It doesn't "remove brightness magically", it just swings the color temperature of ALL colors displayed towards the red end of the spectrum, which studies have shown can decrease the inhibition of melatonin production which means you will be more likely to get sleepy around the time you're supposed to, whereas before you might not have until much later and disturbed your sleep.
What aren't you getting about this? There is leakage of white unfiltered light that comes from from most monitors (the super duper high quality monitors boast about low light leakage, but they still do it a bit). F.lux cannot possibly change the color temperature of this light. So my question is and has been, does that leakage have enough of an impact to negate the effects of F.lux? That's the question worth answering, because if it's, "No, the effects are negated by the white light leakage", then there's absolutely no point in using F.lux whatsoever.
That light is negligible. I have used F.lux for years, I swear by it, and if I bring an iPad or iPhone to bed and stare at it a while before I sleep (note: don't do this), I can totally notice the difference. (Then again I have a diagnosed sleep disorder and have to be extra careful about these things.)
The LCD in your monitor is a filter, that controls what colors from the white back-light reach your eyes.
It is not completely effective - when the LCD is set to block all colors, there is some white-ish leakage as you say. A completely red shape on an LCD would still have some blue in it from the leakage. We perceive color and brightness relative to the other colors close by.
Flux shifts the color balance of the screen towards red at night. After your eyes get used to it, the colors look about right, but the LCD is in fact blocking a higher proportion of blue than normal.
What it does not do is change the brightness of the screen - this is usually done by changing the intensity of the backlight. It is bad for you to look at a screen that is either too bright or too dark for the room you are in, so this is best done with an actual light sensor.
Here is a thread about doing that, but neither of the programs in it work very well. If you fancy making an open source program to do this, you'll be my hero :-)
When your monitor is displaying a white image, lots of white light is coming out. When your monitor is displaying a red image, lots of red light is coming out. F.lux makes white images tinged red. F.lux isn't about removing brightness in general.
Your clarification is confusing me even more, I don't understand what your confusion is at all.
When your monitor is displaying a red image, red light is coming out, but so is white light, because your monitor is not 100% light sealed. The backlight in your monitor is leaking through the edges of your LED screen.
So the question becomes, is that leaked white light enough to disrupt your sleep? No one's done any studies on this question, no one's looked into it, and so until they do, F.lux's effectiveness remains in question.
None of what you just said refutes any of the following: "When your monitor is displaying a white image, lots of white light is coming out. When your monitor is displaying a red image, lots of red light is coming out. F.lux makes white images tinged red. F.lux isn't about removing brightness in general."
Lets get some basics out of the way here, since you seem to be confused about either what "white", "blue", and "red" light is, or you seem to be confused about what f.lux actually does.
"red" light is light that consists largely or entirely of visible light with long wavelengths. "blue" light is light that consists largely or entirely of visible light with short wavelengths. "white light" is light that contains a good but not necessarily perfect mix of all visible wavelengths.
When you decrease the color temperature of an image, you are reducing the amount of short wavelength light, not eliminating it. LCD imperfection is frankly irrelevant as some short wavelength light is still coming through by design. If you removed all short wavelength light you would just be left with reds and blacks. That isn't what f.lux does, or claims to do.
> "So the question becomes, is that leaked white light enough to disrupt your sleep? No one's done any studies on this question"
Many studies have been done on the effects of different intensities of short wavelength light on sleep. This research isn't binary "blue light or zero blue light" For example, any of the studies which include an incandescent bulb cover this. Incandescent bulbs are black body radiators, they emit short wavelength light and long wavelength light, however the proportion of short to long wavelength light that they emit is much different from what florescent bulbs emit. (Specifically, light from incandescent bulbs is said to be "warmer").
By the way, one of polyfractal's links is to a study that used LCDs as the light source. Try reading before making an ass out of yourself.
Bring your laptop under your covers one night. Load up a completely black page. Can you still see (hint: probably)?
There is between almost none and a whole lot of white light that leaks out of monitors, regardless of what each pixel is set to let through.
Does that matter? None of these linked studies have examined that, to my knowledge.