Diamond Color Grading • The Crying Game

1 April 1999
By Richard Hughes
Diamond Color Grading • The Crying Game

A critical examination of the methods used in gemological laboratories for the color grading of diamonds.

 Diamond Color Grading • The Crying Game

Hey there. How many of you have experienced this? You color grade a diamond with your in-house master set and it seems to you to be a solid F. Needing a document to flog the rock, you then send it off to your main laboratory ice house and it comes back one grade different. Kinda makes you wanna cry, doesn't it? If this sounds familiar, it should – arguments on diamond color grades occur with the frequency of a Bill Clinton bimbo outbreak.


Before the sobbing starts, let's look at why this is so. Probably the most important reason for color grade variations is simply the narrow size of individual grades. Let me tell you, these cubbyholes are tighter than Tom Jones' jeans, with differences being literally at the edge of human discernment. While under optimum conditions graders can generally get within a quarter grade of one another, on lab documents the best reproducibility graders can probably attain is plus or minus one full grade. Yep. One full grade. Thus today's F could be either tomorrow's righteous E or disastrous G.

Probably the most important reason for color grade variations is simply the narrow size of individual grades. Let me tell you, these cubbyholes are tighter than Tom Jones' jeans…

Why is this so? One reason is because no major laboratory currently lists whether a stone is in the high or low part of a particular grade's range. Thus if a stone lies near a grade boundary, a quarter grade real difference can translate into a full grade printed difference on a report. For example, a high G is simply printed as a straight G. When it is regraded and becomes a low F, again it appears to the client as a pure F, even though the two internal grades are virtually identical.

Master blaster

Take the master stones themselves. Being creations of nature, no two stones are alike, and these differences are amplified by differences in cutting. De Beers continually preaches this, but all too often we ignore it. As each diamond is different, it is impossible for anyone to produce a master set totally identical to the original Gemological Institute of America-Gem Trade Laboratory (GIA-GTL) set, including GTL. And since GTL has more than one master set (reports suggest upwards of five sets in each of their two labs), variations will result. Some labs/dealers even use cubic zirconia masters, creating further problems.


Guess what happens when graders swap out individual master stones to improve their sets? This produces subtle shifts in the grade boundaries and can result in different grades when the same stone is regraded at a later time.


Just how does GTL grade master sets anyway? They use something called a colorimetry scale, assigning a number to each master. This allows graders to better understand the exact position of each master stone relative to the master GTL master set. The colorimetry of a perfect master set would advance in 0.5 increments, where E is 0.5, F is 1.0, G is 1.5, etc. But not every master set is perfect. For example, a perfect F would be 1.0, but in their system an F master could be anywhere between 0.85 and 1.15. Thus if your F master is 1.15, you could have a stone that is a high F (say 1.10) that you would still grade as an E when compared to your F master.

To top it off, GTL policy allows such colorimetry to be given only to American Gem Society (AGS) members. This leaves everyone else in the trade, including the vast majority of diamond wholesalers, jewelers and labs with one extremely chilly willie.


By definition, masters are supposed to lie only in the Cape series (yellow series, Type 1a), but again, nothing in nature is 100% perfect. So while GTL will reject stones with obvious color contaminants, subtle contaminants (such as browns, grays or greens) might still get through. Due to Australian production, many stones mined today contain a significant brown component. Stones in this brown family are particularly difficult to grade against Cape series yellow masters.


Not only do masters vary, but also the eyes of graders. Even among so-called color normal individuals, variations exist. Not all eyes show the same sensitivity to colors like yellow or brown. Furthermore, some people's moods show as many ups and downs as the presidential staff. These emotional swings, along with little molehills such as the yellowing of our corneas with age, the time of day we grade, the number of hours grading in a day, etc. can become insurmountable mountains when subtle color differences need to be discerned.


Enjoying this? We're just getting started. Diamonds attract grease like a dirty politician, and feathers and bruted girdles provide perfect places for dirt to lodge. Bruted girdles also pick up metal from tweezers and since master stones get handled so often, this is a real problem. Simple cleaning with a cloth will not remove it – only acid boiling will do. Experienced labs generally polish or facet the girdles of their masters, but many traders do not.

So what happens when a stone with a dirty girdle is graded? If the grader does not examine it carefully in all directions (and inexperienced graders may not do this), a mistake may be made. Directional inclusions such as color zoning and graining may have a similar effect and that's something not even the white tornado can remove.

Enjoying this? We're just getting started. Diamonds attract grease like a dirty politician…


Even graders working in the same lab with the same master set often do not standardize the viewing conditions. Some graders prefer to hold the stones within an inch or two of the light source, while others hold the stones down at the bottom of the box, more than doubling the distance from the lamp. As anyone familiar with the inverse square law knows, distance is extremely important. Doubling the distance of the stone from the lamp decreases the amount of light by a factor of four. Tripling the distance decreases the amount of light striking the stone by a factor of nine. While this may not change the color of a stone with weak fluorescence, holding a stone close to the tubes could have a dramatic effect on those with stronger blue fluorescence, masking the yellow body color.

Comparing grades between different labs can occasionally resemble an ecclesiastic housewarming party. BYOF: Bring Your Own Faith.

It gets uglier still. The viewing background is not always consistent; certain labs grade against white plastic and others against cardboard cards, some of which actually fluoresce blue. Each lab believes their methods to be sound, but since labs do not necessarily use the same conditions, comparing grades between different labs can occasionally resemble an ecclesiastic housewarming party. BYOF: Bring Your Own Faith.


Light sources themselves are a can of worms some gemologists and many dealers would just as soon not think about. While most graders use the GIA GEM Instruments DiamondLite, how many regularly change the bulbs? The spectral output of any fluorescent tube changes with time, which means the same stone may appear different at different points in time. Furthermore, the bulbs used for diamond grading all have a UV output, and this may vary from one bulb manufacturer to another.

As an added complication, the GIA GEM Instruments DiamondLite comes with two daylight bulbs, each independently controlled. As demonstrated to the author by Mike Scott and Edward Boehm of White Rose Enterprises, the light output with just one tube turned on (as opposed to both at once) reveals a significant change in color temperature, in addition to a change in the quantity of light striking the stone. The readings, using a Gossen Color Master Color-Pro 3F meter at 3000 lux sensitivity, are shown in Table 1. 

Table 1: Color temperature readings for the GIA GEM DiamondLite
Condition Color Temperature (±50°K)
Rear tube only 6450°K
Front tube only 6280°K
Both tubes together 6410°K
Both tubes together reflected off bottom of DiamondLite 6130°K

By now, having been reduced to blubbering heaps, you're probably reaching for that old chestnut, instrumental color grading. What about getting a machine to grade the stones, right? Nice try. File that along with cold fusion and Uri Geller's mental spoon bending. It doesn't work. Ever since B.W. Anderson first suggested machine color grading of diamonds, it has remained the gemological holy grail. And to date, every instrumental color grader has had problems, insurmountable problems. This should give you some clue: reports suggest that the GIA, which sells a diamond colorimeter through their GIA GEM Instruments subsidiary, does not regularly use the instrument in their GTL.

Unfortunately, color differences between machine and eye are far too common. There are a number of reasons for this. Perhaps the biggest problem is that the machine does not "see" the stone's color in the same way as the human eye. Two stones that are visibly different to the eye may produce the same grade on the machine, while those that are identical to the eye may fall into different categories with machine grading. Since it is the human eye that is the standard, a machine eye that doesn't closely match the human equivalent is useless.

Tears fall down

Since I am not really the sadistic bastard I seem, I won't go into the other problems faced by diamond color grading systems. It is not my intention to scare readers, but rather to help those who use diamond grading reports better understand the limitations of the present techniques.

Qualified graders working under optimum conditions in the same lab with the same master set can generally attain reproducibility to within one quarter to one half grade. However, as you should now understand, such conditions are not always found today, either in dealers offices or grading labs. While many of the above problems can be minimized, none can be entirely eliminated. In combination, the result is that, even at top labs such as GTL, diamond color grades may not be reproducible to better than plus-or-minus one full grade, let alone under the conditions found in most dealers' offices. That's reality. Sad to say, but we've got to get used to it. Until the industry is ready to consider modifications to the existing grading methods and system, all we can expect is more of the same, lots more tears. Now you can start crying.

R S end dingbat



The author thanks Thomas E. Tashey, Andrea McShane, Elizabeth P. Quinn and Craig Slavens for their careful reading of the manuscript and helpful suggestions.

Author's Afterword

One more stab at the subject of diamond grading, written after nearly a year in the laboratory diamond-grading trenches. It first appeared in the GQ Eye in 1999 (Vol. 2, No. 1, pp. 3–5).

See also The Naked Eye: A Diamond's Worst Friend

Back to top