How to design your system

The most difficult part of setting up a security system is usually just figuring out where to start. Most manufacturers either overwhelm with marketing material or provide very little useful information about the features needed to make a decision. Below is a straightforward overview of the key factors that will determine what your system needs.

There is no single approach that works for everyone, but there are guidelines that apply to most situations. The process is: figure out where coverage is needed, determine the best way to get that coverage (wired or wireless), then look at camera-specific features that may matter (night vision, audio). Once those questions are answered, storage and other equipment requirements fall into place.

The basic hardware components of any system are the cameras, a computer to store the footage, and any additional equipment needed for an optimal setup. That additional equipment may include routers, Power over Ethernet (PoE) network switches, and separate wireless access points, which provide performance and security improvements.

How long will it take?

Setting up a security system requires a bit of planning before purchasing any equipment, regardless of the manufacturer or style of system. Following this guide will help clarify what to expect when choosing and installing the physical devices.

StreamShuttle handles the monitoring side, but the cameras still need to be physically installed. With a little planning, most external installations can be completed in a few hours. After that, it's just a matter of configuring the devices and adding them to StreamShuttle. This initial setup only happens once.

Already have a plan? See the recommended hardware.

Start by deciding where coverage is needed

The most obvious areas to cover are entry points. Beyond simply having a camera pointed at the area, it's important to ensure the angle is good enough to capture useful detail: faces, license plates, and other identifiable information.

• For example, a camera on the corner of a house may cover the front door area, but may not be close enough or at the right angle to capture that level of detail. An additional camera right at the door, such as a video doorbell, may be necessary.

Once the coverage areas are identified, the next step is determining how many cameras are needed and what angles will best cover each area.

Determining how a camera will cover an area

It can be hard to picture coverage without some kind of plan, but this is easier than it sounds. A tool to help generate these plans is coming soon.

For the sake of simplicity, the examples here use a basic four-sided house.

If the house plan is more complicated, don't worry. The dimensions don't need to be perfectly accurate. Getting the general angles right matters more than exact measurements.

The approach is to draw a simple floor plan and create paper cut-outs representing the field of view of each camera. Overlaying those cut-outs where cameras are planned gives a clear picture of what will actually be covered.

Camera connectivity types (wired vs. wireless)

Both wired and wireless cameras have advantages and disadvantages. Wired cameras are recommended where possible. They are significantly more resilient to interference and network saturation, and when issues do arise they are much easier to diagnose.

Considerations when choosing wireless cameras

• Wireless cameras can be jammed: It's relatively trivial for someone with the right knowledge to jam Wi-Fi signals, rendering cameras completely useless. This is not a concern with wired cameras.
  Example: Tom's Hardware | Video demonstration
• High-resolution cameras require high bandwidth: Wireless systems are much more limited here. A couple of low-resolution cameras may work fine, but adding more cameras while maintaining a stable wireless network becomes increasingly difficult.
• Wireless cameras still require power: A common assumption is that wireless means no cabling, but power still needs to reach each installation point if it doesn't already exist.

Wireless isn't all bad, it just requires more careful planning to ensure enough bandwidth is available. If more than a couple of wireless cameras are needed, a dedicated wireless router on a separate channel (used solely for cameras) is worth considering. This keeps camera traffic from affecting day-to-day network usage. With a wired setup, this isn't a concern.

Most wired and wireless cameras can be powered using PoE, which sends low-voltage power over the same Ethernet cable carrying network data. This simplifies installation significantly. Instead of running power to each camera location separately, a single Ethernet cable runs back to a PoE switch, which handles both power and connectivity.

Picking the right cameras

With coverage areas and connectivity type determined, the remaining decision is which specific features each camera needs.

• Resolution: More isn't always better. Resolution is one of the primary factors in determining storage requirements and how many cameras the network can handle. Higher resolution doesn't necessarily mean better image quality either.
• Audio: Is audio recording needed? Is two-way audio (intercom-style) a requirement?
• Night vision: Most cameras include infrared night vision by default. It works well for indoor use, but outdoor viewing distance drops considerably in night mode. Supplemental infrared lights can extend that range.
• PTZ (pan, tilt, and zoom): The ability to remotely move and zoom the camera.

To work with StreamShuttle, a camera needs to support RTSP, RTP, or the ONVIF discovery protocol. If the camera advertises any of those it should work. Check the Recommended hardware page to see which brands are known to be compatible.

Many of the cheap, no-name cameras on large consumer retail sites will not work. They require proprietary apps and force reliance on their own cloud services on a per-camera basis. See Why do it yourself? and Why choose StreamShuttle? for more on why that matters. If compatibility is uncertain, send us an email and we'll confirm.

Sticking with one camera manufacturer for most or all devices is worth considering, though it's not a requirement. StreamShuttle supports a wide variety of devices. The reason for the recommendation is purely practical: many manufacturers require their own apps for initial device configuration (timezone, resolution, frame rate), and using multiple brands means learning multiple apps, even if only used once during setup.

Determining what's needed to run StreamShuttle

StreamShuttle will run on most hardware. All that's needed is enough disk space and processing power to handle the number of cameras in use. If picking the right hardware feels overwhelming, the recommended hardware page lists equipment that is known to work well.

A few things to consider when choosing a computer:

• How long does footage need to be retained? In most cases, a few weeks is sufficient. Once the cameras and their resolutions are decided, storage requirements become a straightforward calculation.
  (Frame Rate x Resolution x Number of Days x Number of cameras) + (storage required for archived events)

Use the storage calculator to get a rough idea.

What storage is needed?

Storage disk performance tends to be the first bottleneck, so the number of drive bays available in a machine is worth checking before purchasing.

Ideally, the OS drive is separate from the footage and cache drives. It only needs to be large enough for the operating system (about 120 GB) and should be an SSD.

The footage drive should be at least 7,200 RPM. Most SSDs will also work, but standard hard drives are generally more economical. A single drive is usually sufficient for most home setups (six cameras, more than 30 days of retention, and 2 TB to 4 TB of storage). If future expansion is likely, make sure the machine has room for an additional drive.

A separate SSD cache drive (about 120 GB) is worth adding if losing footage during a power outage is a concern. Even a cheaper USB 3.0 SSD has worked well in testing.

What processor and memory are needed?

As a general rule, one processor core and 750MB of RAM per camera is sufficient to handle object, motion, and audio detection on a pass-through stream. If transcoding is required (adding timestamps, cropping video, changing frame rate), an additional core per camera may be needed.

Other useful hardware

• Power-on after outage: Most servers support this, but many home PCs do not. When enabled, the machine will automatically power back on once electricity is restored after an outage.
• UPS (uninterruptible power supply): A battery that keeps the server running for a few minutes after a power outage, allowing for a graceful shutdown if power doesn't return.
• Dual network cards: With two NICs, camera traffic can be isolated on a separate network with no internet access, without complex firewall configuration.