How Bright Should a Sunrise Alarm Be? Lux Explained

The short answer: what matters is the light reaching your eyes, and the range that keeps coming up, in third-party reviewer testing of popular devices and in the classic dawn-simulation research, is roughly 250–300 lux at the pillow. Not at the lamp. Not at the spec-sheet distance the vendor chose. At your closed eyelids, where you actually sleep. Almost everything confusing about sunrise alarm brightness comes down to that distinction, so let’s unpack it properly.

Lux is measured somewhere, and “somewhere” is everything

Lux is a measure of illuminance: how much light is landing on a surface. It is not a property of the lamp. The same lamp produces wildly different lux readings depending on where you hold the meter. A device that reads 2,000 lux with the sensor pressed against its face might deliver 100 lux at your pillow three feet away.

So when someone asks “how many lux is that sunrise alarm?”, the only honest reply is another question: measured from where? A brightness number without a distance attached is close to meaningless, and it’s the root of most disappointment with these devices.

The inverse square law, told plainly

Light from a small source spreads out as it travels, and it spreads fast. Double the distance and the same light covers four times the area, so each point gets one quarter the illuminance. Triple the distance, one ninth.

Concretely: if a bedside puck delivers 400 lux at one foot, it delivers about 100 lux at two feet and roughly 45 lux at three feet. Most people’s pillows sit two to four feet from the nightstand. That’s the whole trick, and it’s why a lamp that looks blinding in a product photo, shot from twelve inches, can feel like a dim ember from across the mattress.

What a real dawn actually delivers

For calibration: commonly published illuminance tables put the moment of clear-sky sunrise at around 400 lux, with full daylight climbing into the tens of thousands. Even a heavily overcast day sits near 1,000 lux. Your ancestors’ wake-up signal was hundreds of lux arriving from the entire sky: a light source so large that walking around the room barely changed how bright it felt.

That’s the reference point sunrise alarms are imitating, and it explains the research numbers. Dawn-simulation studies, including work by Michael Terman’s group at Columbia, typically ramped up to a peak of about 250 lux at the sleeper’s position. Independent reviewers who bench-test consumer sunrise alarms have converged on a similar figure, generally finding that devices delivering around 250–300 lux at pillow distance produce a convincing dawn, while devices managing only a few dozen lux at the pillow often fail to wake their testers at all. (Whether the whole category holds up is its own question. I dug into the evidence in do sunrise alarm clocks actually work.)

One clarification, because these get conflated constantly: 250 lux is a dawn simulation number. Bright-light therapy boxes are a different animal, specced at 10,000 lux at close range for a reason. I wrote up the differences in sunrise alarm vs. SAD lamp. They’re cousins, not twins.

Why vendor spec sheets mislead

Knowing the above, read a typical product page again. “Up to 300 lux!” Measured where? Usually at some short, unstated distance, sometimes 10–20 centimeters, occasionally not disclosed at all. By the time that light crosses your nightstand gap, the inverse square law has taken its cut, and you may be getting a tenth of the headline number.

I’m not accusing anyone of lying. It’s just that lux-at-the-device is the flattering number and lux-at-the-eye is the useful one, and spec sheets are written by marketers, not by the person squinting at a dim orange dot at 6 a.m. If a manufacturer publishes lux at a stated distance, take them seriously, Philips, to its credit, documents the SmartSleep HF3520 at 300 lux measured at 45 cm in its own manual, which is exactly the honest way to state it (and I can vouch the light itself is genuinely bright; I own one, it’s the device that got me into all this). Many vendors publish no lux figure at all, Hatch’s documentation gives none for the Restore 3, per its own support pages, which is why third-party light-meter tests exist. If you see a bare number with no distance, assume it was measured up close.

A wall of light changes the geometry

Here’s where my own build diverged from the bedside-puck formula. Horizn drives a multi-foot addressable LED strip: in my bedroom it washes light across the wall rather than beaming from a point. That changes the physics in two useful ways.

First, the inverse square law in its harsh form applies to small sources. As a source gets physically large relative to your distance from it, falloff softens. You’re never far from all of it at once. A lit wall behaves less like a lamp and more like, well, a sky.

Second, and I think more importantly: a big diffuse source fills much more of your visual field. Two hundred lux arriving from a glowing wall feels like morning in a way that the same meter reading from a six-inch disc doesn’t, because your peripheral vision is being told the whole world is brightening. That’s the effect a real dawn produces, and it’s the thing a point source at any wattage struggles to fake. If you’re the soldering type, I documented the hardware side in my DIY ESP32 + LED strip build guide.

Full disclosure: I have not put a calibrated lux meter on my own strip yet, so I won’t quote you a number I haven’t measured. When I do the bench test, I’ll publish it, distance included.

Measured brightness vs. perceived brightness: the gamma problem

There’s one more trap, and it cost me real mornings to understand. Suppose you nail 300 lux at the pillow and ramp your LEDs linearly from 0% to 100% power over thirty minutes. Congratulations: you’ve built an alarm that reads as nothing, nothing, nothing, POP.

That’s exactly what my first ramp did. The reason is that human brightness perception is roughly logarithmic: your eyes are exquisitely sensitive to changes at the dim end and nearly numb to changes at the bright end. Going from 1% to 2% LED power looks like a doubling; going from 90% to 100% is barely noticeable. A linear power ramp therefore spends its first twenty minutes in territory your visual system rounds to “off,” then detonates.

The fix is gamma correction: reshaping the output curve so that perceived brightness climbs smoothly instead of raw power. In Horizn’s render engine, gamma correction is applied before brightness scaling, order matters, or dim colors band and shift, and the ramp itself follows a tunable brightness curve with four control points, so I can shape how gradual the earliest, most perceptually-loaded minutes are. My build also runs a separate pre-dawn glow phase, a long deep-ember stage before the main climb, because those barely-there levels turn out to be where a fake dawn either feels natural or feels like a light switch with a delay.

The takeaway for any device, bought or built: peak lux is only half the spec. How it gets there determines whether you drift awake or get startled.

How to sanity-check your own setup

You don’t need lab gear. Phone lux-meter apps are imprecise but fine for relative comparisons: put the phone on your pillow, sensor facing the light, run the device at full brightness, and see what you actually get. If it’s a dim fraction of the box number, move the device closer, aim it at a wall to enlarge the apparent source, or accept that it’s a gentle supplement rather than your primary wake-up.

I built Horizn because I wanted the geometry, the curve, and the schedule under my own control, a room-scale dawn from an ESP32 and an LED strip, running locally with no cloud and no subscription, waking me every morning at 4:34. It’s pre-launch, but if this kind of light-nerdery is your thing, you can follow along at horiznsunrise.com.