Choose low-light cameras using moving footage, not a bright still

A colourful scene is not enough

The useful question is whether the camera captures the detail you need while the subject is moving. A bright static night image can still show a blurred face or unreadable moving vehicle. Define the task and distance first, then test it under the actual light. Do not buy solely on resolution, a low-light label or an image of an empty driveway.

Colour can help describe clothing and vehicles when there is enough suitable light. Infrared night imaging can produce useful monochrome footage where visible light is limited. Neither mode is automatically superior for every scene. Axis's camera documentation describes the interaction between exposure, gain and night operation; settings and conditions are part of the result.

Understand exposure in plain language

The shutter controls how long light is collected for each image. A longer exposure can brighten a dark scene but allow movement to blur. Increasing gain can brighten the image while adding noise. Lens aperture, sensor performance and illumination also affect what the camera can capture.

Do not assume that a high frame rate fixes blur caused by a long exposure. Frame rate is how often images are produced, while shutter time is how long each image gathers light. Ask the installer to demonstrate a walking person at the required distance, then inspect a frame from the recording. A usable face matters more than a smooth-looking overview if identification is the job.

Assess the lighting you control

Walk the site after dark. Note porch lights, street lights, shadows, headlights, reflective surfaces and areas that become dark when staff leave. Consider whether appropriate permanent lighting would improve both safety and footage. Have electrical work assessed and performed appropriately rather than adding makeshift power arrangements.

Built-in white light or infrared illumination can help, but its suitability depends on distance, beam, mounting and the scene. Nearby walls, eaves, spiderwebs or glass can reflect light back into the camera. A manufacturer's illumination distance is not a guarantee of identifying detail at that distance. Ask for a practical result at your proposed position.

Fit the lens and position to the task

A wide lens covers more area but allocates fewer pixels to a distant face. A closer or narrower view may be needed at a gate, doorway or transaction point. Mounting high can protect equipment while making the angle less useful for faces. Use an overview and a task-specific view where the requirements justify them.

Motorised zoom can simplify adjustment, but it does not remove the need to choose the correct position. Confirm the final field of view before installation is accepted. Test the subject walking towards and across the camera, not only standing still under the light. Include backlighting at entrances where daytime performance matters as well as night footage.

Compare footage fairly

Ask suppliers for the exact model and relevant settings, then compare footage captured under similar conditions. A laboratory lux figure, a promotional clip and an on-site recording are different forms of evidence. Do not assume brand-wide performance from one camera or a newer model number.

Use the recorded stream at the intended retention settings. Compression may obscure detail even when live viewing looks good. Check the exported clip on a normal display and pause it at the point where you need information. Record any compromises, such as black-and-white operation, added visible light or a narrower view. A clear trade-off is more useful than an unsupported promise of perfect night vision.

Keep a practical decision record

Retain a short approved sample from the night acceptance test through an authorised process. Note the date, route, lighting state and recording settings. If performance later deteriorates, compare against that baseline before replacing equipment. Check whether a light has failed, the lens is dirty, vegetation has grown or settings changed during maintenance. These practical causes can be cheaper to correct than a camera replacement. If the original sample never met the task, however, routine cleaning will not repair an unsuitable position or lens choice.

Specify acceptance and maintenance

Write the night-time test into the proposal: subject, route, distance, lighting state, recorded settings and required outcome. Ask what adjustment or equipment change is included if the test fails. Include cleaning and inspection because dirt, webs and altered lighting can degrade the result over time.

Use the planning questionnaire to document the scene, then check storage using the storage estimator. Low-light performance is a system outcome involving camera, lens, position, lighting and settings. Choose on demonstrated moving-image usefulness, with the compromises understood, rather than on the brightest promotional still.