RAPIDVISION VIDEO PHONE LINE TRANSMISSION
WHAT OUR EXPERTS SAY...
Sending a high-quality live colour video
image over a conventional telephone line is by any reckoning, a rather neat
trick. The basic technology was developed by GPO engineers more than twenty
years ago, as part of a research project into video phones. It works on the
basis that most of the time there’s comparatively little change in a video
image from one frame to the next. Therefore it follows that the amount of
information needed to convey an image can be drastically reduced by only
updating the parts of the picture that change. Domestic video phones using this
principle made a brief appearance in the late 1980s but floundered on the fact
that it was pointless owning one, unless you knew someone else with a video
phone. At around £200 a throw the market was inevitably very small...
The original GPO prototype equipment, capable
of sending a jerky low-definition black and white picture, filled several
floor-standing racks. Work on video processing and compression microchips has
made enormous strides since then and now The Vision Research Company have
managed to pack all of the technology needed to send and receive up to 16
high-quality colour video images, plus alarm and telemetry information, over a
normal public switched telephone network (PSTN), leased line or ISDN
connection, into a pair of 2U standard rack mounting cases.
Remote site surveillance has become an
increasingly important tool for the security industry, but fixed cable or
microwave communications links are often either impractical or prohibitively
expensive, depending on the distance between the site and the monitoring
station. The Rapidvision system effectively takes distance and cost out of the
equation. The camera and the monitor can be tens, hundreds or even thousands of
miles apart, the only requirements are for a mains supply and telephone line at
Operating costs can be very low indeed.
Continual on-line monitoring via a leased line is possible, though it is likely
many end-users will elect to use a PSTN communications link. The remote site
will only call the monitoring station when an alarm has been activated. This
imposes an additional cost burden on the remote site installation, namely that
each camera must have an associated alarm sensor, or be linked to a system, but
this is unlikely to be a deterrent as by definition in most cases as there will
be few, if any video surveillance alternatives.
The Rapidvision system makes use of the
latest microchip-based digital video processing and compression techniques. The
various components are available in a range of configurations, based on the
number and type of cameras. We’ve been looking at two variants, based around
the CR16 receiver, used with CT4 and CT16, 4 and 16 channel colour camera
The CR16 16-channel receiver is the core
component in a Rapidvision installation, it can be used with any type of
transmitter/camera combination -- colour or mono -- and communicate with
multiple sites, each of which has its own unique address code. Front and rear
panel layouts are reasonably straightforward. It has a pair of composite video
outputs, via BNC sockets. There are three 9-pin D-sub sockets. Com 1 is a RS232
serial port, for hard-wire communications with a transmitter. Com 2 is an
internal communications port, for external control functions and diagnostics,
Com 3 is for an optional keyboard, used for programming. A 25-pin D-sub socket is used to connect the
receiver to a modem or terminal adaptor. A range of settings, to suit most
common makes and models, are programmed into the receiver.
The front panel is divided into five areas.
On the far left side there are four keys for selecting the stored phone
directory and initialising dialling. They are also used in various on-screen
menu operations. The second group of keys are picture controls, for selecting
definition mode, controlling a thumbnail ‘window’ display (more about that in a
moment), plus updating and freezing the display. Next to that are the camera
selector buttons, they are numbered 1 to 8 and arranged in two groups (A and
B), giving a total of 16 switched inputs. This section also contains keys for quad
or octal display mode, and pre-select,
which moves the active camera (on a suitable pant/tilt mount) to a preset
The fourth section has a set of four position
buttons, for controlling pan and tilt; there are two keys for zoom in/out, and
four buttons for auxiliary functions, including wiper and lights, plus two other switched outputs. On
the extreme right is the program button, mains on/off switch, and indicators
for power-on and on-line.
A similar set of camera functions are
available on the CT16 transmitter unit (for local control), along with two
additional switched auxiliary outputs. The rear panel of the CT16 has a bank of
BNC sockets for the camera inputs and monitor output. There are the same set of
Com ports and a modem socket as the receiver, plus a second 25-pin D connector,
for the alarm inputs.
The alternative CT4 transmitter (4 colour
camera inputs) is housed in a 1U rack mount case; other than the mains on/off
switch there are no front panel controls. Around the back there are two 9-pin
D-sub Com ports, and two 25-pin connectors for the modem and alarm inputs.
On both units two alarm conditions are possible.
An off-line alarm is when the transmitter is not connected to the receiver or
under local control. An on-line alarm occurs when the transmitter and receiver
are in communication with one another. If a single alarm is triggered when the
transmitter is off-line the receiver immediately stores a sequence of four
images from the associated camera, (presettable at between 0.5 second and 2.0
second intervals). The transmitter initiates a call to the receiver using the
first number in its telephone directory. As soon as the connection has been
made it uploads the four stored images in a quad format, to the receiver.
Alternatively the fourth image can be live,
though this is really only practical on faster ISDN connections; 30 seconds or
more can elapse between the alarm being set off, and the images being received,
by which time there’s probably nothing to see. If there are two alarm events, a
second set of four images are stored by the transmitter.
Full frame images are subsequently
downloaded, and these can be scanned in rapid succession on the receiver, to
aid movement analysis. The alarm can be
released by pressing the freeze button, the transmitter automatically hangs up
five minutes after the images have been successfully transmitted.
During an on-line alarm event the display
immediately switches to the relevant camera and it remains on the screen for a
pre-set period or until the operator presses the freeze button. Alternatively
the display remains unchanged and display caption reporting the alarm appears
on the screen. To prevent screen burn from fixed images the Rapidvision
receiver has a switchable screen saver facility (moving and morphing ‘V’ and
‘R’), that comes on after a few minutes inactivity
Installation and initial set-up -- using the
factory defaults -- is quick and easy.
Our review system was supplied with a
pair of US Robotics Sportster Flash modems. The simplest way to verify
the operation of the receiver -- once the power, monitor and modem connections
have been carried out -- is to call up the telephone directory display, select
Vision Research’s test number and press dial. If all is well the receiver will
access the manufacturers on-site test system at their UK headquarters in deepest
Havant, Hampshire. At the time of writing this had a total of five cameras
operating, comprising a mixture of colour and monochrome models, indoors and
outside, on fixed and PTZ mounts, some with
switchable wiper and light functions.
After the dial button is pressed the main
display confirms that contact has been made and shows the baud rate (4800 to 38400
bps); shortly afterwards the first image appears, usually in just over 30
seconds. Each camera display incorporates a site ident along the bottom of the
screen; the camera number, time, date, connection and communication status
appear on panels at the top of the screen. Once an image has been formed the
speed at which it is being updated appears in the top right hand corner of the
screen. Depending on the quality of the line, the refresh rate can be as low as
If any camera functions are selected (pan,
tilt, zoom etc.) a small monochrome window appears in the middle of the image.
This has a much higher refresh rate, resolution is very poor, though it is
sufficient to show the change the camera position and zoom setting in near
real-time. There is lag, lasting between half a second and two or three
seconds, between a PTZ button being pressed and the action showing up on the
screen. Some practice is needed, to avoid overshooting the desired position or
zoom setting. When all operations have been confirmed it’s time to check out
Camera and alarm connections on the transmitter
should pose few problems for most installers. Camera telemetry is carried on
the video feed, supported protocols are Videmech, American Dynamics and Pelco.
Transmitter set-up can either be carried out locally, or more conveniently,
from the receiver, once the system has been checked out. The set-up mode is
entered by pressing the program button, this brings up the first screen showing
basic configuration and status information. To access further menus a pass-code
-- up to 12 characters long -- has to be entered. The next menu has three
selections, for receiver and transmitter set-up, and exit. Choices are made
with a shifting highlight, controlled from the PTZ keys, and confirmed using
the dial button.
The receiver programming menu contains almost
20 items, covering everything from pass-code re-programming and setting baud
rate, to switching off the screen-saver. There’s an equally comprehensive set
of adjustments and pre-sets for the CT16 transmitter, including creating site idents,
alarm configuration, setting the time and date and scheduling operations. Once
the necessary changes have been made the save changes option dials up the receiver
and downloads the revised data. The CT4 transmitter has similar remote set-up functions,
though it lacks any local camera controls.
The quality of construction on all three
units is very high. The all metal cases are made from a mixture of steel and
alloy and built to withstand a good deal of punishment. The printed circuit
boards are neatly laid out and servicing access is good. The instruction
manuals are clearly presented, generally well written and easy to follow. A few
more diagrams wouldn’t have gone amiss, though.
The instruction manual goes into considerable
detail about the system’s limitations. These are mainly concerned with avoiding
too much movement on the screen. This can have a drastic impact on transmission
speed as larger quantities of data have to be sent. The advice includes
choosing a wide angle lens setting where possible, since movement will take up
a proportionally smaller amount of screen area. Stable camera mounts are essential,
even slight movement will be regarded by the video processing circuitry as a
scene change. Trees, traffic, flags and smoke should also be avoided if
possible as they will all slow down the refresh rate unnecessarily. The manual
also points out that in low-light conditions most cameras turn up their AGC,
which results in a grainy, noisy
picture, that the system will interpret as movement.
On the default ‘low’ definition setting
digital processing artefacts or ‘blocking’ is normally clearly visible in the
image, colour depth and grey scale are quite limited. Selecting the high
definition or ‘HD’ mode yields a dramatic improvement, the overall amount of
detail and colour fidelity are most impressive, though the refresh rate is
significantly slower. Depending on the camera/lens type and scene lighting
conditions, even in the low definition mode there’s sufficient detail for
critical identification of things like vehicle type colour and even number
plates. Clearly it would be unrealistic to expect the same kind of performance
from Rapidvision as a conventional fixed connection CCTV set-up, but in HD mode
there’s really not a lot in it! Considering the amount of heavyweight
processing taking place, in favourable conditions image quality can be
The delays in PTZ control and the slow
refresh rate can pose problems; tracking a subject or vehicle, for example, can
be extremely difficult, though it does get easier. The lack of any
alarm-triggered VCR control facilities on the receiver could also be a
disadvantage in some situations, and it a little surprising that the designers didn’t
incorporate more alarm functions on the receiver. Even something as simple as simple
warning buzzer would be a step in the right direction.
Quality considerations aside, the convenience,
flexibility, cost savings and security
implications of this kind of remote video surveillance cannot be
underestimated. Rapidvision clearly has much to offer. The complex technology
operates entirely behind the scenes and it is no more difficult to install, set
up and use than a fixed-line multi-camera system, though the need for an
associated alarm system is obviously a consideration.
The reduction in picture quality and lack of
full motion video is something to bear in mind, but against that has to be
weighed the fact that this system can provide a viable video observation solution
at sites where other forms of surveillance are either impossible or uneconomic.
Design and design features ****
Circuitry and components *****
Ease of installation and wiring ****
Range and variety of functions ****
Accompanying instructions *****
Technical advice and backup ****
Value for money ****
Ó R.Maybury 1997 0608