Open Gem Property Dataset

Name: Loupewise Gem Property Dataset
Creator: Loupewise
License: https://creativecommons.org/licenses/by/4.0/

The physical and optical property table behind Loupewise, published in full as an openly licensed dataset. 132 entries across 69 species, in CSV and JSON, free to download, redistribute and build on under CC BY 4.0 — including commercially. All we ask is attribution.

132

entries

species

fields per entry

132

entries carrying source keys

CSV

loupewise-gem-properties.csv · one row per entry · lists separated by “;”

JSON

loupewise-gem-properties.json · typed — numbers stay numbers, lists stay lists · includes a field dictionary

Licence and citation

Released under the Creative Commons Attribution 4.0 International licence. You may share and adapt it for any purpose, including commercially, provided you give credit and indicate whether you changed anything.

Please cite it as:

Loupewise. Loupewise Gem Property Dataset (version 4fbacd51ce04). 2026-10-09. Licensed CC BY 4.0. https://gemid-labs.com/data/

The version string is a content hash of the source data, so it changes when and only when the data changes — two downloads carrying the same version are the same table.

What is in it, and what is not

This is the property table: the measurable physical and optical characteristics a reference work would carry anyway. It is the part our sources are actually sources for.

What is not here is the curation — the judgement about which stones get mistaken for which, how to tell them apart, and how to walk a stone through an identification. That is the product, and it stays closed. We would rather state that boundary plainly than let you discover it by absence:

Withheld	Fields
Confusion pairs and simulant notes	`confusedWith, simulants, common_simulants_for`
Natural/synthetic and treatment flows	`natVsSynTests, nat_syn_tests, natural_vs_synthetic_tests, treatments, treatment_tests`
Curated teaching prose	`key_differentiators, notes`
Inclusion observation lists	`inclusions_loupe, inclusions_darkfield`
Commercial data	`price_info`

Provenance, and where the sources disagree

Each row carries a sources column listing the reference keys supporting its physical and optical properties. Those keys resolve against our reference database. They are cited for properties and nothing else — there are no prices in this dataset, by design.

Where published authorities genuinely disagree about a value, the JSON carries a source_discrepancies note on that entry rather than silently picking a winner. 2 entries do so today.

An empty cell means not recorded — never zero. Sometimes the property is structurally absent (singly refractive material has no birefringence at all, rather than a birefringence of zero; an entry that is its own species has no separate variety name). Sometimes it simply has not been recorded for that entry — the Hanneman filter and thermal probe columns are the sparsest. Do not impute a value to a blank.

Ranges are ranges. Where a property varies across material sold under one name, the table gives a minimum and a maximum rather than a single figure. A stone reading outside a range is not necessarily misidentified — measurement error on ordinary equipment is often wider than the range itself, particularly for specific gravity.

Field dictionary

Field	Meaning
`id`	Stable identifier used across the dataset and site.
`name`	Common trade or varietal name.
`species`	Mineral species the entry belongs to.
`variety`	Variety name within the species.
`crystal_system`	Crystal system.
`optic_character`	SR (singly refractive), DR_U/DR_B (uniaxial/biaxial, + or −), or AGG (aggregate).
`adr`	Whether the material commonly shows anomalous double refraction.
`ri_min`	Lowest refractive index recorded for the entry.
`ri_max`	Highest refractive index recorded for the entry.
`birefringence`	Maximum minus minimum RI. Empty for singly refractive and aggregate material, which has no birefringence rather than a birefringence of zero.
`dispersion`	Dispersion (B–G interval). Empty where no value is recorded — mostly opaque, aggregate or organic material whose fire is not observable. Until 2026-10-09 these 17 entries carried 0, a placeholder rather than a measurement.
`sg_min`	Lowest specific gravity recorded.
`sg_max`	Highest specific gravity recorded.
`sg_typical`	Representative specific gravity within that range.
`hardness_min`	Lowest Mohs hardness recorded.
`hardness_max`	Highest Mohs hardness recorded.
`cleavage`	Cleavage quality and direction.
`fracture_type`	Characteristic fracture.
`crystal_habits`	Habits the material commonly forms in.
`luster`	Luster terms that apply.
`transparency`	Transparency range.
`color_categories`	Broad colour bands the material occurs in.
`pleochroism`	Pleochroism strength: none, weak, moderate, strong_dichroic or strong_trichroic.
`pleochroic_colors`	Twin or triplet colours where recorded.
`facet_doubling`	Whether facet doubling is visible through the crown.
`facet_doubling_strength`	Strength of that doubling.
`magnetism`	Response to a neodymium magnet.
`fluorescence_lw`	Long-wave (365 nm) UV response.
`fluorescence_sw`	Short-wave (254 nm) UV response.
`chelsea_filter`	Appearance through a Chelsea filter.
`hanneman_filter`	Appearance through a Hanneman filter.
`spectroscope_absorption`	Absorption features described for the visible spectrum.
`thermal_test`	How the material reads on a diamond thermal probe.
`sources`	Reference keys supporting the physical and optical properties in this row.

Generated from the same data the reference charts and the identification engine read, so the dataset and the site cannot disagree.