Analog television signals are the first large-scale methods for analog image and sound transmission, which were used, among other things, in the field of television broadcasting . Since the mid-1990s, analog television transmission methods have increasingly been replaced by digital television , which uses various image compression methods and digital modulation methods within the framework of standards such as DVB-T .
This article only deals with the analog variants of television signal transmission.
Video signals ( video technology ) are not a real subset of the television signal because they are application- related and therefore do not have to comply with the television standard. However, there is an economic incentive to be compatible with the devices of the television broadcasting system (country-specific television technology ), which are manufactured in large numbers and are therefore inexpensive, or with the computer technology available recently .
The classic picture tube delivers a B signal (picture signal, not to be confused with the blue signal). This B signal is blanked in the television camera and results in the BA signal. BA means picture with blanking.
The blanking level is the DC voltage that corresponds to the waveform during the blanking period. This blanking level is the reference level for further signal processing and transmission. The BA signal can be understood as a multiplication of the B signal with the logical blanking signal (A signal). It was customary to add the A signal to the BA signal. This signal component is referred to as lift-off and simplified the suppression of visible image and line returns.
BA signals (0.7 V) were processed and displayed within the studio. The pulse signals required for synchronous operation were supplied by a clock generator via a pulse distribution amplifier. The pulse group H, V, A, S was typical (horizontal, vertical, blanking and synchronizing pulses with Uss = 4 V).
The output signal of a studio of black and white technology was formed from the BA signal by adding the S signal (0.3 V) and referred to as the BAS signal.
The pulse signals are not considered video signals, although the S signal can be interpreted as a black screen (BAS). In the same sense, the A signal is a white image (BA signal, i.e. without an S component). With the transition to transistor technology, which largely corresponded to the transition to color television, the impulse group was increasingly replaced by a leading black screen. This black screen then also contained those impulse signals that were additionally necessary because of the introduction of color.
The signals Y, R, G, B (luminance signal and color value channels) used in color television are BA signals, which often also contain the S component and are then BAS signals. The modulated color carrier resulted in the further letter F and the complete color image signal was referred to as FBAS signal or, if appropriate, as FBA signal. Data signals (teletext, possibly sound, control signals for the clock generators) and test lines were embedded in the blanking gaps in the composite video signal.
The signal aspect
The explanations about video signals show that the usual signal concept in physics hardly meets the requirements. A signal is always about the relatively arbitrary interpretation that the receiver makes. A signal is a non-empty set of quantities. In the cases considered here, the variables are preferably electrical voltages, which in particular correspond to image and sound.
As early as 1865, the Pantelegraph was a device that transmitted two images broken down into lines, time-multiplexed line by line, whereby the signal was discrete-value ("on" - "off") but continuous with respect to time. The image source was not charge images, as in the usual pick-up tubes, but conductivity images that were painted or written on a conductive film with insulating ink. This early fax machine wrote electrochemically and used synchronized swinging pendulums with electrical holding magnets. The image signal and the synchronous signals differ only in their parameters from the more modern signals BA, V and H.
Apart from a matrix with 100 × 100 lamps set up for demonstration purposes in 1936, pixels only existed with the introduction of digital storage. Until then, only the continuously scanned lines actually existed (in the vertical direction these images have therefore long been discrete).
With the development of television technology in the 1920s, a way had to be found to transport the image recorded by the camera to the recipient. A parallel transmission of the individual pixels cannot be implemented, since in this way each pixel to be transmitted would require a transmission channel (e.g. a cable). A television picture with today's PAL resolution would therefore require 414,000 transmission channels (575 picture lines × 720 points per line, for example).
Thus, a serial transmission of the television signal was chosen, in which the moving television picture is broken down into individual still images shown in rapid succession and these still images are in turn divided into individual lines that are transmitted one after the other. Only a single transmission channel is required for this. There were various approaches to achieve this, for example the Nipkow disk . In the end, scanning by means of a video tube prevailed .
Today, images in television cameras are scanned in a discrete-place and time-discrete manner using CCD chips or a CMOS sensor .
BAS signal (Bild-Ausast-Synchron) is the German equivalent for the so-called VBS ( Video Blanking Sync ). The BAS signal is the complete television signal for black-and-white image transmission, which is composed of the image signal (B), the blanking signal (A) and the synchronization signal (S) (see above). The FBAS signal ( Color BAS ) or CVBS ( Color Video Blanking Sync ), which also contains the color information, is used for color image transmission . The English abbreviations are often interpreted as Video Baseband Signal (VBS) or Composite Video Baseband Signal (CVBS).
Transmission line by line
Successive single images ( frames , English for "frames") or half images ( fields , images with halved vertical resolution) are transmitted one after the other. Each of these images consists of several lines, which are also sent one after the other, i.e. basically the television signal consists of the sequence of the individual lines.
However, if this signal only consisted of the lines in a row, the receiver would not be able to recognize where a line should be displayed on the screen. Neither could the recipient see where a new line begins. For this reason, certain voltage patterns are added to the television signal at the points where lines of two different (half) images meet and at the beginning of each individual line, which the receiver has to evaluate. These are the sync signals that are embedded in the overall signal. A distinction is made between the vertical and the horizontal synchronous signal . All details on timing refer to the PAL system common in Germany .
The beginning of each line must be marked individually so that the recipient can synchronize again. This is to be shown using the representation of an image line.
The picture shows the timing of a line of the television signal. A four-step gray staircase is used as an example , that is, four vertical bars can be seen in the picture. From left to right they have an increasing brightness, the left bar is black, the middle dark and light gray, the right bar is white. The bars each take up a quarter of the width of the image. The edge durations, which are approx. 200 ns in the video range and approx. 300 ns in the synchronous range, are not shown. In this position, the area called the line length is only available in the studio (leading pulse signals). On the receiver side, the reference times are always in the middle of the leading edge of the synchronous signal.
On the far left you can still see part of the previous line (white level), then the front porch at 0.3 volts (blanking level), then the 4.7 microsecond line sync pulse (sync level). In the illustration, the synchronous level is 0 volts. Behind the line sync pulse is the back porch , which lasts 5.8 microseconds. This is where the actual image signal begins, the voltage value of which corresponds to the brightness, with the black level being 0.32 volts with a cut-off of 0.02 volts. The white level is 1 volt. In another way of looking at things, the voltage level of the black shoulders is assumed to be 0 V. The level values are accordingly −0.3 V for the line sync pulse and 0.7 V for the white value, which should not be exceeded (sound interference due to the difference tone method).
The front shoulder of the blanking area connects to the image content on the next line.
The nominal reference time is the middle of the leading edge of the sync pulse. With direct synchronization, this edge triggers the line sequence. Direct synchronization has been replaced by flywheel synchronization , which is practically indispensable for SECAM decoders. With flywheel synchronization, errors in the detection of the edge are averaged over time.
The electron beam of the receiver first draws the first field ("odd lines" = lines 1, 3, 5, 7, ... etc.) and then the second field ("even lines" = lines 2, 4, 6, ... etc.) . When the display of the line content is finished, the falling edge of the line sync pulse triggers the line return, in which the beam jumps back to the beginning of the next line. This happens very quickly and the electron beam is blanked out. When the first field is finished, there is a so-called image return (vertical return ).
Front black shoulder
If the end of the line is white (level at 1 volt), the level should drop very quickly to 0 volts, which is not possible for technical reasons (signal bandwidth). This is the case in our picture, so there would be a delayed line return. The result would be incorrect synchronization between the transmitter and the TV. To prevent this from happening, insert the front porch with a duration of 1.5 microseconds. However, this shortens the visible line by the same amount.
The presence of the back porch has a circuit-related cause. After particularly fast line return of the line content (trace) occur at the beginning Einschwingerscheinungen on. The rear black shoulder serves as a buffer so that these vibrations have subsided in good time before the start of the picture content. It is also used to determine the black level ( clamping circuit (communications engineering) ).
The line sync pulse, the front and the back porch together form the line blanking interval . This can be made visible on a monitor by moving the television picture to the left and increasing the brightness to a maximum. The front and rear black shoulder can be seen as gray vertical bars and the line sync pulse in between as a black, vertical bar. Professional monitors have their own switch for this function.
The pattern of pulses for vertical synchronization is also in the level range between 0 V and 0.3 V and does not interrupt the sequence of the leading edges of the horizontal synchronization pulses:
In order to achieve a simple recognition of the vertical pulse in the home receiver by integration - for example by an RC element - and comparison with a threshold value known as the “switching level”, the vertical pulse is 2.5 lines (2.5 × 64 microseconds) long, of which in front of each horizontal flank and the middle between them go about 4.7 µs. The pulses of the second field consistently correspond to the frequency of 50 Hz and are therefore shifted by half a line in the line raster with respect to those of the first field. The interruptions of the vertical pulse avoid disturbances of the horizontal deflection during the vertical retrace and transients (flywheel synchronization) at the beginning of the field.
In the studio, the duration of the individual sync pulses related to the reference edge is usually compared with a comparison value (monoflop).
The pattern for vertical deflection consists of the following pulses:
- 5 preliminary riders: Short sync pulses with half the duration (2.35 µs) and half the spacing of the normal horizontal sync pulses.
They charge or discharge the capacitor of the RC element to a defined voltage level so that the time it takes to reach the required switching level always remains the same. If they were missing, the capacitor could already be precharged due to previous image content or possibly existing interference voltages and the point in time at which the switching level is reached would not be predictable.
All falling edges of the synchronous signal (leading edges) are determined by double the line frequency. The position in time of the rising edges contains information about the type of individual pulse. The synchronous level at 0 V applies as the impulse top, the blanking level at 0.3 V as the impulse base.
The satellites cover up the fact that the V-pulse begins in one field at the beginning of the line and in the other field in the middle of the line and are therefore also referred to as compensation pulses. In the case of the leading black image, the beginning of the vertical blanking can be derived from the trailing edge of the first leading edge.
- 5 main pulses : sync pulses that last almost half a line (27.3 µs).
They charge the capacitor and thus synchronize the vertical oscillator, the vertical return is initiated. The time until the switching level is reached depends on the device. There is even a good chance that the switching point will be determined by one of the interruptions between the main pulses. In this case in particular, the temporal position in each field is correct. Errors lead to a pairing of the lines.
- 5 Nachtrabanten: Short impulses like the Vorrabanten.
They discharge the capacitor again, so that the switching level cannot be reached again immediately after the main pulses due to any interference pulses that may occur. Otherwise these interference pulses could lead to renewed (incorrect) synchronization. Because of the interruptions in the vertical pulse and because of the satellites, a counting frequency meter shows 15640 pulses per second.
25 full images are transmitted per second, each full image consists of 15,625 / 25 = 625 lines, of which, however, only a maximum of 575 lines are visible in color; the rest represents the “vertical blanking interval”. On most real televisions only a good 550 lines can actually be seen, the rest of the theoretically visible lines “disappear” behind the picture edge (so-called overscan, see also action field and title field ).
Each individual frame consists of two "fields". The first field contains only the odd-numbered lines of the frame, the second only the even-numbered lines. The two fields are interlaced one after the other. They are displayed like two nested combs. The flicker of the image is reduced by the sluggishness of the human eye and the afterglow of the tube. However, this procedure can lead to noticeable flickering in thin, horizontal lines. Televisions with 100 Hz technology or LCDs first have to combine the two fields into a single image ( deinterlacing ). Due to the higher temporal resolution of the fields, comb effects can occur in horizontally moving objects that cannot be seen on conventional 50 Hz tube televisions.
Depending on the television standard, exactly 15,625 picture lines are transmitted per second in Europe, i.e. a complete picture line takes exactly 64 microseconds.
The tolerance of the frequency was for black and white television and has been improved with regard to the PAL method . Starting from the mother frequency, the distinctive points in time were also obtained for color television with the help of a series of tilting processes ( monostable multivibrator ). In the vertical direction, the relatively imprecise times of the cost-saving astable multivibrators were synchronized by the more precise times of the mother frequency.
Image data is transmitted during the active line duration of 52 µs, plus 1.5 µs on the front porch, 4.7 µs sync pulse and 5.8 µs on the rear porch. The two black shoulders and the sync pulse are collectively referred to as the horizontal blanking interval , which lasts a total of 1.5 + 4.7 + 5.8 = 12 microseconds. The transmission of the burst (see below for color representation) is within the rear porch and begins 5.8 µs after the start of the sync pulse, i.e. 1.1 µs after the start of the rear porch if the timing is correct. The burst or burst pulse lasts 2.25 µs and contains around 10 sinusoidal oscillations in the FBAS signal. At the end of the back porch, the image data of the next line begin again. There is no burst with black and white broadcasts; the receiver detects their absence and turns off its color decoding circuitry. If it did not do that, a black-and-white transmission would be stored with the color noise resulting from the color decoding. However, today almost all television stations broadcast all programs with bursts, including black and white programs. In this case, the color display is prevented by the transmitter by filtering out the relevant frequencies.
The color-picture-blanking-synchronous signal ( FBAS ; English CVBS , Color, Video, Blanking, and Sync. ) Is also known colloquially as the "color television signal ".
Except for the color components (gray in the picture below), it basically has the same structure as a BAS (black and white) signal and can therefore also be reproduced on a black and white receiver, which was very important when color television was introduced ( compatibility ).
This diagram shows the oscillogram of a line of a PAL- modulated television picture for the standard bar sequence with a color saturation of 75 percent and a brightness also reduced to 75 percent in the colored bars yellow, cyan, green, magenta, red and blue. ( EBU - test signal ). It is divided into the following sections:
- The picture starts with a white bar.
- This is followed by further bars with the colors in decreasing brightness. You can see the transmitted color information, colored gray in the picture. The color intensity is expressed in the strength of the color signal (here: height of the gray area in the diagram), the color in the phase position relative to the color carrier (in the diagram, see number 5). The color information is z. B. with PAL in the frequency 4.43 MHz and because of this comparatively high frequency is not recognizable in this representation as a sinusoidal oscillation, because each color bar contains almost 30 full oscillations of the color signal. The vertical and horizontal lines shown in the drawing can only be seen where they are not washed out by the color signal. Examples: Lines are not recognizable to the right and left of the burst pulse (trapezoidal before the modulation!), The representation of the color bars is inconsistent.
- Black and the front porch. No color modulation can be seen with black and white because these colors are "achromatic", that is, their color saturation is zero. Since in the quadrature modulation used by PAL and NTSC no carrier is transmitted, in contrast to the frequency modulation used in SECAM , this feature can be used to distinguish between PAL and NTSC from SECAM. With SECAM, the (unmodulated) color carrier would be recognizable on the brightness signal for white and black; on the oscilloscope these would look just like a colored bar to the naked eye.
- The sync pulse with a length of 4.7 µs.
- The rear black shoulder with the PAL burst (English for "burst into"). With PAL, the color information can only be seen with colored picture content. To decode them, an oscillator is required which is synchronized with the color carrier (which is suppressed with quadrature modulation). The burst is used to synchronize the receiver with the (otherwise suppressed) color carrier. Approximately 10 sine waves of the color carrier are transmitted directly; the circuit that generates the new color carrier in the receiver is synchronized with the transmitter's color carrier in terms of frequency and phase position during this period; for the rest of the line it can then work independently on the basis of this coordination. The oscillogram of a SECAM signal would be similar, since there, in order to recognize that SECAM is present and not PAL or NTSC, the carrier is transmitted unmodulated during this phase. (In the early days of SECAM, special image lines were used for this purpose during the vertical blanking interval, but they were later to be available for teletext , VPS and other services.)
- The beginning of the next line.
Broadcasting the television signal
A video signal (e.g. FBAS) that is transmitted directly (without modulation) on a line is called composite video . For transmission over long distances ( terrestrial , satellite, cable) the video signal, also called the baseband signal here, is modulated onto a carrier signal . In this way, several video signals can be transmitted simultaneously over a route and the costs of the route are divided among the video signals transmitted. Most television standards use negative amplitude modulation - the lowest voltages (synchronous pulses) of the composite signal correspond to the highest field strengths of the radio signal, and conversely, the highest voltages (white areas in the picture) correspond to the lowest field strengths. The advantage of this arrangement, which at first appears illogical, is that typical short interference pulses are then not found in the image as very conspicuous white, but rather as inconspicuous black points. In addition, with negative modulation, the automatic gain control in the television receiver can be implemented more easily in terms of circuitry.
This picture shows a line of a modulated television signal as one could display it with an oscilloscope if a gray bar test image is used. It only shows the positive half-waves of the carrier signal (shown in red) with the modulated BAS signal. The negative half-waves contain the same BAS signal again. Therefore you would have to "fold it down".
In addition to the carrier frequency, amplitude modulation also produces other frequencies, the so-called sidebands. They lie on the frequency axis below and above the carrier frequency and both reach the width of the highest modulation frequency. At a maximum video frequency of approx. 5 MHz, the video signal alone would occupy a bandwidth of 10 MHz. The two sidebands would each individually contain the full information of the signal. Therefore, one could theoretically do without the transmission of one of the sidebands and thus halve the bandwidth requirement. The technology of single sideband modulation required for this is relatively complex on the receiver side, so a compromise was made. On the transmitter side, one of the two sidebands is partially removed. This broadcast signal with partially suppressed sideband ( residual sideband modulation ) allows a denser occupancy of the frequency bands and led to a channel spacing of only 7 MHz in the German VHF band .
This picture shows the frequency spectrum of a television signal : CCIR standard for the image transmitter amplitude frequency response (above) and the receiver transmission curve (below). The specified frequencies refer to the PAL B / G television standard used in Germany and are specified relative to the video carrier.
- The lower residual sideband. The television signal is amplitude-modulated , only part of a sideband being transmitted. The rising edge of the band filter in the receiver is called the Nyquist edge . The lower residual sideband is around 0.75 MHz wide (1.25 MHz are transmitted).
- The image carrier. It is not drawn to scale and has a little more than twice the amplitude of the sidebands. The point at which the Nyquist flank intersects the beam is called the Nyquist point. It is centered on the flank of the filter. (It should not be confused with the Nyquist point , which is an important part of considering the stability of control loops.)
- The brightness signal. This extends to about 5 MHz.
- The color signal is nested in the upper part of the brightness signal . The color carrier is at 4.43361875 MHz and is suppressed - it is regenerated in the receiver.
- The sound signal. The sound carrier is 5.5 MHz. Its level is 12 dB lower than that of the video carrier.
- The second sound carrier if there is a stereo or two-channel sound transmission. It is 242.1875 kHz above the first sound carrier and its level is 6 dB lower than this.
Both sound carriers are frequency-modulated with a frequency deviation of 50 kHz .
- ↑ Expertise for radio mechanics, Jürgen Heinrich and Rainer Ludwig, VEB Verlag Technik Berlin, 1965, page 33
- ↑ Single sideband and residual sideband modulations , Prof. Dr.-Ing. Dietmar Rudolf, TFH Berlin
- ↑ Ohm Lüke, Signal Transfer, p. 364