Article, www.researchgate.net:
Calculation of the Unified Glare Rating based on luminance maps for uniform and non-uniform light sources
Luminaires must be designed with concern for vision discomfort where the luminaire is unavoidably gazed-upon. There are standards, as in requiring UGR less than 19. We will find out what that means.
At 3/21/2016, I sought to find origin of a design error in luminaire Nicor DLS4, too bright to gaze upon by about times 1.5. This light at 3000°K, 671 lumens, 100° beam angle, needs further characterization. What happened? As I learn, I will question and adjust all I have offered in rating luminaires with a proposed Brightness Number, a rating of delivered illumination through side-by-side comparisons with the human eye and my best available digital camera intended to produce photos matching human sensation.
My definition of Brightness Number is that task illumination however measured, is B4 with a good old 100 watt incandescent light bulb. Further, I offer that a 3000°K LED of broad beam angle, about 120°, is B4 at light output of 450 lumens. I use the B4:450 lumens relationship to assign brightness numbers of competing practical LED luminaires, including offer of shame upon those tricked-out foolishly to work as point source light bulbs. It will be best, in my opinion, if we let our opportunity in LED lighting be delivered in directional downlighting; any point-source bulbs, as allowed decor unregulated, not promoted for energy conservation.
Have I missed anything, in the state-of-the-art box marking of Nicor DLS4 and Nicor DLS56, where DLS56 is good, and DLS4 is eye-jarring?
The illuminance numbers not seen for any other product, are interesting and no help. They are only a math exercise from measured lumens output.
What test device might "look" at the luminaire close-up, to observe brightness and pattern of brightness on the luminaire lens? I want to do this with a test device, although eyes and mathematical reason serve without ambiguity. A lux meter would work, held only a few inches from a luminaire lens in a non-standard test. Perhaps it is time I owned one, although no test is defined for it that would better my unambiguous professional-enough side-by-side visual comparisons.
Learn from this EPA resource of year 1995, about established measures of lighting for task illumination:
http://www.mts.net/~william5/library/epalight.htm
Quantity of Illumination
Light Output
The most common measure of light output (or luminous flux) is the lumen. Light sources are labeled with an output rating in lumens. For example, a T12 40-watt fluorescent lamp may have a rating of 3050 lumens. Similarly, a light fixture's output can be expressed in lumens. As lamps and fixtures age and become dirty, their lumen output decreases (i.e., lumen depreciation occurs). Most lamp ratings are based on initial lumens (i.e., when the lamp is new).
Light Level
Light intensity measured on a plane at a specific location is called illuminance. Illuminance is measured in footcandles, which are workplane lumens per square foot. You can measure illuminance using a light meter located on the work surface where tasks are performed. Using simple arithmetic and manufacturers' photometric data, you can predict illuminance for a defined space. (Lux is the metric unit for illuminance, measured in lumens per square meter. To convert footcandles to lux, multiply footcandles by 10.76.)
Brightness
Another measurement of light is luminance, sometimes called brightness. This measures light "leaving" a surface in a particular direction, and considers the illuminance on the surface and the reflectance of the surface.
The human eye does not see illuminance; it sees luminance. Therefore, the amount of light delivered into the space and the reflectance of the surfaces in the space affects your ability to see.
Nothing in this contradicts my observed-illumination mindset. Nothing addresses the luminance close to the lens as need-to-know, in itself.
Should I buy a lux meter? I think not. I doubt I could do better than with dimensional comparisons, as in this table:
Say there is a fixed distance from lens edge to light engine edge. The active area of the lens would shrink as fraction of lens area, by more than the square of the lens diameter. I think the light density at the active portion of the lens might be doubled, for the DLS4. I won’t destroy luminaires to flesh this out with measurements.
Where the available solution for my DLS4 problem has been substitution of 700 lumens Utilitech 0752125 LED disk lights from Lowe's. Please read on there , for now, where "perceived diameter" may be the best dimension for glare assessment, and a 2.5" diameter at B4 brightness of a 100 watt bulb may be the standard for some measure of luminance in gazing upon a luminaire. This post will go on to recognize that LED glare has been a problem even for the B4 4" Glimpse, with 1.5" board diameter highly visible through its lens. In our willingness to gaze-upon, LEDs have not measured up.