Testing A Night Light In A Dark Room
A lamp judged on a bright bench is a lamp judged in the wrong conditions.
For a while we assessed the Glow pieces the way we assess everything else: on the bench, under workshop lighting, in the middle of the day. Every one of them passed, and every one of them was far too bright the first time it was switched on in an actual bedroom.
The problem is that your eyes have already adapted. Under bright light a warm low output looks feeble, so you push it up, and then it lands in a dark room as a floodlight. The only fix we found was to stop testing in the daytime.
So the range gets judged at night, in a room with nothing else on, by someone who has been sitting in that room long enough to have adjusted to it. It slows the work down considerably and it is the only method that has produced a lamp anyone wants to sleep next to.
The side effect is that we now notice things that never came up on the bench — the brightness of an indicator, the edge of a shadow on a ceiling, how much light escapes backwards. All of it invisible at noon.
