Scanning, an Ingenious
Solution
North America's traditional NTSC television system comes in
for a lot of abuse, and indeed it is technically imperfect compared to digital TV, but it
is still what virtually all of us watch most of the time. Considering that the technology
was developed more than seventy years ago, even before broadcast engineers had really
perfected AM radio (and FM was only in its earliest experimental stages), the system has
always struck me as a remarkable combination of ingenious answers to what must have seemed
like insoluble problems at the time.
For example, the developers had to deal with the fact that
our eyes see and process millions of bits of information simultaneously -- our field of
vision is taken in all at once and fed to the brain by the rods and cones in our retinas
as many parallel signals. This is dramatically unlike our sense of hearing, which is
sequential: each ear picks up a single complex signal and sends it to the brain as one
constantly varying series of impulses. Audio equipment can duplicate this process quite
neatly with one channel of transmission (more only if spatial information is necessary).
It might have been possible, in theory at least, to develop
a TV system using separate signal channels connected to each point on the TV screen, and
primitive attempts were made to do this. But to achieve any sort of reasonable resolution,
hundreds of thousands of such channels would have been necessary -- far beyond the
technology of the day. Instead, the television pioneers realized that the system must
"paint" a picture with sequential information, and do it fast enough to fool the
eye into thinking it is seeing a multiplicity of points of light all at once.
Fortunately, the human mind is a willing conspirator in
this. We are very adept at integrating a series of bits of visual information into a
single image, as long as they happen very close together in time. Using this facility,
television's inventors devised a way to analyze a picture by breaking it down into a
series of horizontal lines, whose varying light levels could be "read" from left
to right, rather as one would read a line of type in a book. This information could be
used to control a moving spot of light that would scan across the display screen,
reproducing the light levels of the original. Reproducing each line in succession would
give the impression of a complete picture, as long as the lines were close enough together
and the whole process happened quickly enough for the eye to take it all in as a single
image.
Then, if this process were repeated over and over in quick
succession, the image would appear to move. The development of photographic motion
pictures was based on the knowledge that if a series of still pictures, each slightly
different, were flashed in sequence, the mind would perceive them as moving smoothly -- as
long as the changes happened quickly enough. Slowing down the rate would produce a visible
flicker; slowing it down further would allow the brain to resolve separate pictures.
Experience with movies had shown that flashing 24 separate
pictures per second would give the illusion of smooth motion, but that a flicker could
still be seen at this rate. By flashing each frame twice, however, but for half as
long, would remove the flicker without requiring an increase in the amount of film used.
The same technique was employed with television, although
each such frame was made up of a rapidly painted series of scanning lines. It was
determined more or less randomly that 525 lines would be sufficient to create a credible
picture (although other systems have used different numbers -- about 400 in the original
British system, more than 800 in one French system, and 625 currently in the PAL and SECAM
systems used everywhere except North America and Japan), and 30 frames per second was
chosen both because it would allow smoothness of motion and because it could be derived
easily from the standard North American AC line frequency of 60Hz. But it flickered, so
rather than increase the frame rate to compensate for this, each frame was divided into
two "fields" of 262.5 lines each, which would flash at a rate of 60 per second.
Every other line would be scanned in alternating fields, to make up a complete 525-line
frame, in a technique called "interlacing."
For this very complex system to work, the receiving TV set
has to be kept in perfect synchronization with the camera picking up the signal in the
first place -- at any moment the moving spot on the TV screen must be in the same relative
position as the corresponding spot in the TV camera. To achieve this, a series of pulses
are added to the signal to tell both the camera and the receiving TV set when to start a
new line and when to begin a new field or frame.
These pulses have a secondary function as well: because the
moving spot must return to the left of the screen after every line, and to the top of the
screen after every field, it must be turned off when it is returning, or the picture would
be covered with a pattern of diagonal return lines. This could be done by using signal
notches that could trigger the return function, and would be interpreted as black by the
flying spot. The difficulty is that such notches could be "filled up" by noise
in a less-than-perfect signal, and confuse the sync circuitry. The answer arrived at by
television's developers was simple: to broadcast the signal as a negative, in which the
synchronization would be maintained by a series of strong positive pulses (called
"blanking pulses"). Once the receiver inverted the signal, to achieve a proper
picture, it would still interpret the pulses as black, but they would be less affected by
interference.
All in all, a remarkable series of technical solutions that
continue to serve us well, whatever their limitations.
...Ian G. Masters
ian@mastersonaudio.com
|
