinite exposure should not replace the classification conditions defined by the applicable standard.

Product classification, intended use, and measurement plan

Before testing, the product should be identified as general lighting, signaling, display, disinfection, projection, medical, cosmetic, research, or another use, with its target markets, users, installation, minimum accessible distance, normal viewing behavior, maintenance access, and fault access recorded. Standard editions, product standards, legislation, and risk-management requirements must be considered together. A marketing change in intended purpose—for example, from ambient lighting to direct exposure of the body—can change applicable standards and safety controls.

The plan should list every source, channel, color, brightness, dimming and pulse mode, optical accessory, diffuser, lens, reflector, filter, user mode, and power state and select combinations likely to produce the greatest relevant hazard. Automatic brightness control, distance sensors, timers, and interlocks should enter the assessment as risk controls only when they are reliable in production, not readily bypassed, and appropriately managed under fault conditions.

Measurement geometry must match the hazard and standard conditions, including distance, angle, aperture, field of view, focal plane, source size, scan path, and stabilization time. A laboratory should not select only a convenient point at beam center or test a diffuser while ignoring a bare source accessible during maintenance. If a standard permits transfer assessment based on lamp or module data, the drive current, thermal state, optics, and final installation must be shown not to increase accessible emission.

Samples, instruments, laboratory capability, and quality control

Samples should represent final production configuration, with model, source lot, driver, firmware, optics, enclosure, rated input, manufacturing location, and aging condition recorded. A low-output demonstration unit, engineering current limit, temporary filter, or diffuser different from the marketed product cannot support a formal conclusion. For products with several sizes, channels, or replaceable sources, risk analysis should select worst-case combinations and document the rationale for model coverage.

Spectroradiometers, imaging or nonimaging systems, detectors, integrating devices, and distance and angle instruments must cover the target wavelength band and dynamic range with appropriate calibration and intermediate checks. The laboratory should control dark signal, background, saturation, nonlinearity, stray light, cosine response, bandwidth, spatial response, optical alignment, and thermal drift. Good visible-light calibration does not automatically establish reliable ultraviolet or infrared measurement.

Quality control can include reference sources, system checks, repeats, cross-range comparison, dark spectra, wavelength checks, and review of original spectral plots. Data processing should retain raw spectra, correction files, weighting-function versions, integration formulas, software versions, and manual edits. Accreditation scope should be checked for the exact standard, spectral range, capability, and location. A generic optical-testing accreditation certificate does not show that every photobiological safety activity is within scope.

Risk groups, labeling, and interpretation

Some applicable standards use risk groups or corresponding classifications to express hazard and time to an exposure limit under specified conditions. A classification is a technical conclusion about a particular source or product, configuration, and measurement condition. It is neither a medical diagnosis for an individual nor a simple binary label of safe or dangerous. Different editions or product-specific parts can use different names, conditions, and product requirements, all of which must be identified accurately in the report.

Near a classification boundary, the report should explain measurement uncertainty, rounding, conservative assumptions, and the decision rule. A lower risk group does not authorize unlimited direct viewing at every distance and does not cover photosensitizing medication, ocular disease, aphakia, or every other susceptibility. A higher classification also does not make a product unusable in every controlled application; it can instead trigger engineering controls, distance, time limits, labels, instructions, or professional management.

Labels and instructions should correspond to actual accessible emission, intended users, and controls and clearly describe prohibited viewing, installation distance, de-energized maintenance, protective equipment, or professional-use conditions. A warning cannot replace a feasible engineering control. Vague claims such as “eye protecting,” “no blue light,” or “zero hazard” cannot override measurement limitations. A report number, risk group, or laboratory name must not be detached from sample, standard edition, and annexes to create broad health endorsement.

Blue light, circadian effects, and special-source boundaries

In photobiological safety, “blue-light hazard” refers to wavelength-weighted photochemical injury risk to the retina; it should not be used as an umbrella term for all blue light, screen fatigue, or circadian influence. For ordinary white light, risk assessment requires actual spectral radiance, source size, distance, and exposure time, not correlated color temperature or a blue percentage alone. Fear based on color appearance and claims that all biological influence has been removed both lack adequate boundaries.

Circadian and other nonvisual effects involve spectrum, intensity, timing, duration, previous light history, viewing direction, and individual variation. Their assessment objective differs from exposure limits intended to prevent eye or skin injury. Passing photobiological safety testing cannot establish that a product is circadian friendly, does not suppress melatonin, or improves sleep; evidence for a circadian effect likewise cannot replace eye and skin safety assessment. Flicker, glare, and visual comfort also require separate appropriate metrics.

Ultraviolet disinfection, therapeutic or cosmetic sources, high-output infrared, projection, and laser products require particular care. Products primarily emitting ultraviolet can have specialized risk-group, labeling, and protection requirements. Medical or cosmetic equipment intended to create a biological effect in people also carries basic-safety, essential-performance, and clinical responsibilities. Lasers are addressed through laser classes and accessible emission limits. A general luminaire report cannot replace these specialized evaluations.

Production changes, procurement acceptance, and continuing risk management

When procuring testing, the laboratory’s standard edition, spectral capability, measurement geometry, dynamic range, accreditation scope, sample coverage, uncertainty, and report content matter more than price or willingness to issue a “low risk” conclusion. The contract should define sample selection, worst-case modes, failure notification, access to raw spectra, subcontracting, report revision, sample retention, and dispute review. Pre-compliance characterization and formal classification should remain distinct.

Changes to an LED, phosphor, package, drive current, dimming frequency, lens, diffuser, reflector, filter, enclosure opening, firmware, supplier, thermal management, or installation distance can change the result. Manufacturers need production-consistency checks, monitoring of critical optical parameters, and change reassessment. Filter aging, broken covers, failed interlocks, and maintenance disassembly also belong in risk analysis. A component classification cannot permanently cover a modified end product.

Photobiological safety testing is one part of optical-safety evidence and cannot replace electrical, thermal, mechanical, EMC, material, usability, clinical, or real-use monitoring. Eye pain, persistent visual disturbance, skin injury, or photosensitivity requires an appropriate local medical pathway rather than a lookup of the product risk group alone. Related entries include “circadian lamp spectral testing,” “sleep-space light-environment assessment,” “photobiological safety evaluation,” “luminaire safety testing,” “laser-product safety,” and “product function verification.” Assessment should be selected by source, use, exposure, and responsibility boundary.