The argument · measured, not asserted

We do not sell daylight. We sell 5 003 K, ±70.

ONE RAY, 86 W780 nm380 nm

Every lamp that calls itself “full spectrum” is making a claim about a curve it rarely shows you. Ours is below, measured in our own sphere, next to the ceiling panel it usually replaces.

Plate 04

Spectral power, 380 to 780 nanometres.

Both curves normalised to equal luminous flux, measured 400 mm below the lens at 4 000 K, sphere serial 7741, 23 °C.

R9 BAND1.00.50400450500550600650700750NANOMETRESNord 120 · Ra 98 · R9 96 · Rf 97 · Rg 101Typical 80 CRI ceiling panel · R9 12PLATE 04 · SPD · SPHERE 7741 · 08.01.2026 · WERKLICHT OPTIK, JENA

The same twelve pigments, twice.

Drawn from the measured reflectance of our own test card under each source. Move the pointer across the card: the left half is what an 80 CRI panel returns, the right half is the Nord 120 at D50. The reds are where it shows.

80 CRI PANEL · R9 12NORD 120 · D50 · R9 96
Hover to open the card · test card WL-TC3, measured 08.01.2026
A pool of warm light thrown onto a dark plastered wall
The measuring wall, 400 mm from the lens — every card is read here before it is read anywhere else
Red ochre · PR102Vermilion · PR108Yellow ochre · PY43Naples · PY41Terre verte · PG23Cerulean · PB35Ultramarine · PB29Manganese violet · PV16Raw umber · PBr7Vandyke · NBr8Flake white tint · PW4Lamp black · PBk6

Four things this lamp does not do.

We would rather lose the sale than have you discover these in week three. If one of them is a deal-breaker, write and we will tell you honestly what to buy instead — usually it is a bigger, duller fixture from somebody else.

THE CLAIMS WE STRUCK OUT, 2017–2026

This list has grown twice and never shrunk. Anselm keeps it on the wall above the measuring bench.

It is not daylight

D50 is a standard illuminant, not the sky. A north window on a clear February morning is a different spectrum and you will still see a difference if you look for one.

It will not cover a 2.4 m bench

One lamp lights 1.08 m² to 750 lx. Longer benches need two, hung 1 100 mm apart, and we will say so rather than sell you one and hope.

It does not improve your mood

We make no circadian, wellbeing or alertness claim. There is good research on it; none of it was done on this lamp, so it is not ours to quote.

It cannot be recessed

The fin stack needs 60 mm of still air above it. Bury it in a ceiling and the board runs at 71 °C, which costs you the eight-year warranty and, eventually, the R9.

Three measurements, every lamp.

Test 1detector

Sphere, 90 seconds

Flux, CCT, Ra, R9, Rf and Rg at four drive currents. Anything outside Ra 97 goes back to the board station; about one in thirty does.

Test 211 angles, 0–90°

Goniophotometer, 11 angles

The batwing is only a batwing if it measures like one. We take candela at every 9° and build the IES file that ships with your lamp.

Test 334 °C0 h72 h

Burn-in, 72 hours

Full output, lid on, in the crate it will ship in. Infant failures happen in the first forty hours, in our workshop, not on your bench.

A doorway throwing a long beam of warm light across a dark concrete floor

“Ceramics are the cruellest test there is — you glaze under one light and fire under another. With the Nord at D65 I stopped losing crowns to a shade that only existed in my workshop.”

Ivar Ostrowski photographed in low light against a dark wall
Ivar OstrowskiDental ceramist · Gera · one Nord 120 since batch 11
Corridor outside the glaze room, 19:12

Want the raw files?

The SPD as a CSV, the IES photometric file and the sphere report for any serial we have shipped. Ask and we will send them; you do not have to buy anything first.