The change is easy to miss until you catch it happening. Sometime after sunset, your phone's whites go faintly cream; an hour later the whole screen is unmistakably amber. Nothing is broken, and the display has not aged. The device is running a schedule you approved, probably without remembering it: from now until sunrise, it will emit less short-wavelength—blue—light, and the white point slides from its daytime setting toward the warm end. It is one of the most widely deployed color decisions in consumer history, and it belongs to the same family of quiet evening shifts you can watch in a street lamp that seems to change color over an evening walk—except this one is deliberate, scheduled, and happening inside a screen you control.
What actually changes
A display's white is a recipe, not a fact. Daytime-mode white on most phones is balanced to roughly the color of overcast daylight; when the recipe calls for white, the red, green, and blue subpixels fire in proportions that produce that balance. Night mode rewrites the recipe for a warmer white point—toward the color of candlelight—by trimming how much the blue subpixels contribute. Every color on screen shifts with it: blues darken toward gray-teal, whites go ivory, skin tones warm. The effect is global and uniform, which is why a screenshot taken at midnight looks fine when you view it the next day—the file was never orange; your screen was.
That screenshot fact is the cleanest demonstration that nothing physical is breaking. The shift is a running software transform applied to the panel's output—a remapping of the red, green, and blue drive values—so the image data and the emitted light part ways. A camera photographing the screen at midnight faithfully records amber, because it captures the light that actually left the glass; a screenshot records the file underneath, which stays neutral forever. One object, two records, and the difference between them is precisely the feature.
Because the shift is a subtractive change, what you lose is exactly the part of the spectrum that was loudest in the blue channel. That is the design: the goal is less short-wavelength light reaching your eyes in the evening, delivered by changing the display's entire colorimetric character, not by dimming alone.
Where the sleep story comes from
The rationale is one of the better-studied pieces of popular color science. The body's circadian system, which times sleep through the hormone melatonin, is not driven by vision's ordinary brightness meter. Controlled laboratory work in the early 2000s established that human circadian responses are unusually sensitive to short-wavelength light around the deep blue band—more sensitive, watt for watt, than the cone cells that read ordinary daylight would predict. Later controlled trials extended the finding to real devices: in one well-known study, evenings spent reading on a light-emitting e-reader, compared with evenings reading a printed book under similar room light, delayed melatonin release, shifted the body clock later, and left readers less alert the next morning. Against that backdrop, trimming evening blue is a defensible, mechanistically sane design response—modest in effect, and individual in degree.
What the evidence does not say
Two overclaims ride along with the feature, and honesty requires separating them. First, eye damage: ophthalmology bodies reviewing the question have found no good evidence that blue light from screens harms the retina—the amounts are simply too small—and the eye strain that follows long screen sessions is attributed mostly to reduced blinking and sustained near focus, neither of which an orange tint touches. Second, sleep itself: night mode is not a sleeping pill. The e-reader trials involved bright screens and full evenings of exposure; whether the smaller spectral trim of night mode meaningfully moves the clock when you use it while checking messages with the lamp already off is much less certain. Light intensity, duration, and timing all still matter more than tint alone, which is why sleep clinicians emphasize dimming and putting the device away rather than trusting the amber glow.
The cost nobody schedules
There is also a purely chromatic price, and it is the one color-interested readers should internalize. A screen in night mode is, by definition, no longer displaying what it claims. Photographs edited at midnight will look different the next day; paint colors compared against an online swatch will not match what arrives; anything color-critical, from product photos to fabric choices, deserves a display at its normal white point, in the room's actual lighting. The feature is best understood as a deliberate, reversible bias of your whole visual environment—useful in its window, wrong for the work. Judging whether a blue-leaning color has drifted toward gray under warm white is exactly the kind of fine boundary call a round of Blue or Gray? exercises under neutral conditions, for contrast with your screen's nightly amber interval.
One distinction untangles a common tangle: dark mode and night mode are different axes. Dark mode swaps the interface's light-on-dark scheme for dark-on-light's inverse—white text on black panels—changing how much light the screen emits overall. Night mode changes the color of that light. You can run each without the other: a bright-blue dark mode, a dim-amber light mode, or both, or neither. Third-party utilities popularized the spectral trick in the late 2000s before operating systems built it in at the system level, which is why the feature now arrives switched on by default on devices that never asked you directly.
Each stop is labeled with its approximate white point and what changed in the recipe; the labels carry the information so the strip is readable without relying on the swatch colors.