Method
Lighting and viewing positions
HCA is based on a face-up view in jewelry store lighting with illumination on the crown only (as when a diamond is set). This is because a diamonds performance in this position and environment is critical to most consumers when purchasing diamond jewelry. Consumers also rock diamonds from side to side to observe the way the diamond ‘sparkles’ however in my experience diamonds with exceptional face up beauty also look beautiful when rocked from side to side (with the exception of ‘fish-eye’ diamonds but HCA warns if a diamond will show a fish-eye).
While worn in jewelry diamonds are often seen from oblique positions, but diamond owners can be seen closely examine larger single stones in the face up position.
Method of Analysis
Much of the development of HCA was done using observations of DiamCalc ‘virtual’, Firescope™ and Gilbertsonscope images, as well as other MSU website based virtual diamond tools. Comparing the effects of variations in diamond proportions using virtual diamond analysis eliminates problems with variation in color, clarity and the minor facet groups, all of which are encountered in studies using real diamonds. Diamcalc ray path analysis was also used to gain an understanding of many seemingly inexplicable anomalies that were encountered. This ‘theoretical’ study has been supported by examination of thousands, if not millions of diamonds. During the course of 2001 I have been rejoined by Mr. Christian Pike FGAA Dip DT who has contributed and collaborated impartially and objectively in a continual review and fine tuning process.
Virtual diamond studies were all based on the DiamCalc settings for:
Images were created for 2197 photoreal virtual diamonds, these were then collated in 13 charts for table sizes between 53% and 65% in a range of commonly encountered diamond proportions with pavilion angles 39.5° to 43° in 0.5° steps and crown angles between 28° and 40° in 1.0° steps. The series of images in the GIA review is typical. The minor facets for all comparisons were set at 50% for upper girdle / star facet ratio’s and 82% lower girdle facet lengths.
Brilliance was assessed by comparing DiamCalc Firescope and Gilbertsonscope images within each table sized grid in columns of crown angle and rows of pavilion angle. Each was subjectively rated and each virtual stone compared to the others with that table size. Stones with the least leakage and the brightest images were given a score of zero, and the worst a score of four. The author then compared images with the same values from different areas on the grid and adjusted scores for consistency where necessary.
This comparison was done for each table size grid from 53% to 65%. A mathematical formula (discussed later) was devised as a penalty factor for the deleterious effect of larger and small tables. This was then added to the primary score. For instance a diamond with a 65% table has an additional penalty of 1 compared to a diamond with a table size of 60%.
Fire was assessed using three methods and rated from 0 to 2. Firstly diamonds that actually show fire in the face-up DiamCalc virtual images were given a low score. Secondly diamonds that have substantial amounts of head shadowing were given a low score. Finally images that displayed darkness resulting from light leakage or the unusual extinction previously mentioned were given moderate values. The methods used in this analysis were similar to that used in the brilliance method. The results for each stone in each table size grid are also adjusted arbitrarily for very large or very small table sizes.
After these results for fire had been tabulated MSU released a fire/scintillation option as it is described on the Cutstudy.com website. This tool was used to check the fire results. The tool employs an outline of a virtual diamond with a ray of light directed perpendicular to the table, individual dispersed rays of emergent light are displayed on a virtual screen placed between stone and viewer. Using the same grid analysis system I found a strong visual correlation with the previous results.
Values for scintillation were estimated by comparing the DiamCalc images on all 13 grids. Images with a well formed dark star were given the top score of zero, those with little or no black were given a moderate score of 1. If darkness was poorly distributed then a larger penalty of up to 2 was given.
The face up view by itself cannot display dynamic scintillation – the contrast of snap-snap or sparkling effect that we associate with scintillation. However using the movie mode on DiamCalc it was possible to check that the dynamic effects of many of the proportions, especially those that rate as excellent.
The method of scintillation analysis for various table sizes was similar to that described for fire and brilliance. It to was based on DiamCalc comparisons between stones with the same crown and pavilion angles. Table sizes were compared in steps of 1%.
The lighting conditions used excluded back illumination (except where this was a precondition of the virtual instrument). Comparisons of actual diamonds were made in fluorescent tube flood lit office type environments, daylight filtered through blinds and at a desk in a jewelry store with several 50 watt halogen spot lights mounted in a 12 foot high ceiling.
After plotting each function it was immediately obvious that a diagonal slope confirmed the theory of an inverse crown and pavilion relationship. Pleasing scintillation patterns were also noticed that followed these patterns on the photo grids.