Buy wired cameras wherever you can run a cable, and wireless only where you genuinely cannot. A wireless camera has two failure points a wired one does not: the WiFi link and the battery. Both fail at the least convenient moment, which is usually the moment you needed the footage.
That is the whole answer. Everything below is why, and where the exceptions are.
What does wired actually mean?
Power over Ethernet. One cable carries both power and data, so the camera has no batteries, no separate power adapter and no dependence on your WiFi. It records to a recorder in your house whether the internet is up or not, and with a battery backup on the recorder and PoE switch, through a power outage too.
That last point is worth sitting with. A cloud-only wireless camera stops recording the moment your internet drops, and an internet outage is correlated with exactly the kind of weather and power events during which you might want a camera working.
When is wireless the right call?
Genuinely often, in three situations.
A location where cable cannot reasonably be run and never will be. A detached structure you are not ready to trench to. A rental, or somewhere you want a camera temporarily.
The mistake is not choosing wireless. The mistake is choosing wireless for the whole property because it was easier to install, and then discovering the limitations after something happens.
What actually goes wrong with a doorbell-plus-floodlight setup?
It tells you somebody was there. It frequently cannot tell you who.
Three specific gaps come up again and again. Resolution and positioning that show a shape rather than a face. Retention that depends on a subscription: without one, many doorbells save no recordings at all, or only a few hours, when you often do not know something happened until well after it did. And coverage that watches the front door while the approach, the driveway, the side gate and the shop are all unwatched.
For a house on a standard suburban lot, a good consumer doorbell plus one other camera is a reasonable purchase and you should not feel bad about it. For acreage, it is not enough, and that is where most people in this metro get caught out.
Why is acreage different?
Because the property is the thing being watched, not the building.
Around Coweta, Bixby, Collinsville and Skiatook, a typical property has a long driveway, a gate at the road, a detached shop or barn, and outbuildings a fair distance from the house. A camera on the porch covers almost none of that.
Distance is the design problem. Wireless bridges across a long span depend on clear line of sight, and trees, wet foliage and interference make them unreliable. Integration at Work, AV runs buried conduit with hardened cable or fiber to outbuildings rather than depending on a bridge, so the shop ends up on the same network, the same recorder and the same app as the house.
Conduit rather than direct-burial cable is worth insisting on. It means the next change does not require digging up the yard again.
How much retention do I actually need?
Thirty days minimum, and longer is cheap.
The reason is behavioral rather than technical. People discover problems late. A tool goes missing and nobody notices for two weeks. A neighbor mentions something that happened last month. A short loop, or no recording at all, has already lost it.
Recording locally makes long retention affordable, because you are paying once for storage rather than monthly per camera.
Do these need a subscription?
Not if they record locally. Footage lives on a recorder in your house, it is yours, and there is no per-camera fee. Remote viewing from a phone is included.
Cloud backup is worth adding for a small number of cameras where an offsite copy genuinely matters, but it should be a deliberate choice rather than the default architecture.
Will they work at night?
That depends far more on placement and existing lighting than on the camera specification, which is the opposite of what most product pages imply.
The most common failure is aiming a camera into a bright porch light or a driveway floodlight. The camera exposes for the bright source and everything else becomes a silhouette, so you get a perfectly clear image of a light fixture and an unidentifiable shape next to it.
Integration at Work, AV checks night performance on site after dark rather than assuming it from the datasheet, and adjusts aim and lighting together. This is also where landscape lighting and cameras end up being specified as one project.
What does a system cost in the Tulsa area?
A four to six camera system with a local recorder typically runs $4,000 to $10,000.
Acreage properties with eight to sixteen cameras, outbuildings and a gate generally land between $12,000 and $35,000.
What moves the number is camera count and type, cable distances, whether conduit has to be trenched, and how much retention you want.
What should I decide before calling anyone?
Two things, and they take ten minutes.
Walk the property and note every way somebody could approach it on foot or in a vehicle. That list is your camera plan, and it is usually longer than people expect.
Then decide honestly what you want the system to do. If the answer is see who is at the door, buy a good doorbell camera today. If the answer is know what happened on my property last Tuesday, that is a different system, and it is worth doing properly the first time.
Integration at Work, AV designs and installs security camera systems for homes across the Tulsa metro, including Coweta. Free in-home design consultation, and a firm number before any work begins. Call (918) 494-9675.
Troy Erickson is the owner of Integration at Work, AV in Tulsa, Oklahoma. He has been designing and installing lighting control, motorized shades, audio, video and home networks across the Tulsa metro for over twenty years, and works on everything from 1920s Midtown homes to new construction and acreage properties. Integration at Work, AV builds on Lutron and Control4 and keeps a showroom in South Tulsa.