A Camera Looks Sharp at Night Until Someone Moves

A bright scene can still lose the face

A clear parked car and a blurred moving face can appear in the same nighttime recording. The camera may have enough detail for the stationary background while exposure and processing lose the moving subject.

An RLC-811A owner on Reddit described good daytime video but blurred people on a footpath about 6.2 meters away at night. The owner later reported an improvement after exposure changes. That is a specific scene complaint, not proof that one copied shutter setting will work on another camera or property.

Resolution alone does not determine identification performance. Lighting, shutter behavior, lens, distance, angle, compression, and the subject’s movement all affect the recorded result.

Source: Rlc-811a night time images blurry and ghosting

Separate focus from motion blur

If stationary objects are sharp while moving faces smear, investigate exposure and available light. If the whole image is soft, consider focus, contamination, reflections, or other image-path issues. Axis’s image-quality troubleshooting guidance explains the relationship between low light, longer exposure, and motion blur.

Increasing frame rate is not the same as selecting a suitable exposure time. A faster exposure may reduce blur but need more light or produce a noisier image. Do not promise a setting that removes every tradeoff.

Source: Axis: troubleshooting image quality

Ghost trails can also come from image processing

Temporal noise reduction, often labeled 3D-DNR, combines information across successive frames to smooth noise. If a moving subject changes position between those frames, aggressive smoothing can leave a trail or blur detail. Keenfinity’s camera guidance documents that tradeoff. It is a possible cause of ghosting, not a diagnosis based only on the word ghosting.

Have the installer compare a controlled walking sample with supported noise-reduction, exposure and wide-dynamic-range settings. Wide dynamic range, or WDR, can introduce its own moving-image artifacts. Change one parameter at a time and compare saved footage; a darker or noisier picture may still preserve more useful movement detail. Do not copy another model’s numeric settings.

Source: Keenfinity: camera settings for video analytics, temporal noise filtering · Axis: troubleshooting image quality

Illustrative scenario: a person crossing the driveway

Imagine a camera covers a broad driveway from high under the eave. At night, a parked car looks clear, but a person crossing near the street occupies little of the image and moves quickly across it. A higher-resolution camera in the same position may still fail the identification goal.

A dedicated view nearer the intended decision point, suitable illumination, and a tested exposure strategy may be more useful. Keep privacy, glare, mounting, and environmental requirements in the design. License-plate capture can require its own carefully defined view and should not be assumed from a general overview camera.

Test the recorded stream

Pixel density describes how many image pixels cover a foot or meter at the target distance. DORI means detection, observation, recognition and identification: these are different planning tasks. The AXIS M4348-PLVE datasheet cautions that calculated distances still depend on motion, lighting, focus and compression. Use the density at the actual decision point as a planning check, then demonstrate the moving subject; a DORI distance is not a guaranteed identification range.

Use an authorized person to walk the expected route at ordinary speed after dark. Review the saved primary recording rather than only a low-resolution phone preview. Repeat with representative lights and vehicle headlights where safely arranged.

Agree on what the image must establish: detecting someone, recognizing a familiar person, or supporting identification at a defined point. Retain sample frames with consent and protect them as project evidence. A night demonstration should disclose the tested conditions and any unverified seasonal or lighting differences.

Source: Axis M4348-PLVE: DORI pixel-density definitions and limitations

Put it into practice

A decision table to keep handy