Where to place CCTV cameras in an independent house

A camera in the wrong place produces footage that proves an incident happened without helping you do anything about it. A camera in the right place produces a face. The hardware is frequently identical.

This is the most valuable thing to get right in a home system, and the easiest to get wrong, because placement is decided in about ninety seconds during an installation while somebody is holding a drill.

Placement beats camera count

Four well-placed cameras will out-perform eight badly placed ones, and cost less to run. The reason is that useful footage requires a person to pass through a defined space at a workable distance and angle, and most houses have only a handful of places where that reliably happens.

Everything else is coverage for its own sake — the wide shot of the compound that tells you a figure crossed the lawn at 2am and nothing more.

The five positions that earn their place

• The main gate or entry, at face height. Everyone who enters passes through it. This is the single most valuable camera in the house and the one most often mounted too high.
• The front door approach. Not the door itself, but the two or three metres in front of it, where somebody stands, waits, and looks around.
• The rear or service entrance. The point most often used when the front is avoided, and often the darkest and least considered.
• The car porch or parking area. Vehicles, and the space around them, generate a disproportionate share of real incidents.
• The internal stair or hallway, if the house has more than one floor. One camera at the pinch point every route passes through beats one per room.

Beyond those five, additions should answer a specific question you can name. If you cannot say what a camera is for in one sentence, it is decoration.

Mounting height, and why eye level wins

The instinct is to mount high — out of reach, wide view, less obtrusive. The result is a camera looking down at the tops of heads, which is the one angle from which a human face cannot be identified.

For any camera whose job is identification, aim for roughly two and a half metres: high enough to be out of casual reach, low enough that an approaching face fills a useful part of the frame. Cameras whose job is detection rather than identification — a wide view of the compound, for instance — can and should go higher, because you are asking a different question of them.

The test is simple and worth applying before the bracket is fixed: stand where a visitor would stand and look at the live view on a phone. If you cannot recognise yourself, the position is wrong regardless of the megapixels.

Five placement mistakes we are called out to fix

• Pointing at a bright background. A camera facing a gate with the open sky behind it will silhouette everyone who walks through. Either move it or specify a camera with real wide-dynamic-range handling.
• Mounting under an eave with a light beside it. The light attracts insects, insects attract spiders, and a web across the lens looks exactly like a failed camera at 3am.
• Covering the boundary wall instead of the entry. Boundary footage tells you somebody was there. Entry footage tells you who.
• One camera trying to do two jobs. A camera angled to catch both the gate and the porch usually does neither at a usable size.
• Ignoring where the sun will be in six months. A position that works in July can be washed out in December. The survey should consider both.

What to decide before the drilling starts

Cable routes are the part you cannot change afterwards without visible trunking, so the sequence matters. Before anyone drills:

• Walk the site and name the job of each camera out loud. If two cameras have the same sentence, you need one of them somewhere else.
• Check the live view from each proposed position, at the height it will actually sit, at the time of day that matters most to you.
• Agree where the recorder will live — ventilated, out of sight, and not in the same cupboard a burglar would open first.
• Confirm the cable route for each camera and where trunking will be visible, before it is.

Half an hour spent on this is worth more than two extra cameras.