CAPTER 2 Black and White Television Systems 2.1 ideo signal The purpose of a black and white television system is to broadcast black and white images. It is the most simple television system. A black and white image can be defined as a bidimensional luminance distribution that changes in time. This distribution B(x,y,t) depends on x, y, and t, where x is limited between 0 and and y is limited between 0 and, being the image width and being the image height. This luminance distribution is time variable because in the most general case the images have moving objects. The ratio p=/ is called the image aspect ratio and is 4/3 for standard television (SDT=Standard Definition Television) and 16/9 for high definition television (DT=igh Definition Television). These aspect ratios have been choosed in this way because they are related to the picture aspect ratio evolution in cinema. In cinema the picture aspect ratios are : Classic or Academy with p=1,375 Widescreen with p=1,85 Cinemascope with p=2,35 The first cinema aspect ratio was for Classic format and that is why for standard television the aspect ratio was choosed to be p=4/3=1,33. A Classic picture is well reproduced on a standard television screen. Then, the cinema developed other picture aspect ratios with bigger values. This was necessary because such bigger picture aspect ratios are related to a human visual perception parameter the viewing angle which is bigger in horizontal direction than in vertical direction. An image with a bigger aspect ratio appears more natural for the viewer. The purpose of the black and white televison system is to transform this 3 variable luminance distribution into an electric signal that depends only on time (the video signal or the image signal), to broadcast this signal and to reconstruct on a display at the reception point a luminance distribution that is as similar as possible to the original luminance distribution in front of the video camera. The possible differnces between the original and the displayed luminance distribution must be small enough not to be visible for the human visual system. In order to generate the video signal from the B(x,y,t) luminance distribution the first step is to time sample this function. Succesive frames will be broadcast, these frames representing the luminance distribution at distinct time intervals : B(x,y,0) ; B(x,y,T) ; B(x,y,2T) ;... It is something similar to the display process in cinema, where succesive pictures on the cinema film are displayed with a speed big enough to surpass the time inertia of the human eye (the critical frequency) and in this way the human eye has the
SISTEME DE TELEIZIUNE ALB-NEGRU sensation of a continuous image with moving objects. For white light and medium luminance of the display on the order of hundreds of candela per square meter this critical frequency (or flicker frequency) is around 46 z, but it increases as the decimal logarithm of the medium screen luminance. In cinema the film runs with 24 images per second (24 z that means under the 46 critical frequency) and that explains the flicker effect on the older movies. In order to solve this problem, in today cinema display devices the film images are displayed twice (with 48 z instead of 24 z that is over the 46 z critical frequency) and the displayed image is not flickering. In television systems, frame frequency f=1/t, was choosed over the 46 z critical frequency equal to the AC power line frequency in each country (50 z or 60 z). The purpose was to minimize the visibility of a possible AC hum on the T set screen. The perception of any kind of interference pattern superimposed on the desired image on the screen is less annoying if the interference pattern is stable or is slowly moving on the screen. At the begining of the television systems, the stabilizer circuits where not as good as today, and an AC hum was possible to exist on the video signal and on the displayed image. If frame frequency equals AC line frequency the possible interference hum pattern is stable on the screen or is moving very slow up or down on the screen in the case of a small allowed AC line frequency change from the nominal value. The brightness and contrast of the display devices used in T sets had a constant growth in their evolution. In this way it was possible that for the brighter parts of the image (white, yellow...) the critical frequency may not be 46 z but bigger, perhaps over 50 z. That evolution can give for the brighter parts of the screen a flickering effect. This problem was solved when it was possible to build a frame memory. In the T sets with flicker free parameter, the frames comes with standard 50 z, but they are stored in a frame memory and read and dispyed twice with 100 z or even higher (200 z or even 600 z with frame interpolation for good high velocity movement dispay). The second necessary step in video signal generation is to scan each frame line by line from left to right and from the top to the bottom of the frame with constant velocity, the output voltage of the image sensor being proportional to the luminance on each line (this output voltage is the video signal). From the mathematic point of view, this scaning process is in fact a vertical sampling of the luminance distribution B(x,y,t), becuse in this way the luminance variation in vertical direction is not continuously transmitted. Instead, the luminance of the vertical succesive lines is broadcast. In analog television the sampling stops here. In digital television the sampling will be complete by sampling B(x,y,t) also on x axis. When the image is displayed on the T set it will be a vertical interpolation due to the human eye and to the display device which has a certain aperture. A biger line scanning number will give a better vertical resolution on the displayed image. The line number was choosed corresponding to the human eye resolution. For standrad television (SDT = Standard Definition Television) the standard viewing distance was considered to be D = [ 4 5] where is image height. That is equivalent to 3 times the screen diagonal (if aspect ratio is 4/3, then the ratio between height and diagonal is 3/5). From this distance the human eye must not view the scanning line structure. The eye resolution for black and white details is 1/60 degrees. Line number is determined by dividing the vertical angle for the whole screen (around 10 degrees) to the eye resolution. 2
SISTEME DE TELEIZIUNE ALB-NEGRU 2arctg Line number is Z = 2 D, that is 600 lines. ϕ The exact value of the line number value was choosed in order to simpify the sync generator construction. Any video generator device must have a circuit, called sync generator that gives all the necessary pulses for building the complex video signal. In this circuit are frequency dividers with line number. In order to simplify the sync generator circuits line number was choosed to be a multiple of prime numbers (3, 5, 7 and so on) For the 50 z standard line number was choosed to be Z = 625 = 5 x 5 x 5 x 5 (standard 625 / 50), and for the 60 z standard line number was choosed to be Z = 3 x 5 x 5 x 7 (standard 525 / 60). In DT (igh Definition Television) the viewing distance is 3 times image height which goes to a bigger line number, usually twice, as in standard television. In all television systems the scan is from left to right and up to down on each frame. The time interval equal to T d is called the active part of the scaning line and at the image sensor output it is generated a video sugnal prportional to the luminance of the pixels on that line. On the reverse part of the scaning line equal to T i, much smaller, there is a movement from left to right in order to continue the scaning on the next line. In a similar matter after scaning one frame on the active part of a frame from up to down, there is a reverse vertical time interval from the bottom to the top of the screen in order to continue with the scan of the next frame. In the case of a cathode ray tube sensor or display the horizontal and vertical driving signals have a sawtooth shape on T = T d + T i and T = Td + Ti, as in fig.2.1. in an example with a few scaning lines on a frame. Figura 2.1 3
SISTEME DE TELEIZIUNE ALB-NEGRU 2.2 Interlaced and progressive scaning The scaning described until now is called progressive or continuous scaning. It is not the scaning process used in television systems, due to the large video bandwith that results in this case as it will be described at the video spectrum later. In order to solve this problem interlaced scaning was choosed. 2.3 Interlaced scaning A frame is divided in two subframes or fields, the odd lines field and the even lines, as displayed from top to bottom on a frame. In the first step the odd lines (1,3,5,...) are scaned. And then, in the second step the even lines (2,4,6,...) are scaned. The fields have a repetition frequency f witha bigger value than the critical > f. In this case the video bandwidth is halved because the frame frequency, f C frequency is half the of the field frequency. For example, if the field frequency is 50 z, the frame frequency is 25 Z, and the video bandwidth that is proportional to the frame frequency is two times smaller than in the noninterlaced (progressive) scan. There is also a drawback of the interlaced scan, the vertical resolution of the interlaced scaned image is reduced at 70 % compared to the vertical resolution of the image with progressive scan with the same line number. 2.4 Scan parameters For the 625 lines, 50 z and 2:1 interlaced ratio (625/50/2:1) the scaning parameters are: f = 25 z, f = 50 z, f = 15625 z, T = 18,4 ms, K Ti = 1, 6 ms, Td = 52 μs, Ti = 12 μs. Tje ratio between the inverse vertical interval (1,6 ms) and line period (64 microsecunde) is 25. 25 de lines is the reverse vertical interval after each field. For the odd and even fields there are 50 return lines. For this standard only 575 lines from the 625 lines are active lines and these 575 lines must be considered for the vertical resolution. d 2.5 Black and white televison signal components 2.5.1 ideo signal It is the main comonent of the television signal and has the information regarding the luminance of the pixels on each active line. 2.5.2 Blanking pulses B and B On the return vertical and horizontal interval blanking pulse with a blanking level under the black level in the video signal (3% from the peak to peak televison signal amplitude lower) are used. In this way the horizontal and vertical return intervals are not visible on the display. 4
2.5.3 Sync pulses S and SISTEME DE TELEIZIUNE ALB-NEGRU S Thes pulses are sent on the horizontal and vertical intervals in order to synchronize at the receiver the scan oscillators. In this way the scan will be the same at the image sensor and at the T set display. The level of the sync pulses is below the blanking level. In this way these sync pulses can be recovered from the television signal with an amplitude comparator. Fig.2.3 The sync pulse amplitude is 30 % from the peak to peak television signal amplitude. After the complex sync pulses S + are recovered at the output of the amplitude comparator, the horizontal and vertical sync pulses are separated taking into account their time interval. The horizontal sync pulses are 4,7 microseconds long, and the vertical sync pulses are 160 microseconds long (2,5 line period). 2.5.4 Additional pulses On S, with 2,5T, are introduce 5 pulses at T 2 intervals. In this way during the vertical sync pulse the horizontal oscillator in the receiver has a reference. The beginning of a line is the front part of the horizontal sync pulse and the beginning of a field is the front part of the vertical sync pulse. Before the vertical sync pulse there are 5 pre-equalizing pulses at half line interval and after the vertical sync pulse there are 5 post-equalizing pulses at half line interval as in fig. 2.4. 5
SISTEME DE TELEIZIUNE ALB-NEGRU Fig. 2.4 2.6 ideo complex signal (CBS) By adding all the above signals the complex black and white video signal is generated (CBS = Complex ideo Blanking and Sync). There are 4 amplitude levels in this signal : white level, black level, blanking level, sync level. If the video complex signal has 100% amplitude between the sync level and white level, then video signal has 67% amplitude between the black level and the white level, sync pulses have 30% amplitude between the sync level and the blanking level, and the distance between thw black level and the blanking level (the guard interval) is 3%. 2.7 ideo signal spectrum 2.7.1 ideo signal limits Minimum video frequency The smallest video frequency is 0 z, that corresponds to the average value component of the video signal, depending on the image content at each moment. Maximum video frequency Maximum video frequency is proportional to the horizontal image resolution. This value is limited at the output of the video camera at a value that gives the same horizontal and vertical image resolution in each T standard. ertical resolution is determined by line number selection for each T standard. Maximum video frequency is then given by the formula : N p N ( 1 x y kp k ) 2 fmax = k = k = Z f K 2 T 2T 1 k 2 1 k ( ) ( ) a, 6
SISTEME DE TELEIZIUNE ALB-NEGRU where T K is frame frequency, Z is line number, SDT and 16 9 k T in DT), k is vertical return loss factor, p = is iamge aspect ratio ( 4 3 in Ti Z Za k = =, T Z i = is horizontal return loss factor, k is Kell factor (k=1 for progressive scan and T k=0,7 for interlaced scan)., For 625/50/2:1 standard : k = 0,08 and k = 0,18. Then f max = 5Mz. For 525/60/2:1 maximum video frequency is 4,2 Mz, which is a normal result, because in this standard the vertical resolution is also lower due to the reduced line number compared to 625/50/2:1 stanard. For a 1250/50/2:1 standard (a DT system), maximum video frequency is 27 Mz, because also the vertical resolution is higher due to the bigger line number. 2.7.2 ideo signal spectrum structure For a static image the video is didcrete, with components situated at : fkm = kf + mf, where km, N. This is due to the fact that the periodic scan determines a periodic video signals with periodes related to horizontal and vertical scan periods. For a moving picture the video spectrum has a structure with small bandwidth (maximum 6 z) at frequencies as above. This is due to the periodic scan operation and to the fact that the velocities of the moving objects necessary to be transmitted in a television system is much smaller than the scan velocities. The inertial behaviour of the human visual system limits the velocities of the moving objects that have to be transmitted. So, the spectrum of the video signal has many free places that are used to transmit the colour information in a colour television signal as will be described in the colour television chapter. 7