The Science
GemMetric™ replaces subjective human comparison with instrumented colorimetry grounded in published, peer-reviewed science — and tells you exactly how much to trust every grade.
The problem
Color is the single biggest driver of a colored gemstone’s value — yet for decades it has been graded by eye, against a set of physical master stones, under lighting that varies from lab to lab and desk to desk. Two graders in the same room can disagree. The same stone can earn a different grade on a different day. That subjectivity is baked into the trade — and it costs real money on every transaction.
GemMetric™ starts from a different premise: measure the color, don’t match it — and be honest about how much to trust the result.
The approach
Traditional grading asks a human to compare a stone against a row of printed master swatches under whatever light happens to be in the room. GemMetric™ measures the stone’s body color instrumentally in CIE L*a*b* — the standard color space used in color science worldwide. The result is a number, not an opinion.
Every grade carries an uncertainty band that tells you how much to trust it. A tool that over-claims gets fired after one bad call; a tool that’s honest about its limits gets re-used — and re-use is what builds trust in the trade. The band widens or narrows based on real capture signals, so you always know when to act and when to recapture.
Every scan is linked to the serial number printed on the calibration card used. That means any grade can be traced back to the exact physical reference that calibrated the capture — making results reproducible and auditable, not just a one-off snapshot.
Every calibration is checked against public color standards, so measured color means the same thing regardless of which phone or lighting was used at capture — the same grade on two different devices is the same color.
Perceptual color modeling (CAM16)
Raw sensor data isn’t enough. The same stone photographed under warm LED and daylight reads differently to a camera — but the human eye adapts, and the perceived color stays stable. GemMetric runs a parallel color appearance model (CAM16) alongside the instrumental L*a*b* measurement, accounting for how the eye actually perceives color under the capture lighting.
The result: a grade that matches what you see, not just what the sensor reads. Two scans of the same stone under different lighting converge toward the same perceived color — which is the whole point of a color grade. No ordinary color app does this, because no ordinary color app models human perception.
The uncertainty band, visualized
Every grade comes with a band that shows how much to trust it. A narrow band means the capture was clean — act with confidence. A wide band means lighting or exposure compromised the read — recapture or get a second opinion. The band is computed from real signals, not guessed.
EXCELLENT CAPTURE
±0.5 grade — narrow band, high trust
COMPROMISED CAPTURE
±2.0 grade — wide band, recapture recommended
Tiered evidence database
Not all reference data is created equal. Every centroid in the GemMetric™ reference database carries a verification tier, so you always know how much to trust the number before you act on it:
Published in an academic journal with full methodology and sample data.
Defined by a gem laboratory standard (GRS, SSEF, GIA, LMHC).
Derived from published color ranges when no direct study exists.
See how honest, calibrated colorimetry changes the way you buy and sell stones.