Light (illuminance) calibration
Illuminance meters must match the human eye response across the spectrum, so their calibration is as much about the lamp as about the meter. Distance and stray light are the usual traps.
Instruments and typical accuracy
| Instrument | Typical range | Typical accuracy | Where it is used |
|---|---|---|---|
| Lux meter (silicon with V(λ) filter) | 0.01 lx to 200 klx | ±3 to ±10 % of reading | Lighting surveys, workplace checks |
| Reference illuminance meter | 0.1 lx to 100 klx | About ±0.5 to ±2 % | Calibration labs, transfer standard |
| Luminance meter | 0.01 to several 100 000 cd/m² | ±2 to ±5 % | Displays, road lighting |
| Spectroradiometer | Wavelength 200 to 2500 nm | ±1 to ±5 %, by band | Colour, LED and display testing |
| Standard lamp on a photometric bench | Calibrated luminous intensity | About ±0.5 to ±1 % | Realises the cd and lx for meters |
Accuracy depends on the light source: a meter calibrated with incandescent light can read wrongly under LEDs. The accuracy figures are typical ranges to help you compare instruments. Always use the specification of the instrument in front of you.
Reference standards and traceability
- Photometric units. The candela is defined from a fixed luminous efficacy of 540 THz radiation and the human eye response V(λ).
- Standard lamps. Calibrated incandescent lamps at 2856 K (CIE standard illuminant A) give luminous intensity at a measured distance.
- Meter classification. ISO/CIE 19476 describes the characterisation of illuminance and luminance meters, including the spectral mismatch f1′ and the cosine response f2.
- Quality system. ISO/IEC 17025, scope listing illuminance with the source type and uncertainty.
How it is calibrated
Typical procedure
- Warm up the lamp for at least 30 minutes at constant current.
- Set the meter on the bench axis at a measured distance, with baffles to remove stray light.
- Measure the dark reading with the lamp blocked, and subtract it.
- Read at several distances to confirm the inverse square behaviour and to calibrate over the range.
- Check cosine response by rotating the meter, and linearity with filters or distance.
- Compute the calibration factor with a stated source type, and report with the uncertainty.
Where it goes wrong
- Stray light and reflections. Walls and fixtures add light that the inverse square law does not include.
- Wrong source. The spectral mismatch of the meter shows up under LED or fluorescent light.
- Distance errors. A 1 % distance error is 2 % in illuminance.
- Unstable lamp. Lamp current and warm-up change the output.
- Angle. Light arriving at an angle is under-read if the cosine response is poor.
What goes in the uncertainty budget
- Standard lamp or reference meter uncertainty.
- Distance and alignment.
- Stray light and dark signal.
- Spectral mismatch and cosine response.
- Temperature, linearity and resolution.
Combine the contributions as in the uncertainty budget calculator, then decide pass or fail with a guard band as explained on the calculators page.
Calculators
Everything runs in your browser. Nothing you type is sent anywhere.