Night vision:
what you actually see

It's how far the LED still manages to put something on the image, measured in ideal lab conditions: no glass, no fog and nothing reflecting light back into the lens. A real installation almost never has those conditions: there's a pane of glass, a nearby streetlight that changes the contrast, or the lens simply isn't perfectly aligned with the LED. What really matters isn't that figure but what you can make out at the distance where something will actually happen, which on most façades and yards is a few metres, not the thirty or fifty the box promises. At that real distance you usually see a moving shape, not a face. To find out whether a camera will do the job, test it at night at the exact spot where it will be installed; don't trust the number printed on the packaging.

It depends on the location and on who lives near you; there's no single answer that suits every case. Infrared is discreet, invisible to the naked eye, records in black and white and bothers nobody in the neighbourhood, which makes it the default for a terraced or semi-detached house. A white floodlight gives you colour, which helps to tell clothing or a vehicle apart, but it turns the camera into a beacon that switches on suddenly at three in the morning, and next to a neighbour that's an almost guaranteed complaint. In an industrial unit, a private car park or a plot with no neighbours around, that same white light may be exactly what's needed, because as well as recording it deters simply by coming on. The right question isn't which is better in the abstract, but who is within twenty metres of your camera at night.

It depends entirely on whether there's ambient light where the camera is. With a streetlight nearby, a neighbour's porch light or the glow of a shop window, the colour sensor makes use of that light and the night image comes out in real colour, which helps a great deal in describing someone by their clothes. But if the yard or façade is completely dark, the sensor still needs light to capture colour, so without its own white LED or active infrared the image stays black or full of noise, however much the manufacturer calls it "24-hour colour" on the box. Night-time colour marketing doesn't create light out of nothing: it only makes better use of it when it's there. Before paying extra for this feature, check how much ambient light there really is at your particular spot at the time you care about.

It's the infrared bouncing off something very close to the lens: a fogged-up window on the housing itself, a cobweb, an insect stuck to it, or a wall or railing a few centimetres away that the LED lights up fully and that saturates the sensor. The camera literally blinds itself with its own light, like car headlights in fog. The fix is almost never a camera with more infrared range, because the problem isn't power but what's in front of the sensor. What you need to do is move the framing a few centimetres so the LED doesn't hit any nearby obstacle, or simply clean the housing if the white circle suddenly appears on a camera that used to work fine. Check it every few months, especially after rain or wind, because a new cobweb can render useless a camera that had been recording perfectly for months.

Not particularly, and it's one of the most common buying mistakes with night vision. Resolution measures how many pixels the sensor captures, but at night, in low light, those extra pixels fill up with noise rather than real detail, so a high-resolution sensor that's poorly lit produces a grainier image, not a sharper one. What does make a difference is the quality of the lens, which determines how much light gets in, and the infrared being properly aligned with the field of view rather than pointing at a nearby obstacle that bounces it back. Before paying more for more megapixels, check how that particular camera performs at night in a spot similar to yours; quite often a mid-resolution sensor with good optics and good infrared gives a more useful night image than a high-resolution one with mediocre components.

No, and it's a common mix-up among people trying to save by buying one camera for everything. The infrared on an indoor camera is designed for short distances, the room it's mounted in, and has neither the power nor the range to cover a garden or a porch of any depth. On top of that, and this is what really makes it useless outdoors, an indoor camera's housing isn't sealed against rain or outdoor humidity, so even if the night image were acceptable, the device will fail with the first serious storm or with built-up condensation. Outdoor means outdoor: a housing with the right weatherproof rating and infrared designed for longer distances. Fitting an indoor camera on a porch looks like a saving on day one and becomes a dead camera within a few months.

Not entirely, and it's worth knowing before you decide on a daytime test alone. A spot that looks perfect by day can fill up at night with the glare of a streetlight, the headlights of passing cars, or a shadow that doesn't exist in daylight because the sun is somewhere else. The camera is exactly the same, but the scene it captures changes completely with the light. The test that really counts is the one done at the time you need coverage, which is almost always at night or in the early hours, not at midday in full sun. If you can only check it during the day, at least ask to see real night-time footage from a similar model in a spot like yours before you decide, and don't rely on the datasheet alone.

Yes, when the case calls for it we check at night, especially on porches, industrial units and yards where a nearby streetlight or backlighting can completely undo what looked like good framing by day. A daytime visit is useful for seeing the façade, measuring distances and deciding where the cable will run, but it tells you nothing about how the sensor behaves in low light or whether the infrared will bounce off something you don't notice in daylight. In clear-cut cases, a standard home with no unusual light sources nearby, the daytime visit is usually enough. In doubtful ones, a porch with a streetlight right opposite, a unit with security lights that switch on and off, we'd rather come back at night than install and discover the problem afterwards, with the camera already up and the customer expecting it to work.