Household lighting adventures

A few years ago (2007 or 2008) I got pretty excited about the GE Reveal incandescents with neodymium-doped glass. I bought into GE's marketing term "enhanced full spectrum" (in my mind, I imagined a hotter filament or maybe some form of two-photon fluorescence). When I got them home, however, I saw that they're not much different from normal incandescents, and still not at all like daylight.

I put together a CD spectrometer, which showed the GE bulbs to have a sharp absorption band in the yellow region but otherwise identical to a normal incandescent. Hardly "enhanced" or "full" in any way. So I mailed them an angry letter, citing my spectrometry results, and a few weeks later they sent back a $5 coupon for GE light bulbs.

After that disappointment, I started buying CFLs. It was hard not to; PG&E was (and still is) subsidizing them to $1 for a 23W bulb (100W equivalent). Based on my estimates, it'd take less than a month to recover that $1 on your electricity bill. The lower wattage also means you can put more powerful bulbs in smaller lamps (like this beautiful hanging fixture we have that's made of tissue paper and kindling).

So they're cheap and efficient, but I've had problems with reliability (two have somehow failed) and fit (we have a floor lamp that won't fit the CFLs' wider ballast base). I'm not too concerned about the mercury, but maybe I should be? It's been a couple years and I still haven't gotten around to recycling those busted CFLs.

A lot of people take issue with fluorescents' higher color temperature. I personally prefer it, but I grew up with daylight-balanced T8 fluorescent lights at home, so incandescents all look yellow and dim to me. My problem with fluorescents is the spiky emission spectrum. The human eye doesn't seem to mind (though the 2700K CFLs do look pretty weird), and digital cameras compensate with auto white balance, but analog film is not so tolerant:

iPhone 4

Velvia 50

I've also tried a Fuji film with their "4th Color Layer", which improves color accuracy under fluorescent lighting. It works a lot better, but still comes out slightly green. These were taken under the same lighting as above:

iPhone 4

Fujicolor Superia X-tra 400

And this is not to say that my interest in continuous-spectrum lighting is primarily for photography; I take very few pictures indoors. But maybe there will be some general quality-of-life benefits to having higher CRI artificial lighting.

Recently I got excited about halogen lamps. I'm most familiar with them as dorm room fire hazards and Source Four theater lights, but recent news about the EISA §321(a) incandescent lamp energy efficiency standards and Ikea discontinuing incandescent bulbs brought up halogens as an energy-efficient incandescent replacement. I found this a bit surprising because I always thought of halogens as a comparatively inefficient devices. But that is wrong.

The energy savings come from increased emission of visible (vs. infrared) light due to the higher temperature black-body (the halogen gas makes this possible by mitigating filament evaporation). For the incandescent replacements, the quartz halogen bulb is mounted inside a larger (incandescent-sized) UV-filtering glass bulb, reducing fire hazard (the bulb surface should be cooler than an incandescent of equivalent light output) and removing the possibility of fingerprint-induced bulb explosions.

Anyways, it's an attractive combination: higher color temperature, black-body spectrum, and drops in to a standard E26 screw base. So I decided to buy one to try it out. Impressions:

• I'm not seeing this purported energy efficiency. The Philips Halogena bulb I bought says "Energy Saver 40w = 75w" on the packaging, but it only produces 550 lm, which is far less than a 75W incandescent.  Its energy efficiency is pretty weak, too:
  

  Type	Power (W)	Output (lm)	Efficiency (lm/W)
  GE soft white A19 incandescent	40	490	12.3
  Philips R20 indoor flood halogen	40	550	13.8
  GE soft white A19 incandescent	60	840	14.0
  GE soft white A19 incandescent	75	1170	15.6
  GE soft white A19 incandescent	100	1690	16.9
  Source Four HPL 115VX (3050K) halogen	575	12360	21.5
  Source Four HPL 115V (3250K) halogen	575	16520	28.7
  Feit Electric compact fluorescent	23	1600	69.6

• Perhaps having a reflector (R20) makes the overall light output less efficient than an omnidirectional bulb (A19)? However, the omnidirectional bulbs lose more light in the fixture, so their actual output is less than what's in the table.
• The last entry is the CFL that used to be in the fixture now occupied by the halogen. It's also an omnidirectional bulb.
• The above investigation pointed out a trend I hadn't noticed before: tungsten efficiency goes up with wattage. Maybe there's less conductive/convective thermal losses? Maybe filament limitations only let you go up to higher filament temperatures at higher powers? Perhaps the higher-wattage halogens will be more efficient.
  
• Color temperature is also pretty lame. Not any whiter than an incandescent. This may also increase with wattage, though, so maybe I'll need to get a 100W halogen to see a difference. The 3250K theater lights are definitely whiter than this, but they're also only rated for 300 hours of operation.