cable tech facts issue 107

In This Issue
Back To Cable Tech Facts main index.
Whether you are ready for it or not, digital video is rapidly emerging as the technology everyone will be jumping into within the next 1-3 years. Digital video offers some tremendous advantages for the system operator. First is the improved quality of the video signal itself. Theoretically, digital video is immune to the signal degradation that affects analog video signals. Problems such as signal-to-noise ratio and crosstalk in an analog system are cumulative and difficult to distinguish from the video signal (not to mention difficult to remove), but are not found in a digital video signal.
If you are already working with digital video, this Cable Tech Facts will be a good refresher. For those of you who have had no real experience working with digital signals, this should be a good introduction. This will begin a series of Cable Tech Facts that are dedicated to digital video. This first CTF will cover how a signal is digitized and some of the problems related to reconstructing an analog signal from a digital bitstream.
Converting Analog To Digital
Any real discussion of digital video should typically include a review of analog video, but in order to provide more space for digital video, we will bypass a review of analog signal processing.
The theory of converting an analog waveform or video signal to a digital bitstream involves sampling a waveform at regular intervals and recording the signal amplitude at each sample point as a digital integer value. This information is then transported to its destination and reconstructed through the use of a digital to analog converter (DAC), then displayed in its original analog form. Sounds easy doesn't it, and it is, but there are several pitfalls or possible problems along the way. Let's take a look at a digital signal from the beginning.
[Top]
Digital Sampling
To create a digital signal, you would need to digitize an analog signal. The first step in the digitizing process is sampling. As we mentioned in the theory of converting analog to digital, sampling happens at regular times or intervals. The time or interval between samples are known as sampling intervals. In order to avoid errors or distortions, it is important to capture enough samples so that the signal can be reconstructed. According to the Nyquist theorem, the signal must be sampled at least twice as often as the highest frequency component in order to accurately reconstruct the information. If we know the sample frequency and the values for each point in time, we can reconstruct the waveform by producing a voltage proportional to each value at the corresponding time.
If a signal is not sampled enough, an artifact known as aliasings can occur. This renders a decoder or digital to analog converter unable to correctly reconstruct the original analog waveform. When this signal is reconstructed back to analog, a digital to analog converter would reconstruct a lower frequency waveform than the original. In order to avoid aliasing the incoming signal, it must pass through a low-pass filter that prevents frequencies above the Nyquist limit from reaching the DAC and introducing aliasing in the signal.
A second pitfall in digitization is the size of the "steps" in the digitizing process. In other words, how can we divide the total range of amplitude to be sampled into a set number of steps. These steps are called the quantization levels. 
In Figure 1, we are digitizing a signal that goes from 0 to a value "X" linearly. We will divide this total range of levels into eight discrete digital values. The goal is to express the levels of the straight line ramp digitally. At time T1, the closest digital value that can express the level of our line is "2." At time T2, "3" is the nearest digital value. At T3, "5" is very close. If we use these values and record them, reconverting to analog would yield the curve shown in the dotted line. We have introduced errors by quantizing in steps that were too large. We cannot accurately express the actual values at each sample time.
Another consideration in choosing the number of quantization levels or steps is the available signal-to-noise ratio (SNR). Since each additional bit doubles the available number of levels, each additional bit increases the SNR by approximately 6 dB. For example, an eight bit quantization will be ~12 dB lower than a 10 bit quantization. The important factor to remember is that the number of quantizations must be chosen to provide adequate signal-to-noise performance for the required application.
In the digitizing process, a high sample rate and a large number of quantization levels would appear to be the best method. However, these two parameters do have a direct effect on the resulting data rate. By doubling either the sample rate or the quantization level will double the data rate. In real engineering practices, increased data rates usually equal increased cost. In order to keep costs down, typically the lowest sample rates and minimum number of quantization levels that will provide the necessary quality are used to reproduce the analog signal.
Digital signals have been found to provide excellent benefits to the operator, but in order to provide all the benefits digital video can bring to the table, the technician must understand how digital links break and what impairment levels can be tolerated. Impairments must be diagnosed and estimated from the modulated digital signals before they're decoded into bits. This will be even more important for cable modems and telephone signals where there is no viewable picture to evaluate.
[Top]
Industry News
Newsworthy! According the "Private Cable & Wireless Cable" magazine, Sencore was ranked #1 in test equipment supplier to the MMDS/ Wireless cable industry. They were ranked number three in the hardwire cable industry.
They're Here! BellSouth launched video services in Chamblee, GA. They are offering a 56-channel basic offering for $23.95, with 20 digital multiplexed premium channels and 18 channels of NVOD. The company is working with Americast to secure studio and network programming.
Digital Standards The Society Of Cable Telecommunications Engineers (SCTE) has formed a new sub-committee that is dedicated to digital video. The committee will be exploring the need for SCTE involvement in the development of standards for digital signal delivery through coordination of the efforts with the National Cable Association, the FCC, and other related organizations.
Warning Deadline Approaches The deadline for the federally mandated Emergency Alert System is only a little more than a year away. The FCC says that all cable operators must have a system in place that offers emergency alert information in both video and audio formats by July 1, 1997. Previous alert systems were strictly audio.
[Top]
[Sencore Home Page]
email webmaster@sencore.com for feedback on our site. Or call 1-800-SENCORE(736-2673)/1-605-339-0100 Copyright © 1997 Sencore, Inc. All rights reserved. |