We'd nearly finished the hang when someone stepped back and said the left wall looked hot. It did. The top of the big canvas was blazing, the bottom half sank into murk, and a bright scallop on the plaster above kept dragging your eye up and off the work. I'd already re-hung one print on that wall twice that afternoon, sure the frame height was the problem. It wasn't. Nobody'd bought the wrong fixture either. Somebody had read the sheet wrong, or hadn't read it at all, and honestly that's more of what gallery lighting comes down to than shopping ever is.
A spec sheet tells you how a head will behave, but only if you know what each line means and how it shifts once the thing's up on a ceiling at a real distance from a real wall.
Three things carry most of it. CRI, beam, distance, and distance isn't even on the sheet, which is half the trouble. None of them means much alone. You won't get product figures from me here either; those get checked per model and per setup, every time.
What CRI is, and isn't
So what's it measuring? How faithfully a source shows a set of test colours compared with a reference source at a similar colour temperature. Higher means it tends to show those samples more like the reference does. Handy shorthand, which is why it's on every sheet you'll ever be sent.
Here's the catch. The general index averages a bunch of mostly muted test colours, so two sources with similar headline values can treat a strong red very differently, and strong reds are exactly where a painting tells on you. Lots of specifiers ask for the separate sample values next to the headline, the deep red one above all. Newer metrics try to give a fuller picture. For the formal definitions, start with the Illuminating Engineering Society.
And there's a lot CRI keeps quiet about. Brightness, nothing. Beam quality, glare, how warm the white looks: nothing there either. Whether a pigment's at risk from ultraviolet or infrared is a conservation question for the gallery's own advisers, not something a lighting spec answers. It won't promise that two heads from different batches will match, either. That's another line on the sheet, and it's the one people skip.
Bright core, soft ring
Squint at a lit frame from across the room and you'll see it: a bright middle, then a soft ring, then plain wall. That ring is what the beam angle figure leaves out.
Sheets usually define beam angle as the cone where intensity stays above a stated fraction of the peak, so there's always light outside the quoted figure, and how it falls off decides whether the edge looks crisp or smudged. Two heads with the same quoted angle can have very different edges. Have a proper look before you trust the number. Narrow beams pack light into a smaller patch that stands out harder against the wall; wide ones spread it more evenly and hit softer. Neither one wins. A small, dense painting might want a narrow beam that sits neatly inside its frame, and a big canvas or a textile might want a wide, gentle spread with no hot centre. Pick for the work, not the head.
Pull it back a bit and see what happens
Same head, same beam angle, and a completely different result depending on how far it sits from the wall. Move it back and the patch grows and dims. Bring it in and the patch shrinks and gets brighter. That's just geometry, and you can sketch it on paper with a ruler before anyone buys a thing. Most people don't. I didn't, that day, and should have.
Aiming angle piles on another variable, and that's where our left wall went wrong. A head on a track running parallel to the wall hits it at a slant, so the patch stretches into an oval that's brighter at the top than the bottom, hot above and murky below, exactly what we'd been staring at the whole time. Aim too steep and frames cast long shadows down across the work; I spent ages squinting at the frame line on that print before it clicked that the shadow under the moulding was the giveaway. Too shallow and the light bounces off the glass straight back into people's eyes. Most curators land somewhere in a moderate middle, picked so frame shadows stay short and reflections fall below eye level, but the right answer depends on ceiling height, where the track runs and how high the work hangs. It's worked out on site. You can't carry it from room to room.
Pick the track position first. The beam angle follows from it, not the other way round.
Even wall or bright pools?
Both, usually. Good galleries do both. Washing a wall lights the whole surface evenly, so work can be rehung freely and the room feels calm. Accent picks out single pieces, for drama and to steer people. Trouble starts when the two get mixed carelessly, a few works in bright pools and their neighbours stuck with a dull wash, and the pecking order looks like an accident.
Washing suits a hang that changes a lot, or a collection meant to be read as a whole. Accent suits the one piece that carries the room, or a sculpture that wants light from one side to show its shape. Either way, keep the gap between them on purpose. Spot everything and nothing stands out.
Look up, too. Scallops on the wall above the work pull the eye away faster than almost anything else, and they're easy to miss when you're staring at the canvas.
Shops run on the same logic, funnily enough. The piece on lighting retail without glare argues it from the shop floor, which is a handy counterpoint if your space mixes showing and selling.
Down the sheet, line by line
This is roughly how I read a track fixture sheet, top to bottom. Every value gets confirmed for the exact model, optic and driver you're ordering, because plenty of products carry several variants under one name and the sheet you got sent might be the wrong one. Ours was, once. We didn't notice till the heads were up.
- CRI (Ra) and R9. General fidelity, plus how the source handles strong red. Check the stated value holds for every colour temperature on offer, not just the one in the photo.
- Colour temperature. How warm or cool the white looks. Which options exist, and can heads in one room really match?
- Colour consistency. How closely units match each other. Read the tolerance, then ask if it's guaranteed across batches, because whoever adds heads to this room in a few years will want to know.
- Beam angle. Which optics exist, and can they be swapped on site from a ladder?
- Luminous output. Is the figure for the complete fixture or the bare source? This line flatters products more than any other on the page.
- Intensity distribution. The real shape of the beam and how it falls off. Ask for the photometric file so you can model it.
- Tilt and rotation. Does the aiming range reach from your track position to your hanging height?
- Accessories. Snoots, louvres, spread lenses, filters. Which of them fit this specific model?
- Driver and dimming. Dimming belongs to the driver and controller as a pair, not to the track, and it's confirmed as a set or not at all.
The DALI Alliance standards and the safety requirements published through the IEC are worth knowing. They still won't tell you whether your driver, controller and head get along. Only trying them together does.
Borrow a few heads and an afternoon
No sheet beats a mock-up. If the budget stretches, borrow a few heads with different optics and light one typical work at the real distance, ladder out, nobody rushing you. You'll learn more that afternoon than from a stack of comparison tables. Watch the edge of the beam. Watch the frame shadow. Stand where visitors will stand, not where the installer stands, and go looking for reflections. Swap optics and watch the colours shift.
In the end the left wall didn't need a new fixture. It needed the head moved along the track and a softer optic tried on site, which took less time than the argument about it.
Think about the years ahead, too. A gallery may add heads long after the first install, and matching colour across batches is far easier if the original spec wrote down the exact variants. Keep that record with the building documents. Not in somebody's inbox, where it'll vanish the week they leave.
Read them together
CRI tells you about colour fidelity, up to a point. Beam is the spread. Distance and aim decide what that spread turns into once it lands on plaster. Read them together, check each against the specific model, test where you can, and the hot wall stops being a mystery. It's a problem with a fix. Annoying, but fixable.
Our page on gallery lighting applications covers rail placement in exhibition rooms, and commercial track lighting goes through rail and mounting options. When you're ready to scope a space, the projects guide lists the drawings and details we'll ask for.