|
Home
Up
Video Editing
What's New
Links
Search
Sparky?
Feedback
| |
|
|
Or
press the BACK button
on
your browser
|
VCRs
|
|
The humble domestic VCR is pretty
much taken for granted these days, and it is easy to become oblivious to
the underlying technology and microscopic precision that these machines
employ. The purpose of this page is to explore their fundamentals, and
attempt to explain the principles behind their various features.
Magnetic recordings are made by
passing a tape coated with magnetic oxide over a recording head, (an
electromagnet with a minute air gap in its magnetic circuit). The speed of
the tape, and the width of the air gap (head-gap) determines the maximum signal
frequency that can be recorded. An audio cassette recorder has an upper
limit of about 18kHz, which is about 250 times too low for recording video
signals. If the speed of the tape was increased to achieve an adequate
response it would have to travel at about 11 Metres/sec. Reducing the
width of the head-gap would allow us to drop the speed to about 5 Metres/sec,
but this method, known as linear recording, is hardly practicable.
A C90 Compact Cassette would hold only about 45 seconds of video, assuming
it didn't shred itself in the process.
Believe it or not, the first studio
video recorders worked on this principle. In Britain, during the 1950's the BBC had a system called VERA. (Vision Electronic Recording
Apparatus) that used reels of tape 2 Feet in diameter, holding 15,000 feet
of 1/2 inch tape, which was capable of recording just 15 minute programmes in black and white.
Toshiba are believed to be the first
company to solve this problem in the mid '50's, by developing a technique
known as 'Helical Scanning'. This achieved the high head-writing speeds by
moving the video heads at high
speed along a diagonal path over the surface of the tape in a series of
stripes. The linear speed of the tape is then reduced dramatically, to
speeds comparable with audio recorders.
Video recording lends itself to this
technique due to the 'field structure' of the signals used in Raster
Scanned Television, and the fact that the signals contain timing
pulses that can be used to control the linear and helical speeds in the
recorder.
From the early '70's to the present
day there have been several competing systems that have been developed, using
the Helical Scanning technique. The final domestic standards to emerge were the
Sony Betamax and JVC VHS (Video Home System) systems, of which VHS is the
only one to survive. VHS did not dominate on technical merit, but rather
due to aggressive marketing methods that included exclusive deals with TV
rental companies (VCRs were very expensive to begin with, and renting was
a popular domestic option at the time).
The diagram below shows how the
helical scanning is achieved in a VHS machine. The tape is wrapped 180
degrees around a rotating head drum which is placed at an angle to the
linear axis of the tape. Suitably angled guide pins (not shown) ensure that the tape
enters and leaves the drum area without creasing or stretching, and they
also bias the tape slightly downwards so that the lower edge locates on a
protruding Ruler Edge to stabilise its passage around the drum.
|
|
|
The drum is formed in two parts. The
lower part, that has the Ruler Edge machined into it, is fixed. The
upper part usually has the heads mounted within it and rotates with the
heads. There are a few models however that have a fixed upper part too, in
which case the heads are mounted onto a disc that rotates in a narrow gap
between it and the bottom part.
In either case, as the head spins it
traces out a diagonal trajectory on the tape, which is wrapped around the
drum in a helical fashion as shown in the diagram.
In its most basic form, the drum
assembly contains two heads mounted 180 degrees apart. As one head is just
leaving the tape, the other begins it scan. Due to the linear motion of
the tape this head writes its track in parallel with, and just behind, the
previous track. This operation repeats continuously until the recorder is
stopped.
It may be recalled that TV pictures
are made up of a sequence of still frames, and that each frame comprises
two interlaced fields.
Video recorders are designed to be compatible with this approach, and
signals are stored on tape as two separate fields, each on its own
alternating track as shown in the diagram below. The tracks shown here are
not to scale - they are very much narrower than the diagram suggests. Two of the fields that comprise a frame have been highlighted to show that
odd fields are written first, followed by even fields, just as they
are in the TV case..
|
|
|
There are two linear tracks, one on
the upper, and the other on the lower outer edges of the tape. These are
written to, and read from, by separate static heads placed a little way
past the video head drum.
The upper track carries the normal (LoFi)
audio signals. This is sometimes a single mono track (particularly if the
VCR also supports HiFi stereo sound), but many current machines divide the track in
two, to provide stereo audio. In this case, the right hand channel is
located on the outer edge, which explains why this channel is the first to
suffer if the tape is worn or damaged (often the case with hired tapes).
The lower track carries a control
waveform to synchronise the linear tape speed during replay. This waveform
is repetitive, and is locked in position on the tape relative to the
position of the video tracks, repeating at each frame interval (two video
field tracks). In modern machines this track also provides pulses for the
tape position counter (often calibrated in real time), and the shape of
the waveform can be modified to allow 'Index Marks' to be impressed on the
tape for accurate searching.
When the videocassette is loaded
into the machine, it is located above the mechanism, the front tape flap
is opened, and then lowered into place. This unlocks the tape spools and
locates them on their drive hubs, and places two groups of guide pins and
a drive capstan behind the exposed tape. The guide pins then move the tape
forward out of the cassette housing, and 'laces' it around the Video Head
Drum and various other components in the tape path as shown in the diagram
below.
|
|
|
During
the threading process the tape is lightly tensioned by the natural
back-tension applied from the Take-Up Spool, and a variable braking
force applied to the Supply Spool, controlled by the Tension
Sensor.
When
Record or Play is selected the Capstan rotates, drawing tape from the
Supply Spool. The tape passes over the Tension Sensor, past the
full-width Erase Head, then over an Impedance Roller, which smoothes out
any vibration or jitter. It then passes around a specially angled guide
pin that prepares the tape for its helical path around the tilted Video
Head Drum. Another similar guide pin returns the tape to its upright
condition after leaving the head drum. The tape now passes over another
Impedance Roller then past the linear Audio Head and Control Head before
arriving at the capstan. Tape leaving the capstan is then returned under
tension to the Take-Up Spool which is lightly driven by one of the decks
motors.
When
'Pause' is selected in the Play mode, the capstan drive ceases, and the
tape remains in tension around the head drum. The means of achieving
this varies from one machine to another, but all except the cheapest
models will use some form of electronic control to position the tape for
optimum freeze-frame picture.
When
Pause is selected in the Recording mode, a special form of transport
sequence is used called 'backspace'
control. This involves reversing the direction of the capstan, and
supplying reverse tension drive to the Supply Spool.
In
the 'standby' condition, the Video Head Drum will be rotating and the
Capstan Pinch Roller will be engaged, but the capstan will not be
rotating. If left in this condition, after a few minutes the drum will
stop, the pinch roller will stand off from the capstan (releasing the
tension), and the back tension will be removed from the Take-Up Spool.
Fast
Forward and Rewind is handled differently in most modern machines to
their predecessors. It used to be the case that the tape was unthreaded,
back into the cassette housing before high speed transport was
commenced. Now, probably due to the success of the Betamax system that
used it, the tape remains, at least partially, in contact with the head
drum. This allows the Audio and Control heads to be in contact with the
tape so that Indexed Search can be accomplished, and accuracy can be
maintained in the Tape Counter. It also means that Play can be started
almost immediately after a fast transit. Needless to say during these
operations the pinch-roller is disengaged from the capstan.
Finally,
the beginning and end of tape is sensed by Infra-red light passing
through transparent leaders at the ends of the tape. There is a large
hole in the underside of the Cassette Housing into which two IR
(infra-red) transmitters are inserted when the cassette is loaded, and there are two
holes on the side edges of the housing which allow the IR beams to fall
on sensors when the transparent leaders enter their path.
|
|
When VCRs first
appeared, their drums had only two heads, one for each field. With the
exception of the simplest budget machines, that is no longer true today.
With the introduction of Long Play (LP), HiFi Audio, and Flying Erase
Heads, the head drum has now become a complex high-precision assembly.
The diagram below shows how the
heads are distributed around the periphery of the drum. There are also
slight differences in the vertical position of the heads to compensate
for their angular offsets, ensuring correct tracking of the two video
field tracks.
Details of how these heads are
used will be found later on this page under topics related to the
facilities the heads provide. Note that not all machines have a full
complement of heads as shown.
|
 |
|
All magnetic recording/replay heads
work on the same principle. A magnetic circuit is formed from a suitable
ferrous material, and a coil of wire is wound onto it such that when an
electrical current is passed through the coil, magnetic 'flux' is induced
into the magnetic circuit. If an air gap is introduced into this magnetic
circuit then a concentrated magnetic field is formed in the air around the
gap. Placing the gap in contact with the sensitive surface of the tape
causes the magnetism to be transferred to it. As the tape moves across the
gap, the strength of this magnetism varies as the current through the coil
is changed. In this way, electrical signals can be stored as magnetic
patterns on the tape.
In replay, the action is reversed.
As the magnetic pattern on the tape is passed over the air gap (head-gap), magnetic
flux is induced into the head which results in a small voltage being
generated across the coil.
In an audio tape recorder the heads
are made of metal, but in a VCR, because the frequencies used are so much
higher, they are constructed from Ferrite. This is a hard
but brittle substance produced from ceramic metal oxides, formed at high
temperature and pressure (by a process called sintering). It enables the intricate design of the minute
heads to be be produced economically.
|
 |
|
In order to achieve maximum signal
during replay, especially at higher frequencies, it is necessary that the
head gap is in exactly the same orientation as when the recording was
made. This factor is a vital key to how so many heads can be used in a VCR
without mutual interference between tracks of one type and heads of
another (the significance of this we shall discover shortly). Each head
has its own standard gap angle, known as the Azimuth Tilt, which
prevents it responding to signals recorded by another head.
|
|
The linear tape speed
is too low to provide high quality stereo sound, and a number of methods
were proposed over the years to achieve near-CD quality. The method
chosen for VHS is somewhat novel, but works well, and has no noticeable
effect on the quality of the video image.
The technique is
known as Depth Multiplex Recording, whereby the audio
signals are scanned onto the tape along with the video signals, using
the same physical tracks, but employing different heads. The audio heads
operate at about one-third of the frequency of the video heads, which
allows them to have wider head gaps. This produces a magnetic
field which deeply penetrates the tape oxide layer. The video head,
which follows it, effectively erases the surface layer, but there is
still more than enough signal left on the tape to enable the audio to be
recovered fully on replay.
The reasons why there
is negligible interference between the audio and video signals are
threefold. Firstly, the frequencies used by the two signals are a factor
of three apart, secondly, the head gap azimuth tilt differs widely
between the two heads, and thirdly both channels use FM (Frequency
Modulation) as a recording method. FM, as used for VHF radio
transmissions, has the property of always 'capturing' the strongest
signal, even though a neighbouring channel may leak into its frequency
band.
It should be obvious,
looking at the diagram below, why HiFi sound cannot be dubbed after
recordings have been made. This would effectively erase (or seriously
damage) the video signal. So, whilst it is possible to place the video
over the audio signals, the reverse is not true.
|
 |
|
The standard VHS
specification calls for a full-width erase head. This precedes the head
drum and erases signals on all tape tracks prior to new
recordings being made. This is certainly adequate for normal VCR use,
but is not accurate enough for 'in-machine' editing .
Video Assembly
Editing and Insert Editing requires that existing video (and HiFi audio)
tracks be rewritten with new signals. For this to be successful, the old
helical tracks have first to be individually erased, which the full-width erase head is incapable of
doing. This is where Flying
Erase heads come in. They are fitted to the drum, and rotate with the
scanning heads, erasing tracks in advance of any head that will be
recording new signals.
It is common for
there to be only one flying erase head, which erases both odd and even
tracks in one sweep as shown in the diagram above.
|
|
There are two main motors that
work together to ensure that the helical tracks are written to, and read
from, the tape correctly. One is the Capstan motor, which feeds the tape
through the machine, and the other is the Head Drum motor that drives
the rotary video heads.
Both these motors are controlled
to fine limits - not only are their speeds controlled but also their
'phase'. The term 'phase' may be a little confusing so, by way of an
explanation, consider two cars travelling in the same direction along a
straight road. Their speeds may be identical, but one may be ahead of
the other. By adjusting their phase, the two cars may be bought
side-by-side, and travelling at the same speed.
The position of the video heads as
the drum rotates is known to the VCR by virtue of a PG (Pulse Generator)
pulse that is produced from a precise location on the circumference of
the head drum. This pulse is vital to both the recording and replaying
process.
When the VCR is in RECORD
mode, the capstan motor runs at a constant speed such that the tape is
fed through the machine at the standard VHS specification of 23.39
mm/sec. The Head Drum motor speed is not fixed, but is synchronised with
the frame rate of the incoming video signal. The phase of this
motor is locked (by virtue of the PG pulses) to the field synch pulses contained within the
video signal , such that the heads contact the tape at the correct point,
ensuring that each field fits exactly along the length of each helical
track. At the same time, the linear control head writes PG pulses to the linear control track, to provide a magnetic version
of 'sprocket holes', a familiar feature of movie film.
When the VCR is in REPLAY
mode, the full complexity of the motor control circuits becomes evident,
but different manufacturers use slightly different techniques to achieve
accurate tracking of the video tracks by the rotating video heads. What
is described here therefore, are the general principles involved in successfully
tracking and replaying a recorded tape, rather than intimate details.
Before we start, it is worth
noting that both the odd and even field heads are fed with the same
signals, in parallel, during
the recording process, but on replay it is necessary for them to be
alternately switched into circuit as each field is replayed. This is to
prevent interference during the short periods of overlap between the end
of one field and the start of the next. (By design, both odd and even
heads are in contact with the tape for a short period during
frame-changeover)
This time it is the head drum that
is driven at constant speed (1500 rpm PAL, 1800 rpm NTSC). The head drum
PG pulses are compared
with the control track pulses on the tape, and the capstan speed is
adjusted until these two sets of pulses are occurring at the same rate,
and the phase is such that the rotating heads follow the recorded
helical tracks.
This is not the whole story
however. Providing the VCR and the recorded tape are working within
specification, the video heads will be following the helical tracks well
enough to produce a noise free picture. However, given normal error
tolerances, this is unlikely to be the case (particularly with rented or
pre-recorded tapes), and some form of 'Tracking' adjustment will be
necessary. This is often achieved by introducing a variable delay (phase
shift) between the PG pulses and the control track pulses. In older
machines it was necessary for the user to adjust a knob or slider to
correct the tracking, but most modern machines adjust this delay
automatically by one means or another. How this is done varies with
manufacturers, and is beyond the scope of this discussion.
|
|
Under Construction
|
|
Under Construction
|
|
Under Construction
|
|
Under Construction
|
|
A
VCR cannot be started and stopped instantly so, in order that it may be
paused and unpaused in the record mode, a standard 'backspace' control sequence is
used within the VCR.
When the VCR receives
a Pause signal in the recording mode, it continues for a few more
frames, then backspaces many frames, to a position some way before the
frame on which it was paused.
When Pause is
released, the recorder plays forwards, synchronising itself to the
underlying tracks until it reaches a point on the tape, near where it
was paused. It then switches to the record mode and continues recording
the new video signal, overwriting the tail-end of the previous
recording.
The time (in frames)
between the 'unPause' signal, and the start of the new recording is
known as the Record Delay. The Pause delay is less easy to visualise. It
is the number of frames between the point on the tape where Pause was
commanded, and the position of the first frame of the new recording.
Because the new recording may start on either side of the 'Pause frame',
the Pause delay may be either positive or negative.
There is a more
detailed, illustrated, description of this technique in the New
Glitch Doctor User Guide.
As
you may realise, this whole sequence has to be predictable and accurate. High-end VCRs,
typically ‘Edit’ machines with ‘flying erase heads’ manage this
very well because they use sophisticated electronics and ‘precision’
transport mechanisms to position the tape accurately. On the other hand,
budget machines may rely on mechanical means to ‘shuttle’ the
tape, or some crude form of timer. This results in inconsistent delay
values which may even be found to change, depending on the amount of
tape used in the cassette. These machines were never designed to be used
for editing. The only reason why they use the standard
‘backspace’ procedure is to help the ‘tracking’ mechanism
cope with the joins on playback, the actual frame(s) they cut to is not
considered important.
|
|
Under Construction |
|
