Chapter 21 Analog Video Reception
21.1 Introduction
To a large degree, the modern cable television industry is founded on the ability to deliver premium programming for which subscribers pay, either on a monthly subscription basis or on a per-program basis. Key to offering premium programming is the ability to deny the program to anyone who has not paid for it. The denial technology must offer an adequate level of program denial combined with a high-quality recovered signal and reasonable cost. It must be transparent to the paying subscriber, and it must offer reasonable immunity to defeat. These requirements conflict and have led to many advances in technology.
In the analog era, the cable TV industry depended on equipment manufacturers to provide program denial systems, which resulted in incompatible proprietary scrambling and communications systems. As the industry entered the digital era, it developed sets of standards to achieve a degree of interoperability between equipment made by different manufacturers. Theft of premium services has always been a serious problem in analog transmission, and the hope is that the standards, besides promoting interoperability, will mitigate the theft problem. In this section we shall discuss analog and digital program denial technologies used in the industry. However, we shall stop short of describing any technologies that might be used to circumvent scrambling systems.
The word scrambling is used to mean modification of the signal such that it cannot be viewed on a television receiver. A descrambler, which has some unique piece of information, is able to convert the signal back into a form visible on a television set. This usage of the word scrambling is contrasted to the unrelated usage in the data transmission field. In data transmission, the word refers to the process of modifying a datastream to remove any repetitive bit patterns for the purposes of equalizing the spectral density of the signal and aiding in data recovery. It is common in the data communications field to provide for separate access denial, the technology of which is called encryption.
We use the word pirate to refer to anyone who attempts to receive programming for which no payment has been made. Usually, it is applied to those who either develop or sell the technology to receive such programming. Sometimes, though, the word is used to refer to a subscriber who purchases equipment to receive unauthorized programming.
21.2 Non-Set-Top Resident Program Denial
The first class of program denial technologies to be discussed are those that normally don’t reside in set top terminals. These have also been known as whole-house systems because they operate on signals before they enter the subscriber’s home. A signal enabled with any of these technologies is available to all TV sets in the home that are connected to cable.
21.2.1 Negative Traps
The first technology widely used for denial, and one that still has adherents, is called negative trapping. Signals are sent over the cable in the clear, and a trap is installed at each home that does not take the program. Negative traps are frequently employed for a service with a high take rate, that is, with a large percentage of subscribers who buy a subscription. By protecting such services with a negative trap, the operator needs to supply a trap only for the minority of customers who don’t take that service.
Figure 21.1 illustrates the functioning of a negative trap. The filter has a notch, or trap, at the picture carrier of the channel to be denied. By removing the picture carrier (and lower-frequency sidebands), the TV has no signal to recover. Since no signal enters the home, the technology is rather secure. The biggest security threat is unauthorized removal of the trap from a drop. This may be discovered by physical audit in many cases though in some apartment complexes auditing may be a problem. The other issue in security is that the trap may fail, allowing signals to pass. Traps may drift with time or temperature so that the notch moves off the picture carrier. Enough signal power is then delivered that the TV can recover a picture, although a degraded one.
A disadvantage of a negative trap is that it will often have an effect on the lower adjacent channel. The frequency to be removed is only 1.25 MHz away from the channel edge and 1.5 MHz above the adjacent channel sound carrier. If the trap is used on channel 3, the adjacent sound carrier is separated from the notch frequency by 2.4%, so it is relatively easy to control the rise of the notch. However, if the notch is on channel 13, the adjacent channel sound is separated only by 0.7%, making the notch much more difficult to realize. The nature of a filter is that the transition band is proportional to frequency all else being equal. Also, temperature drift is a constant percentage of the filter frequency. Thus, traps are normally restricted to use in the lower portion of the cable spectrum. As you go to higher frequencies, the effectiveness of the notch decreases, and the problems it causes increase.
A trap will also introduce group delay in the lower adjacent channel. Group delay is introduced near the transition region of a filter: signal components closer to the band edge of a filter are delayed with respect to those further removed. The lower adjacent color carrier is closer to the transition region than is the lower adjacent picture carrier so the color information will be delayed with respect to the luminance information. If the group delay is large enough, it produces the so-called funny paper effect, in which color information is shifted with respect to the black-and-white information in the picture.
21.2.2 Positive Traps
A similar technology is that of positive trapping. The shape of a positive trap is nearly identical to that of a negative trap, but the trap frequency and method of use are different. An interfering carrier is inserted at the headend, and a trap is used to remove the interfering carrier at the home of a subscriber who takes the service. Positive traps are popular for services having low penetration since they are installed only at the homes of the people taking the service.
Figure 21.2 illustrates the operation of a positive trap. One or two jamming carriers are inserted at the headend, usually 2.25 MHz above the picture carrier. The peak amplitude is usually the same as, or slightly higher than, the picture carrier amplitude. The jamming carrier(s) may be modulated with various signals. In a mixing process, such as the picture demodulator in the TV, the second harmonic of the jammer (with respect to the picture carrier) is generated and falls on the sound carrier, affording a large degree of scrambling for the sound carrier.
The positive trap is located at the jamming carrier frequency so as to remove the jammer. Assuming the jammer is completely removed, the TV will be able to recover the signal. Since the trap cannot be made infinitely narrow, it will remove some desired modulating sidebands as well as the jamming carrier. Except at the jamming frequency, this may be at least partially compensated for by preemphasizing the midband frequencies at the headend. Preemphasis accentuates the frequencies that will be partially removed later such that the net result is an approximation of normal sideband levels. Preemphasis may also be applied to the group delay introduced by the trap. Because the amount of power around 2.25 MHz in a typical video signal is low, the picture damage is often considered acceptable. Nonetheless, positive traps have a reputation for “softening” a picture.
Since the positive trap notch frequency is toward the center of the channel, we would expect less effect on the lower adjacent channel than that caused by a negative trap. Since the trap can be located inside the subscriber’s home, the temperature range over which the trap must function is reduced.
A jamming carrier generator is connected to each modulator carrying a signal to be positively trapped. This jamming carrier usually interfaces with the modulator using either the picture IF loop or the composite IF loop of Figure 8.3. In addition, the baseband video signal may be predistorted to compensate for the amplitude and delay effects described earlier. This can be done either at baseband or at IF.
Engineers must be cautious when measuring the level of a signal that includes a jamming carrier. Depending on the instrument used to measure the level, the jamming carrier may or may not interfere with the measurement. The easiest way to determine whether or not a measurement is being thrown off by the jamming carrier is to turn off the jammer while measuring the signal level. If the measured level changes, then the jammer was interfering with the measurement.
A variation on positive trap technology is a system that replaces the jamming carrier with a highly emphasized copy of the visual signal: an emphasis circuit at the headend amplifies the baseband video spectrum around 2.25 MHz so much that the power at that frequency is about as great as at the picture carrier, so it forms its own jamming signal. This “jamming signal” is restored to more-or-less normal level by the positive trap at the home.
21.2.3 Interdiction
Interdiction is a program denial technology that can offer addressable services and good security, along with the ability to provide whole-house service. Signals are transmitted in the clear on the cable plant. The normal tap is replaced with a special tap that includes jamming oscillators whose frequencies are close to that of the picture carriers to be denied. The jamming is applied only to those channels that the subscriber has not purchased. Since the jammer is not intended to be removed, it can be located arbitrarily close to the picture carrier, where removal is more difficult. One system actually allows the jamming carrier to wander from one side of the picture carrier to the other.
Figure 21.3 represents a generic interdiction system, which takes the place of the normal tap. The interdiction system includes a directional coupler DC1 and an amplifier A1. After amplification, the signal is split in CM1 so as to be routed to all tap ports. Directional coupler DC2 routes signals to the data receiver or transceiver of the addressing system. Directional coupler DC1 and splitter CM1 compose a normal tap. (In this text, splitters are referred to using the designation CM for combiner because the same device is used for either task.) Each port has another amplifier, used mainly to provide isolation for the jamming carriers, to prevent them from leaking to other outputs. A series of jamming oscillators is provided, which can be switched in and out under command of the addressing system. When an oscillator is connected to a port, it is set to a frequency close to, and sometimes equal to, the picture carrier. The signal received in the home is rendered useless because of the jamming.
The advantage of the interdiction system is the ability to control a number of channels, allowing the operator to provide advanced programming options such as can be provided using a set top terminal. At the same time, interdiction provides excellent security with no equipment in the home. It does not interfere with the use of VCRs or picture in picture, and the signals are available to all TVs in the house.
The disadvantages include the need to supply equipment for every subscriber, regardless of his or her purchase of premium services. Powering is also an issue: the equipment must be mounted away from a subscriber’s reach so that he or she doesn’t have access to the unscrambled signals, which exist on the plant side of the interdiction equipment. This normally means mounting at the tap location. Powering is from the plant, so the cable operator assumes the power bill rather than its being assumed by the subscriber.
21.3 Set-Top Resident Program Denial-Analog Modulation
The dominant program denial methods involve use of set-top terminals (STTs) in the home. Although many technologies have been proposed to effect scrambling, only two are in common use in North America: sync suppression and video inversion. The security afforded with these program denial methods is not total. However, the industry has decided, at least for analog signals, that other methods are not cost effective. Indeed, a review of the plethora of attacks on signal security in the last 20 years reveals that the most successful attacks against reasonably modern scrambling methods are those commonly called “spoofing.” In spoofing systems, a modified set top terminal is “spoofed” into “believing” that it is authorized to descramble a particular signal. The attack is through the authorization system, not the scrambling system. Though there have been attacks against the scrambling system, they usually either have been minimally successful or have yielded to signal modifications in the headend.
21.3.1 Sync Suppression Scrambling
The most common scrambling method in North America is, by far, sync suppression. Variants of sync suppression have been employed since the earliest practical scrambling systems, dating back at least to the mid-1970s.
Recall from Section 8.5 that the sync tip corresponds to the highest amplitude of the modulated RF envelope. If the sync is somehow modified by moving it to amplitudes occupied by active video information, then a television set will not be able to synchronize. The effect is a screen full of random, more-or-less horizontal, lines.
The Scrambling Process
In order to illustrate certain features of scrambling systems, a reference waveform is used, as shown in Figure 21.4. Scrambling operations will be illustrated with reference to this waveform. (Refer to Chapter 2 for more information concerning the video waveform.) The reference waveform shown in Figure 21.4 is a ramp, which rises from black, at the left side of the screen, to white, at the right. The setup shown is a feature of NTSC video that is left over from the early days of television. Color burst is shown though no color information is shown during the active line. In general, we would expect additional color subcarrier power during the active line.
Figure 21.5 illustrates a typical relationship between a modulator and an RF sync suppression scrambler. The portions of the modulator shown are identified as they are in Figures 8.3 and 8.4. Flags B and D are as shown in those two figures, where they were used to route signals from one figure to the other. Baseband video is looped through the scrambler, where a high-input impedance amplifier couples video to a sync separator. The scrambler must have access to the video sync so that it can synchronize the suppression and restoration information. The process of routing video through a piece of equipment is often called loop-through and was described in Section 8.9.1.
From the loop-through input of the scrambler, the video is routed to the modulator, where it is processed as described in Chapter 8. In the scrambler, the sync separator extracts vertical and horizontal sync, which are passed to timing circuitry. The timing circuitry operates a switched attenuator, AT10, in the visual IF path, and an amplitude modulator, M10, in the aural IF path.
The top left of Figure 21.5 shows the reference waveform of Figure 21.4 modulated onto the IF carrier. As explained in Chapter 2, the modulation format prescribed for NTSC (and PAL) video is negative modulation, in which the sync tips correspond to the highest amplitude of the RF (or IF) signal, and peak white corresponds to the minimum amplitude. Sync can be suppressed simply by switching in an attenuator, AT10, during the sync time. This results in the waveform shown at the top right of the figure. During the horizontal (and, usually, vertical) blanking intervals, AT10 is switched in, and during the active video it is switched out. Now the sync tips no longer correspond to the highest amplitude of the modulated waveform. Two values of attenuation are commonly employed: 6 and 10 dB. The waveform shown has been scrambled using a 6-dB attenuator. Had a 10-dB attenuator been used, the horizontal blanking interval would be of even lower amplitude.
From attenuator AT10, the signal is returned to the modulator, where it passes through the vestigial sideband filter FL2. The process of switching AT10 effectively applies amplitude modulation to the visual IF. This modulation produces sidebands that would extend outside the television channel were they not filtered in FL2. After the vestigial sideband filter, the signal is combined with the aural IF, with the remaining signal processing being as shown in Figure 8.4.
Inspection of the scrambled video waveform at the top right of Figure 21.5 shows that the peak amplitude of the scrambled signal is not constant. The peak amplitude depends on the darkest portion of the picture. Because of this, it is not possible to use a sync-suppressed signal as a pilot to operate the AGC of a coaxial or microwave distribution link. Furthermore, it is not straightforward to measure the signal level of a sync-suppressed signal. Manufacturers normally recommend turning off scrambling before measuring signal level. The manufacturer must then ensure that the scrambled signal level remains correct when scrambling is turned on.
Some sync suppression systems take advantage of the reduced peak amplitude of a sync-suppressed signal to raise the signal level about 2.5 dB when scrambling. This returns the peak amplitude of the scrambled signal to the peak amplitude of the nonscrambled signal, thus improving carrier-to-noise ratio by 2.5 dB during active video.
Synchronizing the Descrambler
Several methods are used to communicate descrambling information to the descrambler. One of the more common is to place pulses on the sound carrier, as shown in the lower portion of Figure 21.5. The aural IF is routed out to the scrambler at flag D. Amplitude pulses are modulated onto the sound carrier as shown by the envelopes in the bottom of the figure, where the left envelope is unmodulated, and the right envelope has the modulation impressed on it.
Recall that the audio is frequency modulated onto the sound carrier. Theoretically, the carrier may be amplitude modulated with an independent signal, such as the scrambling synchronization waveform shown, and the two will not interfere with each other. In practice, a number of mechanisms can cause crosstalk between the amplitude and frequency modulation components. The manufacturer must take pains to control these sources of so-called AM/FM and FM/AM conversion. It is possible, though not common, that the distribution plant can introduce undesirable AM/FM or FM/AM conversion.
The sound carrier occupies a relatively narrow spectrum (see, for example, Figure 8.2). You must carefully control the spectrum that results from impression of the AM modulation on the carrier. For this reason, filter FL10 is used in the path between the timing circuitry and modulator M10. The filter controls the spectrum of the modulating signal and, hence, of the modulated signal. It reduces the rise and fall time of the modulating waveform, resulting in the slow transitions shown in the right envelope. A fast transition would be preferable to allow accurate recovery of timing information from the waveform because noise can distort the recovered waveform. The slow rise time means that such distortion could translate into timing errors. However, fast rise times mean the spectrum will spread, possibly causing interference with either the on-channel or adjacent channel video. The manufacturer must walk a fine line between timing recovery accuracy and spectrum spreading.
Descrambling is effected with an attenuator in the descrambler that operates backward from attenuator AT10: the descrambler attenuator must be IN during active video, and OUT during blanking intervals when AT10 is IN. This operation must be timed very carefully to coordinate the operation of the two attenuators to within tenths of microseconds. Besides adjusting the value of attenuator AT10 to effect variations of scrambling in order to frustrate pirates, some systems also vary the timing of the aural carrier pulses with respect to the video.
A variant of the scrambling system described here is baseband sync suppression scrambling. The sync is suppressed before the video is modulated. Because of the way the systems most logically work, baseband sync suppression is accomplished by offsetting the sync level rather than multiplying it. Baseband sync suppression systems don’t require the visual IF interface shown, and since commercial systems don’t rely on pulses on the sound carrier, they don’t use the aural IF interface either. However, because the sync is not the most negative portion of the waveform as it enters the modulator, the clamp in the modulator (see Figure 8.4) has to be modified to clamp at the modified sync tip level, and a signal must be supplied to the modulator to define the clamp time.
Characteristics of the Scrambled Waveform
Figure 21.6 illustrates the demodulated waveform that would be processed by the baseband circuits of a TV receiving the scrambled signal. Also shown, for reference, is the nonscrambled waveform in dashed lines. Note that the sync, which should be the most negative part of the waveform, has been moved up to a voltage range normally reserved for picture information. The darkest portion of the picture is now the most negative portion of the waveform, assuming the picture is not all “whiter” than the modified sync tip level. Sync circuitry in the TV will try to synchronize to the wrong place in the line.
As illustrated, the most negative portion of the waveform occurs at both the beginning and end of the line, where the picture content was black as scrambling ended and started. With most real video patterns, the most negative portion will be somewhere else along the line and will vary from one line to another.
Notice that the peak-to-peak amplitude of the sync tip and the color burst, which is also suppressed, is reduced to one-half that of the unscrambled signal. This is a characteristic of RF sync suppression systems. Since attenuator AT10 of Figure 21.5 reduces the amplitude of the signal by one-half (for 6-dB suppression), all portions of the suppressed signal are reduced in amplitude by one-half. Baseband sync suppression, on the other hand, operates by shifting the signal level during the HBI using a voltage added to the video signal. The peak-to-peak amplitude of the sync and the burst is not reduced with baseband scrambling.
The Descrambled Waveform
Figure 21.7 illustrates the signal as it would appear in a TV that has received the signal after descrambling. The horizontal (and usually vertical) blanking intervals have been suppressed in the scrambler and expanded in the descrambler. The order of events is very important. If the descrambler ever expanded (by switching out attenuation) a signal that had not been suppressed, the result would be a sliver of signal more negative than sync. This could cause unacceptable operation of the TV receiver.
Because some time must be allowed between suppression and expansion, a small sliver of time remains in the descrambled signal, during which the waveform was suppressed but not expanded. This artifact of the scrambling and descrambling process is characteristic of sync suppression systems, unless removed by another signal-processing step. The artifact is sometimes called a “rabbit ear.”
The rabbit ears of Figure 21.7 can have deleterious effects on the signal. One early closed caption decoder required all the run-in pulses transmitted at the beginning of the closed caption line (line 21). In a particular cable system, the scrambler was not set properly, resulting in the rabbit ear at the beginning of line 21 impinging on the first closed caption run-in pulse. The problem was fixed by moving the timing in the scrambler so that the run-in pulse was spared. (Run-in pulses appear at the beginning of the line of closed caption data and are used to synchronize the decoder.)
The rabbit ears are normally in what is known as the overscan region of the TV picture. That is, the electron beam in the TV scans off the edge of the picture tube screen and does not show the rabbit ear. Occasionally, a TV is encountered that does not exhibit normal overscan: the picture does not cover the entire screen plus a little spilling off the edge. In this case, a white bar may be visible on one side of the screen.
Sometimes electrons emitted during the rabbit ear reflect from the side of the picture tube and strike the phosphors on the screen. The electron beam is badly defocused but can result in a slight lightening of the picture on the edges. Usually, the effect is not noticeable because it is always present. However, in scrambling systems that change scrambling modes such that the rabbit ear changes, the effect can be noticeable at a scrambling mode change. For this and other reasons, scrambling modes are normally changed at scene changes.
Yet another issue relating to the rabbit ear is the minority of television sets that perform dc restoration (clamping) on the back porch rather than on the sync tip. If the set tries to clamp when the rabbit ear occurs, the clamp level will be in error, leaving the picture darker than it should be. A back porch clamp can be identified by changing the timing of the back porch rabbit ear and noting if the picture brightness changes. Such a test is best done using video from a test generator.
The suppression interval is determined solely at the scrambler, which will usually contain some sort of adjustment to set the suppression start and finish times. The expansion interval begins at a time related to the time of a pulse on the sound carrier (or some other time marker depending on how the system works). The end of the expansion interval is then timed from the beginning of expansion. Because of the criticality of the sound carrier pulse timing, it must be set at the scrambler/modulator. Reference to Figure 21.5 will show that the sync-suppressed picture carrier is routed through the vestigial sideband filter FL2. The sound carrier, with the synchronizing pulses, is frequently not routed through FL2 (though some manufacturers may do so).
Modern vestigial sideband (VSB) filters are realized using surface acoustic wave (SAW) technology. A characteristic of this technology is unavoidable delay to the signal traversing the filter. The amount of delay is related to the steepness of the filter transition region. To an extent, it is also related to the present state of the art, but in every practical case, there will be a rather large delay to the signal in the SAW filter. This delay is long, compared with the 100 ns or so timing accuracy desired, making imperative the careful setting of the relative timing of the scrambling and sound carrier pulses. This timing is done in the scrambler, which must be mated with the modulator with which it will be used. If the SAW filter or the scrambler is changed, timing must be reset because a different SAW filter in FL2 may have a different delay.
The scrambler manufacturer will provide a procedure to adjust the scrambler. Usually, the most accurate method involves detecting the video signal and the aural carrier pulses, possibly using a spectrum analyzer, whose output is applied to an oscilloscope. The time between the leading edge of sync and the 50% point of the pulse on the sound carrier is measured and set to a specification supplied by the manufacturer. This technique, though very accurate, is tedious and therefore prone to error. Often a substitute method, using a certified descrambler, is used. This certified descrambler is subject to drift in descrambler timing but is adequate if sufficient care is exercised.
21.3.2 Scrambling by Video Inversion
Many analog scrambling systems used in North America employ video inversion in addition to sync suppression. In video inversion systems, the active video is inverted with respect to its normal polarity. The produces a negative picture image. In addition, since the HBI is rarely inverted, color information is reversed as well: the phase relation between the active color and the burst is changed by 180°.
Figure 21.8 illustrates video inversion of the reference waveform. The nonscrambled waveform is shown as a dotted line. The video in this case is inverted with respect to an axis of 50 IRE. One way to think of the inversion is that the video (but usually not the sync) is rotated about an axis, in this case, of 50 IRE. Mathematically, the operation performed on the active video is given by the following equation:
(21.1)This axis information, Vaxis, must be communicated to the descrambler in some manner. If it is not communicated, then gain errors will create an error in the axis about which the video is reinverted (reinversion applies the same equation), resulting in a picture that is too dark or too light. Some early video inversion systems suffered from this problem, but the issue was quickly recognized, and more modern systems include a means for communicating the axis information to the descrambler. One method is to transmit a level in the vertical blanking interval that is related to the axis of inversion. Another method is modification of the sync pulse in some manner to communicate the axis information. Either method can give satisfactory results.
The illustrated scrambled waveform communicates axis of inversion information to the descrambler by shifting the last half of the unsuppressed sync tip to the level of the axis of inversion. Some scrambling systems use an axis of inversion of 50 IRE, and others use 30 IRE or some other axis. Use of 30 IRE would yield a picture in which the white level is translated to the normal unscrambled sync tip level.
Video inversion is often accompanied by sync suppression since inversion by itself doesn’t necessarily yield a completely unrecognizable picture. By itself, the effect of video inversion is much like that of looking at a photographic negative. Combined with sync suppression, it makes life more difficult for the pirates. Frequently scrambling manufacturers will combine sync suppression and video inversion, using several variations on each, to yield a number of scrambling modes they can switch between so as to cause more problems for the pirates.
21.3.3 Audio Scrambling
Audio scrambling is used at times as a supplement to video scrambling. In certain cases, cable operators are required to deny sound to subscribers who have not purchased a particular program. Some set-top terminals, by the nature of their operation, don’t tune unauthorized channels. Thus, the subscriber never sees a scrambled picture nor hears the audio unless he or she tunes the channel without a set-top terminal. If the subscriber does so, he or she will hear the audio.
The more common audio scrambling methods have involved one of two approaches. One approach is to place monaural audio on the difference channel of a BTSC stereo signal. For a description of BTSC stereo, see Chapter 8. Sometimes a variation of BTSC will be used. A “barker,” or teaser, audio is added where the sum signal goes, both to advertise the program and to mask the audio on the difference channel.
Another method used to deny audio is to place a second audio carrier at a frequency other than that normally used for audio (picture carrier plus 4.5 MHz in NTSC transmissions). Normally, the regular aural carrier is also transmitted, with a barker carried thereon. This method allows stereo to be transmitted, but finding the required “clean” spectrum is difficult, and the set-top terminal cost is increased because of the need for a second sound demodulator.
21.4 The Set-Top Terminal
For many years, the set-top terminal (STT) has been a part of subscribers’ lives, first as a device that enabled more channels than what a TV could tune. Later, STTs added remote control before that feature was standard in TV sets. STTs are also commonly used as a vehicle for program denial. Most scrambling systems rely on a descrambler built into an STT. A very recent ruling by the FCC requires descrambling to be separated from the STT early in the 21st century.
21.4.1 Set-Top Terminal Classification
STTs are frequently considered to come in three fundamental types. The simplest is the plain converter, which performs no functions except channel tuning and perhaps volume control. This type of converter is the one traditionally used for expanding the tuning of a TV that cannot tune all the channels available on the cable. Though still available, plain converters tend to be sold more at retail today, for use with the dwindling number of TV sets that don’t tune all cable channels.
The next class of set top up from the plain set top is the descrambling STT. It is similar to a plain converter, except that it includes some sort of descrambler. Authorization of programs must be embedded in the terminal because the cable operator has no way to communicate with the terminal once it is installed in the home. Older descrambling terminals had programmable read-only memory (PROM) devices installed, which controlled what could and couldn’t be descrambled. Later models usually had a facility to be programmed either by plugging a programmer into a socket on the terminal or by using the infrared remote control link. In every case, the terminal manufacturer provided the programming facility. This type of terminal is rarely sold today though many remain in use.
By far, the most popular class of STTs today are addressable terminals, in which the cable operator maintains some sort of communications with the set top when it is in a subscriber’s home. This communication can be accomplished in several ways, the most popular of which we describe briefly here. The communications link is used to authorize the terminal to descramble services according to what the subscriber has purchased.
The addressable data sent traditionally has included authorization information for the terminal to specify what channels or programs should be descrambled. Some STTs tune only authorized channels. Others tune all channels and search for program identification information (“tags”) to determine descrambling authorization.
Electronic program guides (EPGs) are popular now. In this application, a program schedule is transmitted to the set top terminal, which formats it for display to the subscriber. The subscriber typically moves a cursor on the screen until finding a program he or she wants to view. By pressing a “select” button, the set top is directed to the relevant channel.
Out-of-Band Addressability
Out-of-band addressability uses a separate data carrier to communicate data to the STT. A separate data receiver is contained in the STT. The advantage of out-of-band addressability is that the operator can communicate with the STT at any time. Instructions to change the terminal’s state, such as when a pay-per-view event is ordered, can be processed at almost any time, assuming that the terminal is powered and connected to the cable. Consequently, a message does not have to be transmitted much more than one time to ensure receipt.
In-Band Addressability
Out-of-band addressable terminals must bear the expense of a separate data receiver. In-band terminals use addressing information carried in one or more TV channels. This data may be carried in the vertical blanking interval, as additional pulses on the sound carrier, or using some other method. One advantage is the reduction in cost by not having a separate data receiver. Some practitioners feel security is enhanced by using in-band data. The problem usually cited with in-band addressability is the inability to communicate with the set top except when it is tuned to the proper channel(s). To overcome this problem, some manufacturers place data on many different channels, even if the channel is not scrambled. They also creatively divide the data among channels to maximize the probability that a terminal will receive intended data. However, the operator can never be sure that a particular terminal has received all the data sent to it.
Alternative Authorization Procedures
Several methods may be used to authorize programs or channels to be descrambled. This section describes the most common.
Subscription programming is the oldest method of authorizing pay programs. In a subscription system, the subscriber pays a monthly fee and receives a premium channel all the time. This requires that a terminal be addressed only when the subscriber changes the channels to which he or she subscribes. Only one-way communications is required.
Pay-per-view (PPV) programming is sold on a per-program basis, but the subscriber must call the operator in advance to order the program. Ordering may be done by talking to a customer service representative (CSR) with the cable company, who adds the price to the subscriber’s bill and causes an authorization signal to be transmitted to the terminal. Alternatively, the customer may call an automated system that uses automatic number identification (ANI, popularly known as caller ID) to identify the caller. The caller punches in a number corresponding to the program desired, and his or her set top is automatically sent an authorization.
Impulse-pay-per-view (IPPV) programming is the preferred method of selling individual broadcast programs since the subscriber doesn’t have to preorder: he or she can order a program on impulse, even after it has started. IPPV systems require two-way communications with the terminal because the terminal must inform the headend of programs viewed so as to allow billing and payment for the programs.
Commonly, IPPV systems work in what is called a store-and-forward manner. When the subscriber requests a program by actuating a prescribed key sequence, the terminal allows the program to be viewed without checking first with the headend. (If the headend must provide preauthorization, the resulting time delay is considered unacceptable.) At a later date, the terminal is polled from the headend and reports what programs have been watched. From the time the program is viewed until the time the terminal reports the viewing, the viewing information is stored in nonvolatile memory. If the subscriber interferes with communications to the headend, the operator will not know to bill for the viewing. To minimize losses in such a case, the terminal limits the number of programs (or the cost thereof) that can be viewed before it communicates with the headend.
Video-on-demand (VOD) programming is perhaps the ultimate in pay programming. VOD systems allow the subscriber to select what he or she wants to see, when he or she wants to see it. Programs are stored on a video file server similar to those used to store local advertising (see Section 8.10). VOD requires the use of RF return signals to allow the subscriber to request a program from some sort of menu. Advanced systems include so-called VCR functions of pause, rewind, and fast forward. For this to work, the RF return path must exhibit low delay (latency) so that the experience is truly similar to that of using a VCR.
Near-video-on-demand (NVOD) is a compromise system used when an adequate number of channels are available. Selected movies are transmitted, each movie on several channels. On each channel used to transmit a particular movie, the start time is staggered by 10–30 minutes. This way, a subscriber gets convenient starting times. By tuning to a channel with a later start time, it is possible to simulate “pausing” the movie.
Subscription video-on-demand (SVOD) allows subscribers to pay a fixed amount per month to have access to a library of programs they can view at any time, as if they were VOD programs. The difference is that, rather than pay per program viewed, the subscriber pays a flat monthly fee.
21.4.2 Elements of a Modern STT
This section describes the general elements of a modern STT, including digital STTs. Of necessity, our discussion is not all-encompassing since each manufacturer does things differently, and techniques are continuously evolving as market needs change. We present this material to allow those not familiar with STTs to understand what is in one.
The Signal Path
Although variations abound, Figure 21.9 illustrates a typical block diagram of a modern addressable STT. RF from the cable enters at the top left. A directional coupler, DC1, or some other arrangement is often used to take some power in the downstream direction so as to supply signals to a data receiver if out-of-band (OOB) data is being used. Also, if an RF return path for impulse-pay-per-view service is used, it will be coupled to the cable through DC1.
Following DC1 is a low-pass filter, FL1, used primarily to remove any energy above the highest frequency the set top can tune so as to reduce noise pickup and local oscillator leakage into the distribution plant. An attenuator, AT1, performs a delayed AGC operation. At lower signal levels, AT1 is at its minimum attenuation point, permitting the set top to exhibit the best noise figure it can. As signal levels increase above a certain threshold, AT1 increases in attenuation to protect the amplifier and mixer following it from being overloaded. Chapter 8 describes the delayed AGC function.
From AT1, the signal passes to a broadband preamplifier, A1, which largely sets the noise figure of the converter. The signal is up-converted in mixer M1 to the first intermediate frequency. For terminals that tune no higher than 550 MHz, a common intermediate frequency is 608–614 MHz. This would be channel 37 off-air, but no TV stations in North America are assigned to this channel, which is reserved for radio astronomy. Thus, the manufacturer can assume little likelihood of signal pickup off the air. For terminals that tune to higher frequencies, several intermediate frequencies in the neighborhood of 1,000 MHz have been used.
Up-conversion is accomplished by mixing the incoming signals with local oscillator LO1, a high-side local oscillator. Use of a high-side LO results in the signal at the first IF being spectrum inverted (see Chapter 9). The local oscillator frequency is changed to effect tuning to the desired channel. A phase locked loop is used to ensure that the local oscillator remains tuned to the proper frequency.
After filtering in FL2, the signal is down-converted either to the normal IF (NTSC picture carrier 45.75 MHz) or to channel 3 or 4 if the converter is an RF-only terminal (rare today). The second LO, LO2, may be fixed in frequency, may have limited tuning range, or may be controlled by an automatic fine-tuning (AFT) circuit, as shown here. The functions just described generally compose the tuner in the set top terminal.
The signal next encounters AT2, which does sync restoration for RF sync suppression scrambling systems, as described earlier in this chapter. This is a switched attenuator, which offers, usually, 6 or 10 dB of attenuation during active video, and is switched out during horizontal and vertical blanking intervals. Operation of the descrambler is described in Section 21.3.1.
After sync suppression restoration at IF (for the common RF sync suppression systems), the signal is supplied to video and audio demodulators. A single demodulator recovers video and audio from a digital signal. At this point, the video has had sync suppression removed, but if the video is inverted, that still must be corrected. The video inversion circuitry handles this task. As in the case of sync suppression descrambling, recovered in AT2, video inversion is controlled by the controller circuitry, as appropriate to the scrambling system being used. Digital program recovery is covered in more detail later in this chapter.
Most STTs today have an on-screen display (OSD), which is used to indicate the channel tuned, time, the program service on that channel, and other functions. In ordering pay events, the OSD is normally used. It is becoming popular to include electronic program guides in set-top terminals, and the OSD will also be used for this.
Analog audio is normally demodulated to just the composite BTSC stereo signal, as described in Chapter 8. It is possible to include a BTSC decoder. Volume control is effected at attenuator AT3, which operates on the composite BTSC stereo signal. This does not yield a truly accurate volume control mechanism since you cannot retain satisfactory stereo separation when changing the amplitude of the composite signal. The technique is used as a compromise to keep the cost low. Because it is critical to maintain the correct gain for proper recovery of BTSC stereo, most STTs include some means for normalizing the volume control setting such that the deviation out of the set top equals the deviation into the set top (unity gain). This setting is often called “best stereo,” or something similar. It is the only setting at which stereo can be recovered with optimum separation.
Volume control of the NICAM digital audio signal (see Chapter 8) is not possible without bringing it back to an uncompressed format (either digital or analog). This operation tends to be expensive, and the expense is greatly worsened if it is required to restore the NICAM signal to transmit it to the TV. Since the NICAM signal cannot be volume controlled, it is normally separated in the STT using a filter much like the one used to separate out the analog sound IF signal. After the analog video and audio are processed, and just before the RF modulator, the NICAM signal is recombined with the analog video and audio.
Terminal Control
The intelligence of the terminal is embedded in a block labeled “Control” in Figure 21.9. This block must receive addressable information from whatever means are provided by the manufacturer. If out-of-band data is transmitted, a data receiver as shown will be used to recover the data and pass it on to the intelligence of the terminal. If in-band data is used, then it must be coupled to the control section. Some terminals use data transmitted as pulses on the sound carrier using the same technique used to transmit descrambling information. Others embed control information in the VBI.
The control also interfaces with some sort of nonvolatile memory, which can retain information even when the terminal is not powered. Several technologies are in common usage. The control must receive user inputs, usually from a remote control receiver (infrared or RF), and usually from a limited function keypad on the terminal. Some sort of output must be provided to the subscriber. This output could be some combination of a numerical display on the terminal and, more likely today, an on-screen display.
Finally, the terminal must be able to communicate back to the headend. This communication is necessary to facilitate impulse-pay-per-view functions as described earlier. One way of effecting this communication involves using the subscriber’s telephone line (telephone IPPV), in which case, the terminal includes a simple telephone modem. Alternatively, if two-way cable plant is available, RF-IPPV can be used, in which case, an RF transmitter sends signals upstream to the headend.
TV Display Issues
A problem with display of data on a TV set is that the TV system is optimized for displaying pictures, not small text. The compression methods used to save bandwidth, even in analog TV, don’t allow for transmission of small characters, especially colored ones. The methods used to send display data from a computer to a monitor are so optimized, however. This difference in function is not always recognized, leading to frustration on the part of people who try to place too many characters on a TV screen. Generally, you should not use normal TV transmission techniques to place more than 32–40 characters on one line of text. This is contrasted to the 80 or so characters routinely used in computer work.
21.5 Summary
In this chapter, we have tried to show the most important features of scrambling systems. The most common analog scrambling technique in use in North America is sync suppression. It modifies a normal signal in ways that hide the sync. In the process, it becomes very difficult to measure signal amplitude, so this is usually done with the scrambling turned off. Video inversion is often used in combination with sync suppression to provide a signal that is more difficult to restore. Other scrambling techniques have been used in Europe but have not enjoyed much success in North America.
An older technology, which is particularly friendly to subscribers but has limited capabilities, is trapping — either positive or negative. Interdiction has been used as an off-premises addressable technology. Most scrambling systems today are integrated into STTs, which were also described. We introduced important characteristics models. Chapter 22 will discuss digital STTs. Chapter 23 will deal with compatibility issues between cable plant and consumer equipment.