Skip to main content

What is blue light?

By Spektrum Glasses Editorial Team · Published 2026-08-05 · Updated 2026-08-06 · Facts re-checked 2026-08-06 How this page is written and checked: our editorial method · how we verify claims

Short answer

Blue light is the visible band from roughly 400 to 500 nanometers, the shortest and most energetic wavelengths the human eye detects. Sunlight is by far the largest source; screens and indoor LEDs emit far less. Within this band, wavelengths near 460-480 nm carry the strongest signal to the body’s circadian timing system.
  • Blue light spans about 400-500 nm, the shortest visible wavelengths.
  • Daylight is the dominant exposure source; screens emit far less blue light.
  • The 460-480 nm sub-band drives circadian signaling most strongly.
  • No near-clear lens filters the whole band: ours measures 99.99% at 400 nm, 33.1% at 450 nm.
  • Published testing put the blue-light hazard reduction of filtering lenses at roughly 10-24%.

Where blue light sits in the spectrum

Visible light runs from about 380 nm (violet) to about 700 nm (deep red). Blue occupies roughly 400-500 nm at the short-wavelength end. Shorter wavelength means higher photon energy, which is why this part of the spectrum gets singled out: a 430 nm photon carries meaningfully more energy than a 600 nm one. Just below 400 nm is ultraviolet, which the human eye does not see. UVA runs to about 400 nm and UVB below that. This boundary matters when reading eyewear specifications, because a lens can filter UV completely while filtering very little of the visible blue band, and vice versa. The two are separate measurements. The industry term “HEV” (high-energy visible) usually refers to the short end of the blue band, around 400-450 nm, but it is not a tightly standardized definition. Different manufacturers and labs draw the boundary in different places. Any HEV percentage is therefore only meaningful alongside the exact wavelength range it was integrated over. The same is true of “blue light” itself: 400-500 nm is the common convention, not a legal definition.

Where blue light comes from

The overwhelming source is the sun. Daylight contains a broad, continuous spectrum with substantial output across the entire 400-500 nm range, and outdoor illuminance is orders of magnitude higher than typical indoor or screen illuminance. A person who spends time outdoors receives far more blue light in a few minutes of daylight than from an evening of screen use. White LEDs — in monitors, phones, and room lighting — produce white light by pumping a blue emitter through a yellow phosphor. That gives their spectrum a characteristic narrow peak in the blue region plus a broad hump across the rest of the visible range. So screens do emit a disproportionate share of their light as blue relative to an incandescent bulb, but their absolute output is small compared with daylight. That distinction is often blurred in marketing. The honest framing is that screens change the timing and spectral shape of light exposure — evening light that would historically have been dim and warm is now brighter and bluer — rather than delivering an unusually large total dose.

How the eye responds to the blue band

Two separate mechanisms matter here. The first is image-forming sight. Short-wavelength cones contribute to color perception, and because the eye’s lens and cornea focus short wavelengths slightly differently from long ones (chromatic aberration), blue light focuses marginally in front of the rest. This is a genuine optical effect, though its practical significance at normal viewing distances is small. The second is non-image-forming. A distinct class of retinal cells containing the pigment melanopsin signals ambient light level to the brain’s circadian pacemaker. Melanopsin’s peak sensitivity sits in the 460-480 nm region — roughly the middle of the blue band, not the high-energy end. This is why 460-480 nm is called out separately in lens specifications, and why filtering performance at 410 nm tells you nothing about performance at 470 nm. There is also a photochemical weighting function used in optical safety standards, informally the “blue-light hazard” function, which is weighted toward the shorter, more energetic end of the band. It is a laboratory metric for comparing light sources, not a measure of any individual outcome.

What a lens can and cannot filter across the band

Filtering is not one number. It is a curve, and it falls off steeply as wavelength increases. Our clear PROSPEK lens, measured by COLTS Laboratories (report O-SPG111015, A2LA-accredited to ISO/IEC 17025, cert 1612.01, spectral transmittance per ANSI Z80.3), measures: Visible (photopic) transmission is 91.6%, meaning the lens is near-clear with no heavy tint. UVA and UVB are both filtered above 99.99%. The pattern is the point: a lens that looks clear can filter almost everything at 400 nm and only a third at 450 nm. Our ZENOX clear lens averages about 52% across the blue band with 100% UV filtering. No near-clear lens filters the whole band, because absorbing 460-480 nm strongly requires a visible tint. Tinted lenses do reach across the band, at the cost of color neutrality. The same COLTS report gives our amber evening lens 97.9% across 400-500 nm, 99.9% of HEV, and 98.3% at the 460-480 nm melatonin band. Orange (99.96% across 380-500 nm) and red (99.83%) lenses were measured in 2026 by our lens manufacturer’s optical laboratory. Neither the orange nor the red lens is suitable for driving. For night driving, the Illumin line uses a lens designed to keep road color signals legible.

What the evidence supports and what it does not

The physics above is settled. The clinical picture is not. Leung, Li and Kee (PLOS ONE, 2017) measured commercially available blue-light-filtering lenses and found they reduced the calculated blue-light hazard by roughly 10-24%. That is a laboratory-computed optical quantity, not a health endpoint. It is also a useful reality check on the marketing: a lens that reduces a weighted hazard metric by a fifth is doing something measurable and modest. Where product pages usually overreach is in promising a symptom outcome the measurements cannot support. The evidence for symptom benefit from blue-light-filtering eyewear is limited and contested, and we will not assert one. What our lenses do is documented and independently measured: they attenuate specific wavelength ranges by specific amounts under a defined standard. If your concern is evening light exposure, the interventions with the strongest physical logic are reducing overall light level, moving screens further away, and getting bright light earlier in the day. A lens changes the spectrum reaching the eye; it does not change how long you look at the screen.

How to read a blue-light percentage

A percentage with no wavelength band attached is not information. “Blocks 99% of blue light” is compatible with a lens that filters 99% at 400 nm and almost nothing at 470 nm — our own clear lens is close to that shape, and we publish the curve rather than the headline. When comparing lenses, ask four things: over what wavelength range was the figure integrated; is it an average across the band or a peak value at one wavelength; what is the photopic (visible) transmission, which tells you how dark the lens actually is; and who measured it, under which standard. ANSI Z80.3 defines transmittance measurement for non-prescription eyewear, and ISO/IEC 17025 accreditation (granted by a body such as A2LA) indicates the lab’s competence is externally audited. Our full spectral data is published at https://kb.spektrumglasses.com/lab-results. All PROSPEK eyewear is non-prescription, with many styles offering reading magnification from 0 to +3.0. Products carry a 365-day warranty; Amazon orders follow Amazon’s 30-day return window and store orders 90 days.

Where these numbers come from

Every measured figure quoted here is transcribed from a third-party laboratory report, published in full with the wavelength band and the report number: lab results. Our rule for what may appear on this page at all is on how we choose what to publish.