Despite the benefits of digital video that are obtained simply by interconnecting digital video devices to each other over Internet Protocol (IP) networks, installers and dealers have yet to embrace this new approach to security. And, by doing so, they may not be delivering the most beneficial solution to their end customers. So what does the security professional need to know about IP networking to realize the full potential of digital video for security and surveillance?

Digital video standards offer CCTV systems immediate solutions to the most common constraints of analog (i.e. costly, bulky cabling; unsuitability for long-distance video transmission; and loss of fidelity of video tapes -- see Exhibit A).

And despite industry misconceptions surrounding the challenges of deploying networked digital video on a large scale, the fact is that good-quality, high-capacity and reliable digital video surveillance can be achieved using standard networking equipment (see Exhibit B) and through an understanding of digital video compression standards.

The key to an effective IP-based CCTV system design and installation is the understanding that although the technology has changed from analog to digital, the basic concepts have not changed. A network is a term that describes a series of devices that are connected together by cables and patch-panels. In fact, an analog CCTV system consisting of several cameras, coax cabling, a video-switch and several monitors and DVRs is actually a type of network, albeit an analog one. IP video devices, however, operate over digital networks, and most of the complexity in IP CCTV system design is contained in the network infrastructure design (this holds true for analog CCTV design also, but the industry has been doing this for so long that it now seems like child's play). The most important facet of a digital network, for CCTV purposes, is its bandwidth capacity -- i.e. its video carrying capability.

Digital networks use an electrical interface standard called Ethernet to allow digital devices to physically connect to each other. Although networks have the reputation in the security industry of offering limited bandwidth capacity, these modern Ethernet networks use a technology called 100Base-T Ethernet to transmit data across CAT-5 cables and, in fact, offers a sustained data rate of 60 Mbps -- which is the equivalent of nearly 1,800 telephone lines. By interconnecting these 60 Mbps CAT-5 cables using a network switch, this capacity is multiplied by the number of cables the switch supports through its ports -- therefore a 24 port switch has a capacity of 24 x 60 Mbps = 1440 Mbps.

Although this amount of bandwidth could carry a handful of raw video channels, another vital technique for IP-based CCTV is video compression. By using modern video compression methods, a standard network's capacity is extended to accommodate hundreds and even thousands of video channels.

The compression of digital video signals is made possible through video codecs (codec is a term derived from COder/DECoder). Though many vendors supply their own codecs, the most common internationally standardized codecs are JPEG, H.263, and MPEG-4.

JPEG compression is the equivalent of taking snapshots and sending or storing the single snapshot. To compress a video sequence, many snapshots are taken and transmitted or stored. JPEG is therefore known as a Still Image Codec (SIC), and sequences of JPEGs are also grouped under a non-standardized "Motion JPEG" label.

Codecs like H.263 and MPEG-4 use a method called frame differencing to encode a video sequence. These codecs do not encode single frames -- they encode sequences of frames by looking for the changes between consecutive frames, and are known as Motion Video Codecs (MVC). The effect of using frame differencing is that the frame density in the resulting compressed signal is much higher than that of a SIC like JPEG -- offering smoother, easier to watch video and higher record-densities.

While the right choices have to be made on networking equipment (e.g. CAT-5 cables and Ethernet switches), the choice of the video codec used by the system is also of paramount importance. IP video systems that are based on the JPEG standard will require approximately 6 Mbps per channel of video to deliver full-resolution, full frame-rate video -- i.e. the level of video quality sufficient for both identification and low eyestrain on monitoring personnel.

IP video systems based on H.263, however, require less capacity to transmit video of equivalent quality. A typical H.263 codec can transmit the same quality video at data rates as low as 2 Mbps -- about 66% lower than JPEG systems.

The latest MPEG-4 solutions being offered by some vendors offer even better compression -- with data rates as low as 688 Kbps for matching quality. And future MPEG-4 solutions are promising data rates as low as 384 Kbps, making the decision to move to MPEG-4 an almost given in the industry today.

However, it is important to grasp that not all MPEG-4 codecs are equal. First of all, many vendors utilize the H.263 codec and "wrap" the data so that it looks like an MPEG-4 stream. So although the video has the MPEG-4 format, it has not been compressed to the low rates that a good MPEG-4 codec should deliver. And this is the second point: there are good MPEG-4 codecs, and the not-so-good. MPEG-4 is a standard that describes the compressed video's data format to the outside world and also defines certain properties that the compressed stream must have. The standard does not define how to compress the video or what data rates should result, given a certain scene. Hence, there is great scope for different MPEG-4 codecs to encode the same channel of video resulting in data rates that vary from the low to the very high.

Ultimately, the quality of the MPEG-4 codec defines data-rate and data-rate determines the amount of disk or tape storage required to store and archive recorded video. Despite the spiraling cost-reductions being achieved in the storage sector, the lowest data rate codec will always result in the lowest cost storage system, a fact which system designers and equipment purchasers should be aware of. This fact becomes more and more vital as the number of camera/days of recording starts getting larger and the storage cost starts swamping the equipment and installation costs.

Finally, there are some misunderstandings regarding the legality of MPEG series codecs due to the word "prediction" used within the standards. In MPEG-4 the term is used to describe a process that encodes the actual (i.e. real) changes to a scene -- there is no "guessing" performed in the encoding process: there is no "prediction" in the same sense as a weather forecast. The use of the term "prediction" within those standards may be seen as unfortunate for the purposes of the security industry -- since there is so much emphasis and reliance of video footage being an accurate and untainted recording of actual events -- with no errors or guesses tolerated.

Today's IP-based CCTV devices offer capabilities unparalleled by their analog counterparts. The decision is no longer analog vs. digital, but rather which digital vendor vs. digital vendor. Despite the many factors that need to be considered, the factor with the biggest cost implication is simply the quality of the video compression engine.

Exhibit A: Solutions digital video offers to analog's most common problems.

• Analog: Coax cabling used to transmit analog video signals is bulky and often expensive; Digital: Cheap, telephone-grade twisted-pair cables can be used to carry up to several hundred channels of video


• Analog: Long-distance transmission is expensive, requiring satellite, microwave, or dedicated wiring; Digital: Transmitting digital signals around the world is inexpensive by using the Internet


• Analog: Loss of fidelity of recording occurs as video tapes become "tired" and need to be replaced; Digital: Perfect recording;a tenth-generation copy will be as high quality as the original

Exhibit B: The following list outlines digital devices being used in digital installations and highlights the most important new digital components with traditional analog equivalents.

Component / Analog CCTV / Digital CCTV

• Video Source: analog camera; IP camera /Encoding Video Processor Unit (VPU)


• Video Display: analog monitor; IP monitor / PC / Decoding VPU


• Video Recorder (& media): VCR (VHS tape) /DVR (hard-disk); Networked Video Recorder (hard-disk, CD, DVD)


• System Controllers: CCTV keyboards / PCs; IP keyboards / PCs


• Main Connector: BNC; RJ-45


• Connecting Cable: coax; CAT-5


• Patch-panels: multiplexors / switchers; networking hubs / switches


• Short-distance Interconnect Devices: video repeaters & relays; networking hubs / switches


• Long-distance Transmission Methods: microwave, satellite, fiber, RF-modulation; routers, PSTN, ISDN, ADSL, the Internet, fiber, microwave, satellite