Audio monitoring for TV: Beyond monitors and meters
Numerous articles have been written about our industry's migration to digital and its impact on both the broadcaster and the home viewer. In general, emphasis has been placed on visual elements, such as compression technologies, emerging formats and conversion strategies. However, the impact of the audio elements is equally important — and merits discussion.
The simple days of the all-analog facility are gone and have been replaced with a complex, multichannel digital audio environment that poses several workflow challenges. Not only does the broadcaster need to design and configure a flexible audio infrastructure, but also a comprehensive and effective audio monitoring capability must be readily accessible to ensure the highest possible audio quality to complement the visuals.
Monitoring in the midst of complexity
Because most facilities deal with a complicated mix of multichannel formats, there is no single way to transport and monitor the audio. At different points within the ingest-to-transmission workflow, the audio signal may reside in one or more of the following formats: analog, AES, embedded AES, Dolby D, Dolby E and embedded Dolby. Each of these formats has its own set of monitoring challenges.
Figure 1. TVM-950HD display showing comprehensive embedded Dolby metadata monitoring table.
Even if a broadcast facility attempts to standardize with one particular schema, operators should still be prepared to handle multiple formats at the point of ingest. Because of this, equipment manufacturers (particularly those that design servers, routers and master control switchers) are now building more flexible audio capabilities into their devices in an attempt to satisfy all combinations of possible audio workflows.
If the facility desires to fully monitor audio signals in order to identify and prevent problems, traditional speakers and meters are simply not sufficient in today's digital environment. Problems can arise that are not readily evident — particularly in a noisy master control or machine room situation. Phase errors, for example, can cause audible distortions. And non-synchronous conditions at the clock level (especially with embedded formats) can cause audio pops, clicks and dropouts. The number of possible HD video formats also complicates this situation.
However, with an advanced audio monitoring tool online — one that goes beyond monitors and meters — the broadcaster can locate problems, prevent them and ensure that the highest-quality signal is provided to transmission.
Figure 2. This display shows an example of the Nordic scale, a commonly used regional scale.
Yes, metadata can be monitored
The monitoring challenge is compounded because new digital formats carry both sound and metadata, the digital information that describes the audio's inherent characteristics. (See Figure 1.) Audio metadata exists in both compressed and uncompressed formats, and it should be monitored as diligently as the audio itself.
If the choice is made not to monitor audio metadata (or to ignore it), be prepared to face a host of potential traps. For example, the improper setting of certain flags, such as the validity bit, in the AES stream can prevent the audio itself from being converted to analog. Or, if the resolution is set improperly (e.g., treating a 20-bit stream as 24 bits), unwanted distortions can arise.
The engineering departments at most facilities are comfortable with the use of audio metadata and have invested in compatible equipment, the required procedures and the proper training. However, many facilities nationwide remain novices at metadata implementation. Without the proper audio test, measurement and monitoring equipment in-house, the phone will ring in master control if audio problems find their way to the home consumer.
Figure 3. This audio display shows a spatially mapped set of level meters for multichannel monitoring.
The home QC station
Home theater sales are increasing, and most purchases include a 16:9 display with a compatible surround-sound system. To the viewer, the brilliant HD image is not nearly as impressive when the accompanying audio has phase distortion, audible clicks or a restricted dynamic range. Consumers do not typically understand sample rates and metadata; however, they do know the difference between superb and degraded sound.
Arguably, in a multichannel home environment, the audio is as important to the viewer as the video, regardless of the consumer's level of expertise. If the broadcast facility's monitoring is thorough, the resulting quality reaches the home receiver. The market-inspired demand for multichannel audio has forced broadcasters to increase their use of metadata. With compressed formats, consumers now can use metadata to alter numerous audio characteristics.
By design, the same audio stream can provide multiple listening experiences. Two examples that affect the viewer directly are dialnorm and dynamic range control (DRC).
Dialnorm is a metadata function within the Dolby Digital (AC-3) elementary bit stream that provides a means to ensure consistent loudness. When properly set, audio levels remain fairly constant as channels are changed. If volume changes dramatically, dialnorm is either set improperly or simply not present in the audio stream.
Figure 4. TVM-950HD multi-pane display including audio alarm log and level alert message.
DRC is a metadata function that allows consumers with stereo and 5.1 surround-sound systems to choose the desired amount of compression. With DRC properly enabled at the facility, consumers can intentionally compress the dynamic range using various DRC profiles, such as speech, music light and film standard. When DRC is set improperly, the function is deactivated (and those loud car crashes can wake the family when you're watching midnight movies). In each case, comprehensive metadata monitoring can spot potential problems — provided that the right tools are in place.
Don't ignore the basics
Traditional audio monitoring remains important, including the ability to hear each program's sound on reference monitors and see a graphical representation on high-quality meters. Unfortunately, graphic metering itself has become more complex with the advent of digital (and the arrival of meters that are easily customized). For example, true peak metering uses an extremely quick responding ballistic, while loudness metering is extremely slow — yet both are quite useful in the right applications.
An additional complication is the proliferation of unique regional audio scales. For facilities performing international work, the localization of the scale itself can be a challenge. (See Figure 2 on page 60.) This too can be solved with a comprehensive monitoring tool that provides a library of scales.
Even though analog monitoring is second nature, the facility still needs the proper D/A conversion equipment to turn digital to analog, along with a decoder for the proper monitoring of Dolby formats.
In terms of the responsibility factor, large facilities, such as satellite providers, have dedicated personnel at dedicated consoles assigned to monitoring tasks, but smaller facilities can't afford that luxury. At the local level, at a minimum, two crucial points should be fully monitored by trained staff: ingest (to guarantee the quality of all audio placed on the server) and master control (to ensure consistent and flawless output quality).
With full knowledge of the complexity, ensure that your facility's monitoring, master control or quality control station is equipped with the proper audio test and measurement tools. You need to:
Provide the necessary speakers and amplifiers to monitor dual and multichannel audio configurations with easy ways to select individual or mix-down channel combinations.
Provide a CRT or LCD display that enables audio phase monitoring, using either Lissajous or phase correlation methods.
Provide a way to graphically represent audio levels, peaks and reference scales.
Select a monitoring configuration for surround sound that uses an intuitive design — one that, in effect, mimics the spatial relationships of all speakers.
Consider an audio monitoring tool that offers automatic or visual alarm capability. Particularly in unattended monitoring configurations, conditions (e.g., clipping) can be set to automatically generate alarms and alert engineering via GPI, Ethernet or SNMP traps. (See Figure 4.)
Remember that monitoring metadata is as critical today as monitoring the audio itself. Ensure that you can monitor both Dolby and non-Dolby metadata, including equalization settings, sample rates, timecode and the validity bit. With both AES and embedded streams, choose a method that enables you to see and easily understand each representative value in a tabular display — rather than in a cryptic format.
If budget is an issue, design a migration path that starts with the monitoring basics and adds functionality as required. This includes choosing a manufacturer that offers both firmware and hardware updates and the proper hooks for emerging formats.
Finally, trust your ears. The human ear is still the last receiver in the signal path. With proper amplification and superb speakers, this all-important evaluation step won't be overlooked.
If your facility is not monitoring the full range of audio parameters, you're simply not getting the full picture. There are significant advantages to doing it properly, both within the broadcast facility and for the home consumer.
Mike Richardson is director of product technologies for Leitch's Videotek line.