GIA Brilliance Study Lacks Fire

After eight years of study, the Gemological Institute of America has issued the first of a series of reports on diamond cut (see accompanying article). The study offers some fascinating mathematical data. But with only one-third of the project complete, the report makes some hasty judgments on cut-grading systems already in place. GIA’s initial report literally lacks fire (discussion of spectral colors, or dispersion) and sparkle (discussion of brilliance and fire in motion, or scintillation).

Using mathematical models and computer ray tracing, GIA’s researchers have unleashed an incredible mass of statistics on cutting for brilliance, measured in terms of “weighted light return” (WLR). The results prove that there are many ways to proportion a diamond for maximum brightness, all of them different from Marcel Tolkowsky’s original “Ideal cut.” As a diamond grader, I would agree, since I’ve seen many bright diamonds not cut to Tolkowsky’s measurements.

The researchers show that the highest values for brilliance occur when a diamond is fashioned with very shallow crown angles (less than 29º), small tables (less than 57%), and pavilion angles ranging from 38º to 42º. As crown angles rise, the highest-rated WLRs disappear. Diamonds with crown angles of 34.5º (Tolkowsky’s measurement), tables of 53% to 58%, and pavilion angles of 39.7º to 41.6º are less brilliant than the highest WLRs but are still considered quite bright. However, with crown angles any greater than 34.5º, diamonds will not show a high weighted light return no matter what the table size or pavilion angle.

Surprise! Diamonds with crown angles of 23º (very shallow by today’s standards) return the most light of any stones with crown angles of 10º or higher. Going one step further, GIA’s researchers found that, “Ironically, the highest weighted light return values are obtained for a round brilliant with no crown at all.” I could have predicted that without an eight-year study.

Can cutters use this to make diamonds look better? Not yet. You’re not going to be seeing round brilliants cut “crownless.” In short, brightest is not always best.

What about dispersion? Diamonds with shallow crown angles may be more brilliant, but there is less fire with shallow crowns. Theoretically, the greatest dispersion occurs when light leaves the gem at or close to the gem’s critical angle. Tolkowsky’s two-dimensional mathematics concluded that 34.5º would give the maximum dispersion for light escaping through the crown facets.

GIA’s next step in determining whether a diamond is cut to “Ideal” proportions is to calculate the crown angle that yields the most dispersion through the crown. It is dispersion weighted against brilliance on which Tolkowsky based his elementary results.

So stop while you’re ahead. The GIA report should have ended its discussion after determining brilliance, but it did not. In their closing remarks, the researchers say their results disagree with AGA, AGS, AGT, HRD, and other diamond-proportion grading systems. They say their results do not support the idea that all deviations from a narrow range of crown angles and table sizes should be given lower grades. Some of their WLR ratings actually increase as the grades in these systems worsen.

But didn’t they forget something? Their study was limited to brilliance. The proportions that affect brilliance can also affect dispersion and scintillation. Just what that effect is remains unknown. We can’t know how the established grading systems relate to a standard proportions grading system because there are no results yet to design such a system.

The researchers know this: “Although arguments can be made for down-grading diamonds with lower crown angles or larger tables on the basis, for example, that they do not yield enough fire, there is little documented evidence at present to support – or refute – such claims.” So why does GIA assume that its brilliance study refutes this claim?

I will not stick my neck out and say that Marcel Tolkowsky was completely right. But I will stand up and shout that it is way too soon to determine that he was wrong. If, as GIA has discovered, brilliance is affected in ways that could not have been predicted, why not consider that dispersion and scintillation could do the same?

Give GIA credit for amassing a lot of information in its analysis of light traveling through a diamond. The results of this study are intriguing but only academic at this point. Determining dispersion, scintillation, and how all of these criteria interrelate will resolve the issue of cut and truly benefit the industry. Drawing conclusions at this time solely from brightness measurements is jumping the gun. Before Tolkowsky’s “Ideal” and other cut grading systems are shot full of holes, the rest of the research must be considered. Only then, when you have all the ammunition, can you fire at will.