Motion Triggers vs Continuous Recording

Before jumping into security cameras for your home, it’s worth understanding what you can expect to get out of them. There are a few different ways you can go in terms of features and performance, so aligning your requirements with these will avoid a disappointing experience later. The biggest thing that will impact your experience will be how the camera records.

Consumer security cameras will operate in one of two ways. Either using motion or other sensors to trigger recording when something of note happens (event based) or they can record continuously to some form of video storage device for later review.

Event-based Recording

Most consumer smart cameras will use this mode. When the motion sensor on the camera detects significant movement, the camera will be triggered. The camera or its backend hub or cloud service will then perform some light-weight analysis of the video stream to determine if it should record the event and notify you. This will be dependent on various settings you can configure to tailor your experience.

Pros

✔ Only record what you need allowing quick review.
✔ Simple installation with Wi-Fi and battery options.
✔ Reduced bandwidth and storage needs.

Cons

✘ Can miss some events
✘ Getting settings right can take trial and error
✘ Recording can start late

This approach can be very effective and provide a concise list of short video clips that cover everything that happened over time. How well this works varies by camera, placement, and the quality of your home network.

It is a simple fact that all event-based smart cameras will miss events from time-to-time. Good quality cameras will do this infrequently, but it will happen due to the limitations of the video processing algorithms and network fluctuations beyond your control.

Camera features will determine both the kinds of adjustments you can make to determine the types of objects to filter out, what you want to exclude, and how and when to be notified. The camera and its supporting services will also determine how well these features work. When object detection misses the mark it can fail to record something that it should have, or it can capture too much nonsense motion resulting in decreased trust of alerts and difficulty in finding events of actual importance later.

Placement can impact how well these smart features perform. Things like height above ground, lighting and shadows at different times of day, and Wi-Fi signal strength will all impact object detection and motion detection accuracy. These impacts will also vary by camera model and brand.

Network latency can delay or interrupt the start of recording if there is a latency spike at the wrong moment. Where the camera uses a cloud service or local hub device for storage or processing, good network connectivity is essential for these cameras to work properly.

Features to look for

To make event-based recording work well, there are a few things that I would consider mandatory.

• Custom motion zones

  Motion zones are a way to define which parts of the camera view is important to record. You might think you want to record everything, but there are often parts of the scene that will cause false positive motion events that you want to exclude. This can include large shrubs or trees that can trigger the sensor when it’s windy for example.

• Person detection

  The ability to filter you recording and/or notification alerts to just when a human is present is a powerful tool in keeping things relevant and avoiding notification spam. It’s important that the camera you choose can do this well as it will determine both how many false alerts you get, and how many important events get missed.

• Other object detection

  Most cameras will have person detection in some form, but for more control look for models that can detect other types of objects such as animals or vehicles. This allows you to further tailor recording and alerts for specific camera locations. For example, you probably want to detect vehicles on your driveway and not just people. A camera that can only detect people may miss important events like accidents or mail deliveries.

• Rich notifications

  Getting a text when something happens is a basic requirement, but having a thumbnail of the cause included makes notifications much more useful. You can see at a glance if the event is something you need to investigate or respond to right away, without having to open the app and review the recording or live stream. Is it the mailman? Your kids? A cat? You can move on with what you were doing. This is especially useful on a smart watch as you don’t even need to get out your phone.

Continuous recording

More traditional CCTV type camera systems may employ this method. As you’d expect these cameras just stream everything all the time to a remote storage system. Some may also do object detection and tag timestamps in the recording, and still notify you when motion is detected. These systems have the advantage of not missing anything that happens but can be more time consuming to review the footage later.

This is the way to go for those who absolutely never want to miss anything, but there is a higher up-front cost for the equipment and the installation given the wiring requirements and extra network hardware.

Pros

✔ Record everything, no missed events.
✔ Can be more reliable (no batteries).
✔ Can deliver higher video quality.

Cons

✘ Requires wired power.
✘ Higher up front cost for recording system.
✘ May not be available from preferred brand.

While there are a handful of camera brands that can do this to cloud storage, most of the time you’ll need a Network Video Recorder (NVR). This is a device that sites inside your home and provides a large storage drive for the video to be recorded onto. Cameras that use Wi-Fi for communication can be found for these installations, but generally you’ll be looking at Power over Ethernet (PoE) to provide both power and data connectivity.

Depending on the system, PoE cameras may connect directly to the NVR, or they may simply connect to your network and communicate with the NVR that way. In either case, the endpoint of the camera cable must be a PoE capable device, either the NVR or a PoE Switch in order to deliver power of the same network cable.

Keep in mind the power requirements of your cameras. Your PoE switch or NVR will need to have the wattage capacity for however many cameras you need to connect.

Features to look for

When going for an NVR solution there are some important considerations over and above just the cameras.

• Interoperability

  Some video recorders only support connection to their specific brand of cameras. You’ll want to look for a recorder that supports industry standards like ONVIF or RTSP. This will open up your camera options and allow you to get the best balance of price and performance.

• Storage

  As the NVR will be handling all of your video storage, keep in mind how much space you’ll need. The recorder will overwrite older video as the storage fills up, so this will depend on how many cameras you have, when resolution they record at, and how long you want to be able to review footage. 1TB of storage is a safe recommendation for most home situations.

• Resolution and frame rate

  Just because your cameras are high resolution doesn’t mean the NVR supports recording at that resolution. You’ll need to make sure the resolution you need is supported by both ends of the system, and that the frame rate is not limited at the maximum recording resolution. Some NVRs will sacrifice frame rate for clarity, but that’s a terrible compromise that you don’t need to accept.

Learn more about choosing and configuring security cameras at the learning hub.