I’ve tried both old and new f.lux with the MacBook monitor, basically to check and see what the spectrometer detects — but the screen hasn’t been bright enough for the little $40 kit spectrometer to say much, the result comes out like this:
You can see the bump at the 450nm end; intensity isn’t useful at those light levels though.
The spectrometer software was pretty flaky for a while and is being rewritten now, and I’ve taken the spectrometer apart to build their v3 kit. More news when I have any
Good stuff on ledmuseum. He evaluated the Asus brand of monitor I use (with f.lux.) But I notice the dates are from a while back. (I wish more stuff on the internet was timestamped or dated.)
When I was shopping for my monitor I found this in a review. Quote:
You may notice in the specs [of the Asus monitor being reviewed] a type of backlight we haven’t covered here before: GB-r-LED. The vast majority of LED screens use white LEDs (W-LED) on the top and bottom edges of the panel. A white LED emits blue light through a yellow phosphor, which neutralizes its color temp to around 6500 Kelvin. This is very easy and cheap to implement, and that’s why it’s so common. At the other end of the spectrum, we have RGB-LED which is literally red, green, and blue LEDs arrayed directly behind the LCD panel. This is very expensive and difficult to manufacture, and therefore quite rare.
The compromise is found in GB-r-LED technology. Here, the backlight consists of green and blue diodes coated with a red phosphor. The net effect is that the spectral peaks of the three primary colors are pretty much even. With W-LED, the spectral peak is much higher for blue. So, software (and the panel’s color filters) must intervene to achieve the correct color balance. A GB-r-LED panel is more accurate natively, making software and the color filter layer less critical. And you get the added benefit of the wider Adobe RGB gamut. It is a bit more expensive to manufacture than W-LED, but less so than RGB-LED.
The question is “what does the f.lux software actually do to control these leds?” At any rate, it’s pretty amusing to be on the computer at sunset. The f.lux kicks in, then the monitor seems to compensate, so the screen gradually shifts to deep yellowish-orange, then it comes back to a nice restful hue (to my eyes.) At least the software marks the passage of time by letting me know when sunset arrives.
It’s hard to find out exactly what wavelength is emitted — the key seems to be blocking or omitting 470 to 525nm, which is up to the middle of the green range roughly — per http://press.endocrine.org/doi/abs/10.1210/jc.2004-2062
(that’s from http://www.sleepinthedark.com/540nm.html — I find new websites every time I search on these questions, but most seem to link back to the same research)
This one does a good job of distinguishing color temperature (what we see) from emitted spectra. That’s where it’s hard to find out exactly what wavelength is being produced. It’s much easier to get color temperature information — but that often enough obscures the question what wavelengths are going into producing that color temperature.
I have some friends who’d been doing fine with amber LEDs, for a few years — then someone got them a nice buttery yellow “warm white” LED and they’ve been using that for more than a year. And their sleep fell apart, one of them ended up on sleeping pills. They simply failed to realize that their “warm white” color temperature still had a blue-green emission spike. I got them back to pure amber light sources — 590nm emitters — and, lo, problem solved. Again.
Duh.
I’d like to do more with color-mixing flashlights just to make this sort of thing easier to explain.
The “color temperature” section at that http://www.sleepinthedark.com/540nm.html page — page down a few to get to it — shows four different spectra that are all perceived as “white light” as an example of color mixing.
It’s not intuitive how this stuff works just for color vision. Throw in the newly discovered receptor for blue-green that controls sleep, and it’s way strange.
Thanks
Interesting, did you use the 1200K setting for this measurment in f-lux?
I found a review of your screen @ Anandtech and it has an ANSI contrast of 796, a black level of 0.42 & a stock white level of 6704K.
This explains a lot or should i say most of the differences, with how we perceive the effectiveness of just using altered gamma ramps (like we do when we use a program like f.lux or Redshift).
My monitor (HDTV) Samsung 46C750 has a ANSI contrast of 3000-4000 depending on how high i set the backlight setting, and a black level of 0.03 cd/m2 & a white level of 6500K.
This means my monitor is 14 times better at blocking the backlight with a gamma ramp, to very warm like 1000-1200K with Redshift or f.lux, or a red & warm bias calibration from the 6500K normal white point if i want.
It is all in the black level or maximum contrast that the display can provide, that limit how effective it is at blocking different frequencies of light.
MG, your monitor has a contrast of ~1000 & if you lower the backlight setting it can get as black as 0.0861, that is pretty good for an IPS monitor, the nice thing about IPS monitors is they have the same contrast, or light blocking ability from all angels you view them from, my VA type HDTV has only the highest level when looked at straight on, as soon as you look at it from the sides it losses contrast, and most likely spills some of that blue light out in to the room.
@MG, you ask “what does the f.lux software actually do to control these leds?” And the thing is that f.lux doesn’t do anything to the leds or the backlight in a LCD display, the colour control in a LCD display comes from the colour filters in the red, green & blue sub pixels And how open or close those sub pixels are.
And the higher ability to close or open & let through light or not let through light, is what gives the perceived or measured contrast or light blocking ability in a LCD display.
And you say that “then the monitor seems to compensate” when f.lux kicks in. I think it is your eyes that compensates, i have used f.lux for years now, and recently added on in house lightning with the same effect.
And my experience is that the more you use it, the more the eyes get use to seeing the warmer colours & it almost look like it use to, at first it looked very very red, & without room lightning in a similar colour i never used the warmest setting, but with all lights the same colour, it doesn’t interfere with perceived contrast for me & i can read just as fast in 1200K as in 6500K.
Aside — here’s an illustration of how wavelengths combine to make color temperature:
That’s a red emitter and a green emitter — and when those are combined the color temperature is yellow.
But this light, while the combination result looks yellow, still has no yellow emitter in it — only a green emitter and a red emitter.
That’s how “warm white” lights can still contain the blue-green wavelengths that mess up sleep, while they look nice and yellowish.
That is good way to demonstrate how it works & why people we care about need to consider these light spectrum issues.
With more & more led lightning being installed to save on electricity, the pharma industry :evil: can look forward to selling a lot more sleeping medicines :Sp in the near future :weary:
Thanks again for the useful commentary. More food for thought and further reading. Like many things, simple answers are insufficient. It gets complicated just below the surface. Cajampa, I am now pretty sure you are right about “eyes that compensate.”
Light on the left is a generic red P60 drop-in (*bay $9.64); light on the right is a Killer_Lumens Gen 2 XML with his custom color shifting driver driving the green side. Original picture is much better, the upload resizing did something to the colors.
One related interesting angle on this phenomenon is.
The pain you feel in your eyes when your eyes is dark adapted & you get strong day light or white light in them, is the rhodopsin pigment that is reduced chemically & heats up your retina, and you feel that as pain.
This rhodopsin gets transparent by light and you lose your night adapted vision, and when you remove light frequencys that have the strongest effect the blue & green, rhodopsin slowly recycles to give you your nightvision, red also reduces it but much more slowly.
One cool biohack I recently learned was one of the reason, why i have unusually good night vision is, i eat a lot of black currant every day & have for years, it is a good source of omega 6 GLA among other things, & it contains anthocyanins that actually make the recycling of the rhodopsin faster in the rods in the eye.
And the result is improved night vision & much faster night adaptation
If someone wants to try, remember that it is dose dependant, as in the more you eat the stronger the effect and lots & lots of carrots is always good
slightly offtopic, as i already have a amber thrower, decided to mod my sipik sk68 and drop this emitter in it https://www.fasttech.com/products/1609/10007033/1574702
I found an exhaustively thorough (up to 2013) review article here with extensive references — lots still being learned about this whole optical system:
(search Google Scholar for “melanopsin” if you want to find more like this — that’s the particular kind of “opsin” used by this system that affects melatonin regulation among other things still being discovered)
tidbits: it’s still an open question even how to measure levels of illumination that affect this system, to determine what is produced by an emitter as well as how the body perceives the light. Much discussion of that.
“… peak sensitivity in the short-wavelength portion of the visible spectrum (from 447 to 484 nm) …”
“… it was originally thought that illuminance of 2500 lux was required to suppress nocturnal melatonin in humans, but later studies have shown that under certain conditions, as little as 1 lux or less can suppress melatonin in humans …”
P.S. — look at those typical spectral curves for “white” LEDs and you’ll see why they (or any “fluorescent” light source that uses a phosphor) can interfere with sleep.
That tall peak lies right over the narrow band of the melanopsin receptor.
All the rest of the spectrum should be no problem.
Now if streetlights had the kind of super effective masking that high end monitors do, they could be adjusted. Not likely, too expensive.
and lots & lots of carrots is always good 
