Here is a small scandal of physics: there is no such thing as "how bright a room is" — not as a plain physical quantity. Brightness, as humans mean it, is defined relative to a human. A lux is a watt of light weighted by how well an average eye sees each wavelength: green counts heavily, deep red and violet barely register, infrared not at all. The weighting curve, called V(λ), is an international standard — the CIE's system of physical photometry, currently published as ISO/CIE 23539 — and a professional lux meter is essentially a photodiode wearing a precisely tinted pair of glasses that imitate that curve.
A phone camera is not wearing those glasses. And that single fact decides what a phone light meter can honestly claim.
What the Light tool actually measures
Sentio's Light tool uses the camera as a broad light sensor: it averages the brightness of the whole frame, samples that average twenty times a second for two seconds, and reports the result — mean, range, spread, and the full two-second curve, filed in your catalog. The number on the screen is a percentage of the sensor's scale, not a lux value.
That's a deliberate refusal, and the reason lives in the camera pipeline. Between photons and the number you read sit three transformations a lux meter doesn't have:
- The wrong spectral glasses. The sensor sees through red, green and blue filters designed for pleasing photos, not through a V(λ)-matched filter. Two rooms with the same true illuminance — one lit warm, one lit cool — can read differently to the camera and identically to a lux meter, or vice versa.
- Gamma. Camera output is deliberately non-linear: the pipeline compresses bright values so photos look natural. Double the light and the camera's number does not double. A lux meter's photodiode is linear by design — double the light, double the reading.
- An unknown absolute scale. Even if you undid the gamma and fixed the spectrum, converting sensor counts to lux needs a photometric calibration this camera never received. There is no reference inside the phone to anchor the conversion.
So what is it good for? More than you'd think
Notice what survives all three distortions: order. If spot A reads higher than spot B, measured the same way moments apart, then A really is brighter — gamma is monotonic and the spectrum of daylight in one room doesn't lurch between your two readings. Ranking, comparing and trending are fully intact, and those cover most of what people actually reach for a light meter to do:
- Rank the windowsills before deciding where the fussy plant lives. "Bright indirect light" becomes: east sill > north sill > that shelf, by this much, measured.
- Compare the desk before and after the new lamp, the mirror film, the rearrangement — same spot, same time of day, a real before/after instead of an impression.
- Track a spot across the day. The two-second capture also records the brightness curve, so a flickering source or a passing cloud shows up as spread rather than silently corrupting the mean.
This is the platform's relative-features principle applied to light: absolute levels belong to the device; comparisons belong to the world. The tool is built to make only claims of the second kind.
What a professional illuminance meter buys you
A real lux meter is a simpler and far more disciplined machine: one photodiode, one V(λ)-matched filter, one cosine-corrected diffuser dome (so light arriving at a slant is weighted the way the definition of illuminance requires), and a linear amplifier — each of those properties characterised against ISO/CIE 19476, the standard that defines quality indices for exactly these instruments: how closely the filter tracks V(λ), how faithful the cosine response is, how linear the scale stays.
| Professional illuminance meter ($100–$2,000) | Phone Light tool | Where the gap bites |
|---|---|---|
| V(λ)-filtered photodiode — measures light the way a standard human eye weighs it | Camera RGB sensor with photo-oriented filters | Comparing scenes under different light sources (warm vs. cool, LED vs. daylight) — the camera's spectral error differs per source |
| Linear response, calibrated absolute scale in lux | Gamma-compressed relative scale, no photometric anchor | Any absolute claim: "this desk is 480 lux" needs the real meter |
| Cosine-corrected diffuser measures light arriving from all directions | A lens with a limited field of view, aimed somewhere | Workplace-style illuminance-on-a-surface measurements; aiming changes the phone's answer |
| Calibration certificate; usable for compliance with lighting standards | No certificate, by design | Office/workshop lighting audits, disputes, anything regulatory |
| One number, one job | Always in your pocket; every reading files under an object, with its curve, trendable for free | — this row is the phone's: the meter you have beats the meter you'd have to buy, for every question that's really a comparison |
Exposure: the photographer's version of the same discipline
Photographers solved this problem a century before phones: a handheld exposure meter is a light meter whose output is denominated in camera settings instead of lux. The deeper lesson from that tradition carries over — a meter is only useful when the thing it feeds stops second-guessing it. A camera on auto-exposure is a light meter wired directly to a dial that cancels out whatever the meter saw, which is why every level-sensitive tool on this platform begins by locking the automation. The measurement and the "helpful" adjustment cannot both survive.
The honest one-liner: the phone is a comparator, the lux meter is a photometer. If your question contains the word "than" — brighter than, dimmer than, brighter than last month — the phone answers it well. If your question contains a number with a unit and a consequence, rent the photometer.
The Light tool's sibling measures what the light is falling on — that's the Color tool, and the torch turns both into instruments with a light source you control: reflectance and translucency. For where phone measurement genuinely competes with professional gear and where it doesn't, see the honest accounting.