Hacker News Reader

Comment by snowwrestler | Hacker News Reader

> Older streetlights are high-pressure sodium bulbs, which produce a warm yellow glow around a color temperature of 2,000 K... Many of the LED streetlights on the market have much cooler color temperatures of 3,000 or even 4,000 K.

I feel like referring to 2,000 K bulbs as “warm” and 5,000 K bulbs as “cool” has got to be one of the worst screw-ups between scientific terms and common terms in English. And probably a missed opportunity to embed some latent understanding of thermodynamics in the broader culture.

View on HN

Warm/cool color theory predates the kelvin scale, so it might be more accurate to say you feel Celsius' flipping his original scale was the screwup? But as how it is now makes that much more sense, and since you couldn't really have a kelvin version of such a thing anyways, maybe chalk it down as just one of those things.

In any case no one could convince me to call the color of fire and the sun anything but warm, and that of snow and sea cold.

>In any case no one could convince me to call the color of ... snow and sea cold.

How about the color of a gas stove flame?

Maybe we should have called them "warm white" and "hot white" instead.

Yeah warm to hot is good but again runs into clashing with things established before much had been said about black-body radiation.

Just because it might not be conductive to a neat 1:1 mapping with scientific common sense doesn't have to mean either should change. Quirks are fun!

I'm used to it myself, but I work in architectural lighting and I can't count how many times I've had to explain that "warm white" is the low temperatures and "cool white" is the high temperatures. It's a frequent nuisance for me, but not a huge deal and we're stuck with it.

If we were going to make people put in some effort and redefine fix stupid quirks that exist out of historical convention, I'd put that effort toward adopting metric instead of switching words for shades of white :P

It's not like switching cool and hot would make sense either, 2700K (orangeish) is 4400°F and 5000K (blueish) is 8540°F. Neither of those is anything near a "cool" temperature. "Blazing white" and "inferno white"?

Exactly what I mean, it's never going to make perfect sense, so trying to adapt two separate meanings to gel slightly better seems futile.

I would argue the US still not being on SI is a bit more than a quirk. I wonder whether you'll have completed metrification by the end of the century.

If we're replacing cool white, "blue white" is the best term I can think of... or perhaps "snow white"?

Yeah, it makes total sense, historically, how we got to this point. I just like to take a step back sometimes and appreciate the silliness of language. Not trying to assign blame, but just notice: hey it's kind of screwed up that totally opposite descriptions of a temp are both correct depending on context.

"This 5,000 K light bulb is cooler than this 2,000 K light bulb."

"This 5,000 K molten iron is hotter than this 2,000 K molten iron."

Yeah, some kind of lede/lead type separation would make sense. But in a sense there already is one as K barely sees layman usage outside color temp, so your second sentence isn't likely.

There is no misshap.

On the one hand, we (or perceptual psychology) deem red a warm color, probably by learned or ingrained association with fire. And the blue part of a flame is actually hotter, so OP has a point. The very word blue actually stems from a sense shiny, bright.

On the other hand, the watt/lumen ratio of red is higher than blues', I believe, and both are much higher than greens'. Simply put, if lower wavelengths pass less energy, then waves at the same energy but different frequency must have higher attenuation if closer to red. Alas, I'm not sure what velocity means in optics.

This is more illustrative with Music: A bass that you can hear and feel carrying the high tones must have a lot of energy.

Not sure what you're trying to say.

There's no inherent watt to lumen ratio for anything as far as I'm aware, but mainly, how would the other things you mention negate the fact that cool/warm colors are indeed flipped compared to temp?

You're saying blue, having higher energy and matching red with fewer photons, makes sense to describe as cooler because it can put in less effort?

Definitely haven't heard that one before.

Edit: there's obviously a language barrier but that's what I took "no mishap" to mean, that you disagree with us.

You have no idea how Lumen is defined, do you?

Red embers (~1000K eigen radiation) are hot, blue(-ish) ice (lit by 5500K daylight) is cold.

..go figure!

It's an association problem. Fire=warm and snow=cold. Fighting unconscious links between words like these is near impossible imo :)

Remember that CCT color temp (K) is about the equivalent ideal blackbody radiator. Think "white hot"

I don't remember that, can you elaborate further?

Color temperature refers to the color of an ideal black body radiator at that actual temperature[1]. And the hotter a black body radiator gets, the more blue the light output is. ~5800K is the color of sunlight because that is the physical temperature of the surface of the sun.

Which happens to unfortunately be the exact opposite of the much older color theory notion of warm (red-ish) and cool (blue-ish) colors.

They both use words related to temperature but actually have nothing to do with each other.

[1] http://www.techmind.org/colour/coltemp.html

Reply on news.ycombinator.com