Professional Jeweler Archive: Study Moves GIA Closer to Cut Grade

January 2002


Study Moves GIA Closer to Cut Grade

Second part of cut study shows no single set of proportions best maximizes fire

The second part of the Gemological Institute of America’s study on diamond cut is complete. “With the results of these two major segments [brilliance and fire] ... now published, we are very close to answering the key questions involved in proportion evaluation for round brilliants,” GIA President William E. Boyajian writes in the introduction. GIA has long stated that, absent a better understanding of the effects of cut and polish, it would refrain from establishing a cut grade on its Gem Trade Lab diamond reports. However, that may change. “As more results are obtained from this project, ... a single cut grading system may become feasible,” says Boyajian.

Results of the brilliance and fire studies will be adapted and used in GTL reports during 2002, says Boyajian. GIA will modify its teachings about diamonds beginning with its all-new Graduate Gemologist program in 2002 and may develop cut-evaluation instruments also.

Labs across the U.S. began offering cut grades on their reports a few years ago. The American Gem Society lab was first to do so and now is funding a study on the effects of fancy cuts on a diamond’s appearance. Other organizations offer instrumentation they say helps retailers and manufacturers buy and sell a round diamond based on brilliance, fire and scintillation.

All this research has set the stage for a debate about who is right or which system is best. Boyajian makes it clear he believes GIA’s scientific study of cut will settle many arguments. “While we laud the efforts taken in past decades to uncover the secrets of diamond proportions and fine cutting, we know that assumptions were made that now can either be verified or corrected through modern scientific means.”

What Role Does Fire Play?

In the new study, “Modeling the Appearance of the Round Brilliant Cut Diamond: An Analysis About Fire and More About Brilliance,” GIA now makes a distinction between fire and dispersion, terms once used interchangeably. The study says the distinction is needed because the dispersion value of diamonds is constant, while fire is understood to be a result of dispersion. The new definition of fire is “the visible extent of light dispersed into spectral colors.”

The authors developed a new standard of measurement for fire: dispersed colored light return. To measure DCLR, researchers used the same computer-modeled virtual diamond as in Part 1 of the study. The computer model was used to study values of 26,000 combinations of pavilion angles, table sizes, culet sizes and girdle thicknesses. Results from these combinations were then compared with the performance of real diamonds under similar conditions. Twenty-eight diamonds were acquired for the study, and an additional 400 diamonds submitted to GIA-GTL were observed also.

No Single Best

Much as they discovered in the brilliance study, researchers found that no one set of proportions in a diamond best maximizes fire. In fact, they found a wide ranging set of proportions yielded similar results.

Researchers also described an inverse relationship between the tests. Generally speaking, the set of proportions that yielded better weighted light return (brilliance), tended to yield poorer DCLR (fire), and vice-versa. But the researchers also found a happy medium where numerous diamond proportions yielded average or above-average results for WLR and DCLR.

The standard reference by which GIA measured changes was a virtual diamond with a 34&Mac251; crown angle, 40.5&Mac251; pavilion angle, 56% table, 50% star facet length, 75% lower girdle facet length, girdle thickness of 3% and culet size of 0.5%. DCLR computations varied all these angles and percentages separately.

To compute the actual diamond tests, GIA researchers used a white hemisphere 16 inches in diameter with a small hole (less than half an inch) at the center. The diamond was mounted on the open side of the hemisphere, with the table perpendicular to the center of the hole. Single-source direct light at the opposite side created chromatic flares of fire to beam into the hemisphere to be photographed (for a graphic illustration, see Professional Jeweler, July 2001, p. 26, or the GIA study illustration).

Other Results

One outcome of the study is the understanding of the interdependence of a diamond’s angles and percentages. Changes in the parameters of one tend to affect the others. Here are a few of the findings:

  • DCLR in general increases as crown angle increases.
  • DCLR in general decreases as the pavilion angle increases.
  • DCLR in general decreases as table percentages increase.
  • DCLR is highest for a 34&Mac251; crown angle combined with a star facet length of 64%-65%.
  • DCLR showed dramatic increases as lower girdle facets increased in length 45%-85%.
  • DCLR showed few changes as girdle thickness or culet sizes were changed.

Researchers determined many proportion combinations yielded above-average DCLR values, while others clearly showed lower values. One surprising finding: Lengthening lower girdle facets in relation to the lengths of the standard reference stone yielded greater DCLR. GIA points out that previous studies and most cut grading systems ignore the dramatic effects observed when lengthening or shortening these facets.

GIA says results of both studies affirm its belief that more than one set of cut parameters will yield above-average brilliance and fire. Diamond cutters and manufacturers eventually may use the findings to fine-tune their cutting process and achieve greater yields.

– by Robert Weldon, G.G.

“Modeling the Appearance of the Round Brilliant Cut Diamond: An Analysis About Fire and More About Brilliance” was written by Ilene M. Reinitz, Mary L. Johnson, T. Scott Hemphill, Al M. Gilbertson, Ron H. Geurts, James E. Shigley and technical writer Barak D. Green. It appeared in Gems & Gemology, Fall 2001. To order a single copy call Gems & Gemology, Carlsbad, CA; (800) 421-7250, ext. 7142. (Part I of the study was published in Fall 1998.)

Copyright © 2002 by Bond Communications