Hacker News Reader

Comment by phasetransition | Hacker News Reader

I'm head of engineering for a company that designs outdoor luminaires for industrial environments, but not street lights. We use LED and other technologies. Like so many things in life, there are aspects of the AMA directive with potential cause for concern, but also combined with a mixed bag of scientific underpinnings. If you have specific questions about LED physics or human vision, I'm happy to give them a go. I also can speak to some of the literature behind the AMA position paper.

View on HN

What temperature LED light should I have for my bedroom?

For sleeping? I recommend Zero Fahrenheit Fahrenheit.

I have a Philips sunrise alarm clock that looks like a lamp.

I recommend that. It is a great lamp colour temperature and way to wake up.

I second this recommendation, although it is worth pointing out that they are not that bright even on full. Fine for going to bed but maybe not the only thing you want on a cloudy day. The sunset mode is really nice too, and I really really wish there was a way to hold at the pink color when it is nearly off. If you could get strips of that color to circle the room that would be excellent.

The flip side of lights that disrupt melatonin is that they can be good for you when you first wake up if natural light is not that strong in the morning where you are (or you aren't exposed to much of it when you first wake up).

Essentially any LED bulb that you can buy is a blue diode with a down conversion phosphor. The composition of the phosphor, and the relative amount of blue light vs. phosphor emissions shape our perception of color temperature for these LED sources.

Due to an effect called metamerism, our brains can perceive the same apparent color from sources with different underlying spectral power distributions (power vs. wavelength).

So in a sense, the color temperature of a black body does not have much underlying connection with the spectral power distribution of LED lamps.

With that substantial caveat, the lower the correlated color temperature of the bulbs that you purchase, the less of the overall spectral power is from the diode, and the more from the phosphors.

If there ends up being a link to blue light exposure for circadian effects then bulbs with lower CCT that have more emission from the phosphors are the safer bet.

Some LEDs have phosphors which almost entirely absorb the light from the diode.

For example, Lumileds PC Amber and Lime LEDs employ this trick. They're usually used as a corrective color component and not the main lumen source.

I've not seen the PC Amber data from Lumileds, but I can vouch that their lime color LED has no blue diode emission peak visible in the spectra.

You might be interested in https://en.wikipedia.org/wiki/Kruithof_curve and https://en.wikipedia.org/wiki/Color_Rendering_Index (high CRI).

Reply on news.ycombinator.com