When viewers encounter such difficulties as audio lip sync, blocking or black screens, they turn to another channel. Therefore, it is imperative that television engineers find and fix network, encoding and transmission problems before their viewers become aware of them.
Many broadcast stations use set-top boxes (STBs) and video monitors to confirm that they are correctly on-air. With an analog signal, this might have been sufficient. Digital broadcasting, however, introduces another element to the monitoring equation — software.
Every digital STB has software running on it. Depending on the implementation of the software in a specific STB, the receiver may react differently to a specific noncompliance in the bit stream. Problems that affect users of one type of STB may not be visible to users of another STB brand, or even a later model from the same manufacturer.
For example, one common error found in bit streams is a conflict between Program Specific Information (PSI) and Program and System Information Protocol (PSIP). Some STBs can intelligently differentiate between correct and incorrect information and still operate properly. Viewers using these STBs enjoy uninterrupted service. Other STBs may become confused, fail to display video, or even refuse to tune to the correct channels.
If a content provider is using one of these STBs to monitor its network, the engineers in the media center or control room are essentially blindfolded. They may never see problems in the delivered stream.
ATSC created a recommended practice (A/78) to address this problem. Although ATSC standards strictly define the contents and characteristics of the DTV emission transport stream, there are several interactions and interrelationships among components that can create momentary noncompliances. Successful tuning and display of programs can be ensured if the transport stream adheres to the applicable specifications.
Unfortunately, due to the complexity of an ATSC stream, there may be times when minor noncompliances are inevitable. Another difficulty driven by the complexity revolves around the severity of an error. Some errors, while violations of the standards, do not produce a perceived quality issue for the viewer. When minor errors trigger alarms, they can produce operator fatigue. When red lights frequently flash for inconsequential problems, important alarms often go ignored.
In the development of A/78, this problem was recognized, and addressing it became part of the design. Error types are now classified by severity, ranging from unwatchable to invisible. This kind of distinction allows manufacturers of monitoring devices to use filtering techniques, which show the broadcaster only those problems that matter and hide (while still logging) those that don't.
Figure 1 shows a single reference analysis point for proper signal verification in an idealized system. However, several additional monitoring points should be considered.
A/78 identifies transport stream issues by type, which helps operators better identify the potential root cause of the problem. Here are some of the key error types to monitor:
An ATSC conformant transport stream is also required to be MPEG-2 conformant. Therefore, an ATSC transport stream must include the two mandatory PSI tables: the Program Association Table (PAT) and the Program Map Table (PMT).
The PSIP is the glue that holds the DTV signal together. The purpose of PSIP is to describe data at the system and event levels, and to define an abstract collection of programs (a virtual channel). Problems with PSIP can cause viewing or EPG difficulties.
Timing is the key to the MPEG-2 encoding and decoding processes. The MPEG-2 standard defines a model for the system timing, adherence to which allows independent design of encoders and decoders that can interoperate. An MPEG-2 decoder's 27MHz reference clock needs to be synchronized with the equipment that is creating the encoded stream.
Before a receiver can decode a transport stream, it must identify the relationship between components in the stream. Some components contain audio and video (elementary streams), and other components contain information describing the relationship between them (metadata). The receiver uses metadata to identify each component, determine its function and select an appropriate set of components when the user selects a virtual channel for decoding. Conflicts within the metadata are called consistency errors. Consistency errors can result in broken decoding, missing system components (such as closed captioning) and missing program guide information.
These errors cover a variety of problems and are typically transport-related.
Each error type is provided with a defined error severity:
The station is effectively off-air as the transport stream errors are severe enough that transport-level logical constructs are damaged beyond utility. Receivers cannot tune and decode anything within the broadcast. The complete or repeated absence of sync bytes is an example of this level of error.
A main service (virtual channel) is flawed to the point that that the service is effectively off-air for conformant/reasonable receiver designs. This could involve improperly constructed program elements or incorrect/missing signaling about elements. The absence of an entry in the virtual channel table (VCT) for a service is an example of this type of error.
One of the program components that is signaled by PSIP or the PMT as being present is either not there or cannot be found and decoded. One example is a mismatch between the video packet ID (PID) signaled in the service location descriptor (SLD) and the actual PID used for the video elementary stream.
Parameters are out of specification by such a margin that a significant fraction of the receivers can be expected to produce flawed outputs. In many cases, the broadcast is viewable, but may exhibit some form of degradation to the viewer. An example is the master guide table (MGT) cycle time being somewhat larger than the specification, which causes slower than normal channel-change tuning.
Violates the letter of the standard, but in practice will have little effect on the viewing experience. Errors of this type should be corrected, but do not have the urgency of higher severity errors. An example is a single instance of a 152ms MGT cycle time (with the remainder of the MGTs coming at less than 150ms intervals).
By using monitoring devices that adhere to ATSC A/78, a broadcaster can quickly be informed of any problems that will impact video quality. When there are multiple issues, this kind of top-down monitoring will help the engineer resolve the most important issues first. In this way, a broadcaster can proactively ensure that his transmission will provide a quality viewing experience.
Ralph Bachofen is director of project management and marketing, and Rich Chernock is director of technology for Triveni Digital.