Why photos fail hardest on rough

Cut stones at least present consistent color and luster; rough presents weathered skin, matrix, coatings, and fracture surfaces. Water-worn quartz, calcite, and glass slag can look identical in a photo. Meanwhile the properties that actually separate them — density, hardness, crystal structure — don't photograph at all. That's why the workflow below is measurement-first.

What works on uncut material

Specific gravity

The star test for rough: hydrostatic weighing works on any clean, solid piece regardless of shape. SG alone often cuts 130 species to a handful.

Hardness

A careful point test on an inconspicuous spot brackets the Mohs value. Rough you own is the one place scratch testing belongs.

Crystal habit, cleavage & fracture

Free data that cutting destroys: prismatic vs massive habit, flat cleavage planes, conchoidal vs splintery breaks. Rough shows structure that cut stones hide.

UV fluorescence

Works identically on rough and cut. A dark room and a LW/SW lamp add a discriminator many rockhounds skip.

Refractometer (until you polish)

RI needs a polished flat. Grind and polish a small window and the confirming reading becomes available — worth it before hours of cutting.

Anything on porous or matrix material

SG readings on porous stone or matrix-attached specimens mislead. Break off and test a clean solid fragment instead.

The field-to-bench workflow

  1. Clean the specimen. Clay and coatings skew SG and hide everything else.
  2. Read the crystal: habit, cleavage faces, fracture type, luster on fresh surfaces.
  3. Hydrostatic SG — weigh in air, weigh in water, divide. Full technique guide here.
  4. Hardness point test on a spot that won't matter.
  5. UV response, longwave and shortwave.
  6. Enter it all in GemID's rough mode — habit, cleavage, fracture, SG, hardness — and read the shortlist. That's usually enough to decide whether the material is worth wheel time.
  7. Polish a window, take the RI, and lock the identification before serious cutting.

Common questions

Can an app identify a rock from a photo?

No — and be suspicious of any app that claims it can. Color and texture don't determine identity; measured properties do. GemID doesn't use photos at all: it matches the measurements you enter against a database of 130 species' published property ranges. Here's the full comparison.

Do I need a refractometer to identify rough?

Not to shortlist. SG, hardness, UV, and crystal structure together usually narrow rough to a few candidates — enough to decide whether to cut. The refractometer earns its keep at confirmation time, after you've polished a window.

What's the fastest single test for a bucket of unknowns?

Hydrostatic specific gravity. It needs one scale, works on any shape, and a reading like 2.65 versus 3.5 versus 4.0 instantly sorts quartz-family material from corundum-density and garnet-density finds. The full SG chart shows what each range means.

GemID's rough mode takes habit, cleavage, fracture, SG, and hardness — and keeps a per-piece record from field find to finished stone.

Open GemID →

Studying for the FGA or GIA practical? GemID drills instrument-based identification — refractometer, SG, and UV readings — and runs timed mock exams built around the real exam formats.

Practice exams →

Related