Adding 4K video to replace older video formats has obvious benefits. But there’s more to be done before you get the most out of the technology.
By Duane Richendollar · March 17, 2016
THE EMERGENCE OF 4K VIDEO has many advantages over both the current high-definition (HD) video and the older standard video formats. 4K offers four times the resolution as 1080p HD, giving the user better image quality to detect events and identify what happened during those events. However, there are drawbacks: Failure to adequately plan and design your network infrastructure and storage systems will significantly impact its effectiveness.
In the past year there have been numerous 4K video cameras introduced to the market, and the security industry is showing an increased interest in this technology. In many instances, businesses are still adopting HD technology and making the leap to 4K at the same time.
Here’s the challenge for system designers, though: HD cameras require four times the storage and bandwidth as a legacy 480TVL camera, and a 4K camera requires four times the storage and bandwidth as an HD camera. The math is simple, but the planning is not so straightforward. Let’s take a look at the requirements of some peripheral equipment.
Data Switches: Adding 4K cameras to a video system could quickly overwhelm the existing network and recording infrastructure. Much of our current data network is legacy 100Mbps network and is designed for computing data, not video streaming. Normal user data is called “bursty.” User data is sent in bursts and then is low until the next burst. Data switches for user data have buffer memory built in so that if it receives more data than it can process at one time, it can store the data until it catches up. Video, on the other hand, comes in at a steady stream, and if it sends more data than a switch can handle, the switch never has a chance to catch up. This can cause dropped video and, in some cases, the switch will lock up until it is rebooted.
Consider two scenarios:
Example 1: Take 10 20-megapixel cameras sending at 3.5 fps, requiring about 112Mbps of bandwidth. If these 10 cameras are connected to a 100Mbps switch it will quickly overload the switch.
Example 2: There are two 100Mbps switches that have five 20-megapixel cameras apiece. This would require about 56Mbps per switch and the switch should operate accordingly, but these two switches connect back to a main switch. If that switch does not have the processing power, it will suffer the same fate as the first example.
There are several things to consider when choosing a switch besides the individual port speed. First, where will the switch be located? Is this a core switch or a field switch? Does it have the option for fiber connection? What is the switching capacity and buffer size? Is it manageable? Does it support QoS and VLAN?
Server/NVR: The same holds true for a server-based NVR. If it isn’t designed correctly, it will not be able to support and record at the levels needed to process the 4K or even the HD video. Things like bus speed and read-write speed on hard drives will affect how many video cameras can be supported and recorded on a single machine.
Storage Space: New large-scale video systems have 350 to 600 cameras, with many incorporating 5-megapixel and higher resolution cameras. With 90 days of storage and the space requirements adding up quickly to accommodate the greater resolution, it can easily reach into the petabyte range. We recently had a scenario featuring a system with 220 5-megapixel cameras with video that needed to be retained for 90 days. By recording at 5 megapixels and 7 fps for 90 days, about 600TB of storage was required. All the more reason a qualified security expert is needed to properly plan for these situations.
In most cases, a company’s IT department can help with the planning, but if IT support is not available you may want to reach out to other industry professionals for guidance and assistance if necessary. Having the sufficient infrastructure in place can make the difference between a remarkable video surveillance system and an ineffective one.
Power and Cooling: Power and cooling are often overlooked. Building a big video system indicates a lot of servers, switches (usually PoE) and storage arrays will be in place. These take power to function, air conditioning to keep cool and UPS products to keep them operational if the power were to go out. Planning for the power load, UPS load and cooling load is just as important to plan for as for the server, switches and storage design. The fastest servers and switches will not work without adequate power or proper cooling.
HD and particularly 4K video raise the bar on the level of image quality you can offer customers. But its implementation also impacts the rest of the installation chain (and costs to customers). For effective operation, a well thought-out system must take into account the bigger picture beyond just the cameras.