In times past, adjusting the response of a sound system to match the room and application required a mysterious combination of arcane mechanics and esoteric products whose value always remained unclear at best.

In the tone of the day, the system was cajoled into a tenuous state of “perfection” through the ever-present hovering of a lab-coat cloaked pseudo-scientist. This gruff sage held the key to sonic bliss with his slide-rule and crumpled notepad. The slightest sneeze or whiff of wind could send the entire PA out of balance as quickly as a 1979 Jaguar V-12 engine left outside on a cold night.

Fortunately, the introduction of software-based processors has improved the performance potential of every church sound system. With computer control of each parameter, the addressable detail improves dramatically while the expense of the product is reduced, since one box can replace several and the complete proceedings can be safely locked away with security encryption. Today’s Digital Signal Processor (DSP) devices are robust, easy to master and contain an abundance of sonic enhancement capabilities, so let’s look at the latest offerings and uncover some techniques to wring the most out of these boxes.

DSP Through The Ages
Digital processing in the audio realm began three decades ago with the introduction of digital delay and reverb units for the studio market. By simulating desirable acoustic environments, recording engineers could transport the performance to virtually any place or space. Once the technology developed usable speed, number crunching became a part of every record, beginning with digital storage on tape (DASH, 3M, DAT, ADAT and DA-88) and then moved on to the console where mix automation came into its own.

Finally, the technology morphed beyond the studio glass and into the live arena where it immediately found a home in digital equalizers as an effective force in the struggle to tame unwieldy rooms and poorly placed loudspeakers. Today, the digital landscape is plump with full-menu offerings covering all manner of signal control.
In the worship environment, DSP can be used as a Loudspeaker Management System (LMS) with line level inputs to provide improved response from a given speaker system or as a Digital Mixing Engine (DME) with mic level inputs; in essence, a mixing console without faders, for areas such as fellowship halls and conference rooms. Both applications use the same basic architecture but vary in their configuration and thrust.

An audio DSP is essentially a purpose-specific computer contained in a rack mountable enclosure. While the physical box is the most recognizable aspect of the solution, it is the software programming that bears the brunt of the horsepower duties. As such, several DSP brands allow an external computer to drive the machine, both to speed the programming process along and to reduce the complexity and expense of the front panel interface. Some brands even forego front panel control, preferring to give the programmer a full-size, full-color computer screen instead of a small LCD display.

Most DSP devices have a defined architecture, that is, they have a fixed number of inputs and outputs. Typically, in a live sound LMS device, there are two or four inputs and four, six or eight outputs to cover stereo or LCR ins and biamped or triamped outputs along with auxiliary signal lines for discreet paths to overflow rooms and the like. For instance, the popular dbx DriveRack 260 contains dual inputs and six outputs, enough to control a stereo three-way system or a mono biamp main cluster plus delayed under-balcony feeds. Some recent product innovations include combining the DSP into an accompanying power amp, as demonstrated by the Ashly NE-PE, Crown I-Tech and Peavey IPR series amplifier lines as well as completely configurable designs as exemplified by the Media Matrix NION.

Since the LMS application of a DSP box requires full control of the loudspeakers, a modern DSP replaces many traditionally analog devices in the signal chain, including graphic and parametric equalizers, multi-way crossovers, signal delay units, compressors and limiters.

Though the older analog boxes may appear simpler to operate, DSP interfaces have evolved to follow the same logical path as their analog predecessors. Thus, the crossover section usually appears first in the display, followed by EQ and then delay and finally, dynamics control. A typical DSP unit takes the place of many analog products and their accompanying need for rack space, power outlets and conditioning, interface cabling and security covers and replaces them all with a single device whose sections talk to one another without risk of level or impedance mismatch.

Crossover and Delay
At the head of the signal chain, the crossover provides a way to divide the frequency bands into their appropriate amps and drivers, with lows going to the subs and highs going to the compression drivers. When a loudspeaker is capable of both active and passive band splitting, it is preferable to separate the frequencies prior to the amplifiers for the most efficient use of power. A passive system must divide the signal when it is a rushing torrent as opposed to an active system where the division takes place when the flow is a mild creek.

In an existing installation where replacement of the main speakers is prohibitive, introducing active crossovers through a DSP can make a dramatic difference in the overall system performance. For instance, the proper addition of a Biamp Nexia SP to a fifteen-year old church installation can increase the clarity of the sound by constraining and optimizing the response of each driver within the speaker array while reducing the potential for damage through the use of compression and limiting on the outputs feeding the power amps.

Meanwhile, the low-end response will improve since the programmer can place a steep infra-sonic filter on the entryway to the bass driver while simultaneously incorporating a mild bump in the EQ around 80Hz and then leaving a slight hole in the crossover region around 125Hz to compensate for the room’s exaggerated reinforcement of that area. Further down the line, more EQ can be added, preferably in the form of parametric cuts instead of graphic EQ adjustments, both to more accurately define the filter solution and to reduce the burden on the DSP’s processing engine.

Many sanctuaries rely on delay speakers to provide sound to under and over balcony listeners as well as those in the nursery and cryroom. Correct signal delay amounts are necessary in each location in order for the perceived arrival times to coincide in an intelligible manner. LMS boxes are able to introduce signal delay in millisecond intervals and some models even sport built-in Haas Effect compensation for precedence, where the closest signal is perceived over subsequent arrivals and the listener localizes on the nearest source. In practice, people seated under the balcony will hear sound as emanating from the small speaker directly above their head instead of from the primary cluster at the front of the room. Introducing signal delay to the near speaker changes the listener’s perception to believe the sound is originating at the distant source.

A modern LMS can deliver delay to multiple zones simultaneously and also bring microsecond delay into the chain for driver alignment control. Often, in multi-way loudspeaker designs, the propagation point of the woofer is different than that of the compression driver with its elongated horn, so driver delay compensates for the resulting “smear” by electronically bringing the shallow driver in line with the deeper unit. Finally, LMS delay can be used to orient the main speaker array to a point in line with the band on stage. In common practice, the DSP’s primary inputs are delayed a few milliseconds so the acoustic sound of, say, the snare drum matches its origin at the loudspeaker. When performed correctly, the move creates the sense of a vanishing sound system as the energy on stage mates properly with that of the speaker system at the listener’s position.
Latency, Templates and Presets
All DSP products exhibit latency, the time delay introduced by converting analog inputs to digital signals, crunching the numbers for processing and then reconverting the signals to analog for the outputs. As such, manufacturers expend effort and money on ways to reduce the issues presented by latency, which include loss of clarity and integration between video imagery and the audio system.

The use of faster processor chips, such as Motorola/Freescale’s latest offerings and Analog Devices’ SHARC series, along with improved daughter board functions have reduced the impact of latency in the latest DSP units. End users can also help the system’s response time by using parametric EQ in place of graphic EQ, and bypassing unneeded filters and unused outputs.

Digital devices can be labeled as “thin sounding” or “harsh” and justifiably so if the proper horsepower and programming chops are missing. However, the latest mainstream models as well as higher end LMS products, such as the stalwart Klark-Teknik 9848 and XTA’s new DP426 incorporate high performance 24-bit converters running at 96 kHz in order to push the treble response beyond the usual 20 kHz limit, giving the system back its dynamic range and full-bodied tone.

Most DSP products also provide the user a series of templates with common configurations already residing in the unit and ready to perform with only minor tweaking. There are templates for stereo two-way set-ups, mono three way with balcony delay and other common scenarios to approximate the majority of situations encountered in the worship environment. For example, family-related product groups like JBL and dbx (both members of the Harman conglomerate) cross-pollinate their lines by providing presets for JBL’s SRX loudspeaker series inside the dbx DriveRack units. As a gesture of goodwill, dbx also features presets for Yamaha and other brands of speakers, though every acoustic environment is different and can benefit from on-site adjustments. The purpose of presets and templates, naturally, is to flatten the time curve between opening the box and using the product and for many applications, a preset is all that is needed to improve the sound. However, it is better to venture inside the system architecture in order to maximize the return on DSP investment.

The first aspect of going beyond the factory presets is to start with them, that is, to find a preset configuration close to the desired result and then modify it accordingly. As a caveat, incorrect assignment of the outputs or mis-wiring of the rear panel can result in damaged speaker components, so double-check cable routing and bring all drivers up slowly and deliberately to ward off any accidental damage. Also of note, the automatic programming functions built into many DSP systems are rough estimates at best and are no substitute for manual configuration by a knowledgeable audio professional.

If, for instance, the factory setting provides good broad-band response but lacks “punch” in the low end due to an absence of subwoofers, the main speakers can be EQ optimized in the 100 Hz region but kept from over-excursion through a combination of a fourth order infrasonic filter at 60 Hz and tight compression on the output driving the woofers.

Further enhancements can be made on upstream frequencies as well. For example, the tendency of older two-way fifteen inch systems to sound muffled in the low-mid band can be eradicated with judicious use of a filter around 250 Hz and the notorious unpleasantness of 1.6 kHz can be mitigated with a dynamically-allocated cut set to activate only when the level rises to a certain point, leaving the response unaltered at lower speech-type volumes.

Presets can also be modified beyond the general setting for a church service to accommodate different speaking pastors and their particular voice characteristics. As an example, a controlled male speaker with a pronounced resonance at 160 Hz can be mollified with an EQ variation on the standard preset, while using the original settings altered with a heavy hand on the compression section will quell a speaker who tends to whisper and then shout.
Additionally, the general-use preset can be modified for different service types held throughout the week. For instance, Sunday events may use one version for the traditional service, (with subs muted and vocal clarity given priority) while the blended setting opens the subs but keeps the voice response prominent. Then the contemporary worship slot is handled with the subs brought to full volume and the presence peak of the voice response reduced. Mid-week services can use a preset with the rear delay speakers muted and the front fills brought up in level to better meet the needs of a small group seated near the stage. Finally, the primary preset can be modified with severe limiting to prevent unintentional damage from visiting artists unfamiliar with the system’s capabilities.

Other Goodies
In contrast to the line-only nature of LMS products, DSP mix engines provide additional control in the form of full mic preamps and phantom power, along with full-fledged mixing features akin to a small format digital console. A mix engine is useful in areas like fellowship halls, chapels and overflow rooms where the need for complex mixing is minimal, but occasional live mics must be blended with local playback or remote line feeds. Mix engines also excel as an adjunct to traditional sound systems for use as simplified controllers of audio during weddings and funerals when a large format console is unnecessary.

The Architectural Acoustics Digitool MX is a prime example of an effective mix controller with its eight mic or line inputs and eight line outputs running at 132 MIPS through twin paralleled 24 bit SHARC processors. Programming can be accomplished via the front panel encoder dial and backlit LCD display or by attached PC with the included software.

In a similar vein, Yamaha’s well-regarded mix engines benefit from the company’s storied digital expertise in products as hallowed as the PM-1D and DME64N. The firm’s IMX644 establishes a solid benchmark for small digital mixers integrated into larger environments with its combination of fourteen analog and auto-clocked digital inputs driving a comprehensive feature set of control parameters to a trio of stereo outputs. With the included management software, any host computer can control all the IMX644’s settings other than in and out levels.
DSP has brought audio processing out of the expensive, delicate days of analog and into the cost-efficient, reliable world of digital technology. With DSP, any sound system can be improved through the appropriate use of dynamics control, equalization, delay and crossover functions while digital mix engines provide the blending and grouping necessary for banquet-style events.

Loudspeaker Management Systems allow the replacement of older speaker systems to be put off for some time since many of the pitfalls inherent in clusters, such as lobing, can be mitigated in the DSP environment. In the past few years, manufacturers have delivered excellent products at affordable prices and now the user interfaces and software programs have caught up with the hardware, making DSP no longer the exclusive realm of lab-coated esoterics, but a welcome addition to the worship tech’s everyday arsenal.