SWS Security
White Papers
"In God We Trust. All Others We Monitor."

From the Listening Post: #5

Questions from readers and answers from a specialist
on electronic surveillance and related topics.

Hi all,

This month weíll cover a number of closely related questions pertaining to the relatively new digital video.

Many of us use CCTV for physical security, surveillance, management information and related tasks. CCTV (Closed Circuit TeleVision) is a mature, reliable, inexpensive method of extending our eyes from one location to another via electronics. Its use is increasing, certainly reaching billions of dollars a year, worldwide.

Early on, all video was analog. Humans are analog. Our speech, eyes, ears and all senses are analog. Analog video fit our biological functions well. Analog was cheap. It was simple. It was effective. It was reliable. The average technician could understand it, and repair it.

However, technology is changing. While most CCTV cameras and monitors still are analog deep down, transmission technology and storage media have been migrating to digital.

To a large extent, being as our bodies accept analog inputs, we generally still convert audio and video to analog outputs so our bodies can process them. However, in between, digital processing rapidly is making inroads and soon will be the exclusive method used.

Take a look the next time youíre in your local department store. See if any Ďold styleí VCRs are for sale Ė yeah, the things which use tape. Bet you donít find any, or at best one or two low end models. But there will be an elaborate selection of DVD players, satellite TV receivers and more recent toys. A decent name brand DVD player with fancy features is cheaper than the cheapest VCR was a few years ago.

Analog tape, both for consumer and professional use, is nearing obsolescence. One roughly could compare it to audio cassettes being replaced by CDs.

We havenít installed analog tape in a professional security system in maybe 2 years. While we continue to service analog systems, all recent work we do uses DVRs, or Digital Video Recorders. As the name implies, DVRs are the current storage system for video, where timelapse or realtime analog tape formerly was used.

DVRs are recorders which record video, and sometimes audio or other data, digitally. They need to take the usually analog input from the camera(s), convert it to digital, compress it, and store it on a digital storage medium.

DVRs almost always use hard disk drives for their storage. Unlike tape, which was sequential access (you had to read everything on the tape in sequence to get to what you needed), hard drives are random access. You can go directly to information you need without having to review any other data.

Hard disk storage is inexpensive even in large quantities, and has proven to be extremely reliable. The enormous base of PCs has driven development of large, inexpensive, reliable drives. Hard drives now are mechanical, but I suspect we soon will see completely solid state mass storage devices as the price of silicon memory continues to decrease.

Many DVRs will have several hard drives. The one we use here at the office has four. A single drive is 160 gigabits, and significantly larger drives are available. Generally, in professional DVRs, you simply can add as many drives as you require to give you the storage capacity you need.

Whatís the major differences or advantages of digital over analog?

In analog videotape recorders, the cameraís analog signal is impressed on the tape in an analog format. There arenít really any effective analog compression formats, so the video generally is limited to 6 hours at 30 frames per second (where 30 frames per second is considered realtime, and is what we see on broadcast television).

By using Ďtimelapseí technology, up to 960 hours could be placed on a single analog tape by slowing down the frame rate. Instead of 30 frames a second, youíd shoot a single frame every few seconds. For nearly anything other than a bullet flying through the air, a snapshot image every few seconds was adequate. Timelapse permitted you to go many days between swapping tapes.

Multiple cameras were combined onto a single tape using a sequential switcher, a quad or other pattern device, or a multiplexer. More on these later.

Hundreds of thousands of analog machines still are in service, and will be for some time. They are reliable, easy to use, inexpensive.

Digital, however, offers many advantages.

One, there are fewer moving parts. Tape machines are precision mechanical monsters, and itís only through good engineering and quality manufacturing theyíve been as reliable as they have been. This especially is true with the way the tapes and machines are maintained (not).

Tape is a complicated mechanism. Alignment is critical. Cleanliness is critical. There are a lot of moving parts. There are rubber parts which wear rather quickly. Heads need cleaning and periodic replacement.

With digital, most of these are non-issues. The only moving part in most machines is a self contained hard disk drive, essentially identical to the one in your computer. This drive is a sealed unit which almost always can be unplugged to transport your data or replace the drive.

Digital machines offer compression, which analog does not. The current standard, MPEG-4, is pretty efficient and allows fairly high quality digital video images while still maintaining a reasonable file size.

Iíve evaluated a number of DVRs from different manufacturers, and for the most part you can treat them as drop in replacements for their tape counterparts. Some have gone so far as to have front panels which emulate tape, to minimize any learning curve.

With analog tape, when you had more than one camera to record things got complicated quickly. There are a number of methods which can be used to combine more than one camera on a single tape, but they are kludges for the most part. The ideal thing would be to be able to record all images from all cameras simultaneously with no sacrifices.

With DVRs, you can do this. As I mentioned, Iíve used several brands but have settled on machines from Everfocus www.everfocus.com, distributed by Dan Potts of ATV Research www.atvresearch.com. You also can get them from me, new or used.

Everfocus makes single channel, 4 channel and 16 channel machines.

Iíve had a 16 channel version installed at the office as the primary recording media for over a year now. Operation has been flawless. The system is a fairly complex box, actually a complete computer with operating system, I/O and many of the features you find on a computer. It takes about 2 full minutes to boot, and uses the Linux operating system as do many quality machines to ensure reliability. I just checked the machine, and the last time it was booted was in October, and I believe thatís when the cleaning lady needed to vacuum in the equipment room and I wanted her to clean the filters on the fans in there. The machine needs a mouse, and prefers a keyboard to be connected. Itís got practically every type of I/O a standard computer has, as well as both floppy and CD R/W drives, which Iíll explain later.

16 channels means you can connect that many cameras to the machine.

Unlike in analog systems, each channel can be set up separately for your needs on that channel. You have the simple stuff like channel titles, quality level, etc. You also have the advantage of variable frame rates on a channel by channel basis. Although I can run the machine at realtime (30 frames a second), thereís absolutely no need to generate that much data. So I set slower frame rates for each channel, in our case one frame every ten seconds. If nothing is happening, I only need a snapshot periodically.

However, this machine (as well as many others) has integral motion detection on a channel by channel basis.

You can set, on a per-camera basis, the area the motion detector will cover. You can include or exclude zones.

You can choose the action to be taken when motion is sensed in the field of view of a camera. You merely can log the motion alarm to a file, which can be reviewed later so only motion events need to be reviewed. You can trip external alarms, activate camera presets or tours, and other actions.

You also can change the recording frame rate when motion is sensed. This is what we do. When thereís no motion in front of a camera, Iím recording snapshots at 1 frame every ten seconds. This is very conservative of disk space.

However, when motion is detected, I jump the frame rate up to 5 frames per second. I maintain this for 10 seconds after the motion event terminates. All these parameters are variable.

BUT Ė hereís a real neat feature.

Youíve maybe worked with motion sensors on analog video before. Once you get them working, you may see a door closing behind someone, but the machine didnít start quick enough to record their face. Or whatever.

The DVRs record 10 seconds of video into solid state memory all the time, on every channel. The last 10 seconds of what the camera has seen is recorded in this memory, continually refreshed.

When there is a motion event, this 10 seconds of video prior to the motion event is written to disk. You see what happened for 10 seconds before the event occurred. This has proven to be a very popular and useful feature and in my mind is one of the major selling points of the Everfocus line.

All motion events are logged to a file, which can be reviewed.

A local monitor can be set with various patterns, switching, etc. to show the view of more than one camera.

Remember, now, youíre in the digital domain. The local monitor in many cases will be a VGA computer monitor, not a composite video monitor like youíre used to. I use a quality flat screen monitor and have excellent results. The price is coming down on these flat screen monitors too,. Watch the distributors.

How do you control the Everfocus machines?

Good question.

If youíre sitting in front of the thing, you use the mouse and keyboard to navigate to icons on the local monitor. You log in at one of several variable security levels, and can do anything from simple operation to full system configuration depending on your level of authorization.

There are many parameters to be adjusted on both a system and individual channel level. It takes a while to get them all programmed, but itís relatively intuitive, and I found little need to use the manual. Some things are backwards to what you might think though, like sliders with max function to the left, etc. No biggie.

Earlier I mentioned a floppy drive and a CD R/W drive in addition to the dual 160 gig hard drives. The floppy drive is for saving all the machineís setup info, or its personality. Once you get all parameters the way you want them, you can save the setup file to a floppy for easy restoring in the unlikely event the data would be corrupted internally.

The CD drive is for exporting data. Remember, youíre not messing with tape anymore. You canít pull a tape of an event and take it back to the lab and play it. In a way this is a hassle, but there werenít exactly universal tape formats either, and as most users neglected to maintain their machines and used tapes far beyond their useful life, you may have had garbage to start.

With digital, if you need to export data, you have a number of options.

Using the internal CD R/W drive, you can copy video from the internal disk drive to the CD using either a Windows-standard format which will play on any standard PC using Windows Media Player, or the same using Appleís Quick Time format. Quick Time requires a special player, although it is free, and the Quick Time is slightly better quality.

You can export video based on date and time from known events or the log, from the internal motion log, or many other ways to locate the desired segment of video.

This CD you write can be played on virtually any computer. You can pull hard copies using the screen print feature, even email the files although they will be large. You can make Nth generation copies with essentially no loss of quality.

The Everfocus machines have a watermark feature to help preserve the integrity of the video. It should not be a problem to defend the video as being a true and accurate copy of the original.

Single channel machines have a flash card socket. You can use any of the readily available, inexpensive flash cards (like the ones in some digital cameras) to store video stills or movies.

Most DVRs also come in networked versions. A standard 10 megabit Ethernet connection puts your DVR on your LAN and, if applicable and desired, out over the web via your high speed web connection.

This ability to access and control the DVR remotely can be a major benefit, and will justify a separate discussion of its own. Soon.

With remote access, you assign an IP (Internet Protocol) address to the DVR, and simply use a web browser to visit that IP. You will see a Java applet which will lead you through login, using a user name and password. Once in, youíll be able to view any camera live, put up quad or more patterns of multiple cameras, start and stop the recorder, view the alarm and other logs, play back video based on log events or date and time. You have virtually full control, as if you were sitting in front of the machine Ė and potentially from anywhere in the world where both you and the DVR have broadband web access.

The remote ability of the DVR can give you eyes around the world. You can secure, or surveil, many locations around the world from a single listening post. And many corporations are doing exactly that.

Iím about out of space for this issue. This short description of digital video recorders covers a number of inquiries all similar in nature from readers. If you have a particular interest in or need to use this technology, feel free to email me and Iíll give you an IP address, username and password where you can access and control a working video system using a DVR as described, and experience for yourself the advantages. Youíll need a broadband web connection though Ė no dialup.

Please keep the questions coming and Iíll do my best to share information. Thanks to all who contribute.

Please address your questions on electronic surveillance technology, operations, hardware, etc. to:

Steve Uhrig, SWS Security

Steve@swssec.com

www.swssec.com

Or call 410-879-4035, in Maryland, after 11 AM weekdays for some live chat.

Copyright (c) Feb 2004 by Steve Uhrig, SWS Security.

Steve Uhrig is the owner of SWS Security, an electronic surveillance manufacturing company located in Maryland. Visit the companyís website for more info on the companyís products and services.