| There are several comments here pointing out that RGB devices such as phones and monitors usually cannot make a narrow band of light around 670 nm. But the article ties this to light absorption by mitochondria, which it says absorb [1] 650-1000 mm light. Two questions: 1. If you use a narrow band source of light, does it really need to be around 670 nm, or will anything in that 650-1000 range likely do? 2. Would it actually matter if there is also light outside this band? If it doesn't actually have to be narrow and only contain 650-1000 nm, then it might actually work with some RGB devices. There seems to be significant variation among RGB devices, though. jlokier linked to spectra for iPhone X [2] and iPad Pro 9.7/iPad Air 2 [3]. The iPhone X with Night Shift at maximum has a significant output about 650 nm. The iPads have very little above 650 nm. This suggests that RGB devices might work for this, but there is no easy way of knowing for a given RGB device unless you can get its spectrogram. From what I've been able to find, it looks like Philips Hue bulbs just miss, with their red falling off rapidly near 650 nm. Candles look really good for this [4], and I'd guess similar for fires. I wonder if this means that ancient people tended to keep better eyesight in old age then us, because every significant light source they had (sunlight, moonlight, firelight) had a lot of its energy above 650 nm? [1] OT, but why the heck does the "b" in "absorb" become a "p" in "absorption"? [2] http://www.displaymate.com/Spectra_41a.html [3] http://www.displaymate.com/Spectra_35.html [4] http://dev.informationdisplay.org/IDArchive/2015/NovemberDec... |
Your pupils will adjust to perceived intensity. So if the bulk of the light comes in at shorter wavelengths it'll contract and you'll get even less of the already small fraction in the range of interest.
On the other side in the infrared it's more a question of safety. You don't want to overload your retina.
So you want it visible but long wavelengths only, which boils down to a fairly narrow range.
That's assuming it actually works. Single studies and all that.