FOUR CUT-GRADING RULES OF THUMB, REVISITED: TILT, HEAD SHADOW, ROOF RIDGES, AND HORIZON LIGHT

Between 1998 and 2005, two institutions wrote the modern rulebook for how the jewelry world grades a faceted gemstone.

In three landmark papers published in Gems & Gemology, researchers at the Gemological Institute of America (Hemphill et al., 1998; Reinitz et al., 2001; Moses et al., 2004) built a computer model of the standard 57-facet round brilliant diamond, work that became the foundation of the GIA Cut Grading System launched in 2006. In 2005, the American Gem Society Laboratories launched their own performance-based cut grade built around the Angular Spectrum Evaluation Tool, or ASET (analyzed in the optical literature by Sasian et al., 2007): a color-coded map, viewed face-up, that shows where in the surrounding hemisphere each part of a gemstone draws its light. Steep light from near the viewer's own head (75 to 90 degrees above the horizon) is painted Blue, bright overhead light (45 to 75 degrees) is painted Red, and lower light from the walls and horizon (0 to 45 degrees) is painted Green.

Those studies were genuine milestones over Marcel Tolkowsky's 1919 two-dimensional analysis (Diamond Design, London). Those systems were built for a specific job. GIA's cut grade applies only to standard round brilliant diamonds, and both frameworks were calibrated on that geometry: the 8-fold symmetric, 57-facet round brilliant, evaluated primarily face-up. Within that scope they work well. The trouble starts when shorthand versions of their lessons travel into retailer guides, forums, and cutting folklore, and get applied to fancy shapes, odd symmetries, and colored stones they were never designed for. Four of those rules of thumb are especially common:

1. Rule of Thumb One (Face-Up Tells the Whole Story): That a gemstone's light performance is best judged at dead-center 0-degree viewing, so a stone that looks dark face-up is simply a poor performer.
2. Rule of Thumb Two (Contrast Needs Head Shadow): That crisp dark-and-bright contrast comes from viewer head shadow (a healthy share of Royal Blue on a face-up ASET map, often quoted in retailer and forum guidance, not in GIA or AGS standards, at around 15% to 20% for well-cut rounds) blocking overhead light and painting dark arrows across the stone.
3. Rule of Thumb Three (More Sparkle Needs More Facets): That richer virtual-facet patterns require grinding more physical facets onto the crystal (80 to 100+ facets), the way many branded high-facet-count diamond cuts do while keeping 8-fold even symmetry.
4. Rule of Thumb Four (Green Means Weak Light): That ASET Green (light drawn from 0 to 45 degrees above the horizon) is weak, second-class illumination that should be minimized in favor of ASET Red overhead light.

When you step outside the 57-facet round diamond and trace full three-dimensional skew rays across 92,862 distinct gemstone geometries, each simulated at 19 tilt positions (from minus 45 to plus 45 degrees in 5-degree steps) and spanning 8 refractive indices (from Quartz at RI 1.54 to Moissanite at RI 2.67), every one of these rules of thumb turns out to have important exceptions outside the round diamond it came from.

Here are four physical mechanisms behind those exceptions, illustrated with ray-path diagrams and eight real gemstones from the OffTheDop Showcase and Signature Collections that you can inspect in live 3D right in your browser.

One caveat up front: every number in this article comes from our own GPU simulation and our own catalog-normalized metrics (Brightness, Contrast, Scintillation, Shannon Entropy). GIA validated its metrics against panels of human observers; ours have not yet been tested that way. Treat what follows as strong physical evidence for exceptions, not as a replacement grading system.

1. Rule Of Thumb One: "face-up (0-degree) Brilliance Tells The Whole Story"#

In the foundational 1998 GIA brilliance study, Hemphill, Reinitz, Johnson, and Shigley defined their primary brilliance metric (weighted light return, or WLR) for a stone viewed face-up. GIA's later observation trials did include rocking stones by hand (Moses et al., 2004), but the computed metrics that anchor the grade are face-up quantities. For a well-cut round brilliant that is a reasonable shortcut, because its light return is designed to peak near face-up.

That shortcut holds only if a stone is tuned to peak at zero degrees. What if a crystal is engineered like a set of Venetian blinds?

Look at the 19-angle brightness curves and optical maps for Pomegranate Serene Cantata (a 97-facet Cubic Zirconia curved octagon at RI 2.15) and Subzero Aegean Soliton (an 80-facet Cubic Zirconia round at RI 2.15):

Follow the numbers in the chart above:

• Parked in Shadow at 0 Degrees: If you clamp Pomegranate Serene Cantata into a stationary vertical grading scope at 0 degrees, 40.2% of its face maps to ASET Blue, and Subzero Aegean Soliton parks 57.0% of its face in ASET Blue. At dead-flat zero-degree viewing, large banks of their virtual facets look straight back at the observer's head and camera lens, holding face-up raw brightness down to 0.274 and 0.465. A static face-up meter would mark both cuts down.
• The Tilt Surge: Now pick either stone up and rock it by 15 to 20 degrees, the range a ring or pendant routinely sweeps through as it moves on a hand. Tilting changes which part of the room each virtual facet looks at: the facets that were staring into your own head shadow now look past it, into the bright overhead light. Pomegranate Serene Cantata stays essentially window-free through that whole range (0.00% windowing out to 15 degrees, 0.06% at 20 degrees). Subzero Aegean Soliton gives up a little light to windowing as it tilts (0.6% face-up, rising to about 5% at 20 degrees), but gains far more by escaping the head shadow.
• Nearly Double the Light Return in Motion: Rather than peaking face-up, Pomegranate Serene Cantata nearly doubles in brightness (+99%, from 0.274 at 0 degrees to 0.545 at 20 degrees) and peaks at 25 degrees (0.556), earning a catalog-normalized Brightness score of 1.092. Subzero Aegean Soliton surges by +35% to peak at 0.627 at 15 degrees tilt, driving both its overall Brightness (1.131) and Scintillation (1.131) more than 13% above the catalog benchmark.

A gemstone is never worn bolted to an optical bench at zero degrees. Evaluating brilliance only at dead-center viewing confuses a stone that is parked in neutral with a stone that cannot run.

2. Rule Of Thumb Two: "contrast Requires Obscuration (head Shadow / Aset Blue)"#

Both GIA's cut research and the AGS ASET system treat the observer as part of the lighting: the head and upper body of the person looking at a stone block light arriving from near the vertical, and that obscuration helps create the dark-and-light pattern of a well-cut round (most famously the eight dark arrows of a Hearts and Arrows diamond). On an ASET map, this steep head-shadow light (75 to 90 degrees above the horizon) is colored Royal Blue. The trade rule of thumb that follows is that a stone needs a healthy share of ASET Blue to show strong contrast, and that a stone with almost no ASET Blue will look "washed out" or "flat white."

In an earlier study (Making Contrast Computable) across 4,799 stones, face-up ASET Blue turned out to be almost useless as a predictor of real 19-tilt spatial contrast, with only a weak trend in the wrong direction. Here is the physical reason: at least two other optical mechanisms can generate strong contrast with less than 2% ASET Blue.

Look at the side-by-side optical maps and ray diagrams for Pellucid Aria Heptad (a 50-facet Moissanite heptagon at RI 2.67) and Specular Valhalla Spinel (a 54-facet Spinel mirrored trilliant at RI 1.72):

Examine the ASET maps on the left of the diagram above:

• Almost Zero Royal Blue Ink: Pellucid Aria Heptad contains just 1.71% ASET Blue (58.0% Red, 25.8% Green, and 14.5% Black leakage). Specular Valhalla Spinel contains just 2.03% ASET Blue (59.3% Red, 34.0% Green, 4.7% Black). By the head-shadow rule of thumb, both stones should look like flat, low-contrast sheets of glare.
• Yet Both Score Above 1.03 in Spatial Contrast: In reality, Pellucid Aria Heptad achieves a 1.049 Contrast score (though its Brightness, 0.905, sits below the benchmark because of that leakage), and Specular Valhalla Spinel achieves 1.030 Contrast alongside a strong 1.130 Brightness.

How do they build dark-and-bright spatial checkerboards without borrowing darkness from the viewer's head at zero degrees?

1. Mechanism A (Zenith-vs-Horizon Elevation Steps): Even when no ray hits the viewer's head, the lighting around a gemstone is not a uniform white sphere. Overhead studio and room light (ASET Red, 45 to 75 degrees elevation) is markedly brighter than low-angle wall and horizon light (ASET Green, 0 to 45 degrees elevation). When neighboring virtual facets alternate between roughly 59% Red and 26% to 34% Green across crisp facet ridges, every boundary becomes a luminance step that does not depend on the viewer's head at all.
2. Mechanism B (Dynamic Tilt Occlusion): A stone that shows almost no head shadow at 0 degrees does not stay that way when tilted by 10, 20, or 30 degrees. As you rock Pellucid Aria Heptad or Specular Valhalla Spinel across the 19 tilt positions, their Red facets take turns sweeping into and out of the head-shadow zone, creating moving dark-and-bright contrast waves without parking a large Blue area in the face-up view.

Head shadow is only one way to draw a dark line inside a crystal, and it is a costly one: every percent of static face-up ASET Blue is light that the viewer's own head is blocking.

3. Rule Of Thumb Three: "virtual Facet Sparkle Requires High Facet Counts"#

When diamond manufacturers wanted finer, more intricate virtual-facet patterns than a 57-facet round brilliant provides, the usual path was brute force: grind more physical facets onto the stone. Branded cuts jumped to 73, 81, 88, or 105 facets, almost always keeping even-numbered 8-fold rotational symmetry.

There is a cheaper lever, and it comes from elementary mirror geometry: what a reflected beam meets on the far side of the pavilion.

Compare what happens to a light beam inside an even-symmetry pavilion (8-fold) versus an odd-symmetry pavilion (3-fold or 7-fold) in the diagram below:

Look at the two geometric cross-sections on the left:

• Even Symmetry (4-, 8-, 12-, or 16-fold): A Matching Facet Opposite: In a standard 8-fold round, every pavilion main at azimuth theta sits directly across the center from an identical main at theta + 180 degrees. A sheet of light reflected off one main and sent across the belly of the stone lands mostly on that single opposite facet. Round brilliants still split beams at the edges between neighboring facets (that is where the Hearts and Arrows pattern comes from), but not systematically at the opposite main.
• Odd Symmetry (3-, 5-, 7-, or 9-fold): A Ridge Opposite Every Main: In a 3-fold trilliant or a 7-fold heptagon, there is no main facet at theta + 180 degrees. Directly opposite the center of every main pavilion facet sits the ridge where two neighboring mains meet.
• The Roof-Ridge Beam Splitter: That opposite ridge behaves like a roof. The half of the beam that lands on one slope is sent toward one part of the room, and the half that lands on the other slope toward a different part. The familiar comparison is the Amici roof prism inside binoculars and camera viewfinders, with one key difference: a binocular roof is ground to exactly 90 degrees so that the two halves rejoin into a single image, while a gem's pavilion ridge is nowhere near 90 degrees, so the halves part ways and each becomes its own virtual facet.
• Compounding Splits: The split sub-beams then cross further facet edges on their way out through the crown, and each crossing can split them again, so the number of distinct virtual facets grows faster than the number of physical ones.

Look at the right side of the diagram above to see what this looks like in real crystal:

• Glacial Cerulean Belvedere (Diamond, RI 2.42, 7-fold Heptagon) has only 43 physical facets in total (about 25% fewer than a standard 57-facet round brilliant). Yet its raytrace breaks into a dense mosaic of small reflections, reaching a 1.014 Shannon Entropy score (32-bin tonal information richness) and 1.072 Scintillation.
• Galaxial Cobalt Undertow (Cubic Zirconia, RI 2.15, 3-fold Curved Trilliant) has only 40 physical facets in total. Because its 3-fold main tier places a ridge opposite every main, those 40 facets reach an above-benchmark 1.011 Shannon Entropy score, with near-benchmark Scintillation (0.997) and Brightness (0.975).

Two stones are not proof that odd symmetry always wins, but they show that facet count is not the only route to a rich virtual-facet pattern.

You do not need to spend hours on the faceting bench grinding 100 tiny facets to get intricate sparkle. Choosing odd symmetry turns the pavilion ridges themselves into free beam splitters.

4. Rule Of Thumb Four: "horizon Light (aset Green) Is Weak Secondary Light"#

In the AGS ASET scheme, light entering a gemstone from 45 to 75 degrees above the horizon is colored Red and described as the most intense, while light from 0 to 45 degrees is colored Green and described as lower-intensity light from walls and surroundings. Graders and buyers therefore prefer Red-dominant maps, and a diamond showing a sea of green is read as a weaker performer.

That hierarchy makes sense in Diamond (RI 2.42), where the critical angle is a tight 24.4 degrees. A ray coming straight down through the table of a round brilliant reflects off one pavilion main and then off the opposite one, turning back up by roughly 180 degrees. For a pavilion angle a, those two bounces strike at a and at 180 minus 3a degrees, and both stay in total internal reflection for any pavilion angle from about 24.5 to 51.8 degrees. That wide corridor leaves plenty of room to aim the returning light at the bright overhead Red zone.

Now apply that diamond recipe to Quartz (amethyst, citrine, smoky quartz, rock crystal) at RI 1.54, where the critical angle grows to 40.49 degrees.

Follow the traced rays on the left of the diagram above:

• A Narrow Corridor: In Quartz, both bounces must exceed 40.49 degrees, which squeezes the safe pavilion range down to about 40.5 to 46.5 degrees. That is why traditional Quartz designs cluster around 42 to 43 degree pavilions.
• That Corridor Points at Your Own Head: Trace the path backward from the eye and a problem appears. Inside the corridor, light leaving the table after two bounces comes from 65 to 90 degrees above the horizon; at a typical 43-degree pavilion it comes from about 78 degrees, squarely in the ASET Blue head-shadow zone. In our 2D trace of a classic Quartz brilliant (43-degree pavilion, 40-degree crown), the table area maps to head shadow and the outer crown ring leaks straight through the pavilion. In Quartz, the diamond recipe does not deliver diamond-style Red; it delivers Blue and Black.
• Green Is the Way Out: The best-performing Quartz designs in our catalog escape this trap with facet layouts a classic brilliant does not use: many tiers at varied angles that route light along three-dimensional skew paths, keeping every bounce above 40.49 degrees. The light those paths return comes from lower in the room, between 0 and 45 degrees elevation, and on an ASET map that light is Green. In Quartz, a large Green share is not a sign of weak light handling; it marks light that would otherwise have been head shadow or leakage.

Look at the ASET maps of Hearth Saffron Sunwheel (a 79-facet Quartz curved hexagon at RI 1.54) and Nebular Navy Beacon Pentagon (a 71-facet Quartz pentagon at RI 1.54) on the right side of the diagram above:

• Hearth Saffron Sunwheel is 73.53% ASET Green across its face (18.1% Red, 8.4% Blue, and 0.00% Black). By diamond standards, a 74% green map looks like a weak performer. In Quartz, it comes with 0.915 Brightness, 0.907 Contrast, and no face-up windowing at all.
• Nebular Navy Beacon Pentagon combines 56.27% ASET Green with 30.93% ASET Red, shows no face-up windowing in our tilt simulation (and just 2.9% Black on its ASET map), and reaches a 0.953 Shannon Entropy score in humble rock crystal.
• The Price Shows Up on Tilt: By the lesson of Rule of Thumb One, these face-up numbers are only half the story. Both stones window more and more as they tilt: Hearth Saffron Sunwheel goes from zero face-up to about 9% at 15 degrees and 17% at 20 degrees, and Nebular Navy Beacon Pentagon reaches about 15% at 15 degrees and 20% at 20 degrees. Their Scintillation scores are also low (around 0.18). Green-dominant Quartz is a strong face-up strategy, not a free lunch.

In Diamond, ASET Green is a compromise. In Quartz and Beryl, where the diamond recipe returns mostly head shadow and leakage, ASET Green is how a stone gets light back to the eye at all. Judge it by what it replaces, and always check it across tilt.

5. References And Primary Literature#

1. Hemphill, T. S., Reinitz, I. M., Johnson, M. L., and Shigley, J. E. (1998). "Modeling the Appearance of the Round Brilliant Cut Diamond: An Analysis of Brilliance." Gems & Gemology, Vol. 34, No. 3, pp. 158 to 183.
2. Reinitz, I. M., Johnson, M. L., Hemphill, T. S., Gilbertson, A. M., Geurts, R. H., Green, B. D., and Shigley, J. E. (2001). "Modeling the Appearance of the Round Brilliant Cut Diamond: An Analysis of Fire, and More About Brilliance." Gems & Gemology, Vol. 37, No. 3, pp. 174 to 197.
3. Moses, T. M., Johnson, M. L., Green, B., Blodgett, T., Cino, K., Geurts, R. H., Gilbertson, A. M., Hemphill, T. S., King, J. M., Kornylak, L., Reinitz, I. M., and Shigley, J. E. (2004). "A Foundation for Grading the Overall Cut Quality of Round Brilliant Cut Diamonds." Gems & Gemology, Vol. 40, No. 3, pp. 202 to 228.
4. Sasian, J., Quick, J., Sheffield, J., Caudill, J., and Yantzer, P. (2007). "Evaluation of Brilliance, Fire, and Scintillation in Round Brilliant Gemstones." Optical Engineering, Vol. 46, No. 9, 093604.
5. Tolkowsky, M. (1919). Diamond Design: A Study of the Reflection and Refraction of Light in a Diamond. E. & F. N. Spon, London.
6. Smith, W. J. (2008). Modern Optical Engineering (4th edition, prism and mirror systems, including the Amici roof prism). McGraw-Hill, New York.