CCTV in the Real World:

Giving Users What They Want

by Robert Grossman

Some of the most usable CCTV systems I have ever seen were put together by people who said over and over again, "I don't know anything about this, but shouldn't we …" and then came up with a suggestion based on the specific application and common sense, not years of experience. Conversely, I have seen the same mistakes repeated over and over again by the "experts."

We all live in the real world, but few of us actually work there. Think of that the next time you push a door that you should have pulled to open, try to unwrap the cellophane from a new CD, or fill out a form by writing above the line you should have written below. There is a poster in many engineering offices that depicts a bumblebee flying and has the slogan "Achieve the Impossible" - any engineer can prove to you that a bee can't fly!

This article is divided into two sections. The first focuses on the elements of CCTV systems design that many of us don't think of but annoy the people who use our systems day in and day out. The second explores some of the high-tech features and buzzwords with an eye on what type of user benefits from each.

Understanding CCTV Users' Frame of Mind
To understand where a design can go wrong, it is important to step back and look at the evolution of the CCTV, consumer electronics and entertainment industries. While they are not a combination we ordinarily put together, there are several interesting parallels that brought us where we are today.

The original CCTV systems were built using equipment intended for other uses, primarily broadcast or industrial television. Cameras were large, expensive, power-hungry and required frequent maintenance. Tubes had to be changed every 12 months to 18 months, and the heat those units generated allowed this fledgling industry to look upon service calls as a continuing and lucrative revenue stream. Only the well-to-do purchased video surveillance systems, as the cost of buying and maintaining the equipment would otherwise cost more than the items to be protected.

In the mid-1960s, CCTV started to evolve as an industry. Two inventions made this possible: the "scanner," or motorized pan-tilt device, which allowed a camera to see in many directions, and the motorized lens, which allowed remote control of zoom, focus and iris adjustment. These inventions reduced the number of expensive cameras required to cover an area and legitimized the field.

Meanwhile, in the consumer electronics market, video rentals and the amateur videographer were coming on the scene. VCRs changed from a pricey luxury to a necessity and, inexpensive CCD cameras were married to lightweight portable VCRs to create the "camcorder."

Suddenly there became a mass market for products that had previously only been enjoyed by the well heeled. The dramatic reductions in price and improvements in quality continue today, with mass merchandisers selling do-it-yourself CCTV systems at the low end and major suppliers offering an ever-increasing array of features and bells and whistles at the high end. Now anyone can afford CCTV, including anxious parents who suspect their toddler is up to no good.

Broadcast television and movies were also becoming more and more sophisticated. Cable television, born in 1949 to serve areas where TV reception was poor or nonexistent, became a household word in the mid-1970s, with an easing of restrictions and increasing availability.

Further programming and audience growth came in the early 1980s, when the government moved to deregulate the industry. The plethora of channels available to the public raised the bar in terms of what was expected of video quality.

From this melting pot, steps forth today's CCTV user. He or she is not intimidated by buttons and will happily press them all day long. Crisp, clear video images are taken for granted as are time-shifting and reviewing video recordings. The available choices in equipment increase daily as the prices continue to drop. And, if that wasn't enough of a challenge, the blinking clock on the front of people's home VCRs is a testament to their unwillingness to read manuals.

Much as the consumer audio industry made everyone believe they were an audio engineer, the video boom has created a generation of people with very specific expectations of their video systems and a penchant for instant gratification.

No Matter the Type, Each Job Needs a Plan
In looking at our users and the events that conspired to create them, we must recognize the tremendous variety of situations in which they can be found.

CCTV today does not always mean the spy material of James Bond or protecting the vast sums of treasure found in a vault. It can mean a single camera in a convenience store, a traffic control system on the interstate, viewing license plates at a toll booth, guiding airplanes into gates, the high-tech systems found in prisons, casinos and nuclear power plants, or anything in between.

The more thought given to the intended use of a system, the closer we get to having that system function perfectly for a specific application. But whatever the use, there are several common elements to be addressed.

I divide a system into three functional areas, regardless of size: system viewing and control, device selection and placement, and event documentation. They are equally important to each application and, to complicate the issue, all of them must be addressed simultaneously.

System Viewing and Control Is Goal No. 1
The most critical information needed for designing a system is not what will be covered, how many cameras, or any of the other obvious questions. It is simply, "How will the system be used?"

The selection of devices, number of monitors, recording methods - all of these variables are dependent on the intended use of the equipment. It makes no sense to tape things that will never be played back, install pan-tilt cameras that will never be moved, or provide monitors that will not be looked at. The operation of the equipment is the driving force behind the selection of the equipment and yet it is often ignored.

I always try to see if the system will be used for surveillance or security. While these terms are often used interchangeably, I define them as follows: security watches things, while surveillance watches people.

If you have cameras in a warehouse and you are looking to see why you are coming up short on widgets every month you are using a security CCTV system. If you need to keep an eye on Ed Davis, the infamous widget collector, you are performing surveillance.

Surveillance systems rely more on control of devices, while security systems typically have more fixed cameras and recorders for later review and examination. While a system may be designed to perform both functions or have certain areas leaning toward one method or the other, a system designed strictly for surveillance will perform poorly for security and vice-versa.

To determine the intended use, ask some pointed questions. How will the system be used? Will operators be specially trained? Will the console be manned at all times? How important is the unattended recording of certain cameras?

Once the intended purpose has been determined, the environment may be designed to handle that functionality. But first, let's take a closer look at security vs. surveillance functionality.

Security: Capturing Images on Tape
Security Systems often rely on unattended recording. Since an operator may be called away at any given moment, the ability to go back and see the "tale of the tape" is critical.

Systems that are left unattended are strictly security systems. They are also characterized by more fixed than movable cameras and frequently rely on many monitors with the intent that an operator may see something out of the corner of his or her eye.

A security system will have more automation than a surveillance system. Cameras with associated alarms may automatically pop up on assigned monitors. Monitors can often sequence automatically through a sequence of related images (one monitor may show egress points, while another shows hallways). Integration with other systems, such as access control and intercom, is more prevalent here.

Operators on security systems are rarely dedicated to CCTV functions. This may be a part of their duties or a slot on the schedule rotation. Consequently, they are less familiar with the equipment and rely more on identification labels, graphical user interfaces and other ease of use features. On smaller systems, turnover may be high or system control may be as time permits, further reinforcing the need for ease of operation.

When designing these systems, keep in mind that the video recording is of primary concern and operational features are secondary. They are looking to view an event after the fact and will rarely catch anyone in the act.

The "wall of monitors" approach so familiar to appliance store aficionados is often very effective; the operator is looking for change rather than function and will often spot things amiss from the corner of his or her eye. Pan/tilt/zoom (PTZ) devices are often used as backup, and today's generation of domes with on-board alarm processing are very effective at acquiring a target automatically when events occur.

Integration with other subsystems must be clearly accompanied by the appropriate titling and annunciation. If not, the operator may believe that they did something wrong when the picture suddenly changes and spend the rest of their shift "fixing" the system so it doesn't happen again.

A byproduct of this type of the system is the "tinkerer." In my experience, there is always at least one operator who takes a keen interest in the inner workings of the equipment and seeks to further customize it. This may seem like a lifesaver in that someone will actually read the manuals and be able to perform routine programming or system maintenance. However, this presumes that ability matches desire; otherwise that individual can be responsible for all sorts of bizarre and unpredictable system behavior.

I have seen the tinkerer lock all users out of the system by deleting passwords, wipe out all programming, change the date format so March 1 becomes Jan. 3 (MM/DD/YY changed to DD/MM/YY) and a host of other harmless and not-so harmless puzzlers. A well-documented system configuration printout is critical in this application.

Surveillance: Observing and Taping for Backup
Surveillance systems are as concerned with how it happened as they are with what happened. They generally are looking for deviations from established procedures. When they see them, they watch further to see what is going on behind the scenes.

Fewer monitors are used and screen size is generally larger. Pan-tilt devices are more prevalent, with the overwhelming choice being the discreet dome, to prevent a subject from knowing that they are being watched.

Surveillance operators are generally dedicated to CCTV work and do not wear as many hats. They may not use as many features in the system as security operators, but the ones they use are accessed continuously.

When designing for this application, performance, speed and reliability are critical. Recorded video is used as backup and support, but is not the only tool in the arsenal. The ease of use features are often dismissed as distractions; constant use of the system breeds such familiarity that they are simply not needed.

Needs Drive Device Selection, Placement
Moving into the second of the three main system design categories makes their interaction more apparent. With the selection of devices, we also must consider the way they will be viewed, controlled and recorded.

If the system is to be unmanned for the most part, with a heavy reliance on recorded material, PTZ devices make little sense. Left to performing preprogrammed tours, they will invariably be pointing the wrong way when something happens. Many units will acquire targets rapidly but all are limited to only being one place at a time. Fixed vs. moving is one of the first decisions to be made and will vary from camera to camera, room to room and job to job.

With cameras, color vs. black and white is an easier decision. Monochrome cameras still offer better low-light response and a sharper picture, dollar for dollar.

Many applications will not permit color; sodium vapor lights commonly used in parking lots and garages cast an orange hue that will render any color information useless. Rooms with two strong light sources - sunlight through a window and bright fluorescent lighting - will not allow a color camera to accurately white-balance.

Color, on the other hand, aids tremendously in identification of people and objects. Sometimes these added visual cues can be more hindrance than help.

During a major incident at a well-known casino, I once watched surveillance operators describing a suspect to security guards attempting to apprehend him. The tan jacket, blue-patterned shirt and blue jeans were all accurately described and, after much confusion ("that jacket was gray, not tan"), the suspect was apprehended. What amazed me at the time was that no one even mentioned the eye patch the suspect was wearing, a clue that would have been less prone to personal interpretation.

While the color vs. monochrome debate is still an area of strong personal preference for end users, the trend is clearly toward color. Careful consideration of environmental factors will prevent your choice from becoming a mistake.

I am amazed at how many projects are installed without the use of camera housings. In my opinion, without exception, a camera should always be in a housing. Cost should not be a factor. If the project calls for a camera, it should warrant a housing. If the price is particularly sensitive, there are many low-cost units on the market.

A housing makes the system look better, protects the camera, hides the fact that a camera is missing if it has been pulled for service or maintenance, and makes sabotage much more difficult and time-consuming. While this is always the first sacrifice to budget constraints, it is one that I try my best not to make.

Intended Use Decides Event Documentation
As with device selection, the manner in which the system will be operated determines the method to be used. If the intention is to log events for later review, it makes little sense to have a camera that is not going to be recorded. And if no one will ever play the tape, why have the cameras (or recording) at all?

When the decision is made to record an event, careful consideration must be given as to how. There are many ways to accomplish this task, and many successful designs rely on a combination of technologies to meet their needs.

Since it affects so many parts of the purchase decision, our first step is to evaluate analog vs. digital technologies. The future is clearly digital, but no one can say for sure how long the "transition period" we are currently in will last.

To capture a moving image, the best alternative, from a quality perspective today, is always real-time (30 images per second of video per camera), high-resolution (S-VHS or equivalent) and uncompressed (no digital video artifacts). Any departure from this involves a sacrifice and, since almost all systems must consider expense and physical space, we wind up making a lot of sacrifices.

The first sacrifice is usually resolution or picture quality. For whatever reason, VHS remains the standard despite a 40-percent loss in picture quality (compared to S-VHS or many digital formats) out of the box.

VHS has the advantage of being portable (put the tape in your briefcase), infinitely scalable (buy another tape), less expensive and firmly established as a standard. No one will lose their job by choosing the wrong VHS VCR vendor. A bonus: VHS tape is accepted as evidence in most courtrooms, while digital technologies are still on trial.

In the analog tape world, we have some technology choices to make. Our best quality decision is to record 1:1, dedicating each camera to its own VCR. While all VCR manufacturers endorse this practice, it is rarely practical from a cost or physical space perspective.

Multiplexing Departs From 1:1 Recording
When we depart from 1:1 recording, we have two alternatives: screen-division (or "quads"), and time-division multiplexing.

In screen-division multiplexing, a quad divides the screen into four quarters (or quadrants) and places a separate, full-motion picture in each.

Benefits include the reduction in space required (four cameras on one tape), ease of review, since no decoding is required to play back the image (making these tapes compatible with all video playback systems), and the availability of the full range of motion.

Drawbacks include the loss of picture quality and the limitation of only four cameras on a tape. While still popular, this technology is rapidly being eclipsed by other ones described in this article.

A time-division multiplexer is another alternative that distributes multiple cameras into the full-motion video stream of 30 frames per second (fps), placing one camera in each frame.

A jumble of flashing images when viewed on a standard player, each camera's image is reassembled on playback by a decoder. Depending on the number of cameras recorded, the effect is either very close to full-motion video, or resembles the stop-action motion familiar to silent movie fans.

Benefits include a tremendous reduction in tape storage, with the industry standard being up to 16 cameras per unit (some specialty units are available which record more but they usually require synchronization of camera signals, an expensive proposition), and full-resolution video. These units offer digital manipulation of the picture with freeze frame, digital zoom and multiple camera displays.

Drawbacks include playback compatibility (a tape must be decoded to be viewed and units made by different manufacturers are usually not completely compatible) and the loss of fluid motion. The latter can be minimized by reducing the number of cameras fed to each multiplexer (three cameras are recorded at 10 fps, while 15 are recorded at two fps) and by units that increase the refresh rate (frequency at which a frame is recorded) based on motion or external alarms.

When time-division multiplexing is combined with either 24-hour real-time or time-lapse recording, further compromises to the refresh rate are made as the sidebar on page ?? illustrates.

Digital Eliminates Analog's Limitations
After looking at the various analog technologies available, digital recording starts to look like the proverbial Ginsu knife advertised on late-night TV.

What would you say if I could combine the functionality of a VCR and time-division multiplexer in one box? How about we add the ability to record for extended periods of time - measured in days, weeks and years, depending on your budget? What say we also add the ability to view the image from your desktop over your corporate LAN, either live or playback - and when you're playing back, the unit is still recording? Let's throw in video quality that does not degrade when copies are made. Finally, how about smart searching capabilities, where we ask the unit, "Show me when that PC disappeared off the corner of the desk during that long holiday weekend," and get our answer in seconds, without reviewing 72 hours of video tape?

Digital recording offers all of these capabilities, at ever-decreasing prices. While still high by VHS standards, many security clients believe the benefits have finally matured and can justify the added cost.

A word of caution; while a few vendors have had digital recording products on the market for several years, this is an emerging technology that, because of the relatively low development cost and standardized PC platform, has attracted many small companies that may not be around tomorrow. Be certain when choosing a unit that you also evaluate the manufacturer and its ability to support and enhance the product well into the future.

A Word of Caution About Displays
A final note in the design and selection of recording systems; pay attention to the time and date display. Many industrial VCRs have an internal time/date/title generator that superimposes the time over the on screen image. If you are using more than one video recorder, do not use this internal clock unless it can be synchronized with other units (another selling point for digital).

Virtually every major CCTV systems house manufacturer has an external product to provide synchronized time and date for multiple video sources. These will provide the same time and date generated by the surveillance system or will function as standalone equipment.

One of the main reasons for making a video recording is to provide evidence in case there is a civil or criminal legal proceeding. With the VCR's independently generated clock, the subject being viewed can walk from one room (VCR A) to another (VCR B) and, according to the tapes, arrive before they left. In some cases, they may arrive hours or even days before! If your intent is to gather evidence, this is usually worthless. Be sure to do what is required to ensure your material will hold up in court.

Go the Extra Mile to Benefit System, Client
One of my pet peeves is that too many people do 95 percent of a good job. The software industry has a saying that the first 95 percent of the job takes 5 percent of the time allotted, and the last 5 percent takes the other 95 percent of the time! Sometimes it seems that the work never ends and if you don't stay on top of it, it never will. I have never had a customer thank me for the 300 cameras that are working, but they will invariably tell me about the three that aren't.

I look at the last 5 percent of the job in two ways. There are things that must be done to clear the punch list, and there are the little details that make a system useful and differentiate your business from the competition.

Punch-list items are pretty straightforward - they have to be done if you want to get paid and they won't go away. If you ignore them, you can forget about all the hours you put in. Your reference from the customer won't mention the fine job you did but will dwell on the missing fill panels, scratched faceplate and the one VCR that eats tapes.

Little details will make your design work well. Do the overhead lights cause a reflected glare on the monitor screens? Was the air conditioning system sized properly? Are there work lights behind the racks and extra outlets for soldering irons? Did you leave space for equipment expansion? While none of these are officially your responsibility, you are perhaps the only person who could have brought them all together and prevented a problem.

Designing a system that is effective and fun to use is both frustrating and rewarding.
I recently went back to one of the first major casino systems I designed. I had not been there to see it in more than four years and was surprised to see how little it had changed. Little things were different; some monitors had been moved and equipment had been added. I remembered working shifts as an operator while the final punch-list items were being resolved, and making changes based on the complaints, comments and feedback I received.

The things I learned during that process have helped me with numerous designs in the years that followed; the fact that I was welcomed and not cursed as the designer of this original system showed that I must have done something right at the time.

As the technologies continue to evolve, the marketing and salesmanship of many features will continue to blaze the trail. With each new "next great thing," we must make sure to understand the limits and capabilities before putting these products in front of our customers.

While many people profess to want the latest and greatest, everyone needs their system to work as advertised. And regardless of technology advances, that concept is not likely to become outdated.

Note: There may be slight variations from this version and the one that appeared in Security Sales & Integration Magazine.

Originally Published:
3/29/2002, Security Sales & Integration Magazine