Three quantities, not one
"How bright is it" gets answered by three genuinely different physical quantities, each with its own unit family, and none of them converts into another without extra information about geometry or surface material. Conflating them — asking how many "lux" a light bulb produces, say — is a category error, not a units error, because a light bulb's output is properly measured in a different unit altogether.
Luminous flux: what a source emits
The lumen (lm) measures total light output from a source, in every direction, independent of how far away anything is or what it lands on. A light bulb's box states lumens because that is the one figure describing the bulb alone, not the room it's placed in.
Illuminance: what lands on a surface
The lux (lx) is one lumen per square metre: how much of that emitted light actually falls on a given area, which depends on the source's total output, the distance from it (falling off with the square of distance), and the angle of incidence. The imperial equivalent, the foot-candle (fc), is one lumen per square foot. Because a square foot is smaller than a square metre, the same light spread over it registers a bigger number: exactly 1 fc = 10.76391 lx, derived from the exact relationship 1 ft² = 0.09290304 m² (since 1 ft = 0.3048 m exactly), so 1 fc = 1 lm/ft² = 1/0.09290304 lm/m² = 10.76391042… lx. Illuminance is what a light meter measures and what building codes and horticultural lighting guides specify.
Luminance: what reaches the eye from a surface
The nit — formally one candela per square metre (cd/m²), "nit" is the informal but near-universal name for it — measures light leaving or reflected from a surface toward a viewer, which is what actually determines how bright that surface looks. Two surfaces sitting in identical illuminance can have very different luminance: a sheet of matte black card and a sheet of white paper under the same desk lamp receive the same lux, but the white paper reflects far more of it back toward your eye and looks far brighter. Nits are the unit used for display brightness because a screen is not reflecting ambient light at all — it is emitting its own, so luminance, not illuminance, is the relevant quantity.
Why you cannot convert freely between them
Going from lumens to lux needs the area the light is spread over, which is a geometry question the unit itself cannot answer. Going from lux (light arriving at a surface) to nits (light leaving that surface) needs the surface's reflectance, since a surface can only send back some fraction of what it receives. For an ideal matte ("Lambertian") reflecting surface, the relationship is:
L = E × ρ ÷ π
where L is luminance in cd/m², E is illuminance in lux, and ρ is the surface's reflectance (0 to 1). White printer paper reflects roughly 80% of the light hitting it (ρ ≈ 0.8). Under typical office lighting of 500 lx, that paper's luminance works out to 500 × 0.8 ÷ π ≈ 127 cd/m² — noticeably dimmer than a laptop screen next to it, which self-emits at 250–400 nits even though "the room" is only lit to 500 lux. Illuminance describes the room; luminance describes what your eye actually receives from a specific surface in it, and the two routinely diverge by a factor of several.
Common illuminance levels, for scale
| Condition | Illuminance | In foot-candles |
|---|---|---|
| Full moon, clear night | 0.1 lx | ≈0.0093 fc |
| Living room lighting | 100 lx | ≈9.29 fc |
| Office lighting | 500 lx | ≈46.5 fc |
| Overcast daylight | 1,000 lx | ≈92.9 fc |
| Direct sunlight | 100,000 lx | ≈9,290 fc |
A concrete case of a wrong figure spreading
Moonlight illuminance is a genuinely useful example of how a plausible-sounding number can outrun the measurement behind it. A round figure of "1 lux" for full moonlight circulates widely in lighting-design references and photography exposure guides; careful photometric work puts the real figure roughly an order of magnitude lower. Ecologist Travis Longcore's 2017 essay "How bright the moon" traced one specific, documented instance of this kind of error through the ecological-lighting literature: a 1999 paper had mislabeled a measured illuminance curve for a 75%-illuminated gibbous moon — not a full moon, and not a quarter moon either — as "quarter moon" lighting, and at least two later papers repeated the mislabeled figure without returning to the original photometric data. Longcore's own working estimate for typical ground-level full-moon illuminance, after accounting for the moon's altitude above the horizon and atmospheric extinction rather than an idealised overhead value, lands close to 0.1 lux — the figure this site uses above — well below the "1 lux" shorthand that keeps circulating in secondary sources that never cite a primary measurement.
The mechanism is the same regardless of which specific mislabeling started it: once a number is repeated in a review article or a table without the original data attached, subsequent authors cite the review rather than re-measuring anything, and a figure that was wrong, or right only under narrow conditions nobody restated, survives indefinitely. It is a photometry-specific instance of exactly the caution this site tries to apply to its own figures: a plausible round number is not the same thing as a verified one.
Picking the right unit
Ask what you're actually measuring before reaching for a unit. Total output of a source: lumens. Light falling on a surface: lux or foot-candles. Brightness of a surface or a self-luminous display as it appears to an eye: nits. None of the three substitutes for either of the others, because each one is answering a different physical question.