A specification without a tolerance is an advertisement. Each line below carries the target, the band we accept, the figure measured on serial 14-0007, and the standard we measured it against. Any lamp outside its band does not leave the workshop.
The amber block is the accepted band; the dot is serial 14-0007 as it measured on 2 February. Hover a row to read the standard.
Serial numbers are stamped into the gland cap, not printed. Ours start at 09-0001.
Flux, colour and flicker are stable for the life of the board. Heat is the one figure that decides how long that life is, so it is the one we design the whole extrusion around — 60 mm of still air above the fins, and a board that settles at 34 °C and stays there.
Because 1 400 mm above a 760 mm bench puts the lens at 2 160 mm, which is where a standing adult's head is not. Quote it at a metre and the number looks better and the lamp is in your way.
The darkest of the nine points is 63 % of the brightest. The eye starts noticing a gradient at about 0.4, so at 0.63 the bench reads as one flat field. A bare strip without the batwing lens measures 0.28 on the same grid.
Yes, up to four. They will track within 40 K of each other across the whole range because the drivers come from one reel and we bin them in fours. Tell us at order and we will keep your set together.
L90 at 50 000 hours means ninety per cent of day-one flux after roughly twenty-four years of normal studio use. Colour moves less than flux: we have measured batch 9 lamps at 11 000 hours and they had shifted 18 K.
No. We built one in 2021. It weighed 7.9 kg, held 22 minutes, and the cells disliked sitting at 34 °C. It is not coming back.
It never leaves. If it fails on your bench inside eight years we collect it, measure it here, and either repair it or build you a new one. Twenty-three lamps have come back since 2017; nineteen were drivers.
Then come back and tell us what they said. Half the changes we have made since 2017 came out of that conversation.