Can My Camera Identify a Face at Night?
Probably not at the distance you're hoping for. A camera's advertised night-vision range is almost always measured to Detect — confirming something is present — not Identify, where you can make out a face well enough to recognize a stranger. Under the real EN 62676-4 DORI standard, Identification requires roughly 10x the pixel density of Detection at the same distance. A camera that can "see" motion at 100 feet in the dark may only be able to identify a face at a fraction of that range.
The one thing to know: An advertised IR range (e.g. '100 ft night vision') is almost always measured to the criterion of 'barely detectable' — you can tell something moved — not 'identifiable,' where you can make out a face. Manufacturers rarely state which criterion they used, so two cameras both claiming '100 ft' can have wildly different real-world usefulness for actually recognizing who is in frame. Under the EN 62676-4 DORI standard, Identification requires roughly 10x the pixel density of Detection at the same distance — so 'detect at 100 ft' does not remotely imply 'identify at 100 ft.'
The gap between "detect" and "identify"
The DORI framework defines four escalating tasks with specific pixel-density thresholds: Detect (~25 px/m), Observe (~62.5 px/m), Recognise (~125 px/m), and Identify (~250 px/m). Because manufacturers almost never state which criterion their advertised range refers to, a "100 ft night vision" claim usually means 100 feet of Detection — confirming presence — not 100 feet of Identification. The two claims differ by roughly an order of magnitude in required resolution.
How to estimate your own setup
Divide your camera's horizontal resolution in pixels by the real-world width, in meters, that its field of view covers at your subject's distance. Compare the result against the DORI thresholds. We walk through a full worked example in the DORI Pixel-Density Decoder — treat any such calculation as a derived estimate, not a guarantee, since actual image quality also depends on lighting, aperture, sensor noise, and focus.
What actually helps
- Mount the camera closer to where you need identification-level detail (a doorway or gate), not just a wide driveway view.
- Prefer higher native resolution and a narrower field of view over a wider one at the same resolution — spreading the same pixel count across a wider scene lowers pixel density on any one subject.
- Add light. A brighter scene (spotlight, floodlight, porch light) improves usable detail at a given pixel count and reduces noise, even though it doesn't change the underlying DORI pixel-density math.
- Avoid shooting through glass — see Do Security Cameras Work Through Glass?
This page reports a published international standard and a derived calculation method — it is not a promise about any specific camera's identification range in your specific installation.
Frequently Asked Questions
Only if the subject falls within the pixel-density threshold for 'Identification' under the DORI standard (EN 62676-4) — roughly 250 pixels per meter of the subject — at your camera's specific resolution, lens, and distance. Most consumer cameras are only capable of Identification-level detail at a few feet, not at the 30, 50, or 100-foot ranges often shown in marketing photos taken at ideal, well-lit distances.
Because 'night vision range' in marketing overwhelmingly refers to the Detection task — confirming something or someone is present — which requires roughly 10x less pixel density than Identification under the DORI framework. A camera can genuinely detect motion at 100 feet in the dark while being completely unable to resolve a recognizable face at that same distance.
Use our DORI Pixel-Density Decoder: divide your camera's horizontal resolution in pixels by the real-world width (in meters) the frame covers at that distance, and compare the result to the DORI thresholds (25/62.5/125/250 px/m for Detect/Observe/Recognise/Identify). This is a derived estimate from the published standard, not a lab measurement of any specific camera.
Digital zoom (cropping and enlarging existing pixels) does not add real resolution — it makes the same limited pixel data larger and blurrier, and does not move you up a DORI tier. Only a camera with more native resolution, a longer physical lens (optical zoom), or being physically closer to the subject increases usable pixel density.
Yes, indirectly — better lighting (a floodlight, porch light, or the camera's own spotlight) lets a color-capable sensor produce a sharper, more detailed image at the same pixel count, and can reduce the IR-glare and low-light noise that otherwise degrade usable detail. But it doesn't change the raw pixel-density math of the DORI framework — a camera with too few pixels on target still won't reach Identification threshold no matter how well the scene is lit.