Light colour
Convert a colour temperature to RGB — and see how much brightness is left.
Runs entirely in your browser — nothing is transmitted.
| Hex | |
| sRGB 0…255 | |
| linear what a GLB carries as color on a light node |
| Left of full output | |
| Extra drive for the same flux |
That extra drive cannot be had: the brightest channel already sits at 100 %. This is exactly where an RGB luminaire runs into its limit, long before it catches up with a white one.
at full output, all three channels at 100 %
| of which missing |
| Light share per channel | DMX · linear curve | DMX · gamma curve | |
|---|---|---|---|
|
|
The light share is the physical quantity. Which drive level produces it depends on the luminaire's dimming curve — which is why both DMX columns are here. If the datasheet says nothing about it, only measuring settles the question.
| Kelvin | sRGB | linear | left | extra drive | |
|---|---|---|---|---|---|
Enter a colour from a GLB, a lighting console or a datasheet — hex, sRGB 0…255 or a linear triple. Brightness does not matter, only the ratio between the channels; the swatch below is therefore normalised to full output.
| Colour temperature | ≈ |
| read as | · guessed |
| Distance from the curve |
The colour sits beside the curve. The temperature given is the nearest one, not the right one — with a green or magenta cast it says little about how the colour looks.
What is found is the nearest point on the same curve this tool uses forwards — it is not a colorimetric CIE colour temperature. That is deliberate: the way back then meets the way out exactly, and with it the value the GLB export writes. Rounded to 50 K, because a colour in 8 bits does not pin it down any further.
RGBW is left out
For RGBW the split cannot be computed, only guessed, as long as one number is missing: the colour temperature of the white channel. Whether the W is 3,000 K or 4,000 K decides how much runs through it and how much the colour channels have to contribute. Until that number comes from the manufacturer, this tool computes RGB only and leaves the white channel untouched.
Method
Tanner Helland's blackbody approximation from 2012, a curve fit to Mitchell Charity's table — not a Planck-law evaluation. Accurate to within a few percent between 1,000 and 40,000 K, undefined outside. Helland's fit yields sRGB-encoded values; the linear column is derived from them with the exact sRGB curve, because glTF works in linear space. Luminance Y follows Rec. 709 (0.2126 / 0.7152 / 0.0722).