Lighting priorities? Lumens, Color Temp, or CRI?

[Dan Hintz]

Believe as you wish. Not worth wasting more time on it.

Agreed.....

[Dan Hintz]

Believe as you wish. Not worth wasting more time on it.

Sorry, leaving this hanging bugs the heck out of the engineer in me, so let me try to explain my point a bit further...

A blackbody radiator (ideal), when heated to the proper temperature (above 6000k), is considered to have a CRI of 100... this means the radiation given off by a blackbody radiator (we obviously only care about the visible portions) make an object appear to our eyes, from a color standpoint, as if it was bathed in natural light (e.g., the sun on a cloudless day). An incandescent bulb is generally listed as a blackbody radiator because it has the same general properties and is as close to a blackbody radiator as we'll get in an affordable package. So, a normally powered incandescent bulb has a CRI of 100.

But note earlier I said "when heated to the proper temperature"... this is key. If it's not hot enough, the emitted color spectrum does not cover the range of visible wavelengths (and for the sake of clarity I'm going to ignore our eyes'/brain's ability to perceive colors that aren't truly there via the right micture of wavelengths... similar to how white LEDs use a mix of blue dice and yellow-emitting phosphours to trick us into seeing white). If the temp is low enough, all we see is a hint of red glow (the Draper Point, about 798k... had to look up that value, memory is getting rusty). I think we would all agree if you light your subject with a dim red glow, the CRI of that radiator is no longer 100.

Back to the incandescent bulbs. Bulbs are designed with a certain thickness of filament for the desired wattage. The filament is our blackbody radiator. This thickness varies with wattage because you are trying to reach a certain temperature range for the given power (had to look up that temp, too... seems the average is about 3,300k). Reducing power to the bulb (i.e., dimming it), reduces the temperature of the filament. As noted above, reducing the temperature of a blockbody radiator (our filament) changes its output spectrum, and by consequence, its CRI.

Therefore, a bulb dimmed to a dull red glow does not have the same CRI as one operating at the specified wattage. And that is why you lose out on the blue end of the spectrum if you take a photograph of an object lit with a dimmed bulb.

If you disagree with anything I've written, let me know and I'll try to clarify...

[Aleks Hunter]

I ran this by my wife, a lighting designer with a master's in lighting from Rensselaer Polytechnic Institute's Lighting Research Center, a full member of the Illuminating Engineering Society of North America, and holding the Lighting Certified credential. As she was reading the post half way through she said "oh don't do that!" meaning the T-12 which has been banned from production and will not be available much longer and are less efficient. Also the T-12 bargain fixtures are likely older and flicker and noise may be an issue. That Craig list bargain will be expensive on your replacement lamp (which means bulb in lightign people speak), headache remedy, and electric bill. She recomends T-8 or T-5 in the 4000k - 5000k range which will allow you to see better per lumean that the lower or higer color temp. ranges. T-5 and T-8 are and will continue to be available, are cost effective. Modern electronic ballasts matched to the lamps will all but eliminate flicker. Cooler light in the 4100 range promoptes neateness and efficiency, is appropriate for general visual tasks. Higher temps. 5000k which is called daylight is better suited for galleries, museums and jewelry stores, and does render colors better.

T8 lamps are smaller diameter make using more efficient and more expensive rare-earth phosphors economical. Physicallyinterchangeable with T12, but can not operate on the same ballast. Meaning that the craglist bargains cannot support them.Operate on 265mAon rapid start ballasting system

T5 lamps use triphosphortechnology, available only in metric lengths and mini bipin bases, highersource brightness and better optical control than T8, optimum light output atan ambient temperature of 35 degree C (95F) rather than the more typical 25C(77F) allowing for the design of more compact luminaires, high output versions (about twice the lumens at the same length as the standard version) are available, T5's are unsuitable for most of retrofit applications, shallower luminaires than T8 but are overall more efficient.


One question is why are you so concerned about CRI? If it is for photography, you will get much better results using a flash that is matched to your camera's CCD. If it is for visual assessment and color matching, color appearance models are much better indices. CRI is a bit of an outmoded metric, simply a comparative ratio versus an ideal light source a full spectrum radiating black body, like an incandescent lamp by an ideal "eye". THe trouble with these is that they lose efffeciency to radiating at wavelengths less than and greater than the visible spectrum. CRI is not really a good indicator of visual assessment and color perception at all with color temperatures below 5000k. For close detail work, point source directional lamps like CREE LED's in heat sink housings are fantastic and available in a wide variety of point source lamps ranging from self contained MR-16 through larger PAR sizes with built in ballast which allows them to be connected into standard 115 wiring. They are more expensive up front, but the ir service life is tens of thousnads of hours longer and they are as or more energy (but not really as much as some lower quality manufacturers claim) efficient than any fluorescents. She has tested dozens of different manufacturer's LED, and CREE are absolutely the best so far. the LED technology is evolving rapidly, and costs will come down. When figuring cost, do not look solely at the price of the lamp or luminaire, look also at life span, disposal concerns, energy consumption, and a myriad of human factors consideration.

One such is the best investment you ever made, the only one guaranteed to always go up, your age. Two facets of presbyopia (aging eye syndrome) are yellowing and scattering which can be partially mitigated with higher color temperatures. If you find that you do not need stronger glasses, just longer arms to read a newswpaper, meaning your near point (the closest distance to your eye at which you can focus on an object) is increasing, then go with lamps at 5000k.

Getting back to simplifying things, color temperature is a straightforward albeit counterintuitive description of the apparent color of the light the lamp radiates. Lower color temperature numbers , say 2700 k are reddish, and will lend a nice warm th to the environment they illuminate and are called "warm white" while higher color temperatures, 5000K and above radiate a blue white light and appear cool to cold to the eye and are the so called cool white lamps, despite appearing to be radiated at a temperature 2300 Kelvins higher(Centigrade degrees with zero being absolute zero, or minus 273 celsius.

She recommends that you pick up a pair of lamps at 4100 K and 5000k set up a space where you can do some close very vissual task, like marking and carving carving under a 4 ft luminaire, work under each at different times, find which works better for you, and use those lamps to outfit your shop. If you can get a light meter, and have the flexibility to move the luminaires before locking them in place, arrange tehm so that the light levels on your work surfacea are between 50 and 100 foot candles or 500 and 1000 LUX. Be careful about placement castign unwanted shadows and producing glare on reflective surfaces such as table saw and router tables.

[Dan Hintz]

Aleks,

Any chance you kids live near D.C.? I'd love to pick your wife's brain over dinner. When Rensselaer first added the degree in lighting, I was sorely tempted to sign up... I was deep into LED design, and the color classes would have been great to have at the time. But alas, time being what it is, I ended up doing other things and learned what I could via other classes and books.

[Keith Avery]

Thanks for all the feedback. I already have the t12 fixtures. They are commercial t-12 troffer fixtures with Class a rated ballast(quietest t12s). I have 11 of these. I will probably buy t8 fixtures for the other 10 fixtures I need/want. This is just a hobby shop and I am not going to be using a ton of electricity during the 20 hours a month I will be lucky to spend there. Couple of things I found through google search that lead to several electrician forums that I would love feedback on. It seems electrician are in general agreement that t8 are far less reliable than commercial t12s. They also agreed that they loved t8s as most lighting desingers were changing out t12s and putting in t8s(so they had more work) and they always got to come back within 2 years to replacement 5-10 percent of the electronic ballasts on the t8s. Many of them said that 20 years ago it wasn't unusual for it to be more than 10 years before replacing a single t12 ballast on a 50 fixture job. In my office, I have replaced about 5 of 20 of my t8 ballasts in the last 3 years.
Based on the thread input I think I will go with lumens(priority 1), color temp(2), and CRI(3). It seems that I can buy 34 watt t12 with about 95 percent of the lumen output of most t8s. So I do see that t8 are a little more efficient but it doesn't seem like a huge difference, especially considering my level of use. I know t8s keep the brightness longer but many of the t12s lamps I saw maintained 87 percent of initial output, I think the t8s were around 92. I am guessing that it will be 15 years before I need to replace bulbs and if t12 aren't available I don't see why I can't just reballast with t8s then. It also seems that these metal on the t12 fixtures is about twice the thickness of the sheet metal on the $50 troffers availabe at Home Depot.
Let me know how misguided I am.

[Aleks Hunter]

Hi Keith,

Reballasting is a very good option if you already have nice heavy metal luminaires. the t8 and t8 fittings are identical.

Not all ballasts, like anything else are crateed equally. Back inthe early eighties some manufacturers did have high failure rates, but in the last 20 years there are many manufacturers of T* ballasts with documented failure rates of less than 1% over five years for ten years or longer.Lumens is really what you're buying, color temp is the flavor, and as said int he post, experiment to find your real preference. After all it is you who will have to work inthe space. There are much higher lumen output T8's coming oin the market. There are also far more choices available than what is on the rack in big orange box stores. An LC lighting designer will help guide you. Their code of ethics precludes them from profiting from any sales, so you can get a very objective recommendation that will be optimal for your purposes.

[Matt Meiser]

Closely related, is there any advantage mixing lamps on purpose? Say using one 4100K and one 5100K lamp in the same fixture? Getting ready to buy fixtures and lamps for my finishing room where I'll be using 4-lamp 8' strip T-8 fixtures and since I need to buy all the lamps I can pretty much buy what I want.

Interestingly while shopping lamps yesterday at Lowes I noticed the very expensive GE Reveal lamps that are supposed to show "pure, clean light". I had trouble finding the CRI rating on those but expected it to be very high. When I finally found it in the small print, I realized why--60 CRI, worst of any T8 lamp Lowes sells.

[John McClanahan]

Matt, I have 1 4 lamp fixture lamped with 2 warm white and 2 cool white lamps. I see no advantage to this, and would not do it to the other light fixtures. For me, 4100K is the best compromise.

John