Analog television signals represented video and audio as a continuous range of values that assumed an infinite number of states. Minor imperfections in the channels that distributed or transmitted these signals produced noticeable errors in the picture or sound. Quality decreased steadily with increased degradation in the channel. Analog broadcast engineers recognized the onset of channel impairment by simply watching the television broadcast. With experience, they classified the type and level of impairment and took corrective action before quality degraded to an unacceptable level. Monitoring instruments add precision to this basic quality-control approach.
Unlike analog, DTV signals represent video and audio information as a discrete set of values that can assume only a finite number of states. Minor imperfections in the channels that distribute or transmit these signals generally have no noticeable effect on picture or sound quality. Quality remains high as channel degradation increases until the impairment level reaches a threshold point. At this “digital cliff,” quality decreases to an unacceptable level. Thus, digital system engineers cannot detect the onset of channel degradation by watching the broadcast; they can only react to severe quality problems once they appear.
To overcome this, digital broadcasters need monitoring approaches that let them proactively address channel degradation before it leads to noticeable quality problems. Broadcasters that use monitoring instruments that can detect impairments before they impact quality can achieve the same level of confidence in DTV that they achieved in analog. These monitoring instruments are called confidence monitors. Systems built from these instruments are called confidence monitoring systems. Requirements for confidence monitoring instruments and systems for digital television facilities are based on quality control and system management challenges that broadcasters face. While historically measuring RF and Transport Stream transmission impairments (QoS), today's confidence monitoring systems now also measure Quality of Experience (QoE).
By supporting the convergence of video, voice and data distribution systems, the transition from analog to digital technology also affects digital television system management. Digital telecommunication network operators gain new revenue sources by offering distribution services to broadcasters, and broadcasters can use these services to reduce operating expenses.
However, this complicates the process of maintaining quality by introducing additional transitions in the distribution chain. As one company hands content off to another, broadcasters must rely on other companies to meet contractual QoS obligations.
Technology convergence has also facilitated new approaches to system management. Many broadcasters are using management techniques that resemble the centralized monitoring and management systems seen in telecommunication facilities. These systems rely on network-capable confidence monitoring devices that can report status and send alarms to a central Video Network Operations Center via standard network communication protocols. The digital-cliff effect, the increase in the number of handoffs and new centralized management approaches are factors driving the characteristics of confidence monitoring systems in DTV.
Confidence monitoring solutions must also address the quality control challenges arising from the layered structure of digital television systems. As broadcasters transition from analog to DTV, fundamental differences in these technologies are leading to new approaches for ensuring broadcast quality and reliability. In this primer, we describe quality control and system management challenges digital broadcasters face and the monitoring devices that have been developed to address these challenges.
With DTV networks, broadcasters use digital signal processing and digital data processing techniques to improve quality and efficiency in their broadcast networks. Hence, distribution and transmission channels in digital television systems contain sequences of signal processing and data processing steps. We can best understand how these steps interact and impact broadcast quality by organizing them into a layered model. (See Figure 1.) Specifically, we can use three layers to model a digital television broadcast system.
In the Formatting layer, TV content producers create and format the video and audio that broadcasters will deliver to the end consumer. Signal processing in this layer includes: the sampling, quantizing and formatting steps needed to create DTV signals; conversion between digital formats; and displaying a digital signal on a TV set or picture monitor.
In the Compression layer, content producers and broadcasters compress and aggregate content for storage, distribution or transmission. Signal processing in this layer includes video and audio compression. Data processing in this layer includes: multiplexing programs and system information into a single data stream; fragmenting this stream into a packet protocol; and recomposing programs from packets for decoding.
In the Distribution layer, broadcasters process content for distribution over internal networks or delivery to the end consumer through DTV transmission systems. Signal processing in this layer includes techniques for modulating digital signals onto RF carriers. Data processing includes error correction algorithms for transmission and formatting needed to embed content into network communication protocols used in internal distribution.
Adding digital signal and data processing to broadcasting introduces new sources of errors, with different types of errors in each system layer. At the Formatting layer, broadcasters face challenges in dealing with the wide array of new formats for both standard and high-definition DTV. They need to ensure correct colorimetery and verify conformance to standards. In addition, they may need to convert from one format to another, such as downconverting HD content for broadcast on an SD system. These format conversions can introduce quality errors. Also, separate processing of digital video and audio can lead to synchronization problems.
Compression introduces new types of quality defects, like blockiness. Errors can occur during the complex process of multiplexing programs and system information into a single data stream. Errors in timing and synchronization parameters can compromise the decoding process and lead to noticeable content quality errors.
At the Distribution layer, broadcasters encounter familiar RF technology in the transmission networks; however, these systems use different modulation techniques and offer new challenges in understanding coverage and interference problems. For internal distribution, broadcasters are increasingly relying on IP networks, introducing problems with latency and packet loss.
From source to consumer, program content typically moves through these system layers many times. Transitions between layers can dramatically alter the nature of the digital information — for example, moving between uncompressed digital video at the Formatting layer and compressed digital video at the Compression layer. The additional processing needed to move across layers increases the probability of quality errors at these transitions.
Moreover, errors in one layer can cause errors in a different layer, in some instances masking the original error source. For example, blockiness errors can arise from problems in a compression step (Compression layer) or as a consequence of uncorrected bit errors in the receiver (Distribution layer). Similarly, transmission errors can occur due to failures in the modulation steps (Distribution layer) or from variations in the data rate from the multiplexer feeding the studio-to-transmitter link (Compression layer).
To meet the quality control and system management challenges described above, confidence monitoring systems should have the following characteristics:
- Layer-specific probes that detect the different types of errors seen in digital television systems;
- Extended monitoring capability to give broadcasters advanced notification of system degradations before they become quality problems;
- Multi-layer monitoring that lets broadcasters quickly isolate the root cause of a quality problem; and
- Network control that supports the new system management challenges.
In a confidence monitoring system, we can think of each monitoring device as a probe, monitoring quality at a particular point and layer in the distribution and transmission chain. Broadcasters need to use different probe types for quality control at different layers. At the Formatting layer, digital waveform monitors or rasterizers help broadcasters detect many quality problems. These probes monitor characteristics of the digital signal, including gamut errors, audio loudness and ANC data parameters.
The MPEG-2 standard defines the basic processing steps and techniques used at the Compression layer. Broadcasters need MPEG monitors capable of detecting problems in basic MPEG processing, as well as the additional processing defined in the DVB, ATSC or ISDB-T/Tb broadcasting standards based on MPEG. These MPEG monitors can also decode the elementary streams and detect blockiness and other picture impairments.
At the Distribution layer, broadcasters need MPEG monitors with RF interfaces to detect quality problems in a wide variety of distribution and transmission channels. MPEG monitors can monitor RF transmissions in DVB-S/S2 or ATSC formats. MPEG monitors can also be used at test points with fiber-based GigE IP backbones.
We can also distinguish confidence monitoring probes by the level of monitoring that the probe offers. Basic confidence monitoring probes track a small set of key quality parameters. They act as an “indicator light,” telling broadcasters when something has gone wrong. However, basic confidence monitoring probes do not offer a complete solution. While they can enhance broadcasters' ability to react to a quality problem, they do not give the information needed to proactively address system degradation before it becomes a quality issue.
Extended confidence monitoring probes use more sophisticated analysis to make additional measurements of quality parameters. They act as “indicator gauges,” telling broadcasters when something is going wrong. RF transmission monitoring offers a good example of this distinction.
Basic RF confidence monitors measure bit-error-rate (BER). BER will remain low until the transmission approaches the digital cliff. Then, it will increase dramatically as the transmission falls off the cliff. This gives broadcasters only slightly more time to react than they would have by watching the transmission on a picture monitor.
Extended RF confidence monitors add additional measurements like Modulation Error Ratio or Error Vector Magnitude. These measures will noticeably change as system performance degrades, giving broadcasters early warning of potential quality problems, and an opportunity to make necessary adjustments or seamlessly transition to their backup systems.
To have confidence that their facilities are operating correctly and efficiently, broadcasters generally need to probe at all layers. Probing at only one layer can give a misleading picture of system health. We began with a simple example of this problem. By watching the broadcast on a waveform monitor or rasterizer, broadcasters are probing quality at the Formatting layer. Monitoring at this point before encoding offers little information about the digital transmission system.
Similarly, monitoring just the MPEG protocol or the RF transmission will only yield partial information. To gain a complete picture of system quality, and to quickly detect and isolate quality problems, broadcasters need multi-layer confidence monitoring solutions. (See Figure 2.)
System management concerns also impact confidence monitoring. Broadcasters often need to monitor at geographically separated locations. For example, broadcasters accepting contribution feeds over a telecommunication network may want to install confidence monitoring probes at the network operator's points-of-presence. These distributed probes will need network capability so that they can report status and also alarm conditions to a central location. Network monitoring software can correlate this information to help engineers identify the root source of any quality problems. (See Figure 3.)
PHYSICAL, ECONOMIC FACTORS
Form factor and cost are additional considerations in developing a confidence monitoring system. Large, card-modular solutions may work well in central nodes with a large number of signals and multiplexes, while small single-channel probes may work better in remote locations like transmitter sites.
Because of the digital cliff, broadcasters can only detect the onset of quality problems by viewing the broadcast and, therefore, will need to use confidence monitoring systems. Combining digital signal processing with digital data processing creates layers in DTV systems, driving the need for multi-layer confidence monitoring systems that contain layer-specific probes designed to quickly detect and isolate quality problems. These systems need extended monitoring capability to help broadcasters proactively address performance degradations before they become quality problems. The systems also need network-capable probes in a variety of form factors to integrate effectively into emerging system management approaches.
As DTV evolves, broadcasters will need to increasingly rely on these distributed, multi-layer confidence monitoring systems in order to ensure optimal performance in both their distribution and transmission systems.
Richard Duvall is the senior video marketing manager for Tektronix.