The first practical television broadcast systems used a form of two-to-one bandwidth reduction, or compression, called interlace. Instead of sending 50 or 60 frames per second, each frame is divided in to two fields containing half the total number of lines. The lines in the first field are every other line from the frame, say lines 1, 3, 5... and the lines in the second field fill in the missing lines during the second field as shown in Figure 1. Picture degradation due to interlace is in the form of an artifact known as inter-line twitter, however the quality is quite satisfactory for entertainment video viewed several picture heights away from the display device.

Figure 1. Interlaced scanning

In the 1950s, color television was developed. A single color picture requires three images, specifically red, green, and blue (RGB) for light emitting devices such as cathode ray tubes (CRT). Starting from the full progressive scan picture, this would require a 30 MHz bandwidth to provide the desired picture rate. Again, interlace is used to reduce the bandwidth to 15 MHz for an analog RGB system. Within a studio the signals are carried on three separate cables at 5 MHz or more bandwidth each, as shown in Figure 2. A fundamental compression scheme used in color television is to translate the three color signals into the color-difference domain where the picture is represented by a luminance (equivalent to the earlier monochrome) picture and two color difference pictures, R-Y and B-Y. Another name for this system is YUV, Y for luminance and U, V for the two color difference signals. Again using the limitations of the human visual system, in this case less color than luminance visual acuity, the bandwidth of the color difference signals is reduced by 50% for a total YUV bandwidth requirement of 10 MHz. Today, YUV signals are used in both analog and digital forms and have very little visible degradation compared to interlaced RGB video. Both forms are known as component video with YUV being used for most applications.

Figure 2. Analog television compression

Color television in the 1950s, and until recently, required further compression both to fit in the allocated 6 MHz bandwidth of transmission channels and to be compatible with the installed base of monochrome television sets. To accomplish this task the two color difference signals are further reduced in bandwidth to about 1.5 MHz each and quadrature modulated on a subcarrier which is subsequently added to the luminance signal producing composite video. Composite NTSC and PAL² produce very good entertainment quality color video in a 5 MHz bandwidth, a compression ratio of six to one from the ideal progressive scan RGB video. The final two to one compression from component to composite does bring with it noticeable picture degradation including chroma information seen as incorrect luminance and visa versa.³

Today, using modern digital compression methods, four or more excellent quality digital component television signals can be delivered to the home within the same 6 MHz transmission channel. If derived from a high quality digital component source, the resulting multiple television signals have a noticeable quality improvement over the single 6 MHz bandwidth composite video signal.

Digital Compression Methods. Digital video became a reality in 1973 with the invention of the composite-based digital time base corrector for video tape recorders. In the early 1980s, a worldwide digital component video standard was developed requiring 216 Mb/s or 270 Mb/s depending on the use of 8-bit or 10-bit sample values. This standard is commonly known as Rec. 601⁴. It is the dominant sampling structure for digital television and its use is showing rapid growth for all types of applications. Since the approval of the Rec. 601 standard, much research and development has been directed towards digital video data rate reduction resulting in a variety of video compression methods. Each of these compression methods has its own advantages, disadvantages and picture degradation characteristics. It will be important for any general purpose picture quality measurement instrument to provide a result that is independent of the compression method used.

The compression method that is becoming dominant today is called MPEG-2, defined by the Motion Picture Experts Group and standardized by both the International Standards Organization (ISO) and the International Electrotechnical Commission (IEC). MPEG-2 is based on the Discrete Cosine Transform (DCT) method in combination with powerful temporal compression techniques.⁵ Although some applications may be best served by other compression methods, MPEG-2 is expected to be the most widely used method in the foreseeable future. This is because it is an agreed standard that is either optimum or good enough for a wide variety of applications, a large amount of effort is going into the development of chip sets for lower cost encoders and decoders, and the forthcoming large installed base will be attractive for many equipment manufacturers and application developers.

The Modern Television System. A simplified block diagram of a modern compressed television processing and transmission system is shown in Figure 3. Television nominally consists of audio and video; however, the system may include data and control signals (not shown in the figure) hence may be thought of as a multimedia system. One-direction transmission is shown. A multiplicity of methods are depicted, particularly in transmission, making this diagram an overview of many types of applications. No specific compression method is shown, however the MPEG-2 transport stream is shown in the transmission area since it can be considered a general purpose multiplexing scheme capable of carrying any type of compressed video and audio.

Figure 3. Modern television system.

Analog RGB video is produced in the camera and processed into one or more of several possible formats; analog composite, digital composite, analog component or digital component. Full-bandwidth digital video is an extremely important part of the television system today. Program production processes must be full-bandwidth digital (or analog) in order to manipulate the picture to produce desired artistic results.

Following program production, the television signal may be compressed for storage, efficient transmission or intra-facility interconnection in digital form. Typically this will be MPEG-2 compression resulting in an MPEG transport stream (MTS) that may be multiplexed with other MPEG transport streams for transmission or interconnection.

New systems for RF transmission of television signals use digital modulation schemes which are generally more robust for the same transmitted power and provides the digital channel for compressed television signals. It is important to note, that even compressed digital video broadcasting to the home often starts with full bandwidth digital video to drive the bit-rate efficient, statistically-multiplexed compression system.

The broadband telecommunication system provides a variety of transmission methods. Traditionally these have been voice channel oriented with special data mapping for digital television signals. Although direct mapping of the MTS into the digital telecommunications hierarchy is in the process of being standardized, it is expected that asynchronous transfer mode (ATM) will become the preferred method of inter-facility video transfer.

Video testing in this modern television system is not just a matter of developing new techniques to evaluate the effects of compression. The significant portion of the system utilizing analog and full-bandwidth digital signals requires application of traditional analog and recently developed digital test methods. To determine picture quality impairments caused by compression, a video measurement system must take into account the various signal format changes affecting the video throughout the system.

Return to top