¡@
One lecture of material
Note -- this lecture assumes basic knowledge of NTSC video
Note -- in this lecture, the small images are anchors for larger images -- if you need more detail, click on the small images!
A. Introduction
A video cassette recorder (VCR) encodes standard video NTSC (or PAL or SECAM) signals on magnetic tape. The signals are recorded in analog and no major alterations are made to the video.
The general problem with recording NTSC (or related) video information on magnetic tape is the length of the tape required. Assuming a video signal bandwidth of 5 MHz, a tape head spacing of 0.6 microns, and a minimum wavelength of twice the head spacing -- the minimum writing velocity must be 5 MHz * 0.6 microns * 2 or roughly 6 m/s. Thus a 2 hour movie would take 43 kilometers of tape. (That's a LOT of tape!)
Early VCR units even attempted to use this much tape. (A scary thought!)
However, relatively quickly, a new method was used. The basic idea is to have the video head rotate under the tape. The tape will be traveling along at one speed, and the video head will be rotating underneath it at another (significantly higher!) speed.
¡@
B. Writing the information
There are two write heads mounted on the video head assembly. The write heads are mounted 180 degrees apart. The video tape is wound on the head drum at slightly more than 180 degrees. This allows for a slight overlap in information between the heads.
¡@
The tape is wrapped on the drum at an angle. When the heads spin under the tape, this results in the tracks being written at an angle to the tape.
¡@
This generates tracks that are at a steep angle to the tape (5 - 6 degrees depending on the tape format). As a word of warning -- drawings of video tapes rarely show the tracks inclined at the proper angle -- as it makes the drawing hard to read. Don't be fooled!
¡@
The tracks are scanned alternately by the "A" head and the "B" head. Each track corresponds to one field of the interlaced video signal. Putting this another way, the fields are written on the tape at 59.94 Hz. Thus, two fields constitutes one frame of actual information at the 29.97 NTSC frame rate.
There is some redundancy between the scans (due to the slightly larger than 180 degree wrap of the tape). The last three horizontal lines (down in the blacker than black underneath the image) correspond to the first three lines of the next scan.
¡@
¡@
During the playback of the tape, the video heads are timed to avoid double playing the redundant region. The video head switching pulse falls in the "blacker than black" area below the bottom of the TV. However, it can be observed by misadjusting the vertical hold control so that the vertical sync pulse moves into the viewing area. The video head switching signal will then be observed as a fine line above the vertical sync pulse. (The video head switching occurs before the vertical sync pulse to avoid possible synchronization problems if the sync pulse were disrupted by the head switch ...)
¡@
C. Guard bands and azimuthal offsets
It is important for the heads to read only the signal on their track -- not the signals on adjacent tracks. Therefore, in SP mode, "guard bands" are incorporated. These are areas of the tape, with no signal, located between the tracks.
However, guard bands significantly reduce the playing time of the tape. Thus, a cleverer trick is used to avoid having the heads read adjacent signals. Head A and head B are oriented at slightly different angles with regard to the tape. Head A is typically oriented at -6 to - 7 degrees (depending on the tape type), head B is oriented at +6 to + 7 degrees (again, depending on the tape type). (Don't confuse this angle with the angle the track makes with regard to the video tape!)
¡@
The idea behind the azimuthal angle is to maximize the averaged signal written by the same type of head as the reading head. More specifically, the A head will see a total averaged signal that is largely A. The B head will catch a corner of the A head signal, but the A head contribution will largely average out.
¡@
Use of azimuthal angles permits closer packing of tracks. In LP mode, the tracks are packed with no guard band and with a 9 micron overlap.
¡@
D. Multiple heads
When SLP (super long play) was introduced, the trackswere packed at a very high density. Smaller heads were introduced to accomplish this. However, the smaller heads meant that the quality of the picture was reduced in SP mode. Thus, manufacturers introduced two sets of heads (the 4 head machines) two for SP and two for LP/EP. However, soon after introducing the 4 head machines, manufacturers found that mixing head sizes resulted in better playback. This resulted in machines with differing A and B head sizes for improved performance.
¡@
E. Pause during playback
When the "pause" button is hit -- the tape stops moving, but the head drum continues to move. Thus, the same tracks are repeatedly read and re-read. Under normal conditions, the angle of the tracks is the vector sum of the tape speed and the angle of inclination of the tape on the head.
If the tape stops moving, the angle becomes a little steeper, because the tape speed component is removed.
During pause, the head keeps moving, but the tape is stationary. Thus, the head is tracking over a region that is slightly offset in angle from where it should be. Although the azimuthal angle helps to prevent each head from seeing the magnetic signal written by the other head -- this is not perfect. Thus, the pause image is significantly degraded in quality from the moving image.
¡@
F. Forward and reverse search
During forward and reverse search, the tape speed is increased by a factor of 3-10.
During forward search, several frames appear under the read head as the tape scans. In some sense the "scan angle" is "flipped" in sign, because multiple tracks appear under the head during a high speed fast forward.
During reverse search, several frames also appear under the read head as the tape scans. However, the search direction is such that the "scan angle" is increased.
¡@
The characteristic "fuzz" that appears across the screen during a forward search is generated by the heads crossing the guard band region at the edges of the tracks.
G. Encoding of the NTSC information - the "color under" system
Recall that the NTSC TV signal is encoded with a monochrome Y carrier and a color chrominance C carrier. The intensity of the monochrome carrier is related to the intensity of the signal. The phase of the color carrier is related to the color of the signal.
The television bandwidth is 6 MHz. The sub-carrier for the color is 3.579545 MHz off the carrier for the monochrome information. The sound carrier is 4.5 MHz off the carrier for the monochrome information. There is a gap of 1.25 MHz on the low end, and 0.25 MHz on the high end to avoid cross-talk with other channels.
The problem with the NTSC is that the color information is carried at a very high frequency. Magnetic tape losses and distortions will tend to occur at high frequencies. Thus, the color signal, located at the top end -- will be maximally distorted. So, for VCR recording two major changes are made to the NTSC signal:
The Y signal is FM modulated and centered at approximately 3.5 MHz.
The C signal is moved downward in frequency and centered at approximately 600 kHz
The Y signal is FM modulated with a center frequency around 3.5 MHz (the center is slightly different for different VCR technologies). In FM modulation, the changes in amplitude of the signal are converted into changes in frequency. In VCR FM modulation of a standard NTSC signal -- the peak white signal is at 4.8 MHz and the peak black is at 3.5 MHz. Both positive and negative sidebands are present. However, the positive sideband is largely cut-off due to channel limitations and only the negative sideband is used.
The chrominance signal is then moved downward - centering around 600 kHz. Again, the exact value is a function of the VCR technology.
The chrominance signal is kept largely intact -- only the carrier frequency is changed.
This system presents a number of advantages. To begin with, moving the chrominance signal downhill in frequency is highly advantageous, as it reduces the possibility of it being cut-off by high end filters. The FM modulation is also a highly advantageous way of storing information. The FM modulation results in complex sidebands being retained above and below the center frequency. The sideband frequencies contain much of the high frequency spatial information. They can be filtered out without severely degrading the picture -- or kept in to enhance the quality.
The higher resolution reproduction features of the formats such as SVHS and superbeta formats are obtained by retaining larger spectral regions of the sidebands.
Frequency and Resolution Specifications
for Various Consumer Formats.
Format Sync. tip Peak white Total FM Luminance
frequency frequency deviation resolution
(MHz) (MHz) (MHz) (TV-Lines)
VHS 3.4 4.4 1.0 240
SVHS 5.4 7.0 1.6 400
(Beta 1) 3.5 4.8 1.3 250
(Beta II/III) 3.6 4.8 1.2 240
SuperBeta 4.4 5.6 1.2 285
ED Beta 6.8 9.3 2.5 500