Specifically white light, a white surface is a surface that diffusely reflects all the light that hits it.
“White is all the colors combined” you say, every wavelength of visible light in equal amounts? “White noise” of light?
Is it the combination of wavelengths that evenly stimulate the cones in your eye?
Do monochrome spikes of red, green and blue count as white?
How about 2 spikes of yellow and blue, or cyan and red?
Is the black body locus considered the definition of white? TM-30 and CRI seem to think so.
Your monitor is probably 6500K, is that “white”? Except it probably cheats with red, green and blue spikes (or humps? [digression, but are OLED screens “spikier” than LCD?]).
How about 5000K? 2700K?
What about 10000K blue car headlights (or Olights) that people get for some reason?
How about when you get down to a cherry red 1000K? Is that still “white”?
Is high-CRI light “whiter” than low-CRI light?
Is whiteness defined by what it is illuminating?
If we are in a room or environment that is entirely white, does it matter if the spectrum of the light is continuous or spiky (high or low CRI), since in the absence of color they would both look the same?
Would the act of introducing a colored object into this room cause the spiky light to be less white now because we have another reference for “whiteness”?
What if somebody is colorblind? Would they perceive a certain combination of wavelengths to be white that a normal person would consider to be colored?
What about people with tetrachromacy? Do they see colors when we see “white”?
Is white subjective? Does it even have a definition? Does it even exist?
Is sunlight the true white? Since it has been the most significant reference for white light since we were created. What about fire though? Or overcast sunlight? Sunsetlight? Moonlight? Lightning?
Do the stars all count as white even though they are different colors?
If our perception of white changes, does that mean our definition of white changes?
If you have a white light in front of a brighter white light is the first light still white?
Let’s not even get started on gray.
or brown.
or purple.
or primary colors!
If you’re in, say Arizona in the summer solstice at noon on a clear day, the CCT of that sunlight is apt to be in the region of 10,000K. Everything will seem similar to a more normal clear day at around 6500K: your brain will adjust visual perception and everything will be remarkably similar under both circumstances.
But that same 10,000K (or 6500K) will look progressively more off n the contexts of a stadium or car headlamp or handheld flashlight. The absolute levels of light will not be as high as sunlight and there will be increasing contrast between lit and unlit regions in your vision.
The phenomenon seems to be that the lower the overall intensity the more sensitive we are to the apparent coolness of high CCTs. Probably why task lighting is 5000K-4000K - the best balance of blue to stimulate awareness, good color rendition, while not seeming angry blue.
According to Nichia datasheets, white is something between 5000K and 6500K, below that is warm white. Delta uv depends on the Standard, like ANSI/NEMA C78 377A for example.
In my opinion, spectral distribution has nothing to do with been more or less white, but is sure important for color rendering. A good measurement of that is ASD. According to Bridgelux “Average Spectral Difference (ASD) provides an objective measurement of how closely a light source matches natural light over the visible spectrum, averaging the differences of the spectral peaks and valleys between a light source and a standardized natural light source of the same CCT.”
The point of this whole post was to be a little bit silly, but I don’t think there really is an “absolute” definition of “white” light.
Wikipedia even says “there is no single, unique specification of “white light””. I think white surfaces do have a theoretical, absolute definition.
To me, a true white light would be one of measurable energy at all wavelengths of the visible spectrum (which would be around 5400k and look slightly red from what we’re used to because it wouldn’t taper off in the extreme reds).
Actual spectrums with the most balance are between 5-6k.
Based on study results, there have been some proposals to shift the curve of the ANSI White definition line so it more closely matches human perception instead of matching a blackbody radiator. In practice, this would mean “white” would have a negative duv at CCTs under 5000K or so, and lights which are currently categorized as “4000K neutral” would be considered 4000K with a green tint.
Basically, everything except cool white would need some pink added (or some green removed), to match with what average human perception finds the most pleasant.
I’m a fan of this proposal, because it aligns closely with how I personally see colors… and it would properly categorize the common 4000K “lemongrass” lights as being green instead of neutral.
To see the difference, take a SST-20 95CRI 4000K LED and compare it to a 519A 5700K dedomed LED, at like 20 lm per LED. They both land around 4000K and they both have high CRI, but they look dramatically different. To me, the 519A looks like a very slightly warm white, while the SST-20 looks sickly green… and the standards proposal aims to officially recognize that difference and use it to improve household lighting.
