Evaluating PAR38 Lamps Subjectively

If we can't encapsulate our experience with light objectively, how about mixing in some subjective elements? That's what Pacific Northwest National Laboratory did in its latest CALiPER report.

We have discussed extensively the ways in which existing lighting metrics don't capture the quality of light the way people perceive it. (Here is one recent blog on the subject.) Correlated color temperature (CCT) is a very rough indicator of color, because an infinite number of different illuminants could produce any given CCT, and the light source's (U,V) position on, above, or below the blackbody curve has a perceptible effect. Color rendering index (CRI) speaks only to how well a light source mimics a reference source in reproducing pastel colors. No common metric relates to people's color preferences. Other metrics that fill in at least some of the gaps are not in widespread use.

The US Department of Energy (DoE) has published a new CALiPER report, the first of four supplements that extend the findings of "Report 20: LED PAR38 Lamps," published late last year. The new report is "CALiPER Report 20.1: Subjective Evaluation of Beam Quality, Shadow Quality, and Color Quality for LED PAR38 Lamps." The research was carried out by PNNL in Oregon.

PNNL color quality demonstration setup. Participants were not allowed to position themselves within the beam in order to use their skin as a reference to aid evaluation.

(Source: PNNL)

Demonstration facility

PNNL constructed a mock-up at its facility and invited members of the local IES chapter to view the 26 different lamps and fill out a questionnaire on the quality of beam, shadow, and color.

Here are the top-level conclusions:

• LED lamps generally showed well against halogen in the observers' ratings.


• For color quality, observers generally preferred 3000K LED lamps over those with a CCT of 2700K.


• The quality drawbacks that the observers found most noticeable were poor color consistency within the beam, and stray light outside the main beam.


• LED lamps with a single emitter were generally preferred for their beam and shadow quality.


• LED lamps with a narrow-spot distribution were viewed as having less-acceptable beam quality than those with narrow-flood or flood distributions.

Interestingly, the observers' evaluations of the lamps' color quality did not always accord with the rank order according to the product's CRI.

What's next

The three planned reports on PAR38 lamps yet to appear will cover (20.2) flicker, dimming, and power quality characteristics; (20.3) stress testing; and (20.4) lumen and chromaticity maintenance. I look forward especially to the final one. We have had a look at some of the preliminary data that PNNL is accumulating on the 40 lamps currently being monitored for this research.

How useful do you consider this kind of subjective evaluation of lighting as an extension to what metrics can reveal?

Related post:

— Keith Dawson , Editor-in-Chief, All LED Lighting