A Desktop Reading of 500 lx—Why Doesn’t That Mean the Eyes Receive 500 lx?
Only when we move from horizontal illuminance to the direction of human vision does the space truly enter the calculation.
Author | Lawrence Lin
Chairman, GLGA | Board Member, GLG | IWBI WELL Light Concept Advisor | Founder & CEO, LRS
During final acceptance testing, an engineer places the illuminance meter flat on the desktop and gets a reading of 512 lx. Everyone breathes a sigh of relief. But once seated, the user is facing a dark wall, a bright window, and reflections from the screen. A compliant desktop reading does not necessarily mean that the visual experience or the light exposure at the eyes is also satisfactory.
Horizontal illuminance is an important basis for evaluating the work surface; vertical illuminance at eye level and mel-EDI describe a different geometric direction. They may be related, but they cannot substitute for one another.
Where the Light Goes Depends on the Distribution and the Surfaces
A narrow-beam downlight can concentrate a large amount of light onto the desktop while sending relatively little light into the user’s forward field of view. Large luminous surfaces, bright walls, or side windows, by contrast, can increase illuminance in the direction of the eyes. The reflectance of walls, ceilings, and floors, as well as furniture height, partitions, and curtain positions, can all change the final result.
Therefore, the same 500 lx of desktop illuminance can correspond to completely different vertical illuminance, luminance distributions, and glare risks. If the design software contains only a horizontal calculation plane, these differences can be difficult to identify in advance.
How Should Eye-Level Measurement Points Be Defined?
In office environments, vertical measurement points are commonly established at seated eye height and calculated or measured along the primary viewing direction. For standing environments, classrooms, patient rooms, and transportation spaces, the measurement setup should be adjusted according to the actual posture. If the direction people face is not fixed, multiple representative directions can be selected, or both typical and worst-case conditions can be evaluated.
When measuring mel-EDI, the sensor must have an appropriate spectral response and be maintained in the vertical orientation. Taking a reading with a conventional illuminance meter and then estimating the result based on color temperature can introduce uncontrolled errors. On-site measurements should also document daylight conditions, curtain position, luminaire dimming, and measurement time.
One Continuous Line from Drawings to the Site
During the design stage, establish both horizontal work-plane calculation surfaces and vertical eye-level planes in the model.
During the mock-up stage, use a spectrometer to verify conditions at the predetermined eye positions.
During commissioning, check daytime, dusk, and nighttime scenarios.
During final acceptance, repeat the measurements using the same measurement-point checklist.
If each of these four stages uses different locations and conditions, even highly precise numbers cannot form a reliable chain of evidence.
What We Can Say Today / What We Cannot Yet Say
What we can say today: Horizontal desktop illuminance primarily serves the evaluation of visual tasks; vertical measurements at eye level are better suited to describing the light exposure in the direction of a person’s view. The two should be evaluated in parallel.
What we cannot say: A horizontal illuminance of 500 lx necessarily corresponds to a fixed mel-EDI value, or that a high vertical illuminance automatically means good visual comfort.
Three Things You Can Do Today
- Mark both the work plane and representative eye-level measurement points on the drawings.
- Record the direction, time, curtain position, daylight conditions, and dimming state during on-site measurements.
- Review glare, luminance distribution, and vertical light exposure together within the same lighting scenario.
Evidence label: Established consensus.
Measurement geometry is part of the result; it cannot be omitted.
References
- CIE S 026:2018：https://www.cie.co.at/publications/cie-system-metrology-optical-radiation-iprgc-influenced-responses-light-0
- Brown et al., 2022：https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001571
- ISO/CIE 8995-1:2025：https://www.iso.org/standard/76342.html
In the next article, we will temporarily step away from new metrics and return to the most fundamental principle: Integrated lighting must first and foremost be good lighting.
English 
Chinese (Taiwan) 
Chinese (China) 
German