Skip to main content

SoundField's DSF-2

For years, broadcasters treated audio as a simple stereo add-on to what they considered the real point of their existence — the moving images they transmitted. This is much harder to sustain today, as broadcasters begin the inevitable transition to HD operations. Most broadcasters who have made the switch want 5.1 audio to be the standard transmission format. Even those who cling to stereo audio are making provisions to allow for an easy switch to 5.1 if it becomes necessary over the next couple of years.

Broadcasting in 5.1 can be a complex business, potentially involving costly multi-mic arrays at the recording stage, complex recording techniques and mixing headaches, as the outputs from multiple mics are never entirely phase-coherent. For this reason, several European broadcasters have placed orders with UK microphone manufacturer SoundField. The company's microphones can output simultaneous, phase-coherent stereo and 5.1 audio from just one microphone, if required.

A little history

Sound field microphones have been around since the 1970s, long before there was a company of the same name. The concept was pioneered by the late Michael Gerzon — a gifted mathematician, acoustician and audio equipment designer — as a means of going beyond the stereo and quadraphonic reproduction systems of the day.

From the beginning, the company's mics were capable of accurately capturing a complete 3-D sound field (hence the name). They could also render the captured audio as a listenable stereo or mono signal without any of the audible phase-related problems that can bedevil attempts to render 3-D sound recordings suitable for reproduction on conventional high-fidelity systems. For many years, the microphones' multichannel capabilities were under-used, as there was no widely established commercial audio or broadcast format that could make use of them.

With the emergence of 5.1 surround sound as an audio standard over the past few years, all of this has changed. HD broadcasters now demand an easy-to-use 5.1 recording solution. SoundField microphones can produce high-quality mono, stereo, mid and side (M&S), and surround-sound recordings in a variety of formats from a single mic, and simultaneously if required. This is useful to broadcasters who need to navigate the transition from stereo to multichannel audio carefully over the next few years, without leaving their viewers behind.

Although the simultaneous output of all these audio formats has been possible with SoundField systems for many years, the company's designers spent more than five years refining and improving the technology to produce a single-microphone, surround-ready system for HD broadcasters. The system, which was finally released last year, is called the DSF-2. The DSF-2 is the company's first product to carry out all of its processing and audio format decoding in the digital domain.

The microphones are particularly useful in situations where stereo and 5.1 sound is required simultaneously for SD and HD transmissions. Recent examples include high-profile sports fixtures like UK Premiership football matches and the Ryder Cup golf championships for Sky Sports. So what makes these microphone systems capable of outputting all of these formats at once?

How they work

SoundField systems are all comprised of a patented multicapsule microphone in which four subcardioid capsules are mounted in a tetrahedral array and one of several hardware processors. The systems produce a four-channel signal that is a proprietary form of output known as BFormat, and its four constituent signals are known respectively as W, X, Y and Z. (See Figure 1) The latter three represent the sound field around the microphone in three dimensions, as though recorded by three figure-eight microphones placed at right angles to one another, while W constitutes an overall reference signal, as though recorded by an omnidirectional capsule.

It is, of course, impossible to mount the capsules at exactly the same physical point in space, so some phase differences are inevitable in the output from the different capsules. However, because the distance of the four capsules from the center of the tetrahedral array is small, known and constant, it is possible to process the output of each capsule. This generates four signals that correspond to those that would be created by recording at exactly the same point at the center of the array, eliminating phase problems between the four constituent channels of the B-Format signal. It allows the four phase-coherent channels to be combined without any unpleasant-sounding phase artifacts, the usual drawback of multimicrophone or other multicapsule systems.

By combining the four signals making up the B-Format in different proportions, it is possible to generate an audio output in any format from mono to surround. The systems offer a variety of options into which the B-Format signal may be decoded, including mono, stereo, stereo M&S, and 5.1. Simultaneous decoding from B-Format to several different audio formats is also possible.

If the four-channel B-Format signal is recorded to a suitable multichannel storage medium, the final format can be determined long after the recording by running it through the required BFormat processor at a much later date. This makes the B-Format an excellent potential archive format.

The process of combining the four phase-coherent B-Format channels in different proportions can be controlled from dedicated switches and knobs on the company's software and hardware B-Format processors. This allows users to change the mic's effective pickup pattern continuously from omni through cardioid to figure-eight by simply adjusting a few controls.

The virtual orientation of the mic — the direction it appears to face — can also be continuously controlled from the processor. This is highly useful in OB situations, where the mic often cannot be physically accessed or adjusted once broadcasting has begun. For example, a mic may have its pickup pattern tailored to become more or less directional, and thus shut out or admit more background noise. Or it may even be virtually moved or rotated to face in the opposite direction — all without anyone going near the microphone itself.

By post-processing a stored four-channel B-Format signal with the company's hardware processor or software plug-in, audio zooming and virtual rotation of the mic can still be carried out long after the recording itself has been made.

From B-Format technology to the DSF-2

The DSF-2 was designed specifically to appeal to the broadcast market. This resulted in the use of 75V AES3-id unbalanced coaxial cable to connect the mic decoder to its broadcast vehicle, rather than a standard 110Ω balanced XLR. This allows the reliable transmission of the audio data over much longer cable lengths. Depending on the exact coaxial cable used, the decoder can be up to 1000m away from the broadcast truck. Moreover, the cable supplied with the DSF-2 to connect the microphone head to the decoder can be a further 300m long, allowing the mic to be reliably placed at total distances of up to 1.3km from its OB truck.

Although the concepts of the B-Format, the tetrahedral multicapsule array and its phase-coherent signals remain at the heart of how the DSF-2 works, the mic represents a modern implementation of the original theory produced by Michael Gerzon and his associates. As such, it produces an even better sense of localization, phase accuracy, and high-frequency sonic detail.

The system's decoding and processing takes place in the digital domain, using an Analog Devices SHARC processor. The mic is therefore the company's first product to make full use of the advantages offered by digital signal processing, such as accuracy, channel matching and repeatability.

In practice, the DSF-2's ease of use as a simultaneous 5.1- and stereo-compatible microphone (over long distances where required) is making it a success amongst broadcasters. In mid-2006, the Host Broadcast Services at the FIFA World Cup elected to install DSF-2 mics at all 12 German stadiums, using them to broadcast stadium ambience simultaneously in stereo (for SD and radio signals) and 5.1 (for the televised HD transmissions) to an estimated audience of 40 billion people.

It was often said that Michael Gerzon's inventions were ahead of his time, and the DSF2 microphone fits that description. Over-specified for stereo recording at the time of its invention, it's only now — three decades later — that it's truly coming into its own at the vanguard of surround recording for HD broadcast.

Pieter Schillebeeckx is chief designer for SoundField.