@JamesFadden Thank you for this wonderful overview of the research. It's one of my favorite slow moving debates. I hadn't seen the Wehr paper yet!

To answer your first question, which I think brings up a very interesting point, f.lux aims to do more than detect and match ambient light, which doesn't have anything to do with daily rhythms of human biology and circadian timing. f.lux runs the lights in our house, for example, and we do lots of interesting things with both cameras and sensors. Unfortunately when it comes to supporting millions of users it gets more complicated as all computer light sensors and cameras are a bit different (and lots of computers have no light sensors at all). We've got a number of things in the works, hope you'll like them.

@herf Thanks, I hope this will be soon. At least brightness changes depending on the time of day ;)

Yes we need an adjustment here - and we did tune this a little bit in the newest build, so let me know which version you have.

I do keep adjusting it because I am very worried that we are shortening our day by spending time indoors, by an hour or more. Last year f.lux was trying to stick to the "natural photoperiod" - i.e. more light around sunset - and now I've backed it off a little bit, so it "looks better".

If you can see a window around sunset, when you look outside and it is "getting dark" it is still often 500 lux outside (which would be "really really bright" indoors). The trouble is that only a tiny bit of this light gets through your windows (<1% in most cases), and so this setup means you might see an hour less "daylight" every day. I think this causes all sorts of unnecessary problems for people in the winter especially.

Not being able to see your screen is a bigger problem though, so we should make some more options here.

@TwoCables said in f.lux effects: Grey / Dark Windows. Suggestion:

The problem with the way f.lux works is, EVERYTHING would be changed. So, that means everything would be black and gray, including the icons. It wouldn't be good at all.

Thanks for your kind response. You are totally right, and that was something I thought before posting. But, honestly, I don't care much about the Windows icons and colors, especially at night. If I can activate this effect after 7 pm, will be fantastic!

@TwoCables said in f.lux effects: Grey / Dark Windows. Suggestion:

Or, use the latest version of f.lux and switch the color temperature to 800K using the keyboard shortcut Alt+Shift+Page Up and Alt+Shift+Page Down

This is wonderful! I'm gonna check it out, love hotkeys!

@lorna said in f.lux effects: Grey / Dark Windows. Suggestion:

We are excited that so many app developers are putting in night modes and hopefully this will continue. It sounds like you'd like a version of Darkroom that's closer to grayscale. Thanks for the request.

That's amazing! Gonna wait for it, I'm sure it will be a success. Keep on the good work, lovely app!

This is determined by the color accuracy of the display. F.lux can only do so much. The rest is up to the color accuracy of the display.

Generally, 1900K should be more than good enough for just about any display because that's an extremely warm setting. It's not blocking blue light. It's reducing how much is being sent to the display by simply changing the color temperature. However, again, the display's color accuracy will determine the actual appearance of the color temperature that you have selected, such as 1900K or 6500K, etc. Most displays aren't accurate. F.lux is at the mercy of whatever the display can do. So, it really doesn't matter what color temperature number you select, just as long as it's giving you the physical results in your body that you want. It's kind of like paying too much attention to the volume number when adjusting audio volume, or like paying too much attention to a house thermostat's temperature setting and allowing that number to determine your perception of the actual results.

So really, don't worry about it. Y'know? Don't place too much importance on the color temperature number that you have selected in f.lux. It's more for your reference so that you can know which settings work best for you with your display and your body. You'll never get any display to accurately or perfectly produce precisely 1900K (or whatever number you select), but that's not the point nor is it the goal. The point of f.lux and the goal is to avoid alerting/stimulating light when you're trying to unwind at the end of your day so that you aren't having your melatonin production interrupted or delayed. There's no specific setting that works precisely the same for everyone. You just have to find what works for you, but generally, 1900K is an extremely good starting point no matter how bad the display might be.

So, there's no right or wrong setting or anything like that. Just use what feels right. You'll know when it's right because you'll feel it. If you're in doubt, switch back to 6500K for a few seconds and see how you feel. Compare that feeling to how you felt when you had 1900K selected. I can almost guarantee you that even with a crappy monitor that emits too much blue light, 1900K should be more than enough to feel good and to enable proper preparation for sleep.

Of course, this is all assuming that you're not being exposed to alerting/stimulating light from other sources, such as your indoor lighting or the TV or other devices that can't use f.lux. F.lux is simply a way to enable you to use your computer at bedtime without being blasted by alerting light, assuming you've already done a proper job of avoiding alerting light from everywhere else in your home.

So yeah, it's not a problem that you have two different results from two different displays even though the same color temperature is selected in f.lux, just as long as the f.lux setting you've selected is causing the display to produce light that's not too alerting for you. To that end, make sure you also reduce the display's brightness a lot using the computer's brightness controls. I know that f.lux has a brightness control, but that's only to be used as a supplement to your display's actual brightness adjustments if needed because f.lux's brightness control only adjusts the black level. Your computer's display brightness control should adjust the brightness of the display's backlight.

Thanks for the comments - we are thinking about this a lot.

The main issue is that we are trying to give advice that has taken us years of study to understand, and yet the software is in a state where some people feel like they are losing control (they don't understand it or it feels too inflexible), and that's no good.

We are treating f.lux as a promise to help most people sleep better, and we take this really seriously. I think there is another path where we just do "what looks better" and don't take the sleep science seriously, and then you'd be in control, but we might not be able to do as much good, and we couldn't make it better over time.

As @nothreek notes, we are not doing a good job for shift work yet, so often people on these schedules are more expert at their schedule than the software is today.

Regarding adopting "always-the-same clock time" schedules: people adjust to the seasons very differently based on their biology and location, and this is the number one reason we don't have a "clock time" setting for adjusting nighttime lighting. This has been a really contentious decision - but for most people, we think this number should change throughout the year and based on your location and travel.

If we made a really prominent setting (as some companies have) where you have to choose "change at this time", then we would have to respect that setting forever, even if we later realize it's doing a very bad job in the winter for some people, or for people at the edge of a timezone. And actually, we are about 99% sure this is true - lighting that doesn't adjust with the seasons is going to screw some people up. So when there are things missing, it's because it ties us into a solution we think would be wrong in the long term.

There are indeed people who actually do know better, and we should not prevent full customization for them, but the interface is making that difficult because 90% of people do not dig in like that (they use the defaults), and so we are avoiding settings that would prevent us from doing better in the future. But we should use the wisdom of people who are very aware of this to help solve the problem.

I am sorry we are slow to ship the newer flexible scheduling we are working on, but I hope we can solve many of these things soon.

I suggest you to take a look at "https://fluxometer.com/". Maybe you see what I mean 👍

Also you can reboot after expanding your color range (f.lux may have prompted you a few times by now to do this), and it will also fix the problem.

But we might try to disable f.lux in this mode if nobody else has taken over prtscn by then.

'Reduce Eyestrain' is a specific color effect just like the other ones that are much more dramatic, like Emerald City or Blue Sky. It's just not as obviously different from plain old 6500K like the other color effects.

'Daytime Eyestrain' is just a preset for which color temperatures you'll see during the day, sunset, and during the bedtime mode. The main difference I see is, you get 5500K during the day instead of 6500K.

There are no rules to this, so there's nothing that's correct or there's nothing you 'need' to be doing for any reason. You just do what you like the most, and that's all there is to it. There's no real 'wrong' way or 'right' way or anything like that. Play around and use the settings that gives you the results that you favor the most.

@aleks Maybe go to the main website for that manufacturer? They have a LOT of different models, they make desktop machines (tower systems, called the Mac Pro), they make iMac which is a large 24-27" screen with the computer system built directly into the screen, and they of course make several different laptop styles. Click the "apple" icon in the very top left corner of your screen, and click "about this mac". What does it say?

Yes, this is a real, accurate measurement, and we think the model is reasonably "right" for circadian and alertness applications. It's possible that eyestrain may be helped by less intense tints, but we don't have a model to explain that.

The eye's visual system (the "three cone" system) is very good at distinguishing small changes in color, so there are a lot of colors between white and "no blue light". If you've painted your walls you'll know that one drop of pigment can make you see an "off-white" color, but this doesn't mean there is a lot of pigment there. This is because the three types of cones are doing very careful comparisons to understand small changes in color. And then, the same thing happens with filters -- we are very sensitive to colors that are "nearly clear".

On the other hand, your non-visual system (the blue-green-sensitive melanopsin cells in your retina, with input from the rods and cones) seems to care more about the absolute intensity of light, and it isn't relying so much on these tiny comparisons to decide what "color" a light is. This system is more sensitive to blue-green light, which is why we are talking about "colors" at all, but it's not making tiny color comparisons like your 3-cone system is.

The trouble comes when people try to reason that what they see visually must be the same as what their non-visual system is seeing, and they are just not the same. Small changes in whitepoint are probably not as important as people think.

For now we have to say that pretty noticeable changes are needed to make a difference.

Because these systems are different, as people get better at spectral engineering, some of these non-visual changes may be made so you can't even see them. Two recent examples include BIOS lighting and Soraa Sky:

http://bioslighting.com/bios-skyblue-lighting-technology/
http://www.soraahome.com/

BIOS is engineering lights that stimulate your body to think it's day without being "blue" and Soraa is making lights that are "darker" to your non-visual system while appearing more like a normal lightbulb.

Regarding the iPhone data, in the "Color" tab you can see 61.66 photopic lux.

And then M/P ratio should be 0.98.

For the melanopic calculation, we are using the "equal energy" calibration from Lucas et al 2014. Since Illuminant E has an equivalent CCT of about 5455K, this checks out.

@herf Thanks for that study link (Rahman et al. Sleep. 2014). It helps me think through two topics I've been unclear on: 1) significance of green vs blue light in circadian sleep rhythm; 2) distinguishing between general sleepiness (homeostatic sleep pressure, they call it) and the nighttime sleep drive or rhythm.

It actually doesn't matter. It's whatever is most comfortable for you. There is no right and wrong answer here. It is all up to you and what feels best to you. Remember, most monitors don't have good color accuracy, so your settings in f.lux aren't the actual color temperature that's being emitted by your monitor. The numbers are really just for your information - like reference points.

So, you have to experiment. You should also experiment with your monitor's brightness setting because even 6500K at a low monitor brightness can be far more enjoyable than a bright monitor at a very warm color temperature because of the superior color accuracy.

I don't want to influence your final setting too much, but keep in mind that you don't want the monitor to be brighter than the ambient light and your surroundings. If your monitor is the brightest thing in your field of vision, then it can cause eyestrain and eye fatigue and possibly a headache.

Look at the wavelength like I did. That's how I knew it was UV.