CCTV best practices for apartment communities.

A camera that produces unusable footage is not a security camera. It is a budget line that pretends to be one. The single most common discovery we make when a property calls us after an incident is that the cameras were recording the wrong thing, at the wrong resolution, at the wrong frame rate, with the wrong retention — and nobody knew until the moment they needed the footage. This article fixes that.

Houston security firm. Only multifamily. Nothing else. CCTV advice for an apartment community is not the same as CCTV advice for a warehouse or a retail store. Multifamily cameras have to balance evidence quality, resident-privacy expectations, fair-housing rules around how monitoring is communicated, and the brutal reality that the footage will be requested by a police officer or an attorney who has thirty seconds to decide whether your DVR was worth their time. We will work through placement, the seven angles that solve most incidents, retention and storage, image-quality requirements, privacy zones, and how to operate the system after install.

Start with what the footage has to prove

Before you place a single camera, decide what each camera has to prove. The framing is forensic, not aesthetic. For every camera position, write down the answer to three questions:

1. Identify — can a viewer match a face or plate to a known person or vehicle?
2. Recognize — can a viewer determine that two captures are the same individual at different times?
3. Detect — can a viewer determine that a person or object was present in the frame?

These three are not interchangeable. Identification requires roughly 80 pixels per face for a definitive ID; recognition requires roughly 40; detection requires perhaps 20. A camera placed too far for identification can still serve recognition or detection. A camera placed where you need identification but only captures detection is failing — even if the image looks fine to a human eye on the property manager’s monitor.

Walk every proposed camera position with this lens. Where do you need identification? Where is recognition enough? Where is mere detection acceptable? The answers determine focal length, sensor choice, and exact mounting position.

The seven angles that solve most incidents

Across the multifamily incidents we have worked, seven camera positions appear over and over. If a property has clean coverage at these seven, the footage answers most of what investigators and counsel will ask. Properties that miss two or more of these positions tend to be the ones that call us a day after an incident to ask if we can help find the missing pieces.

1. Vehicle entry — face shot

A camera angled to capture the face of the driver of any vehicle entering the property, at a resolution that supports identification. This is normally a fixed bullet camera mounted on the entry-gate pillar or a nearby light post, aimed across the gate at approximately driver-window height. Identification is the goal — you need to be able to match this face to a known resident, guest, or unknown.

2. Vehicle entry — plate shot

A separate camera, often dedicated, capturing the license plate of every vehicle entering. Plate cameras have different specifications — narrower field of view, faster shutter, infrared illuminator with anti-glare. Do not try to combine face and plate into one camera. The optics make it impossible to do both well.

3. Vehicle exit — plate shot

Often skipped, often the most useful for car-burglary and unauthorized-removal investigations. The plate of every vehicle leaving, regardless of whether the gate registered the exit on the access system.

4. Pedestrian entry — face at gate

Pedestrian gates need their own face-identification camera. Foot traffic at a multifamily property is a major investigatory channel for trespass and resident-dispute follow-up.

5. Leasing office entry

The leasing office or clubhouse entry is where unknown persons meet known residents during business hours. A camera here is forensic gold for fraud and follow-up.

6. Mail room and package room

The single highest-frequency incident location on most properties. The camera must capture face at entry, the cart or bag of anyone leaving full, and the package-locker bank itself. Three angles in this one small room are often warranted.

7. Pool and high-traffic amenity

Not because of premium incidents, but because of liability and after-hours trespass. The pool deck camera should capture the gate and the deck itself with a clear view of who is on deck at any time.

Beyond the seven: secondary positions worth having

• Stairwell exits to parking — the door, not the stairwell itself. The door is where the actionable footage lives.
• Elevator cab — specifically the interior. Lobby cameras alone do not capture what happens between floors.
• Dumpster and trash enclosure — a frequent dumping ground for stolen property and a frequent late-night gathering spot.
• Fitness-center entry — fob-only entry creates the appearance of access control; the camera proves it actually was used by the right person.
• Parking-lot aisles — one camera per aisle if cost permits. Wide-angle works here because detection (something happened in this aisle) is usually the bar, not identification.
• Maintenance shop / vendor entry — vendor sign-in is often the weakest control on a property.
• Move-in / move-out staging area — if the property has a loading dock or staging zone, expect periodic theft from staged belongings.

Resolution, frame rate, and the “looks fine” trap

The camera looks fine on the live-monitor wall. The DVR thumbnail looks fine. The exported clip, viewed on the police investigator’s laptop a week later, looks like mud. This is the most common failure pattern in apartment-community CCTV.

Three settings drive whether exported footage is usable:

• Recording resolution — not the camera’s rated resolution. The actual resolution being written to disk. 4K cameras with their recording stream capped at 1080p will export 1080p footage, no matter what the spec sheet says.
• Frame rate during incidents — many systems record at low frame rates by default to save storage. Low frame rate is fine for empty parking lots. Motion-triggered higher frame rate at the moment of an incident is what produces usable footage.
• Compression and bitrate — aggressive compression saves disk but smears faces. Configure the encoder for evidence retention, not storage savings.

Test the system by recording at the intended settings, leaving the recording untouched for 30 days, and then pulling a random clip and viewing it on a laptop not connected to the DVR. If you cannot identify a face at the chosen distances, the system is misconfigured — before any incident has ever occurred.

Retention windows that match how requests actually arrive

How long the footage exists on disk is a policy decision, not a hardware constraint. The right retention window depends on how requests actually arrive at apartment communities.

In our experience, police requests for footage typically arrive 5 to 21 days after the incident. Resident complaints sometimes surface within hours, sometimes within weeks. Insurance follow-up on slip-and-fall and similar claims can arrive 30 to 60 days out. The retention math is straightforward: 30 days is the absolute minimum, and 60 days is meaningfully better. Anything less than 30 days will routinely lose the footage before the request arrives.

Calculate storage based on actual recording settings, not the DVR’s marketing-friendly “up to” spec. A 16-channel system at usable recording settings typically needs 12 to 24 terabytes of effective storage for 60-day retention. If your install came with 4 TB of storage and a marketing claim of “up to 90 days,” the bullet point and the reality do not match.

Backup power and the gap-in-recording problem

The single worst footage outcome is a gap in the recording at exactly the moment of the incident. Two common causes: power loss to the DVR, and recording-channel dropouts caused by network or PoE issues.

• UPS the DVR and the network switch, not just the DVR. A DVR on battery cannot record cameras whose PoE switch lost power.
• Generator backup on properties where it is justified. At minimum, the DVR closet should have HVAC during outage events — heat is what kills DVRs during a Texas summer power loss.
• Monitor for channel-down events. Most modern systems can email an alert when a camera stops recording. Configure these alerts to a real inbox, not the office account nobody reads.

Privacy zones and how to talk to residents about cameras

Apartment cameras live under a different privacy expectation than retail or industrial cameras. Residents are not visiting your property — they live there. The privacy framing matters for three reasons: it shapes where you can and cannot mount cameras, it shapes how disclosure should read in the lease and in posted signage, and it shapes how attorneys read the system after an incident.

Concrete guidance:

• Never aim cameras into individual unit windows or balconies. Use the system’s privacy-mask feature to black out the portions of any frame that capture unit interiors.
• Public common areas only — hallways, lobbies, pool deck, parking, perimeter. Not gym lockers, restrooms, or laundry-room individual stations beyond entry.
• Post visible camera notices in common areas. This serves both deterrence and consent.
• Lease disclosure — the lease should mention common-area CCTV. Talk to your counsel about exact language. Boilerplate from a national property-management software is not a substitute for Texas-specific language.
• Footage-release policy — document who can authorize a footage release, when residents can request their own footage, and how police requests are handled. Inconsistent ad-hoc releases create their own liability.

Live monitoring versus recorded-only systems

Most apartment communities do not need live human monitoring of CCTV around the clock. The cost is significant and the operational benefit on multifamily is limited compared with patrol presence. What works better is recorded-only systems paired with motion-triggered alerts and a clear “pull and review” process when incidents occur. Properties with concierge desks can layer a live-monitor wall during staffed hours, which gives concierge officers situational awareness of the gates and lobby without committing to 24/7 dedicated monitoring staff. See our service page on CCTV monitoring for how we package this.

Common mistakes we see repeatedly

1. Camera-count theater — 64 cameras of which 12 are pointed at empty walls or aimed past the action.
2. One camera trying to do three jobs — identification, recognition, and detection at the same time. Each is its own focal length and angle.
3. Skipping exit-plate cameras — vehicle exit is where stolen property leaves the property.
4. Retention shorter than typical police request lag — the footage is gone by the time someone asks for it.
5. Aggressive compression to save disk — the footage exists but is forensically useless.
6. No backup power for the recording chain — the DVR is fine; the switch is dead; no recording.
7. No drill — nobody on staff knows how to export a clip until they have to.

Key takeaways

• Decide what each camera has to prove — identify, recognize, or detect — before you mount anything.
• The seven angles that matter most on multifamily are entry face, entry plate, exit plate, pedestrian entry, leasing office, mail room, and pool / high-traffic amenity.
• Retention should be 30 days minimum, 60 days strongly preferred. Less than 30 will routinely lose the footage before it is requested.
• UPS the entire recording chain, not just the DVR. Monitor for channel-down alerts.
• Apartment cameras live under residential privacy expectations — mask unit interiors, disclose in the lease, post in common areas.
• Run a pull-a-clip drill every quarter. Find the gaps before an incident does.