glamoxyn

What the number actually describes

Colour temperature is a measure of how warm or cool a light appears, written in Kelvin. A low figure, around 3200K, gives the amber glow of a traditional tungsten lamp. A higher figure, around 5600K, looks like clean daylight with a slight blue lean. The scale feels backwards at first, because the cooler-looking light carries the bigger number, and that trips up many people new to studio work.

The figure is not a quality rating. A 3200K lamp is not better or worse than a 5600K one; it is simply a different colour of white. What matters is that every source that touches the same face or the same part of the set shares a number, or that the difference is chosen on purpose.

On our lighting plots the Kelvin value is written beside each fixture, next to its position and beam angle, so the crew rigging the truss and the engineer shading the cameras read the same intention from one sheet.

  • Around 3200K reads as warm, amber white
  • Around 5600K reads as cool, daylight white
  • A higher number means a bluer light, not a hotter lamp

Why the camera is less forgiving than your eyes

People standing under mixed lighting usually feel nothing is wrong, because the eye adapts within seconds and treats whatever is brightest as neutral. A camera has no such habit. It is balanced to one colour temperature, and anything off that setting is recorded as a tint, often a greenish or orange cast across the face.

The difference shows most on skin. Faces carry subtle reds and yellows, and a small shift in the light pushes presenters towards looking unwell or sunburnt. Darker and paler skin tones react differently to the same shift, which is why we check a plot against more than one person in the chair rather than a single stand-in.

This is the reason lighting design and camera setup are done together. A white balance set on the cameras can correct one dominant colour, but it cannot untangle two different colours landing on the same cheek from opposite sides.

Where mixed sources creep in

Few studios are lit by one type of lamp. A set may have LED panels overhead, an older tungsten key light kept because the director likes its softness, and a practical lamp on the desk that was bought from a shop. Each can sit at a different Kelvin value, and the set looks fine until the picture reaches the monitor.

Daylight is the other common intruder. A studio with a window, or an event hall with open doors, receives light at around 5600K that changes through the afternoon. In late autumn the problem is sharper than people expect: low, orange sun arrives sideways at a time of day when rehearsals are often running.

The practical answer is a walk-round before any camera is switched on, noting every source and its stated colour temperature. Where a lamp cannot be matched, we flag it, gel it or move it out of the picture.

  • Older tungsten units kept alongside LED panels
  • Shop-bought practical lamps on the set
  • Window or doorway light changing through the day

Gels, filters and fixtures that can be tuned

Colour correction gel is a thin tinted film placed in front of a lamp. A blue gel lifts a warm tungsten source towards daylight, and an amber gel does the reverse. It works well and costs little, but every gel absorbs some light, so a lamp needs enough output in reserve before it is filtered.

Many modern LED fixtures let the operator set the Kelvin value directly from the desk, which removes the gel entirely. The convenience is real, though we still check the result against a meter and a camera, because two fixtures reading 4300K on their menus can differ slightly on screen.

Tunable lights also change how a rig is planned. A single fixture type can serve a warm drama look and a cool news look on different days, which reduces what has to be hired, stored and cabled for a changing schedule of programmes.

Checking the colour on the day

Before recording we place a plain white or grey card in the presenter's position and look at it on the monitor. If it holds a clean neutral tone under every lamp that falls on that spot, the colour temperature is consistent. If one side drifts warm or green, the card shows it at once, before a person is sitting there waiting.

We then compare faces rather than relying on the card alone, because skin responds to the light differently from paper. A short check with the actual presenters, in their actual clothing, catches the problems a measurement misses, such as a bright shirt bouncing colour back under the chin.

Notes from that check go on the lighting sheet so the next session starts from the same values instead of rediscovering them. Repeating a setting is far cheaper than finding it again.

  • Place a neutral card at the presenter's position
  • Compare faces on the monitor, not just the card
  • Write the settled values on the plot

A short habit worth keeping

The simplest rule from this whole subject is to treat colour temperature as a decision rather than a default. Before anyone rigs a light, someone should be able to say what number it is meant to be and why. That one question prevents most of the tinted faces we are asked to diagnose after the fact.

Colour temperature sounds technical, yet the idea is plain: every white light has a colour, and a camera records those colours without the eye's forgiveness. Match the sources on a face, note the figure on the plot, and check it with a neutral card and a real presenter. Done early, it takes minutes. Left to the edit, it takes far longer and rarely looks right.