CCTV-over-IP: it’s not universal

On the face of it, IP surveillance seems like a brilliant idea. One that, potentially, holds many advantages over traditional analogue systems. Although it has some way to go before becoming a universal standard for CCTV, security managers can realise its benefits now without having to give up their analogue systems altogether.

One of the key benefits of IP surveillance is image fidelity. Unlike their analogue cousins, digital signals don’t degrade during transmission and storage, thus evidential images are bit-for-bit copies of the original images generated by the camera(s).

Also, there are significant savings to be realised by end users in terms of cabling and infrastructure. Unlike standard analogue cameras, multiple IP cameras can share the same network so you don’t have to bring as many cables back to your Control Room.

We mustn’t forget that IP network cameras also support remote configuration, saving time and money on adjusting cameras.

The human factor in CCTV systems is one that’s often downplayed at the design stage, but the fact is that a digital video recorder (DVR) can enhance the productivity and efficiency of your monitoring staff, enabling them to work far more proactively. A key feature of a good DVR is its inherent ability to search for video based on camera location and time (or date) at the click of a mouse, thus streamlining the investigation process.

IP surveillance offers you greater flexibility in the architecture of your system. As stated, it’s no longer necessary to bring all of your camera feeds back to a central Control Room, freeing you from the need for expensive leased lines. Place a DVR at your remote site and bring back only those images that are important.

Mixing analogue with IP
As phrases such as ‘IP surveillance’, ‘video-over-IP’ and ‘IP-CCTV’ have become the buzz terms in the industry, so more and more end users want a 100% IP solution. However, there are many issues which almost inevitably push them toward a mixture of analogue and IP network hardware.

Analogue and IP can work quite happily together. There are instances where an analogue camera or transmission solution is preferable to IP, in particular for long cable runs where latency is an issue or you’re unable to find an IP network camera that meets the specification for your particular job.

Depending upon the system manufacturer and the model DVRs can readily accept analogue inputs, so it’s not necessary to use an IP network in realising the benefits of digital video recording. One thing to remember is that, while you can run a DVR solution without an IP-CCTV network, the reverse isn’t true. If you build an IP-CCTV network you’ll almost certainly need an enterprise DVR.

One of the first choices to be made by the security professional as regards any CCTV system is the type of camera to be used for the installation. At the moment, one of the greatest drawbacks with IP-CCTV is the lack of choice in IP network cameras. Unlike the analogue CCTV camera market – in which there’s now a comprehensive range of products including domes, low light units, IR models and so on – the market for network IP cameras remains somewhat underdeveloped.

Specifying an IP network camera gives you the advantage of being able to configure it over the network, saving all the time that would otherwise be required to access the camera(s) physically and fiddle with the settings. In addition, it only requires a single cable for transmission and telemetry and, given the imminent development of power-over-IP, you might even be able to dispense with the power point as well.

For specialist requirements it may be necessary to specify an analogue camera, but that doesn’t mean the security manager has to give up on IP-CCTV altogether. A video server or compression module can link a new or existing analogue camera to an IP network, thus realising the benefits of CCTV-over-IP.

As stated, a distinct lack of any basic standards is a big stumbling block as far as IP surveillance is concerned. Before specifying your cameras, make sure that your DVR manufacturer is able to support its protocols. You may even want to get those assurances in writing before you buy!

In practice, there’s nothing to stop you from mixing analogue cabling with Ethernet cabling in the same CCTV system. Indeed, this may well represent the ideal solution if there are a number of analogue cameras sited at far-flung locations across the site(

Making good connections
IP networks are typically constructed using Category 5 Ethernet cable. Unlike analogue cabling – ie coax or twisted pair – which can stretch to 1 km or more, the longest your installer would be able to run a Cat 5 copper cable would be around 100 metres. If you need longer cable runs there are a number of low cost solutions out there, including fibre optic cables which can easily bridge distances of 1 km or more. Alternatively, it’s possible to link a series of Cat 5 copper cables together with an arrangement of network switches.

In practice, there’s nothing to stop you from mixing analogue cabling with Ethernet cabling in the same CCTV system. Indeed, this may well represent the ideal solution if there are a number of analogue cameras sited at far-flung locations across the site(s) to be monitored. Simply bring the analogue cameras back to the main building with coax and then convert to IP by way of a video server.

One of the great selling points of IP-CCTV – and at the same time one of the biggest myths – is that the system can sit quite happily on a local area network (or LAN), side-by-side with the rest of the ‘corporate IT traffic’. While this could indeed be the case for a small number of cameras on a non-critical LAN, as soon as you start talking about putting a large number of cameras on a mission-critical IT infrastructure your CCTV project is likely to be torpedoed by the IT manager.

There’s nothing IT managers hate more than streaming video over their LAN. It’s a real bandwidth hog! On a typical 100 mbit LAN, for instance, just six cameras can easily gobble up at least 25% of ‘resources’. On peak demand, that figure might rise to 75%. At this level of demand it wouldn’t take much to push the network beyond its capacity, resulting in a system crash. Even on a one gigabit LAN bandwidth is a scarce commodity, and IT managers could object if total bandwidth requirements exceed 10% of network capacity.

If you’re talking about a CCTV system of between 30 and several hundred continuously streaming cameras, you’ll quickly find that the corporate IT network is out of bounds altogether. This doesn’t mean you have to dispense with the advantages of IP-CCTV, though. A separate network using standard Cat 5 cabling is well worth looking at.

Bear in mind, also, that an IP-CCTV network is cheaper to install than an IT network simply by dint of the fact that it’s considerably less complex – regardless of what your IT Department might choose to tell you.

Effective compression: the reality
There are many codecs for compressing video, including JPEG, MPEG, Wavelet and other variants and proprietary codecs. However, judging the actual quality of codecs is something of a minefield, particularly as there are none written specifically for CCTV.

Most codecs work by throwing information away in order to achieve a constant rate of information. The so-called ‘bit rate’. MPEG – the most commonly used codec in CCTV – was created for broadcast or home entertainment applications and, when used correctly, certainly delivers good results.

One of the myths in digital CCTV is that the latest MPEG standard – MPEG4 – is better than MPEG2. The newer version must be more advanced and therefore better than the older standard, right? Wrong. We feel that MPEG4 is processor-insensitive, and not as good at coping with sudden changes in a scene (see panel ‘What is MPEG4? An end user guide’).

DVD-quality video runs at an average of 5 mbits per second, and we can take that as the benchmark for CCTV-quality video. Some people will try to tell you that 0.1 mbits per second is good enough, but there’s no substitute for bandwidth – and at that level of compression you have thrown away another 98% of your information.

The way in which your system responds to movement is important. If you’re only sending a still picture, then you don’t need much bandwidth. If you’re sending a series of rapidly changing images, on the other hand, you need a great deal more bandwidth.

A positive development?
Used sensibly, IP-CCTV is certainly a positive development, but it’s definitely being oversold. There’s not a corporate IT network around that can support a good quality IP-CCTV system, but a little DIY networking will go a long way. The real benefits of the technology are to be found in the Control Room, where there’s both increased efficiency and reliability.