Wed, 05 Jul 2006

Making the case for Imagem diamond grading

one looks for solutions to problems it is best to look for the simplest
possible one. There may be more than one legitimate way to solve a
problem or to accomplish a task, but the simplest method that does the
job correctly is the one to ultimately choose. It is not wrong to try
various methods, but when it comes time to adopt a standard approach to
universally solve a certain problem or task, then the simplest method
that provides the best results or equivalent results must be preferable.

have often said that one would always use a scale to obtain the exact
weight of a diamond in preference to measuring the length, width, and
depth and then using the GIA formula for the particular shape to
estimate the weight. It is good to understand both methods, as one does
not always have a scale or an unset diamond, but when the diamond is
not set and a scale is available, no one should use the more complex,
less accurate, method.

In the past 15 years or so, I have
suggested to many people that the AGA Cut Class system that I developed
for screening the best cut round and fancy shaped diamonds was as good
a method, or better, for judging diamond cut quality. It was totally
based on parametric measures of the diamond and grounded in the
Tolkowsky-like model that jewelers had grown to accept as the basis for
best cut. Through my experience dealing with many diamonds over the
past 39 years, I felt confident that diamonds within the best cut range
would nearly always stand up as well cut and be quite pretty. Now
craftsmanship is one thing and pretty must be considered another. We
can judge craftsmanship based on several objectively measured
characteristics of specific diamonds. However, “pretty” or beautiful
remains somewhat subjective and will always be part of what each
customer uses to select one diamond from many.

One must remember
certain given characteristics of faceted diamonds. We are not talking
about randomly made items, but something that is cut in very
traditional ways and highly repeated and developed styles. Diamond
rough is costly. Cutters must consider several things with every single
piece of rough. They must consider retained weight, the market that
stone will likely be offered into, the crystal characteristics of each
diamond, and the unique placement of inclusions. In addition, they must
have an eye for getting the subjective beauty to show. Believe me, I
have owned a few unusually cut diamonds over the years where I saw
rarity of the cut as desirable, but few others felt the same. These
were very difficult to sell. Therefore, diamonds are cut within quite
narrow parameters of symmetry, facet style, and appearance. This is a
given. You just do not see wide variation in cutting and this will
serve to provide us good assistance in getting a rather simple system
to work for grading the Light Behavior and Craftsmanship Grade of a

ImaGem uses three measures for judging the Light Behavior of a diamond.

These are: Brilliance, Sparkle and Intensity. The definitions are provided below:

a measure of a stone’s overall strength of light return that represents
its average light return in the face-up position. This measure is
arrived at by computing the mean gray-scale value of all pixels within
the girdle portion of a diamond.

Sparkle: a
measure of those spangle-like flashes of reflected and refracted light
that gives diamonds what might be called “life” or “kick”. The greater
the number of these flashes, the higher a stone’s sparkle. This measure
is arrived at by measuring the standard deviation in the gray-scale
value of the light return within the girdle image.

a measure of the number and strength of contrasting light-dark areas in
the girdle portion of a diamond that give it vitality and character.
The greater the stone’s symmetry the higher its intensity will be. This
measure is arrived at by calculating a ratio of bright pixels to the
total number of pixels within the girdle.

Because of
the rather consistent manner in which diamonds are cut, these three
rather simple yet scientific measures serve to well define Light
Behavior in diamonds and can readily act as metrics to delineate the
quality of Light Behavior beyond the ability of even expert human eyes.
Grading diamonds to levels beyond the ability of human perception to
discriminate is the accepted norm of the traditional diamond business.
Few people can tell the difference between one or two color grades or
clarity grades. Experts can do this far better than amateurs, but our
data shows that agreement on color grading among experts is only 65%
under the best conditions in proper laboratory environments. This is
why labs differ in their grading of color and nothing but an
advancement of technology stands to cure the human inability to color
grade to a higher degree of consistency. While we have not yet studied
the clarity issue so closely, our supposition is that human grading is
somewhat inconsistent there, as well. Technology will advance to make
these grades far more reliable.

The measure of Brilliance is
based on the fact that the camera sensor used in ImaGem reads 256
values from total black to total white. In between are shades of gray.
The higher the number, the less black there is with 0 representing
black and 255 representing white. The brilliance reading is the AVERAGE
of all the pixels within the face-up view of the girdle outline, Better
diamonds do return more light, but a diamond hat returned enough light
o show only white pixels would have no beauty or merit. It would be
very bright, but not sparkly or show contrast. The best round diamonds
show 200 to 150 gray scale average values. The best princess cuts show
200 to 130 gray scale average values.

One should keep in mind
that the lighting inside the VeriGem is not truly unusual, but the
situation is highly controlled, not like jewelry store lighting, and
loaded with computerized functionality. We don’t want to see any large
number of pixels reading at 255, so controlled light intensity is used
in order to discriminate all levels of light return present in each

The measure of Sparkle is based on the range of light to
dark pixels present in the diamond. When we measure average light
return there are much brighter and much darker pixels present. Sparkle
is simply a report of how diverse, how far apart, the high and low gray
scale readings of the pixels is within the diamond. For example, 5 is
the average of all of these: 555 – 456 – 357 or 258. The standard
deviation of 555 is 0, for 456 the standard deviation is 0.8, for 357
it is 1.78 and for 258 it is 2.64. Now these are just simple examples
based on a mathematical formula, but I hope you can see the point and
gain a basic understanding of the measure of Sparkle.

The measure
of Intensity also has some connection to the diversity of gray scale
values in each diamond measured. All faceted diamonds, even rather
poorly cut ones, have some degree of variation is gray scale readings
among all their pixels. Obviously, the best performing diamonds would
have a higher number of bright pixels. What ImaGem does is calculate
the brightest range of pixels and reports it as a percentage of the
total number of pixels. When you see an ImaGem Intensity grade of 200,
all you do is move the decimal 1 space to the left and you have 20% as
the percentage reading of the brightest pixels compared to the total.
It is a measure of contrast.

The three measures tend to rise as
diamonds become higher in light return, but patented and un-branded
cuts of great beauty reside well within the norms of high performance.
Not all are at the ultimate limits. You may wonder why that is. Beauty
is a very important factor when it comes to the personal selection of a
diamond for purchase. Beauty is also judged by human cutters. While we
can say that the highest ranges of Light Behavior assure us of
attractive diamonds, we cannot say that everyone would select a diamond
only based on numerical scores. Few, if any, fancy shaped diamonds can
outperform or match the Light Behavior scores of round diamonds, yet
many people prefer them. This is our human nature at work. We don’t
want to change that, but simply provide meaningful data to assure the
trade and consumers that what they are buying is of the quality they
wish to own and can afford. ImaGem is a highly discreet screening
device for most people. We fully acknowledge that for those deeply into
numbers, such as engineers, high performance numbers may be key to
their decision making process. For most other people, the exact numbers
will play a lesser role. For this reason, labs that are using or
planning to use ImaGem may or may not choose to report the numbers/
Labs may opt for word grades derived from these numerical outputs.
Words will work well for most consumers and the trade. They may even be
superior to numerical grading as we won’t assert that higher numbers
alone make a prettier diamond than somewhat lesser numbers within the
same word category. Just like with color and clarity, ImaGem can grade
what the eyes cannot detect. This is the typical grading strategy for
diamonds and we offer the same with Light Behavior.

Behavior grading does a wonderful job of screening the appearance of
diamonds, but there is more to judging a diamond’s suitability for use.
AGA and ImaGem offer the DFS system which is an optional cut
craftsmanship measuring system. DFS stands for Durability, Finish and
Size. Again, these measures are simple, and readily understood by all.
We continue to depend on the consistent nature of how all diamonds are
cut. Of course, the rare exceptional diamond may exist somewhere, but
that does nothing to disprove the reality of how nearly every other
diamond is fashioned.

The Durability measure looks at culet size,
crown angle, and girdle thickness. The Finish measure looks to symmetry
and polish. Finally the Size measure looks at the relationship of
spread, diameter, to depth. Remember, diamonds have natural
characteristics which inherently affect their cutting, structure and
long lasting character. The DFS system is the simplest, most elegant
way, to measure the quality craftsmanship, “look” and safety of the
stone over the long run. It is the final measure of how certainly one
may know that a particular diamond is among the best qualities.

everyone will need, want or can afford the very best diamond. Just like
in other measures of diamond for color and clarity, the best are
invisibly better to the naked eye. Only under high tech equipment can
one discern the nuance differences. This gives consumers far more
choices about how to balance their desire to buy something very nice
with how to get a larger diamond that might otherwise not be in their
price range. More information may seem confusing at first, but there is
good reason to understand the rather simple nature of the information
that ImaGem is providing. It is rocket science applied to diamond
grading, but made simple by ten years of dedicated work to also make it
commercially viable.

The reason this all works is partly due to
good science used in the development of ImaGem products and partly to
the inherent and consistent nature of diamonds. Diamond cutters shape
diamonds into very consistent objects. Right now, we have studied
rounds, princess cuts and marquise shapes. ImaGem only needs a range of
100 diamonds of any other shape in order to also take Light Behavior
measures on them. We see most of the others who hope to grade diamonds
struggle to produce any measures on stones except rounds. Using
predictive tools, which we admit are useful to cutters, as measuring
devices is na?ve, unscientific, and inaccurate. Just like weighing a
diamond, one always uses a scale in a grading, laboratory environment
and never uses the dimensions and a standard formula.

Will the
industry choose the proper fork in the road that has now appeared in
their path to the future? Will informed consumers insist on ImaGem
grading versus a forecast grade done without regard to the exact
situation inside their diamond? What of cloudiness, clarity placements,
physical measuring inaccuracy, and not measuring certain features at
all? We think we know what is best. Some will hold other viewpoints
mostly because they own competing equipment which has been quite
costly. All gemologists who have invested in the future of diamond
grading own high cost equipment that has or will become obsolete. This
is just a cost of doing business and treating consumers right. We urge
all interested parties to learn more about ImaGem, Light Behavior and
the DFS system. ImaGem’s VeriGem equipment will be available for hands
on, private sessions at the upcoming 2006 NAJA Summer conference and at
the San Diego 2006 GIA Symposium.

by David S. Atlas, GG(GIA) Sr Mbr(NAJA) ASG(AGA)

D. Atlas & Co.