Cameras that went dark, and footage that was never there

Repair, in most cases, and we will normally try that first. A drive, a power supply, a switch, a connector, a bubble, a lens: these are parts and labor measured in hundreds of dollars against a replacement measured in thousands. A system that is eight years old and working is not automatically a system that needs replacing.

The honest triggers for replacement are narrower than the industry pretends. Replace when the manufacturer is gone and no parts exist. Replace when a camera can never produce identifiable footage from where it is mounted and there is no better mounting point. Replace when the recorder cannot hold the retention you are legally or contractually required to keep and cannot take larger drives. Replace when you are spending more per year in service calls than a new system would amortize to. Anything short of those, we would fix it and tell you roughly how long the rest of the system has.

Usually not, and it is worth understanding why so you can stop it happening again. A recorder overwrites the oldest data continuously; once a segment is overwritten the space has been reused and there is nothing to recover. Specialist forensic recovery from a physically failed drive is occasionally possible and it is expensive, sits with a data-recovery lab rather than with us, and offers no guarantees.

What we can do is establish exactly why the window was shorter than you expected (nearly always either a drive at the end of its life or a retention figure that never matched the camera count and bitrate) and then fix the number for the future. If a matter is likely to become a claim or a case, tell whoever holds the recorder to export and preserve the relevant footage the same day. Systems overwrite on schedule and do not wait for attorneys.

Four causes, in the order we find them. A spider web or road film sitting directly in the infrared beam, which scatters IR straight back into the lens and washes the whole frame. This is the most common single cause and it is a cleaning item, not a fault. Infrared bouncing off the inside of a dome bubble, usually because the light-seal gasket around the lens has been pinched or reassembled wrong; the permanent cure is a turret housing, which has no bubble at all. A bubble that has hazed, scratched or yellowed over years of weather. Or genuinely inadequate low-light performance for the position.

That last one is a design issue rather than a repair. If the camera has a small sensor packed with pixels, it will always struggle at night, because those pixels are physically smaller and collect less light. A 4 MP camera on a larger sensor beats an 8 MP camera on a small one in a dark corridor. Where the scene simply has no light, the choices are infrared (which costs you color, the single most useful descriptor of a suspect) or a camera with a fast lens and supplemental white light, which keeps color but is visible to everyone. We will tell you which trade you are making rather than quietly picking one.

Often, yes, and that is usually the cheapest way to improve a system. The questions are whether the recorder has free channels and free licenses, whether it will accept the resolution and codec of the cameras you want, whether the storage can carry the extra bitrate at your current retention, and whether the switch has both ports and remaining power budget.

That last one catches people. Adding four cameras to a switch that is already close to its power budget will not fail on the bench in October. It will fail the first cold night when heaters energize across every outdoor camera at once and the switch begins shedding ports. We check the budget arithmetic before adding anything, and if the answer is that you need a bigger switch, that goes on the quote with the numbers shown.

Yes, and it is worth fixing regardless of who does the work. Port forwarding puts the recorder’s web and streaming services directly on the public internet, where automated scanners index them within hours of being exposed. Combined with default or hardcoded credentials, which a great many camera models have shipped with, that is the precise mechanism behind the large surveillance-device botnets, and it is why cameras and recorders remain among the most exploited categories of connected device.

The replacement is not complicated: no inbound ports at all, remote access over a client VPN or a manufacturer’s outbound relay, UPnP disabled on both the router and the recorder so nothing opens a port without anyone deciding to, cameras placed on their own network segment with no route to the internet, every default credential changed, and unused services switched off. On most existing systems this is a few hours of configuration rather than new hardware, and it is the single highest-value thing we do on a takeover visit.

Yes. Most of our camera work is somebody else’s system. We will not, however, quote a repair on hardware we have not looked at, and we will not tell you the previous contractor was incompetent as a sales technique. A lot of what looks like bad work turns out to be a client who was quoted for what they could afford at the time, which is a different thing entirely.

What you get from the first visit is a written condition report: what is installed, where it is, what state it is in, what is failing and what will fail next, what the real retention is, and what it would cost to bring the system to the standard you want, in stages, with the order we would do them in and the reasoning. If the sensible answer is to keep it running for another two years and put the money elsewhere, that is what the report will say.