WE NEED SPECIFICATIONS
FOR COLORED LIGHT BULBS

WHAT DO WE NEED?

We need the ability to measure standard values for colored light:

• We need a unit of measure of visual brightness stimulus used for colored light that corresponds to the lumen used for white light.
• We need an exact measure of the visual hue of the bulb.
• We need an exact measure of the visual saturation of the hue of the bulb.
• We need a spectral plot of the bulb output (some lamps don't actually emit the color that is visually seen).

  We need this for white bulbs too.

• We need an indication of the choppiness of the spectral plot.

We also need to know what kind of power supply the bulb has or needs:

• We need to know what kind of power supply or ballast the lamp has built into it.
• We need to know what kind of external power supply or ballast the lamp needs.
• We need to know the voltage, frequency, and power requirements of the lamp.

We also need a measure of the life of any lightbulb - one that actually works:

• The method of measuring lightbulb life works for incandescent bulbs, but not for other types.
• Often the power supply, instead of the light emitting element, determines the actual life of the lamp.
• The life of a fluorescent lamp does NOT depend on the number of hours the lamp has been on. It depends on the number of times the lamp has been turned on. A fluorescent lamp lasts for about 6500 starts. Power sags use up many starts.
• Compact fluorescent lamps (CFL) also depend on the number of starts (again near 6500), but overheating of the internal ballast further shortens the life of the lamp.
• LEDs are supposed to have a very long lifetime, but that is just for the light emitting element. The power supply in the base rarely lasts that long. In addition, heat shortens the life of the power supply.
• Power surges can cause sudden failure of CFLs and LEDs.

WHY DO WE NEED IT?

People do not know what they are getting when they buy a colored light bulb.

There are a lot of uses for colored light bulbs other than illumination at a party:

• Lighting a stage or a rock band
• Special lighting to compensate for color blindness
• Color blindness simulation
• Science experiments
• Photometry and colorimetry
• Ordinary color photography (adjustment for peaks in the spectrum of the ambient light)
• Special photography
• Adjustable room lighting color
• Blending light for dioramas
• Color from RGB inputs.
• Other special uses

Many of the colored light bulbs offered for sale do not produce light that is usable for many of these purposes. The page author has found at least 4 different kinds of colored LED light bulbs on the market:

• Bulbs designed for maximum pure-color light output per watt (e.g. Feit)
  • Has a wider bandwidth than other pure-color designs.
  • Can possibly be used for scientific purposes, depending what is needed.
• Bulbs designed with narrow spectral bands for purer colors (e.g. illumin8)
  • Has a narrow bandwidth than other designs.
  • Very useful for scientific purposes, but it depends on what is needed.
• Bulbs that simulate the poor color purity of colored incandescent bulbs (e.g. Zilotek)
  • Has a wide bandwidth and puts out some white light.
  • Intended mainly as direct replacements for lights in electric signs.
• White LEDs inside the bulbs with color filters in the outer covers (e.g. Great Value)
  • Has a wide bandwidth and leaks some white light.
  • Has all of the gaps in the spectrum the white LED has.
  • Intended mainly for decorative use and for lights in electric signs.
• The page author found two bulbs with accidental color impurities:
  • The clear outer case of a narrow band blue LED fluoresces a weak yellow when its own blue LED is on (Christmas Valley).
  • The clear outer case of a narrow band green LED fluoresces yellow-green when a nearby blue LED is on (Christmas Valley).

WHAT CAN BE DONE?

• Labeling standards must be devised and set for the labeling of colored lamps that are similar to the standards for white lighting.
• Labeling standards must use the same unit for all lamps. Right now, lamps are sold using candlepower, candelas, lumens, and lux.
• The law must be changed so the labeling for bulb life reflects the truth for each kind of lamp instead of using the misplaced belief that all bulbs age while they are emitting light that the legislators had when the law was passed.

WHAT CAN BE USED?

There are several ways this information can be measured:

Color Map

• A light meter for photography gets a good indication of visual brightness stimulus. A corrected selenium cell gives a nice flat result.
• A unit of measure of visual intensity that corresponds to the actual visual stimulus, instead of some energy value, is needed. It should be something similar to the lumen, but for color (e.g. a chromen).
• If the locus of the light can be found on the CIE Chromaticity map (right), the following can be calculated:
  1. The dominant hue as a wavelength in nanometers (nm) can be found from the direction of the locus from the equal energy point E on the CIE map.
  2. Nonspectral hues get the negative of the wavelength found 180° from the direction of the locus from the equal energy point (E on the map).
  3. The saturation will be distance from the equal energy point to the locus as a percentage of the distance from the equal energy point to the pure hue in that direction.
  4. The XYZ locus on the CIE map can be reported.
• The CIE locus can be found from modified RGB readings using a mathematical matrix.
• Note the following about the CIE map:
  1. The white triangle in the CIE map is the gamut of colors visible using a typical white LED.
  2. The triangle containing lines A and D are the colors visible on a color TV or monitor.
  3. The quadrilaterals containing lines B and D or lines C and D are proposals for 4-color TVs.
  4. The curved line through the middle is the set of loci of the various color temperatures of incandescent sources.
  5. The point E is the equal-energy point where R, G, and B have equal energies for a pure white). It has a color temperature of 5700 K.
• From a spectral plot of the bulb output, we can get the following:
  1. The spread of wavelengths in the light can be found as the distance in nm between the half-energy points on each side of the dominant hue on the plot.
  2. If the dominant hue is not emitted by the bulb, the peaks on each side (in nm) can be obtained.
  3. The standard deviation of the spectral plot intensity is an indication of the choppiness of the curve.
  4. The spectral plot explains many of the anomalies found in color photography.

THE BULBS

Appendix 1: Comparing Colored LED Bulbs

Here are the approximated values of various brands of colored LEDs:

For white lights:
- The BRIGHT column contains lumens, not chromens.
- The HUE column contains color temperature, not wavelength.
- The SAT column contains the Color Rendering Index (CRI).

For colored lights:
- The HUE column is negative if the hue is nonspectral. See the CIE color map above.

Other brands:
- The Enchanted Forest C9 and C7 lamps are the same as the Christmas Valley C9 and C7 lamps.

Appendix 2: Measuring LED Bulbs

Here us the procedure to measure the various parameters of colored LEDs:

Preparing:

1. Get the needed equipment:
   • A computer with a spreadsheet (for recording the data and making the calculations)
   • A room that can be darkened
   • An adjustable lamp or stand with the socket for the bulbs to be tested
     (it should be capable of being rotated to present different parts of the bulb to the light meter)
   • A light meter that reads footcandles or lux
   • A small light used to read the light meter, placed where it will not affect the reading
     (the page author used light from the computer monitor and a cardboard baffle to keep this light off of the photocell)
   • A measuring stick calibrated in US and metric units.
   • A pair of Diffraction Grating Glasses (sold in science and novelty stores)
   • If measurements of how much power the bulb consumes are needed, a wattmeter is a good idea. A good choice is a Kill-O-Watt consumption meter. It can measure watts, volt-amps reactive, and power factor.
2. Set up the equipment:
   • Determine whether or not the light meter uses footcandles or lux.
   • Determine which method will be used to measure and calculate the amount of light.
     • Meter uses footcandles, set up for footcandles
     • Meter uses lux, set up for footcandles
     • Meter uses lux, set up for lux

     Setting up for footcandles makes the setup much smaller.

   • Find out whether the sensor of the light meter will face horizontally or vertically when it is placed on a stable surface.
   • Arrange the light meter and the lamp so the lamp is the proper distance away and is located perpendicular to the surface of the photocell on the light meter:
     • Space them horizontally if the meter looks horizontally.
     • Space them vertically if the meter looks vertically.
     • Space them exactly one foot apart to set up for footcandles.
     • Space them exactly one meter apart to set up for lux.
     • Place a wedge, box, or prop under the light meter if the meter looks horizontally and the lamp is higher than the meter.
   • The page author's setup is shown at right.
     • The page author was lucky to find in a flea market a working Weston Electrical Instrument Corp. model 756 footcandle meter. It was manufactured in the 1930s and 1940s for stage lighting use.
     • It has a separate sensor head on a cable. Two 1.5"-diameter color-corrected selenium cells are in the sensor head. Amazingly, it has retained its calibration after all of these years.
     • The carpenter's square is bolted to the board. It holds up a clamp-on light (with reflector removed) and shows the height of the bulb above the sensor.
   • Make sure the light from the computer monitor and the light from any light used to read the light meter do not illuminate the photocell.

   Be sure the area can be darkened.

3. Calibrate your setup with several known bulbs:
   • Calibrations (sample using footcandles)

     Use the procedures below under the headings:
     - Get information from the bulb
     - Take readings from the bulb
     - Do the calculations

     Calibration: Meter uses Footcandles, Set up for Footcandles
     	A	B	C	D	E	F	G	H	I
     1	PRODUCT	fc	cd	nom lm or cr	shad ang	sphere difference	unshaded part	bright portion	lm or cr
     2	FEIT 40 DL 450 lm	57	57	716	150	0.741181	7.90939	0.62941	451
     3	Formulas for above	57	= B2	= C2 * 4 * PI()	150	= 1 - COS(E2 / 2 * PI() / 180)	= 4 * PI() - 2 * PI() * F2	= G2 / 4 / PI()	= D2 * H2

     The results match the bulb within a small error.

   When in doubt, experiment. Adjust the lamp angle and the entered shadow angle.

Collecting the Data:

1. Get information from the bulb:
   • Read the bulb package for brand and product identifying information.
   • Read the package for any luminance, color, and power consumption information.
   • Measure the angle θ, with the vertex at the center of the bulb globe or light emitter (if visible), of the portion that does not emit light (the base) from edge to edge.
2. Take readings from the bulb:
   • If reading power consumption, connect the bulb to power through the wattmeter.
   • Adjust the orientation of the bulb so the direction to the photocell is approximately halfway between the edge of the base and the apex of the bulb. This usually gives the best reading.
   • Adjust the location of the bulb so the nearest light emitting part is the required distance to the light meter.
   • Take the reading.
   • Take readings from the wattmeter (if used).
   • Look at the bulb through the diffraction grating glasses from a distance of at least 10 feet and note any gaps or bright lines in the spectrum. Compare it to the spectra on this page.
3. Do the calculations:

   Put one bulb in each row of the spreadsheet.

   Enter the following in the spreadsheet:

   • Cell A - The make and name of the lamp
   • Cell B - The light meter reading (in footcandles or lux)
   • Cell C - Calculate the number of candelas.
   • Cell D - Find the nominal lumens of a spherical source by multiplying by 4π.
   • Cell E - The "shadow" angle of the nonluminous zone from edge to edge
   • Cell F - Find the sphere difference by subtracting the cosine of half the shadow angle from 1
     (the spreadsheet requires radians, so the angle is multiplied by π / 180).
   • Cell G - Find the unshaded part by subtracting 2π times the sphere difference from 4π
   • Cell H - Divide the unshaded part by 4π to get the bright portion.
   • Cell I - Find lumens by multiplying nominal lumens by the bright portion.

   When dealing with colored light, substitute the proposed chromen unit of measure for the lumen.

   • Footcandle Calculations

     The setup was purposely designed so the number of footcandles is equal to the number of candelas.

     Meter uses Footcandles, Set up for Footcandles
     	A	B	C	D	E	F	G	H	I
     1	PRODUCT	fc	cd	nom lm or cr	shad ang	sphere difference	unshaded part	bright portion	lm or cr
     2	FEIT 40 Red LED	14.5	14.5	182	150	0.741181	7.90939	0.62941	115
     3	Formulas for above	14.5	= B2	= C2 * 4 * PI()	150	= 1 - COS(E2 / 2 * PI() / 180)	= 4 * PI() - 2 * PI() * F2	= G2 / 4 / PI()	= D2 * H2

   • Lux Calculations

     The setup was purposely designed so the number of lux is equal to the number of candelas.

     Meter uses Lux, Set up for Lux
     	A	B	C	D	E	F	G	H	I
     1	PRODUCT	lx	cd	nom lm or cr	shad ang	sphere difference	unshaded part	bright portion	lm or cr
     2	FEIT 40 Red LED	14.5	14.5	182	150	0.741181	7.90939	0.62941	115
     3	Formulas for above	14.5	= B2	= C2 * 4 * PI()	150	= 1 - COS(E2 / 2 * PI() / 180)	= 4 * PI() - 2 * PI() * F2	= G2 / 4 / PI()	= D2 * H2

   • Lux Meter With Footcandle Setup Calculations

     The setup was designed so lux must be converted to footcandles to get the number of candelas.

     Meter uses Lux, Set up for Footcandles
     	A	B	C	D	E	F	G	H	I
     1	PRODUCT	lx	cd	nom lm or cr	shad ang	sphere difference	unshaded part	bright portion	lm or cr
     2	FEIT 40 Red LED	156	14.5	182	150	0.741181	7.90939	0.62941	115
     3	Formulas for above	156	= B2 * 0.0929	= C2 * 4 * PI()	150	= 1 - COS(E2 / 2 * PI() / 180)	= 4 * PI() - 2 * PI() * F2	= G2 / 4 / PI()	= D2 * H2

   When in doubt, again experiment. Adjust the lamp angle and the shadow angle. The lamp may have an unusual emission pattern.

Appendix 3: About Footcandles, Lux, Candlepower, Candelas, and Lumens

Here is the explanation of these units of measure:

• The STERADIAN (sr) is the metric unit of solid angle. A sphere is 4π sr.

  For photometric use, the steradian is a conic part of a sphere with included angle sr of 65.541 degrees.

  In the diagram at right, the white area shows the conic part of the sphere that makes up a steradian. The yellow part on the surface of the sphere is a circular cap with a solid angle of 1 steradian.

• The CANDELA (cd) is the metric unit of luminous intensity.

  The candela is a measure of light leaving the source within a steradian cone.

  The STANDARD CANDLE and the CANDLEPOWER are US units equal to the candela.

• The FOOTCANDLE (fc) is the US unit of illuminance. It is one candela shining on one square foot.

  This is defined as the steradian cone (white) in a sphere with a radius of one foot and a curved circular area (yellow) of one square foot at the wide end. The light shines on the yellow cap.

• The LUX (lx) is the metric unit of illuminance. It is one candela shining on one square meter.

  This is defined as the steradian cone (white) in a sphere with a radius of one meter and a curved circular area (yellow) of one square meter at the wide end. The light shines on the yellow cap.

• The LUMEN (lm) is the metric and US unit of luminous flux. It measures the total luminous output of a light source.

  The difference between the lumen and the candela is that the lumen indicates the total output of the light source, while a candela indicates only the light directed in the direction of the steradian cone used.

  An omnidirectional source (the entire sphere) with 1 candela in each possible steradian cone produces 4π lumens.

  A lightbulb usually has a shadow zone at the screw base where no light is emitted (darkest area in diagram). For lamps with the same candela value in a bright zone, the lumen output is lower by the area of the sphere blocked by the shadow zone.

  In the diagram, the shadow angle sh is 90°.

• The CHROMEN (cr) is my proposed unit of luminous flux for colored light. It measures the total luminous output of a colored light source, but indicates that it is colored light.

  The difference between the chromen and the lumen is that the lumen is defined for only white light, while the chromen is also defined for colored light. They both have the same value for the same bulb.

  The footcandle or lux reading (depending on the meter design) of a color-corrected selenium cell is used to calculate the chromen value for colored light in the same way it is used to calculate the lumen value for white light.

• Lightbulbs should be sold in lumens. But some are sold in candelas or candlepower.

  To find the number of lumens for a light source rated in candelas or candlepower, use the following procedure:

  • Get the number of candelas (= candlepower).
  • Find the nominal lumens of a spherical source by multiplying by 4π.
  • Find the "shadow" angle sh of the nonluminous zone of the bulb from edge to edge.
  • Find the difference by subtracting the cosine of half the shadow angle from 1.
  • Find the unshaded part by subtracting 2π times the difference from 4π.
  • Divide the unshaded part by 4π to get the bright portion of the bulb.
  • Find lumens by multiplying nominal lumens by the bright portion.

  If the source is omnidirectional, multiply the candlepower by 4π.

Understanding the relationships between these units is essential.