Is the claim of re-usable cabling infrastructure a valid benefit of IP surveillance technology?

The debate goes on around the *killer* reason to move to IP – this time, centring on a very common claim relating to the cost benefit of being able to eliminate proprietary system cabling.

Personally, I don’t think it’s a simple point, but I’ve seen a lot of sales presentations from equipment vendors where it pops up without any clarification as though it’s a “no-brainer” advantage. I’m going to explore the issue in some depth here, looking at the simple practicalities in commercial refurbishments and in new-build projects, and I’m specifically looking at systems over copper, not over fibre. Here in Ireland at least, fibre is rarely used for analogue systems except on very long cable runs and in town centre applications.

My aim is not to act as a *mythbuster* here, but I do want to present the practical side of this point so that people understand what the real benefits are. And I’m not even going to begin to talk about anything other than just cable, because we all know you can’t start streaming video from cameras over the corporate network; so we already know that we’re going to have to install separate patch panels, switches, fibre transceivers, servers, storage etc etc etc….don’t we?

Traditionally, CCTV systems employed a coaxial cable connecting each camera to the recording and control equipment. This would be a dedicated cable that could not be used for any other purpose, with each cable carrying video from just one camera; and although various multiplexing technologies are available that give you a little flexibility, you would generally need additional cables if you wanted to transfer data, audio or any other media to or from the camera.

Un-repeated video over coaxial cable can be run for quite some distance (around 230m for RG59 or 450m for RG6) without significant signal degradation.

In addition, cameras need power to work. Some operate from low voltage and some are mains powered, but in either case a second cable is required at the camera, either from a mains spur nearby or from a low voltage power supply, often installed in a convenient riser or (in smaller systems) back with the recording equipment. PTZ cameras require power to run the motors and cameras installed externally often require some “assistance” dealing with their environment in the form of anti-condensation heaters or cooling fans, and in these cases those additional components are often mains powered for convenience (and to reduce the loading on low voltage power supplies).

Careful consideration needs to be given to power sources for analogue CCTV systems, particularly in larger installations where cameras in different parts of a development might be fed from different phases of the mains supply or where there may be earth potential differences. That at least is not quite such an issue with IP systems.

Fully-functional pan, tilt and zoom cameras also require a telemetry connection from the camera head to the control equipment, and whilst systems are available that transmit telemetry down the coax, it remains common to have a separate screened-twisted pair cable running to each of these cameras specifically for telemetry. IP systems have the great advantage that the telemetry comes down the same network connection up which the video is transmitted.

Actually, it is also possible to transmit video over twisted pair using either active or passive differential line drivers – commonly known in the CCTV world as “baluns” (balanced/unbalanced – like “modem” – modulator/demodulator…groovy…). In fact you can achieve even greater cable distances this way (around 300m on CAT5e with a passive balun and a kilometre or more on CAT5e with certain types of active transceiver). Baluns are so cheap (and coax is so expensive) that it often works out less expensive to cable your analogue CCTV system in CAT5e than it would to use coax; and if you’re using low voltage cameras you can use another of the pairs to power the camera…and another of the pairs to run telemetry.

…oh, hang on. So regardless of whether I’m putting in an IP system there’s a cost advantage to using the same cable I use for voice and data? Well if that’s the case then I can’t count this as a good reason to switch to IP, can I?

Have a look at this document for a good succinct comparison between these cable technologies.

I’ve actually had a lot of clients over the last few years (people with whom I’ve spent a good amount of time extolling the wonders of IP) who’ve decided to run UTP instead of coax so as not to deny themselves the opportunity of moving to IP at a later date, but they’ve installed DVRs and cheap analogue cameras with passive baluns for now, because the system cost is so much lower.

Given the choice we’d still rather cable an analogue system in coax anyway, so let’s have a look at that situation.

In a new-build there will normally be an electrical installation contractor given the job of installing all cabling and containment, including voice and data cables, and security cables.

To the guy pulling a whole pile of cable along a tray it makes no difference whatsoever whether he’s installing CAT5e or coax. RG59 is roughly the same dimensions as CAT5e although slightly more flexible. The biggest cost on small to medium sized projects is the cost of the man, not the cable, even though coax is a lot more expensive than CAT5e. Sure, there is some advantage to having him pull all the same cable from a procurement perspective (not price, just simplicity) and he’s got less to think about as he works (which may reduce mistakes a little bit), but I’d still contend that these are very minor advantages that probably have an infinitesimal bearing on the overall cost of the cable installation in this type of project. So except in very large projects there’s no significant cost advantage in moving to IP over CAT5e if you just want to save money on the cable itself.

Often I’ve seen vendors throw in the one about reducing the number of power points the electrician will have to install. I don’t actually believe that one at all. You can’t power heaters for external cameras or PTZ motors from PoE, so the most efficient way to run them is with a separate mains power supply fed locally to the camera (just like in an analogue system). For low voltage cameras I don’t see that it makes any difference whether you’re providing mains power to a PoE switch feeding a number of cameras or to a low voltage power supply feeding a number of cameras.

In fact, because of the cable length restriction on Ethernet, you may find yourself having to put in more comms closets around the site, each with its own switch and therefore each with its own power…so, it doesn’t necessarily stand to reason that you’re going to save money on mains power either.

Oh yes, while we’re on that subject don’t forget about the cable length restrictions for Ethernet. I don’t find this particularly a restriction – especially not since there are so many cable length extending products on the market now anyway – but I do tend to find myself thinking in terms of fibre when planning IP video network far more than I ever would have with analogue cameras and DVRs; both for backhaul from remotely situated comms closets and for distant external cameras, it is *simpler* to go fibre, but this does add cost and complexity that wouldn’t have been there otherwise.

Perhaps, then, the cost benefit of moving away from coaxial cabling is a scale thing. Perhaps we only really begin to feel the love when we’re reducing all those individual camera cables into a single backhaul connection. But from a risk management perspective, if you’re moving from a big bundle of cables to a single cable with lots of cameras on it you’re increasing the potential impact of a single point of failure event, and as soon as you start putting redundancy into the backhaul connection you add in cost that probably outweighs the money you’re saving on cable.

Certainly, the fact that we only need a single point-to-point network connection from each camera cluster back to the recording server has saved us a lot of effort in some of our more distributed installations. But I’m talking about campus applications where there’s already a major investment in resilient network infrastructure on site. I don’t believe this stands up in more typical commercial security surveillance installations.

So what about when we’re looking at refurbishments and system expansions? Moves, adds’n’changes is what they call ‘em. Once again I’ve seen lots of sales presentations advocating IP that point to this as another *obvious* advantage, but…really?

Those spare data ports all along the office wall are no good to you. You don’t want to fit a camera on the dado looking at someone’s desk, you need a data point up somewhere useful…somewhere where there isn’t one.

Granted, there is likely to be a comms closet within 100m that you can patch a cable into, but really the major expense in cabling an additional camera is in the last few metres of cable where it has to get through a wall or down a partition or across a beautifully plastered, seventeenth century ceiling…and it doesn’t matter whether you’re IP or analogue, that’s going to be painful.

PoE isn’t really an advantage in these circumstances, because with a balun you’d only need to run one cable anyway, and if it was an external or PTZ camera you’d probably have to get mains to the camera too.

Okay. So is that it? Is there no advantage to IP from the cabling perspective? Well, put it this way. I have not once sold an IP based system and won the project on the basis of reduced cabling cost, regardless of the system size.

Axis are under the impression that the cost crossover point between analogue and IP is at 32 cameras. I think that’s nonsense. As of the time of writing, here in Ireland, the installation cost of an IP system is always higher than that of an equivalent analogue system.

That is, where the coverage, resolution and camera quantities are the same, where there are no particular connectivity issues that are easier to overcome with IP (i.e. multiple media types to be connected over a single link or getting connectivity from one side of the city to the other etc), or where there are functions required that cannot be achieved with the analogue system without increasing the quantity or specification of the cameras.

I’ve seen a lot of cost comparisons done on systems, but in general the person doing them is trying to make a point and they compare the cheapest IP system with a much higher class analogue system, thereby skewing the results.

Moving to IP can not be about saving money on the installation.

It can be about reducing lifetime cost, but only with the right choice of equipment. Closed platform NVRs and the like shoot that advantage in the foot.

The advantages of IP relating to cabling and connectivity as I see them are as follows :

• It’s neat. You get an ever decreasing number of cables as you get back towards the control and recording equipment instead of an ever increasing number. Without all the cross connecting associated with cameras powered from discrete power supplies, this makes for a neat and tidy installation. Neatness makes things easier to support

• One cable does everything – even the things you haven’t thought of yet. Video, audio, telemetry, data, power, alarms, you name it – access control, SIP telephony, boiler monitoring, water level alarms, thermostatic control, gate automation. Just one cable. Neatness once again.

• Re-distribution of media to PCs on a network is easy. It doesn’t matter where you decide to put your control room or where you decide you want a monitor, you can display whichever camera you like without having to run additional cable or physically re-patch anything. It doesn’t matter how you decide to re-configure your video matrix, which cameras you want to display or when, you can do all the reconfiguration virtually, without unplugging a single camera.

I am a firm advocate of IP based video systems but I’m also a firm advocate of honesty and reality. Our responsibility as security professionals is to deliver cost effective security solutions to people with security problems, and I get a bit tired of seeing a lot of equipment vendors make unfounded claims.

They’re not stupid, those customers, you know? If you’re going to go out there and sell them something, you can’t base your argument on something that’s demonstrably not true.

Let’s make a concerted effort between us to distil out the truth about some of these technologies so that we’ve got something really compelling to tell our clients.

Tags: cabling, IP CCTV

General comment | RSS 2.0 | Respond | Trackback |