contact us

Use the form on the right to contact us.

You can edit the text in this area, and change where the contact form on the right submits to, by entering edit mode using the modes on the bottom right.

912 Live Oak Park Rd South
Fallbrook, CA, 92028
United States

+1 (760) 728-9121

Pala International has consistently earned its reputation as the direct source for the finest colored stones.

Buying and Selling Gems: What Light is Best? - Part 2

Buying and Selling Gems: 
What Light is Best?


By William J. Sersen, Ph.D., AG (AIGS)
Formerly of the Asian Institute of Gemological Sciences, Bangkok, Thailand

Note: This article is reprinted with permission. It originally appeared in the Gemological Digest, Vol. 3, No. 1, 1990, pp. 45–56. To read Part I of this article, click here


What light source best qualifies as a standard for dealers and graders of colored stones? Natural daylight isn’t the answer, for it varies too much in strength and spectral composition. That leaves only man-made lamps to choose from. But which one? The author tries to answer that question, reviewing the merits, disadvantages and operational theory of a number of commercially available options. He concludes by identifying a lamp that appears to meet all criteria for a universal lighting standard.


There was once a gem dealer named Rick O’Shea (his friends called him Ricochet, for short). Living on a dreary, foggy island off the coast of Ireland, a place where the sun seemed to set more than rise, he grew accustomed to grading his gemstones in dim lighting conditions. “What, me use artificial lamps?” he would say jokingly to his dealer friends who often gathered at their favorite pub, The Loupe & Tweezers. “The stones wouldn’t look natural if I did that, now would they?”

Ricochet thought that all gems were dark-toned. Moreover, he had never seen asterism before, despite what he read of it in his FGA course materials. The light on his tiny island just wasn’t strong enough. Sure, he believed that asterism existed, because the book said so. Pluto was probably the ninth planet too. But light-toned gems: “No way,” was Ricochet’s famous and proverbial reply (a reply his wife learned to echo, nightly, to poor old Ricochet: she liked having a lamp turned on, but he refused to buy one).

Other dealers–those who lived on the same island, but used artificial daylight lamps to do their grading–loved doing business with Ricochet. Secretly, they spoke of him as Ol’ Backfire, referring to his business acumen. Australian customers called him Boomerang for the same reason. His grading techniques were legion: everyone knew that if Ricochet-Backfire Boomerang wanted to sell cheaply a blue sapphire he considered too dark in tone, that meant it was a more saleable medium tone under everybody else’s lighting arrangements. After repeated deals, Ricochet-Backfire-Boomerang went out of business. Ultimately, he became a lonely and forsaken pessimist whose dim view on gems extended to life generally.

To avoid falling into the trap that snared poor Ricochet, let’s read on. Gem dealers who read this article will easily relate it to MONEY. Professional graders may regard it more as a contribution toward standardizing gem-grading lighting conditions. Put simply, retailers will either love it or hate it!

“What light source best qualifies as a standard for dealers and graders of colored stones? Natural daylight isn’t the answer, for it varies too much in strength and spectral composition. That leaves only man-made lamps to choose from. But which one?”

As we saw in Part I of this article, dealers often buy and sell gems under natural lighting conditions. Of those that do, most prefer assessing stones with reflected skylight, not direct sunlight. For latitudes north of the equator, north skylight is frequently used as it contains the least glare; for latitudes south of the equator, south skylight is the general rule and for the same reason.

We also observed that the quality of natural light varies dramatically with latitude, weather conditions, atmospheric pollution, season and the sun’s position in the sky. All contribute to the strength and spectral composition of what we call natural daylight. Consequently, a gem seen under daylight can and does change appearance relative to where we are, what the weather is like, the degree of pollution, the season and the time.

It is therefore incorrect to think of daylight as a standard light source. For something to be standard it should have constant properties, which global daylight does not. If daylight were constant everywhere, that dark stone bought in Bangkok would not look darker in New York or Amsterdam. Nor would a blue sapphire necessarily look better at 5 PM than at I PM – even at the same latitude!

So what light source is best for dealers and graders to view gemstones (Hold tight retailers, we’ll get to you later)? In a sense, any light source is best so long as it has constant properties and is used consistently. Anyone who has spent time color grading gemstones under specific lighting conditions, such as a certain kind of fluorescent lamp, gets used to how stones appear under that light. Abruptly switching to a different kind of lamp, let’s say incandescent or even a cooler or warmer fluorescent model, is tantamount to changing the rules: the grader temporarily loses his bearings or color “reference points.” The stones do not look the same. The only cure for that ill is for our grader to spend a lot of time looking at gems with the new lamp: long enough time to re-establish his bearings. It is sort of like target shooting with a favorite pistol; suddenly switching to a different pistol results in an adjustment period before one’s accuracy is back on par. In other words, it is a question of what one is used to and thus feels comfortable with.

But if a grader consistently uses one light source by which to grade, and his colleagues each use some other light source, then all will differ in their grading results, even though each feels comfortable with his respective lighting arrangement. This applies to gemstone graders, cotton graders and even snakeskin graders (the latter who would be hissing in the wind)!

Therefore, we must distinguish between personal and universal lighting standards. If somebody assesses the color quality of gems under one type of light, and so is accustomed to how stones appear relative to one another under that particular light, then that person has established a personal lighting standard. Only if everybody uses the same or similar lighting arrangement does that standard become universal. The closest thing to a universal lighting standard in the gem industry is the GIA DiamondLite for diamond grading purposes. There is no universal lighting standard for color grading colored stones.

“What light source best qualifies as a standard for dealers and graders of colored stones? Natural daylight isn’t the answer, for it varies too much in strength and spectral composition. That leaves only man-made lamps to choose from. But which one?”

Other industries, such as the cotton and textile trades, have long ago made headway toward defining universal lighting standards for grading their commodities. Comparatively, the gem industry is still in the dark ages. Why is this the case? Lack of discussion on the subject, partly because of a lack of perceived need to standardize, partly also due to the inherent conservatism of the industry. This indifference has only harmed the trade as a whole; with a multitude of personal lighting standards but no universal standard, there can be no universal color communication. After all, the color of any object depends on three things: the object itself, the eye and the light source. Assuming that graders and traders have “normal” color vision, than we are left with only the object and the light source. If you change either, you are apt to perceive color differently. If the object is a colored gemstone, that difference may translate into many dollars and cents.

So again, what light source should dealers and graders use when viewing gemstones? For that matter, does an ideal lamp exist? In order to answer those questions, we must first examine in detail the criteria for an ideal lamp and, equally important, how and why we have chosen these criteria. Later we will look at the lighting options commercially available: those lamps that meet the grade, those that come close, and those that fall short of the mark. In the course of this little discourse, a few technical terms will arise, but – worry not – nothing intimidating or unfathomable. We promise that this brief submersion into the waters of illumination theory and practice will be totally painless!

“…the quality of natural light varies dramatically with latitude, weather conditions, atmospheric pollution, season and the sun’s position in the sky. All contribute to the strength and spectral composition of what we call natural daylight.”

What Constitutes the Ideal Lamp?

Criterion #1: Sufficient Strength of Light 

The majority of dealers polled in Part I of this article said that light intensity contributed greatly to the quality of a stone’s color appearance.

Indeed, what do dealers usually do when another dealer or broker hands them a stone–at least, dealers in Thailand, Burma, Sri Lanka and in parts of Europe and the States? Answer: they hold it about a foot away from a window and examine it under reflected daylight. This reaction is so commonplace that it seems almost instinctive. Usually, there are artificial lights in the same room, but artificial lighting can not (in most cases) duplicate the strength of natural daylight, especially on a clear or partly cloudy day. And part of the reason why dealers bother to take the stone to a window in the first place is because natural light–at most latitudes and in most seasons–is relatively strong.

It follows that a lamp used by dealers for buying/ selling/grading should illuminate gemstones as much as the light entering from a typical window. The trick here is to define “typical window.” You see, the amount of light that falls on a unit area, which is what lighting engineers call the degree ofilluminance, depends entirely on window direction, latitude, season, time and weather conditions.

Table I shows the average annual daylight illuminance from 8 AM to 4 PM at 30°N to 50°N latitude for windows facing north, east, south and west. Clear, partly cloudy and overcast days are included in these averages. Figures are in kilolux. A lux is a unit of illuminance. To help put things into perspective, 1000 lux is roughly the amount of light a common photographic light meter would register at about one foot from an 18-inch (20 watt) fluorescent “daylight” lamp. Three thousand lux is about what a 60-watt incandescent bulb produces at one foot. Forty thousand lux and more is what that same meter registers when placed right next to a Bangkok window facing south on a clear day.

As you can see, there are huge differences in illumination depending on the direction the window happens to face. Predictably, daylight is dimmest with overcast weather, equally so in all directions. Perhaps not so predictably, north daylight is stronger when the weather is partly cloudy than when it is clear. For latitudes south of the equator (30° to 50°S) the same can be said for south daylight.

Table 1: The amount of daylight entering a window depends on the window direction, the latitude, season, time and weather conditions. (average kilolux from 8 A.M. to 4 P.M. for all seasons)
Lattitude Clear
Partly Cloudy
30° 7 28 39 28 11 21 27 21 5
34° 6 27 41 27 10 20 27 20 5
38° 6 26 42 26 9 19 27 19 5
42° 5 25 42 25 9 18 26 18 4
46° 5 23 43 23 8 17 26 17 4
50° 4 22 43 22 7 16 24 16 4




Within these same latitudes the average illuminance for all window directions at all seasons, weather conditions and times is slightly more than 19 kilolux. However, north is the most glare free direction, which is why dealers often use it when examining stones. Average illuminance for a north-facing window at all seasons, weather conditions and times, again for the same latitudes, is slightly over six kilolux. In either case, this is far more illuminance then that of 20-watt fluorescent daylight lamps at distances of one foot or greater.

It seems reasonable, given dealer habits and the above lux figures, that our ideal lamp should similarly produce an illuminance in the range of six kilolux at one foot distance.

Just how important is distance? 

Distance is extremely important. A curious thing about light is that in doubling the distance from a source, let’s say a lamp or candle flame, the resulting illuminance is not halved as one might think. For instance, if a lux meter is held one foot from some lamp, a reading is taken, and the process is repeated two feet away, the lux meter will not register half the illuminance. It instead shows one quarter the illuminance. Three feet away the illuminance drops to 1/9th; at four feet to 1/16th; at five feet to 1/25th, etc. This is known as the inverse square law, and is depicted in Figure 1. The law states that the amount of light falling on a body or surface is inversely proportional to the square of the distance.

This means that a lamp giving six kilolux of illuminance at one foot would register 1.5 kilolux at two feet, because the same quantity of light is spread over four times the area. Six inches from the lamp, the lux meter would read 24 kilolux (4 x 6).

The point is that if one selects a grading lamp partly for its illuminance, the inverse square law can not be ignored. When stones are graded, they should be held at a specific distance from the lamp to help insure consistent results. Deviating much from that distance can affect the stone’s color appearance to greater or lesser degrees.

Table 2 has further examples of the inverse square law. If you feel more at home with footcandles than lux, just multiply the lux figures times 0.10 for footcandles.

Figure 1. Illustration of the inverse square law (after White).

Figure 1. Illustration of the inverse square law (after White).

Criterion #2: Ability To Make Objects Look Natural

Lighting helps dictate how an object appears to our eyes. Lighting may be natural or artificial. Few artificial light sources render color the same way daylight does. As we have seen, not all phases of daylight render color exactly the same way either.

So, what is meant by ability to make objects look natural? For that matter, what does natural mean? If something seems natural to us–it could be a color, a smell or even an experience–all that means is that we are accustomed to that “something.” In this sense natural appearance andnormal appearance are synonymous. Both are subjective impressions; both are relative to what we are used to; commonly, both mean much the same thing.

For many colored-stone dealers, gems look natural when observed in daylight. Since this so often means a glare-free direction of daylight, we can limit the discussion to north daylight. We have already discussed average illuminance of north daylight. But, there is something every bit as important as illuminance, and that is spectral composition.

Table 2: The inverse square law says that the amount of light falling on a body or surface is proportional to the intensity of the lamp and inversely proportional to the square of the distance. The ‘amount of light’ is called illuminance and measured in lux or footcandles.
Lamp output in candle power Distance from lamp in feet Illuminance in Lux Light falls on ‘X’ times area
1.0 0.50 40 0.25
1.0 0.60 28 0.36
1.0 0.70 20 0.49
1.0 0.80 16 0.64
1.0 0.90 12 0.81
1.0 1.0 10 1.0
1.0 2.0 2.5 4.0
1.0 3.0 1.1 9.0
1.0 4.0 0.63 16
1.0 5.0 0.40 25
300 1.0 3000 1.0
300 2.0 750 4.0
300 3.0 330 9.0
300 4.0 190 16
1000 1.0 10000 1.0
1000 2.0 2500 4.0
1000 3.0 1100 9.0
1000 4.0 630 16

A light source that provides the right amount of illuminance but is not spectrally balanced is no good for grading, as it does not render color adequately. It does not do so because it cannot do so. Let’s pursue that point.

Many light sources that look white to our eyes may vary drastically in spectral composition. The more they vary, the more different they render object color. For instance, daylight sources (natural and artificial) and incandescent lighting look somewhat whitish to the eye. Yet, why is an alexandrite red under incandescent light and green under daylight?

Alexandrites transmit (= allow to pass through) and reflect green and red wavelengths. Since the spectrum of incandescent light is strong in red wavelengths, Alexandrites have plenty of opportunity to transmit/reflect red light when illuminated by that source, and so appear red. They are green under man-made or natural daylight because that kind of lighting contains more blue/green wavelengths.

Similarly, rubies seem a deeper red under incandescent lighting than under most fluorescent lamps. This is again because incandescent lamps give out more red wavelengths than fluorescent lamps.

“Similarly, rubies seem a deeper red under incandescent lighting than under most fluorescent lamps. This is again because incandescent lamps give out more red wavelengths than fluorescent lamps.”

Blue sapphires are another example. These stones look bluer under skylight or cool white fluorescent lighting than under a warm white fluorescent or incandescent lamp. In fact, deep blue or medium-dark sapphires can look as black and inky as onyx if displayed under a strongly illuminant incandescent lamp. This is predictable, as skylight and cool white lighting are very rich in blue wavelengths while incandescent lamps are not.

Hence, the perceived color of gemstones depends on the spectral composition of the source. Since a gem transmits and reflects only certain wavelengths, the source must contain those wavelengths or the incident light is absorbed, making the stone appear black or gray. It is for this reason that red paint is red under white light. Red paint only reflects wavelengths that correspond to a red visual sensation, absorbing all others. If that same paint is viewed under a green light source, containing only green wavelengths, it appears black. This is because the paint absorbs the green wavelengths and has nothing to reflect. An essential difference between paint and transparent gemstones is that the latter transmit, reflect and absorb light whereas paint can only reflect and absorb light.

Since we have chosen north daylight as our model light source on the basis that it makes gems look natural to many dealers, let’s examine its spectral composition in detail.

Figure 2. Comparative spectral power distribution of CIE Illuminants D50, D55, D65 and D75. All curves intersect at the same point.

Figure 2. Comparative spectral power distribution of CIE Illuminants D50, D55, D65 and D75. All curves intersect at the same point.

Spectral power distribution

The spectral composition of any natural or man-made light source may be measured and graphed. The graph is called a spectral power distribution (s.p.d.) curve. The curve states precisely how much light the source radiates at each wavelength. This helps tremendously in assessing the color rendering ability of the source.

Figure 2 shows the s.p.d.’s of four phases of natural daylight, the so-called CIE Standard Illuminants D50, D55, D65 and D75. The D50 is a white light with a fairly even curve in the visible spectrum. It is the evenness of this curve that gives this source a white appearance similar to noon sunlight. D55 contains a bit more blue and a bit less red. D65 and D75 are distinctly weighted toward the blue; their color appearance is comparable to a slightly overcast sky. Pure north skylight on a clear day is bluer still. The s.p.d. of north daylight (a mixture of skylight-and a little direct sunlight) is, like the D65–D75 curves, weighted toward the blue. 

The apparent color of these illuminants is referred to as the color temperature. Color temperature is measured in degrees Kelvin. For instance, the color temperature of the D50 illuminant is 5000° K; that of D75 is 7500° K; incandescent lamps start at around 2600° K, which is a slightly yellow appearance.

For a quick appreciation of what color temperature is all about, see Figure 3. Table 3 clarifies the concept in greater detail and with more examples.

Figure 3. Color temperature describes how a lamp appears when lit. The term applies equally to natural light sources. (Courtesy of GTE Products Corp.)

Figure 3. Color temperature describes how a lamp appears when lit. The term applies equally to natural light sources. (Courtesy of GTE Products Corp.)

Table 3: The color temperature of man-made and natural light sources. If an iron bar were heated to the temperatures on the left, it would have a color similar to those sources.
Color Temperature Natural Light Source Man-Made Light Source CIE Standards
28,000° K Very blue clear northwest sky    
20,000° K      
15,000° K Blue sky with thin white clouds    
7,500° K     Standard Illuminant D75
7,000° K Uniform overcast sky   Standard Illuminant ‘C’
6,774° K      
6,500° K (Pale blue)     Standard Illuminant D65
6,100° K   DiamondLite  
5,500K° K   Mitsubishi Daylight 99; Duro-Test VitaLite Standard Illuminant D55
5,300° K (White) “Noon sunlight”    
5,000° K (White) “Noon sunlight” Sylvania Design 50 Standard Illuminant D50
4,874° K (White) “Noon sunlight”   Standard Source & Illuminant ‘B’
4,870° K (White) “Noon sunlight”    
4 200° K   Cool white & deluxe cool white fluorescent  
4,000° K   White fluorescent  
3,100° K   Incandescent lamps  
3,000° K (Yellow)   Incandescent lamps; warm white fluorescent  
2,950° K   Incandescent lamps; deluxe warm white fluorescent  
2,855.6° K   Incandescent lamps Standard Source & Illuminant ‘A’
2,600° K   Incandescent lamps  
2000° K   Candle flame  
1,800° K Direct sunlight at sunrise/sunset    
1,000° K      
800° K (Deep red)   Heating element of electric stove  
290° K (Black)      

Conversion Key —
Fahrenheit to Celsius: F = (9/5) C+32
Celsius to Fahrenheit: C = 5/9 (F-32)
Celsius to Kelvin: K = C+273

If we know only the color temperature of a lamp, then we know little about the way that lamp renders object color. You see, the approximate color temperature of a lamp can be inferred from its s.p.d., but the reverse is not always true: the s.p.d. can not always be inferred by the color temperature. This is important, because s.p.d. determines the color appearance of objects, such as gemstones; color temperature alone does not.

Figures 4 and 5 are extreme examples of how two lamps that look alike when lit, i.e., have similar color temperatures, can render object color quite dissimilarly. Figure 4 is the s.p.d. of a lamp having total output in two bands of blue and yellow light. Figure 5 is a continuous curve spectrum of a lamp with an output at all wavelengths (= colors). In either case these lamps appear white to the eye. But, what happens if we use them to illuminate an object?

A red object appears black or gray under the first source and red under the second. This is because one lamp emits red wavelengths and the other does not. A yellow sapphire would look yellow under either lamp, since both emit yellow wavelengths.

So, the color temperature of a light source is little indication of how it renders object color, and two sources that look the same may render colors differently. On the other hand, two sources with the same or similar s.p.d. must render color the same way.

Figure 4. Spectral power distribution of a lamp with total output in the blue-green and yellow-orange (after Allphin).

Figure 4. Spectral power distribution of a lamp with total output in the blue-green and yellow-orange (after Allphin).

Can we measure a lamp’s color-rendering ability?

Browsing through lighting-manufacturers’ catalogs, one comes across the terms Ra or C.R.I., which stand for Color Rendering Index. This index is supposed to tell us how well a lamp renders color compared with a reference source of the same color temperature.

Color Rendering Index was the brainchild of the Commission Internationale de l’Eclairage, better known as the C.I.E. The idea is simple in principle. A minimum of eight sample colors, consisting of (opaque) Munsell papers, are viewed under each source. If in both cases the color of all the papers looks the same–in other words, there is no color shift compared to the reference source– then the test lamp has a C.R.I. of 100, the highest rating. Reference sources are automatically rated at C.R.I. 100.

Figure 5. Spectral power distribution of a lamp having a similar color temperature to the one in Figure 4 (after Allphin).

Figure 5. Spectral power distribution of a lamp having a similar color temperature to the one in Figure 4 (after Allphin).

As a rule of thumb, this means that a lamp rated at C.R.I. 90 renders color 90% as well as some reference source; C.R.I. 80 indicates an 80% rendition, etc.

In this writer’s opinion there are three problems with published C.R.I. data. The first is that C.R.I.’s are mean indices, averages of how a lamp renders/distorts the color of all the test papers. This is fine and dandy, but the corollary is that a lamp can receive a fairly high mean rating even though it may do a miserable job of rendering the color of one of the Munsell test papers. Which one? Lighting catalogs are silent on that point. Try writing the manufacturers for that information, you say? Good luck!

The second problem with published C.R.I.’s is that the reference source/illuminant against which the lamp is rated is usually not specified. In theory, lamps having a color temperature of less than 5000° K should be rated against a Planckian radiator of the same color temperature. For temperatures above 5000° K, the s.p.d.’s of C.I.E. Standard Illuminants should serve as the reference. Whether manufacturers actually use these references is an open question, as lighting catalogs are again generally silent on that point.

And finally the third problem: even if a lamp gets a rating of C.R.I. 100 with respect to a defined reference source, this rating is only applicable for the Munsell papers used in that particular test. It is no guarantee that other papers of other colors would be color-rendered as faithfully.

C.R.I. ratings are a powerful sales tool in marketing lamps purchased by the print industry and others for color critical applications. But, in view of the above, readers are cautioned to interpret these ratings with a grain of salt.

The Options Available

We have outlined what we expect of the ideal lamp. We have done this in context of basic illumination theory, whenever possible emphasizing color-rendition aspects. Now it’s time to look at specific types of artificial light sources: the commercial options that are actually available to us.

Fluorescent Lamps

First introduced in 1938, fluorescent lamps use a low-pressure mercury-vapor arc to produce ultraviolet (U.V.) radiation at 253.7 nm. The glass tubes of these lamps are coated with phosphors that absorb the U.V. and convert it to visible wavelengths. The chemical composition of the phosphors determines the wavelengths of the generated light and, hence, the lamp’s color temperature.

Fluorescent lighting is often used in office buildings, due to cost effectiveness. This is because these lamps can produce more lumens per watt (= amount of light per unit of consumed power) than incandescent lamps. For instance, a 16-watt fluorescent tube yields as many lumens per watt as a 60-watt incandescent bulb.

Figures 6 and 7 are the s.p.d. curves for warm white and cool white fluorescent sources. The warm white lamp is rich in red energy, while the cool white lamp has more blue energy. Typically, these curves show both continuous and narrow-band energy emissions. The continuous spectrum is emitted by fluorescent phosphors. The narrow-energy bands are due to the mercury arc, and occur at 404.7 (violet), 435.8 (blue), 546.1 (green) and 578 (yellow).

Figure 6. Spectrum of warm-white fluorescent lamp (after GTE 0-341).

Figure 6. Spectrum of warm-white fluorescent lamp (after GTE 0-341).

Figure 7. Spectrum of cool-white fluorescent lamp (after CTE 0-341).

Figure 7. Spectrum of cool-white fluorescent lamp (after CTE 0-341).

Most of the energy emission of fluorescent lamps is in the form of visible light. As we will see, this is not the case with incandescent sources.

Figure 8 is the s.p.d. of a fluorescent “daylight lamp.” You will note that though it is less weighted toward the red or blue than the warm and cool-white lamps, it still contains the same mercury-arc peaks. These peaks or narrow-energy bands do not occur in natural daylight, which is a continuous curve spectrum of greater relative intensity.

Figure 8. Spectrum of daylight fluorescent lamp (after GTE 0-341).

Figure 8. Spectrum of daylight fluorescent lamp (after GTE 0-341).

Incandescent Lamps

Prototypes of this lamp were devised independently in 1878 by Swan in Great Britain and Edison in America. The modern version consists of a high resistance tungsten-wire filament within a glass bulb containing a vacuum or gas. Lamps over 40 watts are usually filled with a mixture of inert gases, such as argon and nitrogen. Pressure from the gas restrains filament evaporation and allows higher filament temperatures. Lamps of less than 40 watts are normally of the vacuum type.

When voltage is applied to an incandescent lamp, the current passing through the filament meets resistance. Power is consumed which makes the filament glow. Put another way, the consumed power heats the filament to incandescence.

Incandescent lamps are commonly used in many households even though they are not as cost effective as fluorescent lighting and do not render color as adequately as, say, daylight fluorescent lamps. Perhaps the reason for this is that some people are so used to having incandescent lamps in their homes that–in the household at least–any other kind of lamp seems unnatural. If so, this is just another example of the subjectivity that underscores the natural vs. unnatural concepts discussed earlier.

Figure 9. Spectrum of incandescent lamp (after GTE 0-324).

Figure 9. Spectrum of incandescent lamp (after GTE 0-324).

Figure 9 is a typical s.p.d. curve for an incandescent lamp. As you can see, energy emission is mostly in the infrared, with only a small amount in the visible region. As with natural daylight, this curve is strictly a continuous spectrum, without the narrow band emissions of fluorescent lighting. Unlike natural daylight (other than direct sunlight at sunrise and sunset), most of the visible emissions consist of red wavelengths.

Discharge Lamps

Electric-discharge lamps radiate light produced by the passage of current through a gas or vapor. The latter may consist of sodium, neon, argon, mercury, helium, xenon or various gas combinations.

Discharge lamps can be expensive. Many have a relatively short life span. Some require out-of-the-ordinary startup circuitry, careful voltage regulation and cooling apparatus. A few models generate enough U.V. to make the Ozone Layer shiver. Sound pretty negative? Give them a chance and read on, for you may be in for a surprise.

Figure 10. Spectral power distribution in nm of a low-pressure mercury vapor discharge lamp (after Cricks).

Figure 10. Spectral power distribution in nm of a low-pressure mercury vapor discharge lamp (after Cricks).

Low versus high pressure

Discharge lamps are made for special applications, including street lighting, searchlights, signal beacons and film projectors. There are as many kinds of discharge lamps as there are applications. Most can be divided into two broad categories: low- and high-pressure discharge lamps.

All low-pressure and most high-pressure varieties lack a continuous spectrum. For some, the total energy emission consists of narrow bands at certain wavelengths. In other words, the output is limited to one or more regions of the light spectrum. Consequently, the color rendition of this type of lamp is, to put it bluntly, lousy.

The s.p.d. of one kind of low-pressure discharge lamp is shown in Figure 10. Notice that the output largely consists of one line in the yellow, one in the yellow-green and two in the violet. There is an almost total absence of red radiation. As a result the emitted light is bluish-violet, which is the additive mixture of these colors.

This type of discharge lamp was a boon to the London police of the 1 930s. Prostitution was then common in London, even more so than now, and all the friendly hookers wore (no pun intended) plenty of red lipstick and rouge. The red lights of the Red Light District made red makeup look redder than ever! When London’s city administration decided to use discharge lamps for street lighting, the situation climaxed for street walkers, in that these lamps made their red makeup appear brownish gray. A sudden drop in business resulted (Kaufman, 1984). 

Figure 11. Spectral power distribution of an OSRAM short-arc xenon lamp (after OSRAM, C.M.B.H.). Compare with the CIE standard illuminants in Figure 2.

Figure 11. Spectral power distribution of an OSRAM short-arc xenon lamp (after OSRAM, C.M.B.H.). Compare with the CIE standard illuminants in Figure 2.

High-pressure discharge lamps include mercury, sodium and short-arc varieties. The mercury and sodium types render color poorly. However, high-pressure short-arc lamps, particularly certain xenon-filled models, offer excellent color rendition. 

Short-arc lamps are called such because the distance between the electrodes is small (1–10 millimeters). Some xenon models generate an arc color that closely approximates natural daylight at 6000° K. Their spectrum is continuous from the ultra-violet into the infrared. As evident from Figure 11, the continuous spectrum in the visible range is very reminiscent of natural daylight at 6000° Kelvin. One West-German manufacturer, OSRAM, G.M.B.H., recommends a 150-watt version of their xenon short-arc lamp for critical color matching purposes, including spectrophotometric applications. The illuminance of that lamp is strong, much more than a 150-watt incandescent bulb or any fluorescent lamp; in other words, lots of lux!

Figure 12 depicts a short-arc xenon lamp of this type. The rendition is actual size.

So What Light Is Best?

An Ideal Lamp for Dealers and Graders?

Is the xenon short-arc lamp the answer to the dreams of colored-stone dealers and graders? Its spectrum is balanced and similar to that of a slightly-bluish phase of natural daylight. The mercury peaks characteristic of all fluorescent lamps–including so-called daylight fluorescent lamps–do not exist in the spectrum of xenon short-arc lamps. Additionally, these lamps produce an exceptionally strong illuminance reminiscent of natural daylight.

Figure 12. Osram short-arc xenon lamp (after OSRAM. G.M.B.H.)

Figure 12. Osram short-arc xenon lamp (after OSRAM. G.M.B.H.)

In theory, the 6000° K xenon lamp should render the color of gemstones every bit as well as natural daylight of the same color temperature. But does it? Unfortunately, xenon short-arc lamps are not available in Thailand and the author could not obtain one from abroad in time for this article to be printed.


How About a Lamp for Retailers?

Up till now, the discussion has been limited to light sources for colored stone dealers and graders. What about retailers

It is common when retailing colored gemstones to display them in the most complimentary light possible. For example, incandescent (including highly illuminant tungsten-halogen) lamps are often used with ruby displays. As we saw earlier, this type of lighting accentuates red stones. When these lamps are tinted red and/or used in conjunction with purplish-red reflectors, complimentaryis no longer the best word to describe the effect: enhancement is more appropriate. Most retail customers do not realize that the beautiful deep-red ruby they admire under such exaggerated lighting conditions may look comparatively washed-out once they step outside the store.

If retailers wish to display stones in a complimentary light, then they might consider one of the following lighting arrangements:

  1. Incandescent or warm-white fluorescent lamps for red/orange/yellow stones (Figures 13 and 14).
  2. Cool-white or daylight fluorescent lamps for green/blue/violet stones (Figures 14 and 15).

If a retailer wishes to display gems in a more balanced (fairer?) light, then daylight lamps are suggested for all colors of stones. Daylight fluorescent lamps, despite their mercury-induced peaks, have fairly balanced color rendition. They do not enhance red stones. Depending on the lamp-model, they offer little or no enhancement to blue stones. Daylight fluorescent lamps are generally inexpensive. They are marketed by several companies, including Mitsubishi (Daylight 99model), GTE-Sylvania (Design 50 and other models) and Duro-Test Corporation (various models). Toshiba and Philips also manufacture such lamps. The aforementioned OSRAM short-arc xenon lamp would appear as an ideal daylight source.

Figure 13. Photograph of a blue-tinted incandescent bulb alongside a common incandescent lamp. (Photo:  Richard W. Hughes )

Figure 13. Photograph of a blue-tinted incandescent bulb alongside a common incandescent lamp. (Photo: Richard W. Hughes)


Specific Recommendations

The Mitsubishi Daylight 99 fluorescent lamp is used for grading gems in two AIGS classes: the colored stone grading class and the appraisal class. Having taught both classes for the past two years, the author has some definite opinions on the Daylight 99. Its color rendition of colored stones is fairly balanced, though it lacks the degree of illuminance obtained from natural daylight. Moreover, it is not quite long enough to fit into standard sockets–at least, sockets readily obtainable in Thailand (which accommodate every other fluorescent bulb the author has come across). It does, however, fit perfectly in Mitsubishi-made sockets, which sell for CONSIDERABLY more than the standard-length socket arrangement.

The Sylvania Design 50 daylight lamp renders the color of gems very much the same as theDaylight 99. In Thailand, the Daylight 99 retails for over US $100 (including the socket-arrangement); the Design 50 currently sells for about $5 and its standard-length socket/holder is available in most local stores for $2–3. Like the Daylight 99, the Design 50 lacks the strong illuminance of natural daylight

The Osram (XBO-series) short-arc xenon lamp may be the answer to everyone’s prayers, wholesaler, grader and retailer alike. One disadvantage, however, is the price: they are much more expensive than most conventional fluorescent or incandescent (including tungsten-halogen) lamps.

Figure 14, from top to bottom:  Sylvania Royal White, 20 Watt (3000° K)  Sylvania Design 50, 20 Watt (5000° K)  Sylvania Daylight ES Deluxe, 18 Watt (6000° K)  A side-by-side comparison clearly shows the difference in color temperature between these lamps.  (Photo:  Richard W. Hughes )

Figure 14, from top to bottom:

Sylvania Royal White, 20 Watt (3000° K)

Sylvania Design 50, 20 Watt (5000° K)

Sylvania Daylight ES Deluxe, 18 Watt (6000° K)

A side-by-side comparison clearly shows the difference in color temperature between these lamps. 
(Photo: Richard W. Hughes)

Figure 15, from top to bottom:  Sylvania Daylight 1040, 20 Watt (6300° K)  Mitsubishi Daylight 99, 20 Watt (5500° K)  Sylvania Design 50, 20 Watt (5000° K)  Note the wide range in color temperature of daylight-type lamps. (Photo:  Richard W. Hughes )

Figure 15, from top to bottom:

Sylvania Daylight 1040, 20 Watt (6300° K)

Mitsubishi Daylight 99, 20 Watt (5500° K)

Sylvania Design 50, 20 Watt (5000° K)

Note the wide range in color temperature of daylight-type lamps.
(Photo: Richard W. Hughes)

In Summary

Artificial light sources, not unlike the diverse phases of natural light, vary considerably in their properties. As a result, some lamps render an object’s color better than others do.

The most important criterion for assessing the color-rendering ability of any lamp is its spectral power distribution curve. Color temperature and C.R.I. information alone tell us relatively little.

Natural daylight varies too much in strength and spectral composition to be taken seriously as a lighting standard for grading and dealing colored stones. For anything to be a standard, it must be constant in its properties, which natural light is not. Yet, many dealers prefer using natural light, often skylight, when buying and selling stones.

For dealers in particular to make the transition from natural light to an artificial light source, that source must offer:

  1. A degree of illuminance at least as strong as the common phases of natural daylight.
  2. Spectral properties identical or comparable to a phase of natural daylight.

A source combining these two things makes gems appear much the same as when viewed under a given phase of natural light. From the viewpoint of many dealers, this corresponds to a naturalappearance.

Of the many kinds of lamps examined in this article – incandescent, fluorescent and discharge – the 6000° Kelvin xenon short-arc lamp appears closest to meeting the criteria for a standard light source. Besides the strong illuminance this lamp affords, its spectrum is very similar to CIE standard illuminants of similar color temperature. Here it is important to remember that CIE illuminants are in turn derived from actual measurements of natural daylight.

If cost is a major concern, one can hardly go wrong with daylight fluorescent lamps. Though they lack the strong illuminance of natural daylight, and are not faithful to the comparatively smooth-curve spectrums of natural daylight or xenon lamps, they nonetheless render gemstone color fairly adequately. Hence, these lamps may serve as a low-cost substitute.

The lighting needs of most retailers differ from those of dealers and graders. Graders want to examine gems in as objective a light as possible – a light that neither enhances nor detracts from the stone’s color. In contrast, the majority of dealers want to at least buy their stones under like conditions, while retailers tend to favor lighting that is more complimentary to the gems they display. This basic difference in philosophy (and business approach) makes it difficult for one standard light source to ever be adopted by graders, dealers and retailers.

Dealers – perhaps more than anyone else – will appreciate the advantages of buying gems under consistent lighting conditions. Standard lighting minimizes guesswork and hence the dollar risk. At the least, it helps set the guesswork and dollar risk at more predictable levels.

For dealers wishing to conduct evening-business, the xenon short-arc lamp may be the lighting answer for buying and selling. A virtual clone of natural daylight, it is certainly worth exploring for this purpose.


The author wishes to thank the following individuals, companies, organizations and schools for their kind assistance and patience, without which this article could not have been written: Mr. Pakorn Borimasporn, General Manager, GTE International, Thailand; GTE Products Corporation, Danvers, Massachusetts, USA; Professor Dr. Pramoht Unhavaithaya, Faculty of Engineering, Chulalongkorn University, Bangkok, Thailand; Mr. Richard W. Hughes, Director, Asian Institute of Gemological Sciences, Bangkok, Thailand; Mr. Ted Themelis, President, GemLab Co., Florida, USA; The International Commission on Illumination (CIE), Paris, France; The Japan External Trade Organization (JETRO), Bangkok, Thailand.
Additional thanks are due the following: Mr. Apichart Assavapokee of Philips Electrical Co. Of Thailand, Ltd.; Messrs. Niyom Raksasutiphan and Sakonchai Adunyanon of Applied Color Systems, Bangkok, Thailand; Mr. Yongyooth Rattana O Pas of Buntanaphan Electric Co., Ltd., Bangkok, Thailand; Mr. Jan Goodman of Centerstone, Ltd., Bangkok, Thailand; M.R. Diaz, Director International Operations, Duro-Test International Corporation, New Jersey, USA; Toshiba Corporation, Lighting Division, Tokyo, Japan.

Tables I & 2 and Figure 2 were generated by the author. The average lux figures in Table 1 were extrapolated from the IFS Lighting Handbook - Reference Volume (Kaufman, 1984). The composite spectral power distribution curves in Figure 2 were derived from data on CIE Standard Illuminants published by ASTM (1987).The information in Figure 4 was derived from the following sources: ASTM (1987), Allphin (1973), GTE Products Corporation (LA-570R, PL-150 and 0-341) and Driscoll (1978).


  • Allphin, Willard (1973) Primer of Lamps and Lighting, 3rd ed., pp. 38–40.
  • ASTM (1987) ASTM Standards on Color and Appearance Measurement. Philadelphia, PA, pp. 151–160.
  • Cricks, Howard (1951) Illumination. The Focal Press, London and New York, pp. 88–146.
  • Driscoll, Walter G. (editor) (1978) Handbook of Optics. McGraw-Hill, Inc., New York, pp. 3/65–3/66.
  • General Electric Co. (n.d.) Engineering Bulletin TP-119.
  • GTE Products Corporation, Sylvania Lighting Center, Danvers, Massachusetts:
  • (n.d.) Color is How You Light It, LA-570R.
  • (n.d.) Fluorescent Lamps, Engineering Bulletin 0-341.
  • (n.d.) Incandescent Lamps, Engineering Bulletin 0-324.
  • (1988) Large Lamp Catalog, PL-150.
  • (1988) Understanding Lighting, LA-656.
  • (1984) Lighting Handbook, 7th ed., pp. 61–64.
  • GTE-Sylvania S.A. (n.d.) Quick Spec. Geneva, pp. 1/22–1/29.
  • International Commission On Illumination (1970) Daylight: International Recommendations for the Calculation of Natural Daylight. CIE No.16/ E-3.2, Paris, pp. 4–20.
  • International Commission on Illumination (1974) Method Of Measuring and Specifying Color Rendering of Light Sources. CIE No.13.2, Paris, pp. 9–14.
  • Kaufman, John E. (editor) (1984) IES Lighting Handbook – Reference Volume. Illuminating Engineering Society of North America, New York, pp. 5/1–5/31, 7/1–7/31, 8/39–8/57.
  • Kueppers, Harald (1982) The Basic Law of Color Theory. Barron’s Educational Series, Inc., New York, pp. 101–161.
  • OSRAM, G.M.B.H. (n.d.) XBO Short Arc Xenon Lamps, pp. 12, 16–21.
  • White, Harvey E. (1959) Physics: An Exact Science. D. Van Nostrand Co., Inc., New York.

To read Part I of this article, click here