Facet Architecture and Proportions
If anatomy tells us what exists, architecture tells us how everything is connected. Facets are not independent parts. Changing one group alters the space available to another and affects the profile, face-up pattern, weight distribution, and the way the stone manages light.
No serious analysis of cut can therefore be reduced to one percentage, one angle, or one table of “ideal” numbers.
[VISUAL 18.1: The standard round brilliant as an interconnected network—table, crown mains, stars, upper halves, pavilion mains, and lower halves]
The Facet as a Building Block
Every facet has:
- a size;
- a shape;
- an orientation;
- an angle;
- a position;
- relationships with neighboring facets.
Architecture arises only from their combined arrangement. Two facets of the same shape can have completely different functions if they are located on different parts of the stone or set at different angles.
Four Broad Organizational Approaches
For diamonds, it is useful to distinguish among:
- brilliant—a facet network that creates many dynamic reflections;
- step—rows of elongated facets that produce broader, step-like returns;
- mixed—a combination of different principles;
- modified brilliant—a brilliant arrangement altered with additional or differently divided facets.
These are architectural categories. They are not quality grades in themselves.
[VISUAL 18.2: Brilliant, step, mixed, and modified brilliant—a schematic comparison of facet networks]
The Round Brilliant as a System of Interdependence
The standard round brilliant is particularly useful for learning because its facet organization is highly regular. But that regularity can create the false impression that reading a few numbers is sufficient.
In reality, the following interact:
- table size;
- crown angle and crown height;
- pavilion angle and pavilion depth;
- star length;
- lower-half length;
- girdle thickness;
- culet;
- the precision of individual facets;
- differences among nominally identical facets.
Two stones can have the same average pavilion angle but a different set of individual angles. Two stones can have the same table percentage but different crown heights and completely different overall geometry.
Table and Crown
A larger table facet reduces the area available to the crown facets if the other elements remain approximately similar. A smaller table facet may allow a higher or more pronounced crown architecture, but the final result depends on angles and other dimensions.
Statements such as “a 57% table is ideal” therefore have no professional meaning without context. It is necessary to know the crown and pavilion angles, girdle, lower-half and star relationships, and the grading system within which that value is being considered.
Crown Angle and Crown Height
The angle of a crown main affects crown geometry, but crown height is not merely another name for that angle. Table size and girdle construction also contribute to the final height.
Thus, two profiles with the same crown angle can have different crown heights if other parameters differ.
Pavilion Angle and Pavilion Depth
The pavilion is a particularly sensitive part of the architecture because small changes in angle can significantly alter the path of light. Yet here, too, one average is not the whole story.
Pavilion angle and depth are geometrically related, but the reported value may be an average of several facets and rounded according to laboratory convention. A real stone may vary around its entire circumference.
Crown and Pavilion Must Be Read Together
One of the most important lessons of modern cut analysis is that crown and pavilion angles must not be read in isolation.
A particular pavilion geometry may function differently with a shallower crown than with a steeper one. A change in table size further changes the arrangement of the crown facets. The limits of a good cut therefore cannot be reduced to two independent columns labeled “good crown angle” and “good pavilion angle.”
[VISUAL 18.3: Matrix of several crown/pavilion combinations—the same individual value, but a different result when its partner changes]
Star Facets
Star facets connect the area around the table facet to the upper halves. Their relative length changes the size and shape of the virtual areas an observer sees face-up.
Shorter or longer star facets are not inherently “better.” Their effect depends on the rest of the architecture.
Lower Halves
Lower halves extend from the girdle area toward the culet alongside the pavilion mains. Their length strongly affects the appearance of the pavilion pattern and the size of the virtual facets.
In general, longer lower halves can create a finer, narrower pattern, while shorter ones create broader elements. But this is not a linear formula in which “longer = more brilliance” or “shorter = more fire.” The optical consequence depends on lighting, other proportions, and the observer.
The Girdle as a Structural Compromise
The girdle must survive manufacturing, handling, and setting, but it also carries weight. Areas that are too thin can increase the risk of damage, while a very thick girdle may retain weight without producing a proportional increase in face-up spread.
In a round brilliant, girdle variation also enters the design and grading logic. However, a single word such as “medium” does not describe every local cross section. Metrology and laboratory reporting are covered in Chapter 20.
The Culet and Lower Termination
The culet is a small element that can strongly affect the face-up center if it is large enough to be visible. In historic cuts, a more open culet is part of the authentic visual language; in a modern round brilliant, a very small or practically point-like termination is usually expected.
This shows why geometry must be read in a historical and design context, not only through today’s preferences.
Architectural Symmetry Is Not the Same as a Laboratory Symmetry Grade
An idealized design may be mathematically symmetrical. A real stone deviates because of manufacturing tolerances, the rough, and the cutter’s decisions.
When discussing architecture, we ask whether corresponding elements are intended to repeat and balance. When a laboratory grades symmetry, it evaluates actual execution and types of deviation. This belongs to Chapter 20.
Proportions of Fancy Shapes
With fancy shapes, the problem is more complex because a single global average often conceals the local geometry that most influences appearance.
Ovals, pears, and marquises are elongated and have different terminal zones. Cushions can have several completely different pavilion architectures. An emerald cut uses a step arrangement and a keel line. Radiants and numerous modified brilliant designs divide the pavilion differently.
Thus, the length-to-width ratio is not a cut grade. Likewise, a single table of “ideal depth percentages” for all ovals or cushion cuts is not equivalent to a researched grading system for the standard round brilliant.
An Average Conceals the Distribution
Suppose a laboratory report shows an average crown angle of 34.5°. This does not mean that all eight crown mains measure exactly 34.5°.
They may be highly uniform. Alternatively, some may deviate more while the average conceals the difference. The same applies to pavilion angles, star length, lower halves, and local girdle thickness.
Geometric precision is therefore not proven by an average alone.
Rounding and False Precision
A laboratory report standardizes data. Values are rounded according to the rules of the specific service. Two stones can receive the same displayed angle even though their raw measurements differ within the rounding interval.
At times, the reported percentages may appear to the reader as though they “do not add up.” This may result from different reference points, local girdle thickness, derived values, and rounding. Such a discrepancy is not automatically a mathematical error.
Tolkowsky: A Reference, Not a Boundary of Nature
Tolkowsky’s 1919 model had an enormous influence on thinking about the round brilliant. It does not follow, however, that only one combination of an approximately 53% table, 34.5° crown, and 40.75° pavilion can produce an exceptional appearance.
Later experimental and computational work showed that multiple combinations can produce very high quality. GIA’s system for the standard round brilliant considers an enormous number of proportion combinations and additional components of appearance, design, and workmanship.
A historical point is therefore not the modern boundary between “correct” and “incorrect.”
GIA Excellent Is Not One Set of Numbers
GIA Facetware and published research clearly show that the GIA cut grade is not determined by comparison with a single ideal proportion point. The system considers the combination of table, crown angle, pavilion angle, star length, lower-half length, girdle, culet, polish/symmetry limitations, and other relevant factors.
This does not mean that all combinations are equal. It means that the quality boundary lies in a multidimensional space, not in a single chart found online.
“Hearts and Arrows”
The hearts-and-arrows pattern appears in very precisely fashioned round brilliants with appropriate optical and geometric symmetry. It is observed with a special optical viewer while the stone is held in a controlled position.
It is important to distinguish among:
- a hearts-and-arrows pattern;
- a laboratory symmetry grade;
- an overall cut grade;
- personal aesthetic preference.
None of these, by itself, automatically establishes any of the others. A precise pattern can be evidence of highly disciplined execution, but it is not independent, universal proof that the stone is the most beautiful or the best purchase.
Facet Count Is Not Quality
Adding facets can create new virtual divisions, change flash size, and give a design a distinctive appearance. But a larger number of facets does not automatically mean better light return.
A facet has value only as part of a coherent geometry. A poorly positioned additional facet can disrupt the pattern just as a well-designed one can create an interesting optical effect.
A Practical Way to Read Proportions
The professional sequence is:
- establish the shape;
- identify the facet arrangement;
- read the dimensions;
- connect table, crown angle, and crown height;
- connect pavilion angle and pavilion depth;
- include star and lower-half length;
- inspect the girdle and culet;
- check actual symmetry, not only the average;
- only then observe the optical result;
- use a grading system that actually applies to that shape and arrangement.
[VISUAL 18.4: “One number is not a cut”—table, crown, pavilion, stars, lower halves, girdle, and symmetry as an interconnected network]
Chapter Summary
- Architecture describes relationships among facets, not only their names.
- Brilliant, step, mixed, and modified brilliant are different organizational approaches, not quality grades.
- The table, crown, pavilion, star facets, lower halves, girdle, and culet function as a system.
- Crown and pavilion angles must be read in combination.
- An angle and relative depth are not the same information.
- Lower-half and star length change the face-up pattern, but neither has a universal, independent “ideal” value.
- An average can conceal significant local variation among facets.
- Rounded laboratory values are not the same as raw measurements.
- Fancy shapes cannot reliably be reduced to a single table of “ideal” ratios.
- Tolkowsky’s model is historically essential, but it is not the only geometry capable of high quality.
- GIA Excellent represents a range of permitted combinations and constraints, not a single point.
- Hearts and arrows, symmetry grade, cut grade, and aesthetic preference are different claims.
- A larger number of facets does not automatically make a better cut.
- The optical result of the architecture is covered in the next chapter.