4 ms·

That's not a bad idea. The newest kobo reader actually has red LEDs (I assume) to fade to red at night: https://us.kobobooks.com/products/kobo-aura-one https://

by azdle 10y ago

That's not a bad idea. The newest kobo reader actually has red LEDs (I assume) to fade to red at night: https://us.kobobooks.com/products/kobo-aura-one https://us.kobobooks.com/products/kobo-aura-one Hopefully amazon copies that feature for their next one.

wlesieutre 10y ago
I can vouch for the Aura One, it's a lovely device. The screen color temperature adjustment goes very warm. Tough to eyeball without a side-by-side reference, but I'd guess at least down to 2200K. I've seen it referred to as RGB LEDs for the backlight, but I think it's more likely to be two-channel mixing with cool white and warm white, as is common on "tunable white" light fixtures. All you get for control is a cool/warm slider or auto, so the hardware is a bit immaterial the average user (though it would affect the blue wavelengths present in the spectrum). I think we have some sort of spectrometer around the office, I'll have to see if I can track that down and check. https://imgur.com/Irw9yhg https://imgur.com/Irw9yhg
- wlesieutre 10y ago
  Checked the spectrum, it's a range from 6300K (cool white) down to 1700K (very warm white, could probably call it amber). I measured it at three points: the ends of the range and a point eyeballed in the middle of the UI slider, which may or may not have been quite at 50%. Points on black body curve in CIE 1931 colorspace: http://i.imgur.com/FCsdxZW.png http://i.imgur.com/FCsdxZW.png That's the curve where we define "white" light, based on the perceived color of a black body object heated to various temperatures. The naming convention is a bit strange; "warm white" refers to the orange end with relatively cool temperatures, while "cool white" refers to the blue end with relatively high temperatures. Relative spectral distributions, each normalized with the peak wavelength at 1: http://i.imgur.com/3Bw6ltt.png http://i.imgur.com/3Bw6ltt.png The "cool" spectrum is pretty typical of a phosphor converted white LED, where you get a strong peak in blue and then a smooth curve across the upper end of the spectrum produced by the phosphor. The "warm" spectrum has two peaks, dominated by red around 630nm and a smaller green peak around 530nm. You can tell just from the black body curve that they've used more than the two-channel LED mixing that I guessed in my parent post. With just two LEDs, you're basically making linear combinations of the two colors and can only move in a straight line between them. If it were only the cool white and red LEDs, the color would move along this path between 6300K white and 630nm saturated red: http://i.imgur.com/d2W3sb4.png http://i.imgur.com/d2W3sb4.png Having a 3rd channel green LED allows them to pull it back up to the black body curve. Interesting to see that they're apparently doing a 3-channel color adjustment with cool white / red / green, rather than the RGB system that people are so used to.

by azdle 10y ago

wlesieutre 10y ago

I can vouch for the Aura One, it's a lovely device. The screen color temperature adjustment goes very warm. Tough to eyeball without a side-by-side reference, but I'd guess at least down to 2200K. I've seen it referred to as RGB LEDs for the backlight, but I think it's more likely to be two-channel mixing with cool white and warm white, as is common on "tunable white" light fixtures. All you get for control is a cool/warm slider or auto, so the hardware is a bit immaterial the average user (though it would affect the blue wavelengths present in the spectrum). I think we have some sort of spectrometer around the office, I'll have to see if I can track that down and check. https://imgur.com/Irw9yhg https://imgur.com/Irw9yhg

wlesieutre 10y ago
Checked the spectrum, it's a range from 6300K (cool white) down to 1700K (very warm white, could probably call it amber). I measured it at three points: the ends of the range and a point eyeballed in the middle of the UI slider, which may or may not have been quite at 50%. Points on black body curve in CIE 1931 colorspace: http://i.imgur.com/FCsdxZW.png http://i.imgur.com/FCsdxZW.png That's the curve where we define "white" light, based on the perceived color of a black body object heated to various temperatures. The naming convention is a bit strange; "warm white" refers to the orange end with relatively cool temperatures, while "cool white" refers to the blue end with relatively high temperatures. Relative spectral distributions, each normalized with the peak wavelength at 1: http://i.imgur.com/3Bw6ltt.png http://i.imgur.com/3Bw6ltt.png The "cool" spectrum is pretty typical of a phosphor converted white LED, where you get a strong peak in blue and then a smooth curve across the upper end of the spectrum produced by the phosphor. The "warm" spectrum has two peaks, dominated by red around 630nm and a smaller green peak around 530nm. You can tell just from the black body curve that they've used more than the two-channel LED mixing that I guessed in my parent post. With just two LEDs, you're basically making linear combinations of the two colors and can only move in a straight line between them. If it were only the cool white and red LEDs, the color would move along this path between 6300K white and 630nm saturated red: http://i.imgur.com/d2W3sb4.png http://i.imgur.com/d2W3sb4.png Having a 3rd channel green LED allows them to pull it back up to the black body curve. Interesting to see that they're apparently doing a 3-channel color adjustment with cool white / red / green, rather than the RGB system that people are so used to.