The Japanese invention is simply a 10x lens with a disc of hot pink card in front of the diamond. By viewing a diamond through a hole in the pink disc you can observe the amount of pink light returned from the front of the stone.  The Firescopetm is a qualitative brilliance gauge even a child can use to assess the amount of white light leaking from behind the stone. The black lens area approximates the darkness cast by an observer in a normal viewing position. The most brilliant diamonds look bright pink/red with some black and minimal white light leaking in the through the pavilion. Some ideal cuts have small 'V' shaped white leakage features at the girdle. The Firescopetm was invented as a brilliance assessment device, but also offers a qualitative assessment of symmetry. A range of proportions cut with perfect symmetry display a black eight-pointed star, but a star does not by itself indicate Ideal proportions. A pleasing distribution of black gives an observer a nice display of scintillation as the stone is rocked from side to side. The Firescopetm helps assess scintillation.  Laboratory Gemologists measure 'features' to define symmetry, but this results in an artificial weighting. The Firescopetm is a 'benefits' system and shows the actual effect of symmetry defects, in a simple and more reliable way. It may be possible to use the Firescopetm in labs to grade symmetry.  There is a design flaw in the Firescopetm. White light is seen through the Firescopetm that has entered via the crown at shallow angles of incidence. This is illustrated in the image of the 40°P - 36°C stone, which has a white 8-pointed star around the star facets. These appear pale blue through Mr. Gilbertson's device, indicating the light came from above at a shallow or oblique angle. The same is true of the pale patches in the very center of the 40.5P row.  The name of this instrument is poorly chosen because it does not measure Fire (dispersion). What I learned with the Firescopetm From looking at hundreds of thousands of diamonds I learned the range of proportions that work best. I noticed: - Diamonds with shallower crowns and deeper pavilions (than Tolkowsky's) are bright red, i.e. they return more light.  I call these Brilliant Ideal Cuts or BIC Stones with steeper crowns and shallower pavilions appear slightly less red but exhibit more fire (dispersion) and scintillation. I call these Firey Ideal Cuts or FIC Sophisticated clients tend to prefer the 'romantic' look of FIC's even though they display slightly less brilliance and have less spread. They appear to have more facets, which makes larger stones appear to have more scintillation. Small BIC's look good in pave settings.  Diamonds with a combination of both steep crowns and deep pavilions leak the most light from behind. Even a single degree steeper and deeper crown and pavilion than Tolkowsky's' leak through the Firescope.  When you grade clarity you block the light from the front of the stone with the loupe, so light leaking in from behind makes these stones appear brighter, while ideal cuts appear dark and dull. I have tested diamond merchants around the world and most prefer the sparkling ring of leaked light just inside the table and the bright outer crown facets of steep/deep diamonds. The sparkle has leaked through the pavilion from their desk lamp! These stones also have better yield or weight retention, so it is no surprise that most diamonds are cut steep/deep. Note - A .10ct diamond of the same proportions as a 2ct stone looks different through a Firescope. This is because of the change of angles and relationships between the firescope and various sized diamonds. The arrangement of red and black patterns change, but leakage of white light from behind is unaltered.     Hearts and Arrows The Firescope spurned a lesser instrument for displaying H&A, the phenomena that became very popular in Japan during the late 80's. I have heard that 80% of engagement rings sold in Japan featured H&A's. The arrows are the same as appear in super symmetrical diamonds through the Firescope, but are white. The hearts are observed by looking at the back of an unset stone, and should be 8 well shaped hearts with Vee shaped arrowheads pointing to the culet. The source of colored light is from a tube around the stone.  The problem with this instrument is it places a white ring around its (lower power) lens. The additional white light source from above removes the value of this toy as a measure of brilliance because you cannot detect light leakage. The device is a simple but effective symmetry tool. But many beautiful diamonds have poor symmetry. H&A are now big in the USA. Here are some of my observations: -   The best are said to have AGS 0 ratings Not all AGS 0 stones display H&A Not all AGS 0's look as good as most H&A There are beautiful H&A's well outside the AGS 0 range I have seen H&A I grade Very Good, outside my ideal range I have modeled many combinations of proportions that display H&A, valuable information to help H&A cutters improve yields. Mr. Al Gilbertson's variation on a firescope is a sophisticated multi coloured firescope. The different colors indicate the direction light has come from before entering the diamond. The device also works well when the stone is tilted.

 
 
 

Diamonds of proportions in the upper left return light supplied from a larger area and explaining why they are more brilliant (Brilliant Ideal Cuts - BIC). More blue indicates light has come from peripheral sources and shows diamonds of these proportions gather light from a full hemisphere. The GIA's lighting model (section 6) favored BIC's.  The images in the lower right have more green red and black colored light returned to the viewer, light from directly above. Had the GIA used normal lighting, with more light coming from above then diamonds with proportions in the lower right (Firey Ideal Cuts -FIC) would have achieved higher light returns (WLR). But FIC's are more affected by the relative size and 'blackness' of the observer, which is why they look very dark through a loupe. Almost all the light being returned perpendicular to the table comes from just 70° of a 180° hemisphere. (Considering this destroys Dr Marty Haske's published arguments about flaws in the GIA study).  5. Rapnet In the early 1990's I bought certified diamonds for my business Precious Metals, on Rapnet's B2B service.  From Australia I could not see the stones so I attempted to quantify my Firescope observations with some simplistic 2D-ray path calculations. Because most rays of light pass once through the crown and reflect twice off pavilion facets I simply settled on a 2:1 relationship between crown and pavilion angles. I have since learned there are more complex factors and relationships but I have saved many clients a lot of money on the premiums for Tolkowsky's proportions. I bought some of my ideals from Martin Rapaport at the 1998 JCK Orlando. He knew I had a system and asked me to write an article for Rapaport Diamond Report, but at the time this was a trade secret. (I have since published an article in Rap. July 2000) After reading the GIA study I decided to publicise the 2:1 principle (I have since adjusted the ratio, but the concept is still valid).  6. Gems and Gemology Study of Brilliance The GIA published an article in the Fall of 1998 based on computer modeled 'virtual diamonds'. The paper drew no conclusions but supported my 2:1 principle between crown and pavilion angles. They were unaware of the significance of their own data.  This study encouraged me to announce my findings in a letter published in Rapaport Diamond Report, May 1999. Using the GIA's data I explained why some of the American Gem Society lower cut grades look better than steepest/deepest AGS 0's. I described the 'single target' flaw in all cut-grading systems.

The white line on the graph represents the highest light return from a range of crown and pavilion angles in Figure 11; page 178 G&G Fall 1998 (Hemphill, Reinitz, Johnson and Shigley). Note the similarity to my dark blue 2:1 line. The upper right corner of the blue box are poorly performing steepest / deepest permitted AGS '0' crown and pavilion angles.  The red line is my proposed intuitive minimum standard, which has been closely confirmed by the MSU studies (discussed below).  I proposed future cut grading systems be concentric ellipses with long axes on this graph.

 

The 'Brilliance' authors saw my graph at the 1999 GIA Symposium in San Diego. They neither refute or agree with the concept and have not been cooperative. Gems and Gemology does not publish commentary papers and have not accepted an offer of a letter to the Editor.  The Australian Gemmologist published a Letter to the Editor in July - September 1999 and a lengthy review of the GIA study in Volume 20 number 8, order at http://www.austgem.gil.com.au    7. American Gem Society Cut Grading In 1996 the AGS added cut grading to its certification service, making it a runaway success. The service is so popular that De Beers Millennium diamonds (all GIA Certified) are being submitted for AGS cut Certs, highlighting the markets cry for the GIA to grade cut on its certificates.  Tolkowsky never designed his proportions for the convenient commercial tolerances applied by the AGS and other labs. AGS 0's and 1's have Crown angles between 33.7° - 35.8° and pavilion angles within the range of 40.16° - 41.2° (tables range from 52.4% - 57.5%). While this seems like a very tight range, stones with the 35.8° crowns and 41.2° pavilion combinations look nothing like Tolkowsky stones. Check this out on the collages at the start.  AGS 0's must have fine symmetry and polish; minor problems rate a lower AGS'1' grade. Yet Peter Yantzer, the director of the AGS lab, tells me he cannot see any difference in beauty as a result of minor polish or symmetry defects.  Features: benefits Cut grading systems use unrelated features to arrive at a grading standard to describe diamond beauty. Sales people who use 'features' to sell - like many watchmakers - never achieve the sales of those who sell 'benefits'. Consumers expect laboratory certificates to define the quality of a diamond. Quality includes 'benefits' such as beauty and rarity of a diamond. Somewhere along the way we lost the plot.  The AGS encourages cutters to send in half polished stones to be sure of 'getting the paper' to sell a stone at a premium. Most sawn stones that 'get the paper' are cut for weight retention i.e. 'steep/deep' near C 35.8° and P 41.2°. The GIA study of 67,621 stones reported that of the 2051 rare stones that qualified for a top cut grade "nearly twice as many had pavilion angles of 41.0° - 41.3°, rather than 40.2°-40.9°". I call on the AGS to share its data or acknowledge the fact that many of its '0's are far from ideal.  Peter Yantzer has known of my findings since he met me at the 1999 GIA Symposium and was one of the first to see the collage of my ideal range. He asked me to be patient while they address the problem. Meanwhile retailers and consumers are paying top prices for so called ideals. A steep/deep1.00ct AGS 0 loses .02ct in re-cutting it to an ideal cut. (6.44 x 3.99mm - remove .03mm from its pavilion to 40.95° and lowering the crown angle to 34.8° with a table of 55.8%.) A .98ct is worth 10% to 40% less depending on its color and clarity. The AGS has embarked on a major study of fancy shapes to establish a grading standard. In my opinion they should be fixing up the round system first and gaining a greater conceptual understanding of basic principles.  Peter and other AGS and GIA people were panelists in an ideal cut forum on GemKey.com from November 1999 to January 2000. There we 'met' two Russian scientists who share many of my views. 8. Moscow State University In November 1999 I discovered the MSU website hosted by Sergey Sivoyolenko and Yurii Shelemetiev www.gemology.ru. These learned gentlemen have the most advanced virtual computer modeling of diamond cut.  They use a variety of lighting environments (the weakness of the GIA study). They give results for both brilliance and dispersion in an easy to use factor; 1 = Tolkowsky performance. Sergey and Yurii examined a wide range of proportions and found several zones of maxima. Tolkowsky's proportions fit within one of these zones. They agree with me - crown and pavilion angles are interrelated.

This chart shows the MSU light return for diamonds with table size of 53%.  The lighting is diffused daylight of the 'office' type. The value 1.00 is based on the light return of a Towlkowsky proportioned virtual diamond, rocked through 30 degrees of tilting. Note the higher values occur at shallow/deep proportion combinations.www.gemology.ru:8111/cut/english/document4_1.htm

This chart presents the MSU results for fire of diamonds with table sizes of 53%. The value 1.00 is based on the fire of a Towlkowsky proportioned virtual diamond as it would appear when viewed in a variety of light environments from 'office' to 'chandelier', rocked through a range of 30 degrees of tilt. Note steep/shallow combinations have more dispersion. www.gemology.ru:8111/cut/english/document4_1.htm

The MSU cut quality function Q. The first 2 charts show high light return does not coincide with maximum fire. Since both characteristics are paramount, the authors proposed a special function 'Q', taking into account the cut parameters of both light return and fire.

These charts are for 53% tables. Sergey and Yurii found 60% tables do not perform as well. Charts for 60% table sizes are posted at the website. Table size is a hot commercial issue in the way diamonds are cut. The GIA found only 76 stones with 53% tables among the 2051 close to Tolkowsky proportions (from a total sample of 67,621).

 
 
 

AGS 4 Shallow/Deep Brilliance    1.05 Fire    1.00 Crown = 32°  Pavillion = 41°  Table = 53%	AGS 0 Shallow Pavillion Brilliance    1.00 Fire    1.04  Crown = 34.75°  Pavillion = 40.4°  Table = 53%	AGS 0  Steep/Deep Brilliance    0.92 Fire    0.96 Crown = 35.8°  Pavillion = 41.25°  Table = 53%	AGS 2  Steep/Shallow Brilliance    0.99 Fire    1.05 Crown = 36°  Pavillion = 40°  Table = 53%

 
 

.96ct Most Brilliance BIC	.98ct Close to TIC	1.00ct  Very Poor Brilliance	.99ct Most Fine FIC

 
 

These are 4 MSU 'virtual diamonds' shown in different lighting environments. For comparisons Tolkowsky's Ideal is given a value of 1.00 for brilliance and fire. There are many more beautiful diamonds than most so-called ideal cuts. Note the weight differences. You can generate your own virtual diamonds like these with free software from the MSU site. They also sell more sophisticated software for less than $200.

 
 

They found some AGS 4's had the best proportion combinations, yet these are four grades worse than AGS 0. The MSU results support my theory of the relationship of crowns and pavilions and the BIC / FIC concept. But as you can see the results Sergey and Yurii plotted form complex graphs.  The function Q on the third graph rates fire and brilliance at equal importance. A range of maxima occur around 31°-36° crown angles and 41.3°-40° pavilion angles, a ratio of around 4:1 between crown and pavilion angles. Sergey and Yurii are now studying how our eyes and minds perceive the beauty of a diamond. The diamond that catches your eye from the other side of a dimly lit restaurant has a sparkle of fire larger than the facet reflecting it, or even bigger than the whole diamond! Why is this so? Can instruments measure this, or is it a purely human perception? There are many complex issues to be resolved.     9. What went wrong with Cut Grades? The GIA brilliance study summarized scientific research into ideal cuts over the past 80 years. In the 19 studies pavilion angles range from 36.5° to 47°, crowns from 25.5° to 48.6° and tables between 40% and 69%. None linked pavilion and crown angles. How did they miss this obvious relationship? Firstly the studies were 'features' driven. Secondly unlike Tolkowsky most failed to 'observe' first. Let me tell a little story:

A man arrived late at a horse auction, the last horse had been sold. He asked the Auctioneer if any were passed in and was shown one for just $2,000.  The buyer was impressed with the horse and its pedigree. "Why was it passed in?" he asked. The Auctioneer replied "It didn't look good". After paying $3,000, he took the horse to his farm and watched as it ran into the only big tree in the field and killed itself! He phoned the Auctioneer who replied, "Told ya it didn't look good!"

The morale - it pays to look good! Da Vinci, Newton and Einstein were 'good lookers', so too was Tolkowsky.   10. Direct Assessment Measuring performance directly is very appealing, especially for use as a retail and marketing aid. But there are some questions about assuming an instrument "sees" and experiences what we see. Measurements are made with a spectrascope at different lighting positions. At the June JCK show I discussed comparing the results and proportions of stones analyzed by the Gemex Brilliance Scope and Diamond Profiles Lab with my ideal cut range.  The Brilliance Scope measures brilliance by analyzing the overall stone area in each lighting angle. It records the percentage of the stone that returns white light. If light is not returned then the assumption is there is light leakage.  The relative quantity of colored pixels counted is said to define dispersion, recorded as the percentage of the stone that is colored. The actual spectral color, position and direction of display are not considered.  Scintillation is a measure of the relative intensity of the colored and white light returned.  I will examine databases from Diamond Profiles and Gemex and see if the spectroscope results match my range of ideals. A digital report will appeal to some customers and I believe we must explore every method to improve the way diamonds are cut. I suspect there are flaws in the concept and assumptions that could result in improvements in the technique. For instance when measuring dispersion allowance must be made for low colored stones, they will appear more dispersive. Presumably the system will not work with fancy coloured diamonds.  If direct assessment does work, it will be the way the best way to grade fancy shapes. The extra variables of fancies make them so complex to grade and we do not yet have a grading standard that works for rounds! The AGS are keen to get in early and establish a standard so they can get more business. Dave Atlas has done a lot of excellent work in this area too, but I think he too agrees we have a long way to go. The AGS are doing a survey to gather many peoples preferences, but as my Russian friends say, "Taking the average temperature in a doctor's waiting room, including dead people and those with a fever, tells you little".  Direct assessment has a place and may be the best way to grade fancies, but for cutters a spread sheet of ideal proportions is better.     11. Conclusion My vision of a website forum for diamond-cut is closer after the trip to JCK Vegas. We need a venue to share and co-ordinate each other's knowledge and experience. We seem to be in a silly winner-takes-it-all "space race". If we work together we can design a global grading system. I dream of the GIA, AGS, MSU and other individuals and companies putting aside their own business interests to create a global standard. The Internet makes this possible. Together we can help cutters improve the beauty of diamonds and make them more attractive to consumers. I have been defining my own range of ideal proportions for 15 years. While many of my peers are purists and focus on the "perfect diamond" I want to raise the standard and beauty of all diamonds. Cutters must maximize dollar returns, but we can help cutters achieve good-looking stones and commercial yields. Here are 2 ways:

1. They say cutters won't change, but retailers and consumers will drive the change. Cutters produce what people value and pay for. Yields drop a lot with smaller tables and shallower angles, but consumers happily pay 10% more for VVS than VS, even though they cant tell the difference. We must educate consumers the way De Beers did with Clarity, Colour and Carat weight. We must put more value on the "look" of a diamond. A 1.00ct steep/deep AGS 0 spreads only 6.44 mm and looks dull. It looks worse still when the customer gets it dirty. By contrast stones with shallower facet combinations handle dirt much better. One trick I use with customers is the Blu Tacktm test (a product for adhering posters to painted walls). Press stones that are being compared into a large piece of Blu Tacktm and instantly they have dirty pavilions, raising the RI at the interface. Then you can compare the stones, as they will appear in a real environment.   2. As I write this in July 2000 I am on my way home from Bombay where I met the manager of a large factory whose Argyle Champagne goods I buy. I can only buy around 1% of their stones - the mistakes. His processes are deliberate and very consistent. He told me a few years ago he reduced tables from 67% to 65% and is aiming for 62%. They produce 41.6° pavilions and 30° crowns, proportions that fall within my ideal range. This is Indian cutting and shows how responsive they are becoming to customer demands! His understanding of both economics and cut will take the business where the market goes. I do not support the large table size, but at 62% and with some combinations of half-length facets, the stones look good.

My charts are for a superior grade not a super or '0' grade. Individual labs could define their own super grades as a way of differentiating their businesses. The AGS 0 - 10 scale has become a de-facto standard, and it makes sense to use the numeric scale (I hope the AGS takes this as a compliment). A minimum standard for promotional stones might be 9 and 10 and could be added later if deemed necessary. Once these boundaries are established then we just 'fill in the blanks' 2 - 8. Value A universal (and effective) grading system for cut will change the way diamonds are valued, beauty will be as prized as rarity. Many diamonds will fall in value, and one in ten will rise. Martin Rapaport will find a way to get the 4th C onto a price sheet. A re-cutting industry will flourish. Cut will be priced like Colour, with a large premium for the very best proportions. The rarity factor between IF and SI will probably continue its 20-year slide. Cutters will use aids like Sarin to design and mark a range of proportions programmed onto computers.  Buying a good looking diamond off the net will be easier and 'Flipping Certs' inevitably is the way most large diamonds will be sold. But as layers of distribution (and jobs) are removed, consumption will rise as people find that a big diamond is more affordable. Conflicting cut grade information confuses consumers. The same applied before the GIA's D-Z Color grades were introduced (blue-white diamonds?).  By removing a barrier that scares consumers diamonds will be safer to buy. Cutters will earn more because the market will value more beautiful diamonds. Diamonds will regain ground from the branded luxury goods. Diamonds are the ultimate possession - forever. I welcome your comments, post them here or e-mail me.

Copyright

Garry Holloway FGAA, DipDT

Please join in or discussions about Diamonds here