The case for luminous flux.
Luminous flux is visually evaluated radiant flux (the total electromagnetic energy that a light source emits across all wavelengths). Luminous flux defines the total light output in all directions. The author mentions in his article that "if you take that same grazer curve and simply shift the peak from 5 to 45 degrees, your lumen values jump from 300 to over 1,700 lumens." This phrase is at least misleading, if not incorrect. Simply shifting the curve only changes the angle of the maximum luminous intensity. The total luminous flux is the same in both cases.
The author also writes that "the amount of lumens is relative to how the intensity is distributed, and, as such, it is an arbitrary, flawed metric." The total luminous flux does not depend on how the intensity is distributed. Luminous flux is the total effect and it is not a directional measure, while luminous intensity (in cd), on the other hand, is the density of luminous flux in space in that direction.
Unfortunately, frequently, one metric cannot tell us "the whole story." For lay people who rarely deal with complex light distributions but rather with the point sources, luminous flux gives an idea of the overall amount of light emitted. For any given application, lighting researchers have to provide guidance on the correct usage of photometric quantities and advocate the appropriate metrics to achieve high quality lighting along with meeting other criteria.
Yulia Tyukhova, Ph.D.
Acuity Brands Lighting
We agree with Don Peifer's general point that there is more to lighting than just the amount of light, but we don't agree with his lumen analogy, or his criticism of efficacy.
First, lumens are analogous to watts, not amps. Amps measure the flow of electrons, regardless of the amount of work they can do. Measuring the flow of photons would be useless for lighting engineers, as the ability of photons to stimulate the visual system depends on their energy or wavelength. Lumens are effective visual watts times the normalization factor of 683 lumens per watt at 555 nm.
Peifer's fixture comparisons in the task efficacy section are not legitimate. Illuminance is candlepower divided by distance squared. If the candlepower is unchanged, the only way total lumen output can go up is if the illuminated solid angle becomes larger. Simply changing the beam direction does not change the inherent lumen output of the source, and does not increase the solid angle, and therefore cannot improve the efficacy of a fixture from 17 lumens per watt to 100 lumens per watt. The only way the comparison can be justified is if the 45-deg beam has a wider azimuthal extent than the 5-deg beam, which is the opposite of what will happen if the fixtures are illuminating the same height and width of the wall.
If the LED source efficacy is 100 lumens per watt, but the LED fixture efficacy is 17 lumens per watt, then the optical efficiency of the fixture is only 17 percent. Either the application is extremely, and possibly unreasonably, difficult, or the fixture really is grossly inefficient. Is it better than the fluorescent fixture? If the task application is 300 lumens in a very narrow horizontal band of light on the wall, then the extra light put out by the fluorescent fixture adds nothing, and the application efficacy of the 54-W fluorescent fixture in this case is at best only 5.6 lumens per watt (300/54). Clearly, designers, and regulators, need to understand the issues of fixture and application efficacy, and not just source efficacy, but this is not a justifiable criticism of the lumen. Candlepower can be an important design constraint, but this does not relieve the designer of the obligation to produce an efficacious fixture within the design constraint.
We agree with Peifer's dissatisfaction with unshielded pixels, as they are very glaring. Quality does matter.
Robert Clear, Fellow IES
Don Peifer replies:
In the interest of educating the readership regarding the confusion around the lumen, I refer to Joseph B. Murdoch's Illuminating Engineering to clarify two points. First, regarding the electrical analogy, Murdoch correctly points out that "flux (lumen) is compared to electric current, in amperes, the time rate of flow of electric charge in a wire." If what Clear and Berman posit is that lumens are analogous to watts, what are we to make of the most popular metric, lumens per watt?
Secondly, I refer readers to the concept of zonal lumens in measuring luminous flux. The total lumens within a zone are the product of the average intensity in that zone multiplied by the zonal constant of that zone. To get the total lumens of a light source, the contribution from all zones are summed.
In the example I used in my column, the zonal constant for the 40- to 50-deg zone (0.774) is much higher than the zonal constant for the 0- to 10-deg zone (0.095). Therefore, it is easy for the readers to see that a tight, identical curve placed in different zones can yield wildly different lumen values.
Zone ([degrees]) Zonal Constant 0-10 .095 10-20 .283 20-30 .463 30-40 .628 40-50 .774 50-60 .897 60-70 .993 70-80 1.058 80-90 1.091
LD+A reserves the right to edit letters for length and/or clarity.